Professional Documents
Culture Documents
Nose Uralica
Nose Uralica
Japanese by visualization
Masahiko NOSE
Reitaku University
mnose@reitakuu.ac.jp
Abstract
This study is an attempt to visualize the case functions of Uralic languages and Japanese. A
typological study conducted by Iggesen (2005) showed that the number of cases among the world’s
languages varies from zero to more than 10 per language. There are many Uralic languages with more
than 10 cases as well as rich case system languages in the Caucasus, Australia, and North America.
This paper examines the case forms and functions of 13 Uralic languages and Japanese as samples.
Using a biological software, the case functions of the languages are visualized in terms of universally
identifiable case functions. Finally, this study claims that the case functions of each language group
(Finnic, Ugric, and Samoyed) were largely similar with some exceptions. The case functions of
Japanese were shown to be similar to those of Mari.
Keywords: case, function, visualization, WALS, Uralic, Japanese
1. Introduction
It is well known that there are rich case systems in the Uralic languages. Finnish has 15
cases, and Hungarian has 18 cases. In contrast, Northern Saami has 7 cases, Northern Khanty
has only 3 cases, and some Uralic languages have very few cases 1 . This paper examines the
functions of cases in the Uralic languages, and attempts to visualize the case functions.
Furthermore, the paper explores the functional relationship and differences between the Uralic
languages in terms of the case functions.
Many previous studies have discussed Uralic case systems in terms of their historical,
descriptive, and cognitive background, and have tried to explain the languages’ diverse and
complicated case systems (e.g., Abondolo 1998, Koizumi 1994). This study selects a sample
of 13 Uralic languages and tries to analyze Uralic case functions using bioinformatic software.
This is an attempt to visualize (a part of) Uralic grammar that previous studies have not dealt
with, and this type of visualization will help to observe functional relationships among the
sample languages.
This paper adds Japanese as the sample language. Japanese is a language isolate and not
part of the Uralic language family. Nevertheless, it is a language with a rich case system.
There are many case forms in some Uralic languages and Japanese, but their case functions
are not considered to be equivalent. This paper will clarify which function(s) Uralic and
Japanese have in common, and especially which Uralic language is most similar to Japanese
using the functionbased visualization.
2. Case functions and classification
The Uralic languages have several case forms and case functions, and it is necessary to
define these case forms and functions crosslinguistically. As Blake (1994:1) describes,
“[c]ase is a system of marking dependent nouns for the type of relationship they bear to their
heads. Traditionally the term refers to inflectional marking, and typically, case marks the
relationship of a noun to a verb at the clause level or of a noun to a preposition, postposition
or another noun at the phrase level.” Following Blake’s definition, Iggesen (2005) examined
261 languages for the number of cases, as presented in Table 1.
Table 1: Number of cases in 261 languages by Iggesen (2005:202205)
No morphological case marking 100
2 case categories 23
3 case categories 9
4 case categories 9
5 case categories 12
67 case categories 37
89 case categories 23
10 or more case categories 24
Exclusively borderline morphological case marking 24
Iggesen’s (2005) study is based on “The World Atlas of Language Structures” (Haspelmath et
al. 2005, henceforth WALS). Figure 1 shows a geographical map generated by WALS using
Iggesen’s (2005) data, where white triangles represent the languages with no case marking
and black ones represent the languages with more than 10 cases 2 .
Figure 1. Languages with no case marking and languages with more than 10 cases by Iggesen
(2005) “Number of Cases” (feature 49, WALS: 202205)
(1) Sample languages in this study:
· Finnic: Mari, Erzya Modrvin, Estonian, Finnish, Komi, Northern Saami, Udmurt
· Ugric: Eastern Khanty, Northern Khanty, Hungarian, Mansi
· Samoyed: Enets, Nganasan
· Other (nonUralic): Japanese
Japanese (nonUralic) is known for its rich case system, and has been chosen to show contrast.
First, I examine the number of cases in each language; then, I examine the forms and
functions of cases according to the descriptive grammars. Table 2 presents the approximate
number of cases collected from descriptive grammars of the individual languages. In all, this
paper examines 13 Uralic languages and Japanese as sample languages.
