Linguistic Politeness at Different Age Levels

Author(s): Giovanna Axia and Maria Rosa Baroni

Source: Child Development , Aug., 1985, Vol. 56, No. 4 (Aug., 1985), pp. 918-927
Published by: Wiley on behalf of the Society for Research in Child Development

Stable URL: http://www.jstor.com/stable/1130104

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Wiley and Society for Research in Child Development are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Linguistic Politeness at Different Age Levels

Giovanna Axia and Maria Rosa Baroni

University of Padova

AXIA, GIOVANNA, and BARONI, MARIA ROSA. Linguistic Politeness at Different Age L
DEVELOPMENT, 1985, 56, 918-927. The general hypothesis is that to be as polite as t
requires, children must be able to evaluate the "cost" of a request according to the soci
In the first experiment it was hypothesized that although most studies show that child
and understand polite sentences at about age 6, they use polite requests in order to
obstacles in natural situations only at a later age level. Spontaneous requests of 58 subjec
and half female, divided into 3 age groups (5, 7, and 9 years), were recorded. The si
structured into 3 levels of difficulty of request accomplishment. The second experiment
with 64 subjects subdivided as before, aimed at examining their ability to attribute
according to polite register in a given scene. As expected, only 9-year-olds fully ma
register.

The object of our study is the genesis and dressee that the speaker recognizes the ad-
development of a particular communicative dressee's need for noninterference by others.
ability: linguistic politeness. Basing her work
Another aspect of the use of politeness in
on literature regarding speech acts and con-
requests comes from Clark and Schunk (1980,
versation (Austin, 1962; Searle, 1965), Lakoff
1981), according to whom politeness formulas
(1973) identified two general rules of prag-
operate within a cost-benefit system: the
matic competence. The first is "Be clear" and
higher the cost of the request, the larger the
corresponds to Grice's (1975) maxims. The
amount of compensatory politeness. A dis-
second, "Be polite," is specified by "Don't
tinction must also be made between po-
impose," "Give options," and "Make A feel
liteness and deference in linguistic interac-
good-be friendly" (Lakoff, 1973, p. 298).
tions (Fraser & Nolen, 1981). Politeness is a
These rules are used when the participants'
complex linguistic means used to maintain
main aim is not only communicating some-
good interactions with other people. Defer-
thing but also maintaining good interac- ence always implies reduction of speaker's
tion.
status and in some social situations is out of

Many studies have been carried out on place and therefore may be impolite.
politeness in requests. Requests are governed The use of polite register in requests has
by special conversational rules (Gordon & been studied widely in relation to the sex of
Lakoff, 1971; Labov & Fanshel, 1977) con- the participants and also with regard to inter-
stituting preconditions for their validity. locutor status (e.g., Crosby & Nyquist, 1977).
These rules involve the needs, abilities, In both men's and women's speech, the de-
rights, and obligations of the participants. In gree of politeness is a marker of social status
addition to the basic syntactic form (e.g., im- within a given interaction (Baroni & D'Urso.
perative), some mitigating expressions are1984).
also used (e.g., interrogatives), mainly in in-
teractions between persons of different status, Substantial agreement exists among de-
in order to avoid damaging personal rela-velopmental researchers that two factors af-
tions. fect the production and comprehension of po-
lite register. The first is knowledge of the
According to Brown and Levinson'slinguistic form of polite requests; the second
(1978) model, orders and requests are speech is knowledge of pragmatic request rules
acts that somehow threaten the addressee's within a given social and situational context
"face" in Goffman's (1967) sense, since they (Ervin-Tripp, 1977). When children achieve
tend to limit his or her freedom of action. A mastery of both these abilities, they become
"redressive" action may take the form of lin-capable of making situational inferences
guistic politeness intended to assure the ad- which lead in turn to mastery of highly indi-

Requests for reprints should be addressed to Giovanna Axia, Dipartimento di Psicologia dello
Sviluppo e della Socializzazione, Via B. Pellegrino 26, 35100 Padova, Italy.

[Child Development, 1985, 56, 918-927. ? 1985 by the Society for Research in Child Development, Inc.
All rights rcserved. 0009-3920/85/5604-0003$01.00]

