DOCUMENT RESUME

ED 060 741

AUTHOR
TITLE
INSTITUTION

PUB DATE
NOTE

EDRS PRICE
DESCRIPTORS

IDENTIFIERS

FL 003 020
Christian, Jane M.
Style and Dialect Selection in Hindi-BhoThuri
Learning Children.
Laval Univ., Quebec. International Center on
Bilingualism.
Nov 71

14o.; In "Conference on Child Language," preprints

papers presented at the Conference, Chicago,
Illinois, November 22-24, 1971, p65-78
MF-$0.65 HC-$3.29
Bilingualism; Child Language; *Children; Cultural
Differences; *Dialects; Early Childhood; Hindi;
*Language Development; Language Research; Language
Role; Language Styles; *Language Usage; Linguistic
Competence; Linguistic Theory; Psycholinguistics;
Social Dialects; *Sociolinguistics; Speech; Urdu
*India

ABSTRACT
In India, the use of language dialect and style, likE
many aspects of Indian thought and life, follows a continuum from the
ritually pure and worthy of respect to the ritually defiled and
unworthy. In North India, according to adult informants, Hindi is
spoken at school, in formal business contacts or government offices,
in formal ceremonies; it is the written language. Bhojpuri is the
language spoken at home and in more informal relationships. Dialect
in India is defined by attitudes, vested interests, and cognitive
assumptions as to the nature of ritual, social and linguistic
context, and kinesic and paralinguistic markers. Studying a child's
development in the recognition and use of the language styles and
dialects indicates some of the learning processes that are involved.
It is possible to see various stages in the language development of
children as they learn the proper usage of the styles and dialects.
(ITYI)

U.S. DEPARTMENT OF HEALTH. EDUCATION

& WELFARE
OFFICE OF EDUCATION
THIS DOCUMENT HAS BEEN REPRODUCED
EXACTLY AS RECEIVED FROM THE PERSON (.R

ORGANIZATION ORIGINATING IT POINTS OF

VIEW OR OPINIONS STATED DO NOT NECES
SAF,ILY REPRESENT OFFICIAL OFFICE OF EDU
CATION POSITION OR POLICY

STYLE AND DIALECT SELECTION

IN HINDI-BHOJPUPI LEARNING
CHILDREN

Jane M. Christian
Universit

of Alabama

Birmin ham

In researching child language and especially child

bilingualism some of the problems met with in linguistic
study of adults are both redoubled, and made more obvious
and inescapable. We are forced for one thing away fro=
the comfortable notion of language as state and to a view
of language as process in the developing child. We thus
may incline towards the explanatory power of mentalistic
models in current psycholinguistics, towardscriteria of
testability and psychological reality, and towards more
emphasis on semantic analysis. Another problem of studying such bilingualism is that of terms definition. Are
we justified in simply transferring concepts inhering in
and surrounding terms such as dialect or style from the
adult Western or European framework a) to children or
b) to a very different social and linguistic complex as
India presents? Should we perhaps re-define the semantic
distinctive features for these terms to increase their
discriminatory and explanatory powers in accordance with
what we find in the field? And can we safely ignore the
importance of paralinguistic, kinesic and other contextual
features in the study of child language acquisition, particularly bilingual? These questions are here illustrated
in the context of dialect-style learning by children of
Banaras in North India.

Necessarily, in order to work at all on discovery procedures

in child language acquisition and in bilingualism, and to integrate
our findings into a more comprehensive field of enquiry, we utilize
and adapt systematic assumptions from the general discipline of
linguistic research with adults and monolinguals and the resultant
theory, both structural and generative. This being a much larger,
better-worked and established field than the subdisciplines concentrating on language learning or bilingual processes, it is only
reasonable that we should expect to bring to bear its current
concepts and methods when working with bilinguals and with children,
to ask questions related to current general questions of theory
and method, and then to relate these matters to the field at large.
However, this has its drawbacks as well as its advantages.
Some
methods and terminology, by the very fact of this extension of

the field of research, may well need to be reconsidered and their

meanings reshaped and broadeiled to fit the larger context. Basic
orientation may also require adaptation to different language and
culture systems, as tneir defining features and the domains based
on these may segment and structure 'reality' in ways unexpected by
the researcher. This again may serve to indicate needs to broaden
theory; and this is, after all, the history and process of any
growing discipline. However, though we are coming to expect this
broad kind of adaptability in bilinguals and in young children, we
as system-oriented scholars may sometimes be less ready or able to
adapt our own thought and research behavior sufficiently to other
systems.

