Professional Documents
Culture Documents
Mentalis Folyamatok A Nyelvi Feldolgozasban
Mentalis Folyamatok A Nyelvi Feldolgozasban
Mentalis Folyamatok A Nyelvi Feldolgozasban
A NYELVI FELDOLGOZSBAN
MENTAL PROCEDURES
IN LANGUAGE PROCESSING
MENTLIS FOLYAMATOK
A NYELVI FELDOLGOZSBAN
Pszicholingvisztikai tanulmnyok III.
MENTAL PROCEDURES
IN LANGUAGE PROCESSING
Studies in Psycholinguistics 3.
Szerkesztette / Editors
TINTA KNYVKIAD
BUDAPEST, 2012
Sorozatszerkeszt
KISS GBOR
Szerkesztette
NAVRACSICS JUDIT
SZAB DNIEL
Lektorlta
FLDES CSABA
GSY MRIA
LENGYEL ZSOLT
HOWARD JACKSON
KEES DE BOT
AVRACSICS
N
JUDIT
ISSN 1419-6603
ISBN 978-615-5219-17-7
A szerzk, 2012
Navracsics Judit Szab Dniel
TINTA Knyvkiad, 2012
A kiadsrt felels
a TINTA Knyvkiad igazgatja
Felels szerkeszt: Szab Mihly
TARTALOM
Elsz ....................................................................................................................... 8
I. PSZICHOLINGVISZTIKA: P S SRLT FOLYAMATOK ........................ 9
The end of psycholinguistics as we know it? ..................................................... 10
Recursion in language, theory-of-mind inference, ............................................ 20
SZELHVS VIZSGLATA ALZHEIMER-KRBAN ...................... 29
Die Adaptierung des GMP-Tests ins Slowakische.............................................. 40
Semantic and Formal Strengthening Processes ................................................ 55
Fonetikai/fonolgiai kapcsolatok ...................................................................... 63
Szelhvs gyermekek szasszociciiban ...................................................... 89
A MENTLIS TRKP S A HELYNEVEK ............................................ 99
A MENTLIS TEREK A JELENTSALKOTSBAN ............................ 106
REGEDS S LEXIKLIS HOZZFRS A KZ-........................... 114
A NYELVI FEJLESZTS LEHETSGEI LOVASTERPIS ........... 122
A HALLGATS SZEREPE S FUNKCII .............................................. 127
A CSEND S A HALLGATS A KULTRAKZI ................................. 139
3WS! WORDS IN THE 21ST CENTURY .................................................. 145
II. BESZDKUTATS ....................................................................................... 154
AZ ARTIKULCI LELLSA A SPONTN BESZDBEN .............. 156
A HEZITCIS JELENSGEK GPI OSZTLYOZSA .................... 169
AZ AGYI MONITOROZS MDOSULSA ZAJHATSRA ............... 178
A MONDATISMTLS SAJTOSSGAI FIATAL, .............................. 190
DISZLEXIS GYERMEKEK SPONTN BESZDNEK ..................... 199
A FOGAZAT S A FOGMEDERNYLVNY ELTRSNEK .......... 208
ELSZ
Tanulmnyktetnkben alcmhez hen a pszicholingvisztika terletrl vlogattunk kutatsokat, s krtnk fel szerzket, hogy osszk meg a szles olvastborral
legjabb eredmnyeiket.
Legnagyobb rmnkre klfldi kutatk is hozzjrultak a gyjtemnyhez, gy a ktet tbbnyelv: a magyar nyelv tanulmnyokon kvl tallunk benne angol s nmet
nyelven rottakat.
A tanulmnyok tmi a beszdprodukci s percepci folyamatainak vizsglatval
foglalkoz kutatsokat lelik fel, hazai s nemzetkzi szntren. A kutatsok nem csak
a magyar nyelvet rintik: az alaptmkrl a hazai szakembereken kvl holland, horvt, nmet s szlovk szerzk is beszmolnak. Az ltalnos pszicholingvisztikai tmkon kvl a ktnyelvsgre s a msodik nyelvelsajttsra vonatkoz tanulmnyok is
egyre nagyobb ltjogosultsgra tesznek szert magyar s nemzetkzi kontextusban.
A ktetet egy gondolatbreszt, majdhogynem provokatv tanulmny nyitja, amelylyel Kees de Bot, a Groningeni Egyetem professzora tisztelte meg az olvaskat. Nzete szerint elavult a pszicholingvisztikban uralkod modellek szemllete, s az egsz
nyelvfeldolgozst, beszdkutatst, nyelvelsajttst dinamikus szemllettel kellene
vizsglni szemben a ma uralkod szeparlt s statikus, nyelvi szintenknt vizsgld
nzettel. A felvets elgondolkodtat, m mg nincs tlet, nincs megfelel mdszer,
amellyel dolgozhatnnk. Ugyanakkor mr maga az a felismers, hogy az eddigi modellekkel csak egyre ellentmondsosabb eredmnyekre jutottunk, valsznleg r fog
knyszerteni bennnket a paradigmavltsra.
Egyelre azonban maradunk a szmunkra megszokott szemlleten alapul kutatsoknl, amelyek jabb rtkes adatokat s eredmnyeket mutatnak be a magyarorszgi
s klfldi pszicholingvisztikai mhelyekbl.
Sorozatunk egyik clja, hogy a kzp-eurpai pszicholingvisztikai kutatsokat a rgiban megismertessk, hogy nemzetkzi szinten fenntartsuk az rdekldst egyms
kutatsai s eredmnyei irnt. Bzunk abban, hogy ezzel a ktettel is sikerl hozzjrulni a nemzetkzi szint tudomnyos koopercihoz.
Veszprm, 2012. mrcius 8.
A szerkesztk
I. PSZICHOLINGVISZTIKA:
P S SRLT FOLYAMATOK
PSYCHOLINGUISTICS:
INTACT PROCESSING AND DISFUNCTIONS
10
The formulator converts the preverbal message into a speech plan (phonetic plan)
by selecting lexical items and applying grammatical and phonological rules. Lexical
items consist of two parts, the lemma and the morpho-phonological form or lexeme.
The lemma represents the meaning and syntax of the lexical entry while the lexeme
represents the morphological and phonological properties. In production, lexical items
are activated by matching the meaning part of the lemma with the semantic information
in the preverbal message. Accordingly, the information from the lexicon is made available in two phases: Semantic activation precedes form activation (SchriefersMeyer
Levelt 1990). Activation of the lemma immediately provides the relevant syntactic
information, which in turn activates syntactic procedures. The selection of the lemmas
and the relevant syntactic information leads to the formation of a surface structure.
While the surface structure is being formed, the morpho-phonological information in
the lexeme is activated and encoded. The phonological encoding provides the input for
the articulator in the form of a phonetic plan. This phonetic plan can be scanned internally by the speaker via the speech-comprehension system, which provides the first
possibility for feedback.
The articulator converts the speech plan into actual speech. The output from the
formulator is processed and temporarily stored in such a way that the phonetic plan
can be fed back to the speech-comprehension system and the speech can be produced
at normal speed.
A speech-comprehension system connected with an auditory system plays a role
in the two ways in which feedback takes place within the model: the phonetic plan as
well as the overt speech are passed on to the speech-comprehension system, where mistakes that may have crept in can be traced. Speech understanding is modeled as the mirror image of language production, and the lexicon is assumed to be shared by the two
systems.
2. Language separation and language choice
In dealing with bilingual speakers there are two aspects that have to be accounted for:
How do these speakers keep their languages apart, and
How do they implement language choice?
Psycholinguistically, code switching and keeping languages apart are different aspects
of the same phenomenon. In the literature, a number of proposals have been made on
how bilingual speakers keep their languages apart. Earlier proposals involving input
and output switches for languages have been abandoned for models based on research
on bilingual aphasia. Paradis (2004) has proposed the subset hypothesis, which he
claims can account for most of the data found. According to Paradis, words (but also
syntactic rules or phonemes) from a given language form a subset of the total inventory. Each subset can be activated independently. Some subsets (e.g. from typologically
related languages) may show considerable overlap in the form of cognate words. The
subsets are formed and maintained by the use of words in specific settings: words from
a given language will be used together in most settings, but in settings in which codeswitching is the norm, speakers may develop a subset in which words from more than
one language can be used together. The idea of a subset in the lexicon is highly com11
patible with current ideas on connectionist relations in the mental lexicon (cf. Roelofs
1992).
A major advantage of the subset hypothesis is that the set of lexical and syntactic
rules or phonological elements from which a selection has to be made is reduced dramatically as a result of the fact that a particular language/subset has been chosen. Our
claim is that the subset hypothesis can explain how languages in bilinguals may be
kept apart, but not how the choice for a given language is made. The activation of a
language specific subset will enhance the likelihood of elements of that subset being
selected, but it is no guarantee for the selection of elements from that language only.
According to the subset hypothesis, bilingual speakers have stores for lemmas, lexemes, syntactic rules, morpho-phonological rules and elements, and articulatory elements that are not fundamentally different from those of monolingual speakers. Within
each of these stores there will be subsets for different languages, but also for different
varieties, styles and registers. There are probably relations between subsets in different
stores; i.e. lemmas forming a subset in a given language will be related to both lexemes and syntactic rules from that same language, and phonological rules from that
language will be connected with articulatory elements from that language. The way
these types of vertical connections are made is in principle similar to the way in which
connections between elements on the lemma level develop.
Activating a subset in the lexicon on the basis of the conversational setting can be
the activation of a particular language, but it can also be a dialect, register or style.
These subsets can be activated both top-down, when a speaker selects a language for
an utterance or bottom up, when language used in the environment triggers and activates a specific subset (de Bot 2004). Triggers on different levels: sounds, words,
constructions, but probably also gestures can activate a subset.
3. Towards dynamic models of bilingual processing
It can be argued that the Hartsuiker & Pickering (2007) article represents the state of
the art at this moment in the sense that this type of model is the most prominent one.
The whole literature on bilingual processes centers on this type of model. It will be
argued in the remainder of this contribution that there may be reasons to move beyond
such models because they have a number of rather serious problems.
The main problem is that such models are based on underlying assumptions that
may no longer be tenable:
Language processing is modular: it is carried out by a number of cognitive
modules that have their own specific input and output and that function more
or less autonomously
Language processing is incremental and there is no internal feedback or feedforward
Elements at different levels in the model can be studied in isolation
Syntax and lexicon are separate modules
Various experimental techniques will provide us with reliable and valid data
on the workings of the model
12
13
through collaborative efforts to achieve shared objectives in cultural surroundings (Salomon 1993: 1).
Van Gelder (1998) describes how a DST perspective on cognition differs from a more
traditional one:
The cognitive system is not a discrete sequential manipulator of static representational structures: rather, it is a structure of mutually and simultaneously influencing
change. Its processes do not take place in the arbitrary, discrete time of computer steps:
rather, they unfold in the real time of ongoing change in the environment, the body, and
the nervous system. The cognitive system does not interact with other aspects of the
world by passing messages and commands: rather, it continuously coevolves with
them.
With these notions in mind, let us look at the main characteristics of the models discussed that are part of the information processing tradition.
Language processing is modular: it is carried out by a number of cognitive modules that have their own specific input and output and that function more or less
autonomously
The most outspoken opponent of a modular approach to cognitive processing at the
moment is probably Michael Spivey in his book The continuity of mind (2007). His
main argument is that there is substantial evidence against the existence of separate
modules for specific cognitive activities such as face recognition and object recognition. For linguistic theories this is crucial since in UG based theories a separate and
innate language module plays a central role. Distributed processing of language undermines the idea that language is a uniquely human and innate because the cooperating parts of the brain are not unique for language, have no specific linguistic
knowledge and work in feedback and feedforward types of structure.
Language processing is incremental and there is no internal feedback or feedforward
One of the problems of this assumption is that many second language speakers regularly experience a feeling of knowing (PeynircogluTekcan 2000). They want to say
something in the foreign language, but are aware of the fact that they dont know or
have a quick access to a word they are going to need to finish a sentence (de Bot
2004). This suggests at least some form of feedforward in speaking. Evidence for
feedforward processes in language production can also be found in Cleland & Pickering (2003) and Herdman et al. (2007)
Additional evidence against a strict incremental view is provided in an interesting
experiment by Hald, Bastiaanse & Hagoort (2006). In this experiment, speaker characteristics (social dialect) and speech characteristics (high/low cultural content) were
varied in such a way that speaker and speech characteristics were orthogonally varied.
Listeners heard speakers whose dialect clearly showed their high or low socioeconomic status talk about Chopins piano music or about tattoos. The combinations
14
of high cultural content and low social status in a neuro-imaging experiment led to
N400 reactions, which showed that these utterances were experienced as deviant. A
comparison with similar sentences with grammatical deviations showed that the semantic errors were detected earlier than the syntactic ones, which is a problem for a
purely incremental process from semantics to syntax and phonology. The semantics
and pragmatics seem to override the syntax in this experiment.
Isolated elements (phonemes, words, sentences) are studied without taking into
account the larger linguistic and social context they are part of
If cognition is situated, embodied and distributed, studying isolated elements is fairly
pointless: we need to investigate them as they relate to other aspects of the larger context, both linguistic and extra-linguistic. For example, work by Eisner & McQueen
(2006) has shown that the perception of ambiguous phonemes is strongly influenced
by the semantics of the context in which that phoneme is used.
Syntax and lexicon are separate modules
There is in linguistics a growing tendency to move away from a strict division of syntax and lexicon. Various authors have argued for Grammar as idiom. In this perspective the difference between sentences like John goes home vs. John went home is
more lexico-semantic than grammatical, they can be viewed as idioms with different
meanings. This kind of thinking can be found in Goldbergs construction grammar
(2003) and Croft & Cruses (2004) Grammatical idioms.
There is also some neurolinguistic evidence to support this view. Csepes (2011)
neuro-imaging data suggest that syntactic processing is essentially a form of semantic
processing.
As Lowie & Verspoor (2011) argue: The emerging picture is one of the multilingual mind as a multidimensional state space, in which fuzzy subsets of symbolic units,
be it words, formulaic sequences or syntactic constructions, are activated in particular
contexts....From this perspective it is not necessary to distinguish procedural lexical
knowledge from a separate syntactic processing component.
Various experimental techniques will provide us with reliable and valid data on
the workings of the model
The existing models of language processing are based on a large body of very sophisticated experimental techniques that aimed at unraveling the complexities of the system. A relevant question is whether we can indeed get a better understanding of the
whole by taking it apart. Spivey (2007) takes a clear position in this debate: The fundamental weakness of some of the major experimental techniques in cognitive psychology and neuroscience is that they ignore much of the time course of processing
and the gradual accumulation of partial information, focusing instead on the outcome
of a cognitive process rather than the dynamic properties of that process.
15
16
As may be clear from the argumentation so far, we may have to review some of the
basic assumptions of the information processing approach in which our current models
of multilingual processing are based. In the previous section we have listed the main
characteristics and the problems related to them. From this it follows that we need to
develop models that take into account the dynamic perspective in which time and
change are the core issues.
As a conclusion, some of the characteristics of dynamically based models are listed
here:
Models should include time as a core characteristic: language use takes place on
different but interacting time scales
Models should allow for representations that are not invariant but variant and episodic
Models should allow for feedback and feedforward information rather than a
strict incremental process
Models should recognize that language use is distributed, situated and embodied;
therefore linguistic elements should not be studied in isolation but in interaction
with the larger units they are part of
Models should recognize that interaction rather than monologue is the focus of
research
The relevance and validity of experimental techniques that assume static representations on various levels needs to be reassessed
Accepting that time and change are the core issues in human cognition implies that
new models are needed, but as Spivey (2007) readily admits, it is difficult to leave
established notions and assumptions behind while there is as yet no real alternative. It
is our conviction that we will move on to more dynamic models in the years to come
but how that will happen is unclear. Model development in itself is a dynamic process.
Acknowledgements
The author is indebted to Ludmila Isurin, Marjolijn Verspoor and Szilvia Btyi for
their comments on an earlier version of this contribution.
References
Cleland, A. Pickering, M. 2003. The use of lexical and syntactic information in language production: Evidence from the priming of noun-phrase structure. Journal of Memory and Language.
49: 214230.
Croft, W. Cruse, D. 2004. Cognitive linguistics. Cambridge: Cambridge University Press.
Cspe, V. 2011. Agglutinating languages Challenges for the human brain? (paper at the 13th Summer School of Psycholinguistics, Balatonalmdi, 2011. 5. 22.)
de Bot, K. 1992. A bilingual production model: Levelts Speaking model adapted. Applied Linguistics. 13: 124.
de Bot, K. 2002. Cognitive processes in bilinguals: language choice and code-switching. In Kaplan, R.
(ed.): The Oxford Handbook of Applied Linguistics. Oxford: Oxford University Press. 287300.
17
de Bot, K. 2004. The multilingual lexicon: modeling selection and control. International Journal of
Multilingualism. 1: 1732.
de Bot, K. Lowie, W. 2010. On the stability of representations in the multilingual lexicon. In Sicora,
L. (ed.): Cognitive processing in second language acquisition. Amsterdam: John Benjamins. 117
134.
de Bot, K. Verspoor, M. Lowie, W. 2007. A Dynamic Systems Theory approach to Second Language Acquisition. Bilingualism, Language and Cognition. 10: 721.
Eisner, F. McQueen, J. 2006. Perceptual learning in speech: Stability over time. JASA. 4: 19501953.
Elman, J. 1995. Language as a dynamical system. In van Gelder, T. (ed.): Mind in motion: Explorations of the dynamics of cognition. Cambridge: The MIT Press. 195225.
Goldberg, A. 2003. Consrtuctions: A new theoretical approach to language. Trends in Cognitive
Sciences. 7/5: 219224.
Green, D. 1993. Towards a model of L2 comprehension and production. In Schreuder, R. Weltens,
B. (eds.): The Bilingual Lexicon. Amsterdam: John Benjamins. 249278.
Hald, L. Bastiaanse, M. Hagoort, P. 2006. EEG theta and gamma responses to semantic violations in online sentence processing. Brain and Language. 96/1: 90105.
Hartsuiker, R. Pickering, M. 2007. Language integration in bilingual sentence production. Acta
Psychologica. 128: 479489.
Herdina, P. Jessner, U. 2002. A dynamic model of multilingualism. Perspectives of change in psycholinguistics. Clevedon: Multilingual Matters.
Herdman, A. Pang, E. Ressel, V. Gaetz, W. Cheyne, D. 2007. Task-related modulation of
early cortical responses during language production: An event-related synthetic aperture magnetometry study. Cerebral Cortex. 17/11: 25362543.
Kormos, J. 2006. Speech production and second language acquisition. Mahwah: Erlbaum.
Larsen-Freeman, D. Cameron, L. 2008. Complex systems and applied linguistics. Oxford: Oxford
University Press.
Levelt, W. 1993. Language use in normal speakers and its disorders. In Blanken, G. Dittman, E.
Grimm, H. Marshall, J. Wallesch, C. (eds.): Linguistic disorders and pathologies. An International Handbook. Berlin: Walter de Gruyter. 115.
Levelt, W. J. M. 1989. Speaking. From Intention to Articulation. Cambridge: The MIT Press.
Myers-Scotton, C. 1995. A lexically based model of code-switching. In Milroy, L. Muysken, P.
(eds.): One speaker, two languages. Cross-disciplinary perspectives on code-switching. Cambridge: Cambridge University Press. 233256.
Nieuwland, M. van Berkum, J. 2006. When peanuts fall in love: N400 evidence for the power of
discourse. Journal of Cognitive Neuroscience. 18: 10981111.
Paradis, M. 1998. Aphasia in bilinguals: How Atypical is it? In Coppens, P. Lebrun, Y. Basso, A.
(eds.): Aphasia in Atypical Populations. Mahwah: Lawrence Erlbaum. 3566.
Paradis, M. 2004. A neurolinguistic theory of bilingualism. AmsterdamPhiladelphia: John Benjamins.
Peynirclog, Z. Tekcan, A. 2000. Feeling of knowing for translations of words. Journal of Memory
and Language. 43/1: 135148.
Pickering, M. Garrod, S. 2004. Toward a mechanistic psychology of dialogue. Behavioral and
Brain Sciences. 27: 169190.
Port, R. van Gelder, T. 1995. Mind as motion: Explorations in the dynamics of cognition. Cambridge: The MIT Press.
Poulisse, N. 1997. Language production in Bilinguals. In de Groot, A. Kroll, J. (eds.): Tutorials in
bilingualism. Mahwah: Lawrence Erlbaum. 201224.
Poulisse, N. Bongaerts, T. 1994. First language use in second language production. Applied Linguistics. 15: 3657.
Roelofs, A. 1992. A spreading-activation theory of lemma retrieval in speaking. Cognition. 41.
18
19
1. The issue
Some researchers claim that the human faculty of recursion is revealed by syntacticstructural embedding (HauserChomskyFitch 2002), while some others claim it is
due to recursive theory-of-mind inferences/embeddings or to pragmatic abilities
(SiegalVarley 2006, EvansLevinson 2009, Everett 2009). Whereas the source of
structural embeddings that show syntactic recursion are the syntactic rules of language
(e.g. Peter asked Mary to watch the film that Helen said John liked the best; In the
moment of hearing of the critical reactions to the reception of the sequel to the performance of the artist), the same rules may merely encode or transmit peoples
awareness of other peoples mental states, i.e. theory of mind inferences (e.g. Peter
believes that Mary believes that Helen can put herself in Brians shoes.). Siegal and
Varley (2006) argue, on the basis of experiments involving aphasic speakers, that
theory of mind abilities may remain unimpaired even in cases of limited language
faculties, that is, they are not grammar-dependent. Theory of mind type reasoning can
be recursive (cf. Takano -Arita 2010). If we assume that individuals with a theory of
mind also consider others and themselves to have a theory of mind, then there should
be a recursive structure here. Hungarian speaking aphasic subjects were able to produce this type of recursive structure that is deeply linked to social intelligence/cognition.
We focused on empirical investigations involving linguistic tests administered to
subjects with agrammatic and Wernickes aphasia. The test sessions involved 5 aphasics, 3 Brocas, 2 Wernickes aphasics as well as 21 healthy control subjects. All
aphasic participants had a left unilateral brain lesion. Aphasic subjects were assigned
to aphasia types on the basis of CT and Western Aphasia Battery (WAB) tests (Kertesz 1982). Three subjects were diagnosed as Brocas aphasic and two subjects were
diagnosed as Wernickes aphasic subjects. The most important results of the linguistic
tests can be summarised as follows.
2. The tests
Photographs representing situations of everyday life were presented to subjects and
questions were asked about them. We used 208 photographs (Stark 1998) for each test,
administered in three sessions. Within the same session, no picture was involved in
more than a single question type.
1
This research has been supported by the National Scientific Research Fund (OTKA), project:
NK 72461.
20
Brocas aphasics 3
Normal control 21
Type 1 question
Number of answers:
(68.8) 50.8
107
(56.3) 45.6
257
(100.0) 100.0
1117
Type 2 question
Number of answers:
(46.0) 40.3
111
(37.1) 31.6
246
(100) 99.1
1115
Type 3 question
Number of answers:
(60.4) 34.3
120
(30.8) 22.8
240
(100) 98.5
1139
21
Type 4 question
Number of answers:
(66.7) 61.1
76
(60.3) 45.6
218
(100.0) 100.0
982
Table 1. Percentage of structurally linked answers (outside the brackets: that of grammatical answers) with respect to the total number of answers by subjects and normal control
According to Table 1 the number of structurally linked and grammatical answers decreased from Type 1 (What is X doing?) to Types 2 and Types 3 (What does X want?
and What is the most entertaining for X?, respectively). With respect to Type 4 questions (What does X say / think / remind Y of / ask Y to do?), requiring a recursively
embedded clause as an answer, the performance of the subjects actually turned out to
be better than with Type 1 questions (What is X doing?); or it was almost as good.
4. Type 4 questions
With respect to Type 4 questions, Wernickes aphasics produced some conjunctioninitial clauses and some clauses involving the subjunctive (i.e. the mood directly indicating subordination).
Brocas aphasics gave few answers beginning with a subordinating conjunction.
One of them did produce some answers involving the subjunctive. However, the majority of structurally linked and grammatical answers produced by Brocas aphasics, as
well as the rest of the answers given by Wernickes aphasics, were statements that
assumed the point of view of one of the characters seen in the picture, rather than being purely descriptive. The subjects answered the question as if they were in the mental state of the characters. These answers are referred to as situational statements
with theory of mind type reasoning. In them, the verb was inflected in the first, rather
than the third, person singular (or second person singular, with reference to the partner
in the situation shown in the picture); their meanings differed sharply from descriptive
statements, as they directly represented the thought or statement of the characters they
cited; most of them did not involve a subordinating conjunction. These answers are
supposed to involve syntactic structural recursion but they contain simple statements
instead, with theory of mind type reasoning. An example of a situational statement:
The picture: A girl is showing her scar to a boy.
Question: Vajon mire gondol a fi? What may the boy be thinking of?
S.T.s answer: Mindjrt rosszul leszek! Im going to be sick.
Possible recursive construction: () arra gondol, hogy mindjrt rosszul lesz. He
thinks he is going to be sick.
In the sense of Takano and Arita (2010), theory of mind type reasoning is recursive.
The subjects, in addition to seeing themselves as able to infer other peoples mental
states, considered other persons (e.g. ones seen in pictures) to be able to infer further
(third) persons mental states, thus exhibiting recursive constructions.
22
Table 2 shows that the share of situational statements increased in Brocas answers
to Type 4 questions:
Wernickes aphasics
(answers :76)
Brocas aphasics
(answers: 218)
Normal control
(answers: 982)
Situational statement
(43.3) 43.3
(74.0) 60.3
(31.0) 31.0
(14.2) 14.2
(10.0) 10.0
Subordinating conjunction
+ situational statement
(12.5) 12.5
(14.5) 8.7
(24.0) 24.0
Subordinating conjunction
+ descriptive clause
(30.0) 30.0
(1.4) 1.4
(45.0) 45.0
Table 2. With type 4 questions percentage of all structurally linked answers (outside the brackets:
that of grammatical answers) in the grammatical categories, in aphasia types and in normal controls
(American Psychiatric Association 2000). In persons with Hungarian speaking Alzheimers disease (AD), as opposed to the case of aphasics, the language faculty becomes limited gradually as the disease progresses. We tried to find out to what extent
structural embedding linguistic operations vs. recursive theory-of-mind inferences
were involved in their case. Therefore, we administered the above pictures and questions (What might X in the picture be thinking of?; What might Y in the picture be
asking Z to do?) to four persons with mild and two with moderate AD.
In the responses of subjects with mildAD, there was no significant difference between them and normal controls in the proportion of replies involving syntacticstructural recursion vs. situative responses. We can infer that in mild AD, both structural and theory-of-mind recursion are unaffected.
In the case of moderate AD, we found a significant difference in the proportion of
responses involving syntactic-structural recursion and simple situative responses requiring recursive theory of mind inference: the ratio of situative sentences was significantly lower than in normal control subjects. On the other hand, in the responses of
persons with moderate AD, the share of sentences involving syntactic-structural recursion (embedding involving hogy that clauses) was not lower than in normal control
responses. Additionally, we also received semantically irrelevant descriptive statements referring to some aspect of the picture. That is, while syntactic-structural recursion may remain unaffected in moderate AD, theory-of-mind inferences seem to be
impaired to some extent.
Subjects
Relevant responses
Irrelevant responses
Mild AD
86.2
13.8
Moderate AD
62.1
37.9
Moderate AD (82)
Normal (982)
13.0 3.8
20.5 13.5
Situative statement
22.0 1.9
2.4 2.4
31.0
3.0
17.0 1.3
7.0
2.4. 1.2.
24.0
55.0 8.1
45.0
Table 4. Irrelevant content in responses from the point of view of the question and/or the picture
In order to support this with additional evidence, we administered a primary (6sentence-long) and a secondary (8-sentence-long) false belief test to our two moderate
25
Moderate AD
Generating
sequences of
figures
50
42
50
90
26
Bracketing
Generate
bracketings
Claculating
brackletings
95
100
Brocas aphasic 1.
11
25
100
Brocas aphasic 2.
16.7
The Hungarian speaking Brocas aphasics did not exhibit any limitation in arithmetical
operations, they calculated correctly, they were able to produce potentially infinite
sequences of numbers, applied recursive arithmetical operations correctly, inserted
parentheses in various combinations, even double ones. They did all that using numerical symbols and operation signs; that is to say, they were not necessarily able to verbalise their otherwise correct operations.
8. Summary
In the case of Hungarian speaking patients with moderate AD, syntactic recursion
involving embedding is relatively unimpaired, as opposed to their limited ability to
tackle theory-of-mind and arithmetical recursion. Conversely, we found limited syntactic recursion but normal theory-of-mind inferences and recursive arithmetical operations in Hungarian speaking agrammatic aphasics.
9. Conclusion
The production differences observed in the tests are explained by the fact that we do
not seem to have to do with a single recursive operation whose application may be
impaired or remain intact at various levels; rather, we encounter separate recursive
operations bound to individual linguistic and non-linguistic subsystems that may be
selectively impaired. We have seen that these operations are not independent of one
another: the impairment of one may trigger the use of another one as a substitution
mechanism or repair strategy. With respect to the relations recursive sentence structure
recursive theory-of-mind inferences, we have found that in cases of a deficit of operations of the left-hand constructions, recursive operations of the right-hand constructions can be used as parts of a repair strategy.
Recursive operations manifested in theory-of-mind inferences may also be dissociated from syntactic recursion. The use of erroneous or impaired theory-of-mind abilities is highly probable to determine person, number, and tense features of a clause, but
does not necessarily prescribe that the clause has to be a recursively embedded one.
The impaired theory-of-mind inferences were not repaired by syntactic structural recursion in the responses by our subjects.
The accessibility of recursive operations is limited in Alzheimers disease for theory-of-mind and calculation but unlimited with respect to linguistic representations;
whereas in agrammatic aphasia, linguistic representations may be disconnected from
the recursion module, theory-of-mind and calculation systems are able to access it
(cf. ZimmererVarley 2010). These dissociations argue for a theoretical model that
27
posits a module of recursive operations in the human mind that is shared by linguistic,
theory-of-mind, and arithmetical performance. This common recursion module is accessible to a limited extent for the theory-of-mind and arithmetical subsystems while it
is fully accessible for representations of linguistic constructions in Alzheimers disease, whereas in agrammatic aphasia, the representations of linguistic constructions
may be detached from the recursion module while theory-of-mind and arithmetical
systems may access it at will.
References
American Psychiatric Association 2000. Diagnostic and statistical manual of mental disorders (4th
ed., text. rev.) Washington, DC: Author.
Bnrti, Z. 2010. Recursion in aphasia. Clinical Linguistics & Phonetics. 2010, Vol. 24, No. 11:
906914.
Evans, N. Levinson, S. 2009. The Myth of Language Universals. Behavioral and Brain Sciences.
32: 429448.
Everett, D. L. 2009. Piraha culture and grammar: A response to some criticisms. Language. 85:
405442.
Folstein, M. F. Folstein, S. E. McHugh, P. R. 1975. Mini-mental state. A practical method for
grading the cognitive state of patients for the clinician. Journal of psychiatric research. 12/3: 189
198.
Hauser, M. D. Chomsky, N. Fitch, T. W. 2002. The Faculty of Language: What Is It, Who Has It,
and How Does It Evolve? Science. 298: 15691579.
Kertesz, A. 1982. The Western Aphasia Battery. New York: Grune & Stratton.
MacWhinney, B. 2009. The emergence of linguistic complexity. In Givn, T. Shibatani M. (eds.):
Syntactic Complexity. Amsterdam: John Benjamins. 405432.
Rosen, W. G. Mohs, R. C. Davis, K. L. 1984. A new rating scale for Alzheimers disease. American Journal of Psychiatry. 1984/Nov.: 356364.
Siegal, M. Varley, R. Want, S. C. 2006. Mind Over Grammar. Reasoning in Aphasia and Developmental Contexts. In Antonietti, A. Liverta-Sempio, O. Marchettio, A. (eds.): Theory of
mind and language in developmental contexts. Berlin: Springer. 107119.
Stark, J. 1998. Everyday Life Activities Photo Series. Vienna: Verlag Peter Poech. (Photo Cards:
Druckerei Jentzsch)
Takano, M. Arita, T. 2010. Asymmetry between Even and Odd Levels of Recursion in a Theory of
Mind. In Rocha, L. M. Yaeger, L. S. Bedau, M. A. Floreano, D. Goldstone, R. L.
Vespignani, A. (eds.): Proceedings of ALife X. www.citeulike.org/user/jasonn/article/7293338
Tariska P. Kiss . Mszros . Knolmayer J. 1990. A mdostott Mini Mental State vizsglat.
(The modified Mini Mental State examination) Ideggygyszati Szemle. 43: 443.
Varley, R. A. Klessinger, N. J. C. Romanowski, Ch. A. J. Siegal, M. 2005 Agrammatic but
numerate. Psychology PNAS Early Edition. www.pnas.org/cgi/doi/10.1073/pnas
Youmans, G. Bourgeoi, M. 2010. Theory of mind in individuals with Alzheimer-type dementia.
Aphasiology. 24/4: 515534.
Zimmerer, V. Varley, R. 2010. Recursion in severe agrammatism. In van der Hulst, H. (ed.): Recursion and Human Language. 393405. Berlin: De Gruyter Mouton.
28
1. Bevezets
A demencia az emlkezkpessg s vele egytt a viselkedsi s gondolkodsi folyamatok progresszv srlsre utal, melyet megfigyelhetnk agyi degeneratv megbetegedsekben, nhny metabolikus megbetegedsben s vaszkulris trtnsek kvetkezmnyeknt. Jelen tanulmnyunkban az agy degeneratv megbetegedsben szenved Alzheimer-kros beteg (AD) s az agyi rkatasztrfa eredmnyezte anmikus afzis (AA) beteg szelhvst vizsgljuk, olyan kpmegnevezsi feladatokban, amelyeknek sszelltst azok a tanulmnyok sugalltk, amelyek a kategriaspecificits
komputcis modelllsban, a szemantikai reprezentci struktrjnak magyarzatban figyelemre mltak.
Hipotzisnket s vizsglatunk cljt, mdszertani alapjt a kategriaspecificits srlsi modelljei ismeretben fogalmazzuk meg (FarrahMcClelland 1991, Devlin et al.
1988, Rogers et al. 1999). Az 1980-as vek kzeptl, azokrl a betegekrl szl jelentsek, akiknek specifikus kategrikban szelektv diszfunkciik vannak, megklnbztetett nehzsget mutatnak az l s nem l szavak elrsben.
Az l s nem l dolgok perceptulis s funkcionlis ismeretekben srltek, amelyek az agy klnbz terletein raktrozdnak (WarringtonMcCarthy 1983,
WarringtonShallice 1984).
A vizsglat clja, hogy megtalljuk azokat a tneteket, amelyek megbzhatan informatvak abban, hogy a szelhv feladatokban megmutatkoz sikertelensget (a) a
szemantikai reprezentcibl elveszett szerkezeti informcik okozzk (raktrozsi
deficit), (b) vagy valamely kognitv kpessg zavara akadlyozza a sz elrst (hozzfrsi deficit).
2. Anyag s mdszer, ksrleti szemlyek
A szelhvs vizsglatban trgykpeket hasznltunk, (i) elfogadva azt a nzetet,
hogy ugyanazt az ismeretet elrhetjk klnbz modalitsokon keresztl; (ii) egy kp
prezentlsa utn a perceptulis ismereten kvl ms informcit (funkcionlis/asszociatv) is aktivlhatunk mg akkor is, ha a trgy eredmnyes megnevezse
nem trtnik meg; (iii) hogy ennek implicit jelenltt igazoljuk s explicitt tegyk kt
nagy feladatrendszert alkottunk: egyik a szelhvs feladatrendszere (tovbbiakban
SZE feladatrendszer), msik a mentlis lexikon aktivls feladatrendszere (tovbbiakban MLA).
SZE feladatrendszer hrom mappba rendezve 3x60 darab vonalkontros, fehr
alapon fekete tollal rajzolt kpet tartalmazott. (Az inger minsget 5-6 ves figyermekkel s afzibl felplt 81 ves nvel teszteltk, nhny mdostst tettnk is29
He
Stc
I. vizsglat
111
34
II. vizsglat
102
III. vizsglat
VL
Fc
So
10
=162/180
38
=160/180
85
46
12
=153/180
He
Stc
Fc
So
31
I. vizsglat
54
33
75
= 173/180
II. vizsglat
87
19
66
= 179/180
4.2. Teljestmnyllandsg
AD (i) vizulis agnzija nyomn perceptulisan hasonl formk aktivldnak (napraforg hmzs; kerk torta, slag), amelyek a tallgatsok szmt nvelik. (ii) A
vizulis percepci integratv zavara azt jelenti, hogy a lmpa megnevezse a perceptulis rszek sszeraksval lesz eredmnyes (vasal (Nzze tovbb!) tml
(Nzze tovbb!) = lmpa). (iii) Kvetkezetes a deficit specifikus grammatikai kategrikban. A kpgyjtemny cselekvst, trtnst vagy a cselekvs irnyt is kifejez
(igekts) igket, cselekvsrnyalatokat, valamint hely (alatt, fltt, kztt) megnevezst is lehetv tette. A cselekvsek megnevezsei elmaradtak a nvszk megnevezsi rtkeihez viszonytva, s a helyzetek egyikt sem tudta felismerni.
32
AA anmikus afzia teljestmnyllandsga neologizmusokban, hangalaki s szemantikai parafzikban mutatkozott meg. Ketts stratgija vgl a hangalaki
parafzik tbbsgt eredmnyezte.
33
Rogers et al. (1999) szemantikai demencia modelljkben kiindul ttel, hogy a szemantikai demenciban szenved betegek igenis hibznak, amikor l dolgokat neveznek meg, szemben a nem l ksztmnyekkel (arte factokkal).
AD esetben az l s nem l lexmk a szelhvsban s a mentlis lexikon
aktivlsban klnbzen voltak elhvhatk. A bal oldali diagramon lthatjuk,
hogy az l kategriatagok helyes megnevezsei elmaradnak a nem lk megnevezseitl. A minsgi rtkelsben tbbet tallgat, felsbb fogalommal jell, s tbbszr ignyli a plusz funkcionlis/asszociatv, azaz szemantikai sttuszt megerst
jelzingert. Csak az els vizsglatban kiemelked a fonemikus jelzinger (10%) tmogat szerepe, amikor felttelezheten a nvszelekci is megtrtnt, m az artikulcis megformls nehznek bizonyult.
Devlin et al. (1998) srlsi szimulcijukban azt tapasztaltk, hogy kis mennyisg srlsnl nagyobb nehzsg mutatkozik a nem l dolgok megnevezsben, s
ahogy a srls felhalmozdik, gy az l dolgok megnevezse lesz nehezebb, majd a
kvetkez stdiumban mindkt struktrban elmaradnak a megnevezsek.
MLA feladatrendszerben az l dolgok tfed tulajdonsgainak egyttjrsa segti AD beteget abban, hogy a kategriatltssal rvidre zrja a sejtsek s tallgatsok sorozatt. gy tmogatst kap az l kategrik kiemelked helyes megnevezshez. A nem l kategrik megnevezsei okoznak nagyobb nehzsget. Mindkt
diagram vizsglatonknt a beteg hanyatl mentlis kpessgeit jelzi. St, a betegsg
elrehaladsval kevesebb helyes megnevezsnl a kt struktrbl csaknem egyazon
arnyban trldnek az l s nem l mintzatok mindamellett, hogy nagyobb mrtk
szemantikai megerstst kvnt az lk specifikus kategrijban.
34
35
37
38
39
1. Einleitung
Der vorliegende Beitrag ist der erste offizielle Versuch der Adaptierung des in Ungarn
seit langem bekannten psycholinguistischen GMP-Tests zur Messung der sprachlichen
Kompetenz. Die Adaptierung des Tests auf das Slowakische ergibt sich aus dem Umstand, dass es in der slowakischen Linguistik bzw. Psycholinguistik bisher keinen
hnlichen Test gibt, und damit auch nicht die Mglichkeit einer systematischen und
komplexen Untersuchung der Sprachkompetenz von Mono- sowie Bilingualen. Bekanntlich ist die heutige Slowakei hnlich wie andere mitteleuropische Staaten in der
alltglichen Sprachpraxis durch kollektive und individuelle Multilingualitt geprgt.
Individuelle Zweisprachigkeit ist dabei oft Ausdruck des Strebens von Eltern, ihre
Kinder mit bilingualer Kompetenz auszustatten, wobei die konkrete Motivation ganz
verschiedener Natur sein kann. Fr die Erforschung der Formen kollektiver slowakisch-ungarischer bzw. ungarisch-slowakischer Zweisprachigkeit steht den Hungaristen mit dem GMP-Test ein bereits bewhrtes und zuverlssiges Instrument zur Evaluierung der Sprachkompetenz des Ungarischen zur Verfgung (vgl. Vano 2005,
2007a, 2007b, 2009). Dieser psycholinguistische Test zur Messung der perzeptiven
und rezeptiven Kompetenz von Kindern ist von hoher Validitt. Dank seiner Komplexitt, Verstndlichkeit und eines problemlosen fehlerfreien Funktionierens, das auf
eindeutigen, allgemeingltigen und durchschaubaren Kriterien beruht, wurde er auch
fr die Evaluierung deutsch- und englischsprachiger Kompetenz adaptiert.
Sicherlich wurde das Original primr zu sprachdidaktischen, sprachtherapeutischen
sowie sprachdiagnostischen Zwecken in Bezug auf Monolinguale entworfen. Bei einer
exakten, breit fokussierten Bestimmung von Evaluierungswerten fr Mehrsprachige
knnte der Test jedoch auch zur Messung der Sprachkompetenz von Zweisprachigen
eingesetzt werden. Die Referenzgre, das tertium comparationis, muss jedoch die
Sprachkompetenz des Monolingualen sein. Erst auf Grund von Standards fr Monolinguale kann mit der Herausarbeitung modifizierter Minimalstandards fr Bi- bzw.
Plurilinguale begonnen werden. Dieses Bedrfnis beruht auf der Annahme Grosjeans,
dass ein Bilingualer nicht die bloe Addierung zweier Monolingualer sei, und dass er
in seiner Sprachentwicklung mglichen Versptungen sprachlicher Natur (d.h. phonetischen, grammatikalischen, lexikalischen, semantischen, stilistischen u.a.) und
sprachpsychologischer Natur (im Bereich von Perzeption und Rezeption) ausgesetzt
ist. Aktuell wird sogar ernsthaft diskutiert, ob die sog. critical period bei Bilingualen
im Unterschied zu Monolingualen deutlich anders angelegt ist, sich das Zeitfenster
muttersprachengleichen Zweitspracherwerbs nicht mit 1315, sondern mglicherweise
erst mit rund 18 Lebensjahren schliet.
40
Bild von der Sprachkompetenz der Testperson sowie von ihren Defiziten. Wegen seines Universalcharakters und seines langjhrigen Einsatzes in der Praxis durch Pdagogen, Logopden und Psychologen sowie seiner stndigen Verfeinerung haben wir
grundstzlich keine Bedenken hinsichtlich seiner Verwendung im Slowakischen.
2.2. Testaufgaben
Die vollstndige Version des GMP-Tests beinhaltet 20 Aufgaben, die sog. GMPAufgaben.
Zur Veranschaulichung dient die folgende bersicht der einzelnen Aufgaben mit
ihrer kurzen Beschreibung.
2.2.1. Richtiges Hren
GMP1 (basierend auf dem speziell dafr entworfenen Gert GOH mit 24 synthetisch
gebildeten Wrtern, die von den Testpersonen wiederholt werden mssen).
2.2.2. Perzeption der Sprache
GMP2 Identifizierung von 10 Stzen, vorgespielt mit Geruschkulisse, d.h. unter
akustisch erschwerten Bedingungen.
GMP3 Identifizierung von 10 Wrtern, vorgespielt mit Geruschkulisse.
GMP4 Identifizierung von 10 Stzen vorgespielt mit verringerter Geschwindigkeit.
GMP5 Identifizierung von 10 Stzen vorgespielt mit erhhter Geschwindigkeit.
GMP6 Identifizierung von 10 natrlichen Stzen.
GMP7 Ablesen von Wrtern von den Lippen (10 Tiernamen).
GMP10 Identifizierung von Lautreihen, Wiedergabe von Logatomen (nicht existierenden Wrtern, die jedoch den grundlegenden phonotaktischen Regeln der jeweiligen Sprache entsprechen und zugleich relativ leicht verwechselt werden knnen).
GMP14 Identifizierung des Wortakzents bzw. der Quantitt (in der slowakischen
Adaptierung).
GMP17 Identifizierung von Lautreihen (Unterscheiden bzw. Gleichsetzen von 25
Logatompaaren).
GMP18 Transformieren der Perzeption (Ersetzen der Laute durch Holzwrfel).
2.2.3. Rezeption der Sprache
GMP12 (Hrverstehen von 4 krzeren Texten fr 36 jhrige, 58 jhrige, 710 jhrige und 1014 jhrige Kinder).
GMP16 (Verstehen von 10 Stzen aufgrund zweier hnlicher Bilder, von denen eines den Inhalt des Satzes widerspiegelt).
43
44
46
47
49
(zamestnala sa vo firme ako tovnka) 7. o sa stalo so Zuzkou? (mala autonehodu, ochorela a oslepla) 8. o radili udia Matejovi? (aby si vzal in) 9.
Ako dlho vydralo nakoniec manelstvo Mateja a Zuzky? (preili spolu cel
ivot) 10. o ns u tento prbeh? (e lovek nesmie posudzova ud len poda
momentlneho vzoru a stavu a e udia asto poruuj suby, ktor dali inm,
len aby ochrnili svoje zujmy).
sk-GMP13 Feststellen der Lateralisierung der Sprache im Hirn: Feststellen der Dominanz von Hand, Ohr bzw. Bewegung/Richtung
1. Chy sa za ucho! 2. Zatlieskaj rukami! 3. Objm sa! 4. Zatl klinec kladivom do
stola!
5. Stiahni si dolu zips a daj ho zasa nasp! 6. Oto oproti sebe oba palce! 7. V
kpke vyhadaj erven spinky a vrich do peaenky!
sk-GMP14 Identifizierung des Wortakzents bzw. der Quantitt (in der
slowakischen Adaptierung)
1. Kde to boli/bol? 2. Musel zmeni tvar/tvr. 3. Babka/bbka je v splni. 4. Uvidel
pokladnku/pokladniku. 5. Mal pred sebou sud/sd. 6. Ltka/latka bola trochu tenk.
7. T vla/vila je prekrsna!
sk-GMP15 Feststellen des Synthetisierungsvermgens einer Information, (jedes
der fnf zusammengesetzten Wrter wird getrennt in zwei Hlften gleichzeitig ins
rechte und linke Ohr eingespielt).
arodejnk, muchotrvka, horolezec, zemegua, drevoruba
sk-GMP16 (Verstehen von 10 Stzen aufgrund zweier hnlicher Bilder, von denen
eines den Inhalt des Satzes widerspiegelt)
1. Myka takmer doiahne na syr. 2. Macko a zajko vyliezli na strom a jeden z nich
spadol.
3. Maka ah spoza stola myku. 4.Dievatko mus da knihu chlapekovi. 5.
Predtm ne sa macko napil, trochu si zajedol. 6. Hovoril o svieke padajcej zo stola.
7. Nie zajko si obliekol kockovan nohavice. 8. Macko be, aby ho nepotpali vielky. 9. Kee vemi sneilo, dievatko sa predsa len nelo snkova. 10. Dievatko by
bolo zjedlo tortu, keby bolo doiahlo na misku.
sk-GMP17 Identifizierung von Lautreihen (Unterscheiden bzw. Gleichsetzen von
25 Logatompaaren)
a) unterschiedliche: 1.krde/kride, 2. kriadnk/kratnk, 3.mevli/mval,
4.mza/moza,
5. du/dou, 6. rfa/rufa, 7. ps/pas, 8. doavil/dosavil, 9. odikovan/otikovan,
10. lozka/lozga, 11. rndo/rnto, 12. zum/vum, 13. vadmo/vano, 14. rask/rask,
15. lodzk/lock, 16. krba/krpa,
b) identische 17. cek/cek, 18. ahla/ahla, 19. vanbu/vanbu, 20. ebeni/ebeni, 21.
hlip/hlip,
22. misel/misel, 23. sb/sb, 24. kial/kial, 25. tlop/tlop
sk-GMP18 Transformieren der Perzeption (Ersetzen der Laute durch Holzwrfel)
Die ungarischen Vokale , mussten wegen der Nichtexistenz des -Vokals im
Slowakischen durch andere ersetzt werden.
b, e, p,
51
vl
obchod
zchod
vlnka
srnka
zhrada
rolda
chlieb
chliev
Aufgabe:
Linkes Ohr
zajac
pajc
syr
vr
kvet
svet
hruka
hska
bbika
babika
koza
kostra
pl
jeda
kneda
hrb
hrb
cibua
cedua
5. Schlussbemerkungen
Dem ersten Schritt des Entwurfs des GMP-Tests im Slowakischen werden weitere
Schritte folgen. Auer dem o.g. GMP1 mit dem GOH-Teil und einem speziell dafr
geschaffenen Gert zum Messen des richtigen Hrens von Kindern bzw. einer technisch bedingten Auswahl von geeignetem Wortmaterial mssen die meisten verbalen
zur Perzeption und Rezeption bestimmten Aufgaben in einem dafr adaptierten Spezialstudio aufgenommen werden. Erst diese empirischen Untersuchungen werden Klarheit darber verschaffen, inwieweit sich unser Anwendungsentwurf kompatibel und
52
adquat zum Original verhlt bzw. sich die Erfahrungen der Autorin in unserem adaptierten Testverfahren besttigen lassen. Seine Funktionstchtigkeit, berprfte Objektivitt, sein groer Umfang mit einer ganzen Flle von unterschiedlichsten Testaufgaben fr mehrere sprachpsychologischen Teilprozesse wrden es ermglichen, auch in
der Slowakei Mono- sowie Bilinguale testen zu knnen. Von besonders groem Nutzen wre der Test fr die Erforschung von bilingualen slowakischen Kindern in den
natrlichen bzw. sog. edukativen Diasporen in der Slowakei und im benachbarten
Ausland, aber auch in Serbien, Rumnien oder etwa den Vereinigten Staaten von
Amerika. Ein solches Vorhaben bedarf jedoch einer Ausarbeitung objektiver Kriterien
fr die Festsetzung von Minimalstandards von Bilingualen.
Literatur
Bulheller, S. Hcker, O. H. 2003. Peabody Picture Vocabulary Test. Manual. Deutschsprachige
Fassung des PPVT-III fr Jugendliche und Erwachsene. Frankfurt am Main: SWETS Test services.
Dunn, M. L. Dunn, M. L. 2003. Peabody Picture Vocabulary Test (PPVT). Manual. Frankfurt an
Main: SWETS Test services.
Gsy Mria Olaszy Gbor Hirschberg Jen Farkas Zsolt 1987. Phonetically based new method
for audiometry: the G-O-H measuring system using synthetic speech. Magyar Fonetikai Fzetek.
17: 84101.
Gsy Mria 1995. GMP-diagnosztika. Budapest: Nikol.
Gsy Mria 2006. GMP-diagnosztika. Budapest: Nikol.
Gsy Mria Imre Angla 2007. Beszdpercepcis fejleszt modulok. Budapest: Nikol.
Gsy Mria 2009. A beszdszlels s a beszdmegrts zavarai. In Tams, M. Fejlesztpedaggia.
Budapest: ELTE. 159192.
Grosjean, F. 1992. Life with Two Languages. An Introduction to Bilingualism. CambridgeLondon:
Harvard University Press.
Imre Angla 2005. Ktnyelv gyermekek beszdpercepcis teljestmnye. Beszdkutats 2005. 123
133.
Imre Angla 2006. Beszdszlelsi s beszdmegrtsi vizsglatok nmet anyanyelv gyermekek
krben. Dissertationsarbeit. Budapest: ELTE.
Imre Angla 2007a. Az anyanyelv-elsajtts vizsglata. In Gsy Mria (szerk.): Beszdszlelsi s
beszdmegrtsi zavarok a anyanyelv-elsajttsban. Budapest: Nikol. 5869.
Imre Angla 2007b. A beszdmegrts s az olvass sszefggse. In Gsy Mria (szerk.):
Beszdszlelsi s beszdmegrtsi zavarok az anyanyelv-elsajttsban. Budapest: Nikol. 184
201.
Imre Angla Grczi Tekla Etelka 2009. Egynyelv, magyar anyanyelv iskolsok beszdpercepcis folyamata idegen nyelven. In Lengyel Zsolt Navracsics Judit (szerk.): Tanulmnyok a mentlis lexikonrl. Budapest: Tinta Kiad. 326335.
Iveta, B. 2008. Fonologick vvin v detskej rei. In Daniela, S. (szerk.): tdie o detskej rei. Acta
facultatis Philosophicae Universitatis Preoviensis. Jazykovedn zbornk 23. Preov: Filozofick
fakulta Preovskej univerzity v Preove. 212250.
Kapalkov, S. Slanov, D. 2007. Adaptation of CDI to the Slovak language. In Proceedings from
the first European network meeting on the communicative developmentinventories. Gvle: University of Gvle. 3241.
53
54
1. Introduction
Linguistic research of morphological change within the last few decades has shown a
tendency to focus on processes of decomposition. Not only did the
Natrlichkeitstheorie (Mayerthaler 1981, Wurzel 1984) concentrate too strongly on
reduction processes, but grammaticalization research adhered to the tenet of
unidirectionality (Haspelmath 1999), even though for some years scientists have been
pointing to processes working in the opposite direction for example an early theory
of secretion (Jespersen 1925) or the evidences for degrammaticalization (Norde 2009).
These areas of research have a focus on composition processes in common, here
named strengthening processes. The schoolbooks take cognizance of these trends
rather slowly. Now they admit at least that these composition processes do exist but
without integrating them into an adequate theory (Szczepaniak 2009). The collection
of articles Prozesse der sprachlichen Verstrkung (Harnisch 2010) tries to illuminate
a grounded theory of strengthening processes. But if you read the individual papers,
you will notice that every author has his own comprehension of the concept. This may
be due to a wide understanding of the term strengthening in daily use and because
the concept has not yet been defined in a strictly linguistic sense (first approaches
BlowMeciarova to appear). Therefore, it is essential to explain what is here meant
by strengthening in linguistic terms.
It is apparent that morphology is a constitutive level for describing strengthening
processes. An appropriate explanation of morphological operations would be possible,
if we knew how language is processed in the mind. An important theory is provided by
Steven Pinker (1999) in his book Words And Rules. The Ingredients of Language. On
the basis of experiments he tries to describe how categories of regularity and
irregularity are mentally processed. This theory will be explained critically in section
2. Afterwards I want to examine the validity of the Words-and-Rules Theory on the
basis of data from strengthening evidence. In conclusion it is necessary to evaluate
how strengthening processes fit into the Words-and-Rules Theory.
2. What is strengthening?
In respect of ordinary language, strengthening means any kind of growing. In a
linguistic sense, however, strengthening is an accession of semantic and/or formal
aspects of a bilateral linguistic sign, leading to an increase in complexity. This implies
that newly created boundaries involve an accession of complexity in the morpheme
chains of a word (*Daum -en (Pl.)). At the same time it is possible that the sign gains
55
formal structure (Igel -s (Pl.)) or a new meaning arises, such as the advertisement of a
travel organization demonstrates (GlO -Bus). The operations involved are: (re)segmentation, (re-)motivation in short, (re-)analysis and upgrading of formal
word structure from (re-)segmented phonetic material. This morphosyntactic
upgrading means an increase on the construction level (phoneme submorpheme
morpheme word phrase). (Re-)Segmentation and (re-)motivation are equal to
reanalysis. However, it is necessary to discuss both operations separately to emphasize
the bilateral character of a linguistic sign. These processes are often based on the
mechanism of proportion analogy, following existing patterns, for example Mensch
(Sg.) : Mensch -en (Pl.) = Daum (Sg.) : *Daum -en (Pl.). Homonymy is an important
factor, too. Structure and sound of the end of the word Daumen is formally identical
with the plural morpheme -en (Mensch -en (Pl.)).
It is necessary to underline that analogy is not the activating motivation. Analogy
shows the how and not the why. It demonstrates a possible pattern according to
which a word has been built formally but it doesnt explain why, for example, a
particular pattern was chosen. Therefore, we have to add to the proportion analogy a
theory of preference to make the predication more adequate (cf. Keller 2003, Koch
2006: 17). At this point we have to ask, whether strengthening processes and analogy
depend on the way language is mentally represented. Therefore it may be helpful to
discuss the Words-and-Rules Theory by Steven Pinker (1999).
3. The Words-and-Rules Theory
In the 1980s and 1990s two major theories tried to explain how language is
represented and processed within our brains. The primary representatives of a strongly
rule based approach are Chomsky and Halle (1968/1991). Their theory was augmented
by McClelland and Rumelhart (1986). They argued that our mind works like a pattern
associator memory or a perceptron. Pinker (1999) demonstrates that both models are
insufficient to describe the difference between regularity and irregularity adequately.
One phenomenon, two models, both explaining too much to be completely wrong,
both too flawed to be completely right. (Pinker 1999: 117)
The modified words-and-rules theory by Pinker (1999) connects these approaches
and combines them into a consistent theory. It tries to explain how languages and their
grammar are memorized in our brains. The human mind is a complex instrument with
different mechanisms, which are optimally prepared for their work (cf. Pinker 1999:
146). The first general assumption is that words are randomly stored as an interaction
of acoustic signals and meaning. The meaning of a word is a link to an entry in the
persons mental encyclopedia, [...]. (Pinker 1999: 3) Furthermore Pinker states that:
Words are stored in the mental dictionary not as haphazard bundles of information
but in a standard format called a root. (Pinker 1999: 152)
The second assumption is that we can combine words and from the constitution of
the combination we can filter a meaning. In Humboldts words: the ability to combine
in respect to languages allows human beings: von endlichen Mitteln einen
unendlichen Gebrauch [zu] machen (Humboldt 1846: 108). It doesnt work merely on
56
the syntactic level, but also on the morphological level, with regard to inflectional
morphology.
The human language system also appears to be built out of two kinds of mental
tissue. It has a lexicon of words, which refer to common things such as people,
places, objects, and actions, and which are handled by a mechanism for storing
and retrieving items in memory. And it has a grammar of rules, which refer to
novel relationships among things, and which is handled by a mechanism for
combining and analyzing sequences of symbols. (Pinker 1999: 12f.)
In respect of inflectional morphology Pinker tries to illustrate the difference between
regular and irregular inflection. On the one hand, there are regular word forms and the
proposal of Chomsky and Halle that this concept is based on rules. On the other hand,
there are irregular word forms and the pattern associator memory meaning that
words or bits of words are mentally stored, associated and recollected by McClelland
and Rumelhart. Word forms are saved as roots and associated as whole forms in the
memory. This memory is not to be confused with a list as in a lexicon. It is rather an
associative memory, which links words with words and also bits of words with bits of
words (cf. Pinker 1999: 118). Pinkers theory is based on the following assumption:
Regular verbs are computed by a rule that combines a symbol for a verb stem with a
symbol for the suffix. Irregular verbs are pairs of word retrieved from the mental
dictionary, a part of memory. (Pinker 1999: 117) This implies that regular and
irregular inflection could be divided neurologically. The question is how to distinguish
them, because both mechanisms are based on general patterns (cf. Pinker: 1999). In
chapter 10 of his book, Pinker tries to give a neurological proof (cf. Pinker: chapter
10).
But first it is necessary to explain Pinkers comprehension of regularity. His
assumptions are based on a psychological comprehension of regularity. Regular here
refers to a rule that speakers treat as the default: an inflectional pattern they can apply
to any word in a category, even if the word has never been stored with that pattern, or
with any pattern, in memory. (Pinker 1999: 214)
It is very important to be aware of the fact that regularity doesnt depend on the
frequency of a pattern but rather on the character of the mental operation (cf. Pinker:
215). Pinker uses the characteristics of the German language to argue for his
understanding of regularity. The -s plural allomorph is, for example, the default
inflection with regard to nouns, although with a glance at the statistics it is the
rarest plural allomorph. Among the 200 most frequent German nouns, only two take -s
(cf. Pinker 1999: 222). The -s plural is the Notpluralendung. It is used when the
memory finds no item, for example in unusual-sounding words borrowed from other
languages (Mokassin -s), for fake words with unusual pronunciation (fneik -s), for
onomatopoetic words (wauwau -s), for acronyms (like AKWs) etc. (cf. Bartke 1998:
207f.).
The heterogeneity of the constructions, and peoples ability to apply the regular
suffix to them the first time they are faced with the choice, show that people do
57
not have to be trained to associate the suffix with each construction separately
the constructions need only be given the mental label noun. (Pinker 1999: 227).
To produce a regular form it is not necessary that our brain operates with associations.
In other words: in the case of absent associations the mind uses a rule to constitute a
word form. Another argument for the -s suffix constituting a rule is the fact that it is
more independent of its phonological environment than other suffixes. However, we
have to notice that there is evidence that the -en plural suffix could also be used as the
default suffix. Children, for example, generalize freely with -s and/ or -(e)n (Igel -s;
Lffel -n). It is remarkable that we can observe this phenomenon when the formfunction-proportion isnt constructionally iconic. Anyhow, the experiments of Pinker
and his colleagues suggest that a distinction between regular and irregular inflection
can be presumed. The human mind has the ability to acquire symbolic rules which
work with regard to virtually every example of a category. A rule is a mental operation
which manipulates variables such as verbs and nouns (cf. Pinker 1999: 202). Pinker
wants to underline his hybrid model with experiments which show: that when people
use an irregular form, they dont need to access memory at all. (Pinker 1999: 123)
Pinkers experiments lead him to the following conclusion: The ingredients of
language are words and rules. Words in the sense of operations that assemble the
words into combinations whose meaning can be computed from the meaning of the
words and the way they are arranged. (Pinker 1999: 269)
4. Strengthening Processes in First Language Acquisition Observations
Based on the noun category and the comprehension of strengthening operations
discussed earlier, it is helpful to give some examples taken from a small qualitative
study which was carried out with five children in Passau and Rostock in the summer
of 2008.
The focus is on cases which are formally undifferentiated in singular and plural.
The word-final structure is also homonymous with a plural suffix. The German
language provides some of these word forms which carry (non-marked) -plural, and
at the same time the structure of the word ends with en. Some examples are Daumen,
Tropfen, Brunnen, Kken, Besen, Socken etc. In child language corpora (CHILDESDatenbank; Anna-Korpus ZfAL; Harnisch-Korpus) I found cases such as the
following: ein Kk or ein Tropf. I want to explain the strengthening process from the
word form Kk.
First, I would like to underline that the difference in the category number between
singular and plural is not constructionally iconic. In general, singular is usually
shorter than plural. Moreover, plural is symbolized with additional phonetic
substance, namely the plural suffix (Mensch (Sg.) vs. Mensch -en (Pl.)). In cases
where words end with a -plural like Kken (Sg./Pl.) a distinction between singular
and plural that is only based on formal information is not possible, especially if there
is no additional inflection with ablaut or umlaut. Moreover, the type and token
frequency is also on a very low level. In addition, Kken usually refers to groups and
the word-final structure of Kken (Sg.) en is homonymous with plural suffix -en.
58
Also, the pattern word-stem + en is highly frequent (cf. Kpcke 1993). The
aforementioned factors indicate that through mechanisms of analogy strengthening
processes are taking place. Children analyze the word-final structure en of the word
Kken as plural suffix -en. They (re-)motivate the meaningless formal substance en to
the suffix -en. An upgrade on the construction level of the word from phonological
substance to a morpheme can be observed (word end structure en > plural suffix -en).
The basic requirement that a morpheme can be built is the awareness that Kk could
be an adequate German word-stem while -en could be an adequate plural suffix. It is
also the creation of a boundary between stem and suffix, namely (re-)structuring.
This demarcation involves an advance of formal and semantic complexity, called
Verkettungskomplexitt. We can conclude that formal substance is searching for
content.
Consequently, children remove the assumed suffix -en to make the singular. Due to
the absence of transparency concerning the plural structure and the existence of analog
proportions a product of strengthening processes comes into existence.
A small qualitative experiment which was carried out with 5 children in 2008
produced interesting findings. The aim was to find out whether the explanations are in
line with the, admittedly few, observations in the corpora. It was proposed to examine
how strong the need of children aged between 3 and 5 years is to create constructional
iconic form-function-relations and the role homonymy plays.
The design of the experiment was very simple. First, the children were shown a
picture with several objects, for example with three thumbs. The question was: What
are these?. All children answered correctly Daumen. Afterwards they were shown a
picture with only one thumb and the question was: What is this?. 3 out of 5 children
answered with: (ein) Daum. I consider this as evidence of (re-)motivation and (re)structuring of en to the plural suffix -en. It is a strengthening process leading from
phonetic substance to a morpheme.
The experiment was pseudo randomized. It was transferred to words that carry the
-plural suffix but have a different word-final structure. For example el in (Lffel (Sg.)
vs. Lffel - (Pl.)) is not homonymous to a plural suffix. Here, it is noticeable that the
first question was frequently answered with the default suffix -s, i.e. Lffel -s (Pl.). By
answering the second question often the opposition Lffel -s (Pl.) and Lffel (Sg.)
arose. In my opinion, it is strong evidence that children have the need to build iconic
form-function-relations. In this case, however, a potential function or content searches
for an expression in formal structure.
5. The Meaning of the Indication for the Dual Mechanism Model
Pinker (1999: 227) points out tellingly: A language is the product of generations of
learners and could reflect, rather than shape, their tastes and propensities. The data
(like Kk or Daum) imply the conclusion that children have a propensity for
constructional iconic form-function-relations. Also children are in a certain sense
intuitive linguists, who apply structures to words, search for these structures and
generalize them. Iconic proportions are very frequent and constitute an undeniably
useful pattern for children. But Pinker (1999: 86) also expresses something different in
59
230). In short: irregularity arises around phonetic similarity. Pinker doesnt explain
anything related to structural similarity or proportional analogy. Hence, it is difficult to
interpret the following statements about accumulated suffixes: If I may be permitted
to psychoanalyze speakers who have been dead for centuries, it probably came from a
widespread human habit: We dont like to put or keep a suffix on a word that looks
like it already has the suffix. (Pinker 1999: 60) Pinkers statement is related to the
English past tense suffix -ed. Another English example, which is close to my German
examples with respect to homonymy, is the ending of the English adjective -ly as in
friendly. Speakers get in trouble converting these adjectives into adverbs with the
suffix -ly (cf. Pinker 1999: 61). Pinker (1999: 62) notes a telling example of avoidance
of prefix doubling by the ex-US-President George Bush: I hope I stand for antibigotry, anti-semitism, anti-racism. Pinkers argumentation again related to the
suffix -ed isnt as clear:
Psycholinguists have offered several explanations. Perhaps speakers develop a
stereotype for past-tense form, namely, ends with t or d, and unconsciously
think that a stem that fits the stereotype has already been inflected and stop
themselves from adding the suffix again. (Pinker 1999: 60)
Yet, the question whether we have to classify structural stereotypes as regular or
irregular isnt answered. On the one hand, children could assume a rule behind the
form, on the other hand, they might avoid a duplicated plural suffix. These results tally
with the findings of the experiment. Therefore, we have to assume that a disyllabic
word with the word-final structure en is generally considered as a prototypical plural
form. Hence, it is unlikely that a plural suffix is being added. The mechanism causing
the avoidance is maybe speculative.
But we can try to answer the question concerning the false singular forms with
Pinkers theory by transferring his statements on the irregular past tense form wrought
to the German singular-plural problem. An experiment has shown that a lot of people
dont know that the present tense stem of wrought is work. Pinker provides a hint:
According the theory that irregulars are pairs of memorized words, an irregular pasttense form could, in principle, survive in memory without a corresponding stem.
(Pinker 1999: 65)
Is it not possible that an irregular plural form is memorized without a stem, too?
What happens if we ask for the stem that is not mentally stored? The data imply that
proportions and rules generate a prototypical singular. Then we would have to accept
the basic rule that on the one hand, forms like Daumen - are irregular, and on the
other hand, we have to assume that the forms are associated as irregular. Against an
irregular process we can also find arguments. Structural similarity implies a rule. We
dont have any sound patterns except for the word-final structure. The suffix -en could
also be considered as default suffix and than we need a stem.
If we state a regular process, we have to accept that we need a rule predicting that
an adequate plural is formally more than an adequate singular. Furthermore, such a
rule implies that a prototypical singular doesnt have a suffix (which is absolutely
true). Probably the singular form of Daumen is not yet registered with its own opaque
61
root (Daumen (Sg.)). This means, according to Pinker, that we have to implement a
rule which creates the appropriate word forms. Such a rule could be read as follows:
An adequate singular which never binds with a suffix is shorter than an adequate
plural which is created by a rule. -en is an adequate plural suffix. Hence, children have
to deduct -en from the plural word form to obtain the singular (*Daum -en (Pl.) >
Daum (Sg.)).
Later the children learn the irregular word form Daumen (Sg.) through input and
corrections. It will be memorized as a root like the irregular plural Daumen -. This
might be the reason for a lower failure rate at an older age.
6. Conclusion
The expanded words-and-rules theory is not able to make definite predictions about
the kind of processing that strengthening operations are based on. Without doubt, a
similar sound pattern associates a certain root. Only if there is not a certain root or a
similar sound pattern then we need a rule. The crux is whether structural patterns
follow regular processing or irregular associations? Pinkers hint of a prototype
phenomenon isnt satisfactory. I have tried to argue for a rule based process whereby
children satisfy their need for constructional iconic form-function-relations.
References
Bartke, S. 1998. Experimentelle Studien zur Flexion und Wortbildung. Pluralmorphologie und
lexikalische Komposition im unaufflligen Spracherwerb und Dysgrammatismus. Tbingen:
Niemeyer.
Chomsky, N. Halle, M. 1968/1991. The Sound Pattern of English. Cambridge: Cambridge University Press.
Harnisch, R. 2010. Prozesse sprachlicher Verstrkung. Typen formaler Resegmentierung und
semantischer Remotivierung. BerlinNew York: de Gruyter.
Haspelmath, M. 1999. Why is grammaticalization irreversible? Linguistics. 37/6: 10431068.
Humboldt, W. 1963. Werke in fnf Bnden. Schriften zur Sprachphilosophie. Bd. 3. Stuttgart: Cotta.
Jespersen, O. 1925. Die Sprache. Ihre Natur, Entwicklung und Entstehung. Hildesheim: Winter.
Keller, R. 2003. Sprachwandel. Von der unsichtbaren Hand in der Sprache. Tbingen: Francke.
Koch, G. 2007. Formengeographie I, Verbum, Bayrischer Sprachatlas. Bd. 5. Heidelberg: Winter.
Kpcke, K-M. 1993. Schemata bei der Pluralbildung im Deutschen. Versuch einer kognitiven
Morphologie. Tbingen: Francke.
Mayerthaler, W. 1981. Morphologische Natrlichkeit. Linguistische Forschungen. Bd. 28,
Wiesbaden: Athenaion.
McClelland, J. L. Rumelhart, D. E. 1986. Parallel Distributed Processing 1. Foundations.
Explorations in The Microstructure of Cognition. Cambridge: Cambridge University Press.
Norde, M. 2009. Degrammaticalization. Oxford: Oxford University Press.
Pinker, S. 1999. Words And Rules. The Ingredient of Language. London.
Szczepaniak, R. 2009. Grammatikalisierung im Deutschen. Eine Einfhrung. Tbingen: Francke.
Wurzel, W. U. 1984. Flexionsmorphologie und Natrlichkeit. Ein Beitrag zur morphologischen
Theoriebildung. Berlin: Oldenbourg Akademie Verlag.
62
FONETIKAI/FONOLGIAI KAPCSOLATOK
A MENTLIS LEXIKONBAN
(MAGYAR SZASSZOCICIS KSRLETEK ALAPJN)
LENGYEL ZSOLT
1. Problmafelvets
A mentlis lexikon (agyi sztr) egysgei (lemmi, mentlis szavai) kztti kapcsolatok feldertse intenzv kutatsok trgya a pszicholingvisztikban. jabban az
asszoci-cis techniktl (is) vrjk a clul kitztt eredmnyt. A szban forg kapcsolatok feldertse a lexiklis szint rtsi/produklsi folyamatokrl szolgltathat
informcit. A lexiklis folyamatok (szfelismers, -lehvs) br klnfle rszmveletek szintzist jelentik a szokvnyos beszdtevkenysg sorn a msodperc tredke alatt zajlanak; jelentsgket nveli az a tny, hogy a szfelismers s -lehvs
nemcsak lexikai mozzanatokat, hanem grammatikai folyamatokat is induklnak.
Az tlagos beszl szkincse tbb tzezer (flretve az aktv s passzv szkincs
krdst, amely befolysolja a szkincs nagysgt). Az emltett lexikai folyamatok
gyorsasga s a szkincs nagysga alapos indok annak felttelezsre, hogy a mentlis
lexikon strukturlt, rendezett; azaz anatmival s fiziolgival rendelkezik. A tnyleges mkds sorn ezek szintzist alkotnak, de a kutats (akr ltalban, akr magyar
viszonylatban) mai llsa e szintzist mg nem teszi lehetv. Annl is inkbb, mert a
kutatsok sem mindig deklarljk: mikor irnyulnak a struktrra, mikor a mkdsre.
Jelen tanulmny inkbb a struktrra, az anatmira irnyul, ezen bell a fonetikai
komponens lexmk kztti cementl funkcijra.
2. Mdszertani megjegyzsek
2.1. Adatbzisok
A cmben jelzett kapcsolatok feltrst hrom adatbzis empirikus adatai alapjn vgezzk: Jagusztinn (1985), Lengyel (2008) s Lengyel (2011). E kutatsok kzs jellemzje, hogy ugyanazt a 200 hvszt hasznltk. Jagusztinn (1985) 300-350, Lengyel (2008 s 2011) 1000-1100 ft vont be a vizsglatba. Jagusztinn (1985) s Lengyel (2008) 1014 vesek, Lengyel (2011) 1826 vesek krben vgezte vizsglatait.