Table 2: Number of cases in the sample languages
Mari 11 Northern Saami 7 Mansi 6
Erzya Modrvin 13 Udmurt 15 Enets 8
Estonian 14 Eastern Khanty 9 Nganasan 7
Finnish 15 Northern Khanty 3 Japanese 11 3
Komi 17 Hungarian 18
The number of cases shown in Table 2 varies according to linguists and descriptive grammars.
It is apparent that Finnic languages have a rich case system, and Ugric ones do not have many
cases, except for Hungarian. Hungarian, in particular, has 18 cases; the case names and forms
in Table 3 are based on Keresztes (1995).
Table 3: Hungarian cases: names and forms (Keresztes 1995)
Nominative Zero Ablative tól/től
Accusative t Adessive nál/nél
Dative nek/nak Allative hoz/hez/höz
Elative ból/ből Terminative ig
Inessive ban/ben Essiveformal ként
Illative ba/be Instrumental val/vel
Delative ról/ről Translative Vá/Vé
Superessive n Causalfinal ért
Sublative ra/re Associative ostul
Nose (2006:112113) identified two problems in collecting and classifying cases: (i)
difficulty in counting the exact number of cases; cf. Borin’s (1986) study of counting cases in
Hungarian 4 ; and (ii) assigning a single function to different case names, or the same case
name to different functions. For example, Finnish, Estonian, and Mordvin have the abessive
case, indicating “without” or “lacking,” whereas Udmurt has the caritive case for largely the
same function. Moreover, Estonian and Finnish have an allative case indicating “onto,” and
the case has a dative function, and Hungarian also has the allative case, but its meaning is
“near” and “to,” excluding the dative function. Although the case name is allative, it has a
slightly different function in each of the languages concerned.
To solve these problems, several functional descriptions are given to each case form
observed in the 14 languages. For instance, Hungarian has 18 cases, and the relevant
functional descriptions have been added to each of them (see Table 5, and Table 6 for
Nganasan). By assigning several nonstrict functional descriptions to the cases, such as
[grammatical], [place], and [with], it is possible to classify them in terms of syntactic and
semantic functions. Moreover, this study introduces 25 universally identifiable case functions,
given in Table 4. These 25 functions are basic functions consisting of grammatical, locative,
and adverbial groups, and they help to classify the functional descriptions crosslinguistically
(cf. Nose 2006) 5 .
Table 4: Universally identifiable case functions
GRAMMATICAL: LOCATIVE: ADVERBIAL:
[1] Nominative [7] Generallocative [17] Abessive
[2] Accusative [8] In: Inessive [18] Causal
[3] Ergative [9] Into: Illative [19] Comparative
[4] Absolutive [10] From inside: Elative [20] Terminative
[5] Dative [11] On: Superessive [21] Essive
[6] Genitive [12] Onto: Sublative [22] Instrumentalcomitative
[13] From ontop: Delative [23] Translative
[14] Adessive OTHERS:
[15] Allative [24] Relationalconcerning
[16] Ablative [25] Vocative
For instance, the nominative case in Hungarian is assigned through functional descriptions
[grammatical] and [subject], and then it is classified as a “[1] Nominative” in terms of the
universally identifiable case functions. In the same manner, the elative case is assigned [place],
[from], and [inside], and then “[10] From inside: Elative.” Some cases, such as the socalled
associative case in Hungarian, cannot be assigned under the universally identifiable case
functions. In Table 4, the associative case is labeled by [together] under functional description,
but there is no corresponding universally identifiable case function. Thus, this study classifies
the associative case as an exception and does not deal with it.
This study gives such functional descriptions and universally identifiable case functions to
Hungarian, Nganasan, and the other 12 sample languages. Thus, the universally identifiable
case functions indicate which kinds of case functions the sample languages do and do not
carry. We can summarize case functions of the sample languages on the basis of
crosslinguistic standards.