In the case of children and bilinguals, and more especially

in the combined case of bilingualism in young children, we are

probably witnessing an extreme of adaptability. If an understanding
of the range and parameters of possible communication strategies
and behaviors is central to the study of linguistics, then the
description and creation of explanatory models for bilingualism in
young children is important indeed for the development of linguistic
theory.
Here we may observe linguistic manipulation and creativity
concentrated, and here models of competence may well be made and
In fact, any powerful and general model of linguistic
tested.

competence must be able to take into account the varieties of

bilingualism and developing language use in children, and of
Bending our
these combined, as perhaps a sort of ultimate test.
energies towards fuller understanding of these phenomena may be
diffi.7:ult indeed, a task with so many dimensions as to be a trap
for the unwary, but one with the ultimate possibility of high
rewards.

In the present case of style-dialect selection,by Hindi-Bhojpuri

speaking children in and near the ancient holy city of Banaras in
North India,there are complexities which beggar the terminology
which this writer had previously learned to use with regard to
dialects, styles and related phenomena, and their analysis-1as well
as the whole question of what we may define as bilingualism in the
adult scheme and especially in the schemes of child learners.
To reap some degree of understanding out of much inEtial confusion
required two things: considerable time spent thoroughly immersed
in the situation, with close attention to what adults and children
actually were saying and doing in a wide variety of contexts; and
,inadeqsIte
a casting aside of numeruusl
econceptions. The object
specifically,then, wasOto define the operative distinctive features
for adults and for children in styles or dialects, and in the contexts for which these were selected, 2) to note what and where were
the markers of communicative behavior, and where their parameters,
and, mos(.. difficult, 3) tc come to decisions as to their basic
meanings or psychological reality,1 and 4) to group and categorize
these behaviors at a higher level of broad cultural meanings and
social functions.
.

The need to view developing child language as process rather

than state seems clear enough. In North India this view of language
is further underlined by the factor of rapid lingcistic change,
1

The question always remains open and theoretically unprovable as

to whether psychological reality has really been captured, and is
further vexed by the question of whose reality and when, and how
many psychological realities may coexist in a social group sharing
a culture and language, and how much these need to and do overlap.
One can be surer by observing and checking carefully with informants
what solutions do not represent psychological reality, or the semantic
set, and at least markedly narrow the possibilities.

change which for centuries has added in complex ways to--at any
time--an already complex situation. Speech differences traditionally
tend to demarcate the enormous variety of crosscut social-and religious
groupings, and emphasize other distinctions made among them. There
is ritual power and bargaining power in lahguage choices. On the one
hand language is conceived as having a divine nature and power, some
types having more mana than others; on the other hand individuals
and groups define and can raise their social status in specific
ways by making stylistic changes in their communicative behavior,
provided, of course, this is done by small increments and discreetly.
Kali C. Bahl of the University of Chicago makes some pertinent
commentf_ along these sociolinguistic lines in a review of M. Jordan-