Az adatbzisok alapjn lehetv vlik a fonetikai komponens szerepnek vizsglata
egyfell egy s ugyanazon (1014 vesek), msfell kt klnbz korosztly (1014,
illetve 1826 vesek) adatai alapjn. Az els irny egy s ugyanazon korosztly (1014
vesek) tekintetben biztost (2030 vnyi) kronolgiai, a msodik pedig (1014, illetve 1826 vesek kzti) genercis perspektvt. (Az emltett adatbzisok megtekinthetk a http://uni-pannon.hu cmen, tovbbhaladva a Karok s kpzsek, Modern Filo-
63
1985
2011
2008
ALAK
alap, falak
balak, falak
kalap, lak
ANYA
baja, banya
APA
kapa
pap
ASZTAL
Antal, atlasz
GY
lgy, vgy
g, ~
gy
LOM
llam, lm
blom
vrom
BABA
abba, bab
bab
bab
BAJ
BART
bann, kart
gart
65
BENYOMS
ugye ms
CSALD
almd
CSOPORT
hoport
port
DOLOG
Dorog
dog
EMBER
EREDMNY
vender
edny
ER
V
nv
FALU
FEJ
Erdly
edny
vi
rv, va, hv
gyalu
fal, gyalu
fel, hej
fel
FNY
kny
FI
Mi j
Fiat
goly, toj
fej
FOLY
folyos, goly
GYEREK
kerek
GYOMOR
komor
GYMLCS
kerek
gymr
klnc
szemlcs
HBOR
borj, bor
HANG
bank
bang
gang, lang
HARAG
Harangi
farag, harang,
harangnt, smaragd,
varangy
farag, harang
HZ
gz
fz, gz, lz
lz
hogy
meggy
Heineken, kehely
haj, hej
HEGY
HELY
HOLD
hord
old, told
ID
IGAZSG
kin, kisz
bigasg, zsk
66
gazsg
JOG
fog
fog
fog, jogging
JEGY
kegy, meggy
KALAPCS
szamars
kalcs
KATONA
batona, lakoda
KLYHA
KENYR
tenyr
KP
p, lp, pk
lp, MIP
kk, pk
KZ
ksz, lp
ksz
kp, ksz, nz
LB
bb
LMPA
LNYA
blna, bnya
LENY
lehny, trehny
MOZI
Rozi, izom
izom
MUNKA
tunka
NAP
pap
NGYZET
pap
kpzeld
NP
pp
fk, kp
p, kp, pp
NV
v, vn
hv, vn
hv, vn
CEN
lcn
OLDAL
dal, oldat
dal, j dal
dalol
RA
orra
da
OROSZLN
ORVOS
RM
kormos
krm, rm
krm
67
PAPR
papiros
PONT
mond, ponty
REND
kend, trend
trend
RSZ
bsz, msz, rz
SAROK
tarot
sorok
s, haras, s
olvas
sz
SZR
kr
SZEM
szemt, voszem
SZK
ksz
fk, ksz
szklet
SZN
szv, min
szn, szv
SZOBA
SZ
s, szsz, sz
pr, tapr
szab
s, szsz
s, sz
SZNYEG
fleg, knyeg
SZL
hll, szl, l
TR
rt, tor
TOLVAJ
tol, toll
vaj, vajsav
rvny
TRVNY
T
kt
VAJ
baj, haj, vj
VROS
sros
VENDG
VILG
kt
baj, haj, jaj, tolvaj, Vajk
baj, haj
r, gyros
kend mg!
virg
virg
68
virg
VIRG
vilg
vilg, vilgossg
VZ
tz
tz/10
rg, vilg
1985
Ady Endre, vad
2008
2011
Ady, vad
Ady
AKAR
BEMEGY
meggy
meggy
ELMEGY
elme
FELEL
FUT
kut
GONDOL
gondola
gandal, gondor
gondola
ISZIK
viszik
viszik
hiszik
GR
egr, vg r
JTSZIK
jcint
ltszik
KAP
KERES
eres
eres
KVN
Ivn
KLD
zd
LT
tl
LEL
fl
MEGBOCST
megtocsl
69
kerek
lt, tl
medvebocst
MEGLT
tl
MEGY
kegy, meggy
MOND
pont
Hold, Kond
STL
tl
sakl, stn
SZERET
veret
szerel, veret
darl, tlal
tl, tlal
TALL
Balzs
TANT
vakt
TART
art, tartrmrts
TESZ
teszt
teszt
leped
taszt
VR
VSROL
rcfol
VESZ
r, jr, kr
Veszprm, vsz
1985
2008
2011
DRGA
Prga
Prga
rgja
EGYSZER
szer
egyszem
keltszer, kpszer,
ktszer
ERS
fens, tets
vers
DES
fmes
les, vres
HES
mhes
mzes
RDEKES
rdes
70
FEHR
cserl
FIATAL
Fiat
GYORS
HANGOS
megr, tenyr
lngos, rangos
gyros, rozs
gyros
lngos
langyos, lngos
oly
nedves, retkes, tetves
KEDVES
nedves
KK
lk
fk
KEMNY
kmny, remny
kmny
kmny
KESER
kesely
kesely
kesely
KICSI
LGY
MAGAS
Magus
MAGYAR
agyar, kanyar
MLY
j, kl, mh
NAGY
hagy
NEHZ
egy g
NYUGODT
dugott
OROSZ
rossz, szr
REG
PIROS
papiros
ROSSZ
orosz
SAVANY
kallanty, kesely
SRGA
SIMA
lnycsaj
oroszln
breg
ger
ossz
kallanty
Prga, Sga, vgva
ima
STT
SZP
agyar
dett
sz, lp
lp
71
TELE
fele, telek
TISZTA
Tisza
Tisza
tszta, Tisza
fuj
ujj
A ksrletben szerepl 46 mellknvbl 35 (76,08%) provokl, 9 (19,56%) nem provokl fonetikai alap vlaszt (FEKETE, HIDEG, HOSSZ, ISMERS, LASS,
RVID, SZABAD, SZOMJAS, ZLD).
HSZ
1985
2008
LLNI
RTENI
HALLANI
vrengni
hal
vrt venni
halni
HALLGATNI
RNI
2011
halogat
brni, Irak
inni, srni
KRDEZNI
mrgezni
NZNI
Knzy
1.4. tblzat. Fnvi igenvi HSZ
72
mssalhangz
74
HSZ
ell
AKAR
kzpen
htul
ell
kzpen
akr
ALAK
alap
ASZTAL
Antal
BAJ
haj, hajj,
raj, vaj
bj
BART
kart
CSALD
cseld
bann
Csand
DOLOG
csaln
Dorog
DRGA
Prga
egyszem
EGYSZER
FALU
gyalu
FEJ
hej
FELEL
delel, kelel,
telek, telel
FNY
kny
FOLY
fej
fel
fejel
telek
goly, toj
GYEREK
kerek
GYOMOR
GYMLCS
htul
komor,
nyomor
gymr
szemlcs
szemlcs
HANG
bank, gang,
lang
HANGOS
rangos
HARAG
farag
HZ
fz, gz, lz
HELY
haj
fej, sej
HEGY
hogy
meggy
HOLD
told
75
hord
GR
egr
JR
kr
JEGY
kegy, meggy
JOG
fog
ja
KLYHA
mlha
knya
KEDVES
nedves,
tetves
tetves
KELL
mell
KEMNY
kmny
kel
remny
KK
fk, lk
KP
lp, MIP
KENYR
tenyr
kk
KESER
kesely
KZ
lp, nz
LB
bb
LGY
ksz, kp,
lp
lgy
LMPA
ppa
lrma
LNYA
bnja,
bnya,
frnya,
knya,
lrva, srga
lrva,
srga
LENY
trehny
MEGY
kegy
meggy
MLY
kl
mh, kl
MOND
fond, James
Bond/Bond,
kond, pont
pont
MOZI
Rozi
76
NAGY
hagy
NAP
pap
NP
NV
hv
NYUGODT
dugott
fk
OLDAL
RA
oldat
ara
orra, da
Oroszlny
orosz lny,
lny
OROSZLN
RM
rm, orom
PAPR
tapr
PONT
hont, kont,
mond
ponty,
mond
REND
beng, bent,
kend
beng, bent
RSZ
ksz, msz,
vsz
rz
SAROK
sorok
STT
sett
SZR
marok,
szarok,
varrok
szar
SZP
lp
SZERET
veret
SZK
fk
SZ
sz
SZN
SZL
TART
panr
s
min
szl
hll
trt
kart, part,
77
szj, szl,
szll
szv
vart
TALL
Balzs
TANT
vakt
TELE
bele, fele
TR
tr, tor, tr
Balzs
teve
br, kr, rg
TRVNY
rg, tl
tltny
VAJ
vj
VR
br, kr,
szr
VROS
gyros,
sros
VILG
virg
VIRG
vilg
VZ
tz, 10
ZENE
fene
2.1. tblzat. Szegmentum cserjvel nyert vlaszok
mgh-vz
msh-vz
ANYA
CSALD
mgh-vz
2011
msh-vz
mgh-vz
baja
LOM
BART
2008
llam, lm
bann
falb, kart
almd
78
msh-vz
ER
ben, ken
ERS
fens, tets
HES
mzes
FEHR
cserl
FELEL
Eger
le-fel, lehel
FELNTT
lentt, leltt
FI
Mi j
GYORS
rozs
HANGOS
lngos
lngos
smaragd,
varangy
HARAG
ID
GR
kin, kisz
sikr
egr
Eger
KALAPCS
szamars
KLYHA
fjhat,
krtya,
mlna,
mlyva, mr
ha, plma,
plya, vrva
mja,
knya,
mlha
KRDEZNI
KZ
mrgezni
lp
LMPA
hrfa, ppa,
prta, pva
LNYA
blna
LENY
teht
MAGYAR
kanyar
MOND
ont
NEHZ
egy g
79
NGYZET
kpzeld
NZNI
Knzy
OLDAL
j dal
ORVOS
kormos
RA
rna
PAPR
fakr
PONT
potty
SAROK
tarot
SRGA
Prga
SZOBA
foga, soma
SZNYEG
fleg
TALL
darl
TART
TR
TOLVAJ
trt
rg
kpzeld,
nzlek
falt
tr, tr
bordal
TRVNY
nlny
VSROL
rcfol
VENDG
kend mg!
2.2. tblzat. Mssalhangz- vagy magnhangzvzon alapul vlaszok
A magnhangzvz valamennyi csoport esetben gyakoribb. Enyhe szmbeli flnyben vannak a csak egy csoport ltal hasznlt VSZ-ak, a kt, illetve hrom csoport ltal
hasznltak egyenl szmban vannak. Hatrozott genercis klnbsg mutatkozik: a
kisebbek jval gyakrabban hasznljk, mint az idsebbek. E mvelet Navracsics anyagban is jelen van (Navracsics 2007).
3.6.3. A HSZ-hoz egy (ritkbban tbb) szegmentum hozzadsa rvn is szletnek
vlaszok; a megtolds trtnhet a HSZ elejn, kzepn, vgn.
80
1985
HSZ
ell /
kzpen
AD
vad
GY
lgy, vgy
2008
htul
Ady Endre
AKAR
ALAK
falak
ANYA
hanyag
anyag,
hanyag
ell /
kzpen
vad
2011
htul
htul
Ady
Ady
gy
vakar
vakar
falak
kalap
tanya, banya
banya,
genya, tanya
LOM
APA
ell /
kzpen
vrom
kapa
BAJ
apad
pap
Baja,
bajszom
baja
Baja
CSALD
csalrd
ER
vel, ver;
erd, Ern
(kzpen)
mer
HES
mhes
LET
mret
nv
vi
FI
Hv; rv
(kzpen)
va
Fiat
FOLY
folyos
GONDOL
gondola
HARAG
Harangi
HELY
harang
kehely
GR
sikr
RNI
brni
ISZIK
viszik
81
Heineken
harang
JOG
jogging
joj
Jkai, jd
KAP
Kapolcs
lehny
(kzpen)
LENY
MEGY
Megyeri
lehny
(kzpen)
meggy
RA
Dra, kora,
lra
RM
krm
krm,
srm
trend
trend
PAPR
papiros
REND
RSZ
mersz
s, haras
SZR
csszr,
sztr
SZEM
voszem
rszeg,
rszlet,
rszeges
rvsz
(kzpen)
olvas
sztr
szemt
SZK
szkely
SZ
TART
szsz
start
start
TELE
telek
TESZ
teszt
fuj
kt
L
VAJ
szklet
teszt
dl, fj, hl,
juj
kt
leped
tolvaj
vajk
82
ell /
kzpen
2008
htul
AKAR
ell /
kzpen
2011
htul
ell /
kzpen
kar
ALAK
lak
BABA
bab
bab
BART
bab
art
BEMEGY
meggy
CSOPORT
port
DOLOG
dog
(kzpen)
EGYSZER
szer
ELMEGY
EREDMNY
htul
elme
edny
(kzpen)
edny
(kzpen)
FALU
fal
GYOMOR
gyom
HALLANI
hal
HBOR
IGAZSG
borj, bor
zsk
83
gazsg
HOLD
old
KALAPCS
kalcs
(kzpen)
KP
KERES
eres
eres
KVN
Ivn
Ivn
lma
(kzpen)
LMPA
MAGYAR
agyar
MEGY
egy
MLY
MOND
egy
ond
NP
NV
agyar
p
v
nzet
(kzpen)
NGYZET
OLDAL
dal
OROSZ
rossz
dal
CEN
lcn
(kzpen)
PAPR
pr
PONT
ont
RSZ
sz
SIMA
ima
SZL
TISZTA
l
Tisza
(kzpen)
Tisza
(kzpen)
TOLVAJ
vaj
vaj
TRVNY
rvny
rvny
84
VROS
VIRG
rg
85
Tovbbi vizsglatra sarkall az a tny, hogy egyes hvszavak kevesebb tpust provoklnak, de egyikket feltnen nagy pldnyszmban, mg a msik sz tbb tpust, de
ezeket nagyjbl azonos pldnyszmban.
4.1.2. A fonetikai felpts szempontjbl hasonlsgot mutat szavak egyike kifejezetten serkent fonetikai alap vlaszadsra, a msik kevsb, pl. TR br, kr, rg,
rt, tr, tl, trdel, tor, tr, vl, vr, de SZK fk, ksz, szklet (Lengyel 2008, Lengyel 2011).
4.2. Pragmatikai befolys
Az ANYA HSZ-ra: anyag, baja, banya, genya, hanyag, jaja, tanya vlaszok szlettek,
az APA HSZ-ra apad, kapa, pad, pap. vatosan fogalmazzuk meg a krdst: mi az
oka a jl rzkelhet mennyisgi klnbsgnek? Fonetikai szempontbl nincs akkora
klnbsg az ANYA s az APA sz kztt, hogy magtl rtetd lenne, hogy az
ANYA bven, az APA pedig alig generl fonetikai alap vlaszt.
4.3. Tovaterjed hatsok: a szomszd szomszdja elv
Vannak tovaterjed hatsrl tanskod vlaszok, azaz amikor a HSZ-ra a szomszd
szomszdja aktivldik. A SZL HSZ-ra szl, szret, frt vlaszok (is) szletnek,
mikzben s ez jelenti a tovaterjed hatst a szret s a frt a szl kzvettst
elfelttelezi; gy, hogy a lncreakci kzps tagja (szl) kiesett. Hasonl jelensg
a SZR HSZ-ra adott szar s kaki (Navracsics anyagban SZR bds fordul el,
ld. Navracsics 2007). E sorba tartozik a HARAGharangnt; TOLVAJvajsav is.
Hasonl folyamatrl rulkodik a KESERkesely, SAVANYkesely, m ebben
az esetben a hvszavak rintkezse okozza a vlaszok azonossgt.
A tovaterjed hats vagy a szomszd szomszdja elv Gsy (2000) anyagban is
elfordul (KILISZPA{giliszta}cseresznye), nla a ketts asszocicis nevet viseli
e jelensg.
5. Eredmnyek
5.1. Bebizonyosodott, hogy a mentlis lexikon adatai kztti kapcsolatokban a fonetikai tnyez jelen van: a HSZ-ak ktharmadnl is tbbnl szlettek fonetikai kapcsolat
nzpontjbl is (olykor csak e nzpontbl) interpretlhat vlaszok. Mindazonltal a
mennyisgi arnyokbl vilgosan kvetkezik, hogy a nyelvi rendszer elemei kztti
cementlsban a fonetika szerepe eltrpl a szemantikhoz s grammatikhoz kpest. A fonetika szban forg funkciban val minimlis rszvtelnek egyik (valsznleg legfontosabb) oka az, hogy a magyar grammatikban a fonetikai elv nem
osztlyszervez. Nyelvtipolgiai vizsglatok derthetik fel, hogy azokban a nyelvekben, amelyekben a fonetika osztlyszervez (pl. fnevek ktelez vgzdse), a
fonetika kapcsolatteremt szerepe ms. A fonetika lexmk kztti kapcsolatteremt
funkcija (a magyarban) nem tnt teht szfajtl fggnek, de tovbbi vizsglatot
ignyel annak a klnbsgnek a kvetkezmnye, hogy az egyes szosztlyokon bell
86
az ersebb (vagy gyengbb) grammatikai, szemantikai ktds szcsoportok esetben minknt jut szhoz a fonetikai komponens. Ugyancsak tovbbi vizsglatot rdemel, hogy a nyelvi manifesztci klnbzsge hangz, rott, jelelt miknt
befolysolja egy s ugyanazon nyelv esetben a szavak kztti kapcsolatokat.
5.2. A fonetika mint a szavak kztti kapcsolatok teremtje nem csak jelen van, hanem viszonylagos autonmit is lvez. Vannak olyan kapcsolatok, melyek tartalmban
a grammatika vagy a szemantika csekly, olykor semmilyen szerepet sem jtszik.
5.2.1. Nincs szemantikai/grammatikai tltete azoknak a kapcsolatoknak, melyek a
klnbz manifesztcik kztti tjrsrl rulkodnak. Anyagunk az rott s a hangz forma tekintetben adott lehetsget ennek megtapasztalsra (GYORSgyros stb.).
5.2.2. Feltntek reverzibilis kapcsolatok (Vnv, NVv stb.), melyekben felfedezhet grammatikai/szemantikai tartalom, de ltrejttkben a fonetikai elv jtssza a fszerepet.
5.2.3. A sz szint alatti s feletti vlaszok (ALAKbalak, illetve RTENIvrt venni
stb.) fonetika rendszervez erejrl rulkodtak, ugyanis ezeket a kapcsolatokat kizrlag a magnhangzvz azonossga alapozza meg.
5.3. A fonetikai kapcsolatok ptst klnbz mveletek szolgltk.
5.3.1. Leggyakoribb mvelet a szegmentum cserje, a tbbi hrom mvelet (magnhangz-, mssalhangzvz megrzse, szegmentum hozzadsa, elvtele) nagyjbl
egyenl arnyban szorult a msodik helyre.
5.3.2. A szegmentumcsere legfrekventltabb szntere a sz eleje. A mssalhangzszegmentumok jval aktvabbak e mveletben, mint a magnhangzk (br bellk
tbb van, mint az elbbibl, teht ez a klnbsg termszetes). A mvelet alkalmazsban nem mutatkozott genercis vagy kronolgiai klnbsg.
5.3.3. A HSZ vznak megrzse rvn adott vlaszok nem szegmentlis sznt analzis eredmnyei; az analzis szintje nmileg seklyebb. E mveletben a magnhangzvz megrzsvel nyert vlaszok vannak tbbsgben. E mvelet kzeli kapcsolatban
van a szegmentum cserjvel, hiszen a vz azonossga azt is jelenti, hogy ms szegmentumok viszont cserldnek. Hatrozott genercis klnbsg mutatkozott: a kisebbek gyakrabban lnek vele, mint az idsebbek. Ez sszhangban van azzal a tbbfle formban is jelentkez tnnyel, hogy a fiatalabbak fogkonyabbak a fonetika irnt.
5.3.4. Szegmentum hozzadsa, illetve elvtele egyms tkrei sem kronolgiai,
sem genercis klnbsget nem mutattak fel, tbbsgben vannak azok a vlaszok,
melyeket legalbb kt, nem ritkn hrom csoport hasznl.
87
88
1. Bevezets
A mentlis lexikon kiplse az anyanyelv-elsajtts idszakhoz kthet, de a trolt
egysgek mennyisge egsz letnk sorn vltozik. Hrom terlett klnthetjk el az
aktv, a passzv s az ppen aktivlt rszt , amelyek kztt a lexikai egysgek mozoghatnak, tkerlhetnek egyik rszbl a msikba (Gsy 2005). A felptst reprezentl
modellek kzl a legelfogadottabbnak az gynevezett pkhlelmletet tartjk. Ennek
alapjn a mentlis lexikon egy hatalmas pkhlhoz hasonlatos, amelynek csompontjain helyezkednek el a lexikai egysgek, kzttk pedig sszekttetsek vannak, amelyek
alapveten szemantikai alapak (Aitchison 2003). Az egysgek sszekapcsoldsnak
egyik mozgatrugja, hogy a gyakorta egytt elhvott egysgek kztt szorosabb kapcsolat alakul ki. Az egyvlaszos szasszocicis ksrletek eredmnyei azt mutatjk,
hogy a mentlis lexikon elemei kztti kapcsolatok klnbz erssgek lehetnek,
ugyanis egy hvsz ersebben hv be egy adott vlaszszt, mint egy msikat (Lengyel
2008). Az egysgek kztt hromfle kapcsolatot ttelezhetnk fel: paradigmatikus
kapcsolatokat, amelyek a felcserlhetsgen alapulnak; szintagmatikus kapcsolatokat,
amelyek szkapcsolatokat hoznak ltre; s asszociatv kapcsolatokat, amelyek esetben
az egysgek vletlenszeren kapcsoldnak egymshoz (v. JacksonAmvela 2000,
Navracsics 2001, Gsy 2005).
A mentlis lexikon felptsvel kapcsolatos krdsek kzl az egyik legvitatottabb,
hogy a sztvek s a toldalkok hogyan troldnak: kln-kln vagy egy egysgknt,
esetleg kevert mdon (pl. TaftForster 1975, Ulmann et al. 1999, Caramazza 1999,
JuhszPlh 2001, Vradi 2009). A kt szlssges elmlet a holisztikus s az analitikus: a holisztikus felfogs szerint kln egysgekben troljuk az egyes toldalkolt
alakokat (Butterworth 1983), az analitikus elmlet szerint pedig csupn a tveknek van
reprezentcijuk a mentlis lexikonban, a tbbi szalakot szablyok segtsgvel generlhatjuk (VannestBoland 1999).
Egy msik fontos krds a szosztlyok szerinti trols krdse, vagyis hogy a trols sorn a nyelvi entitsokat rendszerezzk, meghatrozott szfaji osztlyokba soroljuk. Ez mr az anyanyelv-elsajtts kezdetn is gy zajlik. A gyermeki mentlis lexikon gy gyarapodik, hogy az jabb igk az igk, az jabb fnevek a fnevek osztlyhoz csatlakoznak. A fnevek s az igk osztlynak megklnbztetst igazolja az is,
hogy a magyar gyermeknyelvben nem adatoltak olyan esetet, amikor a gyermek fnvi
toldalkot illesztett volna ighez vagy fordtva (Lengyel 1981, Plh 2006).
A trolt lexikai egysgekrl klnfle informcik is kdolva vannak a mentlis lexikonban, ilyenek a fonolgiai forma s a jelents mellett a szintaktikai krnyezet,
amelyben hasznlatos az adott sz, a r alkalmazhat grammatikai szablyok, a hasznlati felttelei s a ms egysgekhez val viszonya (Fillmore 1971).
89
A mentlis lexikon aktivlsa, vagyis a lexiklis hozzfrs folyamata a beszdprodukcihoz vagy a beszdmegrtshez szksges nyelvi, illetve beszdjel keresst s
megtallst jelenti. A beszdprodukciban a kvnt sz elhvsa, a beszdmegrtsben pedig egy hallott sz megfeleltetse az egyn mentlis lexikonnak egy adott egysge kztt, vagyis (leggyakrabban) a szrts (Gsy 2005). Tbb modell is ismeretes,
amely a folyamat mkdst reprezentlja. Levelt (1989) hierarchikus modellje a kvetkez szintekben kpzeli el a szprodukcit: fogalmi elkszts, lexikai vlogats,
morfol-giai, fonolgiai s fonetikai kdols, vgl az artikulci. A lexikai vlogats
sorn jutunk el a lemma szintjre, amely a fogalom szemantikailag s szintaktikailag
meghatrozott reprezentcija. Ksbb hozzrendeljk a sz fonolgiai alakjt is, ez a
lexma szintjn trtnik. E kt szint kztt idbeni eltrs van, a lemma szint elbb
aktivldik, mint a lexma szint. Ennek altmasztsra szolglhatnak az gynevezett
nyelvem hegyn van jelensgek, amelyek esetben a lemma elhvsa sikeres, de a
lexm valamilyen okbl gtolt. Emellett a lexiklis elhvs problmi a spontn
beszdben mg a tves sztallsok s a tves kezdsek is (Gsy 2001).
Az aktivls a szasszocicis ksrletekben egy specifikus parancs, az adott feladat
vgrehajtsra vonatkozik. Az asszocicis mdszer tbb tpust klnbztethetjk
meg. A teljesen szabad asszocici sorn minden nyelvi vagy egyb megkts nlkl
kell egyms utn szavakat elhvni (lsd pl. Cser 1939, GsyKovcs 2001, 2002,
Neuberger 2008). Az egyvlaszos asszocici esetben egy hvszra kell kimondani
az els szt, ami esznkbe jut rla (lsd pl. Jagusztinn jvri 1985, Ball 1983, Lengyel 2008). Ezeken kvl ltezik mg szktett szasszocici, amikor is meghatrozott kategrij (pl. hasonl hangzs vagy ellenttes) szavakat vrnak egy hvszra.
A nyitott tesztekben pedig az adatkzlnek egy meghatrozott jelentskategrin (pl.
llatok) bell kell annyi szt aktivlni, ahny eszbe jut.
A klnbz letkor gyermekekkel vgzett asszocicis ksrletek a mentlis lexikon mennyisgi sajtossgai (a szkincs nagysgnak becslse) mellett az aktivls
sajtossgaira is fnyt derthetnek. gy jutottak el ahhoz a megllaptshoz, hogy az
angol gyermekeknl a szintagmatikai/paradigmatikai vlts a hetedik letv krl
kvetkezik be, vagyis mg a fiatalabb gyermekek szintaktikai sszefggsek rvn
aktivljk a szavakat (pl. a kutya hvszra az ugat szt mondjk), addig efltt az
letkor fltt azonos nyelvtani kategrin, illetleg jelenskrn bell maradnak (pl. a
kutya szra a macska aktivldik) (Berko Gleason Bernstein Ratner 1998). A magyar gyermekek asszocicis stratgiinak vizsglata azt igazolta, hogy a mentlis
lexikon dnten szemantikai elrendezdse mellett a hangzsi hasonlsgon alapul,
illetve a szintagmatikai viszonyok is jelen vannak (GsyKovcs 2001).
A jelen ksrlet clja pontosabb ismereteket szerezni a gyermekek szelhvsi sajtossgairl az anyanyelv-elsajtts klnbz szakaszaiban, kt beszdhelyzetben:
asszocicis ksrlet sorn s spontn beszdben. A spontn beszd vizsglatakor kitrnk a lexiklis elhvs problmjnak vizsglatra is. Arra a krdsre keressk a
vlaszt, hogy a szabad szaktivlsok s a beszdszndknak megfelel szelhvsok
milyen sszefggseket mutatnak. Hipotzisnk szerint az letkor elrehaladtval
gyarapods mutatkozik a szavak szmban s a szfajok tern is, fejldik az asszocicis stratgia, s ez hatkonyabb teszi a szelhvst, ezrt kevesebb diszharmnia
tapasztalhat majd a lexiklis hozzfrs sorn spontn beszdben is.
90
92
Megvizsgltuk az aktivlt szavak szfaji kategriit. A korbbi szasszocicis ksrletek tapasztalatai azt mutatjk, hogy a szabad asszocici sorn legnagyobb szmban
fnevek jelennek meg, s ez a jelen ksrletben is gy alakult. Az vodsok ltal aktivlt szavak 97%-a fnv, emellett megjelentek mg kis szmban mellknevek (2%) s
igenevek (1%) is. A kisiskolsoknl a fnevek arnya kzel 91%-os, nluk a 871 szbl 54 volt ige, 14 mellknv, 8 nvms s 3 szmnv. A hetedik osztlyosok 96,5%ban mondtak fneveket, 3,2%-ban mellkneveket, 0,3%-ban egyb szfaj szavakat.
Az vodsok mellknevei sznnevekre (zld, kk) s nemzetisgnevekre (nmet, szlovk) korltozdnak, az alssoknl ezek mellett megtallhatjuk a meleghideg, a szomorvidm stb. ellenttprokat is, a felssk pedig a sznnevek (pl. narancssrga,
neonzld) s az ellenttprok (pl. meleghideg, aprnagy) mellett mg nhny fnvbl kpzett mellknevet is elhvtak, pldul poros, kocks, virgos. Igk csak az
alssoknl jelentek meg, k ltalban kijelent md, jelen idej, egyes szm, harmadik szemly igket mondtak, mint pldul: csinl, lakik, eszik, jtszik, nevet, sr. A
nyolc darab nvmsbl hatot egyetlen kisiskols sorolta (n, te, , mi, ti, k).
Feltteleztk, hogy a toldalkolt szavak, sszetett szavak, illetleg szszerkezetek is
lesznek a szlistban, hiszen az asszocicis ksrletekben nem csak tszavakat aktivlnak a beszlk. A hatveseknl 78%-ban, a kilencveseknl 71%-ban, a tizenhrom
veseknl pedig 69%-ban jelentek meg tszavak. sszetett szavak az vodsoknl s a
kisiskol-soknl egyforma mrtkben (az sszes sz 16%-a), a nagyobbaknl 14%-ban
fordultak el (pl. jtsztr, teniszlabda, glyatbor). A toldalkolt szavak kz soroltuk
a klnfle kpzett szavakat (pl. kutyus, poros, sels), valamint a tbbes szm jelvel
elltott alakokat (pl. bogarak, csillagok, tanrok) ezeket legnagyobb arnyban a felssknl adatoltuk (az sszes sz 12%-a), az alssok szavai kztt 9%-ban, az vodsokban csupn 5,5%-ban tallhatk ilyen alakok. Szszerkezet (pl. intelligens lny, kznsges hzimacska, kovalens kts) a legkisebbeknl alig, a kt iskols csoportban pedig
4-4%-ban jelent meg.
A szavak rendszerezsnek harmadik mdja a szfaji kategrik s az alaktani
jellemzk mellett a szkszlet elemeinek jelentsmezkbe, szemantikai mezkbe
vagy fogalomkrkbe val sorolsa. A lexmk az egyes mezkn bell klcsnsen
93
Mindhrom korcsoportban a legtbb plda a szemantikai stratgia mkdtetsre szletett: az vodsok az esetek 31%-ban, az alssok 42%-ban, mg a felssk tbb
mint felben, 58%-ban szemantikai stratgit alkalmaztak. Hangzsi stratgit az
vodsok 3%-ban, az alssok 5%-ban, a felssk 2%-ban hasznltak, szerkezeti stratgit pedig az vodsok 1%-ban, az alssok 3%-ban, a felssk pedig 6%-ban. Az
eredmnyek azt mutatjk, hogy az letkor nvekedsvel fejlds ll be az asszocicis stratgikban.
Az elhvs nehzsgeit a szasszocicikban az ismtlsek is mutathatjk. Ilyenkor a gyermek zskutcba kerl, elakad a szelhvs, s egy, mr korbban aktivlt
sz jra megjelenik az asszocicikban. A hatves gyermekek 50%-nl, a kilencvesek 66%-nl, valamint a tizenhrom vesek 58%-nl azt fedezhettk fel, hogy egyegy sz jra megjelenik egy adott gyermek szsorban. Ez azonban a sok szt aktivl
gyermekeknl is gy volt, nluk az ismtlsek a j teljestmny velejri lehetnek
(minl tbb szt mondott, annl nagyobb volt a valsznsge, hogy egy sz ktszer is
szerepelt).
Elemeztk az elhvs nehzsgeit a spontn beszdben is. A lexiklis elhvs
problmi kzl a jelen ksrletben a tves szavakat, a tves kezdseket s a nyelvem
hegyn van jelensgeket figyeltk meg. Ezeknek a lexikai hibknak az elfordulsa
klnbzkppen alakult a hrom letkori csoportban (4., 5. bra).
95
97
Irodalom
Aitchison, J. 2003. Words in the mind. An introduction to the mental lexicon. Malden: Blackwell.
Ball Larissza 1983. Magyar verblis asszocicik 1. Szeged: Juhsz Gyula Tanrkpz Fiskola.
Berko Gleason, J. Bernstein Ratner, N. (eds.) 1998. Psycholinguistics. Orlando: Harcourt Brace
College Publishers.
Boersma, P. Weenink D. 2009. Praat: doing phonetics by computer. Version 5.1.05.
http://www.praat.org
Butterworth, B. 1983. Lexical representation. In Butterworth, B. (ed.): Language production. Vol. 1.
London: Academic Press. 125153.
Caramazza, A. 1999. Hny feldolgozsi szint van a lexikai hozzfrsben? In Bnrti Zoltn (szerk.):
Nyelvi struktrk s az agy. Neurolingvisztikai tanulmnyok. Budapest: Corvina Kiad. 249287.
Crystal, D. 2003. A nyelv enciklopdija. Budapest: Osiris Kiad.
Cser Jnos 1939. A magyar gyermek szkincse. Budapest: Magyar Pedaggiai Trsasg.
Fillmore, C. J. 1971. Types of lexical information. In Steinberg, D. D. Jakobovits, L. A. (eds.):
Semantics: An interdisciplinary reader in philosophy, linguistics and psychology. Cambrige:
Cambridge University Press. 370392.
Gsy Mria 2001. A lexiklis elhvs problmja. Beszdkutats 2001. 126143.
Gsy Mria Kovcs Magdolna 2001. A mentlis lexikon a szasszocicik tkrben. Magyar
Nyelvr. 125: 330354.
Gsy Mria Kovcs Magdolna 2002. The mental lexicon: Results of some word association
experiments. Acta Linguistica Hungarica. 49: 179224.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Osiris Kiad.
Jackson, H. Amvela, E. Z. 2000. Words, meaning and vocabulary. LondonNew Yrok: Cassell.
Jagusztinn jvri Klra 1985. Magyar verblis asszocicik 2. Szeged: Juhsz Gyula Tanrkpz
Fiskola.
Juhsz Levente Plh Csaba 2001. Tbbmorfms szavak megrtse a magyarban. In Plh Csaba
Lukcs gnes (szerk.): A magyar morfolgia pszicholingvisztikja. Budapest: Osiris Kiad. 1137.
Lengyel Zsolt 1981. Gyermeknyelv. Budapest: Gondolat Kiad.
Lengyel Zsolt 2008. Magyar asszocicis normk enciklopdija I. Budapest: Tinta Knyvkiad.
Levelt, W. J. M. 1989. Speaking. From intention to articulation. Cambridge: The MIT Press.
Navracsics Judit 2001. Ktnyelvek mentlis lexikonnak jellegzetessgei. Alkalmazott Nyelvtudomny. 1: 5161.
Neuberger Tilda 2008. A szkincs fejldse vodskorban. Anyanyelv-pedaggia 3-4.
http://www.anyanyelv-pedagogia.hu/cikkek.php?id=86
Plh Csaba 2006. A gyermeknyelv. In Kiefer Ferenc (szerk.): Magyar nyelv. Budapest: Akadmiai
Kiad. 753789.
Taft, M. Forster, K. I. 1975. Lexical storage and retrieval of prefixed words. Journal of verbal
learning and verbal behavior. 14: 638647.
Ullmann, M. T. et al. 1999. Neurolgiai sztvls a nyelven bell: bizonytkok arra, hogy a mentlis
sztr a deklaratv memria rsze, a nyelvtani szablyokat pedig a procedurlis rendszer mkdteti? In Bnrti Zoltn (szerk.): Nyelvi struktrk s az agy. Neurolingvisztikai tanulmnyok. Budapest: Corvina Kiad. 443467.
Vannest, J. Boland, J. E. 1999. Lexical morphology and lexical access. Brain and Language. 68:
324332.
Vradi Viola 2009. A toldalkok a lexiklis hozzfrs folyamatban. In Lengyel Zsolt Navracsics
Judit (szerk.): Tanulmnyok a mentlis lexikonrl. Budapest: Tinta Kiad. 5768.
98
A publikci elksztst a TMOP 4.2.1./B 09/1/KONV-2010-0007 szm projekt tmogatta. A projekt az j Magyarorszg Fejlesztsi Terven keresztl az Eurpai Uni tmogatsval,
az Eurpai Regionlis Fejlesztsi Alap s az Eurpai Szocilis Alap trsfinanszrozsval valsult meg.
99
szem eltt kell tartanunk azt is, hogy az egyn a vilgot kzssgben tanulja, ismeri
meg, s a kzssg viselkedsmdja, kommunikcija is hatssal van kipl kognitv
rendszerre. A beszdnek teht azon tl, hogy tri ismereteket kzvett, fontos szerepe
van abban is, hogy a tri informcik megfogalmazsa (bizonyos tri vonatkozs
nyelvi formk, pl. raghasznlat) rvn a kategorizcis folyamatok miatt klnbsgeket mutat egyni kognitv trkpeket kzelti egymshoz.
A ktirny kapcsolat abban is megragadhat tovbb, hogy a kognitv trkp mindenkori szervezdse meghatrozza a trrl val nyelvhasznlatot, a terletrl val beszdet. Ez pedig azrt nagyon lnyeges, mert ebbl addan a nyelvi adatokat (kztk a
helynevek hasznlatt) vizsglva is kvetkeztetni tudunk a trbeli orientcit irnyt
kognitv struktrkra, a kognitv trkp felptsre. A tovbbiakban azt igyekszem
megmutatni, hogy ez a kapcsolat milyen jelleg, illetve hogy miknt trhat fel. Minthogy azonban a tri nyelvhasznlat problmakre igen kiterjedt, ezrt (ahogy jeleztem)
kutatsi tmmhoz kapcsoldva ennek csupn egy ezidig igen keveset vizsglt aspektusra, a helynvhasznlat s a tri reprezentcik viszonyra fkuszlok, s a meglehetsen kis szm vonatkoz kutats alapjn a vizsglat irnyait igyekszem vzolni.
2. A beszd sorn a beszl valamire s valahogyan hasznlja a helyneveket, a kognitv szemllet nyelvszeti vizsglatok sorn ezrt a helyneveket nem a nyelvi krnyezetkbl kiszaktva elemzik, hanem azt a nyelvi struktrt is figyelembe veszik,
amelybe a nv begyazdik.
2.1. Heinrich Andrea ebbl kiindulva a Szatmr megyei, romn, magyar, svb s cigny etnikum Szaniszl magyar helyneveit vizsglva igyekezett kpet alkotni a magyar nyelv szaniszliak kognitv trkprl. Feltrta, hogy a kognitv trkpk FENT
LENT, BENTKINT viszonyai nem a vals vertikalits szerint szervezdnek, hanem egy
hierarchikus viszony (kzpont s perifria) szervezi ezek szlelst. A FENT hely (...)
amellett, hogy KZEL hely, a mienk is, hozznk tartoz, a magyar szaniszliak a falunak az aszfaltozott ft (Hajnal utca) kt oldaln elterl rszt emltik FENThelyekknt (itt tallhat a kt templom, a kultrhz s a tancs). Pl. Mr akr a szft
[nptancs], akr a Fnyes, akr a kultrba menc, akr a doktorhoz menc, akr a
milcihoz menc, ezek mind fent vannak. Ha felmegyek a Hajnal utcn, van a bt, az is
csak fent van. A perifrikus, a LENT-hely, az tvolabb esik tlnk, nem a mienk, kevsb hozznk tartoz, LENT-helynek minslnek azok a helyek, ahol ms etnikumhoz tartozk laknak, ilyen pldul a magyar szaniszliak szmra a cignyok ltal
lakott Szke-rt vagy a romnok lakta Bnt (2000: 3132).
A tri informcik teht ltszlag egy sajtos nyelvi szrn keresztl fogalmazdnak
meg nyelvileg. Szilgyi N. Sndor ezt a jelensget a szavakhoz kapcsold rtkjelentsekkel, illetve a kzttk megvalsul szolidris rtkek vonzsval magyarzza, eszerint a pozitv rtkjelents szavak asszociciszeren ms pozitv jelentsekkel kapcsoldnak ssze, mg a negatv jelentsek ms negatv jelentsekkel. A nyelvi (a beszdnkben megjelen) vilgban pedig egy sajtos trstruktra figyelhet meg: amelyben a trbeli viszonyt elemek rtkjelentsk szerinti polaritsuknak megfelelen tmrlnek a nyelvi tr kt trfeln (1997: 1213). Szilgyi N. ezt a jelensget egyrtelmen s kizrlagosan nyelvi jelensgknt magyarzza: a szavak szemantikai tartalm101
sre is hasznlt nevei esetben ugyanakkor nem ritka, hogy egy nvhez mindkt vgzds jrulhat, taln annak megfelelen, hogy pp melyik helyfajtra akar utalni vele a
beszl, pl. Htuls-hegy, -be, -re (Kovts 2000: 92); Remete-hegy, -be, -re (125). A
Pska-hegy (81), Szr-hegy (83), Szeml-hegy (83) nevekhez azonban, noha erdt is
jellnek vele, az adatok szerint mgsem kapcsoljk hozz a benne viszony vgzdseit.
A nvads-nvhasznlat fenti ellentmondsosnak tn vonsai kognitv nzpontbl knnyen magyarzhatk. A helynevek ugyanis a kzszkhoz hasonlan a kognitv folyamatok nyelvi megvalsulsai, ennek megfelelen magukon viselik a megismer folyamat mgtt meghzd osztlyoz, rendszerbe illeszt mozzanatok jellemz vonsait. A megismersi folyamat eredmnyekppen pedig a kiemelkedst s az
azt bort erdsget az egyn nem vlasztja el lesen egymstl, a kt objektumot
valjban egyazon denottumknt dolgozza fel. Ezt a szemlletet pedig a nvhasznlat
maga is tmogatja, vagyis a viszony klcsns: nincs elklnlt neve az sszetartoz
helyeknek, ezrt nem is tekintik kln objektumoknak ket. Amennyiben pedig nem
klnlnek el a helyek a tudatunkban, akkor nem is vrhat, hogy a szvegbefoglalsuk mdjban (raghasznlat) szablyossgot tallunk.
2.3. A nyelvi krnyezet mellett termszetesen a beszlknek a kognitv rendszerre,
ismereteire is tekintettel kell lennnk. A nyelvi rendszerbl nem magyarzhat pldul, hogy a mr idzett Szaniszl hatrban lv fldterletek neveihez mi alapjn jrulnak bels vagy kls helyviszonyragok (hogy a hatrban mirt a Cserre s a
Ptererdre mennek, s mirt a Nyresbe). Heinrich Andrea szerint a vgzdsek
megoszlst esetleg kls krlmny: a beszlknek a fldterletek minsgre vonatkoz ismeretei befolysolhatjk. Taln arrl lehet sz, hogy a kls helyviszonyraggal hasznlt nevek ltal jellt terletek jobb minsgek, ilyen a Csere s a
Ptererd, ezek ugyanis szntk, a belsvel hasznltak pedig rosszabbak, mint az
idzett Nyres, ez ugyanis parlag (2000: 10). Az ilyen jelleg, tletszernek tn felvetsek nyilvnvalan igazolsra vagy pp cfolsra vrnak.
3. Vgezetl egy, a nvtani-nyelvszeti szakirodalomban mintegy hatvan ve megfogalmazdott vlekedst kell jragondolnunk. A helynevek szereprl korbban azt
tartottk, hogy az ember elssorban a trbeli tjkozdsnak megknnytsre hasznlja ket, Lrincze Lajos megfogalmazsban: a fldrajzi nevek keletkezsnek oka
[...]: az ember tjkozdni akar a tjban, s ezt a tj feltnbb felsznalakulatai segtsgvel ri el, azok ltalnos (kznvi) vagy klnleges (tulajdonnvi) megjellsvel
(1947/1967: 3). A kognitv megkzelts vizsglatok azonban cfoljk ezt a felfogst: a tjkozdshoz alapveten nem szksgesek helynevek (Heinrich 2000). Jl
mutatja ezt, hogy egy idegen, a helyet nem ismer szemly tbaigaztsakor ktflekppen jrhatunk el: vagy megmutatjuk az utat, vagy szbeli eligaztst adunk, ez
utbbi mdszer azonban akkor sikeres, ha olyan fogalmakat, kzneveket hasznlunk
fel, amelyek nyelvileg minstenek, kategriba sorolnak, vagyis megrtskhz
elg a nyelvet tudni (v. Szilgyi N. 2004). A helynevek viszont alapveten nem
ilyen jelleg nyelvi elemek: csak akkor segtik a tjkozdst, ha a tjkozdni akarnak mr van egyfajta, nem csak nyelvi tapasztalaton alapul tudsa a terletrl. E
103
nyelvi elemek lokalizlshoz meg kell jegyezni a nevet a vele megjellt terlethez
kapcsol szablyt. A helynevek ismerete eszerint egy olyan plusztudst felttelez (a
nevet a terlettel sszekapcsol jegyeknek az szben tartst), amire a trbeli orientci sorn nem felttlenl van szksg. A helynevek teht nem a tjkozdst irnytjk, alapvet szerepk sokkal inkbb abban van, hogy egy adott beszlkzssg tagjai szmra knnyebb, egyrtelmbb tegyk egy adott terlet megjellst (Heinrich
2000: 5, v. Hochbauer 2010: 108).
A kognitv szemllet vizsglatok alapjn az is megllapthat teht, hogy a trben
val eligazods elssorban nem a helynevek ismerett (a hely s a nv sszekapcsolst) ignyli, hanem szmos ms ismeret (Lynch eredmnyei alapjn: utak, hatrvonalak, szektorok, csompontok, irnypontok, illetve ezeknek az elemeknek a viszonya;
1960: 7) trolst felttelezi, lnyegben teht a kognitv trkp (a tr kognitv reprezentcijnak) felptse irnytja a trbeli tjkozdst. A kognitv trkp mkdst
azonban, amint utaltam r, a beszd is tkrzi, illetve befolysolja is.
4. Dolgozatomban azt igyekeztem szemlltetni, hogy annak ellenre, hogy az ilyen
jelleg vizsglatok mg kezdeti stdiumban vannak, az effle kitgtott (a helynevek, a
helynvhasznlat s a kognitv trkp kapcsolatt szem eltt tart) vizsglatok milyen
haszonnal jrhatnak a tri reprezentcikrl val ismereteink bvtsben, ms oldalrl pedig a nvhasznlatrl val ismereteink is gazdagthatk ezek rvn.
Irodalom
Dll Andrea 2007. A krnyezet hatsa a tanulsi folyamatokra: krnyezet s alkalmazkods. In Cspe Valria Gyri Mikls Rag Anett (szerk.): ltalnos pszicholgia 2. (Tanuls emlkezs
tuds). Budapest: Osiris Kiad. 111153.
Fehr Krisztina 2011. A nyelv modularizld hlgrammatikja s az alkalmazott nyelvtudomny.
Modern Nyelvoktats. 17/2-3: 87100.
Heinrich Andrea 2000. Szaniszl helynevei kognitv nyelvszeti megkzeltsben. (Szakdolgozat).
Kolozsvr.
Hochbauer Mria 2010. Tr- s nyelvhasznlat a mindennapi tjkozdsban. Magyar Nyelvjrsok.
48: 99110.
Jackendoff, R. 1996. The architecture of the linguistic-spatial interface. In Bloom, P. Peterson, M.
A. Nadel, L. Garrett, M. F. (eds.): Langue and Space. Dambridge: A Bradford Book The
MIT Press.
Kovts Dniel 2000. Az abaji Hegykz helynevei. A Storaljajhelyi Kazinczy Ferenc Mzeum
Fzetei 4. Storaljajhely.
Landau, B. Jackendoff, R. 1996/2003. What and where in spatial language and spatial
cognition. Behavioral and Brain Science. 16: 217265. Magyarul: A mi s a hol a tri nyelvben s a tri megismersben. In Plh Csaba Kirly Ildik Lukcs gnes Racsmny Mihly
(szerk.): A tri megismers s a nyelv. Budapest: Gondolat Kiadi Kr. 69125.
Lrincze Lajos 1947/1967. Fldrajzi neveink lete. Budapest: Nptudomnyi Intzet. jrakzlse:
Magyar Nyelvjrsok. 13: 327.
Lynch, K. 1960. The imagine of the city. Cambridge: The MIT Press.
Nnay Bence 2000. Elme s evolci. (Az elmefilozfia s a kognitv tudomny tudomnyos evolcis
megkzeltse). Budapest: Kv Kiad.
104
Plh Csaba Kirly Ildik Lukcs gnes Racsmny Mihly 2003. A tr a szavak vilgban. In
Plh Csaba Kirly Ildik Lukcs gnes Racsmny Mihly (szerk.): A tri megismers s a
nyelv. Budapest: Gondolat Kiadi Kr. 726.
Plh Csaba Lukcs gnes Kas Bence 2008. A sztr pszicholingvisztikja. In Kiefer Ferenc
(szerk.): Strukturlis magyar nyelvtan 4. (A sztr szerkezete). Budapest: Akadmiai Kiad. 789
852.
Reszegi Katalin 2009. A kognitv szemllet lehetsgei a helynvkutatsban. (A metonimikus nvads). Magyar Nyelvjrsok. 47: 2141.
Reszegi Katalin 2010. A kzneveslsrl. Nvtani rtest. 32: 143149.
Szilgyi N. Sndor 1997. Hogyan teremtsnk vilgot? Rvezets a nyelvi vilg vizsglatra. Kolozsvr: Erdlyi Tanknyvtancs.
Szilgyi N. Sndor 2004. A jelentsvilg szerkezete. Elhangzott Kolozsvrt 2004. november 18-n a
Mindentuds Egyeteme eladssorozat keretben. http:// www.mindentudas.hu/szilagyi/index.html
105
1. Problmaflvets
A formlis szemantika a jelents meghatrozsnl azon felttelekre fkuszl, amelyek teljeslse esetn egy kijelents (propozci) igaznak tekinthet (Kiefer 2007).
Kognitv nyelvszeti megkzeltsben azonban az igazsgfelttelek listzsa nem
kpes feltrni a jelents struktrjt, gy pldul a kategrin belli idbeli, trbeli
vagy logikai sszefggseket. Ezt a problmt a kognitv nyelvszet a keretszemantika
segtsgvel prblja feloldani, amely szerint a jelentst rszben a fogalmi keret hatrozza meg (KvecsesBenczes 2010).
2. A fogalmi keret (frame) s a mentlis tr (mental space) Lekpezsek
(correspondences)
A fogalmi keret (frame) tapasztalataink idealizlt, sematizlt reprezentcija, a valsg prototpusokra pl vltozata. Ezt szemllteti Fillmore pldja, a REGGELI prototpusa (Fillmore 1982, idzi Kvecses 2010: 55), melyet a kvetkez kijelentsek
krvonalaznak:
A reggelit huzamosabb alvs utn esszk.
A reggelit a nap elejn fogyasztjuk.
A reggeli sajtos telekbl ll.
A fenti lltsok azonban idealizltak, hiszen tbbfle mdon is eltrhetnk a REGGELI prototpustl. Pldul lehet, hogy egsz jszaka nem alszunk, a reggeli rkban
mgis esznk valamit. Elfordulhat, hogy nem tojsrntottt fogyasztunk a nap els
tkezsekor, hanem egy szelet tortt. Mgis ezeket az tkezseket reggelinek nevezzk.
2.1. A fogalmi keretek olyan kognitv folyamatok alapjt is kpezhetik, mint amilyen a
metafora s a metonmia.
Mg a metonmia egy fogalmi kereten bell fejti ki hatst, addig a metafora kt fogalmi keret kztt teremt kapcsolatot. Az egy tartomnyon belli vagy a kt tartomny
kztti megfelelseket lekpezseknek (correspondences, mapping) nevezzk (Kvecses: 2005).
2.2 A mentlis tr is a tapasztalataink strukturlt halmazait jelenti meg, de az a beszd
s a megrts sorn jn ltre (KvecsesBenczes 2010). Lekpezsek egy mentlis tr
s egy fogalmi keret kztt is megvalsulhatnak. A mentlis tr s a fogalmi keret
kztti klnbsg rzkeltetsre az albbi plda szolgl:
106
Mentlis tr
vev
Kati
elad
Pisti
pnz
(!)
ru
bicikli
tranzakci
vett
107
Fogalmi keret
Mentlis tr
mreszkz
kt serpeny
sly
A fogalmi keret szerint a mrleg sly mrsre szolgl eszkz. A tovbbi vizsglds
trgya, hogy milyen mentlis tr aktivldik a fenti fogalmi kerethez.
Az idzett mondat egy jsgcikk elejn ll, de megelzi egy hrommondatos rvid
bevezet, amely szerint a gazdasgi szakrtk nem rtenek egyet a gazdasgi problmk megoldsban. A vizsgland mondatunk eltt kzvetlenl az albbi mondat olvashat:
A deregulci tegnapi szszli szerint a kormnyoknak passzvnak kellett volna
maradniuk, kzgazdszok szerint azonban az llami kiadsok hatsosak voltak,
s elhrthatv tettk az jabb depresszit. Abban is sokan egyetrtenek, hogy
hiba, ha csak a mrleg egyik oldalt vizsgljuk.
A bevezetben teht a VITA mentlis tere jelenik meg az egymssal szembenll
gazdasgi szakrtkkel, rvekkel s ellenrvekkel.
Ebben az sszefggsben az EGYETRTS mentlis terben tallhat mrleg a
VITA mentlis terhez mint az IGAZSGOSSG mentlis tere kapcsoldik. Ismert
Justicia brzolsa kezben a mrleggel mint az igazsgossg jelkpvel.
A fenti mentlis terek kztti megfelelsek a kvetkezk:
VITA
IGAZSGOSSG
mentlis tr
mentlis tr
mrleg
a kt vlemny
kt serpeny, kt oldal
kpviseli
rvek, ellenrvek
rvek, ellenrvek
108
magnszektorrl). Nemcsak azt kell nzni, hogy egy orszg vagy egy cg mivel
tartozik, hanem az eszkz oldalt is ltni kell.
Az utols mondat a mrleg sz szmra a SZMVITEL mentlis teret hozza ltre.
A tartozik s az eszkz oldal pnzgyi-szmviteli szaknyelvi kifejezsek alapjn az
rt olvas szmra egyrtelm, hogy a tovbbiakban a szmviteli mrleg kt oldalrl, az aktv s a passzv oldalrl van sz.
Teht a mrleg kifejezsre a fenti szvegben kt mentlis tr is kpzdtt, melyek
kztt a kvetkez megfelelsek alakultak ki:
IGAZSGOSSG
SZMVITEL
mentlis tr
mentlis tr
mrleg
mrleg
kt serpeny
kt oldal
rvek, ellenrvek
aktvk, passzvk
109
Input 1 s Input 2 a mentlis tereknek felelnek meg. A generikus trben jelenik meg az
inputok kzs struktrja. Az integrlt tr (blend) a mentlis terek projekcija, amely
magban foglalja a generikus teret, de egy j specifikus struktrt is (Emergent
Structure) (FauconnierTurner 2002: 4042, 47).
A fogalmi integrciban ltalban ez a ngy tartomny mkdik egytt, de szmos
input s tbbszrs elegyts terek (integrlt tr, blend) is kialakulhatnak (Kvecses
Benczes 2010).
4.1. A fogalmi integrci szempontjbl elemzsre kerl szaknyelvi szveg a kvetkez:
Kirly (Jlia) brlta a kilakoltatsi moratrium meghosszabbtst is, vlemnye szerint ennek a lpsnek az lehet a hatsa, hogy meggyilkolja az egybknt is
nehezen mkd jelzlogpiacot. A szocilis hl megteremtse s fenntartsa az
llam feladata, de ebben az esetben nem a kilakoltats felfggesztsvel, hanem
a szocilis brlaksok ltrehozsval kellene a problmt orvosolni. (Napi Gazdasg 07. 10. 2010)
A fenti szveg alapjn az albbi hlzat trkpezhet fel:
Input1
gyilkos
gyilkol
Lehetsg tr
Generikus tr
cselekv szemly
cselekvs
110
Input2
llam
bevezets
fegyver
ldozat
eszkz
hosszabb
irnyuls
a moratrium
jelzlogpiac
Integrlt tr (blend)
cselekv szemly: llam a) gyilkos b) orvos
cselekvs: a) gyilkol b) gygyt
eszkz: a) a moratrium meghosszabbtsa
b) szocilis brlaksok
irnyuls: a problma megoldsa (betegsg)
a cselekvs kvetkezmnye: a jelzlogpiac halla
cl: a szocilis hl fenntartsa
Input3
orvos
gygyts
gygyszer
betegsg
Szksgessg tr
Generikus tr
cselekv szemly
cselekvs
eszkz
irnyuls
Input4
llam
ltrehozs
brlaksok
problma
4.2. A fenti hlzatot a kvetkez mentlis terek alkotjk: A Lehetsg mentlis trbe
gyazdik Input 1, a Gyilkos mentlis tere s Input 2, az llam mentlis tere. Az inputok kzs struktrja a generikus trben jelenik meg, s ezek szerint a gyilkos s az
llam cselekv szemlyknt vannak jelen.
Tovbbi mentlis tr a Szksgessg tere, ahol Input 3, az orvos mentlis tere s Input 4, a szocilis brlaksokat rendelkezsre bocst llam mentlis tere tallhat.
Ezen inputok kzs struktrja is a generikus trben van.
Kzpen helyezkedik el a blend, az integrlt tr, ahov a mentlis terek bizonyos
elemei kivettdnek.
4.3. A fogalmi integrcinak tbb fajtja van. Esetnkben pldul Input 1 s Input 2
kztt egy gynevezett egyoldal hlzat krvonalazdik. Vagyis az integrlt tr csak
egy input struktrjt rkli. A hlzat ezen fajtja a fogalmi metafora, ahol Input 1 a
forrstartomny s Input 2 a cltartomny, s az integrlt trben csak a forrstartomny struktrja jelenik meg. Ezek szerint az llam gyilkos, illetve mivel a Lehetsg
trben vagyunk, az llam gyilkos lehet.
Input 3 s Input 4 szintn egy fogalmi metaforn alapul hlzatot szemlltet, ahol
az llam orvos, illetve a Szksgessg mentlis terben az llamnak orvosnak kell/ene
lennie.
4.4. Ezek az egyoldal hlzatok azonban esetnkben egy tbboldal hlzat alkotelemei, s ebben a hlzatban mr egyidejleg tbb input elemei kerlnek az elegyts,
111
az integrci terbe. Ahogy mr emltettk az integrlt tr magban foglalja a generikus tr struktrjt is. gy a generikus tr olyan elemei, mint cselekv szemly, cselekvs, eszkz s a cselekvs irnyulsa az integrlt trben is jelen vannak.
Az elegyts terben azonban egy specifikus struktra is ltrejn (emergent
structure). Ez az j struktra azltal keletkezik, hogy az inputok elemei kztti kapcsolatok egy sajtos, az eddigiektl eltr sszefggsben szervezdnek. Mindez egy
olyan j jelents kialakulst idzi el, amely egybknt az inputokban nem lehetsges
(composition).
gy pldul az integrlt trbe a cselekvs irnyulsa kizrlag Input 3-bl s Input 4bl vettdik ki, azaz a cselekvs irnyulsa az integrlt trben a problma megoldsa,
vagyis a betegsg orvoslsa.
Az Input 1-ben s Input 2-ben feltrt irnyuls az integrlt trben mr mint egy alternatv cselekvs kvetkezmnye tnik el: a jelzlogpiac halla a kilakoltatsi moratrium meghosszabbtsra vezethet vissza.
j elemknt bukkan fel a cl, tudniillik a szocilis hl fenntartsa.
4.5. Az elemek tszervezdse mdostja az azok kztt fennll logikai sszefggseket is, ami szksgszeren egy j, megvltozott rtelmezst eredmnyez. Az j jelents kialakulst a megrtsi folyamat rsztveviknt httrismereteink is befolysoljk.
Mg ugyanis az inputokban egy-egy szemly jelenik meg, addig az integrlt trben
rtelmezsnk szerint egyrtelmen egy szemly, az llam van jelen. Az inputokban
ugyanis nincs kifejezve, hogy a cselekv szemlyek egyazon cselekvt testestenek
meg. Csak annyit rzkelnk, hogy az egyik szemly meghosszabbtja a kilakoltatsi
moratriumot, ezltal gyilkoss vlik, a msik szocilis brlaksokat hoz ltre, s ezltal orvosol. Az inputokban ezek a szemlyek klnbz cselekvseket hajtanak vgre
ms-ms kvetkezmnnyel. Az integrlt trben ezek a cselekvsek s kvetkezmnyeik sszehasonlthatv vlnak, s ily mdon mentlis tapasztalataink s httrismereteink alapjn tovbb bvtjk, kiegsztjk a jelentst (completion). Az integrlt trben a
cselekv szemly vlaszt eltt ll, el kell dntenie, hogy l vagy gygyt-e. Ennek a
szemlynek az elegyts terben Janus-arca van, egyidejleg hordozza magban a
gyilkossg s a gygyts lehetsgt.
4.6. A teljesen j jelentst teht az eredmnyezi, hogy az integrlt elemek alapjn a fogalmi hlzat cselekv szemlye valjban egy orvos, aki gygytani szeretne, de a nem
megfelel gygymd a beteg hallt okozhatja. Mindazonltal tudatban vagyunk
annak, hogy ennek a szitucinak a fszereplje maga az llam.
Ez az j jelents mindenekeltt azon mentlis kpessgeinkre pl, amelyek az input elemei kztti kapcsolatokat folyamatosan fenntartjk s beazonosthatv teszik.
Ez a folyamat (elaboration) mentlis s fizikai tapasztalataink alapjn mkdik. Ezeknek a kpessgeknek nem vagyunk tudatban, az egyetlen dolog, ami a tudatunkba jut,
a megrts pillanata, the flash of comprehension (FauconnierTurner 2002: 44).
112
5. sszegzs
A mentlis tr egy olyan rszleges fogalmi struktra, melyet minden esetben egy adott
kontextus, egy konkrt kommunikcis helyzet idz el. A mentlis terek kztt a
kommunikci sorn megfelelsek gynevezett lekpezsek alakulnak ki. A megrtst nagymrtkben befolysolja, hogy mennyire vagyunk kpesek nyomon kvetni
az aktivlt mentlis tereket, illetve a kzttk kialakult lekpezseket. Ugyanaz a kijelents eltr mentlis tereket is aktivlhat. A jelents teht nem elre meghatrozott,
hanem az aktivlt mentlis terek megfelelseitl fggen formldik. A mentlis terek
bizonyos elemei integrldhatnak is egymssal, melynek kvetkezmnye az elemek
kapcsolatainak egy specifikus struktrja. Ez az j struktra egy j jelents alapjt
kpezi. Az j jelents teljes kialakulst a kommunikciban rsztvevk httrismeretei s mentlis tapasztalatai teszik lehetv.
Forrsok
Stiglitz, J. E. 2010. Egyelre tl nagyok a deficitcskkents veszlyei. Vilggazdasg. 42/47: 16.
Napi 2010. Kirly szerint bntets a hazai bankad mrtke. Napi Gazdasg. XX/195: 2.
Irodalom
Fauconnier, G. Turner, M. 2002. The Way We Think. New York: Basic Books.
Fillmore, Ch. J. 1982. Frame semantics. In The Linguistic Society of Korea (ed.): Linguistics in the
Morning Calm. Soeul: Hanshin. 111137.
Kvecses Zoltn 2005. A metafora. Gyakorlati bevezets a kognitv metaforaelmletbe. Budapest:
Typotex Kiad.
Kvecses Zoltn Benczes Rka 2010. Kognitv nyelvszet. Budapest: Akadmiai Kiad.
Kiefer Ferenc 2007. Jelentselmlet. Budapest: Corvina Kiad.
113
1. Bevezets
Vizsglatomban arra szeretnk vlaszt kapni, hogy az regeds mely lexiklis elemekhez val hozzfrshez van nagyobb hatssal, a kznevekhez vagy a tulajdonnevekhez. Ezen kvl az is rdekel, hogy a httrzaj mennyire befolysolja a lexiklis elrst.
A krdseim megvlaszolst Budapesten l fiatalok s idsek csoportjnak a
vizsglata szolglja. Az egyik ksrleti csoportot 2240 v krli fiskolsok kpezik,
a msikat pedig 67 v feletti, valaha rtelmisgi munkakrt betlt nk s frfiak. A
vizsglat hrom rszbl tevdtt ssze: (1) Az adatkzlknek ismeretlen szemlyeket
mutattam be kpeken. Ezek a kpek, amelyeken a szemlynevek is rajta voltak (a)
rendes krlmnyek s (b) httrzaj vltakozsval kerltek bemutatsra. (2) Ezt kvette a lexiklis elrsi feladat, amelyben a vizsglati alanyoknak az elzleg bemutatott szemlyeket kellett megneveznik a kpek lttn. (3) Ezutn a vizsglati szemlyeknek kpek alapjn trgyakat s kzismert szemlyeket kellett megneveznik (a)
httrzaj ksretvel s (b) httrzaj nlkli krlmnyek kztt.
Hipotzisek: Az idsek esetben a nyelvemen van jelensg (tip-of-thetongue/TOT) gyakrabban fordul el, mint a fiataloknl; tovbb ez a tulajdonnevek
elrst gyakrabban befolysolja, mint a kznevekt. A httrzaj szintn az idsek
lexiklis elrst rinti negatvabban a fiatalokhoz kpest. A httrzaj viszont nem
rinti kedveztlenebb mdon a bizonyos tpus lexiklis elemek elrst: gy pldul a
tulajdonnevek elrst nem htrltatja jobban a httrzaj, mint a kznevekt.
2. Irodalmi ttekints
Az albbiakban arra vllalkozom, hogy nhny irodalmi hivatkozs tkrben, nagy
vonalakban sszefoglaljam, hogy valjban hol is trtnik a lexikai elemek trolsa, s
ennek a trolsnak, valamint hozzfrsnek nhny jellegzetessgt is kiemeljem.
Ngy angol nyelven vgzett vizsglat rvid bemutatsra is vllalkozom, amelyek az
n vizsglatomhoz hasonlan szintn a kznevekhez s/vagy tulajdonnevekhez val
lexiklis hozzfrst kutattk.
2.1. A mentlis lexikon
Az elsajttott/megtanult s az emlkezetnkben rgztett szavakat a lexiklis vagy
deklaratv memriban troljuk (v. Navracsics 2007). Ezen bell a szavak egy felttelezett mentlis rendszerben, azaz mentlis lexikonban reprezentldnak, amelybl
114
mg a nem modlis s pontatlan, nem rszletezett (unspecified) hivatkozsok elfordulsi szma megntt az idsek esetben. Ezen kvl az idsek tbb nvmst s kevesebb
fnevet hasznltak, mint a fiatalok. Ezek az eredmnyek is azt igazoltk, hogy a fiatalokhoz kpest az idsek tbb sz-tallsi problmval talltk szemben magukat (HellerDobbs 1993).
Jennifer K. Taylor s Deborah M. Burke (2002) arrl szmoltak be, hogy az idseket a szmegnevezses feladatban az elhangzott, szemantikailag sszefgg, figyelemelterel szavak jobban gtoltk a ltott trgy helyes megnevezsben, mint a fiatalokat. gy pldul a vizsglati alanyoknak megmutattak egy labdt, amelynek angol
nyelv megfelelje homofon, mert a ball bl jelentssel is br. Az adatkzlk a kp
megtekintsvel egy idben a bl sz egyik szinonimjt: prom dikbl szt hallottk, amely szemantikailag sszefggsben ll a ball szval. Ez az elhangzott nyelvi
elem negatvabban befolysolta az idseknl a helyes kpmegnevezst, mint pldul a
vele ssze nem fgg frisbee frizbi, vagy hammer kalapcs, vagy a vele fonolgiai
viszonyban ll, hasonl kiejts call kilts sz (TaylorBurke 2002).
3. Mdszer
3.1. Adatkzlk
Mg az egyik ksrleti csoportot 2240 ves fiskolsokbl, a msik csoport 6798 v
kztti felnttekbl llt. Az 1. tblzatban lthat az ids s a fiatal adatkzlknek a
kdja s az letkora. Az (F) azt jelzi, hogy az adott adatkzlk frfiak, akik kzl
mindkt korcsoportban csak egy-egy tallhat, a tbbi tizennyolc alany pedig n.
Idsek
Adatkzlk
Fiskolsok
letkor
Adatkzlk
letkor
A1
98
B1
37
A2
67
B2
29
A3
82
B3
28
A4
80
B4
40
A5 (F)
68
B5
40
A6
82
B6
33
A7
89
B7 (F)
37
116
A8
97
B8
22
A9
80
B9
27
A10
75
B10
37
3.2. Eszkzk
A vizsglat els rszben az adatkzlknek ht ismeretlen szemlyt mutattam be a
szemlyneveikkel (Kati, Istvn, Jnos, Pter, Zsuzsa, Judit, Jnos) egytt. A kpek
bemutatsa alatt csend, httrzaj, majd pedig jbl csend kvette egymst, gy httrzaj csak a hrom kzps szemly (Jnos, Pter, Zsuzsa) bemutatst ksrte.
A vizsglat msodik rszben az adatkzlknek negyvent egyenknt bemutatott
kpet kellett megneveznik. Kzlk huszonhrom trgyat, huszonkett pedig egy
kzismert szemlyt (rt, sznszt, politikust stb.) brzolt. Mint ahogy a pldkbl is
kitnik, a trgyak fogalom al soroltam ismert tereket, pleteket s szemlyekre utal
kzneveket is, mert felttelezsem szerint ezek megnevezse egyttesen knnyebb a
kzismert szemlyek nevnek a lehvsnl.
Trgyak:
fs, piros rzsa, tank, trombita, Hsk tere, villanykrte, hint, tulipn, dobostortaszelet, huszrok, szultn, kalapcs, kukoricacs, fagylalt, Akropolisz, Szent Istvnbazilika, nagydob, Lnchd, grgdinnye, sparhelt, papucs, szemveg, piramis.
Kzismert szemlyek:
Ady Endre, Beethoven, Mtys kirly, Shakespeare, Trcsik Mari, Nagy Imre, Szeleczki Zita, Chaplin, Mricz Zsigmond, Szchenyi Istvn, Kos Jnos, Mikszth Klmn, Petfi Sndor, Latinovits Zoltn, Jkai Mr, Jzsef Attila, grf Teleki Pl, Gobbi
Hilda, Erzsbet kirlyn, Bajor Gizi, Kardi Katalin, Ferenc Jzsef.
A trgyakat s a kzismert szemlyeket kevert sorrendben mutattam be a vizsglati
alanyoknak, gy, hogy egy trgy kpt egy szemly kpe kvette s gy tovbb. A
vizsglat vgzse kzben a fiskolai tanteremben, illetve az ids adatkzlk otthonban csak az adatgyjt s a vizsglati alany voltak jelen. A kpeket csend s httrzaj
vltakozsval mutattam be, s mindkt esetben a kpen lthatkat rgtn meg kellett nevezni. A csend s httrzaj kpekre lebontva egyenletesen vltotta egymst
mindegyik alany esetben. A httrzajt rdimsorok hallgatsval oldottam meg.
117
4. Eredmnyek
4.1. Ismeretlen szemlyek neveinek az elhvsa
A felidzsi feladatban ngyen (A4, B1, B6, B7) tkletesen teljestettek.
Hrom adatkzl az utols vagy a kzps neveket nem tudtk felidzni a htbl:
B3 az utols hrom nevet,
B4 a legutolst,
B5 az 5. s 6. nevet nem tudta sikeresen elhvni az emlkezetbl.
Tovbbi ngy fiskols hallgatnl fordultak el gondok az ismeretlen szemlyek
s neveik trstst illeten:
B2: Csak az els kt nevet tudta a megfelel szemlyekkel trstani.
B8: Olyan nevet is emltett, amivel a szemlyek bemutatsakor nem tallkozott (Jnos helyett Jzsefet mondott).
B9: Kt szemlynevet az elejn felcserlt.
B10: A harmadik szemlynevet nem tudta felidzni (a httrzaj kezdete) s Zsuzst
Katival helyettestette (httrzaj vge).
A lexiklis hozzfrssel kapcsolatos deficitek nem csak azon lexiklis elemek esetn fordultak el, amikor a kpek bemutatsa httrzaj ksretben trtnt, hanem akkor is, amikor csendes krlmnyek kztt tallkoztak a kpekkel.
A 2. tblzatbl jl kitnik, hogy a bemutats, s a kzvetlen utna kvetkez felidzs feladatait ksr nhny percben az idsek ktszer annyi ismeretlen szemlynek
a nevre nem emlkeztek, mint a fiatalok. A kt csoportnl a nvtvesztsek (az arcok
s nevek helytelen asszocicija) kztti klnbsg viszont nem szignifikns.
Idsek
Fiskolsok
nvtveszts
nvhiny
nvtveszts
nvhiny
14
15
11
118
mg?) A tnyron? t van dfve a tnyron egy ecset. (Adatgyjt: Torta, az egy
tortaszelet.) Naht, ez rmes. Ugyan, ht ez egy tortaszelet? A XVIII. szzadba
val.
A 89 ves hlgy egy tbbgombcos fagylalt lttn a kvetkezt jegyezte meg:
A7: Ht ez rzsa akar lenni pamutbl (Adatgyjt: vagy pedig) kosrban fagylalt?
4.2.2. Tulajdonnevek/kzismert szemlynevek
Akiket az sszes adatkzl felismert, nem msok, mint Mtys kirly s Petfi Sndor.
Akiket az sszes fiskols felismert, az Charlie Chaplin s Kardi Katalin. A tzbl
kilenc fiskols felismerte Kos Jnost. Erzsbet kirlynt/Sissyt az sszes ids adatkzl s nyolc fiskols ismerte fel s tudta megnevezni. A msik vglet Szeleczki
Zita s grf Teleki Pl, akiket egyetlen fiskols sem ismert fel.