Table 5: Hungarian cases with functional descriptions and universally identifiable case
functions
Case names Case forms Functional descriptions of cases Table4 number
Nominative Zero [grammatical][subject] [1]
Accusative t [grammatical][direct object] [2]
Dative nek/nak [grammatical][indirect [5]
object][goal][possession]
Elative ból/ből [place][from][inside] [10]
Inessive ban/ben [place][in] [8]
Illative ba/be [place][into] [9]
Delative ról/ről [place][from][on] [13]
Superessive n [place][on] [11]
Sublative ra/re [place][onto] [12]
Ablative tól/től [place][from][by][near] [16]
Adessive nál/nél [place][be][near] [14]
Allative hoz/hez/höz [place][toward][by][near] [15]
Terminative ig [till][as far as] [20]
Essiveformal ként [as][manner[][like] [21]
Instrumental val/vel [with][animate][inanimate] [22]
Translative Vá/Vé [result] [23]
Causalfinal ért [cause][reason] [18]
Associative ostul [together] Not available
Table 6: Nganasan cases with functional descriptions and universally identifiable case
functions (WagnerNagy 2002: 7681)
Case names Case forms Functional descriptions of cases Table4 number
Nominative Zero [grammatical][subject] [1]
Accusative Zero, m [grammatical][direct object] [2]
Genitive Zero, ŋ [grammatical][possession] [6]
Lative tə, -ntə [place][to][into][indirect object] [5][9][11][15]
Locative ntənu [place][in][instrumental] [6][7][12][22]
Elative kətə [place][from][ablative] [10][13][16]
Prolative mənu [place][till][along] [20]
(In addition to the seven cases shown above, the vocative function [25] (zero case form) is
observed in Nganasan (WagnerNagy 2002: 76))
3. Results and visualization
In this section, I present the adjacent relations of the sample languages in terms of the
functions of the case system. We describe the case functions of the sample languages in terms
of universally identifiable case functions and we modify the 25 case functions in “having” or
“not having” (1 or 0), as shown in (2). According to Table 4, if the language has a nominative
case, the first place ([1] Nominative) is 1, and if the language has an ergative case, the third
place ([3] Ergative) is 1.
(2) Functional descriptions for visualization:
Eastern Khanty 1000001011000111100101100
Northern Khanty 1000001000000000000001000
Mansi 1000001010000011000101100
Hungarian 1100100111111111010111100
Mari 1100110110000111101000100
Estonian 1000010111111000100111100
Finnish 1100010111111000100011100
Komi 1100110111100101110111100
Erzya Modrvin 1100110111000010101001000
Northern Saami 1100011010000000000010100
Udmurt 1100010111001011100101110
Enets 1100011010010011000011000
Nganasan 1100011011110110000101001
Japanese 1100110100000011001100110
As shown in Table 5, Hungarian has the following 17 universally identifiable case functions:
[1], [2], [5], [8], [9], [10], [11], [12], [13], [14], [15], [16], [18], [20], [21], [22], and [23].
Thus, Hungarian is described in (2) as “1100100111111111010111100.” Other languages also
can be described with 1 and 0 according to the universally identifiable case functions. These
data of each language in (2) will be used to visualize a functional map in Figure 2.
Figure 2 shows an attempt to visualize the results of the above list using a bioinformatic
software (SplitsTree4; Huson & Bryant 2005). This software can create a phylogenetic tree of
the data for each factor, and recently, it has been used in linguistic areas as well (e.g., Dunn et
al. 2005). It will be helpful in visualizing the data more easily and objectively. In this study,
we try to create the phylogenetic tree of case functions (cf. Figure 2) based on the data of
universally identifiable case functions (cf. (2)). Figure 2 demonstrates similarities between the
languages in the distribution of the universally identifiable case functions. When two
languages share many universally identifiable case functions, they are visualized as closer,
and we can therefore observe linguistic distances in terms of the 25 case functions. For
example, cases functions in Finnish and Estonian are similar, and these two languages are
situated closer in Figure 2. In contrast, Finnish and Northern Saami are situated far away in
terms of case functions, although these two languages are belonging to Finnic group.
Figure 2: Visualization of Uralic languages and Japanese based on case functions using
SplitsTree 4
4. Discussion
In this section, this study will discuss the characteristics of the case functions in sample
languages with contrasting Japanese, and the advantages of the visual representation.
With regard to the visualization of the case functions, we first examine some
characteristics of the Uralic cases. We observed that the languages of the same group (Finnic,
Ugric, and Samoyed) are situated closer together, but it is unusual that Hungarian and Komi
are closer to each other, and Northern Saami and Nganasan are closer to ObUgric languages
(Mansi, Eastern Khanty, and Northern Khanty). There are two kinds of languages in Uralic:
some (e.g. Finnish and Hungarian) have a rich case system, and others (e.g. Northern Khanty
and Mansi) do not. Such differences in the number of cases can be comprehended visually in
Figure 2. The former is situated to the left, and the latter to the right, respectively. Moreover,
the languages with a rich case system can be classified into two groups: some (Hungarian and
Komi) have many adverbial functions, and others (Estonian and Mari) do not.
