You are on page 1of 16

-rk. '~e v~/ofl)).

~ eyt/ ()
~ fhe-dt:K· 11) Ii t'l Individual Differences 293
C H .A PTE R
abiliry in other beginning language learners, These scudies have grappled with [he
problem of how besr ro describe such differences. AIe Johanna and Caitlin anending
ro d ifferent aspects of rheir environment, using language for different purposes, or seg­

Individual Differences menting m e speech [hey hear in different ways? Other quesrions concern where such
differences might originare. Do children orienr [Q objecrs versus people, or exploi[ dif­
ferent learning or processing mechanisms' Or are these differences rooted in charac­
ImpLications for the Stud~ of [erisrics of rhe language meir caregivers use'
These issues have only begun ro be !xplored for children learning English. We
Language Acquisition will need funher evidence from orher languages ro so n our rhe problems [hat varied
languages present and [he array of solurions ma[ individuals bring [Q [he rask of learn ­
ing [Q ralk. These solu[ions, moreove r, carry with [hem implicarions for a number of
recenr [heore[ical proposals, such as how early lexical development is rela[ed [Q gram­
mar, whemer dis[incr modules or general lea[ning mechanisms suppOrt acquisition,
Beverlld A. GoLdfieLd and [heexren[ [0 which nouns and me learning principles [hat apply [Q object labels
Rhode IsLand CoLLege are [he most prominent feacures of earl y lexical development.
This chapter begins by (facing [he history of individual differences research, ex­
Catherine E. Snow ami nes variation a[ different levels oflanguage learning, and considers how data on in­
dividual differences may inform current theoretical debates.
Harvard Graduate SchooL of Education

[ c!igh[een months, Johanna has a substantial vocabulary of single words tha[

A label im po rtant objects and entities in her world. She can, for example, [3.1k
aboU[ food (banana, apple, cheese), clothing (sock, shoes, hat), an imals (birdie, Ctlt),
household items (kqs, light), and roys (do lly, btll!). Nonnominals are fewer: hi, bye-bye,
"p, down, no. Many of her words are learn ed and used in [he course of nam ing games
[hat she plays with her parents. Ba[h rime elicits nose, teeth, eyes, ears, foce, hair, belly. Pic­
[Ure books are also enjoyed as opportunities for displaying Johanna's word knowledge.
Eigh[een-month-old Caitlin also has words for things ro ear, wear, and play with.
However, her lexicon includes many nonnominals (e.g., there, pretty, nice, yuck, ollch,
no, bye-bye, tlh-oh, down, "p) and quire a few phrases (e.g., let's go, bless you, sit down,
hey you guys, lemme see), some of which may be amibu[ed ro [he presence and influ­
ence of an older sibling. Caidin produces th ese phrases with appropriate melodic (e.g.,
Where are you?) or emphatic (Don't touch') inrona[ion. Like Johanna, she enjoys pic­
[ure books, bur [his context el icits more "collversa[ion" [h an labeling. On one such oc­
casion, Caidin sat with Mom and a fami liar book, turned [Q a fa vorite page, and
del ive red a p rosodical ly varied bu[ unintelligible C'Nenty-seven-second discourse, em­
bellished with [he pauses, gesticulations, and gathering momenrum of a narrative.
Johanna and Caidin have each made considerable progress since flrs[ words ap­
peared a[ about C'Nelve months of age. They talk about familiar people, entities, and
events, and have some words in common. Howeve r. [here are considerable individual Picture books elicit labels fom some children and animated conver­
differences in these early lexicons. In [he past [hirry years, researchers have noted vari- sation fom others.
292

~
~
._Q ~ ., .--.",. ' :--:--:---_..

294 Chapter Eight Individtuzl Diffirences Individual Differences in Early w"rdr 295

The History of Individual Differences before the onset of word combinatio ns. As the scope of child language research broad­
ened, departures from a universal acq uisitio n sequence began to be noticed and ac­
in ChiLd Language Research corded some significance. Using larger samples of children to study the meaning and
fonction as well as the form of early language, investigators have si nce observed that
The topic of individual differences has not always been of interest to child language children vary along all three dimensions.
researchers, and early texts would hardly have devoted an entire chapter to its discus­ Another facror contributing ro increased reports of variation is rhe attention now
sion. This change can best be understood by considering the history of child language paid to children and ch ild utterances previously excluded from srudy. Children with
research. Interest in individual differences followed almost twO decades of research poor articulation and early jargon like serlrences now appear in the literature (Adam­
committed to documenting universal patterns of acquisition. Even though researchers son & Tomasello, 1984; Peters, 1977, 1983). Some invesrigarors have made deliberate
during this period typically reported (and dismissed as unimportant) variations in the methodological decisions nor to select children on the basis of a priori decisions about
rate of language development, they placed greater emphasis on similLzrities among chil­ the representativeness of their language or language envi ronment (Bloom, 1993;
dren in the sequence of development. At one level, emphasis on commonalities among Lieven, 1980). We also have the benefit of studies of children from diverse language
language learners grew OUt of a practical need for basic information about the nature communities (e.g., Slobin, 1985, 1992, 1997) and varied cultural and socioeconomic
and sequence of development. Another contributing factor, however, was the influence groups (e.g., Bloom, 1993; Heath, 1983; Hoff-Ginsberg, 1991, 1998; Lieven , Pine, &
of linguisti c theory on chi ld language research. In the early 1960s Chomsky's (1957) Barnes, 1992) .
theory of transformational syntax offered a new and coherent way of accounting for Serious mention to individual differences began in the 1970s, with a few stud­
structural principles of adult linguistic competence that CUt across inter- and intralin­ ies that reported var iation in children's first words (Dore, 1974; Ne lson, 1973) and
guistic diversiry. For the next ten years the kinds of questions asked and the methods early sentences (Bloom, Lightbown, & Hood , 1975; Starr, 1975). Since then, research
used to study chi ld language were direct outcomes of applying the new theory to prob­ into individual differences has exp lored one or more of the following questions: (I) In
lems of acquisition. Child language research focused on questions of structure, with the what ways does language-learning vary) (2) What foetors contribute ro individual dif­
intent of documenting the mles governing children's early sentences; fot example, stages ferences) and (3) What are the implications of individual diffe rences for understanding
were hypothes ized to characterize the acquisition of wh-questions and negative sentences, the process of acquisition, for devising an adequate theory of language development,
(e .g., Bellugi, 1967), and of noun and verb inflections (e.g., Cazden, 1%8). and for clarifying the complex interdependence of cognitive, social, and linguistic fac­
The focus on linguistic unive rs als during this period carr ied with it ce rtain as­ rors in development)
sumptions about the methods that could be used to investigate child language. For in­
stance, since all normally developing children were thought to co nstruct similar rule
systems, longitudinal study of a single ch ild or a few children was a typical research par­
adigm (e.g., Braine, 1%3; Brown , 1973; Miller & Ervin, 1964). Many studies looked
Individual Differences in Early Words
cross-linguistically for common structu res and stages of acquisit ion (e.g., Bowerman, Nelson's (i 973) srudy of ea rl y lexical development was the first ro draw attention ro
1973; Slobin, 1%8). Whi le this paradigm guided much research and outlined the variability among young language learners. Nelson collected diary data on the pro­
major dimensions of language development, it also biased us toward seeing shared pat­ ductive vocabularies of eighteen children (seven boys and eleve n girls). The first fifty
terns of development in the data. Children in some studies were selected for inclusion words of each child were assigned ro form classes (nominals, action words, modifiers,
because of the ease with which the researcher could understand and record theit speech personal-social items, functiol1 words) based on content or the child's first use of a
(e.g., Brown, 1973) . Children with less clear articulation or "messy, " jargonl ike strings word. Nelson found that all of the child ren acquired words for fami liar peo ple, ani­
in their early speech were less likely ro be included (Pe ters, 1983). Similarly, utterances mals, food, roys, vehicles, and household objects. The children varied, howeve r, in the
that appeared to be advanced, imitative, or nonrule-generated in an otherwise pre­ proportion of nominals in their vocabulary. Ten referential child ren had early lexicons
dictable corpus of child speech wete often relegated to the anomalous or miscellaneous that were dominated by wotds for objects. These children moved predictably from sin­
catego ry Jnd were excluded from furthet study. gle words ro a two-word stage. A sudden Spurt of new words near th e fifty-word level
Several factors are responsible for the paradigm change reptesented by a focus on often preceded the appearance of word combinations. An early preference for object
individual differences. Fitst, linguistic theory grew more attentive to semantic and prag­ labels was positively related to talk about objects and negatively related ro talk about
matic aspeccs of adult language, and child language research also began to shift away self in a follow-up speech sample at twenty-four months of age.
from an exclusive emphasis on syntax. In the 1970s investigators became interested in Eight expressive chi ldren followed a different route. They had fewer object labels
the meanings of early words and sentences and in the ways in which language was used but more pronouns and function wo rds than the first group. They also acquired many
- ------,-- - - - - - - - - - - - -- - -..........,.---=-:"-:"::""-::-...--,--:-"- -.... ,- .? -.- •. ---~- -,