Horstmann's 5adani: A Bhoipuri Dialect Spoken in Chotanagar (1969)

in American Anthropologist 73:4:909-10 (August, 1971). It is noted
by Bahl that the author fails to mention anywhere that "wholesale
language-switching has been going on in this area 2 for quite some
time....Several 5adani speaking communities are in the process of
switching over to modern Hindi." Further, "language-switching...
from Sadani to modern Hindi ...serves to signify sociocultural
progress in this area where a particular language or dialect identifies the social status of an individual oi a group in relation to
other individuals or groups." The important commer.t is made that,
"The problems of correlation between language and dialect grouping
along the lines of social stratification canshe fruitfully studied
in North India." The additional comment might be made that definiof languages and dialects in North India is presently, and
understandably, in a somewhat chaotic state.
In this land of overwhelming linguistic diversity and fourteen
official state languages, an enormous amount of writing and verbal
tion

exposition continues to deal with the subject Hindt, but it must

be said that few issues have been settled. Throughout, there is
little agreement abc.ut how many speakers of Hindi there are, who
actually speaks 'true' Hindi, how well and to whom, how much and
what sorts of bilingualism and multilingualism exist, what dialects
are dominant in what ways, just what the Hindi or Hindustani
2

the Ranchi District of Bihar

language consists of, and whether or not scores of dialects and

sub-dialects are part of the Hindi language. Out of this of course
rises the question as to just what is a dialect and how it is to
be operationally defined. It would appear that to some extent each
has been empirically and separately defined on the basis of varying
criteria by people with varying qualifications to evaluate them.
Especial2y has controversy continued as to the relative status of
Hindi and Urdu, for political, communal, religious and regional
reasons more than narrowly linguistic ones.
Extreme separatists in Banaras and elsewhere argue Hindi and

Urdu are two distinct languages, and point for conclusive proof
to their different scripts--devnaorI for Hindi stemming very closely
from Sanskrit, and Persian for Urdu. Ordinary Muslims of course
speak Urdu in Banaras; their Hindu neighbors speak Hindi or Bhojpuri,

they say; aside from a few differences in formal greetings and

prayer formulae, a linguist would be hard put indeed to detect
any difference at all when they converse with each other or among
themselves, in terms of phonolooy and grammar.
It is true that
tnere are some small differences in kinesic and paralinguistic features, and differences in dress, etc., some of which can be consciously exaggerated or poinied out if need be. There is of course some
larger difference in lifestyle: in other words the differences
are primarily social rather than strictly verbal, but it is not
always easy to see where language fades into other aspects of
culture through the communicative devices of such items as gesture
and dress.
But it is interesting and informative to compare the Hindi-Urdu
stylistic differences given in a standard text with actual usage
in everyday speech in this holy city of the Hindus.. By far the
greatest number of stylistic lexical alternateslisted in the text
as Urdu were those in ordinary use among both Muslims and Hindus.
My informants, both Hindu adults and children over eight who were
able to select and identify dialects or styles by name, contended
3
3
these were by no means Urdu, but ordinary Ninth...lit
should be noted
that estegan, motar, pensil, pen, rill, kMp

Ccopybook), sa.kil, redivn,

levt and tavm, connected with the new mechanization and literacy in
North India, were regarded and inflected as Hindi tooa

Many of the words listed as Hindi variants were rejected as either

not known or considered bookish. Some were commonly contrasted with
a lexeme from the Urdu list, but the difference given was that of
respect-religious form versus ordinary. A few typical examples of
the latter are:
orih

god house

ohar

ordinary house
H

pustak religious book

kitMb

ygtr5

pilgrimage

safar trip

guddt-

pure ritually

sMf

sMhMytE divine aid

sth5n sacred place

book

clean

madad ordinary help

laoah place

(e.g., ttrth-kM sth.5n)

ritually purifying pahMn bath

bath (e.a., oanoa snZn)
This question of what is Hindi or Urdu is matched and overlapped
sn-gn

by the question of what is Hindi or Bhojpuri, according either to

adults or children. Bhojpuri is what is spoken at home, say all
informants old enough to be aware of named sorts of speech. Then
they add Bhojpuri is the medium of ordinary bMzMr contacts, contacts
with consanguineal kin, with close friends, with women and children.
One also prays a'id sings for the gods in Bhojpuri, alone at one's
peig or in company at a hhaiAll or Mrtht.