Second, Nose (2006) identified the universally identifiable case functions, and this study
adopts all 25 functions. However, there are no ergative or absolutive cases in the sample
languages, and it is characteristic that there are many kinds of locative and other adverbial
functions in Uralic languages. Thus it is still necessary to construct a better definition of case
functions and to visualize them more precisely. Notably, many Uralic languages (Eastern
Khanty, Northern Khanty, Mansi, Northern Saami, Enets, and Nganasan) have a general
“locative case” ([7] in Table 4), with several functions of location. For instance, Nikolaeva
(1999:13) claims the locative case in Northern Khanty carries place and direction; passive
agent; patient argument in certain ditransitive constructions; adjunct indicating point in time,
measure, instruments or means, distributive and indirect object of some verbs. In contrast,
Hungarian, Finnish, and other languages have many kinds of specific locative cases. This
study has examined the descriptive grammars of the sample languages and confirms that there
are some Uralic languages with a few cases. Languages such as Eastern and Northern Khanty,
Mansi, Northern Saami, Enets, and Nganasan do not have many types of locative case. We
can observe a group of these languages in the visualization (Figure 2, right).
Third, we discuss functionally common characteristics among the sample languages, and
the case functions of Japanese by contrasting it with the cases of Uralic. The sample
languages have relatively rich case markings with grammatical, locative, and adverbial
functions. Some universally identifiable case functions are observed in many Uralic
languages: [1] Nominative, [2] Accusative, [6] Genitive, [8] Inessive, [9] Illative, [22]
Instrumentalcomitative, and [23] Translative. These universally identifiable case functions
are considered to be common characteristics of Uralic languages, and we point out that they
are indicative of a kind of Uralic case preference. In contrast, there are no [3] Ergative and [4]
Absolutive case functions and very few [18] Causal, [19] Comparative, [24]
Relationalconcerning, and [25] Vocative case functions. They are considered to be
uncommon in Uralic, or rare case functions.
Japanese is a typical nominativeaccusative language like Uralic. The case functions of
Japanese are similar to those of Uralic (especially Mari) in that there are dative, genitive, and
some locative case functions. However, there is no “relationalconcerning” function in Uralic,
although Japanese does have the topic wa marker. Koizumi (2007) asserts that made is in the
terminative case of Japanese, and the terminative can be observed in Hungarian, Komi,
Udmurt, and Estonian (cf. see also [20] Terminative function). Moreover, Koizumi
distinguishes between the comitative to and the instrumental de in Japanese, and several
Uralic languages (Finnish, Komi, and Eastern Khanty) also have such a distinction between
comitative and instrumental functions. Thus, Japanese and some Uralic languages are similar
in their case functions, in spite of having no genealogical relations. Japanese was considered
to be similar in case function to Finnish and Hungarian, but this study clarifies that Japanese
is closer to Mari and Erzya Mordvin, according to Figure 2. Future study is necessary to
examine the similarities among the three languages by adding lexical (e.g., Swadesh’s list of
basic vocabulary) and other grammatical data.
5. Conclusion
This paper focuses on one of the WALS features, the “number of cases,” and has applied it
to the numbers and functions of cases in the Uralic languages. We have attempted a
visualization of the case functions of the Uralic languages. There are 3 to 18 cases in the
Uralic languages and Japanese, and as a result of summarizing cases based on the universally
identifiable case functions, the following points have become apparent.
First, the case distributions of Uralic are not always consistent with genealogical or
geographical distributions. For example, Finnish and Estonian are similar, and Hungarian and
Komi are closer, but Finnish and Northern Saami are separately placed, and Hungarian and
other Ugric languages are separate, too. Second, when a language has a rich case system, it
has many locative and adverbial cases. When a language has fewer cases, it has at least one
locative case which sums up several functions.
Finally, this study has added Japanese for the purpose of contrast. Japanese has 11 cases
and is similar to one Finnic language, Mari. The grammatical and locative functions in
Japanese are different from those in Uralic, but the two languages share several locative ([15]
Allative, and [16] Ablative) and adverbial ([19] Comparative, and [23] Translative) functions.