4.2.3. A nyelvemen van jelensg
Ha az adatkzlk memrijukbl azonnal nem tudtk lehvni az ltaluk ismert szemly
nevt, elmondtk rla a vele kapcsolatos ismereteiket, vagy kifejtettk, hogy tudjk,
kirl van sz, de nem jut eszkbe a neve. Legtbb esetben egy kis gondolkodsi id
vagy a vezetknv kezdbetjnek a megadsa elgsgesnek bizonyult a tulajdonnv
felidzshez.
A kvetkezkben az adatkzlk pldin keresztl mutatom be a vizsglatban elfordul nyelvemen van jelensget. A bemutatott szemlyek kzl hat megnevezsekor fordult el ilyen jelensg. Nyolc pldval szolglnak az idsek s hrommal a
fiatalok:
Chaplin:
A6: Naht, meg a Stan s a Pan-bl? Nem. (Adatgyjt: Nem, de hasonl.
Nmafilmekben szerepelt.) Igen, igen (Adatgyjt: cs betvel kezddik Cs Cs
Chap) Chaplin.
A7: Na kicsoda? Tudom, hogy ki. Mondjad csak a nevt. Tudom, tudom (Adatgyjt: Cs Cs Cse) Chaplin, igazi Chaplin a nagy cipivel.
Kos Jnos:
A2: A az Antal Imrvel egytt szokott hlyskedni. Annak a prja. Az egereket
nekeltk. Ht most ez nem, ez nem ugrik be. Antal Imre, Kos Jnos, na. Csakhogy.
B3: Ismerem. Ko, milyen Kos J nem. (Adatgyjt: De, Kos Jnos).
Trcsik Mari:
119
A4: Jaj, ht ez a Mari sznsznnk. A milyen Mari? ppen most emlegettem. lltlag itt van egy lak, aki szegrl-vgrl rokona. Na szval a pttysben debtlt, pttys szoknyjban. Na milyen Mari? Jaj. (Adatgyjt: T betvel kezddik a vezetkneve. T.) Trcsik.
B5: Sznszn. (Adatgyjt: Ismers?) Abszolt tudom, hogy ki az. Krhintban jtszott s nem jut eszembe a neve. Mr reg vagyok (40 ves). (Adatgyjt:
Tr) Trcsik Mari.
Shakespeare:
A1: Jaj-jaj, haj-jaj, Shakespeare.
A3: Na most ez, ez nem tudom, ez Shakespeare.
A8: Jaj, ht ezt hogy hvjk, ez angol, na. Nehezen jutnak eszembe a nevek. Nem
Shakespeare. (Adatgyjt: De. az.)
rdekessg: a fiskolsok kzl csak ketten (B7, B8) tudtk Shakespeare-t megnevezni.
Latinovits Zoltn:
A1: Ht ez a mi nagy-nagy sznsznk. Ez senki nem mlta fell ennek a mvszett. ngyilkos lett a Balaton-parton ez is. Hrman vannak testvrek. A zensz
az ccse, annak se ugrik be a neve, a btyja Bujtor Istvn, meg a Rutkainak a
frje volt. Nem ugrik be a neve. Latinovits Zoltn.
B2: Nem jut eszembe a neve, de tudom. (Adatgyjt: Kicsoda?) Sznsz, de nem
jut eszembe a neve (Adatgyjt: Latinovits) Latinovits Zoli.
Kardi Katalin:
A8: Na a Katalin. (Adatgyjt: Milyen Katalin?) Aki nekel a Kardi Katalin.
Kardi Katalin nevt az sszes fiskols egybl el tudta hvni a mentlis lexikonbl,
viszont az idsek kztt akadt egy, az A8-as adatkzl, akinek az nekesn vezetkneve egy kis reakciid elteltvel jutott csak eszbe.
A httrzaj s a csend egyenl arnyban oszlott meg azon kzismert szemlynevek
felidzse kzben, amelyeket a nyelvemen van jelensg jellemzett. Mg Chaplin,
Kos Jnos s Kardi Katalin bemutatsa httrzaj mellett trtnt, Trcsik Mari,
Shakespeare s Latinovits Zoltn kpeinek megtekintse csendes krlmnyek kztt
valsult meg. Az a tny, hogy ennek ellenre a nyelvemen van jelensg nagyobb
mrtkben jelentkezett az idseknl, nem zrja ki annak a lehetsgt sem, hogy httrzaj nagyobb mrtkben s negatvabban jtszott kzre az idsek, mint a fiatalok
tulajdonnevekhez val hozzfrsben.
5. Kvetkeztetsek
Az idseknl a nyelvemen van jelensg gyakrabban fordul el, mint a fiataloknl. A vizsglatban a nyelvemen van jelensg a tulajdonnevek, azaz kzismert szemlynevek lexiklis hozzfrsnl jelentkezett, s nem a kznevek aktivlsakor. Az
120
121
1. Bevezets
A jelen tanulmny clja, hogy bemutassa egy haznkban mg kevss ismert terpis
eljrs mdszereit s fejlesztsi lehetsgeit a nyelvi s beszdzavarok kezelsnek
alkalmazsban. A lovasterpia hatsmechanizmusnak ismertetse utn t gyermek
nyelvi s beszdfejldsnek eredmnyeit kzljk, akik 0,51 ves idtartam, heti
egyszeri lovasterpis fejlesztsen vettek rszt.
2. A lovasterpia helye a rehabilitciban
A lovasterpia a l, a lovagls s a lval val valamennyi foglalkozs hatsait felhasznl komplex eljrs, amelynek mozgsfejleszt, gygypedaggiai s pszicholgiai
hozadkai vannak (Bozori 2002). Hrom alkalmazsi terlett klnthetjk el: a
hippoterpia neurolgiai alapokon nyugv gygytornakezelst jelent a l segtsgvel
(Gyrgypl 2006), a gygypedaggiai lovagls s lovastorna/voltizsls az rtelmileg
akadlyozott, tanulsban akadlyozott gyermekeknek nyjt kiegszt fejlesztst, mg
a l-asszisztlt pszichoterpia a llek betegsgeit gygytja a l mint koterapeuta segtsgvel. Magyarorszgon a gygytornsz, gygypedaggus, pszicholgus vagy
pszichoterapeuta diploma megszerzse utn vgezhet el a 2 ves lovasterapeutakpzs, a Magyar Lovasterpia Szvetsg ltal koordinlt akkreditlt tanrtovbbkpzs formjban, a nmetorszgi DRA-vizsgarendszerrel egyttmkdve. Azok a kliensek, akik valamilyen kommunikcis, nyelvi s/vagy beszdzavarral kzdenek, tbbnyire a gygypedaggus szakgon vezetett foglalkozsokba integrldnak.
3. A lovasterpia hatsmechanizmusa
A lovasterpia fentiekben ismertetett alkalmazsi terletei nem vlaszthatk el lesen
egymstl. A kliensek problmi gyakran halmozottan jelentkeznek, gy mindig a
vezet tnet (srls, betegsg) dnti el, melyik szakemberhez kerljenek. A
lovasterpia azonban egy olyan egyedlll terpis eljrs, amely egyszerre kpes
hatni a pciens teljes ideg- s izomrendszerre, a mentlis s pszichs faktorokra
ezrt szoks komplex szenzoros/szenzomotoros integrcis trningnek is nevezni.
Munknk sorn alapveten a l lpsjrmdjt hasznljuk terpis clokra. A lps
egy ngytem, ritmikus mozgs. A lbak diagonlis sorrendben, egyenletes idkzkkel kvetik egymst. A hts lbak ltal indtott mozgs kvetkeztben a l gerincben rotci jn ltre, ez a lovas medencjt jobbra-balra billenti, amihez egy elrehtra trtn mozgs is addik. Ahogy a l emeli s leteszi a lbt, egy fel-le hullm122
mozgs jn ltre, ami szintn ttevdik a lovasra (Katona 2006). Ezt a komplex mozgsos stimulcit a lovon l pciensnek elszr a medencje veszi t, az impulzusok
innen haladnak tovbb a gerincoszlopon felfel, eljutva a nyakszirtig, s ingerlik a
megfelel agykrgi kzpontokat is. Ilyen sszetett hatst semmilyen ms mozgsfejleszt eljrssal nem lehet elrni. ppen ezrt figyelembe kell vennnk azt is, hogy a
kliensek ideg- s izomrendszert igen intenzv ingerls ri (a lpsben jr l percenknt kb. 90110-et lp), emiatt a lovasterpis foglalkozsokat ltalban heti egy alkalommal, 2030 perces idtartamban vgezzk.
A lp lovon trtn terpia teht egyszerre, egy idben kpes hatni a l htn l
kliens teljes ideg- s izomrendszerre. A petyhdt izmokat ersti, mg a spazmusokat
laztja. Ahogy a testi grcsk olddnak, azzal prhuzamosan olddnak a pciensben a
lelki feszltsgek, szorongsok is. A l htn jralhetk, azaz bekapcsolhatk az
idegrendszer esetlegesen kiesett fejldsi fzisai is. A lovasterpit szoks jrstipikus
trzstrningnek is nevezni: elsegti a helyes testtarts kialakulst, fejleszti az egyenslyt. A lovagls jtkony lettani hatsai a terpis munkban is rvnyesek: lnkti
a keringst s az anyagcsert, harmonizlja a lgzst (ami a beszdkpzs fontos elfelttele), ersti az nrtkelst, fejleszti az nbizalmat, segtsgvel a pciens rzelmi s kommunikcis szempontbl is nyitottabb vlik a vilgra.
4. A nyelvhasznlat s a lovon trtn mozgsfejleszts sszefggsei
A nyelvi fejlesztsnek elssorban kzs cselekvsen, nem kizrlag kzs beszden
rdemes alapulnia (Frhlich 1989). Mindez a lovasterpiban egyedlll mdon
megvalsthat.
A mozgs jelenlte kzponti szerepet jtszik a nyelvi fejldsben s a ksbbi aktv
nyelvhasznlat sorn, hiszen a beszd ltrejtthez finommotoros tevkenysgek bonyolult kombincijt hasznljuk (Hildebrand-Nilshon 1989). gy is mondhatjuk,
hogy a beszd fejldse s kivitelezse a mozgs fejldstl s ksbbi alkalmazsnak, alkalmazhatsgnak minsgtl fgg. A motoros zavar ugyanis a beszdet ugyangy krostja, mint minden ms mozgst (Haupt 1966). A beszd a mozgs szemszgbl nzve a lgzmozgsok, a ggef-, nyvitorla-, llkapocs-, nyelv- s ajakmozgsok eredmnye. A spasztikus, athetotikus s ataxis mozgszavarok ezeket a mozgsokat is befolysolhatjk. Egy spasztikus izomtnus kliensnek valsznleg nehezebb
lesz a szjmozgsok beindtsa is; a tlmozgsok sokaknl elgtelen szjzrssal, fokozott nylfolyssal jrnak egytt, ami szintn megnehezti a beszdet; s akinek a nagymozgsai koordinlatlanok, annl a finommozgsok, gy a beszdtevkenysgek is
koordinlatlann vlnak.
Tapasztalataink alapjn a lovasterpis foglalkozsokra rkez nyelvi s/vagy beszdzavarral kzd pciensek mindegyiknl elfordul valamilyen fok mozgsneheztettsg, -zavar vagy srls is. Az egyik legfontosabb terpis elvnk ennek szellemben a mozgsfejlesztsen keresztl trtn rehabilitci. Tmaszkodunk arra a
tnyre, miszerint a l ltal kzvettett komplex mozgsstimulci az egsz idegrendszert befolysolja, mely ltal jtkonyan tmogathat az agyi/idegrendszeri plaszticits. Szemlletnk fontos alapjt kpezi tovbb a rendszerszemllet gondolkods. A
123
lovasterpiban nem a nyelvi s beszdzavart kezeljk, hanem a teljes embert, a biopszicho-szocilis lnyt.
5. Hipotzis
A lovasterpis foglalkozsok megkezdsekor feltteleztk, hogy a mozgs jelenlte
kzponti szerepet jtszik a nyelvi s beszdfejlesztsben. A lovasterpia elsegti a
helyes testtarts kialakulst, ami a beszd ltrehozshoz elengedhetetlen lgzs harmonizlsnak elfelttele. gy gondoltuk, a nagymozgsok fejlesztse a finommozgsok, gy a beszdmozgsok kivitelezsre is jtkony hatst gyakorol. Kiemelt szerepet tulajdontottunk annak a tnynek, miszerint a lovasterpis foglalkozsok a szabad levegn, termszetes krlmnyek kztt, oldott lgkrben, llat-asszisztlt formban valsulnak meg, s e pszichs faktorok mellett knnyebben megnylnak a kliensek, tovbb maga a spontn hangads is knnyebben sikerl, ha a pciens figyelmt
nem a fonci kti le. Elkpzelhet, hogy a srlt agyterletek kiesett funkcii is helyrellthatk ezzel a komplex mozgsstimulcis eljrssal.
6. Anyag, mdszer, ksrleti szemlyek
A tovbbiakban t olyan gyermek lovasterpis fejlesztsi eredmnyeirl szmolunk
be, akiknek vezet vagy jrulkos problmja (volt) a kommunikci, a nyelv- s a
beszdhasznlat zavara.
Sz lesz kt kislnyrl, akikben kzs, hogy slyos figyelemzavarral kzdenek. A figyelemzavarhoz trsul hiperaktv tnetek mg tovbb zavarjk a kommunikcijukat.
Egyikk 2005-s szlets, vods kislny, ADHD-diagnzissal, mely az esetben
valsznstheten genetikus eredet. Trsul tnetei: megksett idegrendszeri fejlds, mozgs- s egyenslyzavar. A msik kislny 2002-ben szletett, vezet diagnzisa a rszkpessgzavar s a tanulsi akadlyozottsg. Problmi htterben tbbszri
gyermekkori agysrls, tmeneti afzis llapotok llnak. Jrulkos problmja gyszintn a mozgs- s egyenslyzavar.
A harmadik pciens egy 2007-es szlets, gyermekkori autizmus-diagnzist kapott
kisfi. Vezet problmja megksett beszdfejldse. Mozgsfejldsbl kimaradt a
kszs/mszs idszaka. Repetitv mozgsok (repked kzmozdulatok) jelentkezse
megfigyelhet. Orvosi diagnzisban ezen kvl petyhdt izomzat megjellse szerepel. Hasonl problmval, br az autizmusnak egy msik vltozatval, n. atpusos
autizmussal diagnosztizltak egy 2006-os szlets kisfit. Az anamnzisben
perinatlis srls szerepel. A kszs/mszs szintn kimaradt a mozgsfejldsbl.
Megksett beszdfejlds, m mg a msik kisfi egyltaln nem, mr beszl. Beszdprodukcija inadekvt, echollia jellemzi.
Az tdik kisfi 2004-ben szletett. Jelents mrtk szorongssal rkezett a
lovasterpis foglalkozsokra. A kommunikcit ilyenformn gyakran megtagadta.
Keveset s nagyon halkan beszlt, a krdsekre tbbnyire szk, egyszavas vlaszokat
adott.
Mind az t gyermekkel kb. 1 vig, tbb-kevesebb megszaktssal, de ltalban heti
egy alkalommal tallkoztunk a lovasterpiban. Valamennyi foglalkozsra fejlesztsi
124
125
elrt eredmny magrt beszl: a lovasterpia segtsgvel nem csak az oroszln btorsgt sikerlt megtallnunk.
8. sszegzs
Tapasztalataink szerint a lovasterpis foglalkozsok szmtalan lehetsget rejtenek a
nyelvi s beszdfejlesztsre. Ez a klnleges, komplex mozgsstimulci harmonizlja
az idegrendszert, oldja az izmok merevsgt, ugyanakkor megerst, btort, nbizalmat
ad. Az a pszichs kzeg, ahol a terapeuta s a kliens tallkozsa megvalsul, sszehasonlthatatlan az egyb terpis terletekkel. A fejleszts olyan specilis kntsbe
bjtatott, amely a felsznen egyltaln nem a betegsgtnetekre fkuszl. A l mint
koterapeuta szerepe ugyancsak felbecslhetetlen rtk ebben a folyamatban. az,
akinek a gyerekek mg knnyebben megnylnak, mint a terapeutnak. A l a maga
elfogad trs-mivoltval annyi szeretetet kpes adni, amellyel a nyelvi akadlyok is
legyzhetk.
Irodalom
Bozori Gabriella 2002. Lovasterpia. Gondolatok s vzlatok a gygypedaggiai lovagls s
lovastorna tmakrbl. Szkesfehrvr: Polu-Press Kkt.
Frhlich, A. D. (szerk.) 1989. Kommunikation und Sprache krperhinderter Kinder. Dortmund:
Verlag Modernes Lernen.
Gyrgypl Zoltnn (szerk.) 2006. Hippoterpia. Balogunyom: Unicornis Egszsgforrs Alaptvny.
Haupt, U. 1966. Die sprachlichen Schwierigkeiten zerebral gelhmter Kinder. Schriften zur
Sonderpdagogik.
Hildebrand-Nilshon, M. 1989. Sprachentwicklung des krperbehinderten Kinders. In Frhlich, A. D.
(Hrsg.): Kommunikation und Sprache krperhinderter Kinder. Dortmund: Verlag Modernes
Lernen. 5780.
Katona Enik 2006. A l mozgsnak elemzse, hatsa a pciensre. In Gyrgypl Zoltnn (szerk.):
Hippoterpia. Balogunyom: Unicornis Egszsgforrs Alaptvny. 2835.
126
1. Bevezets
A kommunikcin belli hallgats szerepvel mint az rt hallgats krdskrvel
2010-ben kezdtem el foglalkozni. Egy gyakorlati tantrgy (Kommunikcis kszsgfejleszts 1.) oktatsa sorn figyeltem fel arra, hogy azok a hallgatk sem mindig rtik a
trsaik egybknt zmmel j produkciit, akik ltszlag figyelnek, s rendre maguk is
jl teljestenek. A kutatsom teht konkrt tapasztalatokbl indult ki, s ler/feldert
jelleg volt. Clja az volt, hogy megvizsgljam a hallgats szerept, s esetleg krvonalazzam annak funkciit a kommunikciban. E kutats eredmnyei egy szk mintra
vettve (41 f) egyrtelmv tettk, hogy a hallgats leginkbb informcis hallgats,
valamint hogy az aktv hallgats fontos kommunikcis stratgia. Az eredmnyek arra is
utaltak, hogy a hallgats a beszdhelyzet vagy a kommunikcis szntr fggvnyben
funkcikba sorolhat.
2011-ben kt szervezet munkatrsainak krben vizsgltam meg egy 74 fs mintn
a hallgats szerept, funkciit s azt, hogy honnan tudjuk, mely helyzetekben illik,
vagy kell hallgatni. E kutats ismertetse eltt, a vizsglt tma szakirodalmi begyazsaknt clszernek tnik elszr is a szerepviselkeds s a kommunikci viszonyt
a teljessg ignye nlkl taglalni, msodszor pedig a nyelvi s nem nyelvi kommunikci funkciit vzolni.
2. A szerepviselkeds s a kommunikci
Az ember gyerekkortl szocializldik a szerepekre (ld. a role-play-t s a role-gameet; Mead 1973). A gyerek elbb konkrt beszdhelyzetek szereplinek (pl. orvosbeteg) a brbe bjik, s eljtssza az egyik vagy a msik kommunikcis partner szerept. Itt mr a szerepek szerinti viselkedsek elklntse, azaz rgztse is elkezddik, vagyis a gyerek 35 ves korban elkezdi megalkotni a viselkedsek ltalnostott mintjt. Majd ksbb a felntt szerepviselkeds elzmnyt, a szablyjtkot
(role-game) tanulja meg. Itt csak gy tud azonosulni a szerepvel, ha elsajttja a jtkszablyokat, megrti a tbbi szerepet betlt szemly viselkedst, valamint azzal
is tisztban van, hogy az viselkedst is rtik a tbbiek.
A szerepekre val szocializlds vgigksri az letnket. Minden szerephez tartoznak megrkltt gondolkodsi-cselekvsi modellek, vagyis a vilgot az
egyn/gens individulisan, de mintk segtsgvel ragadja meg. A viselkedsi (nyelvi
s nem nyelvi) mintkat a normk/szablyok rjk el. Az egynt a normk betartsra
a szankcik elkerlse sztnzi. letvilgaiban, azaz azokon a szntereken, ahol meg-
127
fordul, ppen a normk azonossga teszi lehetv a szerepek normk szerinti megvalstsnak elvrst egymstl.
Minl tbb szerepet tlt be az ember, annl jobban betagolt a trsadalomba. A szerepkomplexum azon szerepek sszessgt jelenti, amelyeket az egyn ltal betlttt
sttusok rnak r. A szerepkomplexum zsugorodsa a trsadalombl val kihulls,
illetve visszahzds jele.
Habermas (1981) szerint a szerepelmlet nemcsak a szociolgiai rendszert kpes
magyarzni, hanem az letvilgot is, amelyben az egyn szempontjbl relevns,
ltalban sajtos s egyedi dolgok kommunikatv folyamatokban lebonyoldnak. E
kommunikatv folyamatok sorn az egynek a beszdhelyzettl fggen iparkodnak a szerepeiket az elvrsoknak megfelelen megvalstani. Azonban sajt maguk
megrtetse s msok megrtse nem mindig sikerl, ami tbbek kztt arra vezethet
vissza, hogy:
tudshalmazaik nem mindig vannak kell tfedsben,
tves a msikrl kialaktott kpk, mivel az egyik ember a msik kpnek kialaktsakor sajt magbl indul ki.
Teht attl fggetlenl, hogy az ember az elsdleges s a msodlagos szocializcija
sorn megtanulja, hol, mikor, mit s hogyan kell mondania, vagy hol, mikor, mit s
hogyan kell meghallgatnia/elhallgatnia, a konkrt beszdhelyzetekben mind a mondst, mind pedig a hallgatst rdemes tanulni, illetve tantani.
A tovbbiakban a hallgats ltalam felttelezett funkciit ismertetem a verblis s a
nonverblis kommunikci funkciit alapul vve.
3. A verblis s nonverblis kommunikci s azon bell a hallgats funkcii
A nyelvi kommunikci funkciit annak cljaibl s feladataibl kiindulva lehet feltrkpezni s megrteni. A nyelvi kommunikci funkciit Bhler (1934), Jakobson
(1972) s Rosengren (2008) nyomn mutatom be.
Bhler organon-modellje (1934) szerint minden nyelvi jel egy hrmas relciban
mkdik: egyik szemlymsik szemlydolog. Bhler kifejezsi, brzolsi s felhvsi funkcikrl beszlt. A kifejezsi funkci a jel beszdaktust vgzhz val viszonyt, az brzol a jel s a tnylls viszonyt, a felhvsi pedig a hang s a cmzett
viszonyt fejezi ki. Vagyis Bhler e modell segtsgvel rja le azt az szrevtelt,
hogy a nyelvi jel nemcsak trgyak s tnyllsok brzolsra szolgl, hanem egyttal
hrt ad a jelet felad llapotrl, s befolysolja a jelads cmzettjnek magatartst is.
Jakobson (1972) jra trgyalja a nyelvi funkcikat, kibvti Bhler (potikai,
konatv, emotv) modelljt: a referencilis, a fatikus s a metanyelvi funkcival. gy
vli, hogy a nyelv kpes valamennyi funkcit betlteni, de egy-egy megnyilatkozsban
bizonyos funkcik tlslyra jutnak. A referencilis funkci megnyilvnulsa esetn a
nyelvi kzlsben a kontextus, a metanyelvi esetn a kzls kdja, a konatv esetn a
cmzett, az emotv esetn a felad, a fatikus esetn a kontaktmdium, a potikai esetn
pedig a kzls kifejezsi mdja ll a kzppontban.
Rosengren (2008) a kommunikci funkciinak ngy ltalnos fajtjt klnbzteti
meg, spedig az informatv, a kontroll, a trsas s az expresszv funkcit.
128
129
130
132
Mint lthat, mindegyik vlasz esetben dominl a negatv megtls, azaz a megfigyeli szerepben lv megkrdezettek szerint jobb, ha az ember egy barti trsasgban
kiveszi a rszt a beszlgetsbl.
Wismer (1984: 117) azt mondja, hogy: Egy zenet tartalombl s rzsekbl ll.
Gyakran az egyiket vagy a msikat nem rtjk meg, mert elfeledkeznk arrl, hogy a
jelentseket az emberek hordozzk. Tovbb szl arrl is, hogy a befogadnak aktvan felelssget kell vllalnia a kzlt tartalom s a kzlt rzelmek megrtsrt.
Emltend mg az is, hogy a kommunikciban trekedni kell az egyenlsgre.
DeVito (1984: 112) szerint:
Az egyenlsgnek a beszd s a hallgats viszonya szempontjbl is jellemeznie kell
az interperszonlis kommunikcit. [] Ha az egyik rsztvev egsz id alatt beszl,
a msik meg hallgat, akkor a hatkony interperszonlis kommunikci nehzz, majdhogynem lehetetlenn vlik. Ksrletet kell tenni a kommunikcis hozzjruls kiegyenltsre.
Vagyis nem szabad elfeledkezni arrl, hogy egy zenetnek van direkt s indirekt (tartalmi s viszonyulsi) szintje, s mindkettt meg kellene rteni. A megrtshez szksg van minden kommunikcis szntren a klcsns elfogads kzlsre. Aki hallgat, illetve nem aktvan hallgat, az a beszlhz s/vagy a tartalomhoz val viszonyt
hallgatja el (palstolja az rzelmeit, nem rti a tartalmat, nem rdekli a tartalom stb.)
Egybknt a megkrdezettek kzl azt, ha valaki hallgat a barti trsasgban (14.
krds) 38 f tli negatvnak, 6 f pedig pozitvnak. 20 f szerint a hallgats lehet
pozitv s negatv is, azaz ez szitucifgg. (10 f esetben nincs adat!).
Lssunk erre egy-egy pldt!
A krdveket 1-tl 74-ig szmoztam, vagyis az idzett vlaszok utn zrjelben ezek a szmok lthatk.
133
5. sszegzs
Az eredmnyek egyrtelmen azt mutatjk, hogy a hallgats nagyrszt informcis
hallgats (Baczerowski 2000), azaz olyan, amely a nyelvi informci mdiuma. Ezen
bell a megkrdezettek szerint a jelentst hordoz hallgats funkcii kzl a kommunikcis stratgia funkci a dominns. Ezt az els felmrs eredmnyei is igazoltk.
Mindkt kutats alapjn a hallgats leginkbb a tudatosan alkalmazott kommunikcis stratgia eszkze, s a msikra val odafigyels, azaz az aktv hallgats (41,67,
illetve 46,5%-ban (1. s 2. krdv)) nagyon fontos. Azt, hogy akkor hallgatunk, ha
nincs mondandnk, illetve, ha nem rtnk a tmhoz, a megkrdezettek 22,4, illetve
31,1%-a kzli (jelz funkci).
A hallgats megtlse a (barti) trsasgban fejezzen az ki brmilyen funkcit is
az esetek tbb mint 50%-ban, mindkt felmrs alapjn negatv. Megjegyzend,
hogy a vlaszadk jelents rsze a hallgats pozitv vagy negatv minstst a szitu-
134
citl teszi fggv. (Az els felmrsben a rsztvevk tbb mint 45%-a, a msodikban pedig tbb mint 25%-a vlekedik gy!)
Az albbi diagram kt krdv kt-kt krdsre adott vlaszai alapjn brzolja a
hallgats funkciinak megnyilvnulst.
135
Mellklet
Tisztelt Vlaszadk!
Felmrst ksztek arrl, hogy a hallgats milyen szerepet tlt be/tlthet be az emberek letben.
Krem, segtse a kutatst azzal, hogy kitlti a krdvet. A krdv kitltse termszetesen nkntes,
s nv nlkl trtnik.
Ksznm a kutats segtst!
Krdv a hallgatsrl:
1) letkor:
2) Idegen nyelv ismerete (legalbb kzpfok):
3) Fontos-e, hogy bizonyos szitucikban ne beszljen, hanem csendben maradjon? Ha igen, melyek ezek a szitucik? (Tbb vlasz is adhat!)
Ha ms beszl, akkor aktvan hallgatok.
Ha nem vagyok biztos abban, hogy helytll, amit mondank.
Ha a trgyalsi pozcit rontanm a gyors vlasszal.
Ha az adott helyen, szituciban nem illik beszlni.
Ha nincs mit mondanom.
Egyb, spedig:
4) Milyen hallgatsra vonatkoz szlsokat, kzmondsokat ismer? rja le!
5) Gyermekkorban figyelmeztettk-e, hogy bizonyos helyzetekben maradjon csendben?
Igen, ha (Csak egy, mgpedig a legjellemzbb vlaszt karikzhatja be!):
a felnttek beszlnek
nem rtem, amirl sz van
valakit knos helyzetbe hoznk
Mit jelent az, hogy megtrni a csendet s nem tudja sz nlkl megllni?
6) A csend jelentheti azt, hogy br nem mondunk semmit, ezzel mgis kifejeznk valamit. n
szerint mit lehet leginkbb kifejezni azzal, ha valaki hallgat? (Krem, vonatkoztassa sajt magra, vagyis egy, az nre legjellemzbb vlaszt karikzza be!) Azt, hogy:
136
1. nyugalomra vgyok
2. j egytt hallgatni
3. valaminek a kimondsa bizonyos fajta cselekvsre vagy cselekvssorra ktelezne
4. aktvan, teljes odafordulssal hallgatom a beszlt
5. nem szeretnk msoknak fjdalmat okozni
7) Honnan tudja, hogy mely helyzetekben kell, illetve jobb hallgatnia? Emltsen legalbb kt pldt.
1.
2.
8) Tapasztalta-e, hogy valaki hallgatssal az albbiakat fejezi ki? (Csak egy vlasz adhat, mgpedig az, amivel a leggyakrabban tallkozott!)
szgyenkezst
haragot
valaminek az eltitkolst
msok flrevezetst
9) n szerint mit kzvett a hallgats a beszlgets sorn? (Csak az n szerint legjellemzbb vlasz
karikzhat be!)
koncentrlok arra, amit a partner mond
nem akarok a szavba vgni
nem rtem, gy nem tudok hozzszlni
nem rdekel, amirl sz van
10) n szerint milyen asszocicikat breszt a tanrban, ha egy dik az rn a feltett krds utn
hallgat?
1. Nem kszlt fel az rra, nem tudja a vlaszt.
2. Mssal foglalkozott rn, nem tudja, hogy mit krdezett a tanr.
3. Tudja a vlaszt, de gondolkodik, hogyan kezdje el.
4. Egyb, spedig:
11) Tanult-e n gyermekkorban valamilyen illemszablyt arra vonatkozan, hogy a trsadalom
mely sznterein s mely szitucikban tancsos vagy ktelez inkbb hallgatni, mint megszlalni? Emltsen kt pldt!
12) Milyen rzelmeket kpes kzvetteni a hallgats? Soroljon fel legalbb hrmat!
13) Pozitv vagy negatv megtls al esik, ha valaki hallgat egy barti trsasgban? Indokolja
vlaszt!
137
14) Vannak a magyar kultrban tabutmk, vagyis tmk, amelyekrl nem beszlnk? Soroljon fel
kettt!
1.
2.
138
trkon bell is nagy az eltrs. Ahhoz, hogy meggyzdjnk, hogy mennyire mskpp kezelik a jelensget a klnbz kultrkban, nem felttlenl olyan gykeresen
eltr kultrkat kell sszehasonltani, mint a japnt s az olaszt, elg sszevetni a
japn kultrt a knai vagy a koreai kultrval (Lebra 1987). ltalban a keleti kultrkban a tl sokat beszl emberek gyansak s nem tartjk igazn tisztelettudnak
ket (Snchez et al. 2008).
Mivel a csendet meglehetsen nehz definilni, ebbl addan, mint jelensget, tanulmnyozni is legalbb ilyen bonyolult. Ms nyelvekkel ellenttben sok eurpai
nyelvben a sz tbb jelentssel br, pl. az angol silence, a spanyol silencio, a francia
silence. Hasonltsunk ssze a magyar csend, hallgats szavakkal, vagy orosz ,
, , ahol a inkbb a fizikai jelensget jelli, a
emberi, a kt elbbi szintzist (ld. 2000). A fenti pldkbl
kikvetkeztethet, hogy maga a nyelv is tudja akadlyozni a fogalom definilst,
akrmennyire is paradoxonnak tnik. Pldul az angol nyelv nem tesz klnbsget, de
mg a magyar is csak kt jelentst klnbztet meg, mikzben az orosz hrmat. Az
Oxford English Dictionary (1989) szerint olyan szinonimkkal lehet meghatrozni a
silence fogalmt, mint muteness, reticence, taciturnity, illetve complete quietness
or stillness; noiselessness. Egyes kutatk azt a megkzeltst javasoljk kvetni,
hogy a csend elssorban kommunikcis vlaszadsi kudarc.
Klnbsget kell tenni a hallgats s a sznet kztt. Tannen (1985: 109) felteszi
azt a krdst, hogy mikor tekinthet egy sznet hallgatsnak. A vlasza: amikor a
csend hosszabb, mint a vrhat sznet, vagy a sznet vratlan helyen keletkezik s
nem a szoksos funkcival br, hanem utal valaminek a hinyra (ibid). McLaughlin s
Cody (1982) azt a vlemnyt kpviselik, hogy a hrom msodpercnl hosszabb csend
mr nem pauznak tekintend, hanem hallgatsnak. Nhny frissebb kutats azt mutatja, hogy hrom msodpercnl hosszabb sznet nem igazn fordul el a termszetes
spontn beszdben. A sznet egybknt a beszd termszetes rsze s szintn kultrafgg.
Azokkal a kutatsokkal, amelyekben a csendet s a hallgatst kommunikcis hinynak fogjk fel, nem tudok egyetrteni. Inkbb Jaworski (1993) nzeteit osztom, aki
szerint a csend s a hallgats a kommunikci metaforja.
A csend jelensgnek tbb dimenzija van:
fizikai,
trsadalmi,
kulturlis,
eszttikai.
Nzznk ezzel kapcsolatban nhny pldt. A politikai letben pldul a csend lehet
az ellenllsnak a jele, a hallgats pedig a korrumplt rtegek viselkedsre utalhat. A
mvszetben a csend sokszor egyenl az idillel.
Az sszes kommunikcis eszkz kzl taln a csend s a hallgats a leginkbb
kontextusfgg. Jaworski (1993), hivatkozva Jensen rsra (1973), a csend s a hallgats t klnbz funkcijt sorolja fel. rdekessgk, hogy ezek a funkcik egyszerre pozitv s negatv mdon jelenhetnek meg a kommunikciban:
sszekt funkci: a csend s a hallgats sszektheti a kt embert, de el is tvolthatja;
140
141
143
144
Building on two premises, namely that there are at least twenty-one words and
phrases for the twenty-first century (Luntz 2007) and that as soon as they leave
the pages of a dictionary and go into action, words dont have meaning, they have
uses (Szllsy 2007), my paper aims to prove the hypothesis that words commuting
between cultures via the highly digitalized media yield new patterns of thought. With
the globalized media we are currently facing a situation similar to international jointresearch projects where, while experimenting with new technologies, specific new
knowledge is tightly linked to the linguistic competence of the participants. Usually
project partners from different countries choose a vehicular language common to all of
them, but the choice of a common language might predetermine certain ways of reasoning and different forms of expression.
In this light my paper endeavours to provide a short multilingual comparative analysis to prove that in the process of linguistic borrowing the nature of a particular language influences the habitual thoughts of its new speakers, who thus develop a multiple vision of the world.
1. Media exposure
In the first decade of the 21st century people live in a media saturated society, where
communication via electronic networks constitutes a new reality. Research carried out
in 2005 (Potter 2008: 5) revealed that almost 70% of an average persons day includes
some form of media use (30% media as a single activity, 39% media linked to some
other activity). Television and computers provide the bulk of visual and audience time,
a reality that is fully exploited by the media outlets. The pervasiveness of IT is visible
across all the layers and classes of society. Owing to cheap and handy information
technology most human communication is mediatized and interactivity is also guaranteed, as by their nature, the current media allow various opportunities for public participation in the form of talk shows, political debate programmes or digital means of
instantaneous commentaries. As a result of the continuous levelling of the cultural
differences the media are continuously altering peoples identities. Most contemporary
societies have a multitude of parameters, which lead to people developing multifaceted
identities. But also the new technologies may transcend the limitations of transnational
communication. As Mihelj points out: given the participatory potential and accessibility, online forums, discussion lists and platforms such as You Tube have the ability to bridge national divides (Mihelj 2011:183). The young generation has presumably developed virtual identities as well, so now reality can be approached by postulating that students have already shaped their world view by borrowing elements
among which a rich vocabulary and a large variety of discourse patterns from several
cultures at a time. Meanwhile it is important to continue to provide a sociological ac145
count of language, as this can actually bear relevance for understanding language in
the media.
2. Media and Education
The above mentioned concerns definitely call for conscious and consistent reform in
education. In order to be successful in transmitting valuable knowledge we all have to
turn green in our professional activity and recycle old content. This issue can be best
illustrated by borrowing a quotation from the Core Curriculum of Harvard University
(Luntz 2007: 184), which clearly states that in undergraduate education an important
paradigmatic shift has been experienced lately, moving away from the old or canonized curriculum, which manifests itself in the form of current knowledge in certain
fields and a tremendous amount of information digested from the Great Books, to a
new one, aiming at introducing students to the major approaches to knowledge in domains the faculty considers indispensable. It goes without saying that the philosophy
of all colleges and universities is to form broadly educated graduates also trained in a
particular academic field. Consequently, at undergraduate level nowadays intellectual
breadth includes primarily intellectual skills and habits of critical thinking and judging
assumptions more objectively (Hortobgyi 2010).
In higher education, Hungary has also joined the Bologna system and the faculty
has to cater for the needs of young people in their late teens and early twenties who
see life through contemporary lenses. First of all, students would rather cope with
manageable amounts of knowledge. Their natural curiosity has to be stimulated by
easily accessible information. We must agree that the best way to do this is to design
problem solving tasks relying on elements of popular culture. In a multicultural setting, when engaging in a conversation, one has to presuppose that the members of a
speech group usually share the same code and an entire system of symbols, signs and
meanings. This new educational trend can legitimate why in the current communication background colleges and universities are tailoring their offers not only to needs
but also to the habits of the students. For instance an interesting aspect related to the
growing importance of media is the spreading of media-related subjects in the so
called serious domains of study, namely academic fields.
By now it has become more than evident that the media have to be both investigated
and fully exploited in order to reveal the challenges they encompass when dealing
with societal issues. The omnipresence of the media in our highly digitalized world
displays a wide range of concerns that shape both the present and the future media
landscape and the issues these can create for communication. One such concern relates
to the changes occurring in verbal communication and can be spotted in the widerange of media related topics such as Social Networking, Global Infotainment or Consumer Culture.
As such the title of my paper 3Ws intends to be in line with the growing trend of
using acronyms in all walks of verbal communication, a trend which could be considered an obvious outcome of the need to economize on spatial and temporal dimensions
of message transmission (short messages, emails, chats, etc,). In my understanding the
146
terms 3Ws could easily yield different sources of the acronym, such as: Words Working Well or Words Worth Writing, even Words Worth Wondering (about).
3. Rationale and Research
In my paper I argue that the words used in a media rich environment exert a high impact on the processes of communication in general and language acquisition in particular. The techniques of persuasion used in most media representations stimulate the continuous development and restructuring of the mental lexicon, even in the case of young
adults.
Several decades ago sociologists and sociolinguists had been concerned mainly
with how social differences in spoken language influence meaning, and focused their
investigations on the connotative aspects of language. At the end of the 20th century
Pierre Bourdieu could only account for changes in language if there is also a change
in the social structure (Myles 2010: 35).
Let us illustrate briefly the rationale of a small-scale research project. Undergraduate and graduate students enrolled in two different seminars in U.S. and Anglo-Saxon
Media had been assigned the task to exploit their personal media habits and diets, and
evaluate the impact the media products they are consuming exert on the development
of their vocabulary. Both the Hungarian and the American students enrolled in our
MA programme admitted that most of their active vocabulary stems from both receptive and productive online activity (reading news and blogs, watching movies and
serials, stumbling over commercials vs. writing emails, blogs, comments in chatrooms
and on the social networks).
At the outset of the research students had to read a chapter from Dr. Frank Luntzs
book Words that Work. Its not what you say, its what people hear (2007). The chapter entitled Twenty-one Words and Phrases for the Twenty-first Century makes an
inventory of words that cut to the heart of Americas most fundamental beliefs and
rights, to the core values that do not change no matter how we vote or shop, or what
delivery devices we use to play music, in the year 2020 (Luntz 2007: 239). While
reading Luntzs book on the power and importance of words, they were asked to set up
a list of the most eloquent Hungarian words they encounter in their daily media
activities. In addition they had to find examples that proved the survival of the 21
English words Luntz had predicted for the 21st century U.S. media. Moreover they had
to find current Hungarian contexts in which some of these international words are
being employed.
By now journalism has developed from a news gathering to a news generating profession (Myles 2010: 111). So why are the media so important when tackling issues
related to the development of ones lexis and communicative skills? Remaining within
the scope of this research, we have to add that most of these students also attend lectures and seminars in translation and interpreting where the activation of the mental
lexicon, the retrieval of the best word association or collocation is imperative. Therefore we should not ignore the daily stimuli that reach the students. For example the
students had to translate headlines from a local newspaper. One of the entries read in
147
Hungarian as follows: Szelektv hulladkgyjts (which literally translates into English as Selective Garbage Collection). Some students provided a quick solution: Going
Green. This choice is also a great example of how codes are often revealed in imagery.
When asked to justify their choice, the students admitted that advertisements on CNN
(multimedia ads with powerful visual and audio effects where the text can be both
seen and heard embedded in a pleasant environment) had a more tangible effect than
their relatively poor experience of texts related to environmental issues. Contrary to
what one might have expected, the interpretation of another headline A tavasz els
hrnkei (Springs First Messengers) visibly caused problems in retrieving the proper
term or collocation. The students said that by now the word messenger would more
often relate to instant communication, chat, dialogue than to forerunner.
The small scale research involved 18 students at the University of Pannonia who attended a seminar in U.S. Media. They had been handed Luntzs list of the most powerful words used in the media dating from 2007, and exploiting their own media habits
and diet, they had to set up a similar list of the most determining Hungarian words.
U. S. Media (Frank Luntz 2007)
Imagine
Hassle-free
Extra- (Extra)
Life-style
Sensational (Szenzcis)
Accountability
Results/can-do-spirit
Free (Ingyen)
Violence (Erszak)
Online (Online)
Efficient/efficiency
Career (Karrier)
Wellness/Fitness
10
Financial security
Crisis (Vlsg)
11
Patient-centered
Terrorism (Terrorizmus)
12
Investment
Environmental commitment/awareness
(Krnyezettudatossg)
13
Casual elegance
Top (Top)
14
Independent/independence
15
All-American
Stars/Celebrities (Sztrok/Celebek)
148
16
Peace of mind
LOrel
17
Certified
18
Prosperity
Multicultural (Multikulturlis)
19
Spirituality
Minorities (Kisebbsgek)
20
Balanced approach
21
A culture of
The above list contains the most pertinent Hungarian words found by the students in
advertisement, commercials, news articles in the spring of 2011. Following this, the
students had to find words from Luntzs list that not only survived in English, but also
migrated into Hungarian.
Let us tackle first the 2007 list. On the one hand, the words have proved their continuing presence, as the students found no difficulty in tracing them in their favourite
TV programmes. In what follows I am highlighting a few contexts in which some
elements of the 2007 list are present.
imagine
We spend our whole lives worrying about the future. Planning for the future. Trying to predict the future. As if figuring it out will somehow cushion the blow. But the
future is always changing. The future is the home of our deepest fears... and our wildest hopes. But one thing is certain, when it finally reveals itself, the future is never the
way we imagined it. Greys Anatomy
hassle-free
New hassle-free drinks package options from Royal Caribbean International
Each new package offers guests a no-hassle way to enjoy an all-inclusive experience across all onboard drinking venues. Royal Caribbean International has developed these new packages to enable our guests to make their holiday more relaxing and
enjoyable by paying one set price for a wide variety of drink choices. Commercial
results/can-do-spirit
U.S. Sen. Michael Bennet visits Scott Elementary School in Greeley U.S. Sen. Michael Bennet, ex-superintendent of Denver Public Schools, is visiting schools across
Colorado that consistently show growth under the federal No Child Left Behind Act to
find out what the schools are doing to get results. He chose Scott because it has many
free and reduced lunch recipients but continues to improve in test scores. greeleytribune.com
investment
As surgeons, were always invested in our patients. But when your patients a
child, youre not just invested, youre responsible. Responsible for whether or not that
149
child survives, has a future, and thats enough to terrify anyoneThey say the bigger
your investment the bigger your return. But you have to be willing to take a chance.
You have to understand you might lose it all. But if you take that chance, if you invest
wisely, the payoff just might surprise you. Greys Anatomy.
Students have also found Hungarian contexts where elements of the 21 English words
(or their translations) can be found. In Hungary the can-do-spirit is present more often
in commercials (Lipton tea, Vichy), life-style in entertainment (Mom Park Centre in
Budapest).
As for accountability, in Hungary the concept implies efficiency and investment
and relates more often to finances and the world of banking. For Americans the term
has a wider implication as Americans need and want accountability within their everyday lives. This goes from the elected government officials to the products that are
bought and used. Not only do they expect these aspects to be accountable, but they try
to hold themselves accountable as well. For example, if a person is holding a company
accountable for providing a service, then he/she is accountable for paying for this
service. When thinking about accountability, one way to show that one is accountable
is to show how you can be independent.
This is how Jenna Skiba, a U.S. student enrolled in our joint Master programme in
American Studies, describes the power of the word imagine: This word has been used
since my childhood and now stands for much more in my everyday life. Growing up,
you are asked to look to the future and imagine what you want life to look like. Imagination is a personal meaning and will be different for every person. This individual
meaning of the word imagine is an invitation to look into the future and at how one
might want to live, which leads directly into the world lifestyle. The new global
trends have coined a new term for people working in the media, imagineers (a blending of imagine + engineers), who employ persuasion techniques targeting the public at
large. Some examples taken from the French media also highlight the symbolic power
of the concept. The first sentence is a commercial of the French Rails offering promotional tickets for youth below age 26, a travel card they call Imagine R: La RATP, le
STIF et la Rgion Ile-de-France nhsitent pas faire la promotion de leur carte Imagine R, le titre de transport rserv aux Franciliens de moins de 26 ans.
The Romanian media utilizes the concept in most diverse contexts:
Noile reclame iPhone: lipsa de imaginatie (The new iPhone: lack of imagination)
Reclama outdoor, publicitate stradala cu imaginatie interesanta (Outdoor commercials, billboards with interesting imagination)
Imaginatia face toti banii (Imagination makes all the money)
For the American people investment has been a major ideal throughout the 20th and
21st centuries. Investing shows that one is planning for their future, and therefore investment is part of their lifestyle. Imagining the future and preparing for it can be done
with investment, not just in liquid assets, but also in abstract items as well. Going to
college and getting an education is an investment in yourself, allowing you far greater
opportunities in the future. Having this investment may make one eligible for certain
certifications, which will continue to make their life the way they might have imagined
150
it. Not only through the globalized economy but especially through the products of the
media these concepts get naturalized in host cultures as well. For instance more and
more often do we find these terms explicitly used in the above contexts not only in
dubbed or subtitled movies, but also in Hungarian-made programmes.
Considering that the media-flooded world we live in basically follows an AngloSaxon globalized communication pattern, I have inferred that some of these words
would also be massively present in various other languages, such as French, Hungarian or Romanian.
For instance, if we investigate the concept of investment in French and Romanian,
we can see similar tendencies. Usually the French verb investir uses two particles
(Investir pour/dans lavenir/to invest for/in the future). Nevertheless an online article
dating from October 2008 introduces a site with the following headline: Un site pour
investir sur le futur prsident amricain (A site investing upon the future American
President (Auberger 2008).
The same word shows similar usages in Romanian. A financial homepage invites
people to go on to higher education: Admiterea la facultate, investiie in viitor (Applying to higher education means investment in the future.) Similarly to French the Romanian language also employs different particles: Investiti in viitor/Invest in the future
or Investiti pentru viitor/Invest for the future (financiarul.com).
4. The media and the mental models
The media have an important role in creating social and cognitive distinctions rather
than reflecting them. As mentioned before, the students involved in the research are
also trained in translation and interpreting. In the process of interpretation word recognition is paramount. Professor Philip Hofmeister from Stanford University, in a lecture
in Introduction to Psycholinguistics held on April 7, 2011 discusses the issues of word
recognition in general and semantic priming in particular. According to him, in word
recognition the question is to what extent are frequency, priming and other effects
subject to automatic processes versus attentional processes. In semantic decoding
nowadays associative priming seems to be more frequent. In automatic processing
word recognition is fast, not preventable and not available to consciousness. In attentional processes word recognition is slow, strategic and available to consciousness.
Hofmeister also underlines that ambiguity provides information on the organization of
our lexicon and allows us to test how meanings are accessed when in competition.
Interestingly lexical decision is faster for words with multiple senses. Perhaps surprisingly, when we see or hear a word, we activate all its possible meanings, but context
will constrain the interpretation (stanford.edu).
Currently in language research, focus has been shifted from rules to words. Elman
(2011) claims that language research results suggest that the lexicon is representationally rich, that it is the source of much productive behaviour, and that lexically specific
information plays a critical and early role in the interpretation of grammatical structure; therefore, he suggests an alternative approach, which provides a way to account
for lexical knowledge without a mental lexicon.
151
There are several trendy metaphors on how the media is viewed these days. We definitely live in a media saturated world where our culture is a huge supermarket of media messages. If in the past the main issue was how to gain information at all, nowadays the major problem is how to protect ourselves from the flood of information. In
order to protect ourselves quite often we put our mind onto automatic pilot, which
might not be the best decision, as the danger in the automatic processing of messages
allows media to condition our thought processes. As we have mentioned before, the
daily media diet of the students explores the lexis found at the strategic intersection
between business, politics, infotainment and Hollywood. When developing students
media literacy skills we need to consider that the strategic and tactical use of specific
words and phrases can change how people think or behave. We already have experience related to how in a multinational joint project the working language (be it English, French, German or any other language) influences the narratives and communication strategies people are using while negotiating, drafting documents, disseminating
the end-products of a project. Therefore we consider it of paramount importance to
develop the young generations conscious awareness to decoding the language of the
media. In this highly digitalized world we are experiencing a phoneticization of messages. Therefore we also have to train our students to develop skills to hear and decode
not only text speak but also the new voices present in the media, as voices point to
the possibility of a more readerly and reflexive appropriation of the language of the
media by lay actors (Myles 2010: 51).
5. Conclusion
Language creates and reflects the world; nevertheless we all dissect the world along
different lines. Nowadays these differences can also stem from the fact that language
is often under the dominance of the media, as a powerful institution. Actually language is often dominated by elite discourses that may generate alternative views and
particular conceptualizations of the world. In addition to the re-conceptualization of
already existing entities, new unknown, unfamiliar or abstract events are difficult to
define. Whenever a new concept is created, linguistic strategies operate to create a
mental model which is stored as a new element in our mental lexicon. The media,
which encompass discourses and narratives situated at the intersection of virtually all
walks of life ranging from private to public sphere, from lay to professional, are
strongly influencing the creation of the mental model, as all depends on the discourse
that is created to embed the concept.
Above we have seen the role of the media in building a relevant lexis used by most
post-modern discourses, almost irrespective of the languages which employ them. It is
true that lexicalization is highly conditioned by the particular interests that influence
personal word selection, but quite often discourses select lexical items to express ideologies rather than meaning in itself. The scope of the next phase of the research on the
most representative words dominating the media will focus on creating the conceptualization maps of the analysed database.
152
References
Auberger, O. 2008. Un site pour investir sur le futur prsident amricain. Le Figaro.
http://www.lefigaro.fr/patrimoine/2008/10/13/05001-20081013ARTFIG00475-un-site-pourinvestir-sur-le-futur-president-americain-.php
Elman, J. L. 2011. Lexical knowledge without a lexicon? The Mental lexicon. 6/1: 133.
http://www.ingentaconnect.com/content/jbp/ml/2011/00000006/00000001/art00001
Hofmeister, Ph. 2011. Intro to Psycholinguistics. http://www.stanford.edu/~philiph/teaching/2011
/ling170/lecture2-words.pdf
Hortobgyi, I. 2010. Aspects of Glocalized Popular Culture in Hungary. Bulletin of the University of
Transylvania. Vol. 3(52). Series IV. Philology. Cultural Studies. Brasov: Transylvania University
Press. 217228.
Luntz, F. Dr. 2007. Words that Work. Its not what you say, its what people hear. New York: Hyperion.
Mihelj, S. 2011. Media Nations. Communicating Belonging and Exclusion in the Modern World.
London: PalgraveMacmillan.
Myles, J. F. 2010. Bourdieu, Language and the Media. London: PalgraveMacmillan.
Potter, J. W. 2008. Media Literacy. Sage Publications.
Szllsy, J. 2007. Hunglish into English. The Elements of Translation from Hungarian into English.
Corvina.
http://www.greeleytribune.com/article/20110323/NEWS/703239985/1002&parentprofile=1001129
http://www.financiarul.com/articol_46863/admiterea-la-facultate-investitie-in-viitor.html
153
II. BESZDKUTATS
SPEECH RESEARCH
154
155
1. Bevezets
A spontn beszd relatve folyamatos; az egymst kvet nyelvi jeleket azonban esetenknt azok szksges, akaratlagos s szndktalan kimaradsa jellemzi. Szksges a
jelkimarads, ha annak fiziolgiai oka van, ilyen pldul az oxigncsere, azaz a llegzetvtel. Akaratlagos a nyelvi jel hinya, ha a beszl tudatosan sznetet tart vagy
pldul nevet. Szmos tnyez vezet a nyelvi jel szndktalan kimaradshoz pldul a gondolatok sszerendezshez vagy a beszdtervezsi diszharmnia feloldshoz
szksges nma vagy kitlttt sznetek megjelense. A jelkimarads egyetlen fonma
realizldsnak hinyban is megjelenhet (Gyarmathy 2008, Vradi 2012), de kombinldhatnak is, s nha meglehetsen nehz, hacsak nem lehetetlen annak meghatrozsa, hogy mi induklta azokat. A llegzetvtelt (is) szolgl nma sznet felhasznlhat nyelvi tervezsre, a krkogs vagy khgs leplezheti a diszharmnia tnyt,
avagy az nmonitorozs felszni kvetkezmnyeknt ltrejtt jelkimarads felhasznlhat lgzsre, avagy kitlthet shajtssal.
A jelkimarads egy sajtos esete, amikor a beszl tudatosan vagy nem tudatosan,
de meglltja az addig folyamatos artikulcit, az agy lellst parancsol a beszdszerveknek. Nooteboom nevhez fzdik az alapvet megakads (a beszdfolyam megszakadsa) szablya (Main Interruption Rule, 1980), ami klnsen akkor feltn, ha
az artikulcis lells a sz ejtse kzben trtnik meg. Nooteboom szablya kimondja, hogy ha a beszl a rejtett vagy a felszni nellenrzs sorn szreveszi, hogy hibzott, akkor azonnal megll (megszaktja a kiejtst). Ezt Levelt (1983) azzal egszti ki,
hogy hozzadja az n. megfelelsgi javts szablyt. Ez akkor kvetkezik be,
amikor a kzls nem tartalmaz ugyan hibt, de az elhangzottak nem felelnek meg
egszen a beszl szndknak.
Az artikulci hirtelen megszakadsa, lellsa tbbflekppen osztlyozhat jelensg, attl fggen, hogy mi a folytats. Ha a megkezdett artikulcis gesztus vagy
gesztusok semmilyen mdon nem folytatdnak, akkor tves kezdsrl beszlnk, fggetlenl a kivlt okoktl, pldul (a. a jelkimaradst szemllteti): a ka. a helikoptert
nem mutattk. A tves kezds egy msik formja ll el akkor, amikor a megkezdett
artikulcis gesztus vagy gesztusok jl indulnak, de a tervezsben valahol hiba kvetkezik be, s ez megllshoz vezet, pldul: nem volt olyan, hogy ne kp. kttt
volna bele valakibe. (A pldk a Beszdkutats 2011-es ktetnek Nyelvbotlskorpuszbl valk.) Az artikulci lellsa gy is bekvetkezhet, hogy a kiejtett beszdhang vagy beszdhangsorozat ltszlag megfelelt az eredeti szndknak, a beszl
mgis elkezdte a sz ejtst az elejtl, ez az jraindts, pldul: hallo. hallottam a
156
ni. Hipotzisnk szerint az artikulcis lells olyan felszni jelensg, amely a folytatstl fggen eltr monitorozsi mkdseket felttelez. gy gondoljuk, hogy az
jraindtsok s a sznet a szban jelensgek kontextuselemzse, valamint a szerkesztsi szakaszok temporlis jellemzi vlaszt adnak az azonos vagy eltr monitorozs
krdsre.
2. Ksrleti szemlyek, anyag, mdszer
A BEA magyar spontnbeszd-adatbzis (Gsy 2008) 12 beszljnek (6 n, 6 frfi;
letkoruk 25 s 35 v kztti) narratvjt vlasztottuk vletlenszeren a
megakadsjelensgek adatolshoz. A mintegy 250 perces anyagban sszesen 58 sznet a szban jelensget s 43 olyan jraindtst talltunk, amely a clunknak megfelelt.
A jelen kutatsbl kihagytuk azokat az jraindtsokat, amelyeknek a szerkesztsi
szakasza 0 ms volt, azaz az jraindtst nem elzte meg semmilyen sznet. Erre azrt
volt szksg, mert csak a sznetet tartalmaz jraindtsokat tudtuk sszehasonltani a
sznet a szban jelensggel. (Az adatkzlk kztt jelents eltrsek voltak az elfordulsokban.) A Praat akusztikai-fonetikai programmal (4.2-es verzi, Boersma
Weenink 2005) mrtk meg a szerkesztsi szakaszokat (a fonetikai szakirodalomban
szoksos kritriumok szerint). A statisztikai elemzseket az SPSS 19.0 verzijval
vgeztk el (egytnyezs ANOVA), 95%-os szignifikanciaszinten.
3. Eredmnyek
3.1. A sznetek anyagunkban a sz ngy klnbz helyn fordultak el, szsszetteli hatron, toldalk eltt, igekt (s a leg) utn, valamint a tsz belsejben (Gsy
2010). A jelensget legtbbszr, 44,8%-ban toldalk eltt tapasztaltuk, a tbbi elforduls csaknem egyenlen oszlott meg. Nem foglalkoztunk azokkal a jelensgekkel,
amikor az artikulci lell, s a sznetet kveten folytatdik ugyan a kiejtettek brmely rsznek megismtlse nlkl, de mgsem az eredeti ejts vltozatlan folytatsa
trtnik: az is j volt. de kicserlt. tettk azrt mert ht biztos ami biztos. Az ilyen
tpus megakadsjelensgek lnyegben a tves kezds s a sznet a szban jelensg
egyttes elfordulsai. Nem egyszeren tves kezdsek, hiszen az eredeti lexma ejtse ismtldik, de abban valamifle vltozs, a jelen pldban ragozsi mdosts trtnik, anlkl azonban, hogy a teljes sz kivitelezse megismtldne.
Elemzseink azt mutatjk, hogy a toldalkot megelz artikulcis lells rszben a
szksges toldalk elhvsval, rszben a szintaktikai szerkezet kialaktsval volt
kapcsolatos. Az elbbire plda: s gy a magyar. ral kapcsolatban azt ltom hogy. (A
toldalk eltti sznet idtartama 640 ms.) A toldalk elhvsnak nehzsgt itt valsznstheten a gondolatok kztti vlogats problmja elzte meg. Erre utal a jelkimarads relatve hossz idtartama is (1. bra). A szintaktikai szerkezet megformlsnak nehzsgt szemllteti a kvetkez plda (az a hezitlst jelzi): egy
tanr hirtelen megbetegedett s tartsan tppnz. re kerlt s megkerestek. A
tppnz sz ejtsekor a szintaktikai szerkezet tervezse mg nem volt kszen, a beszlnek idre volt szksge, hogy megformlja a folytatst, ami a kerl ige elhvsval
158
Az igekt (vagy a leg) utni s az sszetett sz els szavnak kiejtst kvet lells
rszben a lexiklis elhvs nehzsgre, rszben annak bizonytalansgra utal. Plda
arra, hogy az igekt kiejtst kveten problma addott a lexiklis hozzfrsben,
vagyis a sz elhvsban: s fel krnk egy ilyen kzvlemny-kutatsi ht nem
tudom cget. Egyrtelmnek ltszik, hogy a beszl tudta, melyik szra van szksge,
az igekt elhvst azonban mgsem azonnal kvette az ige megtallsa. A felszni
tny alapjn az mondhat, hogy a beszl nehezen jutott el a lemma szintrl a lexma
szintre. Nem zrhat ki azonban az sem, hogy a krnk ige ksleltetst a kzvlemny-kutatsi sz aktivlsnak nehzsge okozta, amire az ilyen tltelksz utal. A
kvetkez plda azt a helyzetet mutatja, amikor a beszl nem tallja a megfelel szt,
noha az igektt mr kiejtette (ennek az is oka lehet, hogy nem akart durva szt hasznlni, s egy elfogadhatbbat keresett): mr menjen mr minl kisebb kvetsi tvolsgot tartani s akkor persze hogy bele. durrantottak htulrl gyhogy n ezrt nem
tettem ki a T bett. Vgl olyan igt hv el durrant , amely br rthet, de stlusban eltr az addigi szvegkrnyezettl. Az sszetett szavak els tagjnak kimondst
kveten is elllhat a lexiklis hozzfrsi problma, amint azt a kvetkez pldban
lthatjuk: engem beadtak valami ilyen ilyen csecsem. otthonba amirl persze semmi
emlkem nincs. Az sszettel msodik tagjnak elhvsa tmeneti nehzsget okozott
a beszlnek, amit a knyszer sznet alatt megoldott. A sztben tapasztalt sznet
rendszerint a lexiklis elhvs problmjt, illetve a hatrozott nvel esetben gyakran az azt kvet sz aktivlsnak nehzsgt jelzi. Az s szerintem a prob. lmt az
okozta pldban a jelkimarads tnye ugyan egyrtelmen a lexiklis hozzfrssel
kapcsolatos bizonytalansg felszni megjelense, arra azonban nem ad vitathatatlan
159
160
magra sz els sztagjt kvet artikulcis lells nagy valsznsggel kapcsolatban van az anticipcival, illetleg azzal, hogy a beszl bizonytalann vlt a folytatsban a villa hangsor ejtst kveten.
2. bra. A ma. magra pldban a frfi beszl az els sztagot indtja jra
161
ul a grammatikai talakts tmeneti problmjra, s ekkor az jraindts mint megakads valjban az idnyersi stratgia rsze. A kvetkez plda htterben az egyeztetsi nehzsg problmja llhatott: nlunk is erre j maximum ez. ezek a telefonok. A
felszni jraindts jelzi a rejtett monitor mkdst, a nem normatv grammatikai
szerkezetre vonatkozan. A gondolatmegformls nehzsge idzhetett el jraindtst
a kvetkez pldban: az els vizsgmon n msodszorra mentem t a forgalomba d.
de radsul gy hogy []. A ktsz jraindtsval nyert id segtette a beszlt a
gondolat kivlasztsban, a kvnt megfogalmazsban. A mentlis lexikon hozzfrsi
folyamatban bekvetkez diszharmnia is tbbfle lehet. Az artikulci lellst a
kvetkez pldban az okozta, hogy a beszl nem a megfelel szt kezdte el kiejteni,
de a monitorozs eredmnyeknt mgis elfogadta az elkezdett szt, avagy nem tallt
megfelelbbet, teht jraindtotta a mr aktivltat: nem nem az ton parkolt hanem gy
fe. felesben a jrdn teht gy icipicit kilgott az tra. A felesben a jrdn kifejezssel a beszl maga is elgedetlen, ezrt magyarzza a jelentst.
3.3. A szban tartott sznet s a sznetet tartalmaz jraindts szmos kzs tulajdonsggal rendelkezik. A diszharmnia mindkt jelensg esetben elssorban a lexiklis hozzfrs sorn kvetkezik be, ugyanakkor ms beszdtervezsi szintek, illetleg
a gondolat kialaktsa is eredmnyezhet nyelvi talaktsi problmt. Az artikulci
lellsa mindkt esetben elmletileg a sz brmely pontjn bekvetkezhet. A sznet
mindkt esetben lehet nma, kitlttt vagy ezek kombincija. A szfaj tekintetben
nagy a klnbsg a ktfle jelensg kztt. Az jraindtsok nagyobb mrtkben (mintegy 65%-ban) a funkciszavakat, mg a sznet a szban jelensg a tartalmas szavakat
(kzel 80%-ban) rinti. Jelents az eltrs az artikulcis lells helyben is. A sznet
a szban jelensg anyagunkban legnagyobb mrtkben a toldalkot megelzen fordult el (44,8%). Az sszes tbbi helyzetben hasonl arnyokat tapasztaltunk; igekt
(s a leg) utn 17,4%-ban, szsszetteli hatron s a sz belsejben kzel 19%-ban
tapasztaltunk lellst. Ide soroltuk az els hangot kvet artikulcis megllst is, de
meg kell jegyeznnk, hogy ezek elfordulsa sokszor rintette a hatrozott nvelt. Az
jraindtsok 41,9%-ban az artikulcis lells az els hang kiejtst kveten bekvetkezik. Viszonylag gyakori az els sztagot kvet lells (34,9%), tbb sztag utn
23,2%-ban kvetkezik sznet. Mg az jraindtsok az els hang(ok) vagy az els
sztagot kveten llnak le mintegy 76%-ban, addig a szejts megszakadsa mintegy
27%-ban kvetkezik be az els hang(ok), illetve az els sztagot kveten.
Felttelezzk, hogy a (rejtett s felszni) monitor a szerkesztsi szakaszban ellenrzi
az talaktsi s/vagy a hozzfrsi folyamatokat visszamenleg, azaz jraellenrzs
megy vgbe. A monitor kontrolllja, hogy a kiejtett hangsor rsze-e az aktivlt
lexmnak, s ha nem tall hibt, akkor folytatdik az artikulci. gy jn ltre a sznet a szban jelensg. Ha a monitornak tbb idre van szksge, hogy dntsn az
elhangzott hangsorrszlet helyessgrl, illetve nem csupn az aktulisan kiejtett
hangsort ellenrzi, hanem a valamivel tgabb kontextust is, akkor trtnik jraindts
felsznen. Az artikulci lellsnak helye (a szban) szmos egyb tnyez fggvnye, mint ahogy a szerkesztsi szakaszok idtartama is.
162
ejt ki a beszl a szbl, annl gyorsabban fogja tudni korriglni az elbizonytalanodst, teht a szerkesztsi szakasz ekkor lesz a legrvidebb. Az adatok azonban nem ezt
tmasztjk al. Ennek az lehet a magyarzata, hogy a monitor mr sokkal korbban
hibt jelzett, s az ellenrzs prhuzamosan zajlott az artikulcival, de a monitornak
nem sikerlt megnyugtatan dntst hoznia a hiba krdsrl. A beszl ezrt lelltja
az artikulcit, s jraindtja a tervezst.
A szban adatolt lellsok pontosabban kvetik a morfolgiai struktrt, mint az jraindtsoki. Ezt jl demonstrlja az igekt (s a leg) kiejtst kvet (tlag: 260 ms),
valamint az sszetett szavak sszetteli hatrn tapasztalt lells (tlag: 344 ms). Mint
mr emltettk korbban, az artikulci megszakadst ezeknek az eseteknek a nagy
rszben a lexiklis hozzfrs problmja okozza. Ez lehet egyszer bizonytalansg,
a beszl jraellenrzi az aktivlt lexmt a kiejts eltt; vagy lehet a nyelvem hegyn van jelensg, amikor a lemma szintrl nem jut tovbb a lexma szintre. A toldalk eltt tapasztalt artikulcis lellsok itt is fknt a vonzat, illetve a szintaktikai
szerkezet tervezsi diszharmnijval kapcsolatosak, s jval hosszabbak a szerkesztsi szakaszok (tlag: 351 ms), mint a msik kt helyen adatoltak idtartama (5. bra).
A kis elemszm itt sem tette lehetv a statisztikai elemzseket. A sz belsejben
tartott sznetek idtartamai (tlag: 265 ms) az elbbi kett kztt helyezkednek el; ez
azzal is sszefgg, hogy az az nvelben megjelen artikulcis lells a kvet sz
aktivlsnak nehzsgt vagy egyb talaktsi problmt is jelezhet.
Az jraindtsok s a sznet a szban jelensg eseteiben az artikulcis lellsok
helye nehezen sszevethet, fknt a funkcionlis klnbsgek miatt. Az sszehasonltst az is nehezti, hogy pldul mg az az nvelben a magnhangzt kvet artikulcis lells a sz kzepn trtnik, addig ennek az jraindtsa abba a kategriba
kerl, amikor a sz els beszdhangja ismtldik. Pldk: akkor viszonylag gyorsan be
164
lehet fejezni a. z egyetemet, illetve ez a. az gy amirl beszl. (Megjegyezzk ugyanakkor, hogy a hatrozott nvel jraindtsa a felsznen tbbflekppen is kategorizlhat, hiszen a jelensg maga tbbfle okra vezethet vissza. Erre azonban a jelen
kutatsban nem trnk ki.) Az jraindtsok els sztagjt s a sznet a szban igektt kvet lellsa lnyegben hasonl tpusnak tekinthet; szmos esetben az jraindtsok els sztagja is igekt (vagy a leg). Ha ezeknek a szerkesztsi szakaszait
vetjk ssze, akkor nem is ltunk klnbsget az tlagokban (260 ms, ill. 257 ms), a
szrs valamivel kisebb az jraindtsokban.
165
4. Kvetkeztetsek
Kutatsunkban a spontn beszdben tapasztalhat kt megakadsjelensget, valamint a
szerkesztsi szakaszaikat elemeztk. A beszl mindkt esetben az artikulci egy
adott pontjn megszaktja az ejtst, sznetet tart. Eltrs van azonban a folytatsban; a
sznet a szban jelensg esetben a beszl a szerkesztsi szakaszt kveten folytatja
a sz ejtst, mg az jraindtsoknl ellrl kezdi az artikulcit. Feltteleztk, hogy
ezeknek a bizonytalansgi megakadsoknak a htterben tbb ok hzdhat meg. Lehetsges, hogy a rejtett monitor tvesen jelez, ilyenkor valjban nem volt hiba, tves
riaszts trtnt, a beszl valamirt mgis elbizonytalanodott. Ha a diszharmnia benyomsa nagymrtk volt (mert a problma esetleg tbb tervezsi szintet rintett),
akkor a beszl gy tli meg, hogy jra kell terveznie fonolgiailag s/vagy fonetikailag az elkezdett hangsort, st a monitorozs visszamenlegesen nagyobb idtartamban
is felttelezhet. Ilyenkor kvetkeznek be az jraindtsok. Ha a beszl relatve rvid
id alatt meggyzdik arrl, hogy nem trtnt hiba, akkor nincs szksge a teljes
lexma jratervezsre, elegendnek bizonyul az ellenrzs, s gy csupn a lellst
kvet, mg nem kiejtett rszletet tervezi jra, majd ejti ki. Az artikulci lellsa
jelezhet teht vals vagy vlt diszharmnit, amelynek a pontos eredete sokszor nehezen vagy egyltaln nem meghatrozhat.
Noha a kontextus nem ad minden esetben biztos tmpontot a httrokok meghatrozsra, az ugyanazon lexmt rint megakadsok elemzse rvilgthat a rejtett mkdsek sajtossgaira. Ilyen lehet pldul az az hatrozott nvel, amelyet a beszlk
relatve gyakran indtanak jra, illetve tartanak sznetet a nvel kt beszdhangja
kztt. A ht nem motivlja az embereket a. a. az idelisabb embereket az hogy tanri
plyra menjenek pldban a megfelel jelz kivlasztsa okoz nehzsget a beszlnek. Erre tbb tny is utal, a beszl nem elgedett a jelz nlkli emberek szval, az
a nvel ktszeri ismtlse jelzi a lexiklis vlogats tnyt. A beszl a megfelelnek
vlt lexma eldntse utn jraindtja a mr nyelvtanilag helyes nvelt (az). Az s a.
z zleti biztostk kzlsrszletben az zleti lexma hozzfrsnek nehzsge eredmnyezi az artikulci lellst, a sznet megjelenst. Ezekben az esetben a nvel
mssalhangzja rendszerint tapad a kvet sz kezd magnhangzjhoz a folyamatos
artikulci kvetkeztben.
Elfordulhat (br nem tl gyakori), hogy az elemzett ktfle jelensg egymst kveten fordul el, mint a kvetkez pldban. Radsul itt valsznstheten ugyanazon
diszharmnia okozhatta a nehzsget: jrakezdtem a. pszicholgit jobban mondva fel.
fel. eleventettem a dolgokat s []. A kzlsrszlet s a tgabb kontextus is arra
utal, hogy a beszl nem tudja pontosan megfogalmazni a kzlni szndkozott gondolatot, ez akadlyozza meg abban, hogy a grammatikai talaktst vghezvigye. A
hezitls utn gy folytatja: mr nem intzmnyestett keretek kztt, vagyis magyarzza, pontostja az jrakezdtem a pszicholgit kzlst, amelyet flrerthetnek tl, erre
utal a jobban mondva, ami a korrekcit vezetn be. Ekkor azonban tovbb nvekszik a
gondolati diszharmnia, amely a felsznen jraindtsban s szban tartott sznetben
lt testet. A pldban teht mindkt jelensg a gondolati tervezs szintjnek problmjra vezethet vissza.
166
167
Gyarmathy Dorottya 2009. A beszl bizonytalansgnak jelzsei: ismtlsek s jraindtsok. Beszdkutats 2009: 196216.
Hartsuiker, R. J. Kolk, H. H. J. 2001. Error monitoring in speech production: A computational test
od the perceptual loop theory. Cognitive Psychology. 42: 113157.
Hartsuiker, R. J. Kolk, H. H. J. Martensen, Heike 2005. The division of labor between internal
and external speech monitoring. In Hartsuiker, R. J. Bastiaanse, R. Postma, A. Wijnen, F.
(eds.): Phonological encoding and monitoring in normal and pathological speech. New York:
Psychology Press.