Bhojpuri can also be partly

grammatically defined by children of eleven in that they can deliberately speak in Bhojpuri and contrast this with Hindi speech, and
can give paradigmatic structure of Bhojpuri verbal inflections, etc.
Hindi is said to be that which is spoken at school, in formal

business contacts or government offices, in formal ceremonies either

public or private, in some contacts with affinal kin; and Hindi is
what is written. One uses Hindi if possible to indicate respect given
to another, and one raises the respect to be accorded to himself by
his proficiency in spoken and written Hindi. Religious books are
written in guddh Hindi, a designation given a more formal, Sanskritized, and ritually pure form of the language; religious discourses,
dramas, and some ceremonies are conducted in guddh Hindi. Virtually

every child over eight is aware of this style, and an increasing

number of boys over this age become more or less proficient in its
production as well as comprehension.

Nearly every pandit, pUiEri(priest), or vy-6s (learned commentator

on various scriptures), knows guddh Hindi well and can expound sonorously and dramatically, quoting at times from Sanskrit, for hours.
He is unlikely to use guddh Hindi in his ordinary speech. Very
many serious minded men, whether of dvi4a or twice born varna or
not, know considerable guddh Hindi.
This dimension or continuum with regard to respect or Mdar in

speech is commonly labeled in terms of high, ordinary, and law, or

nirMdar, thaugh finer distinctions can be made if it is considered
necessary in special situations. This is a measure of distinctions

both iinguistic and social, which transcends and complicates very

much of what we are accustomed to think of in terms of dialect or
Here these categories are inadequate to
style throughout India.
describe or explain the interrelationships in a country where some
languages may be ritually high and others low, where paradoxically
thetha can mean both pure and unmixed, and the unwritten language
.
of common people. Hardly any aspect of Indian thought or life remains untouched by this contimuum of the ritually pure and worthy
of respect, to the ritually defiled and unworthy--persons, groups,
objects, ideas and even languages or dialects not remaining constant but rather sliding along the scale according to a multiplicity of factors, and a complex etiquette.
And all this points to the problem of how speech behavior is
conceived and defined by the speakers. Factors such as attitudes,

vested interests, and cognitive assumptions as to the nature of

ritual, social and linguistic context clearly can effect how utterances are produced, received, interpreted and understood. On the
basis of these factors plus kinesic and paralinguistic markers we
can thus sometimes distinguish a 'dialect' in India. Linguistic
distance is generally measured according to social and ritual distance.
For example, a child of eight or more, or an adult, would
quickly and positively state what dialect or style another person
was using or would use, even on the basis of photographs or the
mention of certain categories of persons, where both verbal and
paralinguistic-kinesic features were largely ruled out. It was a
matter of who ought to be speaking what to whom, a matter of established expectations. An informant's more considered decision would
be based not necessarily on listening but on further knowledge of

such factors as agelsex, dress, residence, j-ati, education,

occupation, plus the speaker's relationship to the person spoken

to, his current ritual status, and where the speech act took place:
home, neighborhood, baznr, mandir (temple), school, office, etc.

Even where listening was clearly possible, as in overhearing street

conversations, listening for grammatical constructions or lexical
items proved secondary to the social-ritual considerations, for
which largely visual evidence or non-verbal information stimulated
cognitive classification.
This is not to say that either children or adults were unable

readily to specify styles through listening alone. They classified

easily on the basis of hearing taped samples of speech of individuals
unknown to them, though here I could find no way to separate cues
derived from the semantic content of the taped speech from purely
dialectical or stylistic differences.

Before the age of three years

children could easily recognize their own taped speech and that of
family members, could recognize speech directed to babies by its
style, and usually could pick out the guddh Hindi stylebylabeling
the speaker bMbri-1, a cover term for any sort of holy man, often

used by children as well as adults. Between three and five they

became proficient in picking out Bhojpuri neighborhood-b"gz-gr type

conversations, in which they were already participating daily, and

could differentiate by respect style markers speech of children
and adults to individuals of higher status, outside or within the
family: Also they could recognize the simplified style of an adult
speaking to a young child in simple short sentences with a restricted
set of lexical items and lack of respect forms. Generally they were
familiar too with curt or even abusive language style, recognizing
it as low, bad oMll Or burM-bali. They identified standard Hindi
with the radio broadcasts generally, as most of these Bhojpuri
learning children had little or no contact with standard Hindi
speakers before going to school. School attending children of six
or so identified standard Hindi with school and textbooks, though
their teachers admitted rather unwillingly that most instruction
for the first two years was in Bhojpuri dialect, the teachers speaking
Bhojpuri among themselves and at home as well. Some called the speech
of the children khEribgli, or uncultivated speech, literally bitter.