296 Chapter Eight Individual Differt!?/w Individual Differences in Early Wtmu 29 7

more personal-social expressions, which were usually longer chan a single word. From ges[ a differem pauern [han does the discribucion of different word types in [he child's
early on, chese children used phrases such as, "go away," "swp ie," "don'c do ie," and "1 lexicon.
Want it." Their cransicion intO syntaccic combinations was less clear and noc marked A chird problem conce rns rhe kinds of crireria used to classifY children's words.
by a rapid increase in new vocabulary icems. Nelson's original classification scheme is a mixcure of formal and functional charac­
Alchough there was no difference in the age at which the cwo gwups acquired reris[ics. Nominals, for example, were defined as "words used w refer to [he 'rhing
fifi:y words, children in che referential group included boch early and lace calkers who world' " [hat could be used w label or demand. However, some nominals could be clas­
tended to learn words ac a faster race chan children in [he expressive group, who evi­ sified otherwise, based on the child's repo,rred use of the word. Thus, Mor would be
denced a slower, steadier race of acquisition. classified as a nominal if the child !Jsed th!word when simply tOuching ir, or as an ac­
Nelson argued [hac chese differences reflecced che children's differing hypotheses [ion word if [he child appeared co wane ro go outside. This confounding of formal and
abour how language is used. Children wich referential language were learning to calk functional criteria has led [0 some confusion in how w imerpre[ rhe referential and ex­
abouc and cacegorize che objects in their environment. Children with expressive speech pressive dis[incrion. Do children with more referemial speech prefer co calk abour ob­
were more socially oriented and were acquiring che means w calk abouc chemselves and jec[s? Do children with more expressive speech more ofi:en use language for social
ochers. purposes ' Wichou[ independenc observarional evidence of how children are accually
Alchough Nelson intwduced an imponant new approach w che study of lan­ using rheir words, we can nor confirm Nelso n's suggesrion [hac rhe referemial-expres­
guage development, subsequent research has pointed our a number of problems with sive disrincrion refl ec rs children's differing h ypotheses abou r [he functions oflanguage.
che origi nal referential-expressive discinccion. These include (1) [he use of parental re­ Subsequenc research has looked direcdy ar functional differenc es in children's
pon as a soucce of data, (2) [he composicion of che early lexicon versus the frequency early speech. Dore (19 74) examined how cwo children used [heir words in video raped
wich which children use individual words, and (3) the cacegories and criceria used [0 conversarions with their mothers and nursery school reach ers. One child used clearly
define the kinds of words children acquire. arricula[ed single words to label, repea r, and practice, and her speech in volved ochers
Parental reportS have been a valuable source of child language data for as long as 26 percenc of rhe rime. The second child produced fewer words buc used prosodic fea­
philosophecs and psychologiscs have observed and recorded development. Parencs are cures co communicare in more ways. His u{[erances included'orhers 63 percent of the
with cheir children in varied contexcs in and OLl[ of che home and ate typically che rime . Dore suggests tha[ the first child's language was code-ori ented, concerned with
child's earliesc and mosc consiscent conve rsational parrners. However, research com­ representing rhings in the environment. The second child was message-orienced, more
paring spontaneous and eliciced speech wich parental diary records finds chac parencs of[en using language co manipulate [he social sicuacion. Thus, Dore finds some sup­
repon more nouns and fewer verbs chan children Lise (e.g., Baces, Bte[henon, & Sny­ porr for Nelson's hypothesized functional differences, but ic is not clear if rhese fWO
der, 1988; Benedicc, 1979; Pine, 1992b; Tardif, Gelman, & XLI, 1999). This discrep­ children also differed with respect w rhe kinds of words (e.g. , nominals versus personal­
ancy may be due co bias on che pau of English-speaking parents co nocice and repon social words) they used.
words for objects. On [he ocher hand, even young language learners may know a num­ Pine (1992a) collected diary data on che tlrst one hundred words of seve n chil­
ber of labels [hac [hey h ave racher limited opponunicies w use. Words such as lion, dren and coded audio raped speech samples for va riou s funccions, including auention ,
moon, and peas may be evoked during book reading, bedtime, and dinner, respectively, labeling, descriprion, demand , and prorest. Alrhough ch ildren varied considerably in
bur are less likely ro occur in a recorded play session. [n any case, [he excent w which rhe prop o rrion of common nouns.in cheir lexicon (a range of28 percenc [054 percenc
a child's lexicon is judged co be "referemial" or "expressive" can be expected co vary with a[ fifty words, and 35 percent ro 67 percent a[ o ne hundred words), there was no re­
how [he researcher obtains informacion on [he child's vocabulary. lacionship between referencial vocabulary and any functional caregory at either vocab­
A second problemacic area concerns how frequendy children accually use [he var­ ulary level.
ious words [hac [hey know. Even in cases where nouns accou nt for about half of chil­ Bowerman (1976) also points ouc thac word usage may shift over rime. She cires
d[en's reponed vocabulary, [hey may be used less of[en [han ocher words (Pine, 1992b) . an example from Ferrier (1975), who repom thar her daughter inicially used phew ex­
Three children observed by Lieven (1980), for example, acquired more nominals than pressively, ro greer her mOther in the morning. The word was originally an imica[ion
any ocher word class, and were chus similar ro Nelson's refere mial speakers. However, of her mother's own rourine commenc on rhe odo r she invariab ly encountered on chese
Jane used general nominals more frequendy [han ocher types of words , whereas Ka[e occasions . The same word was later used by her daughter referencially, as a name for
used almost as many personal-social words as general nominals. Bech, on [he orher diapers, clean or soiled.
hand , used man y more names for people, nonclassifiables (ambiguous words such as Thus, although children may \'ary in [he kinds of words they acquire, [here is no
there), and aerion wo rds [han [he ocher cwo chi ldren. Thus, frequency of usage may sug- consis[enr supporr for the no [ion [har c hildren with rela ti vely more nouns use langu age