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Axia and Baroni 919

rect requests such as "hints" (Mitcheli-

in which children produce requests must be
Kernan & Kernan, 1977). Bates (1976a),
controlledfor
as far as possible to create "natu-
example, identified three stages in ral"
thesettings,
devel- that is, spontaneous speech be-
tween persons
opment of children's linguistic politeness in and not with puppets or dolls.
requests. In the first stage, which It
endsis also
atpossible
the to introduce controlled vari-
ations. direct
age of about 4, children produce only
questions such as imperative phrases. In the
The main hypothesis of the present re-
second stage, children aged 5-6 can produce
search
all surface syntactic devices of their is that being polite when making re-
native
language, but are not able to "mask"queststhe is con-
a complex ability because it requires
a combination
tent of their requests in order to reach their of both linguistic and social
skills. To be
aims. In the third stage, 7- and 8-year-olds as polite as the situation requires,
can
children must master the form and content of
vary both the content and form of their re-
quests by producing very indirecttheir requests as well as comprehend their
requests.
Prinz (1982) further suggests thatlisteners' signals, recognize reciprocal status,
politeness
and so forth.
in requests is based either on linguistic as- Although developmental re-
pects or content modifications, according search indicates
to a that by about age 6 children
given degree of directness. Becker (1982) pro- are able to produce indirect requests, it is sug-
poses that the general social rules gested here that only at a later age can chil-
governing
communication are also very important, dren masterbe-polite requests in a natural situa-
cause they are the contextual criteria tion,judging
with social constraints. A time gap is
a request as appropriate (see also Ervin- expected between the ability to produce po-
Tripp, 1976). Every request is actually a social lite sentences and the capacity to use them to
act, and using politeness in making requests balance the social cost of the request. Most
is a relatively sophisticated means of control- studies in sociocognitive growth (Deutsch,
ling social assessment. By means of polite re- 1981; Flavell, Botkin, Fry, Wright, & Jarvis,
quests people keep their interactions alive 1968; Higgins, 1981) indicate that children
and also reach their aims. In this perspective, begin to know the social world adequately
some authors have linked the development ofwhen concrete operational thought is well as-
polite requests to the recognition of an impor- sessed, and that this phenomenon occurs at
tant aspect of social context, such as partici- about 8-9 years of age (Piaget & Inhelder,
pants' status (James, 1978; Mitchell-Kernan & 1955). According to Damon (1981), social
Kernan, 1977; Read & Cherry, 1978). knowledge is of two types: the construction of
a network of interpretative categories of social
In order to study the development of relations, and the ability to understand the
politeness in children's requests, therefore, at messages of the interaction that occurs.
least two factors must be taken into account.
First, theoretical considerations suggest relat- Our hypothesis is that children fully mas-
ing ability to produce indirect requests to de-ter the use of politeness when they are able to
velopment of general social skills such asrecognize the "cost" of a request with refer-
maintaining verbal interactions, overcoming ence not only to the status of both speaker and
obstacles in communication, and so forth. On addressee but also to the interlocutor's reac-
the other hand, methodological techniquestions to the request. The former capacity is
must be found allowing the phenomenon to based on processes through which social real-
be studied in natural social situations. Many ity is adequately classified, the latter on the
studies have indeed investigated requests understanding of the interaction that occurs.
produced by children in natural and free situ-Therefore, two experiments were carried out.
ations not manipulated by experimental con-In the first, sisointaneous requests uttered by
ditions (Bates, 1976a; Dore, 1977, 1978; children of differenit ages interacting with
Eisenberg & Garvey, 1981; Sachs & Devin,adults were examined. In the second, chil-
1976). Other studies deal with more struc-dren were presented with graphic representa-
tured situations, such as role-playing withtions of real scenes showing interactions be-
dolls or puppets (Bates, 1976b; Mitchell- tween children and adults. They then were
Kemrnan & Kemrnan, 1977; Read & Cherry, asked to attribute polite and impolite requests
1978). Both these methods have producedto the different characters. In a natural situa-
many interesting findings and useful sugges-tion children over ages 5-6 were expected to
tions for further research. Nevertheless, theyuse different degrees of politeness depending
on their interlocutor's reactions. Production
seem to be lacking in one point: neither
method considers spontaneous speech and(Experiment 1) and comprehension (Experi-
the introduction of experimental variables in ment 2) were expected to appear at the same
the same setting. The various social situationsage level.