By eight years school attending children were developing some proficiency within this restricted environment in standard Hindi,
though they exhibited a wide range of interest and ability in this.
Only a little guddh Hindi learning takes plac, in the schools, and
boys from this age generally learn more or less formally within the
context of religious instruction from an elder family or outside
preceptor, or, failing this, may pick up some informally by attendance at religious festivalsor other functions where it may be
heard and seen. Within one neighborhood of artisans of several
jnti-s, boys from about ten -co twelve varied widely from little or
no ability to produce guddh Hindi to proficiency at nearly adult
level. The variable mast closely associated with this seemed to
he religious and ritual interests of a traditional sort, and an
interest in myth and narrative in general, in other words a semantic
context. In many families it is considered improper for a girl to
speak anything but Bhojpuri or to attend school, at least beyond
4
the age of nine or ten. Standard Hindi and iuddh Hindi are considered the province of males, especially elders, but this does not
prevent girls from being able to recognize, identify and understand
these styles, and to respond to them appropriately. Within Bhojouri
it is possible for them to produce all of the main patterns along
the respect continuum, and they learn much as the boys do from
religious functions. At the same age as boys, girls develop the
characteristic narrative style of Hindi; beginning with simple conjoined sentences with narrative intonation patterns at five years,
and increasing the length and imbeddedness of the sentences and
overall length and semantic complexity and cohesiveness of narrative

to early adolescence, when they have mastered production of the

adult style. A difference between speech styles of boys and girls
is discernable by the age of seven or eight, each recognizes that
of the other and will not use it. Here again the differences are
largely paralinguistic and kinesic, with a general feature we may
6
call emphasis predominating more in the boys's style,with more vari4Within the last thirty years more girls of educated families are using
the standard language of literacy.
5,
Ihis i s virtually the same for Bhojpuri aE

grammatical and lexical differences.

a style, allowing for the

- 74 -

abi;.ity in intonation patterns, and a wider scope for the same

general postures, gestures, etc., plus a greater overall amount

of talking allowable. Boys may with impunity use some forms like
slang and nicknames which most families will not allow their girls
to use. Most families, again,are quite particular that their children
in general conform to the standards of good, clean Bhojpuri and not
use abusive language. When asked what they most liked to hear,
chiljren varied considerably in their answers; in answer to what
they disliked most to hear, most replied abusive language. Some few
families, it must he said,diverge from this norm97
There is an important difference, currently receiving considerable attentjon,between linguistic competence and performance. This
underlies mcch of relevance in hilingualism and language acquisition,

of course, and as a concept possesses the virtue of testability with

both bilinguals and children.
Children's recognition, understanding, and classification of dialect-stylistic differences, as well

as their apropriate responses to them within this Hindi-Ehojpuri

system can,it is clear, be mapped out in process of development.
Working out the best model to explain the children's changing
distinct:ve feature systems and analytic strategies is more diffirultp
but can be approached through study of their behavioral and linguistic performance, both spontaneous and tested in various ways.
6

5emantic emphasis is signalled by several different means in Hindi

and Bhojpuri, often conjointly used. These include:
a) vowel lengthening beyond the phonemic
contrast,
b) use of the emphatic particloW or /ht/ which ig employed in many
ways and places,such as negative /nal 4. /hI/4/nahi/ emphatic negative,

c) reduplication of lexemes, phrases, clauses or whole sentences,

d) use of a rhymed doublet of the word requiring emphasis,
e) increasing the voice volume,
f) exaggerating the intonation patterns, and employing other paralinguistic devices, and
g) exaggerating kinesic features such as posture, e.Kpression, and
gesture. Children firmly possess all these features by age three.
Communication of emphasis is closely related to that of respect
levels: all its forms enter to some degree in both plus and minus
respect communication, the greater use being correlated with greater
divergence from neutral or ordinary respect. Further, the particle
4ii/expreEsly denotesrespect, as in gan9a-11, bffa-i1 (father), ht6-.ii
(yes sir), and AI-naht (no sir); and pluralization is used to some
extent in Hindi and more often than not in Bhojpuri to indicate
respect rather than literal plurality.