5 ::: ,, ~--, -""--~- "

----
Segmenting the Speech Stream 299
298 Chapter Eight 11Idividzuzl Differences

..-----.- ..- .. - .. -,- .. -.~


Refe rential Lexicon ExpreSSive Lexicon
-+----- - - - - -~
Code·Oriented Speech Message-Oriented Speech

------------------~
-+------
More Common Nouns More Phrases
;
--- --------------
Analytic Srrategy Holisric SU" tegy

-------------
Nominal Combinations Pronominal Combinarions

Figure 8.1
Some dimensiollS of individual differences ill early language
developmen t.

er ai. , 1992). Ir is im porranr ro nore, however, rhar rhe refere nrial-exp ress ive dimen­
sion is nor a dichoro my, bur rarher a co nrinuum along which in divid ual children vary.
Mosr ch il d ren appear ro acquire a relarive balance of re Fere~ ri al and express ive lan­
guage; only a few children acquire a dis rri burion ex rreme eno ugh ro be call ed a disri ncr
style o r srra regy. A close examinarion of rhe extre mes, howeve r, should help us ro dis­
en rangle rhe possible mechanisms and processes rhar contr ib ure ro ea rly lexical devel.
opmenr for all children (see Figu re 8.1).

Segmenting the Speech Stream


T he difficu lries inherenr in describi ng children's ea rly speec h in rerms oHormal andlo r
Code-oriented children are imeresced in che UJorld of objeas. fu ncriona l charJ.cre risrics have led researchers ro alrernarive ways ofco nceprualiz ing in ­
div id ual differe nces . T he rendency of some children ro acquire longer, phrasel ike ur­
rerances du ring rhe single· wo rd srage suggesrs rh ar children may di ffer wirh respecr ro
in more nam in g :\Od fewe r social conrexrs. Many, if no r m o~r, ea rl y words se rve a va­ rhe lengrh of rh e lingu isric unirs rha r rh ey segmen r From ad ulr speech. A samp le of such
riety of func rions, and rh e distribution, fi mccion, l nd frequency of wo rd usage are rela red ph rasa l speech ca n be Fo und in ou r in rrod ucro ry illustrar ion of Cai rlin's ea rly language.
bur separable aspecrs of early lex ical developmenr. No re rhar Cai rl in's phrases are rwo and rh ree words in lengrh, bu r rh e individual words
A numbe r of srud ies have subsequen rl y co nfir med some ve rsion of rhe reFeren­ do nor occu r alone, suggesring rhar rhe enti re "package" has been lea rn ed as a whole.
ri al-exp ressive disrincri on in children's firs r wo rds . Some chil d ren acqui re relar ively Moreover, she typ ically pro duces rh ese phrases wir h adu lrl ike inronarion. (For a more
more co mmon nouns (e.g., Brerh er[On, McNew, Snyder, & Bares, 1983; Lieven er .,1., rhorough desc rip;io n ofCairlin's language, see Goldfie ld, 1985/86) .
1992; Pi ne & Lieve n, 1990), or words rh ar label w d desc ribe rhe properries of obj ecrs A1chough segmenca ri on may be Faciliraced by che shoreer urcerances, exaggeraced
(Goldfield , 1985/86, 1987) , whereJ.s orhe r childre n acquire many words [0 describe intonacion. paLlses, reperirions, and mess paccerns of child·direc red calk, chi ldren musr
their ow n ac rio ns and srares <Lnd use mo re ph rases (Go ldfield , 1985/86, 1987; Lieven build rhe ir lexicons From che raw mare rial of connecced speech. The unics ch ar chey
-.....,-------------~~~ - - ~ ...

300 Chapter Eight Individual Di/firences Individual Differences j" Early Sentences 301

select may thUS be single wo rds or longer phrases. A number ofstudies repon chi ldren remained srable when the firsr fifty words were compared to rhe second fifry, and the
whO orient to syllables and segme nts and others whO at tend to prosodic tunes that rwo measu res were negarively correia red ar both vocabulary levels. This patte rn of re­
unify larger sequences of speech (Echols, 1993; Klein, 1978; Lieven, 1989; Perers, sults suggesrs mar frozen phrases and common nouns may more precisely define rwo
1977,1983). Serh and Daniel rep resent these differing straregies: approaches to early lexical developmenr. Moreover, rhere was no significant correlarion
(FJormula ic chi ldren, like Seth, pay anencion [Q "horizoncal" information such as rhe berween eirher measure and age at which fifty and 100 words were acquired, sugges t­
number of syllables, mess, intonation patterns ... word-oriented children, like Daniel, ing rhar neirher srraregy affords an advantage ar rhis level of development.
pay more arremion [Q rhe vertical segmencal informar ion co nrained in single (usually Perers ( 1977, 1983) describes how tl).e child she observed acquired borh phrasal
messed) syllables focusing on rhe derails of consonancs and vowels. (Perers & Menn, speech marked by stress and intonarion, ak more clea rl y arti cu lared single wo rds. She
1993, p. 745) rermed chese rwo kinds ofspeech gestalt and analytic, respectively. However, as Pine and
Lieven (1993) point OUt, rhe segmentarion of phrases from adult speech is also an an­
A disrincrion rhar may be relared to rhese diffe ring segmentation strategies oc­
alytic process. Whar differs is rhe length of rhe unir that children exrracr from the
curs in child ten's early phonological systems as well. Ferguson ( 1979), for example, has
speech rhar rhey hear. The occurrence of phrasal units in children's early speech raises
described cautious versus risk-raking approaches. Some children's early utterances seem
a number of quesr io ns abour rhe relarionship becween rhe lexicon and syntax. We will
to be generated on rhe basis of an e1eganr and orderly set of phonological rules, such
return to rhis tOpic afrer firsr reviewing rhe evidence for individual differences in ea rly
rhar rhe child form of any adulr word is highly predicrable. Such children (e.g., Smith's synracric developmenr.
subjecr Amah l, 1973) eithet apply rheir rule system consisre ndy to imitated as well as
ro spontaneous fot ms or else resist imitating words that would constitute violations of
the tesrrictions on their output. O thet children, in conrrasr, operate with fairly sloppy
phonological systems, Showin g alternation among several ways of producing most Individual Differences
words and applying their phonological rules optionall y. Typically, the children wirh in Early Sentences
sloppy phonological systems inco rporate imitated (and thuS ptogressive) forms into
their lexicon quite easily and may be more likely ro show improvement in production Two years after Nelson's srudy of variabiliry in early words, 8100m and colleagues
:lS a result of direct modeling (Macken, 1978). (1975) reponed that children also differ in the early srages of multiword speech. They
Plunkett ( 1993) suggests that articulatory fluency and artic ulatory precision are found that the senrences of rhe four children they observed had similar contenr (they
inversely related in early speech. Phras:ll speech rep resents segmen[:ltion that over­ talked about a common ser ofse mantic c:ltegories such as recurrence, negarion, actions,
shOOts a target adult wotd. Such expressio ns tend to be produced t1uenrly but with less and srates) thar emerged in J. similar ord er. However, the form of rheir early sentences
precise arriculatio n of the individual pho netic segments. Arricularory precision is the differed.
outco me of an alternate srraregy that undetshoors the adu lt target word by focusing on All four children used a pivot Strategy to encode negation and nonexisrence (no +
accurate production of sub lexical units. Variation in perceptual acuiry, in verbal mem­ X, no mor!! + X) and recurrence (more + X, 'nother + X), This straregy consisted of com­
ory, andlor in ch:lractetistics of the input may influence the kinds of linguistic units bining one of a small class oflunaion words and anyone of :llarger, varyi ng se r of con­
that are perceprually salient and likely to be used. The so unds learners attend ro may tent words. Eric and Perer Llsed rhis same approach ro exp ress action, locatio n, and
be the sounds that rhey hear often and know they can produce, resu lting in a kind of possession by combi ning all-purpose pronouns wirh comenr words. They produced ut­
arriculatory filter thar may by unique to each child (Vihman, 1993). terances such as [finish, play it, sit here, and my truck. During rhis 3ame period Ka thryn
Lieven and colleagues (1992) have coined rhe rerm frozen phrases for phrasal and Cia expressed rhe same semantic relarions by combining content words, as in Gill
speech that :lppe:lrs before true word combinations. They defined froze n phrases as ut­ pllSh, touch milk, sweater chtlir, and Kathryn sock.
ter:lnces containing rwo or more words that h:ld not previously occurred as si ngle unirs 8100m, Li ghtbown, and Hood ( 1975) claim that the children were using rwo dif­
in the child's speech. They examined rhe numbe r of common nou ns, personal-social ferent combinarorial srrategies. The pronominal2pproach used by rhe cwo boys allowed
words, and frozen phrases in :l sample of twelve children who were observed longiru­ them ro begin encoding relarionships bwveen objecrs and eVe nts withour relying on
dinally using parent word diaries and periodic audio samples of the children's speech. specific lexical ire.ms. Since rhey used a varied lexicon in single-word utterances and in
They found rhar personal-social words declined as vocabulary increased, suggesting sentences wirh no and more, their strategy could not be atuibuted to simply nor know­
that such words may not be a stable defIning cha racrerisric of eKpressi ve sryle. On the ing enough labels. The (WO girls, on rhe orher hand, preferred to ralk abour rhe same
other h:lnd, rhe proporrion of co mmon nouns and frozen phrases in children's lexicons meanings using specific nouns. When MLU approached 2.5, rhe cwo systems began to
302 Chapter Eight IndividuaL Differences Stability ofIndividuaL Differences 303