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
920 Child Development

Experiment 1 by the experimenter to increase the level of

politeness of their requests, as Bates's (1976a)
Subjects
subjects were. In fact, children were pre-
After a pilot study with 20 subjects of dif-
sented only indirectly with the need to
ferent ages, three age groups were formed: (a)
change their previous requests and had to
18 children of mean age 5-1 (range 4-8 to 5-9); make social inferences about how best to do
(b) 20 children of mean age 7-1 (range 6-0 to so from contextual information.
7-10); (c) 20 children of mean age 9-3 (range
8-7 to 9-10). Half the subjects in each age Verbal expressions were recorded. We
group were male, the other half female. All considered three samples of request: the first
the children lived in the same area of Padua, a
sentence opening the "game" by the subject
large town in northern Italy. and the first requests after each of the two
Materials and Procedure resistances. As regards "deaf ear" resistances,
Materials were two plastic pegboards,we did not take into account first re-requests
if they were the exact repetition of the
and a bag containing pegs of four colors. The
subject was asked to sit at a table in a quietunfulfilled request, since we considered
these to be cases in which children wished to
room of his or her own school. The experi-
menter sat opposite. After a short conversa-ascertain whether they had been heard.
tion intended to put the child at ease, the ex- The order of resistances was counterbal-
perimenter introduced the materials andanced across subjects. The experimental de-
asked if the child wanted to use the pegs to sign included three comparisons among data:
make a figure on the pegboard. If the answerbetween first requests and requests after the
was affirmative, the experimenter said she first resistance (within subjects), among age-
would make a figure on her own pegboard at level groups for requests after "motivated re-
the same time, adding, "I'll keep the pegs--
fusals" (between subjects), and among age-
when you want one, ask me." The experi-level groups for requests after "deaf ear"
menter's status was thus higher than the resistances (between subjects).
child's because of both age and control of the
pegs. At this point, the experimenter began toResults
concentrate on her peg figure but was pre- Classifying requests according to a
pared to respond to the child's requests. politeness scale presents serious methodolog-
When the subject appeared to have becomeical problems. Ervin-Tripp (1977) proposes
the use of a general scale based on directness.
familiar with the situation and the spirit of the
game, two "resistances" were introduced.In our case this could not be used because
Their purpose was to signal to the child thatrequests were not free; rather, they were in-
his or her request could have been inter-duced experimentally by communicative ob-
preted as a threat to the interlocutor in somestacles. Also, we wished to evaluate not only
way (in the sense used by Brown & Levinson,requesting but also re-requesting, considering
1978, pp. 70-73). In the case of "deaf ear""the initial requests and re-requests as part of
resistances, the experimenter simply behaveda larger unit of discourse-the request se-
as if she had not heard the child's request. Ifquence" (Becker, 1982, p. 25). We therefore
the child made the request again as a way ofused a scale very similar to that proposed by
checking whether he or she had been heard Eisenberg and Garvey (1981). First, we
correctly, the experimenter continued the classified subjects' first requests into two cate-
same behavior. Further requests normallygories:
were satisfied.
1. Impolite: very direct requests, such as
In the second case, the experimenter didimperatives.
not satisfy the request but supplied some 2. Polite: less direct, mitigated requests,
motivation such as, "I can't give you a redsuch as interrogative or conditional sentences.
peg, because I'm using the red ones." The
motivation was rather arbitrary since the child Requests were rated by two independent
knew very well that there were plenty of judges who knew neither the sex nor the age
pegs. This kind of resistance did not lead to of the subjects. Agreement between judges
an evident break in social relations and was was 94.8%. Disagreements were resolved
called "motivated refusal." These two resis-through discussion.
tances are respectively similar to the "unac- Requests made after resistances were
counted" and "accounted" refusals of Labov
classified into five categories:
and Fanshel (1977, pp. 87-88). Each type of
feedback was aimed at eliciting a new re- 1. Silence: Subjects interrupted interac-
tion.
quest. Thus, subjects were not asked directly