Horvth Viktria 2009. Funkci s kivitelezs a hezitcis jelensgekben. PhD-disszertci. Budapest: ELTE, Nyelvtudomnyi Doktori Iskola.
Levelt, W. J. M. 1983. Monitoring and self-repair in speech. Cognition. 14: 41104.
Levelt, W. J. M. 1989. Speaking. From Intention to Articulation. Cambridge: The MIT Press.
Nooteboom, S. 1980. Speaking and unspeaking: Detection adn correction of phonological and lexical
errors in spontaneous speech. In Fromkin, V. A. (ed.): Errors in linguistics performance: Slips of
the tongue, ear, pen, and hand. New York: Academic Press. 87102.
Oomen, C. C. Postma, A. 2001. Effects of increased speech rate on monitoring and self-repair.
Journal of Psycholinguistic Research. 30: 163184.
Postma, A. 2000. Detection of errors during speech production: a review of speech monitoring
models. Cognition. 77: 97131.
Postma, A. Kolk, H. H. J. 1993. The covert repair hypothesis: Prearticulatory repair processes in
normal and stuttered disfluencies. Journal of Speech and Hearing Research. 36: 472487.
Postma, A. Kolk, H. H. Povel, D-J. 1990. On the relation among speech errors, disfluencies, and
self-repairs. Language and Speech. 33: 1929.
Roelofs, A. 2004. Errors biases in spoken word planning and monitoring by aphasic and nonaphasic
speakers: Comment on Rapp and Goldrich (2000). Psychological Review. 111: 561572.
Slevc, R. L. Ferreira, V. S. 2006. Halting in single word production: a test of the perceptual loop
theory of speech monitoring. Journal of Memory and Language. 54: 515540.
168
1. Bevezets
A spontn beszd szmos megakadsjelensget tartalmaz, amelyek kzl az egyik
leggyakoribb a hezitls, vagy ms nven kitlttt sznet (a kznyelvben zsnek is
nevezik hangzsa miatt, mert leggyakrabban az //-re emlkeztet semleges magnhangzknt realizldik). A hezitls a spontn beszd termszetes jelensge, szmos
funkcit tlt be a beszdben. Egyrszt idt biztost a vlogatshoz a gondolat nyelvi
formjnak tervezse sorn, egyttal reflektl a keressi folyamatra (Beattie
Butterworth 1979). A kitlttt sznet megjelenhet az egyes beszdtervezsi szinteken
fellp hiba ksrjelensgeknt is, egyttal idt biztost a hibajavtsi folyamatokhoz
(Levelt 1989). A kitlttt szneteknek a trsalgsban is szmos funkcijuk van (beszdszndk jelzse, beszlvlts kivitelezse).
A beszlk kztt nagy klnbsgek vannak abban a tekintetben, hogy milyen
gyakran produklnak kitlttt sznetet (Goldman-Eisler 1968, Mark 2004,
FehringerFry 2007, Horvth 2010), de ez nem fggetlen a beszd mfajtl sem. A
spontn narratvkban s a kplersban (Mark 2004) nagyobb a hezitls arnya,
mint a dialgusokban (Horvth 2004). A feladat vagy a beszdmfaj mellett a kzls
tmja is hatssal van a hezitls gyakorisgra. Ha a beszlnek olyan tmban kell
megnyilatkoznia, amelyet kevss ismer, nvekszik a kitlttt sznetek arnya
(Bortfeld et al. 2001, MerloMansur 2004). j informcit vagy tmt nagyobb
arnyban elz meg hezitls a trsalgs sorn, mint egy mr ismert informcit hordoz nyelvi egysget (Arnold et al. 2000). A kzls hossza is eljelezheti a hezitlsok
gyakorisgt: hosszabb kzlsek eltt gyakrabban jelennek meg (Shriberg 1996). Minl hosszabb s komplexebb ugyanis maga a megnyilatkozs, annl nagyobb az eslye
valamilyen diszharmnia megjelensnek a beszdtervezs sorn. A hezitls teht azt
jelzi a hallgat szmra, hogy a kvetkez kzls relatve hossz s komplex lesz
(Watanabe et al. 2008).
A kitlttt sznetek elemzse nemcsak a pszicholingvisztika vagy a fonetika szmra fontos; a beszdtechnolgiai alkalmazsok is fontos tnyezknt kezelik. A hezitlsok ugyanis sok esetben rontjk a beszdfelismers eredmnyt, azon bell a szbeszrsok s -trlsek, illetve a tves elutastsok szmt nvelik (Kaushik et al.
2010). A beszdfelismerk egy rszben ezrt megtallhatk a hezitlsokra vonatkoz modellek (Ward 1991, NakagawaKobayashi 1995, KaiNakagawa 1995). Az
egyik HMM-es beszdfelismerben pldul (NakagawaKobayashi 1995) a gyakran
elfordul kitlttt szneteket hozzadtk a rendszer szkszlethez, mg egy msik
alkalmazsban (KaiNakagawa 1995) a kitlttt szneteket mint a sztron kvli szt
169
170
171
tat; ha egy beszl beszdre hosszabb nma sznetek jellemzk, akkor felttelezheten a kitlttt sznetek is hosszabb idtartamban valsulnak meg (3. bra).
173
A beszlk a spontn beszd sorn mintegy 1015%-ban a szn bell tartanak sznetet, amely szintn lehet nma vagy jellel kitlttt (Gsy 2005, Horvth 2009, Bna
2010). Az igektket vagy sszetett szavak els tagjt kvet szn belli sznet a
lexiklis elhvs problmjra utal, mg a toldalkot megelz sznet a grammatikai
tervezs zavart jelzi. A jelen korpuszban is elfordult, hogy a beszl egy szn bell
tartott kitlttt sznetet (5. bra), felteheten azrt, mert gondot okozott neki a pontos v
felidzse.
174
A kitlttt sznetek s a tbbi beszdhang osztlyozshoz hasznlt 3 llapot HMMeket hasznltunk. Az egyes llapotokban a kibocstsi eloszlsokat Gauss-fggvnyek
slyozott sszegvel szoks lerni. 2, 4, 8 s 16 komponenssel vizsgltuk a kitlttt
sznetek osztlyozsi eredmnyessgt. A legjobb eredmny a 16 Gauss-komponens
modell adta, amelynek eredmnye 91,8% volt (6. bra).
Mssalhangz
Hezitls
Magnhangz
92,33%
0%
7,67%
Mssalhangz
0,13%
89,31%
10,56%
Hezitls
1,67%
0%
98,33%
175
64,28%
35,71%
46,15%
53,84%
4. Kvetkeztetsek
A hezitls a spontn beszd gyakori jelensge, amely szmos funkcit tlt be a tervezsi s nellenrzsi folyamatokban, de fontos szerepe van a trsalgsban is a beszdszndk vagy a beszlvlts jelzsre. A jelen kutats krdse az volt, hogy a hezitlsok hogyan valsulnak meg a spontn beszdben s osztlyozhatk-e automatikusan.
A jelen korpuszban a hezitlsok tbbfle formban, szmos funkciban s nagyon
vltozatos idtartamban realizldtak. A kitlttt sznetek 28%-a irregulris
foncival valsult meg.
A hezitlsok MFC egytthatkkal elfeldogozva 3 llapot HMM-ekkel 98,33%os eredmnnyel osztlyozhatk a spontn beszdben. A hezitlstpusok azonban csak
59%-os eredmnnyel osztlyozhatk automatikusan SVM-mel.
A kutats sorn kialaktott gpi osztlyoz eredmnye kivlnak szmt a nemzetkzi irodalomban lertak tkrben is.
Eredmnyeink felhasznlhatk a beszdfelismers eredmnynek javtsra, illetve
a beszdkutats szmos terletn.
Irodalom
Arnold, J. E. Wasow, T. Losongco, A. Ginstrom, R. 2000. Heaviness vs. Newness: the effects
of structural complexity and discourse status on constituent ordering. Language. 76/1: 2855.
Audhkhasi K. Kandhway, K. Deshmukh, O. D. Verma, A. 2009. Formant-based technique for
automatic filled pause detection in spontaneous spoken English. IEEE International Conference
on Acoustics, Speech and Signal Processing. Taipei. 48574860.
Beattie, G. W. Butterwoth, B. L. 1979. Contextual probability and word frequency as determinants
of pauses and errors in spontaneous speech. Language and Speech. 22: 201211.
Boersma, P. Weenink, D. 2009. Praat: doing phonetics by computer (version 5.1).
http://www.praat.org
Bna Judit 2010. Bizonytalansgi megakadsok idsek s fiatalok spontn beszdben. Beszdkutats 2010. 125138.
Bortfeld, H. Leon, S. D. Bloom, J. E. Schober, M. F. Brennan, S. E. 2001. Disfluency Rates
in Conversations: Effects of Age, Relationship, Topic, Role and Gender. Language and Speech.
44/2: 123147.
Bhm T. Ujvry I. 2008. Az irregulris fonci mint egyni hangjellemz a magyar beszdben. In
Gsy Mria (ed.): Beszdkutats 2008. Budapest: MTA Nyelvtudomnyi Intzet. 108120.
Fehringer, C. Fry, C. 2007. Hesitation phenomena in the language production of bilingual speakers.
Folia Linguistica. Acta Societatis Linguisticae Europae. 41: 3872.
176
177
1. Bevezets
A spontn beszdet szmos tnyez befolysolja. A beszd aktulis megvalsulsi
formja (a kzls szerkesztettsge, a szegmentlis s a szupraszegmentlis szerkezet
stb.) nagymrtkben fgg a beszl genetikai adottsgaitl, anyanyelvi ismereteinek
biztonsgtl, beszdben val gyakorlottsgtl; meghatroz tovbb a beszdtma
(ismert vagy ismeretlen dologrl kell nyilatkoznia), a beszdhelyzet (formlis, informlis szituci) s a beszdpartner szemlye. Mindezeken tl a kzls szempontjbl
fontos az egyn mentlis llapota (pl. afzis, ill. demens betegek beszdre jellemz az
egyes hiba tpus jelensgek tlslya; v. Szpe 2004, 2006, HoffmannNmeth 2006),
aktulis pszichs (dh, rm, bnat, szorongs) s fizikai llapota (kimerltsg, betegsg, alkoholos, ill. drogos befolysoltsg; v. Braun et al. 1992, BraunKnzel 2003,
Gyarmathy 2007, 2010). Kommunikcis szempontbl alapvet jelentsg az az
akusztikai krnyezet, amelyben a megnyilatkozsok elhangzanak. Az rthetsg fgg a
beszd hangerejtl, a beszl s a hallgat kzti tvolsgtl, a krnyezeti zajoktl,
illetve ezek sszhangjtl (Tarnczy 1984). Az emberek kztti kommunikci szinte
mindig valamifle hangz kzegben megy vgbe, mivel minden, ami a vilgban l s
mozog, hangot kelt (Pap 2002).
A krnyezet hangjelensgeit az emberi fl (p halls esetn) a 20 Hz s 20 000 Hz
kztti frekvenciatartomnyban kpes rzkelni (Gsy 2004), ami nem jelenti azt,
hogy az ezen kvl es hangok ne lennnek hatssal a szervezetnkre. Az infrahangok
(16 Hz alatti hangok) nagy intenzits esetn tbbek kztt dezorientcit, rmltsget,
zavartsgot, rossz kzrzetet s egyenslyvesztst okozhatnak, gtolhatjk tovbb az
rzkszervek megfelel mkdst s szlssges esetben epilepszis rohamot is kivlthatnak (Bartha 2004). A magas hangteljestmnyszint ultrahangok (20 000 Hz
fltti rezgsek) elidzhetnek flzgst, hnyingert, szdlst s fejfjst.
A tarts zajterhels mg a lertaknl is slyosabb problmk kivltja lehet. A fizikai definci szerint a zaj vgtelen sok rezgst tartalmaz, szablytalan szerkezet
hang (Tarnczy 1984); mg kznapi rtelemben az adott krlmnyek kztt zavar
hangot rtjk alatta. A krnyezeti zajok hatst gyakorolnak a hallszervnkre, amelynek elsdleges kvetkezmnye a hallskrosods (Pap 2002); emellett azonban slyos
pszichikai s vegetatv tneteket is elidzhetnek (Tulipnt 2004). A zaj zavarja s
htrltatja a gondolkodsi folyamatokat, a szellemi feladatmegold kszsget; magatartsbeli vltozsokat, illetve stressztneteket eredmnyezhet. A krnikus zajterhels
a pszichikai hatsokon tl olyan vegetatv kvetkezmnyekkel is jrhat, mint az
anyagcserezavarok, a keringsi rendszer mkdsi zavarai, a pulzusszm s a vrnyoms nvekedse, kardiovaszkulris rendellenessgek, lgzszervi (pl. bronchitis),
illetve mozgsszervi (pl. zleti gyullads) panaszok, allergia s depresszi (Tulipnt
2004, Niemann et al. 2005).
178
Mind a csendben, mind a zajban rgztett korpuszban a beszdtervezs bizonytalansgbl add diszharmonikus jelensgek (hezitlsok, nyjtsok, ismtlsek, jraindtsok s tltelkszavak) voltak tlslyban. A csendes beszdrszben az sszes adatolt
jelensg 93,17%-t tettk ki; ez az arny a zaj hatsra csekly mrtkben nvekedett
(93,53%). A klnfle zajok hatsra teht arnyaiban valamivel tbb bizonytalansgra utal jelensget tudtunk adatolni, az egyes jelensgek percenknti elfordulsa
azonban cskkent, vagy alig vltozott (2. bra). Kismrtk nvekedst egyedl az
jraindtsok esetben talltunk.
180
181
Az egyes zajtpusoknl eltren alakult a megakadsjelensgek percenknti elfordulsa. A hiba tpus jelensgek arnya a kutyaugats (0,63%-kal cskkent) s a fogfrs
(0,61%-kal ntt) esetben a zaj hatsra alig vltozott, a tbbi zajnl kismrtkben ntt
(lgkalapcs: 2,5%; zene: 1,04%). Jelentsebb vltozs csupn az egyes tpusok gyakorisgban mutatkozott. A kutyaugats kivtelvel minden csoportban tbbfle hiba
jellemezte az adatkzlk beszdt; a fogfrs esetn pldul 18-fle megakads fordult
el. Egy korbbi kutatsban (Gsy 2008) szabad hangtrben kzvettett trsalgsi zaj
segtsgvel (jel/zaj arny: 15 dB) vizsgltk a beszdzaj spontn beszdre gyakorolt
hatst. Az eredmnyek szerint a zajos beszdrszben szignifiknsan tbb
megakadsjelensg fordult el. Az eltr vgeredmny a vlasztott zajok tpusval,
illetleg a mdszertani klnbzsggel magyarzhat. Egy trsalgs, mint httrzaj,
sokkal zavarbb, mivel a beszl nkntelenl is megprblja megrteni, feldolgozni
azt, s ez nagyban htrltatja beszdtervezsi folyamatait. Mindennapos tapasztalat,
hogy ha az ember krnyezetben beszlgetnek, vagy szl a tv, rdi stb., nehzkesebben, lassabban tud olvasni, tanulni, fogalmazni stb. A monoton zajok (forgalom
zaja, kutyaugats, tpts stb.) ezzel szemben hamarabb vlnak megszokott, s taln
a gondolati tevkenysget is kevsb zavarjk. Az ltalunk felhasznlt httrzajoktl a
klasszikus zene klnbzik a tekintetben, hogy hatsa nem azonos a zajval. Az eddigi
kutatsok mg nem tudtk egyrtelmen tisztzni a pszichs teljestkpessgre gyakorolt hatst, ami termszetesen fgg a zene fajtjtl s hangerejtl (egy diszkbl
kihallatsz hangos zene a krnyez lakosok szmra zajknt realizldhat) is. A ksrleti eredmnyek azonban igazoltk, hogy a halk zene kedvezen hat a pszichs llapotra, nem cskkenti, hanem inkbb nveli a figyelmet (Tulipnt 2004). A zene jtkony,
gygyt hatsa mr a Bibliban is megjelenik: Az r lelke elhagyta Sault, s egy
gonosz llek zaklatta, az r kldte. Szolgi gy szltak Saulhoz: Minden bizonnyal
182
Istennek egy gonosz lelke zaklat. Parancsolja ht meg urunk, s szolgi keresnek neki
valakit, aki tud hrfzni. Aztn, ha rd tr Istennek a gonosz lelke, megpendti a
hrokat s jobban leszel. [] S ahnyszor csak rtrt Saulra az Isten lelke, Dvid
fogta a hrft s jtszott. Ilyenkor Saul megnyugodott, jobban lett, s a gonosz llek
odbbllt (I. Sm. 16:1416, 16:23). A zennek a krnyezeti zajoktl elklnl
hatsa a javtsi stratgikban is megmutatkozik, ahogy az a ksbbiekben lthat
lesz.
3.2. A megakadsjelensgek javtsa
Feltteleztk, hogy a krnyezeti zajok negatvan befolysoljk majd az nkorrekcis
folyamatok mkdst. A 4. brrl leolvashat, hogy az egyes csoportok beszli a
csendes beszdrszben is eltr sikeressggel javtottk a beszdbeli hibikat (100%nak az sszes javthat megakadsjelensget tekintettk). A legjobb teljestmnyt
nyjt csoport is tves kivitelezseinek csupn kevesebb mint a felt (45,71%) volt
kpes korriglni. A kutyaugats a javts szempontjbl elklnlni ltszik a tbbi
httrzajtl. Egyedl ebben a csoportban fordult el, hogy a korrekcis folyamatok
jobban mkdtek a zaj hatsra. Az adatkzlk a zajos beszdrszben 14,29 szzalkponttal magasabb arnyban javtottk a hibikat. A tbbi csoportnl az nmonitorozs
sikeressge negatvan vltozott. A fogfrs esetben 6,32 szzalkpontos, a zene esetben 14,58 szzalkpontos, mg a lgkalapcs esetben 13, 89 szzalkpontos cskkens volt adatolhat. Meglep mdon az ltalunk szubjektve legzavarbbnak s
legkevsb zavarnak minstett zajtpusok esetben mutatkozik a legnagyobb mrtk visszaess az nkorrekcis folyamatok mkdsben. A lgkalapcs esetben ez
egyrtelmen a zaj kellemetlensgvel magyarzhat, mg a zene esetben ktfle
magyarzat is lehetsges. Elkpzelhet, hogy mivel adatkzlink fiatal egyetemistk
voltak, szmukra idegenknt, kellemetlenknt hatott a klasszikus zene. A msik valsznsthet ok, hogy a beszlk a megnyilatkozsaik helyett inkbb a zenehallgatsra koncentrltak, ez akadlyozta meg ket a szksges javtsok elvgzsben.
183
hibk (csendben: 0,16 db/perc; zajban: 0,27 db/perc), a tves kezdsek (csendben 0,16
db/perc; zajban: 0,41 db/perc) s az egyszer nyelvbotlsok (csendben: 0,95 db/perc;
0,53 db/perc) elfordulsi arnya a korpuszban nagyon alacsony volt. A kzls rtelmezhetsgt valsznleg nem befolysoltk olyan mrtkben, hogy a beszlk javtsukat fontosnak tartsk. A nyelvi tervezs szintjn ltrejv grammatikai hibkat a
csendes beszdkrnyezetben az nmonitoroz mechanizmus kpes volt felismerni s
javtani, azonban a korltozott mkdse rvn a zajos beszdrszben romlott a korrekci sikeressge. Pldul: ott valaki lent vr, mghozz egy rgi j haverja, akivel
/szndkolt sz: akinek/ meggrte, hogy az rmeit mindig megosztja. Az egyes jelensgekkel kapcsolatos javtsi stratgikat teht gy tnik, a zene vltoztatta meg a
legkisebb mrtkben.
A beszlkben kellemetlen kzrzetet kivlt fogfrs hangja nagymrtkben befolysolta az egyes jelensgek javtst. ltalnossgban elmondhat, hogy az adatkzlk
kevsb voltak kpesek korriglni hibikat, a tervezs kzben felmerlt problmkat
kevsb eredmnyesen oldottk meg (7. bra). A mentlis lexikon tves aktivlsa
folytn keletkezett mindkt diszharmonikus jelensg esetben kzel 20 szzalkpontos
visszaesst figyelhetnk meg. Abbl azonban, hogy az adatkzlk a zajos felvtelrszben a tves kezdseknek 60%-t, mg a tves sztallsoknak mindsszesen 25%t javtottk, arra kvetkeztethetnk, hogy a tves aktivlsok nagy rszt a rejtett
nmonitorozs segtsgvel ismertk fel, s korrigltk. Amikor ugyanis a rejtett monitorozs hibt jelez, az artikulci flbeszakad, s lehetsg szerint megtrtnik a
javts, pldul: a harmadik kpen az lthat, hogy Jzs Bla mg mindig mszik; s
ebben nem tudt nem tudtak megegyezni. A javtott tves sztallsoknl a hibs
lexmt a sz teljes elhangzsa utn vlheten a felszni monitor ismerte fel, pldul:
t valahogy ugye az erklyrl illetve a magas emeletrl le kell vinni a fldszintre. A
korrekcis folyamatok az artikulcis tervezs kzben sem mkdnek megfelelen; a
fogfrs hangjnak kvetkeztben az anticipcik javtsa is cskken. Egyedl a
grammatikai hibknl tapasztalhatunk pozitv irny vltozst, ez azonban a kis
185
A negyedik zajtpus, a kutyaugats elklnl a tbbitl, ennek hatsra ugyanis szszessgben ntt a javtott megakadsjelensgek arnya. Az egyes tpusok elemzsbl kiderl, hogy a tves kezdsek (pl. s a n, akivel idn ezeltt nagy hvvel lelkezett; le is engedi a kle ktelet bartjnak), a grammatikai hibk (pl. a felmsz
embert vagy az emberre puskt fog) s az anticipcik (pl. felttelezhetjk, hgy a
n hogy a n be akarta csapni) javtsi arnya jelentsen megntt (8. bra). Egyrszt
a beszlk a kls zavar tnyez ellenre igyekeztek jobban ellenrzsk alatt tartani
a tervezs s a kivitelezs folyamatait, msrszt pedig elkpzelhet, hogy a kutyaugats volt az a zajtpus, ami a leginkbb illeszkedett a trtnetbe. A kpsor ugyanis egy
nagyvrosi krnyezetet brzol, ahol az egyik laksba pp betrnek. Az akusztikus
visszacsatols hinya azonban itt is megmutatkozik; a felszni monitorozs nem mkdik megfelelen, aminek kvetkeztben a beszlk a tves sztallsaikat nem kpesek javtani. Pldul: egy n cigarettzik, s htrafele (szndkolt sz: httal) l a
frfinek.
186
4. Kvetkeztetsek
A jelen tanulmny a krnyezeti zajok spontn beszdre gyakorolt hatst vizsglta
ngy zajtpus segtsgvel. A 20 adatkzltl szrmaz sszesen 60 perc 49 msodperces felvtelben a nma szneteket is figyelembe vve 2439 darab
megakadsjelensg fordult el. A zaj hatsra megvltozott a hibk s a bizonytalansgok egymshoz viszonytott arnya. A zajos beszdrszekben nagyobb arnyban
fordultak el a beszl bizonytalansgra utal megakadsjelensgek, amibl arra
kvetkeztethetnk, hogy az adatkzlknek a tervezs kzben tbb problmjuk akadt
ugyan, de mivel a zaj hatsra kzlsket fokozottabban kontrollltk, a tervezsi
diszharmnik nem realizldtak hibaknt a felsznen. A hibadetektlst s a javtst a
kzlsben a bizonytalansgok jelzik.
Elzetes felttelezseinkkel ellenttben a zajos beszdrszben percenknt kevesebb
megakadsjelensget tudtunk adatolni. A megakadsok elfordulsi gyakorisga a
legnagyobb mrtkben a fogfrs csoportjba tartoz adatkzlknl cskkent, mg a
legkisebb arny visszaess a kutyaugats esetben volt tapasztalhat. Valsznleg a
beszlk a tapasztalt httrzaj hatsra jobban sszpontostottak a beszdprodukci
rszfolyamataira, hogy a feladatot minl sikeresebben teljestsk, illetleg megnyilatkozsaik a hallgat szmra rthetek maradjanak. A legtbb megakads percenknt a
kutyaugats csoportjba tartoz adatkzlk zajos felvtelrszben fordult el. Vlheten a leginkbb ez a zajtpus zavarta a tervezs s a kivitelezs sszehangolst. Ha a
megakadsjelensgek percenknti elfordulsa nem is nvekedett, a httrzaj negatv
hatsa egyrtelmen megmutatkozott abban, hogy mind a ngy csoportnl tbbfle
hibajelensg volt adatolhat a zajos felvtelrszben. Ez egyrtelmen utal arra, hogy
beszlk a folyamat tbb szintjn is problmval kzdenek. A TOT-jelensg a lemma
s a lexma szint kztti folyamatmegszakadst, a kontaminci a nyelvi tervezs,
mg az anticipci az artikulcis tervezs zavart jelzi.
Az eredmnyek igazoltk, hogy ha az akusztikai visszacsatols valamely oknl fogva gtolt, az krosan hat az nmonitoroz mechanizmus mkdsre. A kutyaugatst
187
188
189
1. Bevezets
A beszdpercepci bonyolult mkdsmechanizmus, amelynek jelents rszrl mg
kevs pontos tudomnyos ismeretnk van. Aktv folyamat, amelynek sorn a hallgat
a beszdjeleket magasabb szinteken rtelmezi. A beszdpercepcis folyamatok mkdsnek brzolsra klnfle modelleket alkottak. Ezek kzl a legelterjedtebb a
hierarchikus, interaktv beszdmegrtsi modell. Ez a beszdfeldolgozs folyamatt
egymssal sszefgg szinteken kpzeli el, amelyek klcsnsen hatnak egyms mkdsre. Ezek a szintek a halls, az szlels, a megrts s az rtelmezs szintjei. Az
letkornak megfelel, tipikus beszdszlels s megrts p hallson alapszik, a beszdmegrts s rtelmezs megfelel mkdshez pontos beszdszlels szksges
(Gsy 2005).
A beszdpercepci sikeressgt az egyes szintek sszehangolt mkdse mellett
komplex memriarendszerek egyttes munkja biztostja. Mind a hossz tv, mind a
rvid tv memriatpus meghatroz a nyelv szempontjbl; s mindegyiknek van
egy elnyben rszestett kdolsi mdja verblis anyagokra. A hossz tv emlkezet
a jelentsen alapszik, a rvid tv memria jellemz reprezentcija akusztikus jellemzk sszessge (Atkinson et al. 1997). Az egyni nyelvi teljestmnyek klnbsgeinek htterben felteheten eltr munkamemria-kapacits is ll. A nagyobb munkamemria-kapacits jobb mondat- s szvegrtst tesz lehetv azokban a feladatokban, ahol a feldolgozs utn krdsekkel vizsgljk a megrtst (Nmeth 2006).
Baddeley modellje szerint a munkamemria hrom rszbl ll, ezek: 1. a tudatos
kontrollt vgz, korltozott kapacits kzponti vgrehajt, amely nagyon hasonlt a
figyelemre; 2. a beszdalap informci trolsra szolgl fonolgiai hurok (eredetileg artikulcis hurok); illetve 3. a vizulis-tri informcikat trol vizulis-tri vzlatfzet. Az artikulcis hurok rszei 1. a passzv fonolgiai tr, amely kzvetlenl a
beszdpercepcival van kapcsolatban; illetve 2. egy artikulcis folyamat, amely a beszdprodukcival kapcsolatos (EysenckKeane 2003). A fonolgiai hurok terjedelmt
ksrletben vizsgltk (Baddeley et al. 1975). A ksrleti szemlyek annyi szt tudtak
kzvetlenl egyms utn felidzni, amennyit kt msodperc alatt hangosan fel tudtak
olvasni. Ez kapcsolatban ll az egyes egynek artikulcis s beszdtempjval is.
Vagyis a munkamemria terjedelmt kt tnyez hatrozza meg: a fonolgiai hurok
terjedelme s az elemek artikulcis ideje. A fonolgiai hurok mkdst mutatja a
fonolgiai hats, a szhosszsgi hats (kevesebb szt tudunk megtanulni, ha azok
hosszabbak, mint amikor rvidebb szavakat tanulunk), az artikulcis elfedsi hats,
illetve a nem figyelt beszd hatsa (Gsy 2005).
5
190
191
Mindhrom letkori csoportbl tz-tz f (2 frfi s 8 n) mondatismtlst vlasztottuk ki. A fiatalok tlagletkora 26,5 v (2032 vesek); az idsdk 61,8 v (60
64 vesek); az idsek 80,7 v (7590 vesek) volt. Mindegyik adatkzl magyar
anyanyelv volt, s egyikknek sem volt klinikai rtelemben vett hallscskkense.
Az adatkzlk kivlasztsnl gyeltnk arra is, hogy hasonl iskolai vgzettsgek
legyenek, s mindnyjan (letkoruknak megfelel) aktv letet ljenek.
A felvteleket a Praat 5.0 szoftverrel annotltuk. Meghatroztuk a pontos ismtlsek arnyt, megmrtk a reakciidket (a ksrletvezet mondatnak utols hangja s
az adatkzl ismtlsnek els hangja kztt eltelt idt), majd elemeztk a tves ismtlseket a bennk elfordul hibk szerint. Helyes ismtlsnek tekintettnk minden
olyan ismtlst, amikor az adatkzl pontosan visszamondta a mondatot, fggetlenl
attl, hogy kzben hezitlt-e vagy hosszabb gondolkodsi sznetet tartott, esetleg
valamilyen (javtott!) hiba tpus megakadst produklt hiszen ezek a jelensgek az
szlelstl s a memritl fggetlenl is megjelennek a beszdprodukciban.
Elemeztk azt is, hogy a mondatok hossza, grammatikai szerkesztse (egyszer
vagy sszetett), illetve a tartalom knnyebbsge/neheztettsge hogyan fgg szsze/sszefgg-e a helyes ismtlsekkel s reakciidkkel.
Az adatokon statisztikai elemzst vgeztnk az SPSS 13.0 szoftverrel.
3. Eredmnyek
A fiatalok tlagosan a mondatok 92,4%-t ismteltk meg pontosan (a szrs: 60
100%); az idsdk a mondatok 84%-t (a szrs 56100%); az idsek a 70,4%-t (a
szrs: 3292%). A fiatalok kzl ngyen nyjtottak 100%-os teljestmnyt, az idsdk kzl ketten, mg az idsek kzl senki sem tudta mind a 25 mondatot hibtlanul
megismtelni. Az idseknl a legjobb eredmny 23 mondat pontos megismtlse volt,
ezt egy adatkzlnl adatoltuk. A leggyengbben teljest fiatal adatkzl fradtsgra
panaszkodott, s mindssze 15 mondatot ismtelt pontosan, a tbbi fiatal 2325 helyes
mondatismtlst produklt. Az idsdknl a leggyengbb teljestmny 14 mondat
pontos ismtlse volt, ketten nyjtottk ezt az eredmnyt; mg az idsek kzl a leggyengbben teljest adatkzl 8 mondatot ismtelt meg helyesen. Gyakori volt mind
az idseknl, mind az idsdknl, hogy miutn a ksrletvezet ismertette a feladatot,
megjegyeztk, hogy a memrijuk mr nem mkdik megfelelen. A helyes ismtlsek szmban az egytnyezs varianciaanalzissel szignifikns klnbsget kaptunk
(F(2, 27)=4,021; p=0,030); a Tukey post hoc teszt szerint azonban csak a fiatalok s
az idsek kztt volt szignifikns a klnbsg (p=0,024), az idsdk eredmnyei
egyik csoporttl sem klnbztek szignifiknsan.
Elemeztk, hogy mely mondatok ismtlse volt a legknnyebb, illetve okozta a legnagyobb nehzsget a klnbz letkor adatkzlknek. Egyetlen olyan mondat volt
(17.), amelyet minden adatkzl hibtlanul meg tudott ismtelni: Megri biztostst
ktni minden klfldi utazs eltt. Ez a rvidebb mondatok kz tartozott, s mind a
szkincse, mind jelentse htkznapi volt. Az 1. brn megfigyelhet, hogy az egyes
mondatok eltr mdon okoztak nehzsget a klnbz letkor adatkzlknek.
192
193
194
Elemeztk a mondatok sztagszma, sszetettsge, illetve a tartalma s a helyes ismtlsek kztti sszefggseket. A Pearson-korrelci szerint a fiatalok helyes ismtlseinek szma nem fggtt ssze egyik tnyezvel sem. Az idsdknl a sztagszm nvekedse s a mondat tartalmi neheztettsge kzepesen ers szignifikns
negatv korrelcit mutatott a helyes ismtlsek szmval: a sztagszmtl fggen
r=-0,585; p=0,002; a tartalom nehzsgtl fggen r=-0,492; p=0,012. Az idseknl
a helyes ismtlsek s a sztagszm kztt kaptunk kzepesen ers szignifikns negatv korrelcit: r=-0,438; p=0,029. Ez azt jelenti, hogy a kt idsebb korcsoportnl a
sztagszm nvekedsvel cskkent a helyes ismtlsek szma; illetve az idsdknl
a nehezebb mondatok esetn is kevesebb volt a helyes ismtls.
A legrvidebb tlagos reakciidt a fiatalok produkltk (457 ms; szrs: 04218
ms); az idsdk tlagos reakciideje 502 ms volt (szrs: 03229 ms); a leghoszszabb tlagos reakciidt az idseknl mrtk (558 ms; szrs: -2092795 ms) (4.
bra). Az idseknl kt esetben is negatv reakciidt adatoltunk: ekkor az adatkzl
mr azeltt elkezdte az ismtlst, mieltt a mondat teljes egszben elhangzott volna.
Ez egyrszt magyarzhat a korltozott memriakapacitssal: az adatkzl mielbb
ismtelni akarta a mondatot, amg az informci elhvhat volt a munkamemribl.
Msrszt lehetsges, hogy a mondat befejezsnek egyrtelm elrejelezhetsge
miatt az adatkzl nem tartotta szksgesnek vgigvrni a kzls elhangzst. A kt
mondat, amelynl negatv reakciidt mrtnk, a kvetkezk voltak: Szerencstlenl
alakultak az esemnyek a tegnapi tszment akci sorn. A kormnyrendelet alapjn
tilos lesz petrdzni szilveszterkor is.
A leghosszabb reakciidt egy fiatal szemly produklta (4218 ms). Az idsdknl s az idseknl a 2 msodpercnl hosszabb reakciidk pontatlan ismtlssel jrtak
egytt. Az egytnyezs varianciaanalzis szerint szignifikns a klnbsg a csoportok
kztt: F(2, 738)=4,837; p=0,008. A Tukey post hoc teszt alapjn azonban a klnbsg csak a fiatalok s az idsek kztt szignifikns (p=0,006); az idsdk reakciidi
egyik csoporttl sem klnbznek statisztikailag.
195
196
197
Irodalom
Atkinson, R. L. Atkinson, R. C. Smith, E. E. Bem, D. J. 1997. Pszicholgia. Budapest: Osiris
Kiad.
Baddeley, A. D. Thomson, N. Buchanan, M. 1975. Word lenght and the structure of short-term
memory. Journal of Verbal Learning and Verbal Behaviour. 14: 575589.
Bna Judit 2011. A beszdpercepci s a beszdprodukci sszefggsei fiatal, idsd s ids
korban. Gygypedaggiai Szemle. 2011/3: 221232.
Czigler Istvn 2000. Megismersi folyamatok vltozsa felnttkorban. In Czigler Istvn (szerk.): Tl
a fiatalsgon. Megismersi folyamatok idskorban. Budapest: Akadmiai Kiad. 11130.
Eysenck, M. W. Keane, M. T. 2003. Kognitv pszicholgia. Hallgati kziknyv. Budapest: Nemzeti Tanknyvkiad.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Osiris Kiad.
Gsy Mria 2008. Magyar spontnbeszd-adatbzis BEA. Beszdkutats 2008. 194207.
Hnath-Chisolm, T. Willott, J. F. Lister, J. J. 2003. The aging auditory system: anatomic and
physiologic changes and implications for rehabilitation. International Journal of Audiology. 42:
2310.
Humes, L. E. 1996. Speech undertstanding in the elderly. Journal of the American Academy of
Audiology. 7: 161167.
Ivn Lszl 2002. Az regeds aktulis krdsei. Magyar Tudomny. 47: 412418.
Janse, E. 2009. Processing of fast speech by elderly listeners. Journal of the Acoustical Society of
America. 125: 23612373.
Janse, E. van der Werff, M. Quen, H. 2007. Listening to fast speech: Aging and sentence
context. In Trouvain, J. Barry, W. J. (eds.): Proceedings of the 16th International Congress of
Phonetic Sciences. 681684.
Kemper, S. 1987. Syntactic complexity and elderly adults prose recall. Experimental Aging Research. 13: 4752.
Nmeth Dezs 2006. A nyelvi folyamatok s az emlkezeti rendszerek kapcsolata. Budapest: Akadmiai Kiad.
Neuberger Tilda 2010. Mondatszlelsi sajtossgok fiatal s ids korban. In Gecs Tams Srdi
Csilla (szerk.): j mdszerek az alkalmazott nyelvszeti kutatsban. Budapest: Tinta Knyvkiad.
220225.
Schneider, B. A. Daneman, M. Pichora-Fuller, K. M. 2002. Listening in Aging Adults: From
discourse comprehension to psychoacoustics. Canadian Journal of Experimental Psychology. 56:
139152.
198
1. Bevezets
A modern trsadalmak minden gyermektl elvrjk, hogy megtanuljon olvasni s rni,
hiszen analfabtnak lenni risi htrnyt jelent a mai vilgban. A legtbb ember szmra az olvass megtanulsa nem okoz klnsebb nehzsget, hiszen csak a beszlt nyelv rott vltozatt kell elsajttani. gy a felnttek jelents rsze egyltaln
nem, vagy alig emlkszik vissza az els osztlyra, a kezdeti nehzsgekre. Szmukra
az olvass elsajttsnak folyamata pozitv volt, gy nem hagyott klnsebb emlknyomot maga utn. Van azonban a trsadalomnak egy kis szzalka, akiknek a betk
s szavak megfejtse nem megy ilyen knnyen s grdlkenyen. k azok, akik rengeteg gyakorls utn is olvassi nehzsggel kzdenek, akik korltokba tkznek p
rtelmi kpessgeik ellenre. Ma a gyermekek mintegy 410%-a tartozik a diszlexisok s gyengn olvask krbe (Meixner 1993).
Az olvassi nehzsg s a diszlexia elklntse fontos, de nehz feladat. A diszlexia ugyanis olvassi nehzsg, azonban nem minden olvassi nehzsg diszlexia
(Gsy 1999). A diszlexia fogalma napjainkban mr nagyrszt egysges a szakirodalomban (Cspe 2006). A jelensget egy kritriumrendszerrel kzeltik meg, amelynek
pontjai a kvetkezk: a) a felnttkorig elhzd olvassi zavar; b) az olvassban s
az rsban/helyesrsban tapasztalhat sajtos hibkkal jellemezhet; c) rkletes jelensg; d) gyakrabban fordul el fiknl, mint lnyoknl; e) norml, illetleg magas
intelligenciahnyadossal jr egytt; f) kizrhat a szociolgiai tnyez, mint pldul a
verblisan ingerszegny krnyezet (Gsy 2005: 387).
A diszlexia s az olvassi nehzsg eltrsnek mibenlte nem egyrtelm, hisz
nem ismernk olyan jl krlrhat olvassi jellegzetessgeket, amelyek a diszlexis
gyermekeket megklnbztetnk a ms okokbl nehezen olvasktl, pldul azoktl,
akiknek problmi a gyakorlat hinyval magyarzhatk (Gsy 1999, Vellutino 1999).
Ltezik azonban egy lehetsg: a Csabay Katalin ltal megalkotott ldiszlexia fogalom
hasznlata. Csabay ugyanis elklnti a valdi diszlexit az ldiszlexitl. Szerinte
elbbinl a gyermek kszlkben van a hiba, utbbinl a gyermeket krlvev
krnyezet mestersgesen idzi el a tneteket (Csabay 1999).
sszegezsl teht elmondhat, hogy a diszlexia olvassi nehzsg, de olyan nehzsg, amelyet nem okozhatnak a gyengn olvask nehzsgeit okoz tnyezk. A
diszlexisok csak olvassi nehzsggel kzdenek, teht nluk nem jelennek meg a
gyengn olvask tneteit okozhat tnyezk, vagyis nem az alacsony intelligencia,
szervi fogyatkossg vagy ms rzkelsi deficit okozza a problmjukat (Gsy 1999).
199
200
201
3. Eredmnyek
A diszlexis dikok tlagosan 620 perc hossz spontn beszdet, a kontrollcsoport
gyermekei tlagosan 840 perc hossz spontn beszdet produkltak. A beszdid
hosszban szignifikns klnbsg van a diszlexis s a nem diszlexis gyermekek
kztt (F(1,19)=11,607; p=0,003). A diszlexis gyermekek teht jval kevesebb kzlst produkltak.
3.1. A KFM-pontszmok
A lejegyzett spontn beszdben elszr meghatroztam a kzlsegysgek fejldsi
mutatjt, majd az adatokat statisztikailag elemeztem. Az eredmnyek szerint (lsd 1.
bra) a KFM-rtkek nem mutattak szignifikns klnbsget a diszlexisok s a nem
diszlexisok beszdben (F(1,19)=1,223; p=0,283).
A diszlexis dikok tlagosan 10,54 (2,43) pontot rtek el. Az eredmnyek azt mutatjk, hogy ngy gyermek beszdnek szintje alig rte el egy iskolba kerl gyermek
beszdnek komplexitst (a nem sztenderdizlt rtk 11), ngyen alig haladtk meg
azt, s csak kt gyermek produklt magasabb pontszmot. A kontrollcsoport tanuli
valamivel jobb eredmnyt rtek el, mint a diszlexis tanulk. A legalacsonyabb KFMpontszmmal, 7,59-el egy tipikus fejlds kisfi rendelkezett. Az alig 8-as pontszm
a jl beszl 4 ves, taln 5 ves gyermekek szintje.
Br a KFM-rtkkel kapcsolatos hipotzisem nem igazoldott, a diszlexis tanulk
kevsb sszetett kzlseket hoztak ltre, mint a kontrollcsoport tagjai. Az albbiakban nhny jellemz plda olvashat a narratvkbl:
Van olyan, amikor azon vesznk ssze, hogy ezt a jtkot most n akarom jtszani vagy , meg hogy hangosan hallgatja a zent, n meg halkan szeretem, meg
202
hogy nem szeret annyira tvzni, n meg szeretek tvzni (a beszl a legmagasabb KFM-pontszmmal rendelkez nem diszlexis kislny volt).
Ht voltunk valahol, gy az ovival? Ja. Ht anyukmkkal le szoktunk menni Balatonra, mert ott van egy ilyen nyaralnk. Na s akkor ott szoktunk nha gy frdni, meg korcsolyzni, hogyha befagy az a kis kikt. gy ht ezeket szoktuk ott
lent csinlni (rszlet egy diszlexis fi beszdbl, aki a msodik legmagasabb
KFM-pontszmot rte el a diszlexisok kztt).
Igen. Mindig nyron elmegynk ilyen ilyen tborba, ahol vannak ilyen csszdk
s nagyon sok csszda. Ne, nem tudom a nevt (a legalacsonyabb KFMpontszm diszlexis fi beszdbl).
Az, hogy ezen a mintn nem igazoldott az els hipotzis, azzal magyarzhat, hogy
nagyok voltak az egyni klnbsgek, illetve a diszlexis dikok egyni fejlesztsben
vettek rszt.
3.2. A tartalmas szavak a beszdben
A diszlexis dikok tlagosan 480 (128) szt ejtettek, a kontrollcsoport tanuli 652-t
(276). A tartalmas szavak szma az sszes szszmhoz hasonlan alakult: a diszlexisok tlagosan 256 szt (55), mg a nem diszlexisok 387 szt (192) ejtettek. A 2.
brn megfigyelhet, hogy a szszm nvekedsvel egyenes arnyossgban n a
tartalmas szavak szma is.
204
205
diszlexis gyermekektl szrmaznak (a zrjelben a nma sznetek idtartama olvashat): Hogy volt egy kutya, s an s volt valami sajt. s s [7783 ms] Igen meg
mg volt vala meg mg volt ja s akkor az a macska megitta. Van ketto barbi, van
ngy vagy ot lovas. Hmm van hmm [9807 ms] ms nin ja mg van ketto singlingi.
4. Kvetkeztetsek
A ngy hipotzisembl mindssze egy bizonyult igaznak. Ez a 4. hipotzis volt,
amelynek bebizonyosodsa igazolta, hogy a diszlexis gyermekek beszdtervezse
sorn tbbszr keletkezik diszharmnia, mint a tipikus fejlds trsaiknl. A gyermekek tbb megakadst produkltak, mint a korban hozzjuk illesztett olvasni tud
trsaik. Nem talltam azonban eltrst a megakadsjelensgek tpus szerinti megoszlsban, a gyermekek hasonl megakadsokat produkltak.
Br az eredmnyek nem mutattak szignifikns klnbsget a diszlexis s a kontrollcsoportban lev gyermekek kztt, tendenciaszer eltrseket talltam. (Hasonlan
Meixner 1993 s HorvthImre 2009 kutatsaihoz; k szignifikns klnbsgeket adatoltak a kt csoport kztt.) Az, hogy az ltalam vizsglt gyermekeknl a hipotzisek
nem igazoldtak, tbb dologgal is magyarzhat. 1. A diszlexis dikok egyni fejlesztsen vesznek rszt. A heti 810 ra egyni s kiscsoportos fejleszts nagymrtkben
javtja a teljestmnyket, gy valsznleg jobban teljestettek, mint egy olyan diszlexis, akit nem fejlesztenek clzottan. 2. A kontrollcsoport tanulinak teljestmnye sem
volt az letkornak megfelel. A gyermekek kzl sokan mg egy korbbi anyanyelvelsajttsi szakaszban jrhatnak. 3. A kis ltszm mellett nagy egyni klnbsgeket
adatoltam.
sszessgben elmondhat, hogy a jelen kutatsban a diszlexisok s a nem diszlexisok beszdben nem voltak jelents eltrsek. Ennek ellenre a tendencik alapjn
kijelenthet, hogy a diszlexia hatssal van a beszdtervezsi folyamatokra. Ennek
pedig slyos kvetkezmnyei lehetnek a dik iskolai teljestmnyre, s gy az egsz
letre nzve is. Mivel a diszlexis dikok az iskolban menteslnek az rsbeli vizsgk all, s csak szban feleltetik ket, egy rossz felels azt a ltszatot is keltheti, hogy
206
a dik nem kszlt. Holott sok esetben valsznsthet, hogy a beszdprodukci elmaradott szintje miatt nem megfelel a felelet sznvonala. Ennek igazolsra mg
tovbbi kutatsok szksgesek.
Irodalom
Csabay Katalin 1999. Az ldiszlexia mint korunk jrvnyveszlye. In Fejleszt pedaggia klnszm. Diszlexia, diszgrfia, diszkalkulia, magatartszavar. Budapest: Nemzeti Tanknyvkiad.
4243.
CspeSzcsOsmann 2000. A fejldsi diszlexira (FDL) jellemz beszdhang-feldolgozsi zavarok eltrsi negativits (EN) korreltumai. In Plh Csaba (szerk.): Magyar Pszicholgiai Szemle.
Budapest: Akadmiai Kiad. LV. 4: 475500.
Cspe Valria 2006. Az olvas agy. Budapest: Akadmiai Kiad. 96190.
Gsy Mria 1999. Az olvassi nehzsg s a diszlexia hatra. In Fejleszt pedaggia klnszm.
Diszlexia, diszgrfia, diszkalkulia, magatartszavar. Budapest: Nemzeti Tanknyvkiad. 3941.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Osiris Kiad. 84116, 307356, 362397.
Grczi Tekla Etelka Gsy Mria Imre Angla 2007. Olvassi nehzsg s diszlexia a beszdfeldolgozs tkrben. In Gsy Mria (szerk.): Beszdszlelsi s beszdmegrtsi zavarok az anyanyelv-elsajttsban. Budapest: Nikol Kkt. 215228.
Herzogn Frst Beta 1996. Kisiskolsok beszdszlelsi s beszdmegrtsi eredmnyeirl. In
Gsy Mria (szerk.): Gyermekkori beszdszlelsi s beszdmegrtsi zavarok. Budapest: Nikol
Kkt. 122142.
Horvth Viktria Imre Angla 2009. A diszlexia tnetei a spontn beszdben. Beszdkutats 2009.
240255.
Imre Angla 2006. Az olvasszavar s a beszdfeldolgozsi folyamatok sszefggsei. Beszdkutats 2006. 160171.
Meixner Ildik 1993. A diszlexia prevenci, reedukci mdszere. Budapest: Brczi Gusztv
Gygypedaggiai Fiskola.
Vellutino, F. 1999. Az olvass egy sajtos zavara, a diszelxia. In Fejleszt pedaggia klnszm.
Diszlexia, diszgrfia, diszkalkulia, magatartszavar. Budapest: Nemzeti Tanknyvkiad. 513.
207
Bevezets
A beszdprodukci sorn a tdbl a toldalkcsvn t kiraml leveg, a gge mdost hatsa utn, a szjregbe rkezik. A ggnek s a toldalkcsnek a hangkpzsben ketts szerepe van: hangforrs s hangmdosts.
A tdbl kiraml leveg fizikai jellemzi mdosulnak a toldalkcsben. A hangszalagok ltal kpzett znge vagy a kiraml leveg a szjregben valamilyen akadlyba tkzik (zr, illetve rs). Amikor a znge vagy a kiraml leveg a rsen keresztlhalad, vagy a zrba tkzik a zrat felpattintja , vagy a zr miatt irnyt vltoztat (pl. az orrreg fel), a szjregben zrej keletkezik. E zngs vagy zngtlen
orlis, nazlis zrejek jellemzik a mssalhangzkat. A mssalhangzk kpzsi helye
vltoz, pldul a nyelv s a fogak, illetleg a fogmedernylvny kztt, a nyelv s a
fogmedernylvny, valamint a kemny szjpad, avagy az ajak s a fogak kztt (Gsy
2004).
Mindezekbl jl lthat, hogy az llcsontok s a fogazat tpusa, mrete, intaktsgnak foka, a fogsorok zrdsnak milyensge jelents hatssal van a beszdprodukci
sikeressgre klns tekintettel a mssalhangzk kpzsekor. A legslyosabb s
legmeghatrozbb vltozsokat a fogak elvesztse, azaz a rszleges vagy teljes foghiny okozza. A fogak elvesztsvel, a fogmedernylvny sorvadsval a nyelv s az
ajak elvesztik artikull prjukat, s az ajkak helyzete, mozgsa megvltozik; a mssalhangzk jelents rsznek kpzse zavart szenved (Fbin et al. 1997).
A szjregnek nemcsak hangkpz, hanem hangmdost funkcija is van. A szjreg mint rezontorreg, mint hangszekrny, trfogatnak, alakjnak megvltozsval
befolysolja a ggben keletkez znge hangsznt, s kialaktja a hangot meghatroz,
arra jellemz felhangcsoportokat (formnsokat). A fogatlanokon a szjreg alakja, kbtartalma megvltozik, mert mdosul a kemny s a lgy rszek arnya, helyzete, valamint a szjreg trfogatt legjobban befolysol nyelv s ajkak helyzete, formja s
funkcija. Emiatt nemcsak a zngs mssalhangzk, hanem valamennyi beszdhang
hangszne megvltozik. Fogatlanoknl teht csaknem valamennyi hang kpzse zavart
szenved, mivel szksgszeren megvltoznak az ajak- s nyelvmozgsok. A beszd
megszokott, tudatalatti, reflektorikus mozgselemei mr nem elegendek a konvencionlis hangkpzshez.
Kiejtsi zavarokat okoz szjregi elvltozsok nem felttlenl kell, hogy ennyire
drasztikus kemnyszvet-vesztesggel jrjanak. Ugyanakkor a fogatlansg sikeres,
funkcionlis rehabilitcija sem megoldhatatlan feladat a korszer fogptlsok alkalmazsval.
Ki kell emelni, hogy nemcsak a kemnyszvetek eltrsei, hanem a lgyrszek anatmiai rendellenessgei, illetve diszfunkcii is mdosthatjk a hangkpzst. Ilyen
208
tpus eltrsek pldul az als s fels ajakfk lenvse, melynek fennllsa esetn a
bilabilis s a labiodentlis hangok kpzse szenvedhet zavart. Hasonl helyzet alakulhat ki a nyelvfk tlzott rvidlse, lenvse miatt, ilyenkor inkbb a perghangok
ejtse neheztett. Mindkt lgyrszelvltozs megoldsa knny, mivel kisebbfajta
szjsebszeti beavatkozssal megoldhat (t kell metszeni). rdemes kiemelni a fels
ajakfk tlzott megvastagodsa miatt kialakul rsessget a kt fels nagymetszfog
kztt. Az llandan fennll, ltalban tbb millimteres rs kedveztlenl befolysolja a labiodentlisok, az alveolrisok s a posztalveolrisok ejtst. Hasonl problmt jelent a viszonylag gyakori, frontfog balesetekbl szrmaz trsek, megpattansok miatt kialakult fogkorona vltoz nagysg hinya. Ma mr ezeknek a helyrelltsa is knnyen, eszttikusan elvgezhet (fogszn tmanyag, kermia korona segtsgvel) s ezzel kik-szblhet az tmeneti rsessg. Tbbszr elfordul, hogy genetikai tnyezk is generlhatnak rses fogazatot (szles fogmedernylvny s ugyanakkor kismret fogak rklse, vagy akr tbbszrs maradfog csrahiny is okozhat
rsessget); ezek megoldsa fogszablyz kezels lehet, azaz fogszablyz kszlkkel ssze lehet zrni a rseket (Surjn et al. 1982).
A fogmedernylvnyok a fogakhoz, a msik llcsonthoz, illetve a nyelv kiterjedshez viszonytott mrete szintn befolysolhatja az artikulcit. Ha tl nagy a mretklnbsg az als s a fels llcsont kztt (brmelyiknek a javra), akkor llandan
meglev, jelents rs lesz kztk, melynek megszntetse a helyes beszdfunkci
elrse cljbl igen nehzkes (a leggyakoribb beszdhangkpzsi zavar ilyen esetekben a diszllia, diszglosszia). A tl szk fels llcsontnak tovbbi, az artikulcis
gesztusokat befolysol kvetkezmnyei lehetnek. Mivel a fels llcsont nemcsak a
szjreg csontos teteje, hanem az orrreg alapja, a nem megfelelen tg, kemny szjpad s a fogmedernylvny potencrozza az orrsvnyferdls htrnyos, a zngekpzsre s a hangsznezetre gyakorolt hatst (orrhangzs beszd). Fontos kiemelni az
ajak- s szjpadhasadkos emberek artikulcis zavarait. Nluk az orrhangzs beszd,
a tbbszrs korrekcis mttek okozta, ersen heges lgyszjpad elgtelen mkdsnek s a nem megfelelen nvekv, rendszerint alulfejlett fels llcsontnak ksznheten jelentkeznek a kiejtsi nehzsgek. Hasadkos embereknl a tbbszrsen
heges garatba torkollik bele a hallkrt (tuba auditiva) is, ami az ilyen embereknl
szintn beszklt, s gyakori kzpflgyulladst s ezltal hallskrosodst is okozhat.
Nluk ppen a percepcis visszacsatols hinya miatt fordulhatnak el beszdszlelsi
zavarok, jellemz lehet a szk szkincs s a hajlam az agrammatizmusra (Surjn et al.
1982). Mindezek msodlagosan hathatnak az artikulcis gesztusok kivitelezsre.
A fogmedernylvny s az llcsontok szerzett mdon is deformldhatnak klnfle balesetek vagy akr rossz szoksok kvetkeztben. Az utbbiak kz tartozik pldul az ujjszops, cumizs, nyelvlkses nyels, ajakszortsos nyels, kizrlagos
szjlgzs. A most emltett tnyezk miatt leggyakrabban n. nyitott haraps alakul
ki, teljesen csukott szj esetn sem rintkeznek bizonyos als s fels fogcsoportok.
Slyossgtl fggen ers selypessg (szigmatizmus) jellemezheti ket, mivel az
alveolris, posztalveolris rshangok s affriktk kpzsekor a nyelvhegy, illetve a
nyelv ells rsze nem a megfelel helyen, hanem annl elrbb helyezkedik el. Megszokjk ugyanis, hogy a fogsorok zrdsi hinyt nyels kzben a nyelvhegy rsbe
nyomsval korrigljk; ers horizontlis mozgs jellemzi a nyelvhegyet a vgfzis209
211
A fogptls behelyezse eltt s utn rgztett beszdben megfigyelhet volt a sziszegk ejtsnek vltozsa (47. bra); a zrtabb oldals nyelvtrnek ksznheten ejtsk knnyebb vlik, rvidl a produkcis idejk, illetve kevesebb mellkzrej regisztrlhat. Mrskldtt a szigmatizmus is. Ugyanakkor a vrakozsoknak megfelelen az affrik-tk ejtse nem vltozott (8., 9. bra), hiszen ez a foghiny nem befolysolta a kemnyszjpad alakjt, illetve annak rintkezst a nyelvvel.
212
213
214
215
A [tsz] hangok izolltabb vltak s kpzsi idejk is lervidlt, ksznheten az oldals fogptls lezrsi kpessgnek. Ugyanakkor nehezti az artikulcit, mivel
jelentsen cskkenti a nyelvteret, s az eddig oldalra benyomul nyelvet vertiklis
irnyba tereli. Ennek megfelelen az affriktk kpzsi helye ideiglenesen glottlis
irnyba mozdul, s az ertlenebb nyelvzr miatt mg a zngssg is gyengbb vlik
tmenetileg (15. s 16. bra). Ezt az artikulcis problmt kedvezen lehet befolysolni a fogsor bels felsznnek megfelel korrekcijval.
216
A [dz] hang intenzitsa cskkent, zrejgcai magasabb frekvenciasvba toldott, valsznstheten a fogptls nyelvtrszkt hatsa miatt. Ilyenkor a nyelv hts s vertiklisan magasabb pozciba szorul a fogptls adaptldsi idejre.
Kvetkeztetsek
sszegzskppen elmondhat, hogy a fogazat, illetve a fogsorzrds mdja, minsge hatssal van a beszdhangok frekvencia- s idviszonyainak vltozsra, a labiodentlis, alveolris s posztalveolris mssalhangzk zrejgcainak frekvencia- s
intenzitsszerkezetre. Ezrt a fogszati beavatkozsok s a fogptlsok tervezse
alkalmval nemcsak a rgfunkci helyrelltst s az eszttikum megtartst, hanem
a zavartalan artikulci intaktsgt is rdemes szem eltt kell tartani.
Irodalomjegyzk
Fbin Tibor Gtz Gyrgy Kan Mikls Szab Imre 1997. A fogptlstan alapjai. Budapest:
Semmelweis Kiad.
Gsy Mria 2004. Fonetika, a beszd tudomnya. Budapest: Osiris Kiad.
Moyers, E. R. 1973. Handbook of Orthodontics. Chicago: Year Book Medical Publishers.
217
Surjn Lszl Frint Tibor (szerk.) 1982. A hangkpzs s zavarai, beszdzavarok. Budapest: Medicina.
218
1. Introduction
1.1. Types of word recognition
Word recognition has been investigated from different aspects. Considering word
recognition in oral language, it is usually associated with the comprehension and analysis of spontaneous speech. Oral language processing requires perceptual analysis,
linguistic and conceptual knowledge and working memory. In decoding the words of
continuous speech, cognitive segmentation is the primary difficulty (Gsy 2005). During the search for the correct word while listening to continuous speech, false lexical
units may also be activated. The cause is mainly perceptual but mismatch may occur at
the level of associations as well (Bna 2004). However, although in oral word recognition listeners need not be aware of the smallest units, the sound segments (phonology),
this ability is a prerequisite to reading via the phonetic route.
In written language, word recognition constitutes a phase of the reading process.
The reading models offer explanations of how readers arrive at word recognition.
Usually two ways are described (Cspe 2005): (1) through the visual analysis of the
word form and (2) via the phonological pathway, in which word recognition is the
result of connecting the elements (graphemes that represent phonemes). This operation
is called blending. In Temples model, blending is the function of the blender, the
connecting unit which carries out the synthesis of letters into a word. In the shallow
orthographic Hungarian language, children achieve word recognition in this way during the process of learning to read. While word recognition in continuous speech requires the mental operation of analysis, the mental operation for reading is the opposite. Therefore, blending is the mental operation of synthesis.
Word recognition is investigated in a third way in current research: the channel is
deemed auditory but the mental operation is deemed synthesis. This type of word
recognition occurs in a frequently used phonological awareness task, namely, blending
(Table 1).
Continuous
speech
Reading
(phonetic route)
Oral blending
of word units
auditory
visual
auditory
The cognitive
segmentation
blending
blending
219
operation to access
the mental lexicon
Required linguistic
knowledge
Phonology
(sequence of
speech sounds)
Morphology
Semantics
Syntax
Phonology
(sequence of
speech sounds)
Morphology
Semantics
Phonology
(sequence of
speech sounds)
Morphology
Semantics
Contextual cues
present
present
absent
WM workload
no backtracking
verification
through
backtracking
no backtracking
220
children showed that syllable blending is mastered by age 5, while phoneme blending
reaches mastery by the end of the first grade (Jordanidisz 2009b).
When children are given a blending task, the following characteristics need to be
considered:
(1) In the input phase:
The acoustic characteristics are different from those of continuous speech. Young
children who are still developing their speech perception lack a distinct mental representation of the phonemes.
(2) In the elaboration phase:
Blending segments increase ambiguity since during the process of synthesizing the
segmented speech sounds, individuals may activate more than one word from the mental lexicon. Therefore, other cognitive functions such as inhibition of irrelevancy or
selecting the relevant cues also have important roles in word recognition.
Further, grammatically marked segments increase the processing load (Gergely
Plh 1994).
The successful blending of sequences of different length depends on the capacity of
the working memory (Sousa 2001). The load on working memory is greater during the
blending of orally presented units than during reading since there is no possibility of
backtracking for verification.
Preschoolers or kindergarten children who lack the understanding that words are
constituted of discrete phonemes are not able to execute mental operations such as
synthesis with a segmented consonant and vowel.
(3) In the output phase:
Children need to present the recognized word verbally. Articulation difficulties may
hinder the correct answer.
1.3. Research questions
The aim of the present research was to examine how Hungarian children in kindergarten and in the first two grades of elementary school are able to blend different lengths
of word units to activate the correct word in their mental lexicon, and what factors
influence this type of word recognition. The research was designed to answer the following questions: (1) How does phonological awareness develop in the area of blending syllables and speech sounds between the ages of 5 and 8 for the task of word
recognition? (2) What causes difficulty in word recognition for words segmented into
syllables or speech sounds? (3) What causes the activation of mismatches during the
blending of syllables and speech sounds? (4) Is there any difference between girls and
boys achievement?
The hypotheses were the following: (1) Word recognition based on phonemes develops mostly from the onset of formal reading instruction. (2) The difficulty level
increases with the number of units, and stem words are easier to recognize than words
with suffixes. (3) Semantic ambiguity can be a cause of mismatch. (4) Since children
with atypical language development were excluded from the research, a significant
difference between girls and boys achievement is not expected.
221
2. Method
2.1. Participants
A total of 90 typically developing children participated in the research with the following age distribution: 30 children were in the age group 56 (average: 5;5), 30 children
in the age group 67 (average: 6;6) and 30 children in the age group 78 (average:
7;6). Children with speech and/or language difficulties were not included in the research. The distribution of gender was 50-50% in each group.
The first age group attended kindergarten, the second age group was divided between kindergarten and first grade students. (The law in Hungary allows children to
start the first grade if their birthday is before 1st June of the year of the onset of primary education. However, parents may decide to hold back their children in kindergarten
for a year if school readiness is questionable. Therefore, it may happen that children
who belong to the first grade age group are still in kindergarten and have not received
formal instruction in graphemephoneme correspondence.)
Children who participated in the research were from the capital city or large country
towns.
2.2. Material
Childrens word recognition was assessed using two subtests of a phonological awareness test (Fonolgiai Tudatossg Teszt) which was developed by Jordanidisz (2009a)
to assess Hungarian childrens phonological awareness development. The first version
of the Hungarian test was the adaptation of the PASS (Barbour et al. 2003) which was
further developed according to the phonetic characteristics of the Hungarian language.
The two subtests detect the ability of syllable and phoneme blending and consist of ten
items each. Five of them are real Hungarian words while five are nonsense words with
the characteristics of Hungarian phonology. In the present investigation the meaningful Hungarian words were used, namely: in the subtest of syllable blending kapu, asztal, dombtl, villamos, viselkedtek (gate, table, from the hill, tram, you behave)
and in the subtest of phoneme blending pld, mos, kapu, stadion, knyvtr (blanket,
wash, gate, stadium, library). The first set includes 3 two-syllable words, 1
three-syllable word and 1 four-syllable word. The first two-syllable word is a root
word with two open syllables; the second is also a root word but with two closed syllables. The third word consists of a root and a grammatical marker indicating movement from a certain place. In this case, the grammatical marker increases the processing load as GergelyPlh explain concerning grammatically marked word segments (1994). The fourth word is a three-syllable root word in which the first two
syllables compose a meaningful word and the third syllable also has meaning itself.
The word is not a compound but an ambiguous stem. In the case of continuous speech,
the context would make backward priming easier after the activation of the mental
lexicon of the three words: villa, mos, villamos (fork, villa, wash and tram).
Without the context, processing this task is more demanding.
222
While the first four words are common nouns, the fifth one is an inflected verb, in
2nd person plural. It contains four syllables. Besides the difficulty of the length and the
grammatical marker, the stem itself may activate another word in the mental lexicon
(wear) which makes the activation of the mental lexicon difficult.
The second set segmented into phonemes, except the first word, where the initial
consonant cluster is pronounced blended. This way, the first two words have three
segments each but the first conveys more information. The third word has four segments while the fourth and the fifth have seven segments. The fourth word starts with
a consonant cluster broken into units. The fifth one is a compound word which may
support the working memory if the first word is activated while listening to the sequence of the speech sounds.
Words were chosen from childrens books. Usually children become acquainted
with them in preschool or kindergarten even if they do not live in cities where certain
words may be more common.
2.3. Procedure
Both phonological awareness subtests were administered individually in a separate
room in the school building. Before the task was given, children were given an explanation of the task, namely: I am going to say words like a robot. Try to understand it
and tell me what the robot said. As practice they solved a similar task with a sample
word before the test items were presented. Childrens answers were recorded on the
test sheet. Results were analyzed by one-way ANOVA and Tukey, PASW Statistics
18. Mismatches were counted and categorized.
3. Results
3.1. Word recognition on the basis of phonological awareness development
Word recognition was assessed on the basis of blending syllables and blending speech
sounds. Blending syllables reaches the 80% mastery level between the ages of 6 and 7.
The 80% criterion level was chosen after Barbour et al. (2003). Other authors use 90%
as the mastery criterion (Chafouleas et al. 1997); however, it is not possible to do this
with subtests containing 5 items. Between the ages of 5 and 6, word recognition on the
basis of syllable blending reached 65%. Between the ages of 6 and 7, the percentage
was 82%, while the 78 year olds could activate the correct words from the mental
lexicon in 97% of trials (Figure 1). Comparing the results of this present research to a
previous one (Jordanidisz 2010) where the 80% mastery level was reached between
the ages of 5 and 6, the conflict is pseudo. In the previous research all the items of the
subtests were used including the nonsense words as well and the task measured the
ability to blend and not word recognition. The mean differences of the present research
were significant (One-way ANOVA: F=20.189, Tukey: p=0.05) (Table 2).
223
Figure 1. Word recognition on the basis of blending syllables and speech sounds
As hypothesized, blending smaller word units, i.e. speech sounds, proved to be a more
difficult task. Less than 20% of the 56 year old children were able to blend a CVC
sequence. This age group achieved a mean of 7.3% in the whole subtest of recognizing
words via blending speech sounds. The second age group showed a significant gain
due to their exposure to formal instruction in literacy: they achieved 30% in this subtest. Findings showed that by the time children learn to read continuously, i.e. between
the ages of 7 and 8, they almost master the skill of blending speech sounds as well.
They were ten times more successful than the first age group with a result of 73%
(Figure 1). The mean difference between the groups was significant (One-way ANOVA: F=50.058, Tukey: p=0.05) (Table 2).
Blending
syllables
Blending
speech
sounds
Mean
Median
Std. deviation
Age 56
0.65
0.69
0.257
Age 67
0.82
0.85
0.185
Age 78
0.97
0.97
0.077
Age 56
0.073
0.038
0.18
Age 67
0.30
0.37
0.25
Age 78
0.73
0.76
0.29
Significance of
mean differences
p=0.05
p=0.05
Since the second age group consists of both kindergarten and first grade elementary
school students, the difference in their achievements was examined. While word
recognition via syllable blending was essentially the same (in kindergarten: 81%, in
first grade: 85%), in phoneme blending first grade students performed three times
224
better than their kindergarten peers (in kindergarten: 19%, in first grade: 57%). This
result also confirms that the exposure to the visual symbols of the phonemes during
reading strengthens the processes of phonemic awareness (Figure 2).
Figure 2. Comparing the achievement of 67 years old kindergarten and first grade students
in word recognition on the basis of phonological awareness
225
Morphology is the other factor that needs to be considered. All the three groups found
it difficult to activate the two-syllable word dombtl (from the hill). This result confirms that morphologically more complex words require a greater effort to activate.
The suffix -tl was recognized by only 60% of 56 year-olds. This indicates that the
mental representation of grammatical markers develops later than that of the root morphemes. The most difficult word not only consisted of four syllables but also contained a grammatical marker, the inflection for 2nd person plural. Since the subtest
contained only this single four-syllable word, the result does not explain the degree of
influence of the grammatical marker.
kapu
dombtl
villamos
viselkedtek
apu
domb t
villa
viselkednek
kamp
domb toll
villa mos
viselkedne
kap
domtl
Vili ma mos
viselked
domb
villa kos
viselkedjetek
toll
viselkedek
viselhettek
viselned kell
visel kend
visel
As for the above mentioned word viselkedtek (you behave) the subtest did not
disclose the contribution of ambiguity into the results. Therefore, the influence of
226
ambiguity in word recognition via blending syllables needs further investigation. Analyzing the mismatches (Table 3), four contributing factors were disclosed: (1) perception, (2) ambiguity, (3) the length of the sequence, and (4) the presence of grammatical
markers. The mismatches of kapu (gate) are clearly perceptual. Interestingly, there
were fewer errors for the same word when it was segmented into speech sounds. An
interesting result is that the second age group made more errors in this task than the
first age group. The reason behind this phenomenon should be further investigated.
The word asztal (table) did not have mismatches. The third item, dombtl generated
four mismatches altogether. The difficulty in activating the correct grammatical marker can be detected in each case. The activation of toll (pen) and t (lake) may also
be due to perceptual problems. However, it is important to note that these mismatches
are root words and nouns. In the activation of the mismatches for villamos (tram),
ambiguity can be detected in all the five cases. Ambiguity alone contributed to a mismatch in the case of villa mos (fork wash), while in the case of villa kos (fork goat),
perceptual influence was also present. The activation of villa was influenced by ambiguity in combination with the length of the sequence. The activation of Vili ma mos
(Vili today wash) was probably the result of the combined factors of the length of the
sequence, the ambiguity and perception. Out of the twelve mismatches of viselkedtek
four may originate in ambiguity besides other factors such as the length of the sequence or perception: visel, viselned kell, viselhettek and visel kend. Since the grammatical marker is located at the end of the long sequence, its omission or substitution
with another grammatical marker may be the complex influence of the load on working memory and the difficulty activating the grammatical marker in the mental lexicon.
Mismatches during blending phonemes happened in fewer cases. In the cases of
stadion (stadium) and knyvtr (library) the mismatches were nonsense words
based on perception. In the cases of pld (blanket) and kapu (gate) mismatches
were also perceptual (Table 4).
pld
eld
kapu
1
stadion
knyvtr
kap
sadion
kap
sadid
ki apu
stadon
kab
stad
kt
stapud
knytr
227
Blending
syllables
Blending
speech
sounds
Mean
Std. deviation
56 y.old boys
0.65
0.245
56 y.old girls
0.68
0.248
67 y.old boys
0.79
0.22
67 y.old girls
0.87
0.123
78 y.old boys
0.97
0.07
78 y.old girls
0.96
0.083
56 y.old boys
0.08
0.182
56 y.old girls
0.067
0.180
67 y.old boys
0.37
0.31
67 y.old girls
0.30
0.26
78 y.old boys
0.70
0.301
78 y.old girls
0.76
0.216
0.088
0.768
1.508
0.230
0.226
0.638
0.041
0.842
0.406
0.529
0.310
0.582
228
However, examining the non-significant differences, results show that girls word
recognition at the phonological level was somewhat better in the first two age groups
(68% vs. 65% at age 56; 87% vs.79% at age 67). The difference had diminished by
the age of 78 (Figure 5).
In the case of the phonemic level, achievement showed contrary results: boys word
recognition was better in the first two age groups (8% vs.6.7% at age 56; 37% vs.
30% at age 67), while girls overtook the boys by the age of 78 (76% vs. 70.%) (Figure 6).
229
Figure 6. Gender based comparison in word recognition via speech sound blending
Since the phonological level of phonological awareness develops earlier than the phonemic level, it seems that girls early language development matures somewhat earlier
than that of the boys. Further, they respond to the direct instruction of graphemephoneme correspondence to a greater extent than boys. On the other hand, it seems
that boys perform better from an earlier age in the more complex task when working
memory tasks and mental operation tasks (synthesis) are combined. However, they
seem to respond more slowly to formal instruction in literacy.
4. Conclusions
The research results answered the questions and confirmed the hypotheses. Word
recognition on the phonemic level develops mostly from the onset of formal reading
instructions. Word recognition is difficult at syllable level if (1) children are between
the ages of 5 and 6 and when words contain grammatical markers or more than two
syllables or are semantically ambiguous; (2) children are between the ages of 6 and 7
and words with suffixes have more than three syllables. Word recognition is difficult
at the phoneme level if (1) children are younger than age 6, (2) children are between 6
and 7 and the word has more than three phonemes and (3) children are between 7 and
8 and the word more than four phonemes. In order to detect the degree of the influence
of the different factors (grammatical marker, the length of the sequence, ambiguity and
perception) in the process of word recognition via syllable blending, a more comprehensive test should be designed to meet this specific purpose.