FOOTNOTES
References
A. Descriptions of sample languages
Mari
KangasmaaMinn, Eeva 1998. Mari, In: Abondolo(ed.): 219248.
Eastern Khanty
Csepregi, Márta 1998. Szurguti Osztják Chrestomathia. Szeged: JATE Finnugor Tanszék.
Enets
Künnap, Ago 1999. Enets. München: Lincom Europa.
Erzya Mordvin
Keresztes, László 1990. Chrestomathia Morduinica. Budapest: Tankönyvkiadó.
Estonian
Oinas, Felix J. 1966. Basic course in Estonian. Bloomington: Indiana University Press.
Finnish
Karlsson, Fred 1999. Finnish: An essential grammar. London: Routledge.
Hungarian
Keresztes, László 1995. A practical Hungarian grammar. Debrecen: Debreceni Nyári
Egyetem.
Rounds, Carol 2001. Hungarian: an essential grammar. London/New York: Routledge.
Japanese
Koizumi, Tamotsu 2007. Nihongo no kaku to bunkei (Japanese Cases and Sentence Patterns).
Tokyo: Taishuukan.
Komi
Nagyezsda, Manova 1998. KomiZürjén nyelvkönyv. Szeged: JATE Finnugor Tanszék.
Hausenberg, AnuReet 1998. Komi. In: Abondolo(ed.): 305326
Nganasan
WagnerNagy, Beáta 2002. Chrestomathia Nganasanica. Szeged/Budapest: SZTE Finnugor
Tanszék/MTA Nyelvtudományi Intézet.
Northern Saami
Sammallahti, Pekka 1998. Saamic. In: Abondolo(ed.): 4395.
Northern Khanty
Nikolaeva, Irina 1999. Ostyak. München: Lincom Europa.
Udmurt
Winkler, Eberhard 2001. Udmurt. München: Lincom Europa.
Mansi
Riese, Timothy 2001. Vogul. München: Lincom Europa.
B. Intext references
Abondolo, Daniel (ed.). 1998. The Uralic languages. London/New York: Routledge.
Blake, Barry J. 1994. Case. Cambridge: Cambridge University Press.
Borin, Lars 1986. Is Hungarian a case language? FennoUgrica Suecana (Journal of
FinnoUgric Research in Sweden) 8: 133.
Dunn, Michael, Angela Terrill, Ger Reesink, Robert A. Foley, & Stephan Levinson. 2005.
Structural phylogenetics and the reconstruction of ancient language history. Science 309:
20722075.
Haspelmath Martin, Matthew S. Dryer, David Gil, and Bernard Comrie (eds.). 2005. The
World Atlas of Language Structures. Oxford: Oxford University Press.
Huson, Daniel H. and David Bryant 2005. Application of phylogenetic networks in
evolutionary studies. Molecular Biology and Evolution 23(2): 254267.
Iggesen, Oliver A. 2005. Number of cases. In: Haspelmath et al. (eds.): feature 49 in WALS:
202205
Ikeda, Testuro 2003. Imiseibun toshite no kaku kouzou (Case structures as sets of semantic
components). Uralica :5368.
Koizumi, Tamotsu 1994. Uralgo tougoron (Uralic syntax). Tokyo: Daigakushorin.
Nose, Masahiko 2006. Diversity of cases: Using The World Atlas of Language Structure.
Gengo Kenkyuu 130: 109123,
要旨
視覚化によるウラル語と日本語の格機能の対照
本論文では,フィン系,ウゴル系,サモエード系から構成されるウラル語 13 言語と日本語
を合わせた 14 言語をサンプルとして選び,各言語の格の数と名前,機能を調査した。言語間
で,格の名前と機能が必ずしも一致しないため,本論文では「普遍的に認定可能な格機能」
を 25 個設定した。その普遍的な格機能の記述に基づいて,視覚化を行った。格機能の視覚化
の結果,フィン系やウゴル系,サモエード系で隣接する言語がある一方,北サーミ語やハン
ガリー語のように格機能の上では語派から逸脱する言語が存在することが判明した。また,
ウラル語と同様に格を多く持つ日本語は,格機能の上ではマリ語と近い関係にあることが判
明した。