10

75 -

It is generally agreed among Bhojpuri speakers that Bhojpuri

seems most natural and comfortable, some parts of the traditional
guddh Hindi next so, the standard Hindi of necessary use third,
and last of all the more formal Hindi of upper castes. It is useful
perhaps to note that this is the same order in which these are acquired by Bhojpuri speaking children. Children also absorb early

a basic set of important knowledge of their culture and how to

behave in it; in fact, it is instructive for a researcher seeking
important patterns to observe what it is that young children are
learning, what they may be imitating and mastering, and what ways
they express creativity within their language and culture. In
studying children themselves it is also often useful to pinpoint
'mistakes' as defined by their elders' system, in that this can be a
guide to developing cognitive patterns and strategies of thought,
or competence within the larger system.
Bypassing the earliest stages of vocal production in cooing
and babbling, and even that of global, one 'word' utterances, we
note that the Bhojpuri learning child at approximately eighteen
months develops pivotal utterances of two component tqordst, has
already mastered most intonation patterns of Bhojpuri, and has a
rudimentary stock of gestures indicating negation, affirmation,
and respect to gods and some elders, among other things. He also
has some of the emphasis markers in his repertoire, has a stock
of verb root imperatives, and generally an impressive list of
He has learned some of the important features of
kinship terms.
family and temple ptii. From one to about three we may say he
speaks as he is spoken to in the family generally, in a style devoid
of formality or respect markers, except that he is early taught
to say namaste as well as perform the gesture, and will definitely
add the/-11/ honorific particle appropriately to his speech, as in
the early morning greeting often extended to me by one two-year-old,
namatEE,behen'l (greeting 4 emphasis,sister
respect). Also he

may early indulge in a bit of abuse as rogannl at 2,6, threatening

his mother's sister's small daughter: m5rE,bgi (beating, brother?)
apparently recognizing that behen cannot be used in such a context
but bh51, brother,can be used in a slang as well as ordinary context.
7 Also abuse language is compulsory under certain circumstances; for
example,at marriages old women must come to sing insulting songs.

11_

- 76 Before three children will be well in command of a stock of minor

expletives, such as hath, (h)E, arg, and others, used appropriately,
for example,to warn off a dog or even another child. They by three
have the particleirvMl5/2which may be roughly translated doer, and
is neutral referring to things or persons of artisan occupations,
but disrespectful for anyone else; and they use it appropriately.
In general they will have the system whereby a child or adult
addresses nonkin persons respectfully by kinship i:erms referring
to elders of the appropriate generation and sex, often with the/-12/

particle added. By three and earlier they know to address kin who
are older by kinship terms only, since it is disrespectful to call
anyone elder by his name. Somewhat later they learn the use of kin
terms is elastic also in that one can use a term belonging t'D the

next higher generation from the person addresselinorder to convey

still more respect in some cases; for example, didg, literally
father's father, for father's elder brother; or cgcg, literally
father's younger brother, for one's own elder brother.
In Hindi and Bhojpuri respect patterns are not equivalent to
polizeness formulae: there is no 'please' as such, nor are words
for thanks used under any but very exceptional circumstances;
expressions such as 'excuse me' are rarely used. Children usually
do not learn these at an early age. They do learn to supply all
relevant inflectional markers as a sign of respectful speech by
the age of four, and that long, involved sentences riither than
abbreviated ones are a sign of respect. A few chi]dren by three,
but nearly all by four appropriately use polite /iye/ verbal re
quest forms, such as baithivg (please sit), caliyg (let's go), and
khaiyg (please eat); and use mat, the negative before polite request
forms.