Considering the gender factor in blending tasks, there was no significant difference
between the achievement of boys and girls. The explanation regarding the minor differences under 3.3 should be explored by specifically designed further research.
The size of this research sample was not sufficient to draw reliable conclusions;
therefore, further research should be carried out with a satisfactorily large number of
participants. The information available from the results can be useful in the pre-phase
of reading instruction and in the preparation of kindergarten children for successful
literacy acquisition.
230
References
Barbour, K. Keafer, K. Scott, K. 2003. Sounds of Speech. Phonological Processing activities.
Norfolk: NILD.
Bna Judit. 2004. A beszdszlelsi folyamat megakadsi. In Gsy Mria (szerk.): Beszdkutats
2004. Budapest: MTA Nyelvtudomnyi Intzete.
Chafouleas, S. M. Lewandowsic, L. J. Smith, C. R. Blachman, B. 1997. Phonological Awareness Skills in Children: Examining Performance across Tasks and Ages. Journal of Psychoeducational Assessment. 15/4: 334347.
Chard, D. J. Dickson, Sh. V. 1999. Phonological Awareness Instructional and Assessment Guidelines. Intervention in School and Clinic. 34/5: 261270.
Cspe Valria 2005. Kognitv fejlds-neuropszicholgia Budapest: Gondolat Kiad.
Cspe Valria 2006. Az olvas agy. Budapest: Akadmia Kiad.
Gergely, Gy. Plh, Cs. 1994. Lexical processing in an agglutinative language and the organization
of the lexicon. Folia Linguistica. 28: 175204.
Gsy Mria. 2005. Pszicholingvisztika. Budapest: Osiris Kiad.
Gsy Mria 1995/2006. GMP-Diagnosztika. A beszdszlels s a beszdmegrts folyamatnak
vizsglata. Budapest: Nikol Kkt.
Goswami, U. 2003. Phonology, Learning to Read and Dyslexia: A Cross-Linguistic Analysis. In
Cspe, V. (ed.): Dyslexia. Different Brain, Different Behaviour. New York: Kluver Academic.
Jordanidisz gnes 2009a. A fonolgiai tudatossg s az olvasstanuls kapcsolata. III. Alkalmazott
Nyelvszeti Doktorandusz Konferencia online ktete
http://www.nytud.hu/alknyelvdok09/proceedings.pdf
Jordanidisz gnes 2009b. A fonolgiai tudatossg fejldse az olvasstanuls idszakban. Anyanyelv-pedaggia. 2009/4. http://www.anyp.hu/cikkek
Jordanidisz gnes 2010. A gyermekek fonolgiai tudatossgnak fejldse 46 ves korban.
http://www.nytud.hu/alknyelvdok10/absz/jordanidisz.pdf
Lerner, J. 2003. Learning Disabilities. Theories, Diagnosis, and Teaching Strategies. Boston: Houghton
Mifflin Company.
Sousa, D. A. 2001. How the brain learns. Thousand Oaks: Corwin Press, Inc.
Zgonc, Y. 2000. Sounds in Action: Phonological Awareness Activities and Assessment. Washington
D.C.: ERIC Clearinghouse.
231
1. Bevezets
A hagyomnyos paraszti trsadalomban az ismeretek, vilgkpek s a kzs kulturlis
hagyomny tadsa a nemzedkek kztt a beszlt nyelv segtsgvel trtnt. A szjhagyomny tjn tadott informcik egyszerre tartalmaztak vltoz (az elbeszl
elkpzelseinek megfelel), illetve a hagyomny ltal konzervlt lland elemeket. A
npmesk, hatalmas mfaji vltozatossgukkal, igen alkalmasak voltak arra, hogy
egyszerre tantsanak, elgondolkodtassanak, formljk a vilgkpet, s nem utols sorban szrakoztassanak, s kiemeljk az egynt a htkznapi nehzsgek sodrsbl. A
mese olyan llandan alakul s vltoz mfaj, amely minden idben a kor szintjn,
az emberek vilgltsnak megfelelen tolmcsolja az si mondanivalt egy, a kzssg ltal elfogadott s megbecslt egyn (a mesemond) segtsgvel. A szjhagyomnyozs (a folklrszveg) egyik fontos jellemzje, hogy a rsztvevk egyfajta kzs
emlkezet, kzs lmnyek birtokban szlnak egymshoz (v. D. Tancz 2002).
A mesk vizsglatban fontos krdskr az emlkezet (egyni s kzssgi), az
emberi memria mkdse, hiszen a szveg akkor ri el tnylegesen a cljt, vagyis
kpes elgondolkodtatni, a tanulsg levonsra ksztetni a hallgatt, ha az minden rszletvel egytt fel tudja dolgozni, meg tudja rteni a mest. A jelen dolgozatban a meseszveg percepcijnak mkdst trjuk fel egy rgi (60 vvel ezeltt elmondott)
mese segtsgvel. Elemezzk a szvegrts s a szvegfelidzs folyamatt mai hallgatknl. A kt folyamat kzl a szveg megrtse zrt, statikus tevkenysg, mg a
felidzs konstruktv, szvegltrehoz folyamat.
A narrci npi vltozatnak felidzse sem pusztn reprodukci, hanem rtelmez,
rtkel s magyarz mozzanatokat is tartalmaz. A konstruls nagymrtkben fgg
az egyn pillanatnyi tudstl (kznapi tudstl, mindennapi ismereteitl, hiedelemrendszertl), lnyegben ez az ismeretrendszer, lexikai sztr hatrozza meg, hogy a
mesei elbeszls megalkotsakor milyen kifejezseket, szavakat hasznl az alkot
(D. Tancz 2002). Az emberi emlkezet az ismers, elfogadhat s megrthet motvumokat, elemeket rzi meg legknnyebben, mg azokat, amelyek tvol llnak az
adott kultra vilgkptl, hamar kirostlja. A tarts hasznlat, a fenntartott trsadalmi
gyakorlat permanensen megjtja az emlkezs folyamatt (Plh 1986).
A mondatokra s a szvegekre val emlkezs, klnsen a felidzs (recall), a lehvs folyamata, alapveten szemantikai jelleg: a logikai terminusokban, propozcikban elraktrozott jelentsekre emlkeznk (van Dijk 1976). A szvegnyelvszet,
diskurzuselemzs kzelebb visz a narratvk ltrehozsnak, mkdsi folyamatainak megrtshez. Ms megkzeltsek szerint a trtnetek feldolgozsa s megr6
232
233
2. Anyag s mdszer
A ksrleti anyag egy archaikus, bukovinai szkely npmese volt, amelynek cme a
Kt koma, elmondta Palk Jzsefn 1952-ben, Kakasdon (Tolna megye). A felvtel
a Hegeds-archvumban tallhat (v. NiklczyHorvth 2007). A mese motvumait
tekintve halotti virrasztsokon meslt ksrtettrtnet, ers erklcsi s oktat jelleggel.
Maga a trtnet nem tl sszetett, lineris esemnyszerkezettel, egyetlen hromszor
ismtld trtnssel, htkznapi s termszetfeletti elemekkel rendelkezik. Rviden
sszefoglalva: kt koma sszevsz a kocsmban, tbbet szba nem llnak egymssal.
Az egyik megbetegszik, haldoklik, de hogy gynhasson, ki kell bklnie a komjval.
A koma eljn, kibklnek, a beteg meghal. Temetni akarjk a halottat, de feltmad,
s kzli, hogy mg nincs eltlve, vrjanak. (Ez az elem hromszor ismtldik.) A
halott vgl kzli, hogy eltltk, de el is krhozott, mert nem gynt szintn, s nem
bklt ki a komval. Felkri a felesget, hogy ne imdkozzon rte, mert ezzel csak
nveli a knjait.
A szveget hangzshen lejegyeztk, majd Praat 5.2 programmal elemeztk. Megllaptottuk az artikulcis s a beszdtemp rtkt, a sznetek jellemzit, adatoltuk a
megakadsokat. A mese teljes idtartama 6 perc 52, az artikulcis temp 12,8
hang/s, a beszdtemp pedig 10,8 hang/s volt. A sznetek arnya a beszdjelhez viszonytva 27%, sszesen 121 db nma sznetet talltunk, tlaguk 771 ms. A mesben
25 db megakadsjelensg fordult el, ezeknek 60% tltelksz (aszongya, ht), 20%
hangnyjts, emellett megjelent mg tves indts mdostssal, illetve jraindts is.
A mesben kitlttt sznet (hezitci) nem volt (a Hegeds-archvum beszlire vonatkozan v. GsyGyarmathy 2009).
A percepcis ksrletben sszesen 50 magyar szakos egyetemi hallgat vett rszt,
36 n s 14 frfi, letkoruk 19 s 25 v kztti. A szveget egyms utn ktszer hallgattk meg. A rsztvevk kzl 25-en 10 megrtst ellenrz krdsre vlaszoltak,
amelyek a cselekmnyekre, ok-okozati sszefggsekre vagy kvetkeztetsekre vonatkoztak (pl. Mi trtnt a kt komval a kocsmban? vagy Mit mondott a halott, mirt nem temethetik el?). A msik csoport, szintn 25 f, azt a feladatot kapta, hogy
rviden foglalja ssze a szveg tartalmt rsban. A kapott vlaszokat sszestettk,
megllaptottuk a helyes s a tves vlaszok arnyt, a szvegrekonstrukciknl pedig
az eredeti szvegben lv kulcsmomentumokat vetettk ssze a lert szvegvltozatokkal, illetve elemeztk a hibs vlaszok tpusait s lehetsges okait. Az adatokat
egytnyezs varianciaanalzissel s ler statisztikval ellenriztk.
3. Eredmnyek
3.1. Szvegrtsi ksrlet
A szvegrtsre kapott tlagos eredmny 84,4% volt, vagyis a ksrletben rszt vev
szemlyek 1 vagy 2 krdsre nem tudtak helyesen felelni. Ez alapjn megllapthat,
hogy ennek az archaikus szvegnek a megrtse nem okozott tl nagy gondot a mai
fiatal hallgatknak, kvetni tudtk a legfontosabb momentumokat, megrtettk az okokozati sszefggseket, megjegyeztk a rszleteket. Ha megnzzk az elrt eredm234
nyek szzalkos megoszlst (1. bra), lthatjuk, hogy a legtbben 100%-os, hibtlan
teljestmnyt nyjtottak (48%), illetve tovbbi 24% csak egy krdsre nem vlaszolt
helyesen, vagyis 72% szmra a szveg feldolgozsa egyltaln nem bizonyult nehznek. Kt-hrom krdsre nem tudott helyesen vlaszolni a ksrleti szemlyek 16%,
kifejezetten rossz teljestmnyt nyjtott, csak 1-3 krdsre vlaszolva 12%. Tekintettel
arra, hogy a vizsglatban magyar szakos egyetemi hallgatk vettek rszt, akiknl a
figyelemhiny nem valsznsthet, az emltett eredmnyeket a szvegrts deficitje
okozhatja.
235
kulturlis jratlansg mint ok, a hallgatk egy rsze nem tudta, mi az utols tlet s a
krhozat fogalma. Ez az ismerethiny tbb megrtsi s rtelmezsi gondot is okoz.
236
dat tnyeleges nehzsgt. Msrszrl itt nem tallunk olyan szlssges megrtsi
problmkat, mint a szvegrtsi feladatnl, ahol a legrosszabb eredmny alig 10%
volt, a rekonstrukcinl a leggyengbben teljest hallgatknak is sikerlt visszaemlkeznik a trtnet fontos mozzanatainak legalbb a felre.
sszestettk s tpus szerint osztlyoztuk a tvesen felidzett vlaszokat is. Megklnbztettnk hinyz, a mesben nem szerepl, illetve tvesen felidzett informcikat. A hinyz informcik a hibk alig 16%-t tettk ki, vagyis a rsztvevk az
esetek tbbsgben igyekeztek valamilyen mdon informcikkal, esemnyekkel,
interpretcikkal kitlteni a szvegben keletkez rt, s egy koherens trtnetet
ltrehozni. A hinyz adatok kztt elfordult ok-okozati sszefggs (a feltmads,
illetve a krhozat oka), tovbb magyarz szveg mirt ne imdkozzanak a halottrt (a mese vge).
A szvegben nem szerepl informcik betoldsa leginkbb a mese kisznezst jelentette: a felesg jajgatott, szrnylkdtt (amikor a halott feltmadt); ltogathat
volt a halott (ehelyett a kifejezs a nyjtpadra tettk, elnyjtottk szerepel a mesben); a halott rszletes vallomst tett; azt mondta, neki mr mindegy. Ez a hibatpus az
sszes tves vlasz 16%-t kpezte.
A tves informcik alkottk a hibzsok legnagyobb csoportjt (68%), itt tbb tpust is megklnbztettnk. A flrehallsra vagy beszdszlelsi problmra utal
tvesztsek viszonylag ritkn fordultak el, ezek oka leginkbb a hangfelvtel minsgben, illetve a szveg archaikus voltban keresend, br a tnyleges percepcis
problmk sem zrhatk ki. Ilyenek voltak az v/j, aludni/llani vagy a halott nem
hlt mg ki. Az id- s trbeli tvesztsek mr a szveg sorrendisgnek tves feldolgozst mutatjk pldul a halott a virraszts alatt szlalt meg (a temets helyett), a
halott a srban volt bds (a nyjtpad helyett) vagy senkit ne temessenek el gy, hogy
haragban van (senki ne haljon meg gy, hogy haragban van).
A jelentsbeli tvesztsek olyan esetben fordultak el, amikor a hallgatk nem voltak tisztban a sz vagy kifejezs pontos rtelmvel, ezrt szmukra rtelmesnek tn
kifejezsekkel prbltk azt helyettesteni. Ilyen volt a meg van tkozva (helyesen: el
van krhozva), a megldotta (megldoztatta helyett) vagy a feltette (feladta) az utols
kenetet. Jellemz, hogy ezek a kifejezsek, mind a keresztny kultra halotti szertart-
237
azonban az a tny, hogy a vizsglt mese esetben a kzps szakasz hromszor is megismtldik, gyakorlatilag ugyanazokkal a szfordulatokkal, vagyis az ott szerepl informcik semmikppen sem kerlhettk el a hallgatk figyelmt. A hibatpusok elemzse azt mutatta, hogy a mai fiatal felntteknek csak igen felletes ismereteik vannak a
keresztny halotti szertartsok s a tlvilgi let trgykrbl. Az olyan fogalmak pontos jelentse, mint az utols kenet, az ldoztats, gyns, utols tlet, elkrhozs nem
igazn ismert a szmukra, gy ez a valsznbb ok a szveg kzpn tallhat informcik tves visszaidzsre. Arra is kerestk a vlaszt, fenntarthat-e a hallgat figyelme
egy hosszabb hallott szveg feldolgozsakor. A ksrleti anyagknt szolgl mese felptse, az ismtld momentumok, illetve a mesl eladsmdja mind elsegtette a
figyelem fenntartst, a 80% feletti szvegrtsi eredmnyek pedig objektve is igazoljk azt, hogy a rsztvevk vgig figyelemmel ksrtk a szveget, illetve a helyes kvetkeztetseket is le tudtk vonni.
A kutatsban kt hipotzist fogalmaztunk meg, amelyeket adataink igazoltak. A szveg felidzse valban nehezebbnek bizonyult, mint a krdsekkel segtett megrts (a
hibtlan vlaszok arnya a szvegrtsben 48%, mg a felidzsben csak 20%), azonban
az eltrs az tlagok szintjn nem jelents, s statisztikailag sem szignifikns. A kulturlis klnbsgek, amelyeket a mai s a 60 vvel ezeltti hallgatk kztt feltteleztnk,
egyrtelmen fennllnak. Br a halotti szertartsok a hagyomnyos falusi trsadalom
legkonzervatvabb szoksai kz tartoznak, a megvltozott letkrlmnyek (a halott
szemlyek eltnse a htkznapokbl) a hallhoz val viszonyt s az azzal kapcsolatos
szertartsokat is megvltoztattk. A mai fiatalok mr csak felletesen ismerik a keresztny hagyomnyokat, ez szkincskben s vgs soron a szvegfeldolgozsi stratgijukban is tetten rhet (viszont sokkal tbb, ms jelleg ismerettel rendelkeznek, mint a
mest elmond, rstudatlan adatkzl). Az letmdvltozs a kultra s kulturlis ismeretek vltozst is maga utn vonja, ezek pedig hatnak a szkincs sszettelre, s vgs
soron az anyanyelv hasznlatra is. Azt, amit nem ismernk, nehezebben is rtjk. Mint
azt a bevezetsben emltettk, az emberi emlkezet az ismers, elfogadhat s megrthet elemeket rzi meg legknnyebben, mg azok, amelyek tvol llnak kulturlis vilgkptl, hamar trldnek (D. Tancz 2002). gy tnk, a rgi meseszvegek mai hallgatk
ltali percepcijra is ez vonatkozik.
Irodalom
Broek, P. van den 1994. Comprehension and memory of narrative texts: inferences and coherence. In
Gernsbacher, M. A. (ed.): Handbook of Psycholinguistics. San Diego: Academic Press. 539588.
D. Tancz Tnde 2002. j paradigmk a nevelstrtnetei kutatsban: a npmese mint a szocilis
reprezentci adekvt mfaja. Valsg. 9: 4254.
Dijk, T. A. van 1976. Narrative macro-structures. Logical and cognitive foundations. Journal for
Descriptive Poetics and Theory of Literature. 1: 547568.
Dijk, T. A. van 1979. Recalling and Summarizing Complex Discourse. In Burghardt, W. Holker,
K. (eds.): Textverarbeitung / Text Processing. BerlinNew York: de Gruyter. 49118.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Osiris Kiad.
Gsy Mria Gyarmathy Dorottya 2008. A nyelvhasznlati vltozs egy jelensge. Magyar Nyelvr.
132: 206222.
239
240
1. Bevezets
Az rt olvass az egyn boldogulsnak s az eslyegyenlsg megteremtsnek
felttele az informcis trsadalomban. Az olvassi tevkenysg lnyege a szveg
megrtse (Adamikn 2008).
A megrts a grafmk s a fonmk megfeleltetst kvet folyamat, amelyben a
szegmentlt sz morfolgiai szerkezetnek felismerse s azonostsa trtnik (Gsy
1996). Megrtskor egy sszetett gondolkodsi folyamat jn ltre az elmben trolt s
a szvegben bennfoglalt j informcik sszevetsvel, valamint a hinyz smk
beptsvel (Adamikn 2008). A megrts teht fgg attl, hogy a diskurzus s az
olvas tudsreprezentcija kztt mekkora az tfeds. A megrts hierarchikusan
szervezd folyamat, amelyben visszahat folyamatok is rvnyeslnek, ennek pldja, amikor a gyenge olvas az ismeretlen szt a kontextushats kvetkeztben kpes
kikvetkeztetni (Cs. Czachesz 1998). Adamikn Jsz Anna (2003) a megrtsnek
ngy fokozatt klnti el: beszl az olvasottak sz szerinti megrtsrl mintegy a
denotatv jelents hozzrendelsrl, az rtelmez, interpretl megrtsrl, amely
mr magba foglalja a szvegben impliklt jelentsek azonostst is. A megrts
legmagasabb fokozata a szveget tovbbfejleszt kritikai olvass, amely lehetv teszi
a szvegre val reflektlst.
A megrtshez az olvasnak megfelel nyelvi tuds birtokban kell lennie szksge van fonolgiai, szintaktikai s szemantikai tudatossgra (Tth 2002). A nyelvi
tudatossg mellett kpesnek kell lennie az olvasnak a kvetkeztetsek levonsra, a
szvegszerkezet felismersre (gy a lnyeg kiemelsre, az alapgondolatok felfedezsre, a szereplk szndkainak s cselekvseinek felismersre), valamint sajt megrtsi folyamatnak monitorozsra (Gsy 2005).
A XX. szzad msodik feltl, az olvasskutats fellendlsvel prhuzamosan
vagy annak kvetkeztben egyre fontosabb krdss vlt vilgszerte, hogy a dikok
klnbz letkori szakaszokban hogyan rtik meg, megrtik-e az olvasott szvegeket. A krdsre klnbz hazai s nemzetkzi kutatsok igyekeztek vlaszokat adni.
Az els nagy fordulat az IEA 197071-es felmrse, amelynek kvetkezmnye lett az
a kzhely, miszerint a magyar dikok nem tudnak olvasni, miutn az eurpai mezny
vgn helyezkedtek el (Cs. Czachesz 2001). A kvetkez sokat reflektlt vizsglat a
2000-es PISA-felmrs, amelyet sokknt emlegetnek. Az OECD 2003-as mrsben
enyhe javuls, 2006-ban stagnls lthat, a legutbbi, 2009-es mrs azonban jobb
kpet, javulst mutat a feladatok megoldsban. A 910 ves dikokat vizsgl PIRLS
felmrs sorn j pozcit foglaltak el a magyar dikok. Gsy Mria (2001) tbb letkori csoportban vgzett szvegrtsi vizsglatai azt mutatjk, hogy az letkorban elvrt szinttl elmarads regisztrlhat. Egyre tbben fogalmazzk meg, hogy a szvegrts fejlesztsnek egyik tja az olvassi stratgik tantsa lehet.
241
A mr emltett negyedikeseket mr nemzetkzi olvassvizsglat az olvass defincijba is beemeli az olvassi stratgikat, az idealizlt olvas eszkzeknt (Tth
2009). A szakirodalom ktfle jelentsben trgyalja az olvassstratgia kifejezst.
Meghatrozza egyfell szvegkezelsi s olvassi mdknt, msfell a jelentsteremtsben meghatroz szerep kognitv s metakognitv folyamatokat rti alatta. A
metakognci fogalmnak els hasznlja Joh Flawell volt, aki a kognitv kpessgrl alkotott tuds, valamint azok megismersesnek jelentsben alkalmazta azt
(JzsaSteklcs 2010). A metakognitv tudst kt nagy terletre osztjk, a procedurlis metatudsra, amely az emberi s a sajt tuds ismeretre vonatkozik, valamint deklaratv metatudsra, amely a sajt tuds nyomonkvetsi s ellenrzsi folyamatait
takarja. A metakognitv tuds funkcija, hogy mdostsa s ellenrizze az alacsonyabb
dekdolsi folyamatokat (Cskos 2008). Az olvas minl fejlettebb metakognitv tudssal rendelkezik, annl jobb az olvasott szvegrtse (JzsaSteklcs 2010). A gyakorlott olvast a kezd olvastl a dekdols automatizldsa mellett megklnbzteti a szvegek feldolgozsnak stratgiai szintje is (Cskos 2007). Tth Beatrix (2009)
a Harvey s Goudvis ltal meghatrozott j olvast jellemz stratgikat mutatja be.
Ezek kztt megtallhat a httrtuds s a szvegben olvashat informcik kztti
kapcsolatok keresse teht a trolt smk felhasznlsa, illetve a hinyok kiegsztse. Megtallhat itt is a monitorozs; a szveg lnyeges gondolatainak megfogalmazsa olvass kzben s azt kveten; a lnyeg kiemelse; az olvassi nehzsg bekvetkezsekor javt stratgik hasznlata; a szerz nzpontjnak s cljainak feldertse,
s a krdseire trtn vlasz adsa; illetve a szereplkrl, a cselekmnyrl, a helysznekrl (stb.) rzkletes bels kp megalkotsa.
Pressley s Afferbach a gyakorlott olvask ltal hasznlt stratgikat gyjtttk szsze. Megllaptjk, hogy az rett olvask az egsz olvass folyamatt koordinljk, a
szvegre rzelmi vlaszt adnak, tudatban vannak, hogy mit s milyen cllal olvasnak,
birtokban vannak olyan stratgiknak, amelyekkel megoldhatjk nehzsgeiket (Kelemen-Molitorisz 2009). Az olvassi stratgik egyik krdves offline vizsglatban a
stratgikat hrom nagyobb kategriba rendszerezik: megklnbztetnek tfog olvassi stratgikat (pldul a gondosabb olvasst ignyl s mellzhet rszek felismerse, a httrtuds alkalmazsa, a tudatos olvassi cl), olvasst tmogat stratgikat (ilyenek a fontos informci technikai/vizulis kiemelse, a jegyzetels, az olvasottak sszefoglalsa), valamint problmamegold stratgikat (ide tartozik a tudatos
jraolvass, az olvassi sebessg szablyozsa).
A jelen vizsglatban kt kzpiskolai csoport stratgiahasznlatt mrem olvasott
szvegrtsi tesztben. Kzponti szerepet kap az nmonitorozs, az alapgondolatok
sszefoglalsa, a lnyeg kiemelse, a httrtuds alkalmazsa, a tudatos olvassi cl s
az olvass kzben fellp nehzsgek problmamegoldsi stratgija. Hipotzisem,
hogy a kt letkori csoportban jelents klnbsgek lesznek. Felttelezem tovbb,
hogy ezek a klnbsgek a stratgai gyakorlottsgbl erednek.
2. Anyag, mdszer, ksrleti szemlyek
A vizsglatban rszt vev szemlyek egy kisvrosi gimnzium kilenc s tizenkettedik
osztlyos dikjai, 1415, valamint 1718 vesek. A kt osztlyban kzs, hogy egy
242
243
A vers humorforrsnak megfogalmazsa jelentette a legtbb problmt. A kt csoportnl a megegyez teljestmny hasonl nehzsgeket mutatott annak ellenre, hogy
a tizenveseknl mr megjelenik a mlyszerkezeti humor elnyben rszestse, 14
ves korban pedig megnvekszik a verblis humor irnti fogkonysg. A humor megrtse fokozott gondolkodst kvetel meg a befogadtl. A humorrzket ugyanis
olyan mentlis kpessgknt fogalmazzk meg, amely lehetv teszi a komikum ltrehozst s befogadst (Lendvai 1996).
Ebben a feladatban a legtbb problma az idsebbik korosztly lnytanulinl volt
tapasztalhat, a csoport tlageredmnye 20%. A megoldsok elemzsekor tbb tipikus
gondolkodsi irny volt rzkelhet, amelyek az elvrt megoldstl eltrtek. Ilyen az
antropomorfizlsra utals a humor forrsaknt. Ennek pldja: A papagj emberi
tulajdonsgokkal lett felruhzva, trsakat toboroz, lzad.
Gyakori a humor indoklsa verstani, stilisztikai jellemzkkel: Ironikusan mesl el
egy egyszer trtnetet, emellett rdekes jelzket hasznl, amelyek csak fokozzk a
humorossgot. A sorok vgn rmelnek a szavak.
Elfordult a vers flrertelmezse is, de tbben felismertk a csattant, s 14%-ban
az elvrt vlaszt adtk. Egy plda a legjobbnak vlt megoldsok kzl: A papagjok
csak ismtlik azt, amit hallnak, s amit gyakran hallnak, azt meg is jegyzik. A humor
alapja az, hogy a papagj beszlni akar a rabtartkrl, de ezt nem tudja elmondani,
mert csak a Gyurrika szp! mondatot ismeri.
A reflexikban 27%-ban rtkeltk ezt a legtbb nehzsget okoz feladatnak. Kifejeztk, hogy nehzsget jelentett a gondolataik megfogalmazsa, vagy a versek
elemzse nem ll kzel hozzjuk, vagy pedig a humornak ms jelentstartalmat tulajdontva inkbb a vicc mfajt jellemz eszkzket prbltk a versben keresni.
A megfelel hirdets kivlasztsa s az arra val jelentkezs bizonyult a legeredmnyesebb feladatnak, a kt csoport teljestmnynek nagyfok eltrse ellenre is. Az
letkor mellett a nem is meghatroznak bizonyult. Az tlag 79%, ezen bell a vlaszts 77%-os, a szvegalkots pedig 80%-os eredmnyt mutat. A megoldsok teret adtak
a tanulk kreativitsnak, kivl rvekkel bizonytottk az lls betltsre val alkalmassgot. A megoldsoknl a hibk egy rsze abbl kvetkezett, hogy tbben nem
244
245
247
III. KTNYELVSG
BILINGUALISM
248
249
Felhasznltunk flig-strukturlt interjkat, melyeket Krptalja 70 magyarlakta teleplsn ksztettnk. Tovbb ksztettnk hanganyagokat: a nyelvvlasztsi krdvek kitltse alatt, rgztettnk ktetlen csoportos beszlgetseket, spontn hangfelvteleket, sajttermkekbl s szpirodalmi alkotsokbl is gyjtttnk anyagot, amelyek szintn lbeszdbl szrmaznak, vagy azt prbljk meg visszaadni, tovbb
rott beszlt nyelvi pldkat is gyjtttnk (online nyelvhasznlat chatszobkban),
melyek szintn kellkppen informlisak s spontnok, hogy ltaluk rdemben lehessen vizsglni a kdvlts (CS) jelensgt.
Beszlt nyelvi korpusz
rott korpusz
a krptaljai magyar
hanganyagtr
interjibl
53
szpirodalom
krdves felmrs
kzben rgztett s
spontn felvtelek,
megfigyelsek
20
sajt
rtekezlet
hanganyagban
28
sszesen:
101
Mindsszesen:
rott-beszlt korpusz
148
14
162
Skype (chatszoba)
Kzssgi oldalak:
facebook.hu; iwiw.hu
40
49
312 megnyilatkozs
1. tblzat. A kdvlts-korpusz megoszlsa
250
hasznlnak azonos jelentsben s kiejtssel: a dumlj, magyar duml (beszl) felszlt md alak, s az orosz/ukrn gondolkodik jelents sz ejtsi hasonlsgt
hasznljk ki. rott-beszlt nyelvrl lvn sz, a cirill bets rsmd megersti a kdvlts szndkossgt, ezzel jelzi a beszl pragmatikai tudatossgt.
BZ: az elbb rohangltam kicsit
MA: akkor mg dumlok,
BZ: aztn kerestem 1 embert
BZ: de nincs a hleyn
BZ: ,
MA: :-)
Spontn beszlgets
V1: Ez mi?
E1: Halva. [Trkmz, dessg]
V2: Az micsoda? lve nem lehet?
E2: !
[Adjl nekik megkstolni!]
Mindkt pldaanyagban fonolgiai s szemantikai tudatossgrl beszlhetnk, ugyanakkor ezt nyelvi jtkknt kihasznljk a beszlk.
b) Morfolgiai szint
A lexmaszint vltsok (angolul: fused lects; Matras 1996) grammatikai fzira,
nyelvtani sszeolvadsra utalnak. Ebben az esetben a lexikai morfma s a grammatikai morfma hatrn trtnik meg a vlts (v. Navracsics 2010: 122). Ehhez a tpushoz nem szksges magas msodnyelvi kompetencia, azok is produklhatnak ilyen
kdvltst, akik alig beszlik a vendgnyelvet.
Navracsics (1999) megllaptja, hogy a lexikai morfma s a grammatikai morfma
keverse gyerekkorban nem tl gyakori, illetve azt is, hogy ezeknek a kontaktusjelensgeknek pragmatikai okai nem lehetnek (v. Navracsics 2010: 122). Azonban lehetnek olyan esetek, amikor a beszdhelyzet elemzsekor kiderl, hogy mgis van
kommunikcis funkcija az ilyen vltsoknak. Az albbi megnyilatkozs is erre lehet
plda. Ebben az esetben a kdvlts funkcija a nyelvi jtk, az irnia kifejezse (illetve ebben a specilis esetben negatv nyelvi attitdt is hordoz; a beszdhelyzetbl,
melynek magam is rsztvevje voltam, kiderlt, hogy abban a sarki hzban ukrnok
laktak, s gy a vltssal nemcsak a hz sznt minstettk, hanem rejtett attitdket
is).
csm, nem semmi az a lila hz a sarkon, megymrtso!
Attl fggen, hogy tipolgiailag milyen nyelvek kerlnek kontaktusba, sokfle vltozata lehet az ilyen kdvltsoknak. Gyakori plda a ketts grammatikai jells (v.
Navracsics 1999, 2010: 122) vagy a bzisnyelvi igekt hasznlata.
Szoktunk ilyet mondani, idu n krhzba.
Mindjrt elubirju, i pojdu domoj.
Szociolingvisztikai aspektusbl az els kt pldban kontextulis kdvltsrl van sz,
mgpedig kivlts, kts trtnik. A morfmaszint vltst mr megelzte egy vlts
253
254
Vratlan:
Do szvidanyije, Kirov utca?
De ez nagyon kultrno.
Nem prepodovtyel, tanr, azt nem volt szabad mondani, az kapitalista cskevny, csak ucstyel, tant.
Frazeolgiai egysgek:
Jelzett:
Az ismerkedst gy kezdi: moj jazik moj vrag. A nyelvem az ellensgem.
Br egyesek szerint a lassan ml id leghatkonyabb ellenszere az alvs:
Szoldt szpit, szluzsb igyot, azaz: a katona alszik, a szolglat telik.
Szkapcsolat, jelzs szerkezet
Vratlan:
Csak gy repltek ki az ablakokon az iratok, trtnelmi dokumentumok,
fasisztszkije matyeriali.
, nemzeti egyetem.
Beszr: inkongruencia.
Vratlan:
Sztalincsi puty volt a kolhoz neve, amikor megalakult, Sztlini t, aztn meg t
a Kommunizmushoz.
Irkutszkj oblszty, az vt a cmem.
Alterncis vltsok sz szinten.
Jelzett:
Hosszan nzegethette, hogy mit csinlok. Vgl csak annyit mondott: riszuj,
riszuj (rajzolj, rajzolj).
Bementem, rgtn azt krdezik cslen partyii?
Vratlan:
Szmolt rszegen, raz, dva.
Mg a milicista meg is fogott, hogy minek akarom tudni, szekretnij.
Az csm rg elment innen, Munkcson lakik, lica Csapjeva.
4.3. Kdvltsok a mondat szintjn
Szintaktikai szinten a mondatrtk megnyilatkozsokat, mint pl. az egyszavas mondatok, egyszer s az sszetett mondatokon belli tagmondatok, illetve a mondatok
255
hvja le a mentlis lexikonbl a lexikai egysget, amin az elhangzott, s erre msodnyelvi kompetencii lehetsget is adnak.
Ltvnyosan tbb plda van beszr kongruencira, mint inkongruencira, a beszl gyakran megismtelte a mtrixnyelven is a mondottakat, ez arra enged kvetkeztetni, hogy elklnlt nyelvi rendszerekrl van sz, de ezek a rendszerek nem kiegyenltettek, hanem abszolt magyar nyelvi dominancia figyelhet meg.
A pldk altmasztjk Poplack (1980) grammatikai knyszereire vonatkoz megllaptsait, vagyis jelzs szszerkezetben a jelz s jelzett sz kztt nem jhet ltre
kdvlts, a szerkezet mindkt tagja azonos nyelven van, a begyazott nyelvbl val.
A pldaanyagban sok szrtk frazeolgiai egysget, lland szkapcsolatot is talltunk. Volt kzttk kzmonds, szls (moj jazik moj vrag) ksznsi formula
(drsztutye, do szvidnyij), kromkods s pohrksznt is.
A vizsglt korpusz egyszavas kdvltsai kztt szfaji gyakorisgot szmszeren
nem szmoltunk, de megllapthatjuk, hogy legtbb esetben fnevek, kznevek:
prepodovtyel (tanr), ucstyel (tant), zzsignyie (gyjts), podpisznyije,
mgyrk, uhork, ncionlnyiszty, tulajdonnevek, intzmnyek, utck megnevezsei, Irkutszkj oblszty, lica Csapjeva, Dopuszk szekretnosztyi, Sztalincsi puty.
Az igk nagy rsze felszlt md: riszuj, igy, dvj, sztvj, gyelj.
Tallhatunk az egyszavas kdvltsok kztt mellkneveket (szekretnij,
ncionlnj), szmneveket (raz, dva), hatrozszt (kulturno, moralno), fnvi igenevet: reklamirovaty, s elfordulnak mondatszk is: vobse, vszjo.
A szkapcsolatoknl leggyakoribbak a jelzs szszerkezetek, ezen bell is a minsg- (szhodnj dokumentcij, fasisztszkije matyeriali, obrzcvij porjdok,
Nyeobhogymij cselovek) s birtokos jelzs szerkezetek: zhod zerna, cslen partyii.
Irodalom
Auer, P. 1999. From codeswitching via language mixing to fused lects. Toward a dynamic typology
of bilingual speech. International Journal of Bilingualism. 1999/3: 309332.
Borbly Anna 2001. Nyelvcsere. Szociolingvisztikai kutatsok a magyarorszgi romnok kzssgben. Budapest: MTA Nyelvtudomnyi Intzete lnyelvi Osztlya.
Csernicsk Istvn 1997. Imja otcsesztvo (Az apai nv hasznlatrl a krptaljai magyarok krben). In B. Gergely Piroska Hajd Mihly (szerk.): Az V. Magyar Nvtudomnyi Konferencia
eladsai (Miskolc, 1995. augusztus 2830). MNyTK. 209. BudapestMiskolc. 1: 1627.
Csernicsk Istvn (szerk.) 2003. A mi szavunk jrsa. Bevezets a krptaljai magyar nyelvhasznlatba. Ungvr: PoliPrint Kft.
Csernicsk Istvn (szerk.) 2010. Megtart a sz. Hasznosthat ismeretek a krptaljai magyar nyelvhasznlatrl. BudapestBeregszsz: MTA Magyar Tudomnyossgrt Klfldn Elnki Bizottsga Hodinka Antal Intzet.
Dussias, P. E. 2001. Psycholinguistic complexity in codeswitching. International Journal of
Bilingualism. 2001/5: 87 100.
Grosjean, F. 1982. Life with Two Languages: An Introduction to Bilingualism. Cambridge: Harvard
University Press.
Gumperz, J. J. Hernandez-Chavez, E. 1969. Communication in Multilingual Societies. In Tyler, S.
(ed.): Cognitive Aspects of Bilingual Communication. Working Paper 28.
258
Lanstyk Istvn 2006. Nyelvbl nyelvbe. Tanulmnyok a szklcsnzsrl, kdvltsrl s fordtsrl. Pozsony: Kalligram Knyvkiad.
Mrku Anita 2009. A krptaljai magyarok ktnyelv nyelvhasznlata pszicholingvisztikai megkzeltsben. In Karmacsi Zoltn Mrku Anita (szerk.): Nyelv, identits s anyanyelvi nevels a XXI.
szzadban. Ungvr: PoliPrint Kft. 107113.
Mrku Anita 2011. Po zkrptszki Ktnyelvsg, ktnyelvsgi hatsok s ktnyelv kommunikcis stratgik (nyelvvlaszts, kdvlts) szocio- s pszicholingvisztikai aspektusai a krptaljai
magyar kzssgben. Doktori (PhD) rtekezs, Veszprm, Pannon Egyetem.
Matras, Y. 1996. Prozedurale Fusion: Grammatische Interferenzschichten im Romanes.
Schprachtypologie und Universalien-Forschung (STUF). 49/1: 6078.
Muysken, P. 1997. Codemixing: Constraints and strategies. Paper presented at the first International
Symposium on Bilingualism. Newcastle upon Tyne, U.K.
Myers-Scotton, C. 1993. Duelling Languages: Grammatical Structure in Codeswitching. Oxford:
Clarendon.
Navracsics Judit 1999. A ktnyelv gyermek. Budapest: Corvina Kiad.
Navracsics Judit 2010. Kdvlts s kdkevers ktnyelvek spontn beszdben. In Navracsics
Judit (szerk.): Nyelv, beszd, rs. Budapest: Tinta Knyvkiad. 121129.
Pfaff, C. 1979. Constraints on language mixing: intrasentential codeswitching and borrowing in
Spanish/English. Language. 55: 291318.
Poplack, S. 1980. Sometimes Ill start a sentence in Spanish y termino en espaol: toward a
typology of code-switching. Linguistics. 18/7-8: 581618.
259
1. Trtnelmi httr
A mai Vajdasg szaki rszt a Magyar Birodalomhoz csatoltk a X. szzadban, gy
akkor meg is kezddtt a magyarok bevndorlsa arra a terletre. A XVI. szzadi
trk hdtsok alatt s utn a magyar lakossg nagy rsze elmeneklt a rgibl,
ahova helykbe szerb nemzetisgeket teleptettek be (Rokai et al. 2002, Jankulov
2003). A pozsarevci bkekts (1718) utn, amikor a terlet Habsburg uralom al
kerlt, a magyarok jra megkezdtk a beteleplst, amelynek az 1920-as trianoni bkeszerzds vetett vget, amellyel a mai Vajdasgot a SzerbHorvtSzlovn Kirlysghoz csatoltk, s ezzel megkezddtt a magyar lakossg elvndorlsa a mai Magyarorszg terletre. A szerbek abszolt tbbsge a II. vilghbor utols hnapjaiban alakult ki, amikor a nmet- s magyarellenes megtorlsok kvetkeztben elpusztult vagy elmeneklt lakossg otthonaiba jelents szm szerb telepes lakossg rkezett.
A dlszlv hbork ideje alatt (1990 s 1995 kztt) menekltek szzezrei telepedtek
le Vajdasg terletn, s ezzel sok vajdasgi lakos tvozott, meneklt klfldre (Gncz
2003).
A npszmllsi adatok szerint 1890-ben a lakossg 24,4%-t, 1910-ben 28,1%-t,
mg 2002-ben 14,3%-t alkottk magyar nemzetisgek (KocsisKocsisn 2001; a
2011-es npszmllsi adatok feldolgozsa mg folyamatban van).
260
261
2. Kisebbsgi jogok
Az els vilghbor utn a dlszlv llam diszkrimincis s erszakos beolvasztsi
politikval nem teremtette meg a feltteleket az alapvet kisebbsgi jogok rvnyestshez, s a magyar npcsoport elsorvasztsra szmos intzkeds irnyult, amely az
oktats terletn is kifejezetten rzkelhet volt. Csak gy indtottak ltalnos iskolai
magyar tagozatot, ha az legalbb 30 fs volt, a tantknak kln vizsgt kellett tennik
szerb nyelvbl (klnben nem tanthattak), akinek a neve nem volt magyar hangzs,
nem iratkozhatott magyar tagozatra.
A msodik vilghbort kvet titi Jugoszlviban a magyarsg helyzett fellendls jellemezte: multikulturlis egyttls, amelyben kialakulhatott a vajdasgi magyar kultra intzmnyrendszere is. A tovbbi intzmnyestsi akcik azonban elmaradtak, s a mr meglvk nagy rszt is beolvasztottk a szerb intzmnyekbe, s a
magyarok hamarosan httrbe szorultak a trsadalmi let minden szntern.
Ami az oktatst illeti, az 1950-es vekben kialakult oktatsi rendszer ltalnos iskolai szinten a magyar tanulk nagy tbbsgnek, kzpiskolai szinten nagy rsznek,
felsfokon azonban csak kis rsznek tette lehetv, hogy anyanyelven tanuljon, amit
tovbb neheztett a terleti iskolk ltrehozsa. Ez azt jelentette, hogy egysgestettk
a szerb s a nemzetisgi tannyelv iskolkat, bevezettk a szakirny oktatst; a kzpiskola els kt ve egysges volt, amit hivatsirny oktats kvetett, kis vlasztk szakmai knlattal.
262
264
Eredmnyek
nyelvhasznlati sznterek (%)
Tmb
Szrvny
Magy
Szerb
Magyar
Szerb
Magnrintkezs
92
68
97*
32
3*
rs
83
17
22
98*
78
2*
Olvass
69
31
34
96*
76
4*
Rdi- s tvmsorok
57
43
38
68*
62
32*
Klnbz helyek
62
38
35
92*
65
8*
Egyb nyelvhasznlat
96
89
98*
11
2*
Isk. nyelvhasznlat
92
91
98*
19
2*
5. Felmrs
5.1. A vizsglat lersa
Muzslya (Zrenjanin/Nagybecskerek klvrosa) szrvnyszigetnek szmt. Eredetileg
magyarlakta telepls volt, magyar oktatsi nyelv iskolval, szerb nemzetisg lakosok ksbb rkeztek. A ma ott l dikok kt oktatsi nyelv ltalnos iskolba jrnak,
ami ebben az esetben azt jelenti, hogy az iskoln bell vannak magyar s szerb tannyelv prhuzamos tagozatok.
A magyar oktatsi nyelv tagozatok dikjai esetben magyar nemzetisg szlkrl
beszlhetnk, a csaldi nyelvhasznlat is 95%-ban magyar, mg a szerb oktatsi nyelv dikok tbbsge vegyes hzassgban szletett, s a csaldi nyelvhasznlat 7585%ban szerb. A dikok mindennapos kapcsolatban vannak egymssal, szmukra termszetes mindkt nyelv hasznlata.
A felmrs, amelyet velk vgeztem, hrom rszbl llt: az els a nyelvhasznlati
sznterek felmrse volt, a msodikban egy magyar nyelv szveget kellett szerb nyelven sszefoglalniuk, valamint egy szerb nyelv szveget magyar nyelven, a harmadikban pedig mindkt nyelven meslnik kellett magukrl, csaldjukrl, szabadidejkrl.
Az els csoportot 16 olyan nyolcadikos dik alkotta, akik magyar oktatsi nyelv
osztlyba jrtak, mindkt szljk magyar nemzetisg, a szerb nyelvet iskolban is
tanuljk, termszetes elsajttssal prhuzamosan, hiszen vsrlskor, orvosnl, bartokkal, Facebookon, utazskor, tvnzskor a szerb nyelvet is rendszeresen hasznljk.
265
A msodik csoportot pedig szintn 16 dik alkotta, akik szerb oktatsi nyelv osztlyba
jrtak, ngy dik esetben mindkt szl magyar nemzetisg, mg 12 tanul vegyes
hzassgban szletett; a magyar nyelvet iskolai sznetekben, nagyszlkkel s szkebb
krnyezetkben hasznljk, ami a mindennapi nyelvhasznlatuk 1520%-t jelenti;
esetkben a magyar nyelvhasznlatot ers kdvlts s kdkevers jellemzi. Iskols
koruk eltt gyakrabban hasznltk a magyar nyelvet, de amikor a szlk gy dntttek,
hogy szerb oktatsi nyelv osztlyba ratjk ket, az addigi nyelvhasznlat csaldon
bell is felbomlott. A szlk gy akartk segteni gyermekeik szerb nyelven trtn
tanulst, hogy otthon is szinte kizrlag szerb nyelven beszltek. Iskolai keretek kztti tudatos tants/tanuls nincs, az esetkben csak termszetes nyelvelsajtts
jellemzi a magyar nyelvet.
5.2. Eredmnyek
5.2.1. Szveg-sszefoglals
Az els csoportot alkot tanulk mindkt szveget teljesen egszben megrtettk,
nem volt szmukra ismeretlen sz egyikben sem. Hezitls s kdvlts jellemezte
azokat a rszeket, amelyekben egy-egy szerb kifejezsnek nem tudtk a magyar megfeleljt (ezek olyan kifejezsek, amelyeket az ott l emberek szerb nyelven hasznlnak egybknt is, magyar ragokkal elltva, pl. Szeretne a medicinszkba iratkozni.).
Nhny esetben megtrtnt a nyelvtani nemek s esetek keverse szerb nyelven (fleg
a kett egyeztetse jelent gondot a nem anyanyelvi beszlk szmra), de ez sszessgben nem akadlyozta az sszefoglal rthetsgt.
A msodik csoport tanuli szmra volt nhny ismeretlen sz a magyar szvegben.
sszefoglalskor ezek a szavak eltt jellemz volt a hezitls. Azt a megoldst vlasztottk, hogy ezeket a rszeket kihagytk, tugrottk az sszefoglalsban, vagy pedig
improvizltak valamit, a szvegkrnyezettl fggen. Az esetkben kevesebb kdvlts trtnt, mint az els csoport esetben, k a kt nyelvet jobban kln tudjk vlasztani.
5.2.2. Irnytott sszefgg beszd
Mindkt csoport tagjai mindkt nyelven komplex s sszefgg szvegeket produkltak, hiszen csak a tma volt adott, vagyis nmagukrl kellett meslnik, brmit, gy
nyugodtan megvlogattk a szavakat, kifejezseket, vagyis kihagyhattk azokat, amelyek hasznlatban nem voltak biztosak. ltalban arra trekedtek, hogy ugyanazokat
az informcikat adjk meg mindkt nyelven. A msodik csoport kt tagjnl trtnt
kdvlts, amikor nem tudtk magyar nyelven megnevezni a szlk foglalkozst.
sszessgben elmondhat a muzslyai nyolcadikos ktnyelv dikokrl az, hogy
ritka s kivteles helyzetben vannak a tmbmagyar s a szrvnyfalvakban l dikokhoz kpest, hiszen mindkt nyelvet magabiztosan hasznljk, annak ellenre, hogy
gyakoriak a kdvltsok, s egy magyar anyanyelv egynyelv szemly szmra ez
egy kevert nyelvnek tnhet, de erre a terletre, mg inkbb erre a korosztlyra ez jellemz, fleg amikor egyms kztt kommuniklnak. Az esetek tbbsgben tisztban
266
vannak azzal, hogy szerb a szt s a toldalk vagy a rag magyar, de szmukra gy
gyorsabb, zkkenmentesebb a kommunikci, hiszen gyis megrti krlttk mindenki, hogy mit szerettek volna mondani, lvn mindenki ktnyelv.
Elgondolkodtat, hogy hosszabb tvon mit jelent majd ez a jelensg a magyar nyelv
alakulsban ezen a fldrajzi terleten, de mivel a szlk nyelvhasznlatt is kevertsg
jellemzi, ez mr termszetesnek tekinthet a gyermekek szmra. Az iskolai rk
nyelvezetre ez termszetesen nem vonatkozik, gy az rai anyagot rkon
kdvltsmentesen tanuljk, illetve produkljk.
Tbbek kztt Andri Edit (2004) is vgzett felmrseket az iskolskor gyerekek
ktnyelvsgnek fokval s az beszdkben megjelen interferenciajelensgekkel
kapcsolatban. Mivel spontn beszlgetseket figyelt meg, az ltala kapott eredmnyek sokkal magasabb fok szkincsbeli interferencira utalnak, mint az ltalam megadott, szablyozott feladatoknl. azt figyelte meg, hogy kdvlts s magyar szavak
szerb ragokkal, kpzkkel val hasznlata leginkbb az albbi tmakrkben fordul
el: iskolai kellkek, telek s italok, klnbz intzmnyek, ltzkds, emberi
tulajdonsgok.
Navracsics Judit (2010) a msodik nyelvi produkcira jellemz tulajdonsgok ismertetsekor kitr a kdvlts jelensgnek magyarzatra is, mely szerint, Grosjeant
idzve, a kdvlts lehet szndkos s vletlenszer is. Szndkosnak mondhat abban
az esetben, amikor egyfajta nyelvi szksgletet elgt ki (hinyzik a megfelel lexma
az egyik nyelvben, knnyebben elrhet egy sz az egyik nyelvben, mint a msikban,
lehet az identits hangslyozsnak trsadalmi vagy pszicholgiai eszkze).
Megkezddtt egy olyan jelleg felmrs, amely rinteni fog tbb tmbmagyar s
szrvnyteleplst a Vajdasgon bell, s az interferenciajelensgek s a
megakadsjelensgek megfigyelst clozza meg a tizenves ktnyelv fiatalok beszdprodukciiban. Remnyeink szerint a felmrs adatainak kielemzse utn tfogbb, globlisabb kpet fogunk ltni a vajdasgi magyarszerb tizenves ktnyelvek
nyelvi kompetencijval kapcsolatban.
Irodalom
Andri Edit 2004. Iskolskori ktnyelvsg Vajdasgban. In Papp Gyrgy (szerk.): Mi ilyen nyelvben
lnk. Nyelvszociolgiai s korpuszvizsglati tanulmnyok. Szabadka: MTT Knyvtr 9. 249267.
Bartha Csilla 1999. A ktnyelvsg alapkrdsei. Budapest: Nemzeti Tanknyvkiad.
Gncz Lajos 1999. A magyar nyelv Jugoszlviban (Vajdasgban). Budapestjvidk: Osiris Kiad
Frum Knyvkiad MTA Kisebbsgkutat Mhely.
Gncz Lajos 2003. Kisebbsgi nyelvi jogok a Jugoszlv Szvetsgi Kztrsasgban A pszicholgus szemvel. In Ndor Orsolya Szarka Lszl (szerk.): Nyelvi jogok, kisebbsgek, nyelvpolitika
Kelet-Kzp-Eurpban. Budapest: Akadmiai Kiad. 163165.
Grosjean, F. 1996. Living with Two Languages and Two Cultures. In Parasnis, I. (ed.): Cultural and
Language Diversity and the Deaf Experience, CambridgeNew York: Cambridge University Press.
Jankulov, B. 2003. Pregled kolonizacije Vojvodine u XVIII i XIX veku. Novi Sad: Panevo.
Kocsis Kroly Kocsisn Hodosi Eszter 2001. Ethnic Geography of the Hungarian Minorities in the
Carpathian Basin. Safety Bay: Simon Publications.
Navracsics Judit 2010. A ktnyelv beszdprodukci jellegzetessgei. Iskolakultra. 2010/10: 316.
267
268
1. Introduction
Psycholinguistic research on bilingualism is more and more focused on the organisation of the bilingual mental lexicon. It has a crucial importance in models of the bilingual memory. In this context we have to mention three models that mirror and summarize our knowledge of the bilingual mental lexicon up till now. The Revised Hierarchical Model (RHM) created by Kroll and Stewart (1994) suggests that in early stages
L2 words are more strongly connected to their L1 translations than to concepts (lexical
mediation) and with increasing proficiency in L2, the links between L2 words and
concepts become stronger, i.e. conceptual mediation takes place. However, Pavlenko
(2009) claims that the RHM does not accommodate the cases of partial and complete
non-equivalence (that is a linguistic category of one language does not exist in the
other language; e.g. the English word privacy doesnt have an equivalent in the Russian language). The next model is the Distributed Feature Model (DFM) by de Groot
(1992, 1993). It focuses on cross-linguistic differences and proposes that concrete
words and cognates share many features across languages unlike abstract words and
this count for the translation-time differences shown in translation tasks (abstract
words are translated slower). The DFM is strongly criticised as it lacks the developmental component and relies on the feature-based approach. Pavlenko (2009) provides
evidence that the assumption that concrete words are translated quicker than abstract
words is in contrast with findings that reveal that there are cross-linguistic differences
between the naming of household objects (Ameel et al. 2005; GrahamBelnap 1986,
Malt et al. 1999, 2003, Kronenfeld et al. 1985). The most recent model was proposed
by Pavlenko (2009), the Modified Hierarchical Model (MHM) that preserves the developmental progression from the RHM and the notion of shared and partially shared
representation from the DFM (this issue has already been addressed by Poulisse in de
Groot Kroll 1997), but at the same time it asks new questions. A central point of the
model is that conceptual representations across languages may be fully shared, partially overlapping or fully language-specific. This differentiation may give answers to the
question raised by researches using the word association method with equivalent stimulus words, whether some trigger more common responses than others.
The word association test is an effective method to reveal the semantic links and
networks within the mental lexicon. In the case of the bilingual lexicon an interlingual
version of the test is needed to reveal the structure of the lexicon.
The interrelation of words within languages and the mental lexicon can be analyzed
by beginning with two basic principles or classes of word association: syntagmatic
(chain) and paradigmatic (choice). Syntagmatic is a horizontal textual relationship that
can be analyzed by what comes before or after a chosen word. For instance, the syn269
tactic word associations with the stimulus word TV could involve such responses as
program, station and watching. The syntagmatic responses to the left or right of the
word would allow for a sequential relationship and the construction of a grammatical
sequence (Carter 1998: 198). Paradigmatic refers to words that might have been chosen instead on a vertical axis. For example, the stimulus word coffee could produce
responses that include tea, milk and sake. The paradigmatic relations include the following types of connections: hyponymy, hyperonymy, synonymy, antonymy, lexical
equivalence. Syntagmatic relations include constructions.
Studies of bilingual childrens lexical-semantic knowledge can provide muchneeded information about the simultaneous development of two linguistic systems. A
developmental phenomenon, the syntagmaticparadigmatic shift, is observed in
monolingual childrens responses in word association tasks around middle childhood
(Nelson 1977), that is why the research is going to be described below was conducted
among older pupils.
2. Subjects and methods
469 HungarianUkrainian bilingual students (age: 1213) who study in 31 schools of
Transcarpathia (Ukraine) served as subjects. They are considered as bilinguals, because during their everyday life they meet and use another language (the state language), the Ukrainian language.
Transcarpathia is situated in the north of Ukraine. Hungarians here constitute a minority. Concerning the multilingualism of the people we must mention that the territory belonged to several countries during the 20th century: till 1918 to AustroHungary,
then to Czechoslovakia till 1938, a big part of the territory belonged to Carpathian
Ukraine between 3 November 1938 15 March 1939, then from 1938/39 till 1944 to
the Hungarian Kingdom, after that till 1991 to the Soviet Union, and from 1991 to
Ukraine (Csernicsk 2009). The linguistic diversity de jure was acknowledged by the
governments but they were trying to force an asymmetric bilingualism to the population which meant that the minority had to study the language of the state but not vica
versa. All these aspects specify the multilingualism and the language knowledge of the
Transcarpathian Hungarian minority. The main part of Hungarians live in villages
64,6%, 8,8% live in so-called city-type settlements and 26,4% live in cities (Molnr
Molnr D. 2010). The main part of people lives in a settlement where Hungarians
constitute the majority.
The representative sample was chosen according to different aspects: type of school
(secondary or grade school), ratio of Hungarians (whether Hungarians constitute a
minority or majority in the settlement), language of the school (Hungarian monolingual or HungarianUkrainian bilingual school in our case bilingual school means
that Hungarian and Ukrainian classes work simultaneously), administrative units (in
Transcarpathia there are 6 administrative units where Hungarian schools can be
found). The control groups were Hungarian monolingual students from Hungary (102
subjects), monolingual Ukrainian students from Ukraine (90 subjects) and bilingual
Hungarian-Slovak students (138 subjects) from Slovakia. The control groups are not
270
representative, the main aspect according to the monolingual groups was chosen was
the closeness to the border (the experimental group live near the border too).
A continuous word-association test on the interlingual level was used as a research
method. The continuous version of the word association task provides a meaningful
way to gather responses. One of the advantages of the continuous approach to the oneword approach is that it allows a large variability in associations for instance, in the
case of a very strong first association (e.g. blood and red). Before the word-association
test subjects filled in a questionnaire that was designed to examine their linguistic
background.
10 words were used as stimuli words. They were chosen from the frequency dictionary of the Hungarian National Text Store (Magyar Nemzeti Szvegtr): boldogsg
happiness (984,167); krdez to ask (7349,667); nagy big (219407,976); lom
dream (5529,333); ad to give (106395,833); kenyr bread (2980,833); vz water
(15760,00); rdekes interesting (8809,667); anya mother (6447,000); szz hundred (20875,35).
The lexical equivalents were applied as Ukrainian stimuli words (Ukrainian frequency dictionary doesnt exist): , , , , , , ,
, , .
And lexical equivalents were applied as Slovak stimulus words: tastie, ptatsa,
velk, sen, dat, chlieb, voda, zaujmav, matka, sto.
The procedure of the word association test was the following. The language of stimuli words was varied through schools, that means in one school Hungarian stimuli
words were applied in the other school Ukrainian/Slovak stimuli words and after 2
weeks the research was repeated in the other language. The instruction of the test was
the following: Write all the words you think of when you look at the stimuli word, the
language is not restricted. Do not write down the stimuli word. For each stimuli word
you have 30 seconds. When you hear the instruction Stop! you have to stop writing
because the next stimuli word comes. There were no other restrictions like word-class
or anything.
The hypotheses were the following:
the organisation of the mental lexicon of the Transcarpathian students is different
from that of their monolingual peers, but similar to that of their bilingual Hungarian-Slovak peers (as both group live in minority circumstances with similar settings);
language dominance is a determining factor in storage;
the examined sociological/sociolinguistic factors have an impact on the number
and type of responses;
paradigmatic relations are predominant within semantic relationships;
the organisation of L1 and L2 is different, i.e. equivalent stimulus words trigger
response words in different ratio.
271
3. Results
Before the results of the experiment are presented it is important to concretize the
phenomena of word. We need to do this because subjects sometimes wrote responses
that consisted of more than one word. That is why each activation is considered as a
mental word as Gsy and Kovcs (2001) named them in their study. According to
the above definition of responses the 20 stimuli words triggered 29702 mental words.
This means that the average number of responses was 63, 33.
The analyses of the data of the questionnaire revealed that the subjects are Hungarian dominant bilinguals (see Table 1 and Table 2). This result is confirmed by the
number of associations triggered by Hungarian stimulus words that significantly
(p<0,001) outnumbers associations of Ukrainian stimulus words (Table 3).
Language of communication in the family
only Hungarian
68%
mostly Hungarian
15%
15%
mostly Hungarian/Russian
2%
only Ukrainian/Russian
1%
Table 1. The ratio of subjects according to the language used in the family
mothers mother tongue
closer language
HUN 94%; UA 5%
language of kindergarten
Mean
Std. Deviation
Minimum
Maximum
Magyar ASSZ
469
33,61
13,663
91
Ukrn ASSZ
469
27,59
14,311
76
Test Statisticsb
272
Descriptive Statisticsa
Magyar ASSZ
Ukrn ASSZ
Magyar ASSZ
Mean
Std. Deviation
Minimum
Maximum
469
33,61
13,663
91
-10,316a
,000
Table 3. The difference between the responses of Hungarian and Ukrainian stimulus words
(a Based on positive ranks; b Wilcoxon Signed Ranks Test)
For the purpose to reveal the strong and weak associates as well, responses were divided into two groups. The first group contains the first responses for the stimulus
words and the second group contains all the other responses that have a connection
with the stimulus word but a weaker one.
First responses were categorized into three groups according to their relation to the
stimulus words: paradigmatic, syntagmatic and associative. Those responses were
considered as paradigmatic where the relationship could be described by one of the
subcategories, like synonymy, antonymy, hyponymy, hyperonymy etc. (e.g.
BOLDOGSG HAPPINESS szeretet love was not considered as paradigmatic).
The syntagmatic group contains those responses which create a whole syntagm with
the stimulus word (e.g. KENYR BREAD enni to eat was not considered as
syntagmatic for in Hungarian the noun requires an object suffix to create a whole syntagm with the infinitive). All the other responses belong to the associative group as
they cannot be labelled as paradigmatic or syntagmatic.
3.1. First responses
The majority of first responses (45,8%) triggered by Hungarian stimulus words are in
associative relationship with the stimulus words (see Figure 1). Paradigmatic responses constitute a lesser part and the difference between associative and paradigmatic
links is statistically significant (t=-2,347; df=9; p=0,04). 16,1% of the responses are in
syntagmatic relation with the stimulus words. The category repetition means that
subjects repeated the stimulus word in their responses, that occurred in 7,1% of the
responses. The no answer category when the subject did not responded to the
stimulus word occurred in a small percent and illegible words constitute 0,5% of the
responses.
273
Figure 1. The ratio of semantic relations between Hungarian stimulus words and responses
Associative links predominate in the responses of the Ukrainian stimulus words (see
Figure 2). Paradigmatic relations occur in a similar number (the difference between
paradigmatic relations of the Hungarian and Ukrainian stimulus words and responses is
significant t=-2,775; df=9; p=0,02), then follows the no answer category with 15,3%.
The ratio of syntagmatic relations constitute a smaller part than in the responses of
Hungarian stimulus words, the difference is statistically significant (t=4,764; df=9;
p=0,001)
Figure 2. The ratio of semantic relations between Ukrainian stimulus words and responses
274
Hungarian
Ukranian
hyperonymy
39%
13%
hyponymy
26%
5%
synonymy
23%
9%
lexical equivalence
7%
70%
antonymy
3%
2%
morphological
2%
1%
275
Figure 3. The ratio of semantic relations between Hungarian stimulus words and weak associates
Figure 4 presents clearly that associative links are predominant in weak associates of
Ukrainian stimulus words. Compared to Figure 3 here the ratio of paradigmatic links is
higher than syntagmatic links.
Figure 4. The ratio of semantic relations between Ukrainian stimulus words and weak associates
72%
krdez/ to ask
48%
nagy/ big
72%
lom/ dream
49%
ad/ to give
36%
kenyr/ bread
78%
vz/ water
76%
rdekes/ interesting
52%
anya/ mother
63%
szz/ hundred
73%
276
4. Discussion
The analysis of the data brought to the surface that associative links are predominant
in our subjects bilingual mental lexicon that can be explained by the functioning of
the associative memory. Comparing responses in the two languages paradigmatic responses are lesser in number for Hungarian stimulus words than associative responses,
while the ratio of paradigmatic responses are high among associations of Ukrainian
stimulus words due to the high number of lexical equivalents. The presence of lexical
equivalents in the data is typical for stimulus words of concrete nouns.
Navracsics (2011) conducted a word-association research among 90 bilinguals and
her results show that the ratio of lexical equivalents is higher in the responses triggered
by concrete nouns than amongst those of abstract nouns. However, L2 responses in
both languages are mostly lexical equivalents. In this context our data shows an interesting picture. Hungarian stimulus words triggered only 2% of the responses from L2
(mainly lexical equivalents), while the responses triggered by Ukrainian stimulus
words are 81% Hungarian.
Van Hell and de Groots (1998) studies prove that bilinguals recalled associations
quicker for concrete nouns than for abstract nouns and in the between-language condition concrete nouns activated mostly lexical equivalents.
If we compare responses activated by equivalent stimulus words we discover differences in the ratio of various links. Paradigmatic and syntagmatic links are significantly
different in number in the two languages. This confirms that the organisation of the
two languages in the mental lexicon is on the one hand different because of different
ratios, on the other hand it is partly similar for in both languages responses predominantly are in associative relationship with the stimulus words that is followed by paradigmatic then by syntagmatic links.
It appears from Navracsicss (2007) results that bilinguals who acquired the two
languages simultaneously store words paradigmatically unlike late bilinguals. The
majority of our subjects are Hungarian dominant bilinguals who acquired the two
languages successively (most of them live in an environment where Hungarians constitute the majority and they meet the L2 only at school) that serve as an explanation
for the dominance of associative links and for the presence of lexical equivalents
among responses for L2 stimulus words.
One of the aims of this study was to examine weak associates, for this reason the
continuous word association test was applied as a research method. Associative links
seems to be the dominant in both languages that proves the network approach
(Aitchison 2003), that is, words in the mental lexicon are connected as a web, elements may be connected directly (like the first associations in our study) or through
other words (like weak associates). This means that besides first associations, weak
associates are not necessarily activated by the stimulus word but by previous associations. McEvoy and Nelson (1982) call this phenomenon as chaining.
As the percentage of common associates doesnt seem to prove any word-class difference (e.g. concrete nouns are not the ones to trigger the most common responses),
we can suggest what de Groot (1993) has already pointed out when examining the
277
278
Kroll, J. Stewart, E. 1994. Category interference in translation and picture naming: Evidence of
asymetric connections between bilingual memory representations. Journal of Memory and Language. 33: 149174.
Kronenfeld, D. Armstrong J. Wilmoth, S. 1985. Exploring the internal structure of linguistic
categories: An extensionist semantic view. In Dougherty, J. (ed.): Directions in Cognitive Anthropology. Champaign: University of Illinois Press. 91110.
Malt, B. Sloman, S. 2003. Linguistic diversity and object naming by non-native speakers of English. Bilingualism: Language and Cognition. 6: 4767.
Malt, B. Sloman, S. Gennari, S. Shi, M. Wang, Y 1999. Knowing versus naming: Similarity
and the linguistic categorization of artifacts. Journal of Memory and Language. 40: 230262.
McEvoy, C. L. Nelson, D. L. 1982. Category name and instance norms for 106 categories of various sizes. American Journal of Psychology. 95: 581634.
Molnr Jzsef Molnr D. Istvn 2010. Akik a magyar nyelvet fenntartjk: magyarok s magyar
anyanyelvek Krptaljn. In Csernicsk Istvn (szerk.): Megtart a sz. Hasznosthat ismeretek
a krptaljai magyar nyelvhasznlatrl. BudapestBeregszsz: MTA Magyar Tudomnyossg
Klfldn Elnki Bizottsg Hodinka Antal Intzet.
Navracsics, J. 2007. Word classes and the bilingual mental lexicon. In Lengyel, Zs. Navracsics, J.
(eds.): Second Language Lexical Processes: applied linguistic and psycholinguistic perspectives.
Clevedon: Multilingual Matters. 1739.
Navracsics Judit 2011. Szaktivci kt nyelven. Budapest: Gondolat Kiad.
Nelson, K. 1977: The syntagmatic-paradigmatic shift revisited: A review of research and theory.
Psychological Bulletin. 84: 93116.
Pavlenko, A. (ed.) 2009. The bilingual mental lexicon. Interdisciplinary approaches. Clevedon:
Multilingual Matters
Poulisse, N. 1997. Language Production in Bilinguals. In de Groot, A. M. B. Kroll, J. F. (eds.):
Tutorials in Bilingualism: Psycholinguistic Perspectives. Mahwah: Lawrence Erlbaum. 201224.
279
1. Introduction
Gender linguistics, which undertakes research into the discrepancies between male and
female discourse, is coming to the fore. It clusters data in connection with the application of diverse modalities speech, writing, computerized communication used by
the two sexes via various channels, in different cultures, subcultures and public life
(Huszr 2009).
My research focuses on the relevance and the discrepancies of mono- and bilingual
discourse on diverse linguistic levels as regards gender.
This study touches upon the communicative strategies of the two sexes, i.e. the occurrence of exclamatives, swear words, question tags, minimal responses, overlaps,
compliments, hedges, directives, topic changes and problem-solving.
The focus of the research centres on HungarianGreek bilingual discourse, as it is
indispensable in shedding light on the results of other Greek corpus linguistic and
cultural-anthropological aspects.
1.1. Greek linguistic and anthropological references
If a researcher is examining bilingualism as entwined with genderlinguistics, it is indispensable to study the norms, social expectations and cultural-anthropological aspects of the subculture, along with the research of linguists and anthropologists concerned with the minority.
Several anthropologists have examined the communicative strategies of Greeks,
such as Friedl, who noted that Greek people are keen on verbal quips and niceties of
expressions. He conducted most of his research in Boetia a village in Central Greece
and concluded that arguments characterized their discourse (Friedl 1962). According
to him, Greek villagers identify Greek ethnicity with an affection for freedom and they
are reluctant to get instructions from anyone. Their maxim 12 Greeks, 13 commanders conveys this message as well (Kakava 2002).
According to Aschenbrenner, passionate debate, frenzied verbal duels and the free
expression of emotion, opinion and disagreement are all common in Greek discourse.
As an anthropologist, Aschenbrenner carried out fieldwork in one of the Peloponnese
villages of south-west Greece, and concluded that they manifest their emotion, agreement and disagreement freely, which gives village life a unique spiritual character.
Disagreement is nurtured in children from their early childhood. Villagers not only let
themselves exhibit their emotions and disagreements explicitly, but are fond of others
doing so as well. They expect social interaction to exceed a relatively tranquil verbal
280
exchange of views. Thus, not only gossip and chatting can be heard but heated arguments as well (Kakava 2002).
Vasillou highlighted emotionally active, competitive and rousing arguments in
Greek discourse (Kakava 2002). The same is reflected in Mackridges research, who
calls attention to (heated, fierce public debate in public places). The
fierce debate, which is used by Greeks in the heat of disagreement, is natural for
them. Mackridge also highlights impassioned, antagonistic debates as well (Kakava
2002). According to Kochman, Greek men outthink, outtalk and outstyle the other
participant, they preserve polarity, grab the floor in communication, and use irony
quite frequently (Kochman 1981: 24).
Alexandra Georgakopoulou highlighted the usage of floor-bidding, floor-holding,
analogy, personalization, delaying disagreement, question repeats (which help to remain neutral), interrupted questions, politeness markers and convoluted turns with
rhetorical acts as general phenomena in Greek discourse. She shed light on the frequency of ironic yes or no questions, mitigated and indirect disagreements, subtle
speechcraft, and minimizing self-committment. Both the usage of discourse markers
which initiate the debate, and the repetition of questions which hinder the disagreement are rhetorical questions, which represent a challenge for the present interlocutor.
As a reaction, by using sudden, interruptive questions, the speaker calls attention to
previously uttered remarks which are faulty, in his/her view (Georgakopoulou 2000).
Georgakopoulou concludes that both delayed and mitigated disagreements are the
preferred and frequently used strategies in Greek interaction. All the above mentioned
strategies are applied for hindering explicit disagreement. When studying code switching, she claimed that the setting for recording the utterances is significant, because the
leisurely atmosphere and conviviality that comes from food and drink consumption as
well as the informal interaction around the table are integrally linked with the participants seamless shifts into liminal scenarios. It also shapes the salience of humorous
talk and joke-telling sessions as the primary activity (Georgakopoulou 2009: 477).
Christina Kakava also conducted research from a genderlinguistic aspect in connection with the linguistic discrepancies of the two sexes. Carrying out classroomresearch, she experienced monological arguments, whereas within families, members
matched wit, used figurative kinship terms and interactional rituals. According to her,
disagreement is a social practice, which is preferred, expected and allowed in Greek
culture. It conveys positive value, just like in Eastern-European Jewish communities.
From a genderlinguistic perspective, in her opinion, Greek women have an inclination
to react, and use nicknames and mitigating strategies. They personalize (if you
were him, what would you do?), and use competitive overlaps. She claims that Greek
females tend to oppose and use sarcasm, competitive overlap and sustained disagreement. They use challenging view alignments, endearment forms, and contrastive repetition and interactional rituals. Firstly, they disagree, and only subsequently do they
give account. They put the other speaker into an analogical situation and raise ironic
yes or no questions. They tend to change deictic centre, and personalize. On the
other hand their male counterparts are inclined to use more interjections, elliptical
questions and upgrade assessment. Males retain position, compete for interactionally
negotiable goods and maintain an opposing stance (Kakava 2002: 1557). They use
281
282
Despite many proverbs suggesting that women are more gossipy, some linguists
claim that men also contribute to the flow of gossip information to a great extent.
Nicholas Emler examined the discourse of 300 interlocutors, and claimed that it was
men who used twice as much gossip information as women (Emler 1994).
In respect of swear words, Gomm examined British participants and highlighted
three times more usage of swear words from men than from women, just as in Coates
New York corpus.