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Axia and Baroni 921

2. Maintenance: Subjects either for- (category 3). In this case, a kind of "face-
mulated the request in the same way or rein- oriented end" (Brown & Levinson, 1978, p.
forced it by raising the voice or adding new 63) was satisfied, in the sense that potential
words. For example, "Rosso" (red). Resis- damage to the addressee's "face" (in this case
tance. "Uno rosso voglio" (I want a red one).' freedom of action) is limited.
3. Reduction of request: Subjects low- Expressions of deference and negotia-
ered their voices or only modified the content
tions were considered to be categories of lin-
of their requests, that is, asked for a peg of aguistic politeness. In contrast, maintenance
different color while maintaining the same was considered typically impolite. Silence
surface linguistic structure. For example, and reductions showed that subjects were
"Dammi un chiodino verde" (Give me a aware of some obstacle in the interaction but
green peg). Resistance. "Dammi un chiodino
were unable to cope with it by linguistic
blu" (Give me a blue peg). means. For this reason categories 1 and 3
4. Expressions of deference: Subjects were considered linguistic impoliteness cate-
gories.
modified the surface linguistic structure of
their previous requests but maintained the
Requests after resistances were rated by
same content. As Labov and Fanshel (1977)
two independent judges with the same proce-
point out, the identical repetition of a request
dure used for rating subjects' first requests.
may be interpreted as a challenge, but the
Agreement between judges for requests after
introduction of "mitigators" (p. 95) changes
"deaf ear" resistances was 84.3%; for requests
the form of the request. In our case, mitigators after motivated refusals it was 82.7%.
could be inserted in the form of courtesy for-
mulas: "I1 giallo" (The yellow one). Resis- Nonparametric statistical tests were used:
tance. "I1 giallo per piacere" (The yellow one, McNemar's Test for dependent samples and
please). Or mitigators could appear with the the chi-square test for independent samples.
introduction of an interrogative form, for ex-The rejection region for these analyses was a
ample, "Uno rosso" (A red one). Resistance. = .05. The first result was that the first re-
"Me ne dai uno rosso?" (Will you give me aquest almost always tended to be impolite in
red one?). Or they could include a condi- all age groups, but the occurrence of a resis-
tional, for example, "Uno verde" (a greentance significantly increases politeness for
one). Resistance. "Uno verde vorrei" (I would higher age levels (see Table 1). Five-year-
like a green one). olds almost never increased politeness level
5. Negotiations: Subjects produced re-after resistances; however, 7-year-olds did
(first request compared to requests after re-
quests that, although not necessarily deferen-
tial, took into account the interlocutor's be- sistances for two politeness categories:
havior and also assured her that her refusal toMcNemar Statistics = 7.11, p < .01). Nine-
satisfy their requests had not caused anyyear-olds also increased politeness (first re-
quest compared to requests after resistances
problems. This is the case of bargaining. The
children did not change the content of theirfor two politeness categories: McNemar Sta-
tistics = 7.36, p < .01). These data show that
previous requests but tried to maintain a good
interaction by producing very indirect re-it is only from the age of 7 that children can
use the polite register to overcome interaction
quests, such as: Resistance. "Qual'e quello
difficulties when the interlocutor raises the
che ti serve che non mi ricordo?" (Which one
do you want, I don't remember?). "Un altro cost of request accomplishment. In the fol-
blu" (Another blue one). Resistance. "Alloralowing tables the groups are slightly smaller
basta-un altro me ne serve e basta" (That'sthan those in Table 1, because six subjects
were not able to continue the interaction after
enough then-I just want one more, that's
the first resistance, showing reactions of si-
all). This category included requests in which
lence and embarrassment.
the children modified the content of previous
requests but also tried to reassure their ad- A second result illustrates ability in
dressee. For example, "Dammi un chiodinoreacting verbally to motivated refusals (see
verde" (Give me a green peg). Resistance. Table 2). There is a significant difference
"Ah, be', non importa, dammi un chiodinoamong age groups in the use of politeness af-
blu" (Oh well, it doesn't matter, give me ater motivated refusals (three age groups com-
blue one). This last kind of request is com-pared for two politeness categories: x2[2,52]
pletely different from "reduction" requests = 14.00, p < .001). Each age level was dis-

1 Children's answers were not always morphosyntactically correct in Italian, but the transla-
tions in parentheses do not necessarily reflect this.

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 922 Child Development

TABLE 1

DIFFERENCES IN USE OF POLITE REGISTER AT DIFFERENT AGE LEVELS IN

FIRST REQUESTS AND IN REQUESTS AFTER RESISTANCES

REQUESTS

AGE LEVELS First After Resistances

5 years old:
Polite subjects ..................... 2 3
Impolite subjects .................. 16 15
7 years old:
Polite subjects ....................... 4 13
Impolite subjects .................. 16 7
9 years old:
Polite subjects ....................... 5 16
Impolite subjects ................... 15 4

TABLE 2 did not increase their politeness, 7-y

DIFFERENCES IN USE OF POLITE REGISTER AT did (5-year-olds compared to 7-year-o
DIFFERENT AGE LEVELS AFTER THE Two two politeness categories: X2[1,31]
RESISTANCES with Yates corrections, p < .10).3 Nin
olds behaved like 7-year-olds.
RESISTANCES
In each age group, polite and im
Motivated Deaf requests after the two resistances we
AGE LEVELS Refusal Ear compared. In order to have two indep
samples, each age group was subdivid
5 years old: subjects who had received a "motiv
Polite subjects ......... 3 1 fusal" first and those who had received a
Impolite subjects ....... 12 11 "deaf ear" resistance first. The reduction of
7 years old:
the number of subjects made Fisher's exact
Polite subjects ......... 10 9
Impolite subjects .......
probability test preferable to the chi-square
8 10
9 years old: test. The rejection region for these analyses
Polite subjects .........
16 9 was a = .05. Results show that the 5-year-olds
Impolite subjects .......
3 11 hardly ever used polite register. Seven-year-
olds also did not show different reactions
when the experimenter's resistance changed,
but they used polite register in about 50% of
their requests.
tinct from the next one (5-year-olds comparedNine-year-olds used polite reg-
to 7-year-olds for two politeness categories:the type of resistance by the
ister according to
interlocutor
X2[1,33] = 4.31, p < .05; 7-year-olds indication (two resistances for
com-
two politeness
pared to 9-year-olds for two politeness categories: Fisher's exact prob-
catego-
ries: X2[1,37] = 3.61 with Yates ability test: p < .05).
corrections, p
< .10).2 In general, there were hardly any po-
The data were subjected to post-hoc anal-
lite requests in the 5-year-old group; they be-
ysis to see if the types of polite (category 5 vs.
gan to appear in the 7-year-old group and be-
6) and impolite (category 2 vs. 3) request
came predominant in the 9-year-old group.
varied with the different types of resistance
There was also a significant("motivated
differencerefusal" and "deaf ear") (see
Tablenumber
among the three age groups in the 3). The two
ofpoliteness categories ("ex-
pressions
polite requests after "deaf ear" of deference" and "negotiations")
resistance
(three age groups compared for appeared
twoto be used in different ways accord-
po-
liteness categories: X2[1,47] =ing to type
6.22, p <of.05)
resistance. Nine-year-olds
(see Table 3). Here too, althoughtended to use "expressions of deference" if no
5-year-olds