overlap. Children using a pronominal strategy combined mo re content words, whereas . Functional diffirences that extend from single words ro word combinations have
me nominal strategy children incorporated m ore pronouns inro their u((erances. also been reponed. Star r (1975) observed twelve children from one ro two-and -a-half
The child observed by Goldfield (J 982) also exh ibited an early pronominal style, years of age and found tha t children who preferred ro label objects in single-word
but close examinarion of her multiword speech over time tevealed that a few specific speech tended ro produce two-word sentences encoding object-attribute relations. A
patterns accounted for the relative dominance of pronouns. For example, the roles of second group of children described objects less frequently but used more interjections
agent and possessor were initially limited ro the ch ild herself and were encoded by Pto­ (conventional social responses such as hi, bye, ouch) and made more self-references
nouns J and my, respectively. These semantic roles later broadened ro include orhers, (e.g., want ball) in their early sentences. ,
and in these cases nominals were used ro encode the constituent. Earliest action Utter­ Similarly, Lieven (1980) reportS tharfearly sentences of the three chi ldren she ob­
ances (I'll do it, J found it) were initiall y rote phrases. Later action u((erances rook the served appeared ro derive from characteris tics of their single-word speech. For one
form of agent + acrion + it and appeared ro have evolved from the earl ie r pattern. A child both single words and earl y consuuctions (e.g., there, mommy, there Julian, and
number of researchers have recencly argued that children's earliest sentences reflect spe­ there mommy) were used to gain adult attention rather [han to convey reference. The
cific lexical combinarions rarher m an instantiations of m ore abstracr semantic roles o r other fWO children were more likely to describe attributes and actions of people and
syntactic rules (e.g., Childers & Tomasello, 20 01 ; Lieven, Pine, & Baldwin , 1997; things in both their single- and multiple-word utterances.
T heaksron , Lieven, Pine, & Rowland, 200 2; Tomasello, 1992). If this is the case, we As we saw earlier, some chi ldren, such as Caitlin, acquire a number of phrases in
would expect considerable variation in ea rly gramma tical development, with children the early stages of learning to talk. Nelson's exp ressive speakers produced go away, stop
differing in rhe words and patterns used ro express their sentential meanings. it. don'tdo it, J want it. Suc h phrases are the resulr of segmenring larger unirs ftOm adulr
These data also suggest that a global m easure such as MLU may obscu re differ­ speech. Ir is not clear, however, how rhese early units are related to larer grammatical
ences in how children lengrhen meir utterances. Some children may PUt rogether sen­ developme nt. Some research suggests that early phrasal speech is unrelared to later an­
tences rhar emphasize semantic content, whereas o the rs focus on grammatical alyzed productions. Bares and colleagues (1988) followed twenty-seven children from
morphology (Rollins, Snow, & Willet[, 1996). Chi ldren who are language matched using first words to early senrences. A number of observarional and parenr inrerview mea­
MLU may thus exhibit profoundly different knowledge of the grammatical system. sures were used to assess language progress when children were ren, rhirteen, twenty,
We will prese nt more evidence for varied approaches ro early syntax in the next and twenty-eight monrhs of age. Bares an d colleagues report thar an analytic style,
section, which reviews those studies rhat have looked at the stability of individual dif­ which included high levels of comprehension and flexible noun production at thirteen
ferences in children who have been observed from early words ro first se ntences. months, predicred advance d grammatical developmenr ar twenty-eight months. Var i­
ables rhat sugges ted an early rote or holistic style, on the other hand , were unrelated
ro brer grammarical progress.
StabiLity of IndividuaL Differences T his line of research has been criticized by Pine and Lieven (1990, 1993), who
argue rhar cross-secrional, age-relared measures of individual differences co nfound strat­
The var iation we have seen in srrategies children use for segmenring rhe speech they egy wi rh variation due to differences in developmentalleve!. Because the ptOportion of
hear, for expressing early meanings, and for in troducing srruc rure inro rheir language nouns increases as vocabulaty tOrals rise, children assessed at the same age but diFfetent
may simpl y re ptese nt different entty points inro rhe language sysrem, or may be early vocabulaty levels will differ in the number of nouns in th eir lex icon, regardless of their
signs of differences that persist across development. Do children wirh a preference for parricular style or srrategy. Thus, assessmenrs of individual differences shou ld be based
object labels tend ro develop a nominal strategy for their early word combinarions' Do on comparable vocabulary tOtals (as was the case for Nelson's original study).
children with relatively more expressive speech prcfer a pronominal approach? However, using conrrolled vocabulary level s, Bates, Marchman , ThaI. Fenson,
There is some evidence thar children's early lexical preferences are reflected in the Dale. Reznick, Reilly, and Hartung (1993) continue ro find evidence that early phrasal
fo rm of thei r first wo rd combinations. Nelson ( 1975) followed rhe larer language de­ speech is unrelated to later grammatical development, whereas Pine and Lieven ( 1993)
velopment of her original refere nti al-expressive sample. Usi ng rranscripts of speech argue that early frozen phrases predict Iatcr productive wo rd combinations. The ap­
recorded whe n the chil dren were twenty-four and rhirty monrhs of age, she found that parent di sc repancy, however, may be due in part to differences in age and the measures
referenrial speakers began with a high propo rtion of nouns in early sentences. W ith in­ used to assess borh Style differences and grammarical progress. Bares and her colleagues
creasing MLU, rhe use of pronouns increased while no uns decreased for these children. (1993) used data from 228 children whose parents complered a checklisr assessing vo­
Children from rhe original expressive group began with a balance of noun and pro­ cabulary and grammar at twenty monrhs, which was repeared six mon rhs Iarer. They
noun use. Pronoun use changed very lirrie for this group, but nouns inc reased wi th ad­ found no rela tionship between rhe number of closed-cl ass words (preposirions, articles,
vanced MLU. aLLx ilia ry verbs, question words, pronouns, qllantifiers, and connecrives) rhar children
304 Chapter Eight Individual Differences Sources ofVariation 305