West, Engle and Milroy focused on the frequency of directives in discourse. West
analysed doctorpatient interactions, and drew attention to imperatives used mainly by
male doctors, whereas females applied mitigating strategies and used lets. Milroy
also claimed that, even in kindergarten, most girls are inclined to compromise and
maintain interpersonal harmony, whereas boys use more imperatives and tend to be
tyrannical. This attitude is replicated in problem solving, which was researched by
Leet Pellegrini, who emphasized tactful, collaborative, mitigating attitude from women in conflicts, as against assertive floor-holding and dominance from men.
1.3. Bilinguals and their communicative strategies
According to Grosjean, bilinguals are those persons who use two languages
separately or together for different purposes in different domains of life, with different people. Bilinguals cannot block out either language, since these languages can crop
up anytime in any interaction. Bilingualism is the use of two (or more) languages in
ones everyday life (Grosjean 1992). The functions of codeswitching are the adaptation of the interlocutor to the new circumstances, and social norms, and to make the
message more successful, effective and authentic. Regarding communicative strategies, contextual, metaphorical and situational codeswitching can be experienced as
strategies, for instance in case of Sub-Carpathian Hungarians. These are conscious
strategies in order to express solidarity, humour or linguistic defiance against the interlocutor (Mrku 2010).
2. Research material and methods
My research is based on gender discrepancies in the spontaneous speech of two case
studies One of these case studies comprises the discourse of Hungarian monolingual
participants and the other HungarianGreek bilingual participants. My objective was
to examine the extent to which bilingual male and female utterances differ in respect
of the phonetic, lexical, syntactic levels and communicative strategies and whether
results correspond to the monolingual outcomes, as well as the revelance of the related
literature. I wondered whether it is men who tend to make more phonetic and syntactic
mistakes, more interruptions, overlaps, use more swear words, directives and assertive
style, or whether it is women who are inclined to use more feminine adjectives,
more exclamatives, question tags, minimal responses, compliments, hedges and gossip
information, as most of the related gender studies claim. I also wanted to know whether I will have the same frequency of usage of the above mentioned factors regarding
the two sexes in case they are monolinguals or bilinguals. My hypothesis was that
283
there will be less gender discrepancies in case of bilinguals because they are more
adaptable, and tolerant individuals than the monolinguals. The attitude and the views
of those persons who are exposed to two languages from early childhood are more
flexible, and they are more capable of adapting to each other, because they are
forced to consider more point of views during communication. I conducted a case
study by recording a two hour long spontaneous manifestation of discourse by means
of a dictaphone of HungarianGreek participants. The population consisted of five
friends three men and two women in their twenties. They were well-known to each
another, with some of them living in Beloiannisz, and the rest in Budapest and they
each have at least one parent who is Greek. The Greek parents migrated to Hungary in
the fifties, escaping from the Greek civil war. Their children were born in Hungary
and were exposed to the Greek language from their early childhood due to one Greek
parent, nursery and primary school, where the Greek language was taught. Greek
grammar and vocabulary were consciously taught to them in these institutions. (Alekos, Nikos, Benji acquired the language in Beloiannisz, while Diamandula and Eleni
did so in Budapest.) After finishing primary school they did not participate in Greek
education, although, due to their interaction with the Greek parent and the everyday
life of the subculture of Beloiannisz village, they were exposed to Greek language
stimuli.
I recorded their discourse in a car, and in a restaurant, as I had previously recorded
the interactions of five two women and three men monolingual participants as
well. Both the monolingual and the bilingual participants were in their twenties,
friends who knew each other well. Since my objective was to compare the results deriving from the monolingual corpus to those from the bilingual one and to the related
literature, I had to ensure equivalent circumstances during the recording of the bilingual discourse. The results I intended to compare was the ratio between the the utterances of women and men in respect of phonetic, lexical, syntactic and communicative
strategy differences. The size of the population does not reflect a representative sample because the space of a car is restricted, nevertheless I tried to counterbalance this
situation by analysing the differences between the two genders on multiple levels, i.e.
phonetic, lexical, syntactic and communicative strategy. In this study, I focus solely on
the last of these, namely: exclamatives, swear words, minimal responses, question
tags, overlaps, compliments, hedges, directives, topic changes, gossip information,
interruptions and problem solving used by the two sexes. After recording the sample
of discourse, a bilingual transcription was carried out, which was followed by the
comparison of the above mentioned aspects in respect of monolingual outcomes and
the related literature.
3. Results of the bilingual corpus
When analyzing communicative strategy, I first examined the frequency of exclamatives in respect of gender, of which women used more (see table 1).
Exclamatives
Female
284
Male
Jaj! [ow!]
H! [ugh!]
! A! [ah!]
F! [wow!]
! ! (!) [oh!]
Total
26
20
Swear words
22
Table 1. The number of exclamatives and swear words in the bilingual corpus
Whereas in the monolingual corpus, men competed for floor-holding while women
were pushed backwards, and barely contributed to the flow of communication, in the
bilingual corpus, highlighting the genuine Greek, Mediterranean virtue, women did
not remain silent and participated actively in the discourse, which is reflected in the
proportion of interruptions.
Gossip and verbosity have a negative connotation, and basically it is women to
whom these are attributed- as cited in many proverbs and sayings throughout history.
In contrast to this, both in my monolingual and bilingual corpus, males used twice as
much gossip information as females (see table 2).
Shedding light on communicative strategy, I examined question tags, the usage of
which may express uncertainty only in some cases calling someone to account but
basically it refers to hesitation and requires reinforcement. Women used question tags
more, together with the expression: -meaning yes, expecting positive feedback
from the other participant (see table 2).
Minimal response itself conveys positive feedback and empathy and only in some
cases indifference. It suggests that the listener is paying attention and reinforces the
interlocutor. Females more emotional and empathetic endowment is reflected in their
greater usage of a minimal response, which is six times more than their male counterparts (see table 2).
Communicative strategy
Female
Male
17
10
39
Overlap
17
28
15
49
68
Directives
36
15
Topic change
99
99
Interruption
289
326
Gossip information
21
55
Overlap does not manifest as assertive a style as interruption, since the listener anticipates and finishes the sentence of the previous interlocutor. Regarding emphatic ability, one may think that it is basically women who apply more overlaps, yet both the
literature and the results of the two corpora demonstrate that men use more overlaps:
286
287
debate, the interlocutor applies an evading technique by using a topic change in order
to lessen tension.
D: Durva, akkor n nem hallgatom. [If its obscene, I wont listen to that]
B: Nem szabad. [We shouldnt]
N: Nem illik. Nem illik. Halljuk a vonaton, (lk a vonaton),
vagy ... . (tudom, a szntfldn
beszlnek gy, ott lehet.) [It is not in a due manner. It is not in a due manner.
We can hear it on the train, Im sitting on the train, or, I know, they speak
such a way on the plough-land, it is allowed there]
N: . . (Ott lehet. Az asztalnl nem lehet.) [It is
allowed there. At the table, you are not allowed to do so.]
D: Engem srt. Ha nem vagyok tolakod. . (Mg
sok tanulnivaltok van.) [It offends me. If Im not indiscreet. Youve got
much to learn]
E: Tudod a mondst: fiatalsg bolondsg [You know the saying: crazy
young.]
E: Milyen fa van a Belban a futcn, ami olyan illatos? [What kind of tree is
there in Belo, in the main street which is so fragrant?]
B: Ecetfa. [Sumac]
E: Menj mr innen! [Come on!]
B: Mondom. [True]
E: Nem ecetfa, ht az ecetfa az nem ez! [It is not a sumac, the sumac is not like
this]
B: De! [It is!]
E: Ez jzmin illat, ht az ecetfa az nem ez! [It smells like jasmine, the sumac is
not like this]
B: Minden a, vagy akc. [Everything, a, or, acacia]
E: Ne!Maradj mr Benji, az akcot ismerem, az ecetet ismerem, egyszeren
olyan illata van mint a jzminnak! [Dont! Come on Benji, I know what acacia looks like, I know sumac, it simply smells like jasmine.]
N: Lehet, hogy hrsfa, van htul egy hrsfa. Benji lehet, hogy keveredik a sok
szag [It may be a linden-tree. Look Benji, fragrance may mix]
E: Ezt most komolyan csinljtok, hogy nem ismerem a hrsft? Nem, a amikor
megynk ki a trrl a domb fele. [ Are you kidding and suppose I dont know
linden-tree? No, when we leave the square and approach the hill.]
B: A domb fel? [The hill?]
N: Az lehet, az lehet, hogy dzindzi. [It might be, it might be dzindzi]
E: s jzmin illata van. [And it smells like jasmine.]
N: Az dzindzi, dzindzi [It is dzindzi, dzindzi]
E: Mi az a dzindzi? [What is dzindzi?]
N: Olajfa, Beloiannisz-i olajfa.[Olive-tree, tree from Beloiannisz]
288
E: Maradj mr! Most hlyre vesztek? n ismerem az olajft is. [Come on! Do
you think Im an idiot? I know what olive-tree looks like]
N: Na most ahogy msz a faluban, [ Well, as you walk in the village, ]
D: , ; (A falurl, a falurl beszltek?) [About the village, are you talking about the village?]
N: . (Beloiannisz-rl.) [About Beloiannisz]
4. Conclusion
Focusing on communicative strategies, the results of the corpus only with slight
differences- justify the literature on gender discrepancies. Bilingual men used more
overlaps, interruptions and swear words than women. There was floor-bidding, and
they wanted to be heard. Women used more compliments, exclamatives, question tags
and minimal response. Contradiction with the literature could be traced in respect of
having more directives from women and men gossiping. In respect of culturalanthropological aspects, irony, repetition, politeness markers, interjections, analogy
and verbal quips were present. Nevertheless, the number of fierce, heated debates and
unsolved conflicts was not significant. Both men and women were inclined to compromise and employ a mitigating and evading strategy. Anytime they provoked with
the help of analogy, repetition, agreement and wit, in the end they reconciled. Bilinguals treated conflicts in a flexible way, they manifested their disagreement, and even
women were active participants of the interaction, which is supported by their equivalent numbers of topic changes with the men, and the only slight difference between the
number of interruptions. Therefore, in the bilingual corpus, the number of interruptions and topic changes were not proved to be clearcut gender markers as in the case of
monolinguals.
Summing up, we might state that micro and macro contextual aspects and the cultural-anthropologial attributes of the population examined have to be considered as
well. The Greek/Mediterranean virtue, in line with the expectations of society in favour of open expression of disagreement, can result in different outcomes in respect of
monolingual gender discrepancies, regardless of whether the circumstances in the two
corpora were the same.
Last but not least, let me finish with an anecdote experienced by a Greek linguist,
Christina Kakava, when talking to an American professor married to a Greek wife,
Christina Kakava: I am investigating whether disagreement is a dispreferred action.
American professor: Thats what I thought until I met my wife. (Kakava 2002:
1537)
References
Coates, J. 1993. Women, men and language. Edinburgh: Pearson Educ. Ltd.
Emler, N. 1994 Gossip, reputation and social adaptation. Wichita: University Press of Kansas Publications.
Friedl, E. 1962. A Village in Modern Greece. New York: Holt, Rinehart and Winston.
289
290
When a parent addresses his/her child, s/he does not think about how his/her choice of
language code affects the life of the child. Even children of ethnically homogeneous
families have to learn that we do not say Szia (Hello) to Uncle Jzsi, but
Cskolom (Good morning), that we should not say bad words etc. to become full
members of society. Moreover, those children being brought up in ethnically heterogeneous families have to learn that they need to use a completely different language when
talking to their mum, or to their dad, that they have to say J napot kvnok! (Good
morning!) to Uncle Jzsi, while to Uncle Sergey, that is to (Uncle
Sergey in Ukrainian) they have to say !. That is, while in ethnically
homogeneous families the language socialization process is oriented towards one language, its variants and towards the rules of its use, in cases of children growing up in
ethnically heterogeneous families, it means there is a choice of two or more languages
and their variants and/or the constant switch between them varying from one situation
of language use to another.
In ethnically heterogeneous marriages the language socialization process of the
child is greatly influenced by the quantitative and qualitative language use indices of
the parents, by the attitude they show towards their native language and towards the
language of their spouse etc.
In the present study, relevant characteristic features of the language socialization
processes of 14 ethnically heterogeneous SlavonicHungarian married couples will be
introduced.
The research aimed at exploring strategies of language choice by children growing
up in ethnically heterogeneous families. During the investigation a questionnaire was
filled in by parents. The items included were designed to shed light on the family
background relating to strategies of language use, from which we can draw conclusions about the ongoing language socialization situations in the family. Furthermore,
in the course of family visits we often asked respondents about factors affecting language socialization, such as grandparents living together with the family, language use
in unusual communication situations etc.
As the SlavonicHungarian ethnically heterogeneous families show variance not
only in ethnicity but also in terms of religion, therefore the religious composition of
the families will be introduced first. The Transcarpathian Hungarian minority citizens
mostly belong to the reformed or to the roman catholic church, where Hungarian is the
language of the liturgy, while the Slavonic people (Ukrainians, Russians, Ruthenians)
belong to the Greek Catholic and Orthodox Churches, where the language of the liturgy is Russian and/or Ukrainian. In addition, while the Protestants and the Roman
Catholic people celebrate the main religious holidays (for example: Easter, Christmas,
Pentecost etc.) according to the Gregorian calendar, the Greek Catholic and Orthodox
291
people, according to the ancient Slavonic custom, prefer to use the Julian calendar. In
practical terms, it means that in ethnically heterogeneous families there are/can be two
occasions for Easter, Christmas etc.
Regarding the religious affiliation, heterogeneity can be observed in 10 families,
while in the case of three families we can speak about religious homogeneity. There is
one family where one of the parents claims to be member of a Ukrainian religious sect,
and in this case it is a completely different issue, compared to the cases of traditional
churches (See Table 1).
Family
(number)
Yes
Orthodox Protestant
Yes
No
Yes
No
Yes
???
Religion
When communicating, mothers and fathers may use Hungarian, Ukrainian and/or
Russian. Perhaps, mothers proportionately use Hungarian only more often than
Ukrainian/Russian. It can also be stated that half of the parents use both (in some cases
all three) languages equally in their communication.
In the field of language use in the family, the parents address their partners by the
dominant use of one language. Equal use of the mother tongue of the parents was mentioned only three times by the respondents. Interestingly, men chose the only in
Ukrainian/Russian option more frequently than women (See Table 2.).
Husband
Wife
Total
Only in Hungarian
Mostly in Hungarian
Mostly in Ukrainian/Russian
Only in Ukrainian/Russian
292
293
an. (Munkcs, woman, Hungarian/Hungarian). It is worth noting, that the family lives
in the family house of the wife.
An interesting case can be observed, on the basis of responses given by a married
couple from Beregszsz, as both of them replied that they use their spouses native
tongue more often: Because she is Ukrainian (Beregszsz, man, Hungarian/Hungarian), and Because he is Hungarian (Beregszsz, woman, Hungarian and
Ukrainian/Hungarian and Ukrainian).
Husband
Wife
Total
Only in Hungarian
Mostly in Hungarian
11
Mostly in Ukrainian/Russian
Only in Ukrainian/Russian
Table 3. Responses given to the question: What language do you use when talking to your child?
The parents exclusively or mostly use Hungarian when talking to their child. Among
the spouses it is the husbands who talk to their children mostly in Hungarian.
Among the reasons, the presence/absence of language knowledge, language learning, communication practices etc. repeatedly stand out.
The wife who lives in Oroszi justifies her Ukrainian language use in her interactions
with her children by claiming that it is their native tongue, and additionally mentions,
as a final thought, the ideas so frequently repeated by Ukrainians, namely, that it is a
state language and they liven in Ukraine (Oroszi, woman, Ukrainian/Ukrainian).
The responses given by the children adjust to the communication of the adults, indicating that they use only Hungarian, mostly Hungarian or approximately equally use
the two languages when replying to their parents (See Table 3.).
In interactions with children, the one parent-one language strategy does not prevail
in any of the families. The parental language socialization applied in the childrens
upbringing mostly relies on habit. They could even say: we got used to this, it happened like this. The lack of language knowledge, however, pushes the language socialization process towards the dominant use of one of the languages in family interactions.
One of the important aspects of family interactions is the choice of the medium of
instruction, when registering the child at a school, which later on will mean the extension of family language socialization in the institutional language socialization. It
defines the childrens ethos, acquisition of values, and identity, not only languagespecifically but culturally as well.
294
Medium of
instruction
(kindergarten)
Medium of instruction
(school)
Hungarian
Ukrainian
Ukrainian
Ukrainian
Hungarian
Ukrainian
Hungarian
Ukrainian
Hungarian
Ukrainian
Ukrainian
Ukrainian
Hungarian
No decision yet.
Hungarian
Hungarian
Hungarian
Ukrainian
Hungarian
Hungarian
Ukrainian
Ukrainian
Ukrainian
Hungarian
Ukrainian
Ukrainian
Ukrainian/Hungarian
Child
Table 4. Educational institutions chosen for the children listed according to their language
of instruction
Altogether, eight of the families participating in the study registered their child in
Hungarian kindergartens, and six in Ukrainian kindergartens. But we have to admit
that three out of the latter families live in the village of Visk, which is considered to be
part of the Hungarian diaspora. (See Table 4).
The reasons why parents chose certain kindergartens were quite trivial as in the case
of the two families below:
because there is only a Hungarian kindergarten in the village (Oroszi, woman,
Ukrainian/Hungarian)
because it is closer to our home (Macsola, man, Hungarian and Ukrainian/Hungarian)
my wife decided to do this (Munkcs, man, Ukrainian/Russian), wife: As usual.
295
Certainly, there are more conscious parents, too, who when choosing the kindergarten
already paid to attention to selecting the most effective, the most useful kindergarten from the point of view of the child:
So that he would know Ukrainian (Kovsz, woman, Hungarian/Hungarian).
(registered her child into Ukrainian kindergarten): So that s/he got on in life, as
I went into Hungarian school, and it was difficult (Beregszsz, woman, Hungarian/Hungarian)
(registered her child into Hungarian kindergarten): So that they could learn and
speak in their native tongue (Beregszsz, woman, Hungarian and Ukrainian/Hungarian and Ukrainian)
Seven families think that after the nursery they will register their child in a school with
Ukrainian as the language of instruction, while three families will choose a Hungarian
school, and four families have not decided upon the institution where the child will
continue his/her studies. As we can see in Table 4, in the case of two children going to
Ukrainian and Hungarian kindergartens, respectively, their families are uncertain
about the proper language of instruction at school, and there are only two families who
will register their child into a Hungarian school after finishing the Hungarian kindergarten. The rest will choose a Ukrainian school for his/her child.
And why?
Because I am working there, and it will be easier for both of us (Visk, woman,
Hungarian/Hungarian)
When choosing the language of instruction at school, the hidden educational aims of
the state come to the foreground. These aims foreshadow studying further in ones
mother tongue in the case of minority learners as valueless, and a dead-end process,
which can be proved by the responses given by parents quoted below:
Because we live in Ukraine (Visk, woman, Ukrainian/Ukrainian)
Because we live in Ukraine. This is an advantage for the child. There is a
chance that we are heading in a Hungarian direction (Beregszsz, woman,
Hungarian/Hungarian)
It was a big mistake that I did not go to Ukrainian school, as it was quite hard
to get on in life (Beregszsz, woman, Hungarian/Hungarian)
In the last response, the mother directly refers to going to Hungarian school as a mistake, which was committed against her by her parents, and this is what she does not
want to commit against her own child, and this serves as a reason why she wants to
register her child in a Ukrainian school. She underpinned her choosing the Ukrainian
kindergarten by her own experiences in Hungarian school, and the difficulties that
followed from it.
296
These reasons have their roots in the hidden educational policy exercised in the Soviet era. But Ukraine, which became an independent country in 1991, since 2007 has
passed several laws, made several decisions, and accepted several decrees concerning
educational policy, which have brought minority education to a more difficult situation, and because of which minority citizens are deterred from registering their child in
a Hungarian school but instead in Ukrainian schools. Introducing exams with enhanced education in 2008, or the decree 461 issued by Ivan Vakarchuk, Minister of
Education, serves as good examples for describing the this situation (Karmacsi 2009).
To sum up the results of the present investigation, it can be concluded that, in the
families examined no planned, consciously executed language socialization can be observed, as for instance, the one person-one language method. In cases of parents in family socializations it can be observed that, if the spouse does not know his/her partners
mother tongue, that leads towards the dominant use of their language between the spouses.
When choosing the language of instruction at the kindergarten, or at school, the
parents appear to be more conscious. The choice is usually influenced by personal
experience, or the hidden social effects exercised on minorities.
Out of the preliminary analysis of the conversations recorded at family visits, and
participant observations, such socializing characteristics become evident when the
parent commands the child to use his/her mother tongue or in an indirect way tries to
plead with him/her to use the given language etc. But we can report on it in detail only
after analyzing the tapescripts.
References
Karmacsi Zoltn 2009. Nyelvi attitd: gyengt s erst tnyezk Krptaljn a magyar nyelv
esetben [Language attitude: Strengthening and weakening factors in the case of Hungarian in
Transcarpathia]. In Borbly Anna Vanon Kremmer Ildik Hattyr Helga (szerk.): Nyelvideolgik, attitdk s sztereotpik [Language ideologies, attitudes and stereotypes]. Budapest
DunaszerdahelyNyitra: MTA Nyelvtudomny Intzete Gramma Nyelvi Iroda Konstantin
Filozfus Egyetem Kzp-Eurpai Tanulmnyok Kara. 415422.
297
LOANSHIFTS OF CANADIANHUNGARIAN
AND ANGLOHUNGARIAN LANGUAGE CONTACT
FORINTOS VA
1. Introduction
This essay discusses one aspect of a large-scale study that investigates how the written
language (Hungarian, i.e. L1) of two minority groups functions outside its traditional
setting in Central Europe in an environment where another language is used (English,
i.e. L2 in Canada and in the United Kingdom). They are intraregional language contact
situations where Hungarian immigrants live among the dominant language(s)-speaking
population of Canada and the United Kingdom. The two languages involved are genealogically non-related and structural-typologically non-identical languages.
This study of lexical contact phenomena reveals differences between Standard
Hungarian (SH) and Canadian Hungarian (CH) as well as English (Anglo) Hungarian
(EH) as a result of the influence of the English language on the above mentioned Hungarian community languages; for example it identifies lexical items that are present in
CH and EH but not part of SH.
This research focuses on the loanshifts found in the corpus with special regard to
semantic extensions, where the SH word form is used with the meaning of an English
word. The scope of the present paper is confined to the qualitative survey of the corpus.
Standard Hungarian is represented by the Hungarian National Corpus (HNC) created by the Department of Corpus Linguistics of the Research Institute for Linguistics of
the Hungarian Academy of Sciences under the supervision of Vradi (2002), see also
Sass (2009). HNC includes 187.6 million words. It is divided into five subcorpora by
regional language variants; and into five subcorpora by text genres also: press, literature, science, official and personal, which is very similar to spoken communication in
particular cases (corpus.nytud.hu/mnsz/index_eng.html last visited June 26, 2011).
2. Data and Method
Scholarly literature provides convincing evidence that, in the field of language contact
research, spoken data have always been considered superior to written data, although
examples of both spoken and written language mixing have been identified since the
beginning of humankind. Kurtbke (1998) also criticises the tendency of contact linguistic studies to neglect investigating written language since this field of linguistics
was introduced. The lack of interest in studying written language to gain wider
knowledge on languages in contact stimulated the author to investigate the subject.
Additionally, in contact linguistic research, the study of written discourse is of fundamental importance in the 21st century.
298
Engwall (cited in Kurtbke 1998) suggests among others, that newspaper texts provide an adequate basis for a linguistic study of general language use; as do literary or
specialised texts. If newspapers in general can offer a solid basis for linguistic studies,
then community newspapers of minority groups of different countries are especially
suitable for this function. Since the language of Hungarian migrants in Canada and in
the United Kingdom unlike that of their counterparts in the United States of America
or Australia as well as the language(s) of Hungarian minorities in the Carpathian Basin
has not been the subject of comprehensive research (cf. Csernicsk 1998, Fenyvesi
1995, Forintos 2008, Gncz 1999, Hatoss 2004, Kontra 1990, Lanstyk 2000 ), this
study employs the corpus of written language samples taken from the Anglo
Hungarian communitys quarterly published newspaper titled Angliai Magyar Tkr
(AMT), which is the Journal of the National Federation of Hungarians in England
(issues of 1996, 1999 and 2006) on the one hand. Additional sources include 2005,
2006 and 2009 issues of Kanadai Magyar Hrlap (KMH), as well as 2006 and 2009
issues of Magyar Krnika (MK), which are the on-line newspapers of the Hungarian
community in Canada.
The authors coding scheme creates simplified information for easier comprehension of the research. It can be interpreted in the following sequence:
2000/1/
2000 the year of publication
1 the issue number
3. Results and Discussion
According to Winford (2003: 2959) lexical borrowing must be seen as one aspect of
a creative process of lexical change under contact, which builds on both native and
foreign resources. The results of linguistic interference on the level of lexis of the
receptor-language are manifested in the form of lexical borrowings, mainly modelled
on the donor language, and native creations. The process of borrowing can be very
selective, adopting a foreign form but assigning it a new meaning, or adopting a foreign meaning or concept and assigning it to a native form. Many of the outcomes of
lexical borrowing involve innovations or creations that have no counterpart in the
donor language. Some of these innovations may be created out of donor materials,
others may be created out of native materials, and still other creations are blends of
native and foreign items.
In the relevant literature, numerous definitions exist for the different types of direct
and indirect loans (Haugen 1950, Kontra 1990, Thomason 2001, Weinreich 1953).
Linguists are not only incapable of arriving at an agreement about the dividing criteria
of each term, but also the terminologies differ from each other to a great extent.
Winford (2003: 42) analyses the different attempts made to establish a coherent
framework for dealing with contact-induced changes in the lexicon. He states (2003:
42) that the most comprehensive of the early frameworks may have been that of Betz
(1949), whose basic distinction between loanword and loan-coinage still forms the
basis for current descriptions, and his very detailed and refined terminology describes
word-borrowing in many aspects. According to Winford (2003: 43):
299
300
(zene)szm generally refers to popular music and the term zenedarab would be the
appropriate choice for the given context.
Winfords states that loanshifts may take the form of pure loan translations or
calques in which the foreign model is replicated exactly by native words (Winford
2003: 43); for example rdi-szndarab [radio play] {AMT/2006/4} (vs. SH rdijtk,
hangjtk) and Pannon Egyetem elnke [University of Pannonia President-POSS]
{MK2009/05} (vs. SH rektor).
In Winfords (2003: 44) words, creative word formation involving imported items
is another by-product of lexical borrowing, which Haugen includes in his category of
native creations . Pure native creations means innovative use of native words to
express foreign concepts; for example, az szakiak esetben a dliek esetben [(in
the case of (Hungarians living) in the north and in the south] {AMT/1999/4},
bnyszklntmny [miners detachment] {AMT/1996/4}. Hybrid creations are
blends of native and foreign morphemes to express foreign concepts, for example,
Lord Mayor rezidencijt [Lord Mayor residency-POSS-ACC] {AMT/2006/4} and
fundraising vacsora [fundraising dinner] {MK/2006/09/25}.
Winford (2003: 42-43), with his abovementioned classification, expanded Haugens
category of native creations to include a third subcategory: creations using only
foreign morphemes, which was not included in Haugens classification. Haugens
divisions are not often observed today, as most discussions refer only interchangeably
to either borrowings or loanwords instead; although loan translations or calques are
recognised. What Haugen called semantic borrowings are generally subsumed under
the rubric of convergence (Myers-Scotton 2002). When considering the different lexical borrowings, in this study, Winfords (2003: 4243) classification is observed because it fulfils the purpose of this research most adequately.
In what follows, the linguistic manifestations called loanshifts are provided. Their
existence proves that the countries dominant language has an influence on the minority community language. Since the scope of the present paper is confined to the qualitative analysis of the data, the list is not exhaustive.
(1) hozza nyilvnossg el
{KMH/2005/02}; (vs. SH jelents)
riportjt
[make
public
report-POSS-ACC]
301
to go on foot for pleasure or exercise; stroll. The Hungarian word sta [walk] is
used to express the second activity.
(10) az urnkhoz vonulnak [the polls-ALL go-3PL] {KMH/2005/05}; (vs. SH szavazni megy, az urnkhoz jrul
In example 10, the Hungarian verb vonul [proceed (to a place), go] is used; however, within this context its synonym (jrul) is the more appropriate term because context refers to general elections, that is, az urnkhoz jrul szavazskor [(they) go to the
polls]. The English term to which the extension is applied is: go, since both meanings can be interpreted from the reading of the text.
(11) Tmogatja X pozcijt [support-3SG position-POSS-ACC] {KMH/2006/09};
(vs. SH llspont)
The English word position must have motivated the choice in example 11 because
in English, among its several meanings, it can refer to a point of view or attitude on a
certain question. As opposed to this in comparison the Hungarian term is defined as: a
place or location; the right or appropriate place; or the manner in which something is
placed but it never denotes an opinion.
(12) kriminlis szekta [criminal sect] {KMH/2006/06/}; (vs. SH trvnybe tkz,
bns)
Example 12 is the extension of the adjective kriminlis [criminal] on the basis of
the English adjective criminal, which can similarly to the Hungarian language
mean shameful, disgraceful, dreadful and awful on the one hand, but it can also carry
the meaning of guilty of crime, involving, or having the nature of crime in English
this definition is inappropriate in Hungarian.
(13) X (miniszterelnk) trzza az szaki terleteket [X Prime Minister trip-3SG
the northern territory-LinV-PL-LinV-ACC] {KMH/2006/08}; (vs. SH jrja, beutazza)
The Hungarian verb trzik [to take a trip] generally refers to an excursion, typically a short pleasure trip, a walk outdoors. A prime minister, when visiting locations
within his country, travels from one place to another; which movement is part of his
job and is not made for pleasure, moreover it is not made on foot. Either the English
verb to make a trip or to journey can serve as a basis for extension in the above
example, since they can denote both meanings in English.
(14) ppa megegyezst kttt [Pope agreement-ACC make/bring-PastT-3S Prep
about] {AMT/1999/4} (vs. SH megegyezsre jutott, megegyezst kttt)
In the case of example 14, the English word agreement has to be considered. It
can mean the act of agreeing and harmony of opinion; accord, as well as an arrangement between parties regarding a method of action; a covenant. In the Hungarian language when referring to the process of making an agreement a proper collocation can
be to bring about an agreement [egyezsget/megllapodst hoz ltre], to arrive at an
agreement with sy [megllapodsra/megegyezsre jut vkvel, megllapodik vkvel]; or
to conclude an agreement with sy [megllapodst/szerzdst kt vkvel] but the noun
303
and the verb included in the example do not collocate with each other. The proper
form would be megegyezsre jutott or megllapodst kttt.
(15) a cm nem mutat karcsonyra [the title not refer Christmas-SUB]{AMT/2006/4}
The investigated verb in 15 is the English verb to denote, which can mean to
mark; indicate, to serve as a symbol or name for the meaning of (megjelli, (r)mutat
in SH); and signify directly; refer to specifically (utal in SH). Similarly to the former
example, the context would require its second meaning.
(16) kr hasonl prblkozst kezdeni [useless similar attempt-ACC start-INF]
{AMT/2006/4}
The extension of the English verb to introduce can be found in example 16 since it
can mean to bring forward (a plan, for example) for consideration (bevezet,
kezdemnyez in SH Hungarian, as well as to open or begin (meg)kezd in SH Hungarian). The context suggests that the Hungarian word carrying the first meaning should
have been preferred.
(17) Bartk s Kodly szmokat jtszani } [Bartk and Kodly piece-LinV-PL-LinVACC play-INF] {AMT/1996/4
The English phrase musical composition must have motivated the choice in example 17 because in English it denotes both classical and popular pieces of music. As
opposed to this in comparison the Hungarian term (zene)szm generally refers to popular music and the term zenedarab would be the appropriate choice for the given context
4. Conclusion
In conclusion it can be stated, that the newspapers of the Hungarian minority community in Canada (Kanadai Magyar Hrlap, Magyar Krnika), and the quarterly published Journal of the National Federation of Hungarians in England (Angliai Magyar
Tkr) which are one version of these communities written language used as corpus in the present essay do include semantic extensions, which are linguistic manifestations that come into being when a native word undergoes extension of its meaning
on the model of a foreign counterpart. Although linguistic manifestations of this kind
are the result of the influence of the English language, they can be regarded as part of
the minority communities language maintenance efforts in their respective countries,
as they avoid using an English word as a pure loanword. As a substitute, Hungarian
terminology with an extended definition is utilized and the reader of the text must
interpret the correct meaning according to its context.
References
Akadmiai MoBiMouse Plus Angolmagyar, Magyarangol Akadmiai Nagysztr CD-Rom 2003.
Budapest. Akadmiai Kiad.
304
Csernicsk Istvn 1998. A magyar nyelv Ukrajnban (Krptaljn). Budapest: Osiris Kiad MTA
Kisebbsgkutat Mhely.
Fenyvesi Anna 1995. Language contact and language death in an immigrant language: The case of
Hungarian. University of Pittsburgh Working Papers in Linguistics. 3: 1117.
Forintos va 2008. Aspects of Language Ecology: EnglishHungarian Language Contact Phenomena in Australia. Veszprm: Pannonian University Press.
Gncz Lajos 1999. A magyar nyelv Jugoszlviban. Budapestjvidk: Osiris Kiad MTA
Kisebbsgkutat Mhely Forum Knyvkiad.
Hatoss Anik 2004. Identity formation, cross-cultural attitudes and language maintenance in the
Hungarian Diaspora of Queensland. In Kerr, L. (ed.): Cultural Citizenship: Challenges of Globalisation. Melbourne: Deakin University. 7177.
Haugen, E. 1950. The analysis of linguistic borrowing. Language. 26: 210231.
Kontra Mikls 1990. Fejezetek a South Bend-i magyar nyelvhasznlatbl. [The Hungarian language
as spoken in South Bend, Indiana.] Budapest: MTA Nyelvtudomnyi Intzete.
Kurtbke, P. 1998. A Corpus-driven Study of TurkishEnglish Language Contact in Australia. (Unpublished PhD dissertation without pagination.)
Lanstyk Istvn 2000. A magyar nyelv Szlovkiban. BudapestPozsony: Osiris Kiad Kalligramm
Knyvkiad MTA Kisebbsgkutat Mhely.
Myers-Scotton, C. 2002. Contact Linguistics, Bilingual Encounters and Grammatical Outcomes.
Oxford: Oxford University Press.
Random House Websters Electronic Dictionary and Thesaurus. College Edition Version 1.0 Copyright 1992 Reference Software International.
Sass Blint 2009. Mazsola eszkz a magyar igk bvtmnyszerkezetnek vizsglatra. In Vradi
Tams (szerk.): Vlogats az I. Alkalmazott Nyelvszeti Doktorandusz Konferencia eladsaibl.
Budapest: MTA Nyelvtudomnyi Intzete. 117129. http://corpus.nytud.hu/mazsola
Thomason, S. G. 2001. Language Contact: An Introduction. Edinburgh: Edinburgh University Press.
Vradi Tams 2002. The Hungarian National Corpus. Proceedings of the 3rd LREC Conference.
385389. http://corpus.nytud.hu/mnsz
Weinreich, U. 1953. Languages in Contact: Findings and Problems. New York: Linguistic Circle.
Winford, D. 2003. An Introduction to Contact Linguistics. UK: Blackwell Publishing.
305
1. Introduction
Translation Studies (TS) has traditionally been concerned with linguistic products, i.e.
translated texts. Psycholinguistics is concerned with mental processes. Recently, however, there has been some cross-fertilisation between the two disciplines: as TS is
taking an increased interest in explanatory models of translation, it has come to devote
more attention to the processes involved in translation, taking psycholinguistic methods on board. Psycholinguistics, in its turn, is showing increasing interest in bilingual
language production, simultaneous interpretation and the processes of translation in
general.
The translation of culture specific items (CSI) has enjoyed quite some popularity in TS. Research has centred on the problems of definition, classification, translation procedures and the issue of translatability (Forgcs 2002, 2004, Horvth
2004, Klaudy 2003, Lendvai 2005, 2007, MujzerVarga 2007, Papp 2004,
Tellinger 2003, VlahovFlorin 1980). What has generally not been addressed is
the processing aspects of CSI. Studies of how potential culture specific meanings
are activated in a given text are yet to come.
CSI are difficult to define. Narrow definitions focus on words denoting objects
of material culture, food, drink, clothes, buildings etc. (also known as realia), i.e.
words for objects, places and persons whose referential meanings are missing from
other languages since the things themselves are missing, being specific to one
particular language culture. Broader definitions include names for institutions,
organisations, professions, musical genres, games, proper names, place names,
names of animals, plants, geographical formations, historical events and figures,
cultural events, literary genres, etc. (see VlahovFlorin 1980). These things either
do not exist in other cultures, or do not have the same central significance. Allembracing definitions, even broader than this, include practically everything from
food names to idiomatic phrases, proverbs, sayings, similes, animal metaphors for
people, address forms, interjections, actions, properties, emotions not lexicalized
in other languages, etc. (so-called linguo-realia; see Forgch 2002, 2004, Lendvai
2005, 2007).
In this paper we have no space to discuss definitions of CSI in detail, therefore
we shall confine ourselves to stating our own approach and justifying it to some
extent. We believe that CSI is a relative concept: it depends on the two cultures
involved (cf. Vermes 2004), the centrality of the item in the SL culture (Pedersen
2005), and its function in the text (Klaudy 2003). We hold that potential culture
specificity is different from actual culture specificity. In addition, we tend to believe that culture specificity may concern various aspects of a word: a word may
306
be culture specific in respect of its denotative (referential) or connotative (associative) meaning, or in respect of the encyclopedic knowledge that the members of a
cultural community possess, or any combination of these.
In support of our view let us cite a comparatively recent definition of culture
specific items:
Those textually actualized items whose function and connotations in a source
text involve a translation problem in their transference to a target text, whenever
this problem is a product of the nonexistence of the referred item or of its different intertextual status in the cultural system of the readers of the target text.
(Aixel 1996: 58)
This definition also recognizes the fact that CSI is a relative concept: it is actualized in a text, and depends on the two languages involved in the translation. We
may add that in the process of translation any word might become culture specific
(cf. Vall 2000, Papp 2004, Vermes 2004, Heltai 2007).
2. Translation of CSI
Much of the literature on CSI is concerned with the transfer operations or translation
shifts that take place in translating CSI (e.g. Klaudy 2003, Tellinger 2003, Papp 2004,
Pedersen 2005, Lendvai 2007). Many authors have pointed out the relative untranslatability of culture-specific items, especially in respect of their connotative meanings.
Tellinger (2003), for example, emphasizes that the words troika, samovar, steppe, taiga
etc. have connotations in Russian that are difficult to understand for people from another culture.
On the other hand, Pedersen (2005), studying the translation of subtitles, does not
think that CSI (in his terminology ECR, i.e. extralinguistic cultural references) are
untranslatable. Listing the various factors that may facilitate or hinder their translation
he concludes that
[...] if all the factors influencing translatability and translation strategies in subtitling (extratextuality, monoculturality, intersemiotic redundancy, co-text, mediaspecific constraints, paratextual considerations) work against the translator, then
none of the strategies may offer a solution and you may have an untransferable
ECR on your hands. However, after analyzing about one hundred thousand subtitles, the present author has still to come across such an ECR.
Our own position is close to Pedersens: we believe that the difficulties of translating
CSI have been overstated. Our arguments are as follows.
Firstly, denotative meanings of CSI representing objects of material culture are not
very difficult to render in another language: borrowing the SL word, for example, is
always an option. Pizza and sushi and tequila might have been culture specific, but
today, as far as their denotative meaning is concerned, they are transcultural.
307
We believe that it is only when we know all this that we can get a realistic picture of
the problems of translating CSI and when we can begin to ask whether they can be
rendered in another language and whether they have been successfully rendered in a
given translation.
This study was undertaken to explore two of the above issues. It consisted of two
parts: a free association test with selected CSI from a Hungarian novel and textual
analysis of the co-texts of the items in the original novel and its translations into three
target languages, English, German and Russian.
In this study we could not address the issue of connotations triggered by CSI in context with various classes of SL and TL readers. Notwithstanding, we believe that the
results are interesting in themselves and may open up new lines of research.
4. Free association tests
4.1. Materials and methods
The study was based on the novel A Pl utcai fik by the Hungarian writer Ferenc
Molnr, translated into English as The Paul Street Boys. This novel is about schoolboys in Budapest at the turn of the twentieth century and is perhaps the most famous
Hungarian novel in the world.
About 300 potential CSI were identified by the present authors in Molnrs novel.
This was done on an intuitive basis: reading the novel, the authors recorded the items
that looked likely to have culture specific potential. Through a gradual process of selection and elimination the list was reduced to about 70. Since it was impractical to
test 70 odd items, the 22 most likely items were selected for the tests. The list, with
the English, German and Russian translations, is given in Appendix 1.
The selected items were used in a written free association test with several age
groups. The total number of subjects was 127, including 51 primary, 44 secondary and
13 college students, as well as 13 adults aged 25 to 50 and 6 German university students. The subjects were selected on a convenience basis. (See Appendix 2 for exact
details.)
The instruction given was Read these words one by one and write down the first
three words that comes into your mind in the space provided next to each word, without pausing to think. All the groups had the same instruction. The German students
had to respond to the German translations of the CSI as they appeared in the German
translation.
4.2. Results
Presenting the results of the free association tests in full would occupy many pages, so
we can present only samples and provide a general summary of the rest.
309
4.2.1. abcg
The responses given to the cue abcg (E. down with, G. Abzug, R. ) are given in
Table 1. The numbers indicate the total number of responses given by all the participants (first, second and third responses were not kept separate). The responses considered to reflect cultural associations are marked in grey.
A large number of the responses given to this word reflected associations related to
a well-known Hungarian nursery rhyme. There was also a significant number of responses reflecting awareness that this word is of German origin. Some respondents
associated it with the German word Zug (train, draft). A few responses indicated
that abcg is not often used in present-day Hungarian, and some responses related it to
a rock opera in which the word occurs.
Most frequent responses
45 (apacuka)fundaluka/gyerekvers
19 nmet
German
9 vonat
train
8 abc
abc
7 Istvn kirly/Koppny
7 Micimack
6 idegen
foreign
5 le (vele)
down with
3 iskola
school
4.2.2. alfldi
Many of the responses given to alfldi (E. wide open spaces, G. Flachland, R.
) have to do with its denotative meaning: synonyms of flat and plain
dominate. Many responses reflect cultural associations: names of animals associated
with the Great Hungarian Plain (puli, a dog breed and szrkemarha, a cattle breed) as
well as the cultural symbol gmeskt (draw well) have a significant presence. The
name of the poet Petfi as well as well-known place names Hortobgy and Tiszntl
are also remembered. The associations magyar, Magyarorszg (Hungarian, Hungary) show that respondents consider the Alfld (lowland) a concept specific to Hungarian culture. Many associations suggest typical collocations (landscape, dialect,
310
people, folk costumes), and a few are related to the name of a well-known actor
(Alfldi Rbert) and the name of a company manufacturing toilets. The responses are
shown in Table 2.
Most frequent responses
22 rna
plain (n.)
18 tj, szls
region, dialect
16 sk
plain (adj.)
11 ember
people
9 zld
green
9 Alfld
Alfld (Lowland)
8 kenyr
bread
8 Petfi / Sndor
Petfi / Sndor
8 fld etc
7 (szarvas)marha, szrkemarha.(hs)
cattle, beef
7 magyar. M.o.
Hungarian, Hungary
7 lapos
flat
6 puli
6 tehn
cow
5 (gmes)kt
(draw)well
5 np/viselet
folk/costume
5 Hortobgy
Hortobgy
5 Tisza, -ntl
Tisza
WC
Table 2. Responses given to alfldi
4.2.3. grund
The responses given to grund (E. ground, G. Grund, R. ) can be divided
into two large groups. Some responses are related to encyclopedic knowledge about
grund: it is a place where children can play. Other responses are related to the novel
itself: the word grund acquired culture specificity because practically everyone in
311
Hungary has read the novel, and this word triggers associations of both the novel and
the concept.
Most frequent responses
9 tr
space
8 jtk
play
7 gyerek(ek)
children
6 banda, bandzs
band, gang
6 foci
soccer
6 labda
ball
5 csapat
team
312
become culture specific in Hungarian exactly because most Hungarian have read this
novel: einstand and gittegylet may be used as allusions in social and political contexts.
The culture specific connotations of some of the items are related to the fact that
they are of German origin, carrying connotations of early twentieth-century Budapest,
where many people spoke German and many German words were also used by the
Hungarian-speaking part of the population. This culture specific feature (of German
origin, typical of the language of Budapest at the time) arises in the case grund, einstand, boxenli and abcug. Except for einstand, where the author explicitly states what
this ugly German word means, the connotation cannot be rendered in German (the
German translation, perhaps not by chance, drops the word ugly).
In some cases differences appeared in the responses of different age groups. In general, the number of clang associations was higher in the below 20 age group. In the
case of abcg, 23 clang associations occurred (e.g. abc, ablak, abnormlis, abszurd).
The responses given by the German subjects were in many cases not comparable to
the Hungarian results, since they had to respond to German words whose meaning was
more or less equivalent to the Hungarian original only in the given context. Thus, in
the case of Abzug their associations centred around chemistry (Chemie, Chemieunterricht, Experiment, Foto, Foto), weapons (gefhrlich, Gewehr, Gewehr Gewehr, Krieg,
Militr, Pistole, schieen) and cooking (Dampf, Dunst, Herd, Kochen, Kochen, Kche,
Kche Kche/kochen) and going away (Ausmarsch, Wegfahren, ziehen). As for Grund,
in isolation this word activated completely different associations, mostly something to
do with the sea: Meer(esboden), Meeresgrund, Ozean, 2 Sand, 2 Schiff, Schiffswrack,
Schlucht, See, Sinken, Teich, tief, Unten. The responses given to Flachland (H. alfldi)
clearly reflect that they were thinking of the north-German plain (Norddeutschland,
Norden, Wattmeer, Wiese, Windmhle), which is obviously different from the Hungarian Alfld.
5. Text analysis
After exploring the potential associations of the selected CSI in isolation, we examined the text in the novel to see whether potential culture specific associations may be
expected to be realized in context.
It appeared that many of the culture specific associations identified in the free association tests are likely to be irrelevant in the context of the novel. Thus, it is rather
unlikely that the name of a present-day actor or the name of a toilet manufacturer will
be activated when someone reads the word abcg in the novel, or that s/he will think
of the rock opera in which the word occurs. More importantly, even if it might occur
with some readers, it is not an association intended or suggested by the author. (The
novel appeared some seventy years earlier than the opera.) In a similar manner, when
one reads alfldi, one might indeed activate connotations having to do with particular
dog breeds (puli) or the dialect spoken there, but such associations fall outside the
main drive of the text, which compares the limited space of the grund to the wide
open spaces (as the English translation aptly puts it) of the Alfld. Many respondents
associated the word krajcr (penny) with a folk tale (A kiskakas gymnt flkrajcrja). In the novel, however, the word is simply used as the name for a small amount of
313
money, and associations with the folk story will probably not arise in SL readers, and
if they do, they are irrelevant for this context. We have a similar situation with egy
korons (one crown), also a kind of money. The word magyar nta (Hungarian song)
is potentially also culture specific, conjuring up images of restaurants where gypsies
play emotional, so-called Hungarian songs (not identical with Hungarian folk songs)
and patrons sing, as if to illustrate the proverb Hungarians cry when they make merry. All these melodramatic associations are neutralized in the novel. The important
thing here is that the hurdy-gurdy was playing a merry song, not some serious classical
stuff, making all the boys happy. We must also add that different Hungarian readers
may have different connotations about magyar nta. In this way, readers of the English, German and Russian translation do not lose too much if they are not given a
background to the role of Hungarian songs in Hungarian culture.
In the case of historical references, Hunyadi, Mohcsi vsz, and Tomori rsek the
TL reader may again lose some connotations, yet the novel can be appreciated and
enjoyed without detailed knowledge of Hungarian history. Young Hungarian readers
do not always know more about Hungarian history than what they learn from the novel
itself. We must remember that this novel is mainly about young people and for young
people, for whom the story itself is more important than the associated culture. From
the context of the novel it becomes clear that Hunyadi was a great soldier whose military abilities adolescents may admire and may try to emulate, that Mohcs was the site
of a lost battle and Tomory was an archbishop who fell in that battle. Thus, compared
to young Hungarians, readers of the English, German and Russian translations are not
really at a great disadvantage, and the Russian translation provides footnotes for Hunyadi and Mohcs.
It must be noted that apparently some associations that appear in the tests can be
traced back to linguistic structure: the components of compound words or metaphorical items. Zongoraverkli (hurdy gurdy), a compound containing the word zongora
(piano) generated many associations with that instrument. It remains to be further
explored whether such linguistic associations play a role in the interpretation of the ST
by SL readers. The same question arises in the case of such compounds as szentjnoskenyrmag (literally, seeds of St. Johns bread) and trkmz (Turkish honey).
It is interesting to note that some associations must have emerged recently: the word
abcg (of German origin) is definitely obsolete in present-day Hungarian, and many
young people would not know it at all, had it not occurred in a song in the popular
rock opera Istvn, a kirly (Stephen the King). Again, the question arises whether
this connotation would arise when young people read the novel.
None of the purported CSI seems to generate an implicature or inference. The only
exception is perhaps the name Hunyadi, where readers must infer that this was a great
soldier and a role model. However, this inference can also be made on the basis of the
translated versions.
314
6. Lost in translation
Some cultural associations are lost in translation. The connotation German word in
words of German origin cannot be rendered in German, and is only partly rendered in
English and Russian. This means that readers of the translations may lose the association that at the beginning of the 20th century Budapest was a town where many German words were used.
The connotations of tt (Slovak) are lost in all three translations. The Russian
translation does best: it renders some of the connotations by reproducing to some extent the (funny) foreign accent of the character, while the English translation translates
it as Slovenian (anyway, does it matter, if you look at East Europe from London?).
However, the connotations that used to be linked to this word seem to be disappearing
with the younger generations, since the word tt has been ousted in present-day Hungarian by the word szlovk (Slovak).
As indicated above, the potential connotations of and encyclopedic knowledge
about names of historical figures are to some extent lost in the translations. We emphasize potential, since the responses indicated that many young people in Hungary do
not know much more about the historical background of the names than TL readers.
The connotations of mta are that this word is no longer used and the game itself is
old-fashioned, no more practised. The English translation, baseball, does not appear to
render this connotation.
7. Conclusions
The bark of CSI is worse than their bite. Our analysis appears to confirm Pedersens
contention that CSI are usually translatable: some purported CSI are in fact transcultural, some can be more or less adequately interpreted by a secondary audience because they occur in a self-explanatory context, some are no more culture specific for a
TL audience than for a SL audience, some can be explained in translators notes, and
many potential culture specific connotations are irrelevant and probably do not arise in
a given context. Apparently, understanding a text and getting a full aesthetic experience does not depend on activating the full or range of potential connotations of the
SL item.
Some connotations that may be activated by SL readers cannot be rendered in the
translations. Yet we must remember that many culture specific associations are not
shared by all members of a SL community. In spite of this fact, most of them can enjoy reading the novel. From this it follows that TL readers can also enjoy reading it,
even if some of the culture specific references are lost. In the given case, following the
adventures of the young heroes is more important for young readers of any nationality
than relating the story to its cultural foundations.
Sources
Molnr Ferenc 1967. A Pl utcai fik. (23rd ed.) Budapest: Mra Ferenc knyvkiad.
315
Molnr, F. 2005. Die Jungen der Paulstrae. Translated by Edmund Alkalay. Budapest: Corvina.
Molnr, F. 2005. The Paul Street Boys. Translated by Louis Rittenberg. Budapest: Corvina.
, . 1986. . .. :
. http://publ.lib.ru/ARCHIVES/M/MOLNAR_Ferenc/_Molnar_F..html#01
Steinbeck, J. 1962. Travels with Charlie. New York: Viking Press.
References
Aixel, J. F. 1996. On the Cultural Aspects of Translation. In lvarez, R. Vidal, M. C. (eds.):
Translation, power, subversion. Philadelphia: Multilingual Matters. 5278.
Forgcs Erzsbet 2002. A relik fordtsi nehzsgeirl szpirodalmi szvegekben.
Fordtstudomny. 4/2: 6382.
Forgcs Erzsbet 2004. Relik s fordtsuk Garaczi Lszl mveiben. Fordtstudomny. 6/2: 38
56.
Gutt, E-A. 1991. Translation and relevance: cognition and context. Oxford: Blackwell.
Heltai Pl 2007. Ekvivalencia s kulturlisan kttt kifejezsek a fordtsban. In Heltai Pl (szerk.):
Nyelvi modernizci. Szaknyelv, fordts, terminolgia. A XVI. Magyar Alkalmazott Nyelvszeti
Kongresszus eladsai. 3/2. ktet. Gdll, 2006. prilis 1012. PcsGdll: MANYE Szent
Istvn Egyetem. 643653.
Horvth Pter Ivn 2004. Kzigazgatsi relik a hiteles fordtsban. Fordtstudomny. 6/2: 3137.
Klaudy Kinga 2003. Languages in Translation. Lectures on the theory, teaching and practice of
translation. With illustrations in English, French, German, Russian and Hungarian. Budapest:
Scholastica.
Lendvai Endre 2005. Reliafelfogsok napjaink magyar fordtselmletben. In Dobos Csilla et al.
(szerk.): Mindent fordtunk, s mindenki fordt. rtkek teremtse s kzvettse a nyelvszetben. Budapest: Szak Kiad. 6772.
Lendvai Endre 2007. Translating Culture-bound Units: Strategy, Tactics & Techniques. Wien: Praesens Verlag.
Mujzer-Varga Krisztina 2007. A reliafogalom vltozsai s vltozatai. Fordtstudomny. 9/2: 55
84.
Papp Nndor 2004. Terminolgia s relik. In Drth Jlia (szerk.): Szaknyelv s mfordts. Gdll: Szent Istvn Egyetem. 7991.
Pedersen, J. 2005. How is Culture Rendered in Subtitles? Challenges of Multidimensional Translation:
Conference
Proceedings
2005.
118.
http://www.euroconferences.info/proceedings/2005_Proceedings/2005_Pedersen_Jan.pdf
Tellinger, D. 2003. A relik fordts a fordt kulturlis kompetencija szemszgbl.
Fordtstudomny. 5/2: 5870.
Vall Zsuzsa 2000. A fordts pragmatikai dimenzii s a kulturlis relik. Fordtstudomny. 2/1:
3449.
Vermes Albert Pter 2004. A relevancia-elmlet alkalmazsa a kultra-specifikus kifejezsek
fordtsnak vizsglatban. Fordtstudomny. 6/2: 518.
Vlahov, Sz. Florin, Sz. 1980. Nyeperevogyimoje v perevogye. Moszkva: Mezsdunarodnije
otnosenyija.
316
Appendix 1
Potential CSI in the Hungarian text and their English, German and Russian translations
Hungarian
English
German
Russian
abcug (Kolnay)
Down (with
Kolnay)
Abzug (Kolnay)
()
alfldi
Flachland
bcsiesen
Viennese
bermtig
boxenli
citadellban
citadel
Festung
egy korons
revenue stamps
einer Krone
einstand
einstand
Einstand
gittegylet
Putty Club
Kitt-Verein
grundon
grund
Hunyadi Jnos
Jnos Hunyadi
Johann Hunyadi
korzn
Danube Drive
Donaukorso
krajcr
penny
Kreuzer
magyar nta
magyar melody
ungarische Melodie
mta
baseball
Schlagball
Mohcsi vsz
The Disaster at
Mohcs
nkpzkr
reading club
Selbstbildungsverein
szentjnoskenyrmag
Johannisbrot
Tomori rsek
Abbot Tomori
Erzbischof Tomori
trkmz
halvah
trkischer Honig
tt
Slovenian
ein Slowake
zongoraverkli
hurdy-gurdy
Leierkasten
317
zsebtintatart
pocket inkwell
Taschentintenfa
Appendix 2
Data for the participants
Age
Male
Female
Total
Place of study
1014
18
33
51
1519
11
33
44
1924
13
3551
13
adults (Veszprm)
2024
Total
42
85
127
318
1. Introduction
This case study/essay discusses one aspect of a large-scale study that investigates how
the spoken language (Hungarian, i.e. L1) of a member of a minority group functions
outside its traditional setting in central Europe in an environment where another language is used (English, i.e. L2 in Australia). It is an intraregional language contact
situation where Hungarian immigrants live among the dominant language-speaking
population of Australia. The two languages involved are genealogically non-related
and structural-typologically non-identical languages.
This study of lexical contact phenomena reveals differences between Standard
Hungarian (SH) and Australian Hungarian (AuH) as a result of the influence of the
English language on the above mentioned Hungarian community language represented by the subject involved in the research. It identifies phonological as well as
lexical items that are present in AuH but not part of SH. The aim of the study is to
provide both the qualitative and the quantitative analysis of the data.
Standard Hungarian is represented by the Hungarian National Corpus (HNC)
created by the Department of Corpus Linguistics of the Research Institute for Linguistics of the Hungarian Academy of Sciences under the supervision of Vradi
(2002), see also Sass (2009). HNC includes 187.6 million words. It is divided into
five subcorpora by regional language variants, and into five subcorpora by text
genres also: press, literature, science, official and personal, which is very similar
to spoken communication in particular cases (corpus.nytud.hu/mnsz
/index_eng.html last visited June 26, 2011).
As far as language contact research in Hungary is concerned, the varieties of
Hungarian spoken in the neighbouring countries and overseas have been the subject
of research (cf. Csernicsk 1998, Fenyvesi 1995, Fenyvesi 2005, Gncz 1999, Kontra 1990, Lanstyk 2000). The language contact of Australian Hungarian has been
studied by Kovcs (1996, 2001a, 2001b), Hatoss (2004), Vszolyi (2003) and Forintos (2008). This line of more sociolinguistically oriented research is complement-
This article was parially made possible by the financial support of the Faculty of Modern
Philologies and Social Sciences of the University of Pannonia. We are thankful to Zsuzsanna
Rednik for her administering the interview serving as the basis of the case study, to Anita Gere
and Hanna Hofstetter for transcribing the interview and to the audience of the 11th Biennial
European Association for Studies on Australia (EASA) International conference in Preov in
Slovakia.
319
ed by this study, which is concerned basically with the linguistic context of language
contact.
2. Data and Method
The corpus of the research is a tape-recorded interview of approximately 60 minutes
made in July, 2010 with a thirty-year old Australian citizen, a woman both of whose
parents were born in Hungary and decided to settle in Australia at an adult age as firstgeneration Hungarians; she was born in Australia, so she is a second-generation Hungarian in Australia having had her education at primary, secondary and tertiary level in
English. She learnt and she speaks Hungarian almost exclusively in the family and
with wider family members living in Hungary via the phone. She has never attended
any Sunday schools or any Hungarian courses, which has greatly contributed to the
fact that she can only speak Hungarian, but she is unable to read and write it.
The interview was carried out by a colleague of ours and the topic was generally the
lifestyle of Australians in Australia; it included 3,345 of her words. As for the method
applied during the research, the linguistic manifestations were collected manually.
During the research, when studying the interlingual linguistic manifestations, the Winfordian (2003) classification was followed.
In order to be able to decide on the exact meanings of the Hungarian and the English words, dictionaries were consulted.
3. Results and Discussion
According to Winford (2003: 2959) lexical borrowing must be seen as one aspect of
a creative process of lexical change under contact, which builds on both native and
foreign resources. The results of linguistic interference on the level of lexis in the receptor-language are manifested in the form of lexical borrowings, mainly modelled on
the donor language, and native creations. The process of borrowing can be very selective, adopting a foreign form but assigning a new meaning to it, or adopting a foreign
meaning or concept and assigning it to a native form. Many of the outcomes of lexical
borrowing involve innovations or creations that have no counterpart in the donor language. Some of these innovations may be created out of donor materials, others may
be created out of native materials, and still other creations are blends of native and
foreign items.
In the relevant literature, numerous definitions exist for the different types of direct
and indirect loans (Haugen 1950, Kontra 1990, Thomason 2001, Weinreich 1953).
Linguists are not only incapable of arriving at an agreement about the dividing criteria
of each term, but also the terminologies differ from one another to a great extent.
Winford (2003: 42) analyses the different attempts made to establish a coherent
framework for dealing with contact-induced changes in the lexicon. He states (2003:
42) that the most comprehensive of the early frameworks may have been that of Betz
(1949), whose basic distinction between loanword and loan-coinage still forms the
basis for current descriptions, and his very detailed and refined terminology describes
word-borrowing in many aspects. According to Winford (2003: 43):
320
322
or native stem+imported affix (no example found in the corpus); and compound
blends, e.g. imported stem+native stem (no example found in the corpus).
Furthermore, there are loanshifts (also called loan meanings) in which the morphemic composition of the item is entirely native, though its meaning derives at least
in part from the donor language. Each of these categories can be further subdivided,
according to the types of importation and substitution involved (Winford 2003: 43).
Loanshifts do not actually include surface-level alien morphemes, but instead influence L1 material. They can be divided into the following subtypes. When a native
word undergoes extension of its meaning on the model of a foreign counterpart, these
are cases of extensions or semantic loans. For example:
3.3. Loanshifts
Semantic loans
(1) erd [forest, wood, bush] (vs. SH Ausz erds-boztos terlet, az ausztrliai serdk)
In example 1 the English word bush has to be considered because it basically
means a low shrub with many branches and a thick growth of shrubs, a thicket, but it
can also refer to land covered with dense vegetation or undergrowth more specifically
to land remote from settlement, e.g. the Australian bush. In the Hungarian language
however, erd means forest, e.g. a dense growth of trees, plants, and underbrush covering a large area, or wood, and it definitely does not refer to the typically Australian
phenomenon, the bush.
(2) monds [saying, expression] (vs. SH kifejezs)
The two meanings of expression relevant to example 2 are: a particular word or
phrase expressed, stock phrase, idiomatic expression/phrase (szls, szlsmonds in
SH) and to designate, to call, to express in words (kifejez in SH). In the given context the subject is looking for the right term for the concept she wants to express, so
the equivalent of the second meaning should have been used.
(3) kinyit zletet [he opens a shop] (vs. SH zletet nyit/megnyitja zlett)
The semantic extension of the word to open can be found in example 3, since
among its several meanings in English it can mean to release from a closed or fastened position, e.g. to open a/the door as well as to commence the operation of: open
a new business, e.g. open ones shop/store. In the Hungarian language however, the
meaning can vary depending on what preverb is attached to it. The Hungarian word
kinyit generally carries the first meaning mentioned above and megnyit or nyit refers to
the second one, which the context would require.
(4) a vros kzepbe [into the center of the city] (vs. SH kzpontjba)
Example 4 is the extension of the noun centre inasmuch as a meaning of it can be
a point or place that is equally distant from the sides or outer boundaries of something, i.e. the middle, e.g. the centre of town on the one hand, which is kzpont in
323
Hungarian. Also it can mean a point equidistant from all points on the circumference
of a circle or on the surface of a sphere, a part of an object that is surrounded by the
rest; a core on the other, which is kzp/kzepe in Hungarian. When one mentions the
city centre, definitely the first form is to be used.
(5) van nekik maga hzuk [they have their own house] (vs. SH sajt)
The (possessive) adjective own is to be considered in example 5, which can mean
of or belonging to oneself or itself, (ones) own, proper e.g. they have their own
house. The Hungarian equivalent of this meaning can be the synonyms: sajt and
maga. Usually the choice depends on the context, i.e. what noun it is followed by, to
have an appropriate collocation. The collocation to be used in the given context is
sajt hzuk.
(6) eukaliptus fk vannak legtbben [the majority/most of the trees are eucalyptus
trees] (vs. SH leginkbb, legnagyobb szmban eukaliptusz fk vannak)
In the English language the terms most and majority meaning greatest in
number, greatest in amount, extent, or degree and in the greatest number of instances
of, the greater number or part; a number more than half of the total can refer to both
animate and inanimate things. As opposed to this, in the Hungarian language one of
their Hungarian equivalents legtbben generally denotes animate creatures/living
things.
(7) rtkes papr [valuable paper] (vs. SH dokumentum, okmny, okirat)
(8) llami paprok [state-owned papers] (vs. SH llami jsgok)
In examples 7 and 8 the semantic extensions of the noun paper are to be considered. In the English language some of its several meaning are the following: a material made of cellulose pulp, derived mainly from wood, rags, and certain grasses, processed into flexible sheets or rolls by deposit from an aqueous suspension, and used
chiefly for writing, printing, drawing, wrapping, and covering walls and a single sheet
of this material. On the other hand it can denote a written or printed paper that bears
the original, official, or legal form of something and can be used to furnish decisive
evidence or information as well as a newspaper. In the two contexts provided by
the subject the latter two meaning are involved which are not characteristic of the
Hungarian language. The two languages share for instance the first meaning of the
word paper mentioned above.
(9) drmt vagy romantikus filmeket nem nagyon kzvettenek [dramas and romantic
film are not very often broadcast] (vs. SH rdin/televziban
ad/kzvett/sugroz/vett)
The English term used in example 9 is broadcast, which can mean to transmit a
radio or television program for public or general use involving any kind of programs.
As opposed to this, in Hungarian the verb kzvetti, with radio or television providing
the context, it generally involves a live broadcast, e.g. a sports event, whereas dramas,
plays and films are shown or played, i.e. vett, sugroz.
324
Winford states that loanshifts may take the form of pure loan translations or
calques in which the foreign model is replicated exactly by native words (Winford
2003: 43); for example:
Loan translations or calques
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
In Winfords (2003: 44) words, creative word formation involving imported items is
another by-product of lexical borrowing, which Haugen includes in his category of
native creations . Pure native creations means innovative use of native words to
express foreign concepts; for example:
3.4. Native creation
Purely native creations (no examples found)
Hybrid creations are blends of native and foreign morphemes to express foreign
concepts, for example:
Hybrid Creations
(1) red back pk [red back spider] (vs. SH vrs ht pk)
(2) history knyvekben [history book+LV+PL+INE] (vs. SH trtnelemknyvekben)
(3) 0record dolgokkal [record thing+LV+PL+INS]; (vs. SH irattri, archvumi dolgokkal)
(4) amerikai sitcom [American sitcom]; (vs. SH amerikai helyzetkomikumra pt
sorozat, tvkomdia)
(5) digital csatornk [digital channel+PL] (vs. SH digitlis csatornk)
325
High
+round
short
long
short
long
i [i]
[i]
[y]
[e]
[]
Mid
Low
Back
-round
short
+round
short
long
[y]
u [u]
[u]
[]
o [o]
[o]
e []
long
[a]
a []
326
harmonic include the facts that there are no non-alternating suffixes with this vowel, as
opposed to [e], [i] or [i], e.g. -rt and -ig and that there are no antiharmonic roots
containing this vowel only as opposed to antiharmonic roots like [hi:d] and [tse:l].
(Antiharmonic roots like hd bridge or cl aim contain neutral (front) vowels but
they take back vowelled alternating suffixes, e.g. hdnak bridge DAT. or clba aim
ILL.)
The literature usually divides the Hungarian vowel harmony phenomena into two
major groups: root and suffix harmony. Native Hungarian roots contain either all back
or all front vowels, though neutral vowels freely cooccur with either. Recent loan
words sometimes violate root harmony and contain both front harmonic and back
harmonic vowels, e.g. [t:ityd] attitude, [rno:] Renault.
The facts of Hungarian suffix vowel harmony are given below for binary, ternary,
and quaternary suffixes respectively. (For the sake of simplicity, the charts list only
monosyllabic roots.)
Back
Front
Back+Neutral
kr+nk
arm
fej+nek
head
rdir+nk
eraser
haz+nk
house
kez+nek
hand
bber+nk
laurel
kor+nk
age
kr+nek
circle
pli+nk
Paul dim
kor+nk
illness
br+nek
skin
koti+nk
cart
juk+nk
hole
sirt+nek
cliff
bui+nk
panty
kut+nk
well
viz+nek
water
putser+nk
naked
ylt+nek
roast
hoher+nk
hangman
tyz+nek
fire
ftser+nk
single
As we can see, stems with back vowels always take back suffixes and generally stems
with front vowels take front vowel suffixes as do stems with only neutral vowels, e.g.
[ke:z], [vi:z], and [sirt] in the second column. Polysyllabic native roots select alternating suffixes according to the backness of the harmonic vowels of the root,
[brat+nk] friend, while loanwords select the alternant according to the backness
of the last harmonic vowel of the root, [t:it:dnek], [rno:nk]. If, however, a back
vowel or a sequence of back vowels is followed by a neutral vowel, the suffix will
often have a back vowel showing that neutral vowels are transparent to backness harmony. (For the behaviour of with words with back vowels followed by one or several
neutral vowels see Ringen and Kontra 1988.)
Back
Front unrounded
327
Front rounded
kr+hoz
arm
fej+hez
head
kr+hz
circle
haz+hoz
house
kez+hez
hand
br+hz
skin
kor+hoz
age
sirt+hez
cliff
ylt+hz
roast
kor+hoz
illness
viz+hez
water
tyz+hz
fire
juk+hoz
hole
kut+hoz
well
Back+Neutral
rdir+hoz
eraser
bber+hoz
laurel
pli+hoz
Paul dim
koti+hoz
cart
bui+hoz
panty
put er+hoz
naked
Ternary harmony only differs from its binary counterpart as far as front roots are concerned. The choice of a back or front ternary suffix is governed by the same rules as in
binary suffixes. Stems that take the back alternant of a binary suffix, i.e. back vowel
stems, antiharmonic stems and stems with sequences of back vowels followed by a
neutral vowel, take the back alternant of a ternary suffix, too. Front roots fall into two
groups based on the final vowel of the root: roots with a final rounded front vowel take
the front rounded alternant, cf. right column, while roots with a final front unrounded
vowel, whether harmonic or not, i.e. [i], [i:], [e:] vs. [], take the front unrounded
alternant. Disharmonic loans behave exactly the same way as native roots.
As we have shown above, while ternary harmony involves the same mechanism as
binary harmony intertwined with a roundedness alternation, quaternary alternations
differ from ternary harmony in that they involve the phenomenon of lowering, a result
of adding quaternary suffixes to members of a morphologically marked closed lexical
class, the so-called lowering stems.
Back
I.
kr+ok
arm
fl+k
wall
bar+ok
bar
haz+k
house
kor+ok
age
hold+k
moon
kor+ok
illness
lov+k
horse
II.
328
rum+ok
rum
juk+k
hole
bur+ok
Boer
kut+k
well
ir+ok
fat
hid+k
bridge
tsel+ok
goal
hej+k
crust
Front
III.
fej+ek
head
kv+ek
book
kez+ek
hand
t +ek
sirt+ek
cliff
viz+ek
water
IV.
kr+k
circle
udder
br+k
skin
ylt+ek
roast
kyrt+k
horn
tyz+ek
fire
byn+k
sin
V.
As we can see in the upper part of Table 4, one of the differences between the ternary
and quaternary suffixes is that there are two different back alternants of the quaternary
plural suffix: a mid alternant, i.e. -ok, and a low alternant, i.e. -k. A similar distinction can be drawn in the case of front alternants: we find a mid, i.e. -k, and a low, i.e.
-k, alternant. However, while the back alternants only differ in height, the front alternants also differ in roundedness, a phenomenon already observed in ternary harmony.
Roots in Group I take the back mid alternant of quaternary suffixes. As can be seen
from the data above, the same vowels, including neutral vowels as well, can be found
in these roots as in stems that take the back alternant of the binary and ternary suffixes.
The same is true of Group II, in which we find roots taking the back low alternant.
Because of the overlap in root vowel quality in the two groups it seems that roots fall
into two discrete morphological classes. Although Group II contains a large number of
roots, it is still much smaller than Group I, which also contains all recent loanwords
and nonce words taking back quaternary suffixes. We also have to note that all foreign
words fall into Group I when used in a Hungarian discourse. Thus we can conclude
that Group I is the default; hence Group II, being anomalous, must be marked in the
lexicon. Front stems are divided into three groups in the above chart although there are
only two front alternants of quaternary suffixes. Of these three groups, two are unmarked and one is marked together with Group II. Roots in Groups III and V behave
in the same way as they do with ternary suffixes: they select the appropriate suffix
based on backness and then choose the one with the roundness value matching the last
vowel of the stem: stems with final rounded vowels select -k, those with final unrounded vowels select -ek. Hence, there is nothing unusual about their behaviour.
However, roots in Group IV behave differently: although containing a final front
rounded vowel they take front rounded ternary suffixes, but they take front unrounded
quaternary suffixes, i.e. rounding harmony is blocked in their case. This phenomenon
329
can be accounted for by claiming that this group is marked for taking low quaternary
suffixes.
Thus we can conclude that stems in Groups II and IV, constituting a grammatically
marked class of stems called lowering stems as opposed to the unmarked stems
shown in Groups I, II, and V, behave differently when taking quaternary suffixes. For
this reason this class must be marked in the lexicon.
4.2. Suffixed derivational loanblends and their possible explanations
In the first group of suffixed derivational loans we find those taking binary suffixes
below:
(5)
driveway-ba
ILL.
suburb-ba
ILL.
degree-jt
culture-k
3PL. POSS.
opportunity-ja
3SG. POSS.
Of these words, opportunityja and drivewayba behave exactly like similar Hungarian
roots would under the same circumstances. For instance, many Hungarian roots containing back vowels followed by at least one neutral vowel will take the back vowel
version of the 3SG. POSS. suffix, e.g. vmpr+ja vampire 3SG. POSS. It seems that
the fact that driveway contains a diphthong in both syllables does not make a difference for the informant as backness harmony skips the /e/ diphthong in the second
syllable just as it would skip a neutral Hungarian front monophthong like /e/ or /i/.
The rest of the words in this group, however, are peculiar in one way or another:
degreejt seems problematic as it is a word with only neutral front vowels and such
words are all to take a front vowel suffix, i.e. degreejt should be the expected form. It
is also important to notice that there is no vowel in the word that could be the source
of this backness feature in the suffix as the root only contains front vowels.
The words culture-jk and suburbba behave in an unexpected way, as they take suffixes with different vowels even though they contain exactly the same non-front stem
vowels only: both stems contain a stressed central-back // followed by a central
unstressed //. In suburb, the back suffix is what one would expect as phonetically
central vowels behave as phonologically back vowels in Hungarian, too cf. the behaviour of the central open /a/ vowel in words like hz+ban /hazbn/ house INE.