2 In order to evaluate this result with greater precision (see Lehman, 1970, p. 62), the critical
level corresponding to the value obtained by statistics was calculated: &(X2 = 3.61) = 0.05778 >
0.050 = oa. The difference between ao and & is clearly very small.
3 In this case, too, the critical level was calculated: &(X2 = 3.49) = 0.06149 > 0.050 = a.

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Axia and Baroni 923

TABLE 3

PERCENTAGE OF RE-REQUESTING FALLING INTO FIVE CATEGORIES OF RESPONSE

CATEGORIES

AGE LEVELS Silence Maintenance Reduction Deference Negotiation

5 years old:
Motivated 6.6 6.6 66.6 6.6 13.3
refusal ....... (1) (1) (10) (1) (2)
Deaf ear ....... 8.3 75.0 8.3 8.3
(1) (9) (1) (1)
7 years old:
Motivated ... ... 44.4 11.1 44.4
refusal ....... (8) (2) (8)
Deaf ear ....... ... 26.3 26.3 31.6 15.8
(5) (5) (6) (3)
9 years old:
Motivated ... 26.3 5.7 68.4
refusal ....... (5) (1) (13)
Deaf ear....... ... 20.0 35.0 35.0 10.0
(4) (7) (7) (2)

NOTE.-Numbers in parenth

answer motivated
was givenrefusal, thus showing
to their ability
th
"negotiations" when
to react to their interlocutor's behavior,fa
which
refusal ("expressions
signals an increase in the cost of request ac-
"negotiations"
complishment. Thefor "deaf ear" response
tw is a
exact probability
clumsy, ostentatious way of nottest
satisfying a
strongly, request;
7-year-olds
younger children are incapable of
same way. coping
Yet, with this difficulty
5-year- and merely ask
polite requests
again with more or lessbut
determination. te
Seven-
same request and 9-year-olds
afterchoose to increase the defer-
"dea
modifying ence it of their language. In contrast,
after mot "mo-
nance" andtivated refusal" requires children to use a
"reduction"
sistances: much more sophisticated reply,
Fisher's and, in fact,
exact
.025). after motivated refusal, "negotiations" pre-
Discussion
dominate among polite requests. However, it
is only at about 9 years of age that children
The most important result of this experi-
show complete mastery of the situation by in-
ment shows that children master polite regis-
creasing the politeness level of their requests.
ter to overcome difficulties in natural interac-
tion at about the age of 9. It is only at this age
These data may be summarized as fol-
that they can grade linguistic politeness ac-
lows:
cording to type of resistance to satisfy an in-
terlocutor. In contrast, 7-year-olds always in-1. Impoliteness level: characterized by
crease their politeness whatever the impolite
cost requests, with qualitative distinction
signaled by the interlocutor, either when among
the different resistances (5-year-olds).
interlocutor pretends he or she has not heard
2. First politeness level: characterized by
or when the interlocutor gives a motivated re-requests, with qualitative but not quan-
polite
fusal. Five-year-olds hardly ever use titative
polite distinction between the two resis-
register, even after resistances. tances (7-year-olds).
Second, we observe that requests ad-
3. Second politeness level: characterized
dressed to an interlocutor "turning a deaf ear"
by polite requests, with qualitative as well as
as well as to one expressing refusal verbally
quantitative distinction between the two re-
and motivating the refusal are qualitatively
sistances (9-year-olds).
different for all age groups. In fact, although
in an extremely impolite way, even younger Although the ability to distinguish
children tend to modify either the contentamongor the interlocutor's reactions occurs
the tone of their request when faced with rathera early, reacting politely to an inter-