But even though isolated and increasing incidents of utterances

appropriate to a definite style occur in the speech of children as
young as two and three, we have little reason to suppose that they
have as yet any abstract concept of two separate stylistic systems.
It would be more faithfu,l to the data and to children's capacities
to judge novel contexts to suppose they have internali7ed bits and
pieces as yet too scanty to form any coherent broad pattern on an
adult style. Furthermore they combine elements of different styles
in the same utterance often up to the age of about eight, and often

12

interestingly reduce the respect forms in sentences they choose

spontaneously to imitate from older children or adults. For example,
Sitg, 6,2 returned to me rather unwillingly my pen, with the /O/
particle related to the sacred syllable 3m and thereby respectful,
but signifying half consent: kalamt lg laobehenil; then later fol
lowed with rOlE lff lata ha.(The ruler, all right, is being brought.
which her four year old brother echoed without the /ii/ as rill le
Or Hanumgn, 6,5 included the standard HindiAthaWmarker
9
old
brother
immediately
after his numbers, while his four year
n'ta ha.

afterwards failed to do soi Hanumgn: hamar pEc thah Toto ho.(I have
five photos.), and Bh'5gavandgs: naht, -ek kar le. (No, one bring.)

BindesvarI at five was well in command of such respectful utterances

dikh. and dEkhaivg..(please look.), but
as, calive-,
sometimes dispensed with them, as when it began to rain and her
baithivg. (You people please
mother respectfully said, andar
be eiated inside.), Bindesvari hurriedly insisted, oharmE caln.
(come in the house!)
By about six it appears a rudimentary sort of systematization
ON/

of styles is taking place, perhaps catalyzed by school and other

experiences outside the home, but still children of this age can
rarely sustain production in the less familiar dialect or style
for over a very few utterances at a time. Here their recognition
greatly outstrips their ability to reproduce0
ome children of
this age can imitate teachers and even holy men in production of
standard Hindi and of guddh Hindi, but generally exhibit shyness
over doing so in the presence of adults--a different situation from
their bold imitation of street vendors at three. By six they could
produce a haughty style of formality for semantic effect, as
Hanumgn's ah aongk-O bahut calgkh ha. (To the conceited one himself
he is very clever.) They continue with their peers to indulge in
abusive speech at times, as Pannalgl, 4,8 to Gitg, 4: aur mattl
_
khaveoe, ngk caoatarg. (And you will eat dirt, flattened nose.)
Systematic instruction, of course, could produce a clear de
marcation of styles or, even more clearly, languages by this age.
At six, the son of the mahgrgja of Banaras could publicly recite
from a vast store of memorized Sanskrit glokas, and knew guddh

Hindi. At ten, an apprentice to his dgdg, a piri, could recite

Sanskrit ond use guddh Hindi easily, while the eleven year old son

13_

of a clerk and particularly pious man follovia his father in con

ducting his own daily home pi-Jig in Sanskrit and c.iddh Hindi, separ
ating these clearly from the Bhojpuri he spoke at home ordinarily,
and the standard Hindi he spoke at school. Another not unusual
eleven year old boy could easily recite myths with almost a full
command of 4Liddh Hindi style in all features, keep his school Hindi
separate frcm this for the most part, and keep his home Bhojpuri
entirely separate.
It would appear that,by ten or eleven certainly,these children

exposed to different styles in different contexts have almost entirely

separated them according to different sets of distinctive features
into integral patterns, and that they are thus able to do what
many adults within their same social groups have not completed.
There seems to be considerable elasticity in the system itself,
which allows many to overlap their styles, yet encourages some to

separate them more fully. And the very closeness of these styles
on a respectlevel continuum makes their study interesting, and
their development :in children revealing, as it shows the types of
confusions and the kinds of separations made during the process of

learning, as well as sometimes indicating criteria and strategies

used for developing systematizations.