Culture, on the other hand, is the odd one out here as it takes a front vowelled suffix
regardless of its back-central stem vowels. There might be two possible explanations
here: according to the first, it might be the spelling of the word that intervenes here as
the final silent vowel letter E would be normally pronounced as /e/ in English, definitely a front vowel. According to the other much weaker explanation, the selec330
tion of the suffix vowel may be due to the typical way schwa, //, is taught in the
English classroom in Hungary: EFL language teachers often describe it as a sound
very close in its articulation to Hungarian //, but it is unrounded an incorrect description as // is front while // is central. However, one might note that this explanation is very unlikely as the informant of the case study was not formally taught English but is a native speaker of the language, so she simply does not have access to this
information. So either the first explanation is correct or some other presently hidden
reason is hiding behind this choice of suffix vowel.
(6)
archive-okkal
PL. + INS.
costume-kbe
PL. + ILL.
recyle-ozzl
supportozza
In this group of derivational blends there are two suffixes following the stem, the second one of which is always completely regular as their backness/frontness values are
determined by the backness/frontness value of the vowel in the preceding suffix: either
the second suffix is back because the first suffix (closer to the stem) is back cf. archive-okkal, recyle-ozzl, supportozza or the second suffix has a front vowel because
of the front vowel of the first suffix cf. costume-kbe. The choice of the first suffix
will be described later below as it is always either a ternary or quaternary suffix.
In the following group we find derivational loans taking ternary suffixes:
(7)
barbicue-n
SUP.
historyhoz
ALL.
charton
SUP.
In this group of words the suffix choice always perfectly conforms to the Hungarian
rules both regarding whether a linking vowel must be present or not cf. the difference between barbicue-n and chart-on and their similarity to native Hungarian stems
like varj+n raven SUP. and sport+on sports SUP. and also regarding the quality
of the linking vowel. In historyhoz, the only non-neutral vowel (in Hungarian terms, of
course) is the unstressed schwa // in the second syllable /hsti/. As we have
already pointed out above, schwa being a central vowel is expected to behave as Hungarian central-back vowels do, i.e. it will attract a back vowelled suffix.
(8)
recyle-ozzl
support-ozza
331
The stems in (8) are followed by a ternary derivational morpheme -oz/ez/z marking
a verb and selecting the suffix vowel just like a native Hungarian root would cf.
monitoroz monitor 3SG. PRES. IND. koproz copy 3SG. PRES. IND. The second
suffix attached after the derivational suffix copies the backness value of the preceding
suffix vowel and thus also behaves completely regularly.
In the last group of derivational loans we find those taking quaternary suffixes:
(9)
restaurant-ok
PL.
apartement-ok
PL.
suburbs-ok
PL.
security guard-ok
PL.
shark-ot
ACC.
camper trailer-t
ACC.
protest-et
ACC.
The stems in this group all behave as Hungarian stems containing identical or similar
vowels both with regards to the presence or absence of the linking vowel cf. camper
trailer+t and kvker+t quaker ACC. as well as shark+ot and park+ot park ACC.
and also with regard to the quality of the suffix vowel cf. suburbs+ok and
kanyar+ok bend PL. vs. protest+et and teszt+et test ACC. One interesting feature
is the double marking of the plural in suburbs+ok where both an English and a Hungarian plural marker are present. Note that as we claimed above in the discussion after
table (4), foreign loans behave regularly and they belong to class I in table (4), i.e they
do not belong to the morphologically marked class of lowering stems.
In the last class of examples there is a quaternary suffix attached to the stem followed by yet another (binary) suffix.
(10)
costume-kbe
PL. + ILL.
archive-okkal
PL. + INS.
The second form in (10) is perfectly regular as the stem with only back vowels, i.e.
// and /a/ is followed back vowels in both the quaternary plural suffix and then the
instrumental. The stem again does not behave as an exceptional lowering stem and
takes a back mid vowel in the suffix. The other stem, costume-kbe, on the other hand
displays some quite peculiar behaviour as the stem containing back vowels, // and
/u/, only takes front vowelled suffixes. This behaviour is exactly the same as the one
332
we saw in culture-k in (5), where we claimed that the most likely reason for this behaviour is the presence of the final silent letter E in the spelling of the word.
5. Conclusion
In conclusion it can be stated that the 3345-word corpus contains 73 interlingual linguistic manifestations, which is approximately 2% of the entire corpus. It cannot be
considered a high percentage of course, but with the exception of the compound
blends and native creations examples for all the other classes can be found in the corpus, which means that the influence of the English language on Hungarian can be
identified in the speech of one representative on the Hungarian community in Australia.
As for the morpho-phonological behaviour of the derivational blends, we can claim
that most of them follow the pattern of native Hungarian words while there are a few
that behave exceptionally, a pattern most likely to be due to the influence of the final
silent letter E in the spelling of these words. Whether or not this is really an effect of
the spelling is to be checked in a follow-up study on a much larger sample.
References
Akadmiai MoBiMouse Plus Angolmagyar, Magyarangol Akadmiai Nagysztr CD-Rom 2003.
Budapest: Akadmiai Kiad.
Csernicsk Istvn 1998. A magyar nyelv Ukrajnban (Krptaljn). Budapest: Osiris MTA
Kisebbsgkutat Mhely.
Fenyvesi Anna 1995. Language contact and language death in an immigrant language: The case of
Hungarian. University of Pittsburgh Working Papers in Linguistics. 3: 1117.
Fenyvesi, A. 2005. Hungarian Language Contact Outside Hungary. Amsterdam: John Benjamins.
Forintos va 2008. Aspects of Language Ecology: EnglishHungarian Language Contact Phenomena in Australia. Veszprm: Pannonian University Press.
Gncz Lajos 1999. A magyar nyelv Jugoszlviban. Budapestjvidk: Osiris MTA
Kisebbsgkutat Mhely Forum Knyvkiad.
Gray, T. S. 2001. Language Contact: An Introduction. Edinburgh: Edinburgh University Press.
Hatoss Anik 2004. Identity formation, cross-cultural attitudes and language maintenance in the
Hungarian Diaspora of Queensland. In Kerr, L. (ed.): Cultural Citizenship: Challenges of Globalisation. Melbourne: Deakin University. 7177.
Kontra Mikls 1990. Fejezetek a South Bend-i magyar nyelvhasznlatbl. [The Hungarian language
as spoken in South Bend, Indiana.] Budapest: MTA Nyelvtudomnyi Intzete.
Kovcs Magdolna 1996. Tulajdonnevek hasznlata az ausztrliai magyar nyelvben. In Mszros Edit
(ed.): nnepi knyv Mikola Tibor tiszteletre. Szeged. 198201.
Kovcs Magdolna 2001a. A szkincs az nekem nem annyira b. Az ausztrliai magyarsg nyelvnek nhny lexikai sajtossga a nyelvfenntarts szempontjbl. Szeged. Nprajz s Nyelvtudomny. 2001/2: 135147.
Kovcs Magdolna 2001b. Code-Switching and Language Shift (in Australian Finnish in Comparison
with Australian Hungarian) bo: bo Akademi University Press.
Lanstyk Istvn 2000. A magyar nyelv Szlovkiban. BudapestPozsony: Osiris Kiad Kalligramm
Knyvkiad MTA Kisebbsgkutat Mhely.
333
Random House Websters Electronic Dictionary and Thesaurus. College Edition Version 1.0 Copyright 1992 Reference Software International.
Sass Blint 2009. Mazsola eszkz a magyar igk bvtmnyszerkezetnek vizsglatra. In Vradi
Tams (ed.): Vlogats az I. Alkalmazott Nyelvszeti Doktorandusz Konferencia eladsaibl.
Budapest: MTA Nyelvtudomnyi Intzete. 117129. http://corpus.nytud.hu/mazsola
Vradi Tams 2002. The Hungarian National Corpus. In Proceedings of the 3rd LREC Conference.
Las Palmas. 385389. http://corpus.nytud.hu/mnsz
Vszolyi Erik 2003. Magyusztrl. Hungarolgia vknyv 4. Pcs.
Weinreich, U. 1953. Languages in Contact: Findings and Problems. New York: Linguistic Circle.
Winford, D. 2003. An Introduction to Contact Linguistics. UK: Blackwell.
Abbreviations
Cases (abbreviated by first few letters capitalized)
Case
Marker
English equivalent
ACCusative
-at/ot/et/t/t
ALLative
-hoz/hez/hz
to
ILLative
-ba/be
into
INEssive
-ban/ben
in
INStrumental
-val/vel
with
SUPeresive
-on/en/n/n
on
Other abbreviations
DerMor
derivational morpheme
InfMor
inflectional morpheme
LV
linking vowel
PL
plural suffix
POSS
possessive suffix
IMP
IND
indicative mood
PRES
present tense
334
1. Hipertextulis szvegrtelmezs
A kltszet, ezen bell klnsen a modern kltszet egyik alapvet problmja a
versszvegek egyre sszetettebb inter- s hipertextulis szervezdse. A modern versek egy jelents rsze esetben a vers jelentstartalmnak megrtshez egyre nagyobb mrtk httrtuds szksges, ennek kvetkeztben a nyelvi szint tuds mellett megnvekszik a metanyelvi szint tuds jelentsge, mivel az rtelmezs sorn
mintegy kilpnk az rtelmezend szvegbl, hogy a szvegek kztti kapcsolatokat vizsglni tudjuk (BodaPorkolb 2011).
A termszetes nyelv szvegek feldolgozsnak alapproblmja gy is megfogalmazhat, hogy egy termszetes nyelv szveg csak a jghegy cscsa, nmagban
nem vagy csak nagyon korltozottan rtelmezhet. A szvegek megrtshez a szksges httrismereteket tbbflekppen is megadhatjuk; pldul verselemzsekkel s rtelmezsekkel, klnsen sszetett mvek esetben pedig (ilyenek pldul a Biblia,
Dante Isteni sznjtka vagy T. S. Eliot tokfldje cm verse) a megrtst elsegt
vagy lehetv tev kommentrokkal. A httrtuds struktrjnak s kapcsolatrendszernek lersra szolgl a jghegy modell, amelynek vzlatos lersa az 1. tblzatban
lthat.
1. Az rtelmezett szveg (a jghegy cscsa)
zenei/vizulis rtegek (a szveg audiovizulis megjelensi formjbl kvetkez, ismtldseken
alapul, nem szemlletes struktrkat tkrz tuds)
hangtani zeneisg (hangok s/vagy sztagok szintjn megvalsul zeneisg: ritmus, rm,
alliterci stb.)
szemantikai vagy paradigmatikai zeneisg (hangok nlkli zene; a szavak szintjn
megvalsul zeneisg: szemantikai hl vagy hlzat, gondolatritmus, egyes gondolatalakzatok
stb.)
intratextulis rteg (szvegen belli kapcsolatok; kohzi, pl. korreferencilis kapcsolatok a
szvegben)
alaprtelmezett kontextus (alapvet rtelmezsi sma, elfeltevsek)
nyelvi szint tuds (a nyelv szintjn llandsult, azaz a nyelvbe beplt jelentsek)
A publikci elksztst a TMOP 4.2.1./B-09/1/KONV-2010-0007 szm projekt tmogatta. A projekt az j Magyarorszg Fejlesztsi Terven keresztl az Eurpai Uni tmogatsval,
az Eurpai Regionlis Fejlesztsi Alap s az Eurpai Szocilis Alap trsfinanszrozsval valsult meg.
335
horizontlis kapcsolatok
(a konkordancikhoz az eredeti szveg
hozzkapcsolsa)
rs 2001-es monogrfijban tbb mint tz oldalon sorolja fel a m vlogatott szakirodalmt). A kutatk a m egyes rszeinek, ill. egsznek rtelmezsre klnbz (s
esetenknt egymsnak ellentmond) smkat s rtelmezseket javasoltak, amelyek az
ltaluk vlasztott paradigmban alighanem egyarnt alkalmasak az tokfldje rtelmezsre taln ppen azrt, mert valban ennyifle gondolhat rla, ennyi potencilis
jelentst tartalmaz (Kappanyos 2001: 14). Ez felveti azt az alapvet krdst, hogy
melyek azok a kritriumok, amelyek alapjn egy javasolt smt elfogadhatunk (vagyis
az rtelmezs szempontjbl rvnyesnek tekinthetjk). Mindez szorosan sszefgg
azzal a krdssel, hogy a hasznlt rtelmezsi smk s a segtsgkkel lert fogalmiszimbolikus kapcsolatrendszerek az rtelmezett mben milyen informciszervezsi
s visszakeressi mechanizmussal rhatak le.
A m eredeti (angol) szvegnek hipertextulis reprezentcija tbb helyen is megtallhat a vilghln (pl. Green 2011, Lancashire 2011, Parker 2011). Azoknak a szvegeknek a listja, amelyek a jelenlegi kutatsok alapjn Eliot mvhez kapcsolhatak,
rszint rendelkezsre ll, rszint folyamatosan bvl (mivel a m jellegbl fakadan
valsznleg sohasem lesz, sohasem lehet teljes). Sokkal nagyobb problma viszont az
eredeti m s a kapcsold szvegek egyttes rtelmezse, azaz a kialaktott
hiperszveg koherencijnak biztostsa az tokfldjvel foglalkoz rsok j rsze
sokkal inkbb az ttteles vagy nylt hivatkozsok puszta felismersre trekszik, s a
m ltalnos rtelmezst, potikai/eszttikai hatst helyezi a kzppontba, mintsem a
hivatkozott szvegek jelentstartalmnak tnyleges hozzkapcsolst, beemelst az
eredeti szvegbe. Az ltalunk kvetett hipertextulis szvegrtelmezs clja az, hogy az
eredeti szveget s a szveghez kapcsolhat szvegeket egyttesen, koherens szvegknt rtelmezzk.
Az albbiakban kt pldn keresztl szeretnnk bemutatni azokat a lehetsgeket,
amelyek egy koherens hiperszveg kialaktsban rejlenek ktnyelv krnyezetben.
Az albbiakban a magyar nyelv szvegrszleteket Vas Istvn fordtsban kzljk
(ld. Eliot 1986).
2.1. Trtnelmi utalsok
And when we were children, staying at the
archdukes,
My cousins, he took me out on a sled,
And I was frightened. He said, Marie,
Marie, hold on tight. And down we went.
Az ismert, hogy Marie Marie Larisch grfnra utal, akivel Eliot szemlyesen is
tallkozott (Eliot 1973: 1984, Eliot 2003: 97); azonban a (valszn) hivatkozs Rudolf koronahercegre s a mben ksbb Erzsbet (Sissi) kirlynre (ltem s srtam a
Genfi-t vizeinl, III/182) mr kevesebb helyen szerepel. Pedig az utalsok a Habsburg-hz tragikusan meghalt tagjaira abban a rszben, amely A halottak temetse
338
alcmet viseli, aligha lehetnek vletlenek. (A fent idzett kommentrokban egy trgyi
tveds is elfordul: a Genfi-tnl Erzsbetet gyilkoltk meg, nem Marie Larisch-t.)
A szlltunk lefel Thomas Mann metaforja alapjn is rtelmezhet: Mlysges
mly a mltnak ktja. A trtnelmi szvegekbl emellett kt gondolatot emeljnk ki.
Rudolf katasztroflis hzassga hozzjrult korai hallhoz; Erzsbetnek pedig, aki
elszr lnya, majd fia halla utn mly depressziba esett, a krtyajsls (v, Tarot,
I/4359) is segtsgre volt a tllsben. Azaz az tokfldjhez kapcsolt tudstartalmak a kollektv, trtnelmi s a szemlyes emlkek mellett a m kt alapvet szemlyes problmjra is felhvjk a figyelmnket: a frfi-n kapcsolat kudarcra s az
tlt tragdik vagy traumk okozta depresszira.
2.2. Bibliai utalsok
Son of man,
You cannot say, or guess, for you know only
A heap of broken images, where the sun beats,
And the dead tree gives no shelter, the cricket no
relief,
And the dry stone no sound of water.
Embernek fia,
Nem mondhatod, nem sejted, mst sem ismersz,
Csak egy csom trt kpet, ahol a tikkadt nap
S a holt fa menhelyet nem ad, tcsk sem
enyhlst,
S a szraz k se csrgedez vizet.
A hivatkozott bibliai szvegrszleteket maga Eliot adja meg (Eliot 1986: 61). Az els
idzetben az embernek fia vagy emberfia Ezkiel prftra utal, a narrtor pedig
Istenre, aki a prfthoz beszl (forrs: Biblia, protestns fordts):
Ezt mondta nekem: Emberfia, llj a lbadra, beszlni akarok veled! Mikzben
beszlt, llek radt belm, talpra lltott, n pedig hallottam, hogy beszl hozzm.
(Ez. 2:12)
Az Eliot ltal megadott msodik idzet az regsgrl s a kzelg hallrl szl:
Mg egy kis emelkedtl is flnek, s ijedeznek az ton. A mandulafa kivirgzik
[a hajad megszl, mint a virgz mandulafa], a sska nehezen vonszolja magt
[ertlenl vonszolod magad, mint a haldokl sska], s mit sem r a fszer [a kapor mr nem kelt tbb szexulis vgyat], mert elmegy az ember rk otthonba,
s az utcn krs-krl siratk jrnak. (Prd. 12:5)
Ebben a pldban mr alapvet szerepet jtszik a ktnyelvsg. Az adott sor ltalunk
ksztett, szgletes zrjelek kztt megadott magyar nyelv fordtsa ugyanis nem
339
rdekes, hogy egyes kommentrokban (pl. Macrae 2001: 15) a trt kpeket utalva
Ez 6:6-ra nem emlkekknt, hanem darabokra trt blvnyokknt rtelmezik (ti.
Isten felszltja Ezkielt, hogy trje ssze a blvnyokat, faragott kpeket). Ez
azonban a szveg egy teljesen ms, pl. vallsi tartalm rtelmezshez vezetne.
A flelem (fear) Vas Istvn fordtsban iszonyatknt jelenik meg ez azrt rdekes, mert A sttsg mlyn fszereplje, Kurtz halla eltti szavait juttathatja
esznkbe Joseph Conrad regnybl: Iszonyat! Iszonyat! (ami Vmosi Pl fordtsban: Borzalom! Borzalom!) Az angol The horror! The horror! ugyan nem egyezik az Eliot ltal az tokfldjben hasznlt fear (flelem) szval, de mellette szl,
hogy Eliot eredetileg Conrad sorait akarta az tokfldje mottjnak vlasztani (Eliot
1997: 150).
Az emberi trtnelem emlkei s az elz rszben kiemelt szemlyes problmk izgalmas paradigmt jelentenek a m tovbbi rszeinek rtelmezshez ami a hivatkozott szvegrszletek miatt alighanem tbb szz oldalt is jelenthet. s mivel vlemnynk szerint Eliot mve klnsen az ltalunk vlasztott hipertextulis szvegrtelmezs megkzeltsmdjt vlasztva szinte kimerthetetlen, gy alighanem itt az
ideje a m hipertextes kiadsnak immron az internet magyar szegmensben is.
Forrsok
Eliot, T. S. 1986. Versek Drmk Macskk knyve. (Jegyzetekkel elltta Takcs Ferenc.) Budapest: Eurpa Kiad.
Green, M. 2011. The Waste Land by T.S. Eliot as hypertext. http://eliotswasteland.tripod.com
/index.html
Lancashire, I. (ed.) 2011. T. S. Eliot: The Waste Land. [Online text with notes.] Toronto: University
of Toronto, Department of English. http://rpo.library.utoronto.ca/poem/790.html
Parker, R. A. 2011. Exploring The Waste Land. http://world.std.com/~raparker/exploring
/thewasteland/explore.html
Magyar Bibliatrsulat jfordts Biblija. (Protestns fordts.) http://www.kereszteny.hu
Online Parallel Bible: Weaving Gods Word into the Web. Biblos.com, 20042010. http://bible.cc
341
Irodalom
Boda Istvn Kroly Porkolb Judit 2011. Egy versszveg rtelmezshez szksges adattartalmak
hipertextulis modellezse. Elhangzott: XIII. Dunajvrosi Nemzetkzi Alkalmazott Nyelvszeti s
Kommunikcis Konferencia. (13th International Conference of Applied Linguistics and
Communications.) Dunajvros, 2011. mrcius 31.
Eliot, T. S. 1973. The Waste Land. In Kermode, F. Hollander, J. (eds.): The Oxford Anthology of
English Literature. Volume II. 1800 to the Present. Oxford: Oxford University Press. 19801998.
Eliot, T.S. 1997. Elveszett kltemnyek. (Fordtotta, szerkesztette, a bevezetst s a jegyzeteket rta
Kappanyos Andrs.) Budapest: Orpheusz K.
Eliot, T. S. 2003. The Waste Land and Other Poems. (Edited with an Introduction and Notes by
Frank Kermode.) Penguin.
Kappanyos Andrs 2001. Ktsges egysg. Az tokfldje s amit tehetnk vele. Budapest: Janus
Kiad Osiris Kiad Balassi Kiad.
Macrae, A. D. F. 2001. Notes to T.S. Eliot: The Waste Land. New York: Longman York Press.
Navracsics Judit 2011. A korai gyermekkori tbbnyelvsg s a kognitv fejlds. In Navracsics
Judit Lengyel Zsolt (szerk.): Lexikai folyamatok egy- s ktnyelv kzegben.
Pszicholingvisztikai tanulmnyok II. Budapest: Tinta Kiad. 95107.
Virgos Zsolt 2008. The Modernist and Others. The American Literary Culture in the Age of the
Modernist Revolution. Debrecen: University of Debrecen, Institute of English and American
Studies.
342
1. Einleitung
Zweisprachigkeit bestimmt das Leben und den Alltag der durch Migration bilingual
gewordenen Menschen. Die erfolgreiche Integration in die Gesellschaft verlangt nach
ausreichenden mndlichen Kompetenzen. Literale Gesellschaften, wie in Ungarn und
in Deutschland, erfordern jedoch auch das Vorhandensein der Schriftlichkeit. Es stellt
sich die Frage, welche Rolle die Literalitt im Falle des Bilinguismus spielt.
Obwohl die ethnische Vielfalt in den Schulen inzwischen als kulturelle
Bereicherung wahrgenommen wird, entstehen im pdagogisch-praktischen Bereich
des fteren Probleme, die zur Bildungsbenachteiligung der Schlerinnen und Schler
auslndischer Herkunft fhren knnen. Die Ursachen vermutet man entweder in dem
soziokulturellen Hintergrund der Zuwandererfamilien oder auch in den oft nicht
zufriedenstellenden sprachlichen Kompetenzen. Deshalb ist es wichtig, die Ressourcen
der Kinder zu evaluieren, die sie fr den Sprachausbau, insbesondere fr
schriftkulturelle Leistungen, nutzen knnen (Maas 2009: 146). Studien zufolge wird
bei dem Prozess der Frderung der Herkunftssprache eine besondere Rolle
zugeschrieben (vgl. Auernheimer 2010).
In dem vorliegenden Beitrag verdeutliche ich die Probleme, die bei der Gestaltung
des ungarischsprachigen herkunftssprachlichen Unterrichts in Deutschland in der
Regel auftreten. Ferner werden einige Transfererscheinungen, die auf das
unterschiedliche Grapheminventar des Ungarischen und des Deutschen
zurckzufhren sind, dargestellt. Die sprachspezifischen Merkmale der PhonemGraphem-Korrespondenz und die Charakteristik der zum differierenden Schriftbild
fhrenden Didaktik werden anhand einiger Beispiele ebenso erlutert.
2. Integrationsmanahmen in Deutschland
Die Daten aus der Erhebung des Baden-Wrttembergischen Statistischen Landesamts
zeigen, dass Deutschland mittlerweile ein attraktives Einwanderungsland geworden
ist. Der Migrantenanteil der Gesamtbevlkerung lag in 2011 bei 19,2%. Die
Eingrenzung der Betroffenen gestaltet sich schwierig. Laut Definition des Statistischen
Bundesamts gelten folgende Bevlkerungsgruppen als Personen mit
Migrationshintergrund:
Bei der Bevlkerung mit Migrationshintergrund handelt es sich um Personen, die
nach 1949 auf das heutige Gebiet der Bundesrepublik Deutschland zugezogen sind,
sowie alle in Deutschland geborenen Auslnder/
343
Der Definition zufolge ist der Besitz eines auslndischen Passes nicht
ausschlaggebend (Jeuk 2010). Sowohl die eingebrgerten Auslnder als auch die
Sptaussiedler haben einen Migrationshintergrund. Entscheidend ist nur, ob in der
Familie (auch) eine andere Sprache als Deutsch gesprochen wird.
Als unumgngliche Voraussetzung fr eine gelungene Integration gilt der
Spracherwerb. Ohne deutsche Sprachkenntnisse haben die Eingewanderten keine
guten Chancen. Die Daten aus PISA 2009 zeigen, dass inzwischen fast 26 Prozent der
fnfzehnjhrigen Schlerinnen und Schler in Deutschland aus zugewanderten
Familien stammen eine Tatsache, die eine immense Herausforderung fr das
deutsche Bildungssystem darstellt (Klieme et al. 2010). Die erworbenen Kompetenzen
der Jugendlichen mit Migrationshintergrund waren zu dem Zeitpunkt der Erhebungen
zwischen 2000 und 2009 zwar gestiegen, aber sie blieben immer noch unter der
Leistung der Schlerinnen und Schler ohne Migrationshintergrund (vgl. Artelt et al.
2001; Stanat et al. 2002; Terhart 2002; Klieme et al. 2010).
Als Reaktion der Bildungsverantwortlichen und um den Betroffenen die
Mglichkeit zu geben, ihre Sprachkenntnisse zu verbessern und sich in das Schulleben
reibungslos eingliedern zu knnen, sind Frdermanahmen eingefhrt worden.
3. Der ungarische herkunftssprachliche Unterricht
Um der wirkungsvollen und effektiven Gestaltung des Unterrichts im schulischen
Rahmen entgegenkommen zu knnen, mssen ausreichend groe Lerngruppen
vorhanden sein. Da die Ungarn in Deutschland nicht so stark vertreten sind wie die
trkische oder die italienische Population, ist die Organisation des
herkunftssprachlichen Unterrichts an den Schulen geradezu unmglich. Angesichts der
Tatsache, dass eine Nachfrage besteht, sind die sogenannten Samstagsschulen
entstanden. Mit Untersttzung des Ungarischen Kulturinstituts (UKI) in Stuttgart und
des Bundes Ungarischer Organisationen in Deutschland (BUOD) wurde im Jahre 2008
ein Versuch gestartet die ungarischen Schulen in Deutschland zu lokalisieren (vgl.
Ills-Molnr 2009). Die Recherchen haben ergeben, dass von den elf ermittelten
ungarischen Schulen diejenigen, die sich in den im Sden liegenden Bundeslndern
befinden, besser besucht werden als die Organisationen im Norden des Landes. Die
Vermutung liegt nahe, dass der Grund dafr die geographische Distanz zu Ungarn ist:
In den sdlichen Regionen des Landes leben mehr Ungarn als im Norden.
Ferner wird die erfolgreiche Vermittlung der ungarischen Sprache auch durch die
Quantitt der Unterrichtseinheiten beeintrchtigt. Eine vorlufige Untersuchung hat
ergeben, dass sich die Gruppen, wenn der Unterricht trotz erschwerter Bedingungen
doch zustande kommt, im Durchschnitt 1 bis 2-mal pro Woche treffen (als Vergleich:
Die Schler haben in der Grundschule durchschnittlich 6 Wochenstunden
Deutschunterricht). Um effektiv arbeiten und die Motivation der Kinder
344
aufrechterhalten zu knnen, mssen das Lehrmaterial und die Thematik auf die vorhin
erwhnten speziellen Bedrfnisse der jeweiligen Lernergruppen abgestimmt sein.
4. Didaktischer Ansatz
4.1. Begriffsbestimmung
Im Zusammenhang mit dem Bilinguismus trifft man in der deutschsprachigen
Fachliteratur des fteren auf die Termini Muttersprache-Fremdsprache/Zweitsprache,
Herkunftssprache-Zweitsprache, Erstsprache-Zweitsprache (L1-L2). Es herrscht kein
Konsens ber den Gebrauch der Begriffe. Gegenstand der vorliegenden Arbeit ist es
nicht, die Begriffe detailliert zu definieren, jedoch ist es unumgnglich zum Thema
Stellung zu beziehen.
Nach der in der einschlgigen Fachliteratur meist erwhnten Definition der
Muttersprache (Skutnabb-Kangas 1997) ist die parallele Existenz mehrerer
Muttersprachen im Leben eines zwei- oder mehrsprachig aufwachsenden Individuums
keine Seltenheit. Die Abgrenzung der Sprachen ist aber aus linguistischer Sicht
notwendig.
Da Fremdsprachen in einer Umgebung gelernt werden, wo sie gewhnlich nicht
gesprochen werden (Henrici et al. 2001), kann man diesen Begriff hinsichtlich der
migrationsbedingten Zweisprachigkeit nicht anwenden. Die Bezeichnung Erst- und
Zweitsprache drckt in dieser Relation eine Rangfolge aus, wobei die Feststellung,
welche Sprache des Zweisprachigen als Erst- oder Zweitsprache zu betrachten ist, ein
Problem darstellt.
Die Herkunftssprache zeichnet spezifische Charakteristika aus. Ihr Gebrauch
konzentriert sich auf die Mikroebene des Sprechers und dient damit nicht der
gewhnlichen Kommunikation in der jeweiligen Gesellschaft (Giay 1997). Der
Zweitspracherwerb erfolgt im Rahmen der Zielgesellschaft (Henrici et al. 2001) und
der Gebrauch fungiert als sprachliches Mittel zur Bewltigung des Alltags.
Ich vertrete die Meinung, dass im Falle des migrationsbedingten Bilinguismus den
Definitionen zufolge die Anwendung der Begriffe Muttersprache, Erstsprache und
Fremdsprache nicht relevant ist. Infolgedessen verwende ich des Weiteren das
Begriffspaar Zweitsprache-Herkunftssprache.
4.2. Die sprachliche Sozialisation
Die sprachliche Konfiguration eines durch Migration zweisprachig gewordenen
Individuums erfordert besonderes Augenmerk (vgl. Navracsics 2004, 2010, Bartha
1999, Grosjean 1989). Wenn man ber Zweisprachigkeit spricht, steht die
Notwendigkeit der Beherrschung der mndlichen Sprache im Vordergrund. Um den
berblick ber die alltglichen Situationen behalten, und sich in der Gesellschaft
zurechtfinden zu knnen, ist das Vorhandensein der Mndlichkeit von existenzieller
Natur.
Bilingualitt bedeutet nicht notwendig zugleich auch Biliteralitt (Siebert-Ott 2006:
536). Dereli (2006) vertritt die Meinung, dass solange die erste Generation der
345
Vorbereitungs-
simultan
Lesen
(rezeptiv)
sukzessiv
simultan
sukzessiv
346
Ein weiterer Punkt ist: In welcher Reihenfolge sollen die Schlerinnen und Schler
in der Anfangsphase die ungarische und die deutsche verbundene Schrift und die
Druckschrift lernen? Sollen sich die Lehrkrfte berhaupt mit dem Thema befassen
oder reicht es aus, wenn die Kinder eine auf der deutschen Schreibschrift basierende
eigene Handschrift entwickeln?
Der Lndervergleich macht deutlich, dass es gravierende Unterschiede in der
Handhabung der Vermittlung der Schreibschrift gibt. In den meisten Bundeslndern
dient die zusammengesetzte, formorientierte Druckschrift als Ausgangsschrift. Mit der
Begrndung, dass der dynamische Charakter der Schreibschrift einen gesonderten
Schreibschriftlehrgang erfordert (Schenk 2009: 150), lernen die Kinder erst im zweiten
Schuljahr die Schreibweise der verbundenen Schrift dazu. Bis dahin knnen sie sie
selbstverstndlich auch nicht lesen. In Ungarn dagegen ist das Erlernen der verbundenen
Schrift und der Druckschrift schon im ersten Schuljahr fester Bestandteil des
Curriculums.
Die Vielfalt der Schreibschriften in beiden Lndern erschwert die Auswahl der
richtigen Methode. Neben der Druckschrift werden in Deutschland drei verschiedene
Ausgangsschriften angeboten: die lateinische Ausgangsschrift (LA), die vereinfachte
Ausgangsschrift (VA) und in den stlichen Regionen des Landes die
Schulausgangsschrift (SAS). In Ungarn unterrichtet man die seit 1956 eingefhrte
vereinfachte moderne Stehschrift mit c-Bindung und die seit 1991 nach Virgvlgyi
benannte leicht nach rechts neigende Schrift. Die Nachforschungen deuten darauf hin,
dass die Notwendigkeit des Erlernens einer ungarischen Schreibschrift mindestens aus
zwei Grnden fraglich ist. Erstens fhrt es zu unntigen Interferenzen in der deutschen
verbundenen Schrift, was gegebenenfalls auf geringe Toleranz seitens der deutschen
Lehrkrfte stt. Zweitens sprechen die wchentliche Stundenzahl und damit die
bungsmglichkeit der ungarischen verbundenen Schrift dagegen.
Die folgenden Beispiele lassen erkennen, wie sich die Handschrift eines auf zwei
Sprachen alphabetisierten Kindes im Laufe der Zeit entwickelt.
Die Abbildung 2. zeigt einen Beispielsatz, der aus dem ersten Schuljahr eines
zweisprachig (ungarisch-deutsch) alphabetisierten Schlers stammt. Das Kind hatte in
seiner Zweitsprache (Deutsch) bis dahin nur die Druckschrift kennengelernt. Das
ungarische Alphabet und damit die Schreibweise der ungarischen verbundenen Schrift
erlernte es vor der Einschulung. Die klaren Strukturen deuten darauf hin, dass zu dem
Zeitpunkt keinerlei Beeinflussungen durch diverse Faktoren bestanden.
Abbildung 2. Beispielsatz 1
Der zweite Beispielsatz (Abbildung 3.) entstand aus dem zweiten Schuljahr, in dem
der Schler die vereinfachte Ausgangsschrift in der deutschen Schule gelernt hat.
347
Schon auf den ersten Blick erkennt man, dass Elemente sowohl aus der deutschen als
auch aus der ungarischen Schreibschrift gleichzeitig vorhanden sind.
Abbildung 3. Beispielsatz 2
Als Vergleich steht hier der dritte Satz von demselben Kind (Abbildung 4.), verfasst in
seiner Zweitsprache, nachdem es die vereinfachte Ausgangsschrift kennengelernt hatte.
Abbildung 4. Beispielsatz 3
Die Form und die Verbindung der Buchstaben zu Wrtern sind im ersten und im
dritten Satz in den meisten Fllen grundverschieden. Der zweite Beispielsatz
unterstreicht die Hypothese, dass das Erlernen verschiedener Schriftarten zu unntigen
Interferenzen fhrt. Die Wahrscheinlichkeit, dass dem Kind mit mehr bung und
gezielter Frderung beide Varianten erhalten bleiben knnen, ist nicht auszuschlieen.
Hinsichtlich der wchentlichen Stundenzahl des herkunftssprachlichen Unterrichts
besteht aber die Gefahr, dass der Vorgang auf Dauer eine berforderung darstellt.
5. Interferenzerscheinungen in der Zweitalphabetisierung
Der Erfolg eines Kindes beim Lesen- und Schreibenlernen hngt letztendlich davon ab,
inwieweit es das alphabetische Prinzip der Schriftsprache begreift. Gem diesem
Prinzip ist die Sprache in eine Anzahl kleinster Lautsegmente (Phone) zerlegbar, die
wiederum durch Schriftzeichen (Grapheme) reprsentiert werden knnen (Kspert
Schneider 2008: 13). Diese Reprsentation uert sich im Regelsystem der PhonemGraphem-Korrespondenz, fr deren korrekte Herstellung unter anderem die
phonologischen Verarbeitungsprozesse verantwortlich sind. Im Falle des zweisprachigen
Schriftspracherwerbs muss man die Koordination von zwei Regelsystemen beherrschen.
Nachdem die gehrten Laute einer der Sprachen zugeordnet werden, ist das richtige
Grapheminventar zu aktivieren und sind die Befehle graphomotorisch korrekt
auszufhren.
Corvacho Del Toro (2004: 37) behauptet, dass es bei der Zweitalphabetisierung von
bilingualen Kindern Fehler bei der Umsetzung von Lauten in Schrift zu beobachten
sind, die auf das erstgelernte Schriftsystem zurckgehen. Im Falle der von uns
untersuchten ungarisch-deutschen bilingualen Kinder spielt eher die Quantitt des
348
Richtige Schreibweise
wis
vz (Wasser)
teisinhab
est kiwanom
medwe
medve (Br)
ttsk
tcsk (Grashpfer)
Auf der Ebene des Satzes findet man ebenso Fehler, die zweifellos auf die
Beschaffenheit der Morphologie der deutschen Sprache zurckzufhren sind
(Abbildung 5.).
349
Schreiben und Lesen auf Ungarisch beizubringen meistens von den Eltern ausgeht,
mssen sie sich darber im Klaren sein, welche Ziele sie damit verfolgen wollen.
Es ist von grter Wichtigkeit, die deutschen Lehrkrfte ber die Biliteralitt der
Kinder in Kenntnis zu setzen. Durch ihr Verstndnis und ihre Akzeptanz lsst sich den
Schlern der Druck nehmen und somit eventuell die bergangsphase, in der
Interferenzen auftreten knnen, deutlich verkrzen.
Der Schriftspracherwerb von Bilingualen unterliegt besonderen Voraussetzungen.
Ihre Schreibfehler besitzen eigene Charakteristika (Corvacho Del Toro 2004: 95), was
beim Alphabetisierungsprozess und Orthographieerwerb unbedingt bercksichtigt
werden muss. Die Zweitalphabetisierung erfordert besondere didaktisch-methodische
Konzepte, deren Grundlage die Analyse und Diagnostik der Fehler in den
Schreibmaterialien der bilingualen Schler bilden.
Um wissenswerte Erkenntnisse gewinnen zu knnen, ist die Durchfhrung weiterer
empirischer Untersuchungen unumgnglich.
Literatur
Artelt et al. 2001. Pisa 2000. Zusammenfassung zentraler Befunde. Berlin: Max-Planck-Institut fr
Bildungsforschung. http://www.mpibberlin.mpg.de/Pisa/ergebnisse.pdf
Auernheimer, G. (Hrsg.) 2010. Schieflagen im Bildungssystem: Benachteiligung der
Migrantenkinder (interkulturelle Studien). Wiesbaden: VS Verlag fr Sozialwissenschaften.
Bartha Csilla 1999. A ktnyelvsg alapkrdsei. Budapest: Nemzeti Tanknyvkiad.
Corvacho Del Toro, I. M. 2004. Zweitalphabetisierung und Orthographieerwerb. Deutsch-spanisch
bilinguale Kinder auf dem Weg zur biliteralen Kompetenz. Frankfurt am Main: Peter Lang
Europischer Verlag der Wissenschaften.
Giay Bla Ndor Orsolya (szerk.) 1998. Magyar mint idegen nyelv. Hungarolgia. Tanknyv s
szveggyjtemny. Budapest: Osiris.
Grosjean, F. 1989. Neurolinguistics, Beware! The Bilingual is Not Two Monolinguals In One
Person. Brain and Languages. 36: 315.
Henrici et al. 2001. Lernen und Lehren von Fremdsprachen: Kognition, Affektion, Interaktion. Ein
Forschungsberblick. Zeitschrift fr Fremdsprachenforschung. 12/2: 1145.
Ills-Molnr Mrta 2009. Minden anyanyelv szrmazsi nyelv, de nem minden szrmazsi nyelv
anyanyelv. Anyanyelv-pedaggia. 2009/3.
Jeuk, S. 2010. Deutsch als Zweitsprache in der Schule. Grundlagen Diagnose Frderung. Stuttgart: Kohlhammer.
Klieme et al. 2010. Pisa 2009: Bilanz nach einem Jahrzehnt. MnsterNew YorkMnchenBerlin:
Waxmann.
Kspert, P. Schneider, W. 2008. Hren, lauschen, lernen. Sprachspiele fr Kinder im
Vorschulalter. Wrzburger Trainingsprogramm zur Vorbereitung auf den Erwerb der
Schriftsprache. Gttingen: Vandenhoeck & Ruprecht.
Maas, U. 2009. Sprache in Migrationsverhltnissen: Sprachausbau (Schriftsprache) vs.
mehrsprachige Kommunikation. In: Gogolin, I. Neumann, U. (Hrsg.): Streitfall
Zweisprachigkeit The Bilingualism Controversy. Wiesbaden: VS Verlag fr
Sozialwissenschaften. 145161.
Navracsics Judit 2004. A ktnyelv gyermek. Veszprm: Pannon Egyetemi Kiad.
Navracsics Judit 2010. Egyni ktnyelvsg. Alkalmazott nyelvszeti mesterfzetek 03. Szeged:
Szegedi Egyetemi Kiad Juhsz Gyula Felsoktatsi Kiad.
350
Schenk, Ch. 2009. Lesen und Schreiben lernen und lehren. Eine Didaktik des Schriftspracherwerbs.
Baltmannsweiler: Schneider Verlag Hohengehren.
Siebert-Ott, G. 2006. Entwicklung der Lesefhigkeiten im mehrsprachigen Kontext. In Bredel et al.
(Hrsg.): Didaktik der deutschen Sprache. Band 1. PaderbornMnchenWienZrich: Ferdinand
Schningh. 536548.
Skutnabb-Kangas, T. 1997. Nyelv, oktats s a kisebbsgek. Budapest: Teleki Lszl Alaptvny.
Stanat et al. 2002. PISA 2000: Die Studie im berblick. Grundlagen, Methoden und Ergebnisse.
Berlin: Max-Planck-Institut fr Bildungsforschung.
Terhart, E. 2002. Nach PISA. Bildungsqualitt entwickeln. Berlin: Europische Verlagsanstalt.
Vkony Valria 2011. Biliterci s ktnyelvsg gyermekkorban. In Navracsics Judit Lengyel
Zsolt (szerk.): Lexikai folyamatok egy- s ktnyelv kzegben. Pszicholingvisztikai tanulmnyok
II. Budapest: Tinta Kiad. 108114.
351
352
1. Introduction
Different languages may contain language-specific elements in their orthography by
means of which they are easily visually distinguished from other languages. Either
their script can differ completely (Latin and Cyrillic alphabets, Chinese characters
etc.), or orthographic elements can vary only slightly within the same script. In this
paper two languages visually differing in the latter manner are studied, namely
Croatian and English. The basis for the script of both these languages is the Latin
alphabet, yet they maintain some language-specific differences. Such differences in
the two orthographic representations of these languages open the possibility for
studying language-selectiveness of the lexical access of speakers who use them. This
is the question of whether the elements of only the appropriate language are accessed
upon the visual presentation of a word in one of the languages a speaker knows
(language-selective lexical access), or whether the elements of both languages sharing
their scripts are activated (language-nonselective lexical access) (Dijkstra 2005). This
problem has been discussed with regard to the nature of lexical processing in bilingual
speakers. Another rather recent group of non-monolingual speakers has emerged in the
past few decades, namely the speakers of Global English. This is considered to be a
growing subdivision of English language speakers in the world emerging due to a
wide use of English as the language of education and business. In this paper an
experimental psycholinguistic research is described in which language-specific
orthographic cues pertaining to either Croatian or English have been used to test the
selectiveness of lexical access in Croatian speakers of Global English.
2. Theoretical background
Previous studies have confirmed that speakers react faster to the stimulus presented in
only one language than to the stimulus presented in two languages (mixed lists) due to
the language-switch cost or the time it takes to switch from one language to another.
The question posed is whether lexical access to the stimulus presented in the mixed
lists can be directed by means of its formal language-specific characteristics so as to
reduce the cost of switching between languages. Experimental design using elements
common to the two languages of a speaker and elements specific to only one speakers
language has been used for this purpose. Such language-specific orthographic cues
were postulated to have the ability to reduce participants reaction times by directing
lexical access to the elements of only the appropriate language, while stimulus lacking
353
such cues would fail to direct lexical access and would therefore cause a prolonged
search of the elements of both participants languages. Conflicting results have been
obtained, but the bulk of recent literature has come to support the nonselective access
hypothesis for bilingual speakers. Different authors have used various differences in
the orthography of the languages they studied to create language-specific cues in their
stimulus. In their often cited study, Grainger and Beauvillain (1987) used
combinations of graphemes specific to French and English and found a reduction in
language switch cost in reactions to language-specific stimuli in the mixed language
condition. However, Thomas & Allport (2000) have shown that Grainger &
Beauvillains study was flawed due to a missing control condition. The 1987 study did
not have language-specific pseudowords in the stimulus used. Therefore, Thomas &
Allport argued that the participants could have developed a strategy by means of
which they would answer positively (word) each time they encountered a languagespecific stimulus as there were no language-specific pseudowords presented in the
experiment. Having controlled language-specific orthographic cues, Thomas & Allport
(2000) found no effect of orthographic cues on lexical access. Hence, they claimed to
have confirmed the language-nonselective access hypothesis. Later studies have
manipulated language-specific cues on the letter level, arguing that this is a more
obvious visual cue than are grapheme combinations (OrifandouSumner 2005).
The methodology used for studying bilinguals can be used to study lexical access in
speakers of Global English as well. Due to the fact that these speakers are not bilingual
and their language processing does not resemble that of bilinguals, they need to be
treated as a unique group of speakers and studied separately. Moreover, their
processing has already been found to differ from that of bilingual speakers as it seems
that their language activation does not change much in different points of the bilingual
continuum (Cergol Kovaevi, in press). When studying the selectiveness of lexical
access of these speakers, the reasoning is the same as in bilingual research: should
language-specific orthographic cues in the stimulus allow for faster reaction times in
the general lexical decision task compared to the reaction times to the stimulus without
these cues, such a finding would provide evidence for the possible language-selective
lexical access in speakers of Global English.
3. Methodology
3.1. Research Aim and Hypotheses
The aim of this study was to investigate whether there were any differences in
participants reaction times to the stimulus with and without language-specific
orthographic cues in the mixed lists. There were three null-hypotheses posed:
Reaction times to the stimulus with language-specific orthographic cues will not
be faster than reaction times to the stimulus without language-specific
orthographic cues.
There will be no more erroneous answers given in the category without
language-specific orthographic cues than in the category with language-specific
orthographic cues.
354
Common
a, b, c, d, e, f, g, h, i, j, k, l,
m, n, o, p, r, s, t, u, v, z
English-specific
q, w, x, y
double letters
The Croatian letters lj and nj and English double letters are not graphically languagespecific in the same way other letters are. In fact, it is the combination of the
graphemes they consist of that is language-specific. Therefore, language-specific
orthographic cues used in this study consist of language-specific letters (as have been
used by OrifandouSumner 2005) and language-specific combinations of graphemes
(as used in GraingerBeauvillain 1987; and ThomasAllport 2000). Nouns containing
language-specific letters or grapheme combinations were used to create lists of stimuli
355
with language-specific orthographic cues, while they were avoided in the lists without
language-specific cues.
Five to eight letter long nouns consisting of two syllables were chosen for the
stimulus. The first syllable of the words carried word stress. Their frequency was 100
and more per million (MoguBrataniTadi 1999; LeechRaysonWilson 2001).
The words chosen did not have a more frequent orthographic neighbour (BNC 2007,
HNK-v25 2005). As cross-language influence of Croatian on English was expected
(LemhferDijkstra 2004), the English nouns chosen were checked for orthographic
neighbourhood in the Croatian language. Only English nouns without a Croatian
neighbour were used in the study. The nouns chosen were divided into two groups for
each language: words with language-specific orthographic cues and words without
language-specific orthographic cues. Nouns with language-specific orthographic cues
did not begin with a language-specific letter due to the greater role the first letter plays
in lexical access (Erdeljac 1997). In the group of stimulus without language-specific
orthographic cues care was taken to ensure that combinations of phonemes and
graphemes at the beginning, at the end, and within a word were possible in the other
language so that access to the elements of both languages in participants mental
lexicons would be stimulated. Cognates and semi-cognates (de Bot et al. 1995) were
avoided. The participants carried out a familiarity test for the words which matched
the above stated requirements. Only the most familiar words were used for the study.
Pseudowords were designed by substituting the first phoneme of the existing words
used in the stimulus with a phoneme differing from the original one in four distinctive
features. Again, two groups of pseudowords were produced for each language:
pseudowords with language-specific cues and pseudowords without them. In
pseudowords without language-specific cues the combinations of letters / phonemes
produced in such a way were legal in both languages. In general, the design of
pseudowords reflected the design of words in the stimulus so as to provide equal
conditions in both word and pseudoword part of the stimulus. The design of the
stimulus is presented in table 2.
WORDS
CROATIAN
ENGLISH
PSEUDOWORDS
WITHOUT language-specific
orthographic cues
12
WITHOUT language-specific
orthographic cues
12
WITH language-specific
orthographic cues
12
WITH language-specific
orthographic cues
12
WITHOUT language-specific
orthographic cues
12
WITHOUT language-specific
orthographic cues
12
WITH language-specific
orthographic cues
12
WITH language-specific
orthographic cues
12
TOTAL
48
TOTAL
48
356
Table 2. Stimulus design and number of items (N) per each group of stimuli
3.4. Task
A general (mixed) visual lexical decision task was prepared in which participants were
presented with Croatian and English stimulus which contained language-specific
orthographic cues and the stimulus without such visual cues. The participants were
seated at computers separated by booths so as to ensure undisturbed performance at
the experiment. After a set of instructions, they were presented with a number of
stimuli used for practice, to which RTs were not measured. Further on they were
presented with the experimental stimulus in Font Courrier New 18, which came up on
screen in random order, preceded with a 1000 ms fixation. Upon the presentation of
the stimulus, the participants were required to decide whether the stimulus on the
screen was a word in one of their languages or not by pressing keys 1 and 2 with
the index and middle fingers of their more dexterous hand (right or left if they were
left-handed). Half of the participants were instructed to use key 1 for word and 2
for pseudoword and the other half used reversed keys. E-prime psychological tool was
used for the presentation of the experiment and measurement of reaction times to
participants answers (SchneiderEschmanZuccolotto 2002).
4. Results and discussion
4.1. Hypothesis 1
In data analysis t-test and Wilcoxon test were used depending on whether data
followed normal distribution or not. Kolmogorov-Smirnov test was used to check the
normality of data distribution. The first hypothesis was confirmed as there was no
difference in reaction times to any of the categories studied: Croatian stimulus (both
words and pseudowords) with language-specific cues and Croatian stimulus without
the cues (t=0.207; Sig.=0.836), English stimulus with the cues and English stimulus
without the cues (t=-0.492; Sig.=0.624), Croatian words with the cues and Croatian
words without the cues (t=-0.842; Sig.=0.403), English words with the cues and English words without the cues (t=-0.900; Sig.=0.371), Croatian pseudowords with the
cues and Croatian pseudowords without the cues (t=0.908; Sig.=0.367), English
pseudowords with the cues and English pseudowords without the cues (Z=-0.488;
Sig.=0.625).
In speakers of two languages, both languages are active in the course of mixed lists
processing. In fact, it has been shown that even when processing only one language,
the other language of such speakers is still active and can cause interference or
facilitation (e.g. in cognate processing) (Dijkstravan Hell 2003). Despite this parallel
activation of two languages, speakers are perfectly capable of keeping their languages
separated. Therefore, it is natural to assume the existence of some kind of an indicator
which could facilitate mixed language processing by directing it to the target language.
However, the results of hypothesis 1 support the hypothesis of language-nonselective
access as language-specific orthographic cues did not direct lexical access to the
357
elements of a particular language and so did not reduce the language switch cost.
There was no facilitation of language switch despite the rigorous controls applied in
stimulus selection and creation as well as the balanced stimulus design in which all
aspects of words chosen for the study were reflected in the design of the pseudowords
as well. It appears that the target language was chosen only in the later stages of
processing. The explanation of the language switch cost phenomenon provided by the
Bilingual Interactive Activation model (van HeuvenDijkstraGrainger 1998) is that
the elements of the currently used language have an advantage over the elements of
the currently inactive language. This is because the activation of the element
belonging to the active language spreads to the rest of the elements of the same
language. The supporters of this view see the cause of the language switch cost as
stemming from within the lexicon.
On the other hand the supporters of the view that the language switch cost stems
from outside the lexicon believe that it is the product of the control structures which
coordinate lexicon activation. This view is woven into the phenomenon of the
language task schema in Greens Inhibitory Control Model (Green 1998). Language
switch takes more time as it presupposes a change in the language task schema. In
order for one language task schema to function, the other one needs to be inhibited and
the time needed for the recovery of the inhibited language task schema produces the
switch cost (ThomasAllport 2000).
Both views are based on similar grounds. According to the first view, one set of
presentations has an advantage of the other set within the lexicon, while according to
the second view, it is one language task schema that has an advantage over another
outside of the lexicon (ibid.). It is also possible that both types of processes are under
way in lexical access.
In reaction times to pseudowords, which lack semantic and conceptual content,
orthographic representation was expected to have a greater role in processing.
However, no such effect was found. In reactions to the words used in the study, the
lack of statistical difference between reaction times to the stimulus with and without
orthographic cues could have been interpreted as resulting from the high frequency
and familiarity of the stimulus, which could have enabled a direct access to the lexical
entry, thereby cancelling the possible effect of the orthographic cues. Words have a
lexical entry and while contacting the appropriate entry the semantic information
which the word carries gets activated as well. It is possible that the role of the
orthographic information gets diminished in the process. Due to their lack of semantic
information, this does not apply to pseudowords. Therefore, the results showing no
difference in reaction times to pseudowords confirm the language-nonselective access
hypothesis in the early stages of lexical access.
4.2. Hypothesis 2
According to the Chi-squared test results, there was no statistically significant
difference found between the numbers of erroneous answers made between the
stimulus with and without language specific orthographic cues for the following
categories: Croatian words ( 2=0,474; p>0.05), Croatian pseudowords ( 2=1.806;
358
p>0.05), and English words ( 2=0,726; p>0.05). There was a statistically significant
difference found for English pseudowords where there were more erroneous answers
given in the category of English pseudowords with language-specific orthographic
cues than in the category without the cues ( 2=4.32; p<0.05). This is curious as it
shows that language specific orthography has hindered the processing of English
pseudowords, resulting in a statistically greater number of errors in this category. It is
interesting to notice the pattern of error analysis results; there were fewer mistakes
made in reactions to both English and Croatian words without language-specific
orthographic cues, while the opposite was true for pseudowords in both languages
(table 3). Results of the second hypothesis seem to suggest that language-specific
orthographic cues do not seem to facilitate the processing of pseudowords.
Words
Croatian
English
Sd
Without cues
11
611.60
104.12
With cues
879.00
532.55
Without cues
13
613.95
153.15
With cues
504.97
87.48
sd
Without cues
661.83
246.12
With cues
14
807.32
499.75
Without cues
25
952.59
514.86
With cues
42
1101.72
640.76
Pseudowords
Croatian
English
Table 3. Error analysis number (N), mean (M), standard deviation (sd) of stimulus groups with and
without language specific orthographic cues in Croatian and English words and pseudowords
4.3. Hypothesis 3
The third hypothesis was confirmed for most of the categories studied. There was no
statistical difference in reaction times between the stimulus with and without
language-specific orthographic cues found in the following categories: Croatian
stimulus (both words and pseudowords) with language-specific cues and Croatian
stimulus without the cues (t=-1.873; Sig.=0.073), English stimulus with the cues and
English stimulus without the cues (t=-1.306; Sig.=0.197), Croatian words with the
359
cues and Croatian words without the cues (t=-1.574; Sig.=0.138), Croatian
pseudowords with the cues and Croatian pseudowords without the cues (t=-0.656;
Sig.=0.523), English pseudowords with the cues and English pseudowords without the
cues (t=-0.874; Sig.=0.387). However, there was a statistically significant difference
found in reaction times to the erroneous answers to the English words with the cues
and English words without the cues (t=2.119; Sig.=0.046). Participants reactions to
the English words with the cues were faster than their reactions to the English words
without the cues.
According to these results, orthography does seem to play a role in lexical access
after all. However, it seems that all (or most) information about a word (orthographic,
semantic, phonological) is available already in the early stages of processing and one
should not forget the great influence frequency and familiarity of the word play in
lexical access. The mentioned (and other) factors combined might diminish the role of
orthographic cues in lexical access. The role of orthography has been confirmed on the
example of the erroneous answers in the results of the third hypothesis. It is curious
that such results should be achieved only for English. The explanation might lie in the
description of participants as speakers of Global English. Vaid and Frenck-Mestre
(2002) report on two facts concerning the processing of language-specific
orthographic cues in bilingual speakers: 1) that speakers are more efficient in
perceiving the orthographic cues of their non-dominant language, as is English in the
case of this study; and 2) that late bilinguals are particularly sensitive to such cues.
Participants in this study had all started learning English before puberty, which might
have influenced their lack of sensitivity to language-specific orthographic cues in the
case of correct answers. Moreover, the fact that the difference between reaction times
in erroneous answers to the stimulus with and without the cues was found in English
(the non-dominant language of the speakers) and not in the (dominant) Croatian,
supports Vaid and Frenck-Mestres (ibid.) first statement.
5. Conclusion
The role of the language-specific orthographic cues in Croatian speakers of Global
English has been confirmed but is not directly obvious. The results of the first
hypothesis support the hypothesis of language-nonselective lexical access for the
studied group of speakers in the early stages of their lexical processing of mixed
language lists. However, error analysis shows a (conditionally) positive influence (as
errors are analysed) of language specific cues on the processing of both English
(hypothesis 2 and 3) and Croatian words (hypothesis 2) and the inhibitory influence on
the processing of both Croatian and English pseudowords (hypothesis 2). Generally, it
seems that in the processing of pseudowords, language-specific orthographic cues
confuse the speakers. In the processing of words, semantic, phonological and other
lexical information seems to be available already in the early stages of lexical
processing, rendering the orthographic cues unnecessary and causing parallel access to
elements of both speakers languages, at least in the early stages of lexical processing.
360
Sources
Hrvatski nacionalni korpus (HNK-v25) 2005. Zavod za lingvistiku Filozofskog fakulteta Sveuilita u
Zagrebu. http://www.hnk.ffzg.hr/default.htm
Leech, G. Rayson, P. Wilson, A. 2001. Word Frequencies in Written and Spoken English (based
on the British National Corpus). London: Longman.
Mogu, M. Bratani, M. Tadi, M. 1999. Hrvatski estotni rjenik. Zagreb: Zavod za lingvistiku
Filozofskog fakluteta u Zagrebu i kolska knjiga.
Schneider, W. Eschman, A. Zuccolotto, A. 2002. E-Prime Users Guide. Pittsburgh: Psychology
Software Tools, Inc.
Schneider, W. Eschman, A. Zuccolotto, A. 2002. E-Prime Reference Guide. Pittsburgh:
Psychology Software Tools, Inc.
The British National Corpus, version 3 (BNC XML Edition) (2007) Distributed by Oxford University Computing Services on behalf of the BNC Consortium. http://www.natcorp.ox.ac.uk
References
Cergol Kovaevi, K. (in press). Global English on Language Mode Continuum. In Karabali, V.
Aleksa Varga, M. Pon, L. (ed.): Applied Research Today: Research and Perspectives. Osijek:
Filozofski fakultet Sveuilita J. J. Strossmayera.
Council of Europe 2001. Common European Framework of References for Lanugages: Learning,
teaching, assessment. Cambridge: Cambridge University Press.
de Bot, K. Cox, A. Ralson, S. Schaufeli, A. Weltens, B. 1995. Lexical processing in
bilinguals. Second Language Research. 11/1: 119.
Dijkstra, T. van Hell, J. G. 2003. Testing the Language Mode Hypothesis Using Trilinguals.
Internatinal Journal of Bilingual Education and Bilingualism. 6/1: 216.
Dijkstra, T. 2005. Bilingual Visual Word Recognition and Lexical Access. In Kroll, J. F. de Groot,
A. M. B. (ed.): Handbook of Bilingualism: psycholinguistic approaches. Oxford: Oxford University Press. 179201.
Erdeljac, V. 1997. Prepoznavanje rijei. Zagreb: SOL/Ibis.
Grainger, J. Beauvillain, C. 1987. Language blocking and lexical access in bilinguals. The
Quarterly Journal of Experimental Psychology. 39/A: 295319.
Green, D. W. 1998. Mental Control of the Bilingual Lexico-Semantic System. Bilingualism:
Language and Cognition. 1/2: 6781.
Grosjean, F. 1994. Individual Bilingualism. Reprint. In Lengyel, Zs. Navracsics, J. Simon, O.
(eds.): Applied linguistic Studies in Central Europe. Vol. I. Veszprm: Faculty of Foreign
Langugaes, University of Veszprm. 103113.
Lemhfer, K. Dijkstra, T. 2004. Recognizing cognates and interlingual homographs: Effects of
code similarity in language specific and generalized lexical decision. Memory & Cognition. 32:
533550.
Orfanidou, E. Sumner, P. 2005. Language switching and the effects of orthographic specificity and
response repetition. Memory & Cognition. 33/2: 355369.
Thomas, M. S. C. Allport, A. 2000. Language Switching Costs in Bilingual Visual Word
Recognition. Journal of Memory and Language. 43: 4466.
Vaid, J. Frenck-Mestre, C. 2002. Do orthograhic cues aid language recognition? A laterality study
with French-English bilinguals. Brain and Language. 82: 4753.
van Heuven, W. J. B. Dijkstra, T. Grainger, J. 1998. Ortographic neighborhood Effects in
Bilingual Word Recognition. Journal of Memory and Language. 39/3: 458483.
361
362
1. Introduction
The present study is concerned with two main questions of bilingual lexical processes.
The first, and one of the most debated questions is whether lexical access is language
selective or not. The second, highly related question is to know how the coactivation
of both languages influences lexical activation and identification.
Among the cognitive models accounting for empirical findings on bilingual lexical
access, Interactive Activation models are viewed as the more developed ones. Models
of this kind of visual and auditive bilingual word recognition give different answers to
our questions. Two competing models will be presented here briefly as possible
theoretical backgrounds for the present research: the Bilingual Interactive Activation
model of visual word recognition and the Bilingual Model of Lexical Access, a model
of bilingual speech perception.
The first of the two models, The Bilingual Interactive Activation model, developed
by Dijkstra and Van Heuven (GraingerDijkstra 1992, van Heuven et al. 1998,
Dijkstra 2005, Thomasvan Heuven 2005) is an algorithmic model of bilingual word
recognition that implements non-selective bottom-up processing. That means that
letters activate words from both languages in an integrated lexicon. The model also
implements a language-specic top-down processing: language nodes, situated at the
top of the model, selectively inhibit the activation of words of the other language.
According to the model, L2 lexical access is strongly determined by proficiency. In
the case of unbalanced, L1 dominant bilingualism, L2 words can be activated more
slowly during both perception and production. Although the L1 language node inhibits
more the activation of L2 words than vice versa, experimental results show that the L2
lexicon cannot be switched off even if experimental instructions require the use of L1.
Cross-language interference might be explained by the language non-selective nature
of bilingual lexical processing.
As opposed to this, the other model, the Bilingual Model of Lexical Access
(BIMOLA), assumes separate lexicons for each language, which means that during
speech perception, L1 words compete only with other L1 words and L2 words with
other L2 words. Thus bilingual word recognition is language specific (Thomasvan
Heuven 2005, Grosjean 2008).
In the BIMOLA model, there are no language nodes, but network activation is
distributed over all the nodes of a language:
if a word in a given language is activated, it is probable that it is that language
which is being spoken. Thus, each word sends a small positive signal to the other
words in the same language. The more words in a language that are active, the
363
more these signals will be exchanged, and therefore the greater the activation of
the word subset as a whole (Grosjean 2008: 207).
In the present study the visual recognition of L1 and L2 words was investigated in
language-specific and generalised lexical decision tasks. The most important research
aim was to find out how the recognition process of L1 and L2 words was influenced
by the co-activation of both languages. (The experiments and results presented here
are part of a broader research project.)
The hypotheses were the following: (1) In a generalised lexical decision task when
both L1 and L2 words are presented, response latencies will be longer compared to
language-selective experimental situations when only L1 or L2 words are presented.
(2) The increase in reaction time is larger in the case of L2 than in the case of L1
words.
These hypotheses are more compatible with the first of the two models presented
previously, i.e. the Bilingual Interactive Activation model.
2. Method, participants and stimuli design
The experimental technique applied in the research was visual lexical decision. To
measure the response latencies and decision accuracy in the visual word recognition
tasks I used a dedicated software tool that I developed myself. Participants were asked
to decide whether the letter string appearing on the screen was an existing word or not.
Positive responses were communicated by pressing the mouse button, while for a
negative response participants had to press the spacebar. Reaction times and accuracy
were measured and saved after each experiment and were subject to further analysis.
Stimuli were presented with lower case letters on a light green background.
Participants always received immediate feedback on the speed and correctness of their
response.
40 native speakers of Hungarian with English as their second language participated
in the experiment. They had all learned English at school or university and they all had
Hungarian as their dominant language. Their age was between 15 and 37 years with a
mean age of 25.6 years. They all had normal or corrected to normal view.
As the repetition of stimuli had to be avoided I divided the participants into two
groups. Items presented in the mixed list of the generalised lexical decision task in one
group appeared in the language specific Hungarian or English lexical decision task
of the other group. Half of the participants encountered English and half of the
participants encountered Hungarian words in the first, language specific task.
This way the effect of different experimental situations can be examined by
comparing reaction times to the same words and non-words tested in different tasks.
For the three different experimental conditions Hungarian, English and
generalised lexical decision 6 lists composed of one or two syllable words and wordlike non-words were elaborated. For the L1 and L2 language-specific tasks I also
developed two shorter lists for the trial sessions. In the L1 lexical decision task only
L1 words, in the L2 lexical decision only L2 words were presented. Lists of items used
in the generalised lexical decision task were composed of words and non-words
364
appearing also in the language-specific experiments. English words and half of the
non-words used in the present study were selected from the database of the English
Lexicon Project that investigates the recognition of English words in word naming and
lexical decision tasks (Balota et al. 2007).
3. Results
Comparing the results of the language specific tasks, mean response latencies were
significantly shorter in L1 lexical decision in the case of both words and non-words,
see the frequency histograms shown in Figure 1 (difference between Hungarian and
English words: group A: 29 ms, one-sample t-test: t(19)=2,929 (p=0,0086)**; Johnson test: J(19)=2,511 (p=0,0213)*; group B: 55 ms, one-sample t-test: t(19)=6,795
(p=0,0000)***; Johnson test: J(19)=5,707 (p=0,0000)***; difference between nonword responses in Hungarian and English tasks: group A: 60 ms, one-sample t-test:
t(19)=5,310 (p=0,0000)***; Johnson test: J(19)=6,834 (p=0,0000)***; group B: 53
ms, one-sample t-test: t(19)=4,058 (p=0,0007)***; Johnson test: J(19)=6,178
(p=0,0000)***). These results coincide with the findings of a Dutch-English visual
lexical decision task when Dutch participants responded faster in the L1 experiment
(de Groot et al. 2002).
In the generalised lexical decision task when both Hungarian and English words were
presented, comparing the reaction times for the three different kinds of stimuli
(Hungarian words, English words and non-words), it appears that response latencies
were shorter in the case of Hungarian words and standard deviation was smaller than
in the case of English words and non-words (Figure 2). Comparing the mean reaction
times, participants responded significantly faster to Hungarian words than to English
words or to non-words (difference between Hungarian and English words: group A:
99 ms, one-sample t-test: t(19)=7,272 (p=0,0000)***; Johnson-test: J(19)=6,764
(p=0,0000)***; group B: 92 ms, one-sample t-test: t(19)=9,922 (p=0,0000)***;
365
Figure 2. Relative frequency of reaction times in the HungarianEnglish generalised lexical decision
task
366
367
The increase in reaction times compared to the Hungarian lexical decision task affects
non-words that are also orthographically Hungarian-like (e.g. vgylcs, abrty, pny).
It seems that orthographic cues (such as language specific word start, word ending or
diacritics) do not facilitate the rejection of non-words in a generalized lexical decision
task.
The direction of the difference is the same when comparing response latencies for
the same non-words in English and generalised lexical decision, but the 24 ms
difference did not reach significance (Figure 4).
There are two exceptions which might require further attention. Although the
experimental situation did not have a significant effect on the recognition of L1 words,
two of the Hungarian words were recognised significantly more slowly when
presented in a bilingual list in generalised lexical decision. In the case of the
Hungarian word hely (place, location) the activation of the L2 lexicon might have
had a slow-down effect on the recognition of the orthographically neutral L1 word.
This effect might be explained by the existence of frequent English neighbours (help,
holy, hell) and the English-like word ending which is rare in Hungarian.
As for the other exceptional word, the Hungarian word kerek (round) was
recognised more slowly in the generalised lexical decision task. It was presented a few
items after the word kerk (wheel). The similarities of the word stimuli presented in
the same list might have produced a slow down effect as well.
4. Control study
A FrenchHungarian control study was designed to confirm the most important
findings of the main research. The method and the stimulus material are highly
comparable to those of the EnglishHungarian study presented earlier, but this time
only 16 subjects participated in the experiment, all native speakers of Hungarian who
had learned French as a second language.
368
This time only those items that were correctly responded to by at least half of the
participants were included in the analysis.
As shown in Figure 5, the experimental situation did not have a significant effect on
the activation of L1 words based on reaction time latencies, whereas L2 (that is,
French) words were more slowly responded to when presented in a generalised lexical
decision task (t(18)=4,162 (p=0,0006)***; J(18)=5,048 (p=0,0001)***). This is in
accordance with the results of the EnglishHungarian experiments.
Non-word responses of the control study show the same pattern as in the English
Hungarian experiments (Figure 6). While non-words presented in the Hungarian
lexical decision task were more slowly responded to in generalised lexical decision
(t(19)=6,722 (p=0,0000)***; J(19)=6,244 (p=0,0000)***), the experimental situation
did not have a significant effect on non-word stimuli presented also in the French task.
As in the main research, in the case of non-words the reaction time results of the
generalised lexical decision task are quite similar to those of the L2 lexical decision
task.
369
5. Conclusion
In conclusion, L2 word responses require more time in generalised lexical decision
than in L2 language specific lexical decision tasks. Consequently, the identification of
L2 words is more difficult when the L1 lexicon is also kept activated. The same
experimental condition does not have a significant slow down effect on the recognition
of L1 words.
As for the non-words, there is a significant difference between the mean reaction
time for non-words in Hungarian versus generalised lexical decision tasks. In the case
of L2 no such relationship between non-word responses and experimental conditions
could be proven. Non-word responses in the generalised lexical decision task are
based on the searching of the L2 lexicon: a correct non-word response is possible only
if the presented stimulus could not be found in either L1 or L2 lexicons. The increase
in reaction times, compared to the Hungarian lexical decision, affects also
orthographically Hungarian-like non-words. The response latencies for non-words do
not seem to depend on the orthographic characteristics of the stimuli.
Based on the results, we can say that for second language learners lexical process
is language specific, in accordance with the Bilingual Model of Lexical Access.
Reaction time differences suggest that the base language during the generalised
lexical decision task was Hungarian and the L1 and the L2 parts of the lexicon were
searched separately.
A possible decision mechanism of the generalised lexical decision task can be
formulated based on the reaction time data. When a letter string appears on the screen,
subjects decide first whether it is a Hungarian word or not. If it is, they can make a
positive decision. That is why L1 word responses do not require more time in
generalised than in language specific tasks. If the stimulus appears not to be a
Hungarian word (i.e. it cannot be found in the L1 part of the lexicon), then the
verification in the English lexicon starts. If the presented letter string cannot be
370
Hungarian word?
NO
YES
word
response
English word?
NO
YES
non-word
response
word
response
In the case of the Hungarian word hely, the recognition process requires more time in
the generalised lexical decision task. The difference might be explained by the coactivation of orthographically similar English neighbours (holy, hell, help). However,
it is also possible that because of the language-neutral orthography and the more English-like word ending, participants decide to search for it among the English words
before it could reach its activation threshold in the Hungarian part of the lexicon.
When participants make sure that the stimulus is not an existent English word, they
give a non-word response (that is why there are much more errors in the case of the
word hely in the generalised than in the Hungarian lexical decision task), or they reexamine the Hungarian part of the lexicon and, after the identification, they give a
word response. Thus a more complicated verification process is reflected in longer
reaction times.
References
Balota, D. A. Yap, M. Cortese, M. Hutchison, K. A. Kessler, B. Loftis, B. Neely, J. H.
Nelson, D. L. Simpson, G. B. Treiman, R. 2007. The English Lexicon Project. Behavior Research Methods. 39: 445459.
de Groot, A. M. B. Borgwaldt, S. Bos, M. van den Eijnden, E. 2002. Lexical Decision and
Word Naming in Bilinguals: Language Effects and Task Effects. Journal of Memory and
Language. 47: 91124.
Dijkstra, T. 2005. Bilingual Visual Word Recognition and Lexical Access. In Kroll, J. F. de Groot,
A. M. B. (eds.): Handbook of Bilingualism. Psycholinguistic Approaches. OxfordNew York:
Oxford University Press.
Grainger, J. Dijkstra, A. 1992. On the representation and use of language information in bilinguals.
In Haris, R. J. (ed.): Cognitive Processing in Bilinguals Amsterdam: Elsevier Science Publishers
B.V. 207220.
371
372
This paper describes a new picture set created for the study of lexical and
grammatical aspects of language learning. The set consists of digitized cartoon
drawings of 20 animals performing 10 different picturable actions in dyadic pairs. The
animals and actions can be combined freely to create a large number of different
scenes corresponding to independent clauses of the type The dog hugs the lion. In
the present experiment, twenty Finnish-speaking university undergraduates learned
novel names for these animals and their actions, as well as two morphosyntactic rules
(object marking vs. grammatical gender marking) embedded in the new language.
Unsupervised learning was accomplished by viewing scenes with accompanying
sentences like Garomi poh+a tunuke+r. Both lexicon and grammar were learned
simultaneously, thus aiming at a more natural language learning process.
Passive vocabulary developed fast but early on language learners had difficulties in
producing the new words and the rules. Overall, syntactic learning was more difficult
than lexical learning. An additional metalinguistic task showed that more participants
became aware of the native-like morphosyntactic rule than the gender-marking rule.