used at twenty months and the extent to which pa rents judged that children's sen­ dren with expressive speech imitate large units and social expressions, chi ldren wim ref­
tences included words and inflections thar indicare g rammarical complexity (preposi­ erentiallangu age tend to imitare object labels, parricularly mose they do not already
rions, arricles, auxiliary verbs, copulas, modals, possessives, plu rals, and rense markers) know (Leonard, Schwam, Folger, N ewhoff, & Wilcox, 1979; Nelson, Baker, Den­
ar twenry-six months. Thus, rheir analysis found no conrinuity betwee n grammatical ninger, Bonvillian, & Kaplan, 1985).
words that may appea r in ea rly phrasal speec h and a related ser of words and inflec_ The evidence mat individual preferences in language learning are somewhar sta­
tions that playa role in larer sentence constr ucrion. ble over me roddler period raises me possibility that more extended longitudinal stud­
Pine and Lieven (1993), on the omer hand, looked at data from seven children, ies would find even more subsrancial srabil\ty. For example, do child ren show me same
who were observed using maternal report and periodic audio recordings, from eleven preferences if fa ced with the task of learni~g a second language later in childhood, or
to twenty months of age. They found rhat the proporrion of frozen ph rases children when faced wirh language-related tasks such as learnin g ro read) Bates's daughter Julia,
used was positively related to the number of productive word combinations, in con­ who had been a highly referential child in learning English, was reported ro ado pt a
trast to the proportion of common nouns, which was unrelated ro sen tence produc­ much more holisric and expressive strategy in acquiring Italian, bu t it must be noted that
tion. This analysis, then, finds continuity be tween early phrasal speech and the she was exposed ro Italian during large group interac rions quite different from the pri­
frequency and pa[(er n of word combinations. Thus, the two studies are concerned marily dyadic social ci rcumstances in which she learned English. Buss is, Chi[(enden,
with so mewhat diffe rent aspects of grammar at two different periods of development. Amarel, and Klausner (1984) found rhat early readers seemed to split into two groups.
Children's acquisi tion of frozen phrases has led researchers to reconsider the So me children used more meaning-driven reading strategies, often previewing texts and
processes that underlie early syntactic development. Phrasal speech allows some chil­ studying illustrations to get a sense of the content before reading, skipping eas ily over
d ren to derive combinatOrial pa[(erns by segmenting phrases intO frames with slots that passages they co uldn't read, and sometimes making up substitute passages for bits they
can then be filled with other lexical items (Pine & Lieven, 1993). Peters (19 77, 1983) could not decode. Other beginning readers were highly faithhtl to the text, sounding out
also suggests rhat phrasal urterances may be stOred, retrieved, and used as single lexi­ unknown words carefully, and building for rhemselves highl y sequenced text reptesen­
cal items thar are later analyzed and btOken down intO productive co mponents. She tatio ns. Long-ter m longitudinal s tudies may help us SOrt out so me day whether, indeed,
suggests the term fission fo r the eventual breakdown of phrasal speech, as distinguished the young language learners who prefer conservative, data-gad'l ering, analytic strategies
from the complementary process of building sentences by combining units, which she pers ist in the use of these strategies for second language learning and for reading, while
calls fw·ion. Individual children may fav o r one or the other approach as rheir entry into their more risk- taking hol istic peers maintain their preferred strategies across age as well.
syntax: one rhat begins with single words :tnd co mbines them to build phrases, or one
rh ar proceeds from whole ph rases to co mponent parrs, with rhe parrs larer re co mbined
in novel utterances . Detailed longitudinal analyses reveal rhat child ren form ulate many Sources of Variation
of their early sentences by building up d is rributional pa[(erns around speci ft c lexical
items (e.g., wanr x, there's tl x, what's x doing). As we mentioned earlier, specifiC pa[(erns Given rhat the individual ditferences we have described above exist and are robust, the
may more accurarely characterize early grammatical development than m ore general question that arises is what might explain rhem. D o they reRect child factors- fo r ex­
seman ric (e.g., agenr + action) or syntacric (e.g., verb + objecr) relations. Th at is, chil­ ample, degree of sociabili ry, dispariries in pe rceptual processing mechanisms, or cogni­
dren ftrsr iso late individual segments and look for parterns in what comes before or tive sryle? Do rhey perhaps teRect differences in how ca regivers interact with or talk to
after. Children differ, howeve r, in rhe size of the items (words, formulae , or partially children ' Do they anticipate differences that characterize the range of different languages
analyzed phrases) rhar they isolare and combine (Thai & Bares, 1988) . within rhe human community? In other words, is the lear ner of English who co ncen­
The rendency to imitate adult speech is ano ther dimension of variation that pre­ [fates on single words JUSt revealing an expectation that her lan guage will be one like Chi ­
dicts some stabi lity across language levels. Child ren who have a stro ng tendency to im­ nese, with little morphology, whereas the more ho listic learn er who in clu des syllab ic
itate prosodic parterns may, as a result, acqui re phrasal unirs as wel l as single words slots where adults produce affixes is expecting a langu age more like Turkish or Po lish)
during the one-word srage, produce more high-frequency items of low semantic value
(such as pronouns) du ring rhe early se nrence stage, and have messier phonological sys­
tems all along . Relationships between imirariveness and the tendency ro be exp ressive,
Child Factors
to produce longer urretances, and to show high levels of unintelligibility have been re­ Perhaps [he most obvious difference among child langudge learners is in rate of learn­
ported (Bloo m et aI., 1975; Ferguso n & Fa rwel l, 1975; Nelson, 1973), but other stud­ ing. So me children begin talkin g close to thei r first birthday, while others wait another
ies have fou nd rhar children wirh refere ntial language imitare 11l0re and suggesr that six months or more be fore words appea r. Most children acq u ire rheir ftrst words slowly
th ese differences lie more in whar children imirare than in how much. Whereas chil- and sporadica ll y. In this ea rly stage , words may be learned case by case, with eac h case
306 Chapter Eight b,dividual Differences Sources a/Variation 307