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
924 Child Development

ble-we did not want to vary other factors

locutor signaling the cost of request accom-
plishment comes later. The younger child such
has as social distance and familiarity by us-
an inventory of polite sentences as well asing
thean over-polite expression. Nevertheless,
ability to distinguish interaction difficulties
in an Italian speaker's mind, no doubt could
arise regarding the polarization of the two
but cannot combine the two abilities suitably.
The use of linguistic politeness implies sentences.
not Sentences such as those used here
only an evaluation of the feedback by whichwith different politeness values had to be
chosen, partly because we could not com-
the interlocutor shows the cost of a request
but also sufficient capacity to categorize pletely
the control another important politeness
two statuses in the interaction. variable, that is, intonation.
It could be hypothesized that children Immediately following the request, the
master politeness when they clearly under- experimenter, pointing to the character who
stand the social setting of the interaction,
had spoken, asked: "Who is he speaking to?"
even though they are not aware of it. Only After the subject had answered, the experi-
within this context are they able to match menter asked why the child had given that
answer by saying: "What makes you think
their interlocutor's reaction with their polite
that?"
inventory. In order to test this hypothesis, we
prepared a second experiment, this time deal-
Every subject was presented with one
ing with the understanding of linguistic
character uttering one request. The four ex-
politeness as a signal of hierarchy in a given
interaction. In this case the situation was not perimental conditions included adult impo-
lite, adult polite, child impolite, child polite.
natural as in Experiment 1. Social interactions
Each subject was asked to indicate the ad-
were represented through pictures, and sub-
dressee (adult or child) in one of the above
jects were asked to identify the addressee of a situations.
request among the characters illustrated.
Results
Experiment 2 Because the aim of this experiment was
Method to check children's comprehension of polite
versus impolite register as influenced by ad-
Subjects.-Subjects included 64 chil-
dressee's status, results were ordered accord-
dren, half male, half female, divided into
ing to this variable only. Fisher's exact proba-
three groups: (a) 25 subjects of mean age 5-4
bility test was used. The rejection region for
(range 4-4 to 5-10); (b) 20 of mean age 7 (range
these analyses was at = .05.
6-2 to 8-0); (c) 19 of mean age 9-6 (range 9-0 to
10-0). No difference was observed in the at-

Material.-The materials included one tribution of polite and impolite requests at the
age levels of 5 and 7. The first significant dif-
picture showing an adult and a child sitting
ference occurred at the age of 9 (see Table 4).
on a two-place sofa in front of a television.
At this age the polite sentence tended to be
One of two other pictures showing an adult
spoken to adults and the impolite one to chil-
and a child, respectively, could be inserted
into the principal scene. All characters were
male.
TABLE 4
Procedure.-Subjects were presented DIFFERENCES IN ATTRIBUTION OF POLITE AND
with one scene and asked to say what it repre- IMPOLITE REQUESTS ACCORDING TO SPEAKER'S
sented. This was done in order to make sure AND ADDRESSEE'S STATUSES AT DIFFERENT AGE
that they had correctly understood the roles of LEVELS IN EXPERIMENT 2
the participants (adult-child) and the purpose
(watching TV). At this point one of the two ADDRESSEE
other characters (adult or child) was inserted
AGE LEVELS Adult Child
and, through the experimenter's voice, spoke
one of the following sentences: "Spostati!"
5 years old:
(Move over!) or "Vorresti spostarti, per fa- Polite request .......... 8 4
vore?" (Would you move over, please?) Impolite request ....... 7 6
Although these two sentences are evi-7 years old:
dently polarized, we checked their position Polite request .......... 6 4
Impolite request ....... 4 6
on an 18-degree politeness scale with an adult
9 years old:
Italian sample (Baroni & Axia, 1984). The Polite request .......... 8 2
chosen polite sentence falls at a high level of Impolite request ....... 3 7
politeness but is not the most polite possi-

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Axia and Baroni 925

TABLE 5

PERCENTAGE OF ANSWERS FALLING INTO FOUR CATEGORIES OF MOTIVATION AT

DIFFERENT AGE LEVELS IN EXPERIMENT 2

CATEGORIES

No Answer General
Egocentric Incidental Social
AGE LEVELS Motivation Aspects Rules Register

5 years old ...... 36 44 20

(9) (11) (5)
7 years old ...... 20 60 20
(4) (12) (4)
9 years old ...... ... 78.9 10.5 10.5
(15) (2) (2)