This study attests to the efficacy of the language learning system in adults: limited
exposure to novel items is sufficient for lexical learning and acquisition of some
embedded regularities, while explicit metalinguistic knowledge of the regularities does
not develop until later. More generally, our research demonstrated that the Animal
Farm can be used to study basic aspects of lexical and grammatical learning in a wide
variety of setups.
Humans produce correct grammatical sentences in their native language most of the
time, but very often they are not able to explain to a foreigner what the exact
grammatical rules are. This reflects the difference between implicit and explicit
learning. Implicit learning is a very important aspect of learning as it is needed not
only in motor, but also in social and cognitive skills (Lieberman 2000; Ullman 2004).
There has been an explosion of psychological investigations into implicit and explicit
learning in recent decades (e.g. Dienes et al. 1991; Braine 1987; Braine et al. 1990;
Hulstijn 2005; Robinson 2005). Implicit learning occurs without any explicit
instruction and in such a way that acquired knowledge is difficult to express verbally
(Reber 1989). However, in everyday life, when performing various tasks (e.g typing,
learning a foreign language) implicit, nonconscious and explicit, conscious processes
I am very grateful to the co-creator of the Animal Farm picture series, Professor Matti Laine,
for his helpful comments on this paper. Many thanks to Anastasia Leikas for her precious work
during this research.
373
375
376
The training session was followed by several tasks testing language learning. The
lexical learning tasks included a picture-naming task and a picture-word matching
task. Lexical and syntactic learning was followed up by a picture-sentence matching
task and a picture description task.
For the picture-naming task 30 pictures were used, including all the animals one-byone (naming of nouns in the new language) and pictures of a single animal performing
the 10 actions (naming of verbs in the new language). For the picture-word matching
task 60 pictures were used: 20 depicting the nouns (animals) matched with the correct
name of the noun, 20 depicting nouns matched with an incorrect name (wrong
animal), 10 depicting the actions matched with the correct name of the verb, and 10
depicting actions matched with an incorrect name (incorrect verb).
The picture-sentence matching task included altogether 80 correct picturesentence
matches and the same number of incorrect matches. The correct matches consisted
partly of training stimuli (20 SVO and 10 SOV sentences) and partly new sentences
(20 SVO, 10 SOV, 20 VSO). VSO sentences were included in the material to test the
participants sensitivity to word-order effects. Half of the sentences included a
feminine noun, the other half a masculine noun. Each verb was shown 4 times overall
in the sentences, each noun as a subject also 4 times, each noun as object 4 times as
well. The types of mismatched sentences were as follows: a. grammatically
mismatched sentences where one of the grammatical rules was violated (20 sentences
with object marking mismatch, 20 sentences with gender marking mismatch); b.
lexically mismatched sentences (20 sentences) and c. grammatically-lexically
mismatched sentences (10 with object marking mismatch and lexical mismatch and 10
with gender marking mismatch and lexical mismatch) (see examples for the different
mismatched sentences in Table 1).
Type of mismatch
Grammatical mismatch
1. object marking mismatch
2. gender marking mismatch
Lexical mismatch
Grammatical-lexical mismatch
1. object marking mismatch +
lexical mismatch
2. gender marking mismatch +
lexical mismatch
* ungrammatical sentences
377
All mismatched sentences were new, i.e., the participants had not seen the
corresponding correct versions during training. The picture-description task comprised
10 new pictures with two animals in action.
1.3. Procedure
Both the training and follow-up tasks were run on a computer using the Superlab program (SuperLab Experimental Laboratory Software, version 2.0.4, Cedrus Corporation), which recorded the participants reaction times in milliseconds (the time from the
appearance of the stimuli to the pressing of the reaction time key) and the correctness
of their responses. The participants were tested individually in a quiet room.
The experiment proper consisted of four sessions run on different days: Session 1
included filling in the informed consent form, the subject questionnaire and the
language background questionnaire, followed by the pre-training session. In the pretraining session participants were asked to carefully watch pictures on the screen of
the computer appearing one after another and label them in their mother tongue,
Finnish. If they committed an error, the experimenter corrected them and gave the
correct label of the picture in Finnish (mean correct was 97.9 %, SD = 2.5, range
between 93 and 100 %). The pre-training session was followed by three training
sessions with the new language. After training sessions the experimenter interviewed
the participants about the rules of the novel language presented (How do you think
this language is constructed?). Session 2 and 3 consisted of a test-phase including all
experimental tasks followed by one training session. In Session 4 all experimental
tasks were run and subjects were asked again about the rules of the novel language.
Also the nonword span and digit span tasks were administered to participants in the
final session. A maximum of two days elapsed between the sessions. Overall, one
session took about one hour.
During the training session, all the pictures were presented for 7 seconds along with
the sentences describing the pictures. Stimuli were preceded by a fixation point and
followed by a 1.5 seconds pause. Participants did not have to give response to any of
the stimuli presented, just read the sentences aloud. The duration of one training session was about 15 minutes.
A test-phase included production tasks and silent tasks presented in a
counterbalanced Latin square order. The silent tasks were the following: picture-word
matching task and picture-sentence matching task. Production tasks included in the
test battery were the picture naming task and the picture description task.
In case of the picture-word matching task, the instruction was the following: press
the yes button if the word presented below the picture matches the picture of the
animal or the action, and press no if pictures and words are mismatched. There was
a 2 seconds pause between different stimuli. In case of the picture-sentence matching
task, pictures and sentences were presented one by one to the participants and they had
to indicate by pressing the right button whether they were correctly matched or
mismatched. Stimuli were preceded by a fixation point for 1 second and were followed
by a 2 second pause.
378
During the picture-naming task pictures were presented one-by-one on the screen
and subjects had to name the animals or actions seen as quickly as possible in the new
language. In all of the tasks mentioned we included an upper limit of waiting for the
answer of 10 seconds. In the picture-description task, the subjects saw 10 new pictures
not presented before and had 12 seconds to provide the sentence in the novel language
describing the pictures. All the answers were tape-recorded.
2. Results
All incorrect responses and reaction times that differed more than three standard
deviations from the individual mean latency were discarded from the data set.
Statistical analyses focused on RTs and error rates yielded from the tasks performed
on the three testing occasions. Data were analyzed using analyses of variance
(ANOVAs), followed by post hoc Tukey tests where applicable.
2.1. Acquisition of active vs. passive vocabulary at the word level
Two-way ANOVAs (test session x task) were performed for RTs and percentage of
correct responses in the picture-naming and the picture-word matching task. Analysis
of the correct responses showed significant main effects for test session (F(1.83,
142.65)=222.5, p<0.001) and for task (F(1, 78)=135.31, p<0.001). These main effects
indicated better performance with more training and also a difference between the two
tasks, one active and one passive. Picture-word matching yielded fewer errors than
picture naming. There was also a significant interaction between the two factors
(F(1.83, 142.65)=123.96, p<0.001), reflecting the fact that the difference in error rates
between naming and picture-word matching was far greater on test 1 than on test 3.
However, the interaction has to be treated with caution due to the ceiling effect on
picture-word matching, as shown in Figure 2.
379
The reaction time analysis yielded a significant main effect for test session (F(1.27,
93.96)=53.25, p<0.001) indicating that on test 1, performance on the word-level tasks
was slower than on test 3. Also a significant main effect for task was found (F(1,
74)=53.98, p<0.001), stemming from the fact that matching pictures to words was
performed faster than naming pictures. The interaction term in the reaction time
analysis was also significant (F(1.27, 93.96)=9.61, p<0.01), indicating a greater
discrepancy between the tasks on the first test occasion than on the second or the third
occasion (see Figure 3)
Figure 3. Reaction time results on word-level tasks (picture-word matching and picture-naming)
380
indicates learning of the grammatical marker options without knowing their correct
slots in this new language (see Figures 4 and 5).
Figure 4. Percentage correct on picture-sentence matching task (lexical and grammatical learning)
Figure 5. Percentage correct on the picture description task (lexical and grammatical learning)
381
Figure 7. Picture description object marking rule and gender marking rule percentage correct
This result may be due to the limited choice of verbs (there were only two possible
forms of verbs). During the picture description task both rules were produced correctly
in around 50 % of the cases.
The results of the interview showed that when participants were asked to formulate
the underlying rules of the languages, more of them were able to tell the object marking rule on both test 1 and test 3 (see Table 2).
Test 1
382
Test 3
12
Table 2. Number of participants formulating the rules of the language on test 1 and 3
The gender marking rule may be difficult for several reasons: it is not characteristic of
Finnish, it is non-adjacent, it is not essential for determining the logic of the language
or because the verb was never shown in unmarked form. Further studies are needed in
order to state which of these assumptions is true.
2.4. Sensitivity to word order frequencies
Subjects committed more errors with the types of sentence that were not seen during
training, namely VSO (comparison of SOV, SVO sentences and VSO sentences: (F (1,
98)=66.6, p=.000). In addition, SVO sentences yielded better performance compared
to SOV sentences. Moreover, there was a tendency for a difference between SOV and
SVO sentences, F(1,78)=2.89, p = .093. These results reveal the learners sensitivity to
word order frequencies.
3. Discussion
By using a miniature artificial language, the present study attempted to shed light on
lexical and grammatical acquisition of a new language in adults. The language to be
learned consisted of a set of lexical items and two embedded syntactic rules used for
creating grammatically well-formed sentences. The analyses focused on three aspects
of language learning: acquisition of active vs. passive vocabulary, lexical vs.
grammatical learning at sentence level, and acquisition (both implicit and explicit) of
the two grammatical rules.
All our subjects had difficulties with the production tasks compared to the silent
tasks. Passive vocabulary developed fast but early on language learners had difficulties
in producing the new words. However, the acquisition of words was mostly
successful, subjects were able to perform the lexical level tasks with few errors and at
high speed when they concluded the three day training phase.
Some researchers (e.g. PerruchetPacteau 1990) have argued that participants in an
artificial grammar study could well have memorized some of the original training
items, rather than learn the underlying structure. Our results at sentence level showed
that subjects were not just memorizing specific sentences but instead they were
learning a new language. Moreover, our learners were effectively picking up wordorder frequencies of the new language as well.
Syntactic learning was more difficult than lexical learning. The participants of the
study learned the grammatical marker options but showed no awareness of their
correct use in the new language. Regarding the rules of the new language, we can posit
that subjects had more difficulties with the gender marking rule, a rule that is not
383
characteristic of their mother tongue. When asking about their explicit knowledge of
the regularities of the new language the same pattern was found: a non-Finnish rule is
less frequently explained by the participants compared to the other rule, which is also
used in their first language.
When adults attempt to learn a new language, they start with an already established
grammatical system. Learners usually transfer segments of their L1 over to the new
L2. Positive transfer occurs when the two languages are similar. Tokowitz and
MacWhinney (2005) found, using event-related brain potentials, that learners were not
sensitive to violations for constructions that differ in the L1 and the L2, but were
sensitive to violations in L2 for constructions that are similar in their two languages.
Our results suggest that learners are able to implicitly process some aspects of L2
syntax even in early stages of learning. However, this knowledge depends on the
similarity between the L1 and the L2. The conclusion is that the structure of the
mother tongue limits our grammatical learning ability: a grammatical rule that follows
the mother tongue is learned more efficiently than a rule that is different.
Our study raises the question whether different learning systems are at play when
learning a language. Electrophysiological, lesion, and functional neuroimaging studies
show that the functional anatomy of semantic and syntactic processing is different
(Osterhout 1997; Caplan 1999; Dapretto and Bookheimer 1999; Friederici et al. 2000;
and Newman et al. 2001). Moreover, studies presented patients affected by grammarspecific deficits, where vocabulary remained unimpaired as well as patients showing
the opposite pattern of impairment (Caplan 1992; van der Lely et al. 1998 and van der
Lely and Christian 2000). Based on these findings researchers argued for different
memory systems for lexical vs grammatical processing and learning (for reviews see
Paradis 1994; Lebrun 2002). According to Ullman (2004) the mental lexicon relies on
the declarative memory system, while the mental grammar is subserved by procedural
memory. The declarative memory system has been implicated in the learning,
representation and use of knowledge about facts and events. The procedural memory
system is needed when learning skills, habits and other procedures (e. g. acquiring a
language, riding a bike, walking). The acquired knowledge and the learning process
itself are implicit, meaning that they are generally not available to conscious access. In
our study the difference found between lexical and grammatical learning might show
that grammar is acquired incidentally and the lexicon is processed by the declarative
memory, but we cannot know whether or not overall difficulty of lexical vs.
grammatical learning in the sentence level tasks is comparable.
A great deal of empirical research on L2 acquisition has focused on the question of
ultimate attainment (e.g. Birdsong 1992; HanOdlin 2006; Lardiere 2007), on what may
or may not be acquirable in an L2 (e.g. HawkinsFranceschina 2004), and whether the
acquisition of certain L2 features are subject to a critical period for language acquisition
(e.g. Birdsong 1999; BirdsongMolis 2001; HyltenstamAbrahamsson 2003; JohnsonNewport 1989; SingletonLengyel, 1995). Few studies have addressed the rates of
attainment or the stages of L2 development, especially the early stages of L2 learning.
Our results indicate that the early phase of adult L2 learning is not uniformly slow
and labourious; some aspects of the language are acquired with remarkable speed.
This study attests to the efficacy of the language learning system: limited exposures to
384
sentence-picture pairs are sufficient for lexical learning and acquisition of some
embedded regularities, while explicit metalinguistic knowledge of the regularities does
not develop until later.
More generally, the study presented here demonstrated that the Animal Farm can be
used to study basic aspects of lexical vs. grammatical learning in a wide variety of
setups.
References
Baddeley, A.D. Gathercole, S.E. Papagno, C. 1998. The phonological loop as a language learning
device. Psychological Review, 105(1), 158173.
Berry, D. C. Broadbent, D. E. 1988. Interactive tasks and the implicit-explicit distinction. British
Journal of Psychology, 79, 251-272.
Birdsong, D. 1992. Ultimate attainment in second language acquisition. Language, 68, 706755.
Birdsong, D. 1999. Introduction: Whys and why nots of the Critical Period Hypothesis. In D.
Birdsong (Ed.), Second language acquisition and the Critical Period Hypothesis, 122. Mahwah,
NJ: Erlbaum.
Birdsong, D. Molis, M. 2001. On the evidence for maturational constraints in second-language
acquisition. Journal of Memory and Language, 44(2), 235249.
Braine, M. D. S. 1987. What is learned in acquiring word classes: A step toward an acquisition
theory. In B. Macwhinney (Ed.), Mechanisms of language acquisition, 65-87. Hillsdale. NJ:
Erlbaum.
Braine, M. D. S. Brody, R. E. Brooks, P.J. Sudhalter, V. Ross, J. A. Catalano, L. 1990.
Exploring language acquisition in children with a miniature artificial language: Effects of item
and pattern frequency, arbitrary subclasses, and correction. Journal of Memory and Language, 29,
591-610.
Brown, R. M. Robertson, F. M. 2007. Off-line processing: Reciprocal interactions between
declarative and procedural memories. Journal of Neuroscience, 27, 10468-10475.
Caplan, D., 1999. Activating brain systems for syntax and semantics. Neuron 24, 292293.
Caplan, D., 1992. Language: Structure, Processing, and Disorders, MIT Press, Cambridge, MA.
Dapretto, M. Bookheimer, S.Y.1999. Form and content: dissociating syntax and semantics in
sentence comprehension. Neuron 24, 427432.
DeKeyser, R. M. 2003. Implicit and explicit learning. In C. J. Doughty M. H. Long (Eds.),
Handbook of second language acquisition, 313-348. Oxford, MA: Blackwell.
De Vries, M. H. 2009. Implicit learning of artificial grammars: Its neural mechanisms and its
implications for natural language research, Doctoral Dissertation, Westfalishen WilhelmesUniversitat zu Mnster.
Dienes, Z., Broadbent, D. Berry, D. 1991. Implicit and explicit knowledge based in artificial
grammar learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 17,
875 - 887
Ellis, R. 2004. The definition and measurement of L2 explicit knowledge. Language Learning, 54,
227-275.
Friederici, A. Steinhauer, K. Pfeifer, E. 2002. Brain signatures of artificial language processing:
Evidence challenging the critical period hypothesis. PNAS, 99, 529-534
Friederici, A.D. Meyer, M. von Cramon, D.Y., 2000. Auditory language comprehension: an
event-related fMRI study on the processing of syntactic and lexical information. Brain Lang. 75,
289300.
385
Hakulinen, A., Karlsson, F. 1980. Finnish Syntax in Text: Methodology and Some Results of a
Quantitative Study, Nordic Journal of Linguistics, 3, 93-129
Han, Z. Terence Odlin, T. 2006 (Eds.), Studies of Fossilization in Second Language Acquisition.
Clevedon, U.K.: Multilingual Matters
Hawkins, R. Franceschina, F. 2004. Explaining the acquisition and non-acquisition of determinernoun gender concord in French and Spanish. In P. Prvost J. Paradis (Eds.), The acquisition of
French in different contexts. Amsterdam/Philadelphia: John Benjamins.
Hulstijn, J. 2005. Theoretical and empirical issues in the study of implicit and explicit second
language learning, SSLA, 27, 129140
Hyltenstam, K., Abrahamsson, N.2003. Maturational constraints in SL. In C. J. Doughty M. H.
Long (Eds), Handbook of second language acquisition 539599 Oxford: Blackwell
Johnson, J. S. Newport, E. L. 1989. Critical period effects in second language learning: The
influence of maturational state on the acquisition of English as a second language. Cognitive
Psychology, 21, 6099.
Lardiere, D. 2007. Ultimate Attainment in Second Language Acquisition: A Case Study. Mahwah,
NJ: Lawrence Erlbaum Associates
Lebrun, Y., 2002. Implicit competence and explicit knowledge. In: Fabbro, F., (Ed.) Advances in the
Neurolinguistics of Bilingualism, Forum, Udine, 299313
Lieberman, M.D. 2000. Intuition: A Social Cognitive Neuroscience Approach. Psychological Bulletin, 126, 109-137.
Newman, A.J. Pancheva, R. Ozawa, K. Neville, H.J. Ullman, M.T., 2001. An event-related
fMRI study of syntactic and semantic violations. J. Psycholinguist. Res. 30, 339364.
Niemi, J. Laine 1997 Syntax and Inflectional Morphology in Aphasia: Quantitative Aspects of
Wernicke Speakers Narratives. Journal of Quantitative Linguistics, 181-189
Osterhout, L. 1997. On the brain response to syntactic anomalies: manipulations of word position
and word class reveal individual differences. Brain Lang. 59, 494522.
Paradis, M., 1994. Neurolinguistic aspects of implicit and explicit memory: implications for
bilingualism. In: Ellis, N. (Ed.). Implicit and Explicit Learning of Second Languages, Academic
Press, London 393419.
Perruchet, P. Pacteau, C. 1990. Synthetic grammar learning: Implicit rule abstraction or explicit
fragmentary knowledge. Journal of Experimental Psychology: General, 119(3), 264-275.
Polonyi Tnde Abari Klmn Ntin gnes 2009. Mestersges nyelvtanuls els benyoms alapjn, Alkalmazott Pszicholgia, XI, 1-2, 5-26.
Reber, A. S. 1989. Implicit learning and tactic knowledge. Journal of Experimental Psychology:
General, 118, 219-235.
Robinson, P. 2005. Cognitive abilities,chunk-strength, and frequency effects inimplicit artificial
grammar and incidental l2 learning:replications of Reber, Walkenfeld, and Hernstadt (1991) and
Knowlton and Squire (1996) and their relevance for SLA, SSLA, 27, 235268
Service, E. 1992. Phonology, working memory, and foreing-language learning. Quarterly Journal of
Experimental Psychology, 45A, 21-50
Singleton, D. Lengyel Z. (Eds.) 1995. The age factor in second language acquisition: a critical
look at the Critical Period Hypothesis. Clevedon: Multilingual Matters
Song, S. Marks, B. Howard, J. H., Jr. Howard, D.V. 2009. Evidence for parallel explicit and
implicit sequence learning systems in older adults. Behavioural Brain Research, 196, 328.
Tokowicz, N. MacWhinney, B. 2005. Implicit and Explicit Measures of Sensitivity to Violations
in Second Language Grammar: An Event-Related Potential Investigation. Studies in Second
Language Acquisition, 27, 173-204
Ullman, M. 2004. Contributions of memory circuits to language: the declarative/procedural model.
Cognition, 92, 231-270
386
van der Lely, H.K. Christian, V., 2000. Lexical word formation in children with grammatical SLI:
a grammar-specific versus an input-processing deficit? Cognition, 75, 3363
van der Lely, H.K. Rosen, S. McClelland, A., 1998. Evidence for a grammar-specific deficit in
children. Curr. Biol. 8, 12531258.
Wechsler, D. 1981. Wechsler Adult Intelligence Scale-Revised (WAIS-R). New York: The
Psychological Corporation.
387
1. Introduction
In a time of ever-changing nationwide educational reforms, universities face the
challenge of more diversified student populations and the introduction of new study
programs. It is a general observation of instructors that each year a larger population
enters higher education unprepared for their chosen field of studies. Many of these
students can be classified as underachievers since they do not meet the requirements:
they withdraw from or do not pass courses, oftentimes receive low grades, therefore,
need more time to finish the program or eventually drop out of it. Underachievement
may have a number of reasons, including, but not limited to, a mismatch between
student expectations and study requirements, unpreparedness for the given study program, lack of study strategies, laziness, inability to manage ones time, or social and
financial problems (Crosling et al. 2008, Dor 2010, Dor 2011a, Moxley et al. 2001).
Moreover, in higher education, more than in secondary school, many courses build on
previously acquired subject knowledge or skills. If students lack these, they have less
chance to proceed with their studies even if they do their best in the new courses. As
there is a wider and larger pool of the younger generations who enter higher education,
instructors have to face new types of student problems: general low cognitive abilities,
low problem solving skills, undeveloped social skills, time management problems,
attendance problems, inability to follow explanations, not doing homework, and
postponing deadlines and exams. We need to predict, however, that in some cases
specific learning disabilities are responsible for student failure rather than laziness or
unwillingness to study.
For foreign language majors, who carry out their studies predominantly in the target
language, inadequate screening prior to entering the university often means they enroll
in the program with insufficient language proficiency and subject knowledge. The
final examination at secondary school, which also serves as an entrance exam, has
proven not to be a good indicator of preparedness for undergraduate studies in a target
language. Nevertheless, there are signs of problems that cannot be directly related to
low proficiency, typical spelling mistakes or shyness in communication. This study
seeks to answer the questions how it is possible to recognize learning disabilities (LD)
and other special needs conditions of students, why it should be important to have
standardized forms of accommodation to the requirements, and how students
themselves perceive their conditions and study experiences.
The sections that follow review the educational and research issues of learning
disabilities in reference to LD in higher education and foreign language learning. It
also discusses the specific signs of study problems that are detectable in student
performances at a selected Hungarian institute of higher education. Then, two case
388
studies of students of English are presented which are based on classroom observation,
informal talks with students, and a structured interview with them. Conclusions and
implications of this qualitative study are drawn at the end of this research report.
2. Learning disabilities and study problems
2.1. Definition of key terms
The 2011 World Report on Disabilities of the World Health Organization defines
specific learning disability as follows: impairments in information processing
resulting in difficulties in listening, reasoning, speaking, reading, writing, spelling, or
doing mathematical calculations for example, dyslexia. (p. 309). Nichols (1995)
indicates five main types or learning disabilities: visual, auditory, motor,
organizational and conceptual. While LDs due to visual and motor impairment are
most of the time easily recognizable, the last two categories often remain
unrecognized or misinterpreted, and therefore, it is more difficult to accommodate the
teaching process to these students needs. Adults with LD may experience the
following symptoms: poor or uneven academic achievement, difficulty with language
usage, poor organizational and/or sequencing skills, and poor or inappropriate social
skills (Nichols 1995: 36). There is a disagreement in the literature concerning the
existence of a specific foreign language learning disability (FLLD), and most of the
systematic data come from higher education institutes in the United States offering
accommodation or substitution for students who have difficulties in passing foreign
language courses (e.g. Ganschow et al. 1998, Hatzes et al. 2002, Sparks 2006).
Although Sparks (2006) argues that there is no FLLD, people with LD often
experience problems with literacy skills, following conventional teaching methods and
materials, which influences their foreign language studies. Learning disabilities should
not be confused with inadequate learning styles or strategies, laziness or low level of
intelligence (Denhart 2008). Many people with LD have average or above average IQ
and develop compensation strategies to balance for their weaknesses (McNulty 2003,
Kirby et al. 2008). The literature also points out that there are adults, including
students in higher education, who prefer to hide their disabilities, which makes it
difficult to estimate the real number of students with LD. Heiman and Precel (2003)
review some of the late 1990s figures available and report on a 3.8% LD prevalence
among first-year undergraduate students in the United Kingdom, and an even higher
and growing rate for the United States. The British Dyslexia Association uses the
umbrella term specific learning difficulties, including the often co-occurring
dyslexia, dyspraxia, dyscalculia, attention deficit/hyperactivity disorder (AD/HD), and
auditory processing disorder. The Association estimates a 4% prevalence rate for
severe dyslexia and a 10% rate for mild dyslexia. Gyarmathy (2007) also estimates
similar figures for Hungary among elementary and secondary school students, with
figures showing a gradual increase. It is most likely that students with LD enter higher
education in larger numbers than a few decades ago when the more thorough
screening processes selected out the least academically fit.
389
390
stress, low self-esteem, anxiety or anger due to their fear of perceived stigma or
unsuccessful coping with their condition.
While some kind of accommodation is available for students with special needs in
the Hungarian higher education system, the higher education act and the special needs
regulations of higher education institutes seem to focus more on students with physical
impairment and fewer specific recommendations are given as to who should be
classified as LD and how accommodation should be carried out. This is a tendency
that exists also elsewhere (Hatzes et al. 2002). Moreover, there could be a number of
students who receive partial or overall exemptions from evaluation in foreign language
classes in elementary and secondary schools (often as a result of diagnosed dyslexia),
but do not have the same option when enrolling in foreign language studies at a
university, and therefore, meet unexpected challenges.
It is a major dilemma how foreign language study programs should recognize these
students and their special needs and to what extent the requirements can and should be
modified to meet their needs.
The need for this research study is in line with departmental and university goals to
improve education, discover the reason for achievement gaps of students and assist
them in becoming more successful students or, if necessary, to redirect them towards
other fields of studies more suitable for them.
3. The empirical study
The purpose of this qualitative case study was to gather information from students
with signs of LD in order to explore possible courses of action to improve student
performance in a foreign language program.
3.1. Participants
The specific study population was two students of English who agreed to discuss their
study experiences with the author of this article, who served as their teacher in small
group seminars. Students were identified as possible special needs students as the
course proceeded because they showed signs of having major difficulties in keeping
up with the rest of the group.
Student A was a male student in the first year of his BA level English studies, while
student B was a male student in his last year of the 5-year MA level English studies.
3.2. Methodology and procedure
A qualitative multiple case study approach was adopted for this inquiry, including
classroom observation, analysis of written work done by the participants, informal
talks with them and a semi-structured interview. The rationale for choosing multiple
data collection methods is that the qualitative data set and results provide a general
picture of the research problem, namely the situation that potentially LD students face.
They are labeled as potentially LD as no formal diagnosis is available for them, but
are identified through coursework as students whose performance or academic
391
achievement make them distinct from other students with general language proficiency
or motivational problems.
The primary data collection technique was an in-depth semi-structured interview
carried out individually with the two subjects, in a quiet room and recorded for later
transcription. The participants were asked to give consent to record their answers, and
they were informed prior to the interview about the types of question asked. Student B
was given the list of questions beforehand in order to compensate for his attention
problems. The interviews were done in Hungarian, the native language of the
participants, and included questions from the following main areas:
1. general experience at the university
How can you summarize your study experiences at the university? What kind of
problems have you encountered? Do you proceed with your studies slower than
expected?
2. following courses, taking exams
Can you concentrate for 90 minutes in class? Can you take notes when you
understand the instructors explanation? Do you prefer written or oral exams?
3. doing assignments
What makes homework difficult? Do you read the assigned reading materials?
Do you have reading/spelling/writing problems in Hungarian or in English? How
long does it take for you to write a one-page essay at home? What compensation
strategies do you have to overcome/balance for your problems?
4. feedback from others
Have you noticed that you are behind your peers or do not meet expectations? In
what ways? Have your instructors/peers/family members noticed and told you
that you are behind or have special needs? Have you ever asked for special
accommodation or help from your instructors (extra preparation time, changing
deadlines and changing tasks, asking for notes, extra tutoring)?
5. previous studies
Did you have extra difficulties in any of the classes in elementary and secondary
school? Were you given exemption in any of the classes in elementary and
secondary school?
6. diagnosis vs. self-identification
Have you ever been diagnosed with learning difficulty, attention deficit,
dyslexia, etc.? Have you ever had the feeling that you might have any of these?
The performance of the two students was also observed for a semester in small study
groups. For student A this was done in a language proficiency class that involved both
oral and written assignments, regular vocabulary tests and a reading journal. Student B
was observed in an applied linguistics seminar which required him to follow the
instructors explanations aided with Power Point presentations and short videos,
reading and commenting in writing and in speech on research articles, being an active
participant in short discussions in English on classroom topics and writing a 610page long research report.
392
vocabulary). These seemed to be avoidance strategies and not the real causes of his
problems.
He reported on having tried out certain strategies for vocabulary learning, and
speeding up oral and written production, some of which produced results, while others
not. One of the most interesting and failed methods was what he called drawer
technique. As he considered himself a visual type of learner, he tried this method,
which involves the imagining of a chest of drawers where you put different sentences
to remember them. When you open the drawer, the information is there the way you
put it in. Student A reported that when he opened the imaginary drawers, single letters
and punctuation marks were floating out, which made it impossible to reconstruct the
original text.
Data gained during this interview are in line with the observations made in class
and during the reading of the students texts. He seemed to show a variety of
difficulties both with oral and written language use in English. He reported to be
aware of some of his weaknesses and had tried various compensation strategies, but
most of these had failed. His low level of general English proficiency, the slow
improvement in his reading, writing, listening and speaking skills suggested that he
had little success in academic achievement in even the first-year courses. Although he
was willing to and had tried to compensate for his problems, he was unable to meet the
requirements.
4.2. Student B
Student B stood out of the group as the late comer, the only one not taking notes and
seemingly not paying attention either to the instructors explanation or the Power Point
slides. If he turned in weekly assignments (summary of and comments on research
articles, individual tasks), the written work was much shorter than expected, but
language-wise and content-wise good. After the first few weeks he asked for the
Power Point slides, which were than made available for all the students in the course.
As other potentially LD students were also identified in the class, homework
assignments included a variety of sources (oral, written and visual) to balance for
possible reading and writing problems; furthermore, reading in class was limited to
short texts followed by pair or group discussions on the content, and individual
accommodation for the final written project was offered. Since special education needs
of university students was one of the course topics, Student B discussed with the
researcher the option of writing about his own experiences as a follow-up of the
interview he was asked to participate in.
During interview and in the follow-up essay he noted that his problems had become
more disabling as he was progressing through the education system. For years he had
not realized he had special needs, relied on what was heard in class, but was unable to
study from books. He wrote I was born in a family who though I was special. I was,
but probably not in the way they thought.
He was labeled as a good, but misbehaving student, which was tolerated in the nonstate schools he attended. Even at the university, as he wrote, he felt an average
student who developed survival techniques as a reaction to the growing challenges. He
394
rarely studied for classes and exams, and relied on the notes of peers or those of the
instructors. Altogether he had good English skills, but it took him much longer than
others to do readings or write longer texts. He admitted not being able to read long
assignments, or writing reviews; instead he adopted the technique of reading thesis
statements, highlighting key words, and relying on these during class discussions. As
he was unable to prepare for exams, he cheated at times. when there was no other
choice, I cheated on the exams with a pure soul, because knew I could have never
learned that material, and that I would probably never need to use it in the future.
He believed his difficulties had not been recognized by instructors and peers, and
even by himself for years. Only in the final year of the university did he notice
strange things about himself. He wrote as until the last 1-2 years, I did not think
that I may have learning impairments, it was never an issue, and even now, no one
realizes that I have difficulties in studying, probably because through the years, I
learned how to present or explain my shortcomings even to myself, and the most I
usually need to do is ask teachers to delay deadlines for me, which they usually do
without asking why. He had never been tested or diagnosed with any LD, and
problems had multiplied as the study demands got overtly challenging for him. He
identified himself as a person with attention deficit and with learning impairments.
He also chose not to disclose his weaknesses to others, as he worded I never told a
teacher about my impairments, and not because I find it a shame, but simply, because I
think I would feel ashamed if I had to do less than others for the same grade. Id rather
get a worse grade or be thought to be lazy or a layabout, instead of being treated as a
student with some kind of handicap, who needs help and special treatment.
He had also adopted what he called survival techniques, such as focusing on the
task he was doing and excluding everything else, imagining a loved person as a reader
for his assignments, and forcing himself to keep to deadlines.
4.3. Discussion of the results
The purpose of this study was to describe the study experiences, areas of difficulties,
strategies for academic coping, and self-identification of two potentially LD students.
Data obtained through semi-structured interviews with the participants provide empirical
support for the authors observation of special needs students in courses. The finding
regarding students need for some accommodation in terms of assignments, tasks and
deadlines can be viewed in light of Heiman and Precels (2003) report on the academic
strategies of university students with and without LD. Both students were able to report
on some of their major problems and the unusual survival techniques and compensatory
skills used. Neither of the two students, however, can be considered academically
successful without adequate study and literacy skills. It is important to note here that
Student A was unable to compensate for his problems in lack of a good English
proficiency as a basis, whereas Student B could survive challenges and even hide his
problems with the help of his advanced language skills. This also implies that while
Student A could be identified by a multiple of instructors as a weak student, Student B
was able to pass unnoticed or mislabeled. In light of these results, it is very much
395
questionable, how these and similar students are able to proceed with their studies
without some minor or major accommodation offered to them.
It is also important to pose the following questions: Whose task is it to identify and
assist foreign language majors with learning difficulties? Do they know about their
impairments, and if yes, are they willing to admit them? Do they want to be identified
and labeled? To what extent can they cope with the challenges of a foreign language
major even with additional help or good strategies? Whose failure is it if they fail?
And finally, how can instructors, who are not trained to identify LD, prepare and be
prepared for a diverse student population that includes students with LD? These and
related questions should be systematically reviewed before a rationale for
recommended and commonly accepted accommodation can be provided to students.
Individualized tutoring and small modification of tasks were done by the researcher
for the participants, but these cannot be considered a regular or commonly accepted
practice or option that is available if larger prevalence rates of LD students appear on
top of non-LD students with different language and academic difficulties.
When students are not successfully meeting the requirement, the focus often turns
to instructors, who experience more pressure in todays academic environment from
parents, educational professionals, and decision makers to teach and prepare all
students for exams. The growing number of dropouts and the pressure to keep in the
system even the academically unfit suggest that teaching practices of years past must
be changed. Before doing so, however, it is necessary to discover instructors,
students, policy makers and university staff members perceptions of the situation.
A general first-year population of English majors has been found not to be able to
make good judgments about their level of preparedness for their studies or seek help
when needed (Dor 2011b). All students, including those with LD or other academicrelated problems, should be encouraged to be self-regulated learners who are able to
understand their needs, study goals, and use adequate study strategies (Zimmermann
1990). This is especially important in light of the results that students are unwilling to
disclose their problems, probably for fear of being labeled and stigmatized. They seem
to prefer the protective mass of peers, even when unable to cope on their own with
their failures.
It also needs to be considered and clearly stated that any accommodation to the
requirements should not result in the compromise of academic standards, loss of
credibility for the instructor, the introduction of a certain inequity of treatment
between student performances or the victimization of students.
A major shortcoming of this study comes from the inability to generalize the
findings to larger student populations. Yet, they show striking evidence for the
existence of students with major and co-occurring problems that go beyond those of
their non-LD peers and which hinder their chance for academic success.
5. Conclusion and implications
This study discussed the need for a better understanding of foreign language majors
who show serious signs of underperformance compared to their peers. It was shown,
through the example of two students, that some of the potentially LD students can be
396
detected on the basis of their class performance, but it remains a question to be solved
how to approach these students, how to raise awareness among staff, and how to
accommodate requirements without jeopardizing academic standards. It was also
pointed out that most of the special needs students are reluctant to ask for help or
accommodation, or do not even know how to modify their own study strategies and
methods. The current general approach of no accommodation seems to meet both
sides immediate interest, namely the students unwillingness to disclose problems and
the academics unpreparedness to cope with the situation. This, however, is not a longterm strategy to be adopted, as professionals in education should search for a common
approach to assist learners with special needs.
References
Crosling, G. Thomas, L. Heagney, M. 2008. Improving student retention in higher education:
The role of teaching and learning. London: Routledge.
Denhart, H. (2008). Deconstructing barriers: perceptions of students labeled with learning disabilities
in higher education. Journal of Learning Disability. 41: 483497.
Dor Katalin 2009. Advanced vocabulary for foreign language majors: An advantage or a must? In
Vradi Tams (szerk.): I. Alkalmazott Nyelvszeti Doktorandusz Konferencia Vlogatott Ktete
[Selected papers from the 1st Applied Linguistics Doctoral Conference]. Budapest: MTA Nyelvtudomnyi Intzete. 2438.
Dor Katalin 2010. Meeting the language barrier: The experience of first-year students of English. In
Hegeds, I. Martsa, S. (eds.): CrosSections. Vol. 1: Selected papers in linguistics from the 9th
HUSSE conference. Pcs: University of Pcs, Institute of English Studies. 289297.
Dor Katalin 2011a. Lemorzsolds nyelvszakon: helyzetelemzs egy szegedi angolos vfolyam
tkrben [Dropout rates in foreign language studies: A case study among first-year students of
English in Szeged]. In Boda Istvn Mnos Katalin (szerk.): Az alkalmazott nyelvszet ma:
innovaci, technolgia, tradci. XX. Magyar Alkalmazott Nyelvszeti Kongresszus, Debrecen,
2010. augusztus 2628. BudapestDebrecen: MANYE Debreceni Egyetem. 404409.
Dor Katalin 2011b. Students perception about their preparedness for undergraduate studies of
English. In Szab Gbor et al. (eds.): UPRT 2009. Empirical studies in English applied
linguistics. Pcs: Lingua Franca Csoport. 8191.
Ganschow, L. Sparks, R. Javorsky, J. 1998. Foreign language learning difficulties: an historical
perspective. Journal of Learning Disability. 31: 248258.
Gyarmathy va 2004. Research on Dyslexia in Hungarian. In Smythe, I. Everatt, J. Salter, R.
(eds.): International Book of Dyslexia. A cross language comparision and practice guide.
123132.
Gyarmathy va 2007. Diszlexia. Specifikus tantsi zavar [Dyslexia. Specific Teaching Difficulties].
Budapest: Llekben Otthon Kiad.
Hatzes, N. Reiff, H. Bramel, M. 2002.The documentation dilemma: access and accommodations for
postsecondary students with learning disability. Assessment for Effective Intervention. 27: 3752.
Heiman, T. Precel, K. 2003. Students with learning disabilities in higher education: academic
strategies profile. Journal of Learning Disability. 36: 238248.
Kirby, J. Silvestri, R. Allingham, B. Parrila, R. La Fave, C. 2008. Learning strategies and
study approaches of postsecondary students with dyslexia. Journal of Learning Disability. 41:
8596.
McNulty, M. 2003. Dyslexia and the life course. Journal of Learning Disability. 36: 363381.
397
398
1. Introduction
Being a twin can be characterised by unfavourable conditions in their intra-uterine life
because of the lack of space for their developing organisms (Mtneki 1997).
Consequently, the weight and length of each twin at birth may be different and one
may appear to be the dominant child, and this may be influenced by the order of their
birth. The twins show different development in their movement compared with a
single child and they also have irregular development in terms of speech.
Backwardness in the development of their communication skill is also claimed. It
can be characterized by late development in speaking (in terms of pronunciation of
first words) on the one hand, and a different kind of development in their speech
compared with a single child on the other. The typical symptoms occurring in their
speech can be as follows: they produce shorter sentences, words or syllables are
missing from their sentences, and they can have different types of speech defects in
terms of articulation. There is a special secret language called cryptophasia used by the
twins in the communication process with each other (Vassn Kovcs 1969, Mrei V.
Bint 1985). It may contain special signs, symbols, or words and may be
understandable only to them, consequently it is not suitable for communication with
those around them. They may have limited vocabulary which can cause problems in
the mental lexicon.
Among the reasons for their special speech development is the different speech
model, and the twin isolation process is also emphasized. In the first case the mothers
speech sample is divided between the children by comparison with the close
interpersonal relationship with a single child. The twin isolation process can appear
especially among identical twins. It means a diminution in the relationship between
twins and their surroundings, consequently their smaller number of communication
mechanisms may also have a negative effect on their language development. However,
it is also emphaisized that their language backwardness and the differences and
disadvantages in their language development may be equalized by the time they start
school and during the compulsory school years because of their cognitive development
and school studies. The dominant twin may also develop later, especially during the
period of adolescence, which can be seen in terms of their studying process,
communication behaviour and influence of it on their personality. Consequently, they
can have differences from their peers in terms of communication skill.
On the basis of these findings the question is what differences occur, how and to
what extent, in their linguistic development in terms of their vocabulary compared to
their peers. Furthermore, there is no experimentally supported knowledge concerning
the size of their mental lexicon in the second/third foreign language versus their
399
400
1st foreign
language
2nd foreign
language
377
300
135
408
330
140
401
180
109
965
739
275
Table 1. The total number of activated words of twins in the foreign languages
versus in the mother tongue
Figure 1. The average number of activated words per person in the languages
The average number of activated words in the different foreign languages, comparing
them to mother tongue, is summarized in Figure 1. Although the twins results were
better than the results of the control group in terms of the mother tongue in the special
class (paired sample test: t(1)=6,065; p=0,104), the number of activated words for
twins studying in the normal class was lower than in the control group (one sample t
test: t(1)=4,782; p=0,131). The tendency was similar to that in respect of the foreign
languages. In the special class both the members of identical twin boys and girls had
almost the same result as the control group. But in the normal class the number of the
students activated words was also lower than in the control group (one sample t test,
t(1)=9,385; p=0,068). Comparing these results to the data obtained in the mother
tongue there were also higher differences than in the control group for the twins
studying both in the special and in the normal classes. In terms of the second foreign
language the differences between the twins and the control group can be seen mainly
in terms of the girls achievement, however it is not significant (one sample test:
t(1)=3,232; p=0,191).
1st language
English
twins
control
2nd language
German
twins
402
control
English
twins
control
German
twins
control
148
165
149,6
67,5
92,5
160,6
70
103,6
Table 2. The number of activated words in terms of the types of foreign language
403
difference in terms of second foreign language English, the opposite tendency was
seen in German. The twins activated words were a little bit longer than in the control
group.
Figure 3. The ratio of words in terms of the number of syllables in the mother tongue
The analysis of activated words in terms of the ratio of their different lengths showed
the following. Our mother tongue is a kind of agglutinative language, so it is
characterized by a high ratio of longer words of two or three syllables containing more
morphemes. Consequently in the mother tongue (Figure 3) the highest ratio was of the
words including two syllables independent of the type of class, and the ratio of these
words among the twins was higher than in the control group. The ratio of the shortest
words consisting of one syllable was also higher among the twins both in the special
and in the normal classes on the one hand, and the ratio of longer words was
significantly lower on the other. The differences between the twins and the control
group regarding the ratio of words with 3, 4 and 5 syllables were significant both in
the special and in the normal classes (in the special class: one sample test t(9)=3,367,
p=0,008; in the normal class t(9)=73,227, p=0,010). On the basis of these findings it is
also important to conclude that the twins words in the mother tongue seemed to be
not only shorter but much more simple than in the control group.
The differences between the twins and the control group were also confirmed in the
first and in the second foreign languages (Figures 4 and 5).
404
Figure 4. The ratio of words in terms of syllables in the first foreign language
Figure 5. The ratio of words in terms of syllables in the second foreign language
As we expected, in the English language the highest ratio was of words consisting of
only one syllable. There were significant differences in terms of the ratio of these
words between the twins and their peers (one sample t test: t(3)=25,791; p=0,000). In
terms of words consisting of two syllables, the differences were also significant
between the twins and the control group (one sample t test: t(3)=19,514; p=0,000).
Among the twins the ratio of the longer words especially with four and five syllables
was really low, compared to the control groups results. German is characterized
mainly by longer words, consequently in this language the ratio of short words is
much lower than in English. This tendency was followed in our examination of the
German language, where significant differences between the twins and the control
group in terms of the words with different syllables were also confirmed. Although the
405
ratio of two syllable words was the highest among the twins, as it was in the control
group, their value was much lower than in the control (t(1)=24; p=0,027). The ratio of
two syllable words was higher than in the control group (t(1)=13,476; p=0,047), but
the ratio of longer words showed the opposite tendency: they were found to a greater
extent in the control group. It also confirms the possible differences regarding the
structure of stored mental words of twins and their peers.
In the second foreign language the tendency was the same as it was in the first
foreign language on the one hand, and the differences between the twins and the
control group in terms of the ratio of various syllables were also significant on the
other. These findings have confirmed that the mental words activated by twins are
very simple.
verb
noun
adjective
spec class
80,5
10,5
spec. contr.
84
7,5
normal
99,6
0,4
control
4,4
85,2
8,3
adverb
pronoun
partic/
gerund
0,6
0,5
0,6
0,4
0,3
0,3
0,3
0,1
0,8
0,2
0,6
0,4
numeral
other
The analysis of the word classes showed the following. Although in the mother tongue
the highest value was the ratio of nouns both among the twins and in the control group
(Table 3), the differences between them were significant (one sample t test
t(1)=31,214; p=0,020). It is also important to mention the difference between the twins
studying in the special and in the normal classes in terms of the ratio of various word
classes. The twins in the normal class were able to activate almost only nouns,
however their peers studying in the normal class showed the opposite tendency, as
they also activated the other word classes. The twins verbs and adjectives were lower
than the control groups, and there was also a difference in terms of the ratio of
adjectives regarding the types of class for the twins (t(1)=5,320, p=0,118 verbs;
t(1)=5,118, p=0,123 adjectives).
In the first foreign language (Table 4.) the differences between the twins and the
control group was also confirmed.
noun
adjective
numeral
adverb
spec.class
57,4
12,8
4,9
2,6
5,9
6,4
spec.contr.
6,4
74,8
10,8
0,5
0,4
0,7
4,6
1,8
406
pronoun
partic/
gerund
verb
other
normal
norm.contr.
3,3
88,3
5,2
1,6
1,6
15,9
63,8
12,1
1,2
0,9
2,5
Although the rato of nouns was highest both among the twins and in the control group,
the ratio of them is much higher among twins (t=40,6; p=0,016). The ratio of verbs in
the control group was twice that among the twins. The ratio of adjectives was a little
bit higher in the control group; however the difference between them and the twins is
not significant (t1=8,5; p=0,075). The opposite tendency was seen in terms of other
kinds of word class, especially regarding the ratio of numerals, pronouns and adverbs.
The ratio of them was higher among the twins compared to the control group, however
these were not significant (t(1)=1,708; p=0,337 numerals; t(1)=4,364; p=0,143
pronouns; t(1)=2,000; p=0,295 adverbs).
verb
noun
adjective
spec.class
6,1
52,1
spec.contr.
9,4
66,6
8,8
numeral
adverb
pronoun
partic/
gerund
other
18,1
3,7
1,6
1,9
7,5
4,4
0,4
2,9
3,9
3,1
In the second foreign language (Table 5) the differences can be seen in terms of the
ratio of verbs, nouns, and pronouns in the same way, as the ratio of them is lower than
in the control group, but in the case of numerals and the other kinds of word class it is
higher. The ratio of nouns among twins was much lower, but the ratio of numerals and
in the other kinds of word class the ratio of conjunctions was much higher in the group
of twins than in the control. These data obtained have also confirmed the possible
difference in terms of the size of the mental lexicon of twins compared to their peers.
4. Conclusions
In this investigation our aim was to analyse the size and structure of identical twins
mental lexicon in the first and second foreign languages and to compare the data
obtained to the results in the mother tongue and in their control group consisting of
peers who are single children.
In our preliminary examination there were significant differences in terms of the
size and structure of twins mental lexicon in the mother tongue. Consequently our
hypothesis was that differences can also be found in the process of first and second
foreign language acquisition and they may occur to a greater extent than in the mother
tongue.
407
The results have confirmed the hypothesis. Regarding the average number of
activated words in the foreign languages, the difference between the twins and the
control group has depended on the type of the class the twins belonged to.
Consequently, the number of the twins activated mental words was less in the foreign
languages only in the normal class. But regarding the length of words, the ratio in
terms of syllables and the ratio of various word classes, there were differences
independently of the type of the class. The twins achievement was weaker compared
to their peers. The activated words of twins were much shorter, the ratio of longer
words in the mother tongue and in German was lower, however the ratios of one and
two syllable words in English were much higher than in the control group. There were
differences in terms of the ratio of various word classes in the mother tongue and the
foreign languages as well.
On the basis of these findings it is important to emphasize that the size of the
lexicon may be less than that of their peers. It is also important to conclude that the
twins mental lexicon in foreign languages may also be different from that of their
peers. It may contain very simple words, not only in terms of the length of them but
also in terms of the word classes. Although the investigation was carried out with the
participation of three pairs of identical twins of teenage years, a similar tendency was
exhibited in both the mother tongue and the foreign languages. Consequently, the
backwardness in terms of their mental lexicon is not presumably equalized by the
influence of school studies. In order to be able to make this claim, an investigation is
needed with the participation of a greater number of identical twins.
References
Aitchison, J. 1987. Words in the mind. Oxford: Blackwell.
Cser Jnos 1939. A magyar gyermek szkincse. Budapest.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Corvina Kiad.
Gsy Mria Kovcs Magdolna 2001. A mentlis lexikon a szasszocicik tkrben. Magyar
Nyelvr. 125: 330353.
Laczk, M. 2010. About the mental lexicon in mother tongue and in foreign language. Tudomny az
oktatsrt oktats a tudomnyrt. Science for Educaion Education for Science. Univerzita
Konstantina Filozofia v. Nitre. Fakultastredoeurpskych stdii. Nitea. 1528.
Laczk Mria 2011. Ikergyermekek nyelvi fejlettsgrl az anyanyelv-elsajtts zr szakaszban.
Alkalmazott Nyelvtudomny. XI/1-2: 89114.
Levelt, W. J. M. 1989. Speaking. From intention to articulation. Cambridge: A Bradford Book.
Mrei Ferenc V. Bint gnes 1985. Gyermekllektan. Budapest: Gondolat Kiad.
Mtneki Jlia 1997. Ikrek knyve. (A fogamzstl a felnttkorig). Budapest: Seneca Kiad.
Navracsics Judit 2007. A ktnyelv mentlis lexicon. Budapest: Balassi Kiad.
Singleton, D. 1999. Exploring the Second Language Mental Lexicon. Cambridge: CUP.
Vassn Kovcs Emke 1969. Logopdia jegyzet. Budapest: Tanknyvkiad.
408
1. Introduction
Bilingual language processing is always more complicated than monolingual, and this
is influenced by a number of factors (cf. Grosjean 1998) among which linguistic
typology might have a crucial role. Typologically similar languages (e.g. English and
Dutch, Spanish and French) might trigger different processing procedures than
typologically distant ones (e.g. English and Hungarian).
In order to understand meaningful utterances, a number of pieces of heterogeneous
linguistic information must be processed and integrated by the human language
comprehension system. The goal is to extract the meaning of a given message, i.e.
seemingly the semantics of a given word string. However, the comprehension of
language is more than the identification of word meanings. Without syntactic
processing, without the recognition of syntactic relations between the words we will
not go far since this is a necessary precondition for normal sentence comprehension.
1.1. Models of Language Comprehension
Two mainstream models have been elaborated for language comprehension. Both agree
that syntactic and semantic information have to be integrated within a minimal duration
of time in order to achieve rapid understanding. What they differ in is the time allotted
to information processing. Interactive models suggest that the semantic and syntactic
information are processed as soon as they are available, in an interactive manner
(MacWhinney Bates 1989, Marslen-Wilson 1984). On the other hand, modular
models assume that the parser constructs the syntactic structure bottom-up, independent
of semantic information, based on the syntactic structure required by the word
categories, and semantic retrieval comes only second (SmithTsimpli 1995, Frazier
1988, Fodor 1983).
In Friedericis neurocognitive model (Friederici 2002) three phases are separated.
During Phase 1, at the initial stage, local phrase structure building based on word
category information takes place. Only in Phase 2 do semantic and further syntactic
processes come into play, when syntactic and semantic as well as thematic relations
between the words are checked in parallel. The final stage (Phase 3) is the result of
these parallel processes, which is gained by the interaction of these information types,
and which allows for the final interpretation.
Regarding Hungarian, it has been shown that morphosyntactic constraints are of
primary importance in processing (Plh 1998, Gsy 1999, 2005). However, Gsy (1999)
10
suggests that the temporal aspects of syntactic and semantic processing have to be
studied.
1.2. Behavioural Studies in Bilingual Language Processing
Psycholinguistic research into L2 acquisition has been focusing mainly on storage
hypotheses, on language representations, on the organization of the lexical and
conceptual systems and on bilingual lexical access (cf. Pavlenko 2009, Singleton
1999, KrollStewart 1994, de Groot 1992). Little has been discovered about how
language comprehension takes place in L2.
Recently, bilingual language processing has been investigated with behavioural
studies and ERPs. Behavioural studies have found evidence for both the modular and
the interactive models. Two types of behavioural studies are introduced in the
literature: auditory, and visually presented, comprehension tests. In such tests,
acceptability judgments and reaction times are analyzed. Here we provide some of the
results gained from visually presented materials.
Rossi et al. (2006) found that L2 syntactic processing was age and language
proficiency dependent. They carried out a sentence comprehension test on German
learners of Italian and Italian learners of German, and they found that native-like
syntactic processing profiles could be seen in late proficient L2 learners. Kotz et al.
(2008) studied SpanishEnglish early bilinguals L2 (English) syntactic knowledge and
examined to what extent their performance is comparable with that of native readers of
English. The offline behavioural results showed no significant differences between L2
readers and native readers of English. However, the online test demonstrated some
discrepancies, which, according to the authors, might be due to the time pressure during
the acceptability judgment.
Proverbio et al. (2002) in an ERP and reaction times study of acceptability
judgments investigated and compared ItalianSlovenian bilinguals and Italian
monolinguals. The reaction times were much faster for the monolingual group. For the
bilingual group RTs were significantly slower in respect of semantic violations in both
languages. In syntactic violation conditions, judgments were faster in Slovenian than
in Italian, but there was no difference between the languages in the judgments of
semantic violations. The accuracy analysis showed no difference between the
languages in the correct response percentages, but did show faster and more accurate
judgment of semantic violations in Italian and syntactic violations in Slovenian.
Weber and Lavric (2008) tested 18 GermanEnglish highly proficient bilinguals
and compared their results with those of monolingual English subjects. Their results
support the previous results with judgment accuracy higher for the syntactically
anomalous sentences than for the correct and semantically anomalous sentences. In all
conditions, L1 scoring was higher than L2 results. Weber and Lavric, just like Kotz
(2009) emphasize the importance of syntactic transfer between L1 and L2 as it may
critically influence L2 syntactic knowledge acquisition.
In this paper, we check the reliability of the proposal regarding the differences in
bilingual processing when the bilingual configuration is composed of an agglutinating
(i.e. Hungarian) and a non-agglutinating (i.e. English) language. We propose that the
410
morpho-syntactic constraints are perceivable sooner in the case of Hungarian, and thus
the syntactic processing is faster and more accurate than in English. As for semantic
processing, we hypothesize that in both languages it is a longer process when compared
to syntactic comprehension with decisions in Hungarian being more accurate than in
English.
2. Subjects. Methods
2.1. Participants
64 Hungarian dominant HungarianEnglish voluntary subjects (42 female and 22 male)
participated in the study. The age range was 18 to 57, mean age: 32.2 years. They are all
but seven right handed with no known neurological history. All participants had normal
or corrected to normal vision. They all live in Hungary. 45 of them belong to the late and
19 to the early bilingual category, considering the fact that most behavioural data find
critical ages 10 or 11, in terms of the definition concerning early vs. late bilingualism.
All of them have spent a certain amount of time in their lives in an English speaking
country. They all are exposed to English on a daily basis, as the younger ones are
university students; the older ones are teachers or researchers at universities, who use
English in their everyday professional life. Reading in English is an important activity
for the older generation, the younger ones prefer the oral manifestation of the language
(i.e. speak or watch and listen to films, music, lectures etc.). Only three of them became
bilingual in natural circumstances, not in school or kindergarten, two of them in the USA
in their childhood, one as an adult.
2.2. Material
240 sentences were processed by the participants, 120 English and 120 Hungarian. In
both languages, 60 sentences were correct both semantically and syntactically, they
served as control sentences. In addition, there were 60 sentences in each language, 30
of which were semantically and 30 syntactically anomalous. A proportion of the
Hungarian violated sentences was taken over from other tests (Gsy 1996, Gsy 1997,
OsmannBnrti 1997, Nmeth 2006).
English syntactically violated sentence types included: word order, number
agreement, person agreement, prepositions, personal pronoun, and tense agreement.
Hungarian syntactically violated sentence types included: word order, number
agreement, verb argument structure. English semantically anomalous sentence types
included: semantically incongruent critical word in final position, in internal position.
Hungarian semantically anomalous sentence types included: semantically incongruent
critical word in mid- and in final positions.
In this study, we concentrate on the processing of violated word order and
agreement. We will analyse the results of the following sentences, and use the
abbreviations given in each category:
411
413
No significant difference was shown between the reaction times in the decisions about
syntactically correct and anomalous sentences, and nor was one shown between those
in the decisions about semantically congruent and incongruent sentences.
The mean reaction time was significantly shorter in the recognition of syntactically
correct sentences than in the case of semantically meaningful sentences (2.21 (0.68) s
vs. 2.35 ( 0.71) s, p=0.0000), as shown in Figure 2.
414
415
416
However, there was no difference between the correct acceptability judgment rates; in
both languages the accuracy rate of the decisions was 93%.
3.6. Violations according to languages
There was a significant difference between the reaction times of the processing of
syntactic violations and semantic anomalies in the English sentences (2.38 (0.76) s,
vs. 2.55 (0.70) s, p=0.0060), as shown in Figure 8. However, there was no such
difference regarding the Hungarian violated sentences.
417
4. Discussion
Our starting point was to hypothesize that syntactic processing is faster than the
semantic one. This has been confirmed by our results which are in accord with the
findings of Proverbio et al. (2002). With the linguistic typological differences between
the Hungarian and English languages, it was believed that structural violations were
sooner recognized than semantic incongruence. Considering Friedericis
neurocognitive model, in Phase 1, where the local phrase structure is checked,
Hungarian with its morpho-syntactic constraints must help the reader with the decision
about the acceptability of the sentence as correct (cf. Lengyel 1984, Gsy 1996, Plh
1998). At the same time, the word order differences between the languages must also
contribute to the successful solution of the task. Hungarian is an SOV while English is
an SVO language. In addition, English has a strict word order while Hungarian is quite
flexible, and pragmatic rules (focus, topic-comment etc.) may overwrite the basic
rules.
In our results, we find that in spite of the fact that English seems to be stricter in
terms of word order, the results of the Hungarian sentences were better, with shorter
reaction time and higher accuracy rate on average than those of the English ones (cf.
WeberLavric 2008). This might be explained by the English word order violated
sentences being mirror translations of the Hungarian word ordered sentences, which
could be viewed as cross-linguistic interference, and that is why they were accepted
incorrectly. The violated English sentence patterns follow those of Hungarian, i.e.
SOV. On the other hand, the Hungarian violated sentences do not follow the pattern of
the English word order, since from a pragmatic point of view they could as well have
been correct in Hungarian. The Hungarian violated sentences are distorted to such
extent that they are absolutely wrong without any hesitation, e.g. the predicate
precedes the subject.
In the sentences with agreement violation conditions, the difference between the
mean reaction time of the judgements about the two languages is smaller but still
significant. At the same time, the acceptability judgement rate shows a greater
difference (30%). This underlines the hypothesis that unlike in English, in Hungarian
the checking of the local phrase structure makes it easier to decide whether the phrase
is correct or not. Violations in the English sentences might be misleading as the
structures might function correctly in other linguistic contexts. Wrong prepositions as
complements can also trigger false decisions on account of the time pressure in the test
situation.
Semantic processing is slower than syntactic, but the position of the critical word is
crucial in the judgment. In both languages, semantic anomaly was recognized faster if
the critical word was in end-position, which shows that if there is a hold-up in
processing in the middle of the sentence, it takes longer to check the meaning and
restart processing.
It took longer to make decisions in the English sentences, which suggests that
semantic processing in L2 is more complicated and a longer process than in L1.
418
Table 1. and 2, show a summary of the data, ordered in Table 1. from the smallest
to the greatest reaction times, and in Table 2. from the highest accuracy rate to the
lowest.
Violation type
RT
AJ
WOH
1.99
96.87%
EPH
1.99
92.96%
AGH
2.07
94.92%
MPH
2.08
94.53%
WOE
2.35
85.15%
AGE
2.40
65.23%
EPE
2.44
93.35%
MPE
2.78
68.35%
From Table 1. it is obvious that languages create clusters in terms of reaction time.
The fastest decisions were made in the Hungarian sentences, irrespective of the
violation type. Even the longest mean reaction time in Hungarian semantically
anomalous sentences is shorter than the fastest reaction time measured in syntactically
violated English sentences. While in Hungarian, the same RT value was registered in
the word order violation and semantic anomaly in end-position, and there is hardly any
difference between the two other categories, in English, the reaction time differences
are greater between the categories, and syntactic violations are recognised faster than
the semantic ones.
Violation type
RT
AJ
WOH
1.99
96.87%
AGH
2.07
94.92%
MPH
2.08
94.53%
EPH
1.99
92.96%
EPE
2.44
93.35%
WOE
2.35
85.15%
MPE
2.78
68.35%
419
AGE
2.40
65.23%
If our data are ranked according to the accuracy rate in acceptability judgments, we
see the clusters between the languages again. However, we found significant
differences only in two cases: at the comparison of judgments of agreement violation
(65.23% and 94.92%, Chi-square test, p=0.005) and at the comparison of semantic
anomalies in the two languages with critical word in mid-position (68.35% and
94.53%, Chi-square test, p=0.014). The best results were achieved in Hungarian.
5. Conclusions
Our data have provided evidence that in accord with the findings of Indo-European
paired bilingual tests syntactic processing takes less time than the semantic one. It
also shows that the language of the immediate environment is processed faster, no
matter how frequently the other language is used on a daily basis. We find it
particularly interesting that sentences with semantic anomaly in end-position are
processed faster in both languages, and with no difference in judgment accuracy in the
two languages. This underlines the importance of the context as an aid in semantic
processing.
References
Bnrti Zoltn 1997. Megjegyzsek a neurolingvisztikrl. In Bnrti Zoltn (szerk.): Nyelvi struktrk s az agy. Budapest: Corvina Kiad. 758.
de Groot, A. 1992. Bilingual lexical representation: a closer look at conceptual representations. In
Frost, R. Katz, L. (eds.): Orthography, Phonology and Meaning. Amsterdam: Elsevier. 389
412.
Fodor, J. 1983. The Modularity of Mind. Cambridge: Cambridge University Press.
Frazier, L. 1988. Grammar and language processing. In Newmeyer, F. (ed.): Linguistics: the Cambridge Survey. Cambridge: Cambridge University Press. 1534.
Friederici, A. D. 2002. Towards a neural basis of auditory sentence processing. Trends in Cognitive
Sciences. 6: 7884.
Gsy Mria 1996. GMP. Beszdszlelsi s beszdmegrtsi teljestmny. Budapest: Nikol Gmk.
Gsy Mria 1997. Listening to English. Tesztcsomag az angol nyelvi beszdszlels s beszdmegrts vizsglatra. Budapest: Nikol Gmk.
Gsy Mria 1999. Pszicholingvisztika. Budapest: Corvina Kiad.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Osiris Kiad.
Grosjean, F. 1998. Studying bilinguals: Methodological and conceptual issues. Bilingualism:
Language and Cognition. 1: 131149.
Kotz, S. A. 2009. A critical review of ERP and fMRI evidence on L2 syntactic processing. Brain and
Language. 109: 6874.
Kotz, S. A. Holcomb, P. J. Osterhout, L. 2008. ERPs reveal comparable syntactic sentence
processing in native and non-native readers of English. Acta Psychologica. 128: 514527.
420
Kroll, J. Stewart, E. 1994. Category Interference in Translation and Picture Naming: Evidence for
Asymmetric Connections Between Bilingual Memory Representations. Journal of Memory and
Language. 33: 194174.
Lengyel Zsolt 1984. A 3-10 ves kor orosz s magyar gyermekek verblis magatartsnak sszehasonlt vizsglata. Kzirat. Budapest.
MacWhinney, B. Bates, E. 1989. The cross-linguistic study of sentence processing. New York:
Cambridge University Press.
Marslen-Wilson, W. 1984. Function and process in spoken word recognition: A tutorial review.
Attention and Performance. Control of Language Processes. 10: 125148.
Nmeth Dezs 2006. A nyelvi folyamatok s az emlkezeti rendszerek kapcsolata. Budapest: Akadmiai Kiad.
Osmann Sgi Judit (Hungarian adaptation) Western Aphasia Battery (WAB)
Pavlenko, A. 2009. Conceptual representation in the bilingual lexicon and second language
vocabulary learning. In Pavlenko, A. (ed.): The bilingual mental lexicon. Interdisciplinary
approaches. BristolBuffaloToronto: Multilingual Matters. 125160.
Plh Csaba 1998. A mondatmegrts a magyar nyelvben. Budapest: Osiris Kiad.
Proverbio, A. M. Cok, B. Zani, A. 2002. Electrophysiological measures of language processing
in bilinguals. Journal of Cognitive Neuroscience. 14: 9941017.
Rossi, S. Gugler, M. F. Friederici, A. D. Hahne, A. 2006. The impact of proficiency on
syntactic second-language processing of German and Italian: evidence from event-related
potentials. Journal of Cognitive Neuroscience. 18: 20302048.
Singleton, D. 1999. Exploring the second language mental lexicon. Cambridge: Cambridge University Press.
Smith, N. Tsimpli, I. 1995. Language Learning and Modularity. Oxford: Oxford University Press.
Weber, K. Lavric, A. 2008. Syntactic anomaly elicits a lexico-semantic (N400) ERP effect in the
second language but not the first. Psychophysiology. 45: 920925.
421
hasznos lehet. Az a kutats, amelynek a hipotziseit, valamint a szksgessgt a kvetkezkben vzolom, beleillik abba sorba, amelyet mindenekeltt Gsy Mria, Lengyel Zsolt, Navracsics Judit, Gncz Lajos pszicholingvisztikai kutatsai, Kovcs Lszl nyelvi hlzatokat feltr vizsglatai, Brdos Jen nyelvpedaggiai kutatsai alapoztak meg. A szasszocicis vizsglatok httert alapveten Lengyel Zsolt s Kovcs Lszl munki jelentik, valamint az a flves doktori szeminrium, amelyet a
Pcsi Tudomnyegyetem Nyelvtudomnyi Doktori Iskoljnak keretben tartottam a
magyar mint idegen nyelv j szempont vizsglatai irnt rdekld hallgatknak. (E
trgykrben rott munkik egy rsze megjelent a THL2 cm szakfolyirat 2010. s
2011. vi szmaiban, egyebek kztt Marthy 2010.) A felvetett tma s megkzelts
fontossgt megerstettk azok a magyarnyelv-tanr kollgk is, akiknek Helsinkiben
egy tanrtovbbkpzs keretben (Te-Le-Hu program, 2009) a mentlis lexikon megismersnek hasznossgrl beszltem.
2. A pszicholingvisztikai kutatsok hasznossga a magyar mint idegen s mint
szrmazsi nyelv tantsa terletn
A magyar nyelvvel kapcsolatosan szmos mtosz kering. Ezeket rszben mi magunk
tartjuk letben, rszben a nyelvnk tanulsra vllalkoz klfldiek terjesztik. Magban vve rdemli meg a nyelvnk, hogy tanuljk ezekkel a szavakkal vezeti be Bl
Mtys a magyar nyelv gazdagsgt, szpsgt, nemes hangzst ecsetel gondolatait
az 1729-ben kiadott Der Ungarische Sprachmeister cm nyelvknyvnek elszavban (magyarul: Bl 1984: 169173).
A fonetikai bzis elsajttsa, a hosszrvid magn- s mssalhangzk
jelentsmegklnbztet szerepe, a krdsz nlkli krd mondat intoncija, a
nyelvtanul ltal ismert nyelvek szkincshez nehezen kthet lexika, az igekts
igk formai s jelentsbeli gazdagsga, a mondat nehezen megfejthet szrendje
okozza a legtbb gondot, s tpllja folyamatosan nyelvnk nehz voltrl, vgs
soron pedig a megtanulhatatlansgrl kialakult mtoszokat. De mennyire valsak
ezek a mtoszok? Tehetnk-e valamit annak rdekben, hogy a nyelvet tlthatbb,
knnyebben tanulhatv tegyk? A nyelv kls szemlletn alapul, de a
pszicholingvisztika anyanyelvi bzisokra tmaszkod eredmnyeinek alkalmazsa a
nyelvpedaggiai folyamatokban (egyebek kztt a nyelvtanrok kpzsben), a tanknyvek felptsben, a tananyag sorrendjben, az egyes nyelvi szintekhez rendelt
lexika kivlasztsban, a tanknyvek dialgusaiban s narratv szvegeiben, valamint
a hangz s multimdis anyagokban hozzjrulhat a sok vszzada jelen lv mtoszok eltnshez. Igaz, nhnyan feltehetik a krdst: ha mr mtoszaink sem maradnak, mirt vlasztank a nem indoeurpai nyelvek tanulsra htozk ppen a magyar
nyelvet, hiszen a nyelvtanulk jelents szzalkt mg mindig a nehzsg, a klnlegessg, az egyedi logika kihvsa (Lanstyk 2009: 155) vonzza a nyelvtanfolyamokra
rdemes kutatni a msodik, illetve az idegen nyelvi beszdprodukci kialakulsnak folyamatt a klnfle els nyelvekkel trtn sszevetsben: vannak-e hasonlsgok az anyanyelvt elsajtt gyermek grammatikai stcii s a nyelvtanul grammatikai tudsnak formldsa kztt (idertve a hibzsok, tvesztsek jellegt is)?
423
425
nulk ezeket a nyelvi formkat (tanulj, rtsd meg, olvasd stb.) nll egysgknt troljk, szksg esetn aktivljk, de csak a hallott s rgztett formban.
4. A kutatsrl
A magyar mint idegen s mint szrmazsi nyelv tanuli mentlis lexikonjnak vizsglatt Lengyel Zsolt 197 elembl ll hvszlistja segtsgvel (2008, 2009) vgeztem el
kzel szz nem magyar anyanyelv egyetemi hallgatval, valamint kontrollknt 50
magyar szakos egyetemi hallgatval. Az adatkzlket tbb csoportra osztottam: a)
magyar, ill. hungarolgia szakos egyetemi hallgatk, akik magyarorszgi rszkpzsben vettek rszt; b) klfldi egyetemen (forrsnyelvi krnyezetben) tanul dikok; c)
magyar szrmazs, a nyugati diaszprbl rkezett, magyarorszgi intzmnyben
tanul dikok. Terveztem mg a hatron tli magyar egyetemi hallgatk bevonst is,
de az ms krdsek felvetst indokoln, pldul a tbbsgi nyelv kontaktusjelensgeinek, a helyi s a magyarorszgi nyelvvltozatok megjelensnek vizsglatt, gy azzal
ksbb szeretnk foglalkozni.
Felttelezhet, hogy a B1 s B2 szinten lv magyarnyelv-tanulkkal elvgzett,
egyvlaszos szasszocicis teszt esetben is fontos szerepet fog jtszani a tanult
nyelvknyvek szkincse. A B1 szinten lvknl nagyobb szmban fordulnak el fonetikai hasonlsgon alapul elhvsok, a magasabb szinteken viszont a szfaji ktds
(nvszra nvsz, igre ige) lesz a jellemz. Valsznleg lesznek olyan esetek, amikor nem egy vlaszt rnak. Ez alkalmat ad a fentebb vzolt, nem azonos sly asszocicik hlzatnak vizsglatra. A magyar clnyelvi krnyezetben, heti 20 rban tanulk esetben B2 szinten legalbb 50-60%-os egyezsnek kell lennie a magyar anyanyelv kontrollcsoporttal, a B1 szinten ez 40% krl lesz. Ebben a specilis esetben
nem hanyagolhat el a nyelvknyvek lexikai anyagnak ttekintse sem, hiszen ez lesz
az egyik viszonytsi pont mind a clnyelvi, mind a forrsnyelvi krnyezetben tanulk
vlaszszavainak feldolgozsnl. Jelenleg a szavak rgztse folyik a Kroli Gspr
Reformtus Egyetem magyar mint idegen nyelv szakos hallgatinak bevonsval, gy
a jv vi konferencin remlhetleg mr az eredmnyekrl is szmot adhatok.
Irodalom
Aitchison, J. 1987. Words in the Mind. Oxford: Blackwell.
Brdos Jen 2006. Mdszertan vagy nyelvpedaggia? THL2. 1-2: 514.
Bartha Csilla 1999. A ktnyelvsg alapkrdsei. Budapest: Nemzeti Tanknyvkiad.
Bl Mtys 1984. Hungaribl Magyarorszg fel. Szerk. Tarnai Andor. Budapest: Szpirodalmi
Knyvkiad.
Giay Bla 2008: A magyar mint idegen nyelv / hungarolgia oktatsnak mdszertani alapkrdsei.
In Hegeds Rita Ndor Orsolya (szerk.): Magyar nyelvmester. Budapest: Tinta Knyvkiad.
125144.
Gsy Mria 2000. Az letkor hatsa a mentlis lexikon mkdsre. Magyar Nyelvr. 124: 410423.
Gsy Mria 2005. Pszicholingvisztika. Budapest: Corvina Kiad.
Gsy Mria Kovcs Magdolna 2001. A mentlis lexikon a szasszocicik tkrben. Magyar
Nyelvr. 125: 330354.
426
427