an independenr relarionship berween so und panern and referent(s). However, as rhe it is clear that such shy children are likely to be slower second-language learners (fill­
lexicon nears fifty words, many children exhibit a "vocabulary spurr" that signals a pe­ more, 1979), as well as less obviously competent communicatOrs in rheir first language.
riod in wh ich rare of word learning accelerates dramarically (Goldfield & Reznick, Orher child factors to consider include Nelson's (1973) proposal that differences
1990,1996; Reznick & Goldfield, 1992). By eighteen months, the average child has in children's prelinguistic conceprual organizarion contribute ro their early preferences
a vocabulary of abour sevenry-five words, although perfecdy normal rwo-year-olds may for a referenrial or expressive vocabulary. She hyporhesized rhat some babies organize
have far fewer (Fen so n, Dale, Reznick, Bates, Thai, & Pethick, 1994). their wo rld around objeccs, whereas others focus on people. Children's differing hy­
The kinds of sryle or srraregy differences rhat we have desctibed, however, are nor potheses about what language is for (to org;anize and categorize objects or ro talk about
simply rhe result of observing chil dren with differing rares of development. Children self and others) derive fro m rhese differingibrganizations of experience. Mothers of ref­
with relatively more referential or expressive speech achieve language milestones at rhe erential children more ofcen reponed rhar their children favored manipulative roys,
same age (Hampson & Nelson, 1993; Nelson , 1973; Pine & Lieven, 1990), and there supporcing rhe notion that preexistent cognitive differences may influence children's
appears co be no clear advantage for eirher sryle when correlared with later vocabulary speech sryle.
or grammatical measu res (Bares et aI. , 1993; Hampson & Nelso n , 1993). Subsequent srudies of children's language and play have not found consisrent
As we noted ea rlier, children may be differentially sensitive co the prosodic runes suppOrt fot linguistic differences rhat map OntO objecr and social preferences. On the
rhar unify whole phrases or co rhe syllables and segments rhat make up single words. one hand, there is evidence rhar children who acquire more referential speech are more
The source of these differences may be found in developmental asymmeuies in rhe artentive to tOys (Rosenblatt, 1977) and excel at object manipulatio n and spatial con­
multiple mechanisms that supporr language acquisition. Thar is, differences in the rare srructions (Wolf & Gatdner, 1979). Similarly, children wirh more noun + noun com­
at which arrenrion, perceprion, and memory marure and are available co parse, store, binations have higher levels of performance on objecr caregorization tasks (Shore,
and analyze rhe inpur srream may affecr rhe size and form of rhe linguistic unics chil­ Dixon, & Bauer, 1995). Children wirh more exptessive speech have been found to ori­
dren produce. Bares and colleagues (1993) and orhers have nored rhat younger learn­ ent more coward adults (Rosenblarr, 1977) and to engage in more social-symbolic (e.g.,
ers musr approach rhe problem of language acquisirion wirh less memory capaciry and puppets and tOy telephones) play (\Vo lf & Gardner, 1979).
fewer analyric skills rhan older learners, often producing forms rhey do nor full y un­ Goldfield (1985/86, 1987), however, suggests rhar episddes of shared attention to
demand. On rhe orhe r hand, rhere is rhe possibiliry rhar some children are more cau­ objecrs may conuibure more ro the acquisition of referentialianguage rhan the sheer
rious, apprehensive "bout making misrakes , and disinclined ro ralk-in shorr, shy. Shy quanriry of object or social behavior. Goldfield observed rwelve children during play
children have been shown ro ralk less and less complexly in a way rhar ren1ains srable sessions in the home at rwelve, fifteen , a nd eighreen monrhs of age. Mothers kept a
rhrough kindergarren and first grade (Evans, 1993). Ho rgan (1981) also suggesrs rhar diary record of the children's firsr fifty words. Children who acquired relatively more
differences berween yo unger and older language learn ers may be a funcrion of rheir nominals did nor differ from their less referential peers on measures of time in tOy play
rendency ro be clurious or ro rake risks. Fifreen pairs of children were marched on or frequency of social behav iors. Children with more nominals, however, mote often
MLU bur differed in age by ar leas r six months. The fasrer (younger) learners rended initiated episodes of joint attention to objects by showing, giving, or bringing roys [0
ro use more nouns and more complex noun phrases. Th ese children also rended co their mother. Children wirh more expressive speech, on rhe other hand, were not nec­
make more grammarical e rrors. Th e slower (o lder) learners used fewer and less elabo­ essarily less interesred in obj ecrs nor more sociable rhan their peers. These children were
rare noun phrases bu r more and more complex verb phrases. Th ese same children, more likely to interrupt or leave their roy play to seek social attention, rather [han ro
however, were more advanced on co mpre hension rasks. share or show a tOy. Differences in the use of objects ro mediate social interaction m ay,
Horgan suggesrs rhar rhe slower children were more caurious language learners, in rum , influence the language parents address to children. Child pointing rypically
wirh good receprive abiliries bLlr a more guarded approach ro displaying rhei r verbal elicits a maternal labd (Masur, 198 2). an d children who more frequently point to ob­
skill s. They may also have arrended more co rh e derails of language s[(uc rure, as evi­ jects acquire more nouns (Goldfield, 1990) .
den ced by rheir use of more auxiliaries and more kinds of cons[[ucrions. Th e faster
children. wirh rhei r more Frequent errors, were "more willing ro rake risks" (Horgan,
Input Factors
1981, p. 636). especially wirh rhe finer poinrs of grammarical sr[llcrure. Similar dif­
ferences in rhe rendency to take risks wirh new syntactic forms verSLlS ro proceed cau­ Children learn to talk in rhe COLlrse of their in[eracrions with any number of conver­
tiously Llnril masrer y is achieved have bee n found for children observed by Kuczaj and sario nal parrners, including parents, day care providers, siblings. and pee rs. Each so­
MaratSos (I983) and by Ramer (1976). cial contact provides a unique source of language variance. lr is likely that Crystal
We do nor yet have the longitudinal dara rhar would p rovide the basis for claim­ learned some phrasal speech from her older brother (e.g., femme see, you know what),
ing rhat slow staners at language acquisition rum our ro be the children called shy, but and other researchers have noted rhat brerborns tend to acquire more phrasal speech
308 Chapter Eight Individual Diffirmces Sources ofVariation 309