NOTE.-Numbers in parentheses

dren (polite requests compared

contains only two examples, both referring to
quests for two
"Spostati!" addressee's
(Move over!): "Perch6 non si pub s
exact probability
chiederlo ai genitori,
test, perch6 i genitori
p sono < .05
plicit motivations for
grandi e si deve portare rispetto" their
(Because
among you can't ask parents, because
the following parents are
categori
grown-ups, and you must be respectful);
0. No answer: Children gave
"Perch6 sono tutti e due bambini. Non poteva
tion, or only answers of the
dirlo al papa, non era a casa sua" (Because
don't know).
they are both children. He couldn't say that to
1. Egocentric motivations: Children gave the daddy, he was not in his own house).
motivations referring to themselves, for ex-
No statistical difference was found
ample, "L'ho capito perch6 ci ho pensato" (I
understood because I thought about it); "L'ho among the three age levels. In general, chil-
capito da sola" (I understood it by myself). dren of all age levels tended to give category
2 motivations. Egocentric motivations de-
2. Incidental aspects: Children referred creased from 5 to 7 to 9 years, and motivations
to incidental aspects of the interaction, for ex- referring to linguistic markers of the ad-
ample, "Perch6 il papa era troppo vicino" dressee's status began to appear at age 9 (see
(Because his daddy was too close) or made Table 5). Because the task required good
inferences which were not justified im- awareness of the metalinguistic rules of re-
mediately, for example, "Dice al papa di spos- questing, it was not surprising to find a very
tarsi perch6 deve parlargli al bambino" (He small number of motivations that took request
tells his daddy to move over because he must register into account.
speak to the child).
3. General social rules: Children referred Conclusion
to some general social rules without any refer- The aim of this research was not to estab-
ence to request content, the most common be-
lish at what age children produce and under-
ing the following: children generally speak
stand indirect requests, but to start an investi-
with children, adults with adults. Some exam-
gation of the conditions under which they are
ples include, "Lo dice al bambino perch6
able to resolve communicative problems us-
amico del bambino" (He says that to the child
ing language. The accentuation of the interac-
because he is a friend of his); "Lui un bam-
tive aspect of requests in Experiment 1
bino e si rivolge a un bambino della mia etk"
showed that the capacity to react to the cost of
(He is a child and is speaking to a child of my
the request in relation to the interlocutor's be-
age); and "Perche questo e un signore e
havior is quite an early acquisition. On the
quello e un signore, e allora i signori parlano
other hand, the capacity to maintain good in-
con i signori" (Because this one is a grown-up
teractions by increasing the politeness of the
and that one is a grown-up, and grown-ups
request occurs at a later age. The results of
speak with grown-ups).
Experiment 2 reveal one aspect of the devel-
4. Register: Children gave motivations opment of the capacity to formulate requests
referring to request register. This category using polite register. Only from the age of 9

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
926 Child Development
Brown, P., & Levinson, S. (1978). Universals in lan-
do children use linguistic politeness as a cri-
terion in judging the appropriateness of a re-guage usage: Politeness phenomena. In E. N.
quest according to the addressee's status. Goody (Ed.), Questions and politeness (pp. 56-
324). Cambridge, MA: Harvard University
In conclusion, handling the politeness ofPress.
requests in context is a complex ability going
Clark, H. H., & Schunk, D. H. (1980). Polite re-
beyond the simple "deferent" formulation of
sponses to polite requests. Cognition, 3, 111-
requests. In daily situations, requests are of-143.
ten reformulated since various types of obsta-
Clark, H. H., & Schunk, D. H. (1981). Politeness in
cles intervene in the interaction. In other
requests: A rejoinder to Kemper and Thissen.
words, there are various extralinguistic factors
Cognition, 9, 311-315.
that modify the way in which requests are
Crosby, F., & Nyquist, L. (1977). The female regis-
formulated or reformulated. In this research
ter: An empirical study of Lakoff's hypotheses.
we studied the development of the capacity to
Language in Society, 6, 313-322.
take two factors into account: (a) cost of re-
Damon, W. (1981). Exploring children's social cog-
quest accomplishment as signaled by the in- nition on two fronts. In J. H. Flavell & L. Ross
terlocutor; (b) appropriateness of register ac- (Eds.), Social cognitive development (pp. 154-
cording to addressee's status. Many other 175). Cambridge, MA: Harvard University
factors are being studied for further infor- Press.
mation on the capacity to use linguistic Deutsch, F. (1981). Cognitive processes and social
politeness in natural settings. For example, cognition in kindergarten children. Journal of
interlocutor's status versus that of the speaker
Genetic Psychology, 1, 63-73.
(subsets of this factor are age, sex, role, and
Dore, J. (1977). "Oh them sheriff": A pragmatic
other situational variables); cost of request
analysis of children's responses to questions.
accomplishment, which must be counterbal- In S. Ervin-Tripp & C. Mitchell-Kernan (Eds.),
anced by the degree of politeness; aim of Child discourse (pp. 139-164). New York: Aca-
maintaining the interaction, that is, the gen- demic Press.
eral aim of politeness rules; linguistic compe-
Dore, J. (1978). Requestive system in the nursery
tence, that is, possession of an inventory school conversations: Analysis of talk in its so-
of indirect request formulas; and social dis- cial context. In R. N. Campbell & P. T. Smith
tance, that is, degree of acquaintance, famil- (Eds.), Recent advances in the psychology of
iarity, and so forth. Our results suggest tying
language (Vol. 4A, pp. 271-292). New York:
study of the development of polite discourse Plenum.
with that of knowledge of social realities.
Eisenberg, A. R., & Garvey, C. (1981). Children's
use of verbal strategies in resolving conflicts.
References
Discourse Processes, 4, 149-170.
Ervin-Tripp, S. (1976). Is Sybil there? The structure
Austin, J. L. (1962). How to do things with words.
Oxford: Oxford University Press. of some American English directives. Lan-
Baroni, M. R., & Axia, G. (1984). Linguistic guage in Society, 5, 25-66.
politeness: A psychological and social model. Ervin-Tripp, S. (1977). Wait for me, roller skate. In
Unpublished manuscript. (Available from S. Ervin-Tripp & C. Mitchell-Kernan (Eds.),
M. R. Baroni, Istituto di Psicologia, UniversitA Child discourse (pp. 165-188). New York: Aca-
di Padova, Piazza Capitaniato 3, 35100 Padova, demic Press.
Italy) Flavell, J. H., Botkin, P. I., Fry, C. L., Wright, J. W.,
Baroni, M. R., & D'Urso, V. (1984). Some experi- & Jarvis, P. E. (1968). The development of role-
mental findings about the question of po- taking and communication skills in children.
liteness and women's speech. Language in So- New York: Wiley.
ciety, 13, 67-72. Fraser, B., & Nolen, W. (1981). The association of
Bates, E. (1976a). Acquisition of polite forms: Lon- deference with linguistic form. International
gitudinal evidence. In E. Bates, Language and Journal of Sociology of Language, 27, 93-109.
context: The acquisition of pragmatics (pp. Goffman, E. (1967). Interaction ritual: Essays on
255-294). New York: Academic Press. face to face behavior. Garden City, NY: Dou-
Bates, E. (1976b). Acquisition of polite forms: Ex- bleday.
perimental evidence. In E. Bates, Language Gordon, D., & Lakoff, G. (1971). Conversational
and context: The acquisition of pragmatics postulates. In Papers of the Chicago Linguistic
(pp. 295-326). New York: Academic Press. Society (pp. 63-84). Chicago: University of
Becker, J. A. (1982). Children's strategic use of re- Chicago, Department of Linguistics.
quests to mark and manipulate social status. In Grice, H. P. (1975). Logic and conversation. The
S. A. Kuczaj II (Ed.), Language development William James Lectures at Harvard University,
(Vol. 2, pp. 1-35). Hillsdale, NJ: Erlbaum. Lesson II, 1967. In P. Cole & J. L. Morgan