(Nelson et aI., 1985; Pine, 1995). Day care offers even more varied language models. in physical ac tivities and social games. The latter ch ild's early utterances were dom i­
Most of the availab le research on individual differences has focused o n the speech of nated by requests for and expressions of these games and routines.
mothers who have been the children's primary caregivets. The extent to which characteristics of me input language influence the course of
Although adult speech to ch ildren shares many featu res, there are also clear dif­ acquisition has teceived mixed empirical suppOrt and continues to provoke th eoretical
ferences in how parents talk to children and encourage their children to talk. Stable dif­ debate. Recently, Hampson and Nelson (199 3) demonstrated a relationship between
ferences have been noted in maternal conversational style, including mothers' preferred maternal language at thirteen monms and ch ild grammar at twenty months when chil­
use of language to direct behavior, el icit conversation, or instruct their children (O lsen­ dren were grouped on the basis of tbe ir laqguage style. They fou nd a significant, pos­
Fulero, 1982). Moreover, at least some aspects of maternal style are related to variation itive telationship between nouns in mate~aI speech and child MLU only fot those
in child speech. children with more than 40 percent n ouns in their spontaneous speech. No such rela­
Fewer nouns, more social expressions, and more verbs are related to maternal tionsh ip was found for children wich more expressive speech (40 percent or fewet
speech that refers to persons o r that directs the child's behaviot in some way (Della nouns). The two groups had simil ar numbers of early and late talkers at thirteen
Corre, Benedict, & Klein, 1983; Furrow & N elson, 1984; Goldfield, 1985/86, 1987; months and did not differ on MLU at twenty months. Thus, children with diffeting
Nelson, 1973; Pine, 1994). More nouns, on the other hand, are associated with ma­ approaches to learning to talk may make differential use of selected aspects of the
ternal utterances that refer to and describe objects and that request and reinforce names input. This sugges ts that it is crucial to consider individual differences, and the extent
for things (Brown, 1973; Della Corte et aI .• 1983; Furrow & Nelson, 1984; Goldfield, to whicb child strategy and caregiver speech style overlap, in studies that examine the
1985/86, 1987; Hampson, 1988; Nelson & Bonvillian, 1972). A1thougb the simple efficacy of the inpUt fo r acquisitio n.
frequency of nouns in maternal talk is unrelated to chi ldren's speech style (Furrow & Some of the input diffetences we have described have also been noted to vary
Nelson, 1984; Nelso n, 1973), maternal descriptions that include nouns are related to along with family socioeconom ic status (SES), especially when SES is indexed by level
the proportion of nouns in tbe tlrst fifty words (Pine, 1994). It is important, then, to of maternal education. That is, parents from high SES families tend to talk to their
consider specific linguistic forms, such as nouns. in relation to tbe pragmatic focus children using more object labels and tewer directives than parents from low SES fam­
(what parents talk about and for what purpose) of parental speech. ilies (HoffcG insberg, 1991 , 1998; Lawrence & Shipley, 1996):' When compounded by
Correlations, of course, cannot tell us the extent to which parental speecb affects dramatic differences in the amount of talk directed to children (e.g., Hart & Ris ley,
chilJ language. or vice versa. They do suggest, though, that mothers and chi ldre n may 1995), less talk and fewer labels may disadvantage low SES children on measu res of vo­
be seeking OUt different opportunities for interaction and conversation. A good deal of cabulary development and larer success on academic tasks tbat depend on this skil l.
chilJren's referential bngu:tge. for example, may origina te in certain routinized nam­
ing games. Dore (1974) found that most of tbe labeling and repeating of the code-ori­
ented baby he observed occurred in verbal routines established by the caregiver: Linguistic Factors
M's nlOrher set up rourines in which she wou ld pick up one item , label it, and encour­
Chi ldren may starr learning language with thei t own preferences and tendencies, and
:tge her daughter co imi(;tce (h e label. There were Jn imal.naming routines . .. ucensil­
caregivers may emphasize certain aspects of language or provide richet input abOU t
naming and people-naming rourines also occurred Frequently. (p. 348)
some domains than others. In add iri on, [hough, languages diffet from one anoth er in
Nelson (1973) also observed that 28 percent of tbe tlrst fifty words acquired by refer­ the problems they pose to the learner. These differences may interact with lear net and
ential chi ldren referreJ to body partS-al most surely learn ed in this kind of rou tine­ input factors to exaggerate individual differences.
whereas none of tbe expressive child ren had acqu ired hbels for pares of the body. Each language can be seen as, in effect. exploiting in its own particular way the
Expressive chi ldren , on the otber hand, learn many conventional socia l expressions capacities for elabo ration, generalization. and rule learning that human beings possess.
(e.g.• hi. bye, please. thtlnk YOlt, lets go. and oh dear) tha t typically mark events such as Both prosodic tun e and segmentation in ro words are relatively accessible to English
arrivals, Jepatwres, 'lnd excha nges. Mothers of ch ildren with rJlore expressi ve speech language learners. Slavic languages such as Po lish, h owever, challenge speakers to learn
tend to use many such stereotypical utterances (Lieven. 1980; Nelson. 1973; Plunkett, a couple of dozen noun endings, including markings for six cases that in singu lar bave
1993). Urwin (1978) observed tbat tbe parents of [WO visually handicapped ch ildren distinct forms for.masculine. fem ini ne. and neuter (Smoczynska, 1985). This pa[tern­
differed in the activi ties they organized and the language they used with their children. ing is made more complex by its synme ti c character, tbat is, that a particular suffix tha t
The parents of one chilJ utilized his limited vision by encou ragin g attention to and la­ reflects [he syn thesis of, fot example, masculine , singular, and genitive might have no
beling of objects, wbereas the parents of a tOtally blind child more often engaged bim phonetic tela[ion ro the fe minine singular genitive or the O)asculin e singular dative.
!. ",.--.--_.

310 Chapter Eight I"dividunl Differences Context: The Interaction ofChild, Caregiver, and Language 311

Since no uns in Slavic are never produced wirhour suffixed case and gender markings, nouns man English learners (Choi, 2000; Gopnik & Choi, 1991, 1995; Kim, Mc­
ir may be almosr inevirable rhar children choose a risk-raking srraregy rhar leads ro Gregor, & Thompson, 2000); rhus, highly referenrial children acquiring rhese lan­
many errors, since rhey can nor learn rhe entire paradigm insranriy. Turkish, on rhe guages may have absolu re levels of nouns much lower rhan relarively expressive
orher hand, also is characrerized by many suffixes, bur rhey are agglucinated rarher rhan children learning English. Iralian shares some characterisrics of pro-drop languages
syntheric, rhar is, added on in a predicrable sequence. Turkish learners may benefir mar highlighr verbs, bur Iralian verbs are more morphologically complex and variable,
from a rendency ro use a prosodic srraregy rhar incorporares dummy syllables for rhe and, unlike Mandarin, Iralian is nor a null- objecr language. Iralian-speaking morhers
affixes nor yer learned, as rhar srraregy creares precisely rhe slors inro which rhe ro-be­ use more verb rypes and cokens rhan noun rypes and rokens (Camaioni & Longobardi,
learned marerial musr fir. Thus, rhe use of fillers and dummy syllables may be more 2001). Overall, rhe inpur co Iralian childreh is more like English man Mandarin, and
common in learners of some languages rhan orhers, as well as varying across learners like English learners, children learning Lralian pcoduce more nouns bur fewer verbs
of rhe same language (Perers, 2001) . man Mandarin speakers (Bassano, 2000; Tardif, Shaa, & Naigles, 1997).
Unfo[(unarely, rhe dara on acquisirion of languages orher rhan English are rela­
rively sparse, lacking in rhe large sample srudies rhar make seeking individual diFFer­
ences Feasible. Borh word and rune babies have been observed For Danish (Plunkerr,
1993), German (Srern & Srern, 1928), and Norv-tegian (S imonsen, 1990). Holisric,
Context: The Interaction of
prosodic learners are also Found in Hebrew (Berman, 1985), Hungarian (MacWhin­ Child, Caregiver, and Language
ney, 1985), French (E. V. Clark, 1985), Iralia n (Cipriani, Chilosi, Borrari, & Poli,
1990), and Porrugese (Scarpa, 1990). As was rhe case For Perers' (1977) subjecr, Minh, As we have seen, borh child and caregiver facrors predin diFferences in children's ap­
rhe German rune baby observed by Elsen (1996) used some reFerenrial speech during proach co the problems rhar varied languages present co rhem. The challenge for re­
naming rourines and when looking ar picrure books. Many languages use intonarion search is co undersrand how available lea rning mechanisms inreracr wirh
ro signal diFFerences in meaning (e.g., rhe diFFerence berween a sraremenr and a yes-no environmenral supporrs. One approach is ro consider rhat language is learned and used
quesri on), suggesring rhar rhis is an individual sryle rhar is common and perhaps as parr of a myriad of conrexrs rhar make up rhe daily life ofi:hildrcn and their con­
equally comparible wirh all languages. The rendency of some children learning Eng­ ve rsarional parmers. As Nelson (1981) has observed, rhe co nrex r in which language is
lish ro sci2e upon whole words rarher rhan on morphological modificarions of words used will derermine the form and funcrion of rhe input. Episodes of joint objeCt atten­
raises rhe quesrion whar such children would do iF rhey were learning !nukrirur or He­ tion, for example, are associared wirh more child lab els and more marernal comments
brew, where "words" as such can hardly be idenriflcd rhrough rh e massive morpho­ (Tomasello & Farrar, 1986; Tomasello &Todd, 1983). Book reading may be a parric­
logical changes rhar eve ry shiFr in meaning imposes. Cross-linguisric analyses have also ularly eFFecrive conrexr for acquiring obj ecr labels (Ninio, 1980; Ninio & Bwner,
yielded more diversified parrerns rhan rhose identified For English speakers. Three par­ 1978). Orher siruarions in rhe child's life-earing, dress ing, playing wirh siblings and
rerns (emphasizing nouns versus phrases and fillers versus verbs, adjecrives, and gram­ peers, playing wirh roys, rough-and-rumble playing, and lisrening co and singing nurs­
marical words) have been observed For French language learners ar rwenry mOl1rhs of ery rhymes and songs-provide quire differenr conrexrs For inpur and acq uisirion. Each
age (Bassano, Maillochon, & Ern e, 1998) . co nrexr, rhen, provides a unique opporru niry ro learn some aspen of language: whole
Lan guages also diFFer in rhe ease wirh which children may exrracr specific pam words or phrases, objecr labels or words for actions and srares, labeling or demandin g,
ofspeech such as nouns or verbs. [n English. nouns may be more salienr rh an verbs be­ prosodic or segmental accuracy. Thus, as rhe range of conrex rs varies, opporruniries for
cause children hear rhem ar rhe beginning and end oFSVO senrences and nouns occur language learning will diFFer for ind ivid ual children.
wirh Fewer grammarical inflecrions rhan verbs; pragmaric Facrors also favor nouns over The inreresrs of borh child al1d caregiver, moreover, will influence rhe kinds of
verbs in producrion (Goldfield. 1993, 1998.2000). Japanese, Korean, and Mandarin conrexts rhar make up rhe daily evenrs and rourines of a parricular morher-child pair.
Chinese. on rh e orher hand, are languages rhar hi ghlighr verbs by Frequenrly delering Go ldfi eld (1987) Found rhar children's lexical diFFerences were besr predicred by a com­
nom inal reFerenrs and posirioning verbs ar rhe ends of se nrences (Clancy, 1985; binarion ofchild and caregiver variab les. Johanna was rhe mosr reFerenrial ch ild in rhis
Rispoli, 1989; Tanouye, 1979; TardiF. 1996). These inreriinguisric diFFerences appear sample. She gave clear evidence rhat shared arrenrion ro objecrs was a familiar and en­
ro have consequences For language learners. As we have seen, children learn in g English joyable inreracr ive·con rext. During rhe play sessions, alm osr half (48 percenr) of her
diFFer in rheir emphasis on nouns in rhe firsr fiFry words; neverrheless, noun learning ar remprs to engage her morher involved showing or giving a toy. Moreover, her
is a prominenr aspecr of vocabubry acquisirion. and rhe proporrion of nouns rypically morher's speech clearly and consistcnriy sllpporred the exrracrion and acquisi ri on of
increases berween fiFry and rwo hundred words (Bares er al ., 1993). [n conrrasr, Japan­ single words rhar named objecrs. Talk abour roys was rhe largesr \caregory of marernal
ese. Korean . and Mandarin learners rypically srarr wirh many more verbs and fewer speech (41 percenr), and her morher highlighred names For rhings during all rypes of
Implications ofIndividual Differences for a Theory ofLanguage Acquisition 313
312 Chapter Eight Individual Differences

Children such as Johan na, whose early language consists of clearly aniculated sin­
play, from book reading ("egg in the hole book," "Look! See the uee?") ro ball games
gle words rhat are predominantly nouns, are well represented in the bulk of language
("Ayy it's a ball") and pretend play (" Here's a woman ," "Yo u can put the woman in the
development research. Cross-linguistic studies, however, reveal that nonnominals fig­
truck") . Nouns made up 76 percent ofJohanna's first fifty words, and only one of these
ure more pro minently in the early speech of children learning languages orher than
was a phrase (get you). English. Italian children appear ro learn fewer nouns than their English-speaking peers
Other children may experience relatively more contexts in which the focus is on
(Camaioni & Longobardi, 1995), and Japanese, Mandarin, and Ko rean speakers are
the child's behavior, performance, or non-roy play. Caitlin, the child with highly ex­
precocious verb learners (Clancy, 1985; Gopnik & Choi, 1991,1995; Tardif, 1996).
pressive speech, included a roy in only 18 percent of h er social iniciations. She was
Because English speakers (parentS, child(en, and researchers) value nouns, much of
more likely ro pause in her play ro look and smile at her mother. Caitlin's mother,
our research effom have been li mired ro undemanding the principles that govern the
mo reover, ofte n engaged her baby in social play, using conversational formulae and
acqu isition of obj ect labels (see, fo r example, Golinkoff, Mervis, & Hirsh-Pasek, 1994;
routines more than any other mother in the sample. Almost half (48 percent) of her
Markman, 1989; Mervis, G olinkoff, & Bemand, 1994). The emphasis on nouns may
unera nces were questions and directives used ro prompt her daughter's performance
be misleading. For example, unlike nouns, nonostensive rather rhan ostensive COntextS
and ro engage her participation in shared play. Sixty-o ne percent of Caitlin's first fifty
appear ro be more conducive ro the acquisirion of verbs (Gold fiel d, 1998; Tomasello
wo rds consisted of expressive speech, and many of these, as we pointed out earlier,
& Kruger, 199 2) . By limiting our explanarions o f word learning ro nouns, we may be
were phrases. m issing many of the cogniti ve a nd linguistic resources thar children bring ro the task
Most children, however, are likely ro learn a more balanced mix o f nouns,
of learn ing ro ralk.
phrases, and expressive speech, acquired through participatio n in a range of contexts.
Expressive speakers such as Ca itlin, who acquire numerous phrases, ap pear less
Peters ( 1983) observed that Minh's use of analytic and gestalt language was often tied
frequently in the literarure. H owever, rhe early use o f phrasal unerances may be more
ro specific contexts. One-word unerances were likely in siruations such as naming pic­
common rhan has been previousl y acknowledged . Longer, expressive phrases occur
rures in a book, whereas gestal t language was often copied from so ngs and srorybook
rhroughout the one-word period (Branigan, 1977; Lieve n et aI., 1992; Srokes &
rhymes. Holden, 1980) . Lieven and colleagues suggest that research' would reveal m ore non­
referential ch ildren if samples were more often expanded ro include a wider range of
social backgrounds. Five out of rwelve children in their sample of bmilies from varied
Implications of Individual Differences
socioeco no mic gro ups had rwenty-fi ve or more phrases in rheir first one hundred
wo rds. Anenrion ro phrasal speech has resu lted in improved methodolo gies and crite­
for a Theory of Language Acquisition
ri a for derermining the lengt h and p rod uc rivi ty of children's linguistic units (Lieven et
aI. , 1997; Plunkerr, 1993). Connectionist models have also revealed neutal networks
The fac t of individual differences in language acquisition has implications for theories
rhar seg ment units larger rhan