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms
 Axia and Baroni 927

(Eds.), Syntax and semantics-Speech acts

Mitchell-Kernan, C., & Kernan, K. (1977). Pragmat-
(pp. 41-58). New York: Academic Press. ics of directive choice among children. In S.
Higgins, E. T. (1981). Role-taking and socialErvin-Tripp
judg- & C. Mitchell-Kernan (Eds.),
ment: Alternative developmental perspectives
Child discourse (pp. 189-210). New York: Aca-
and processes. In J. H. Flavell & L. Ross
demic Press.
(Eds.), Social cognitive development (pp. 119-
Piaget, J., & Inhelder, B. (1955). De la logique de
153). Cambridge, MA: Harvard University
l'enfant a la logique de l'adolescent. Paris:
Press. Presses Universitaires de France.
James, S. L. (1978). Effect of listeners' age and situ- Prinz, P. M. (1982). "Requesting" in normal and
ation on the politeness of children's directive. language disordered children. In K. E. Nelson
Journal of Psycholinguistic Research, 4, 307- (Ed.), Children's language (Vol. 3, pp. 139-
317. 203). Hillsdale, NJ: Erlbaum.
Labov, W., & Fanshel, D. (1977). Therapeutic dis- Read, B., & Cherry, L. J. (1978). Preschool chil-
course. London: Academic Press. dren's production of directive forms. Discourse
Lakoff, R. (1973). The logic of politeness: Or mind- Processes, 1, 233-245.
ing your P's and Q's. In C. Corum, T. C. Smith-Sachs, J., & Devin, J. (1976). Young children's use
Stark, & A. Weiser (Eds.), Papers from the of appropriate speech styles in social interac-
Ninth Regional Meeting of the Chicago Lin- tion and role-playing. Journal of Child Lan-
guistic Society (pp. 292-305). Chicago: Uni- guage, 3, 81-98.
versity of Chicago Press. Searle, J. R. (1965). What is a speech act? In M.
Lehman, E. L. (1970). Testing statistical hypoth- Black (Ed.), Philosophy in America (pp. 221-
eses. New York: Wiley. 234). London: Allen & Unwin.

This content downloaded from

202.43.95.117 on Tue, 08 Sep 2020 15:00:21 UTC
All use subject to https://about.jstor.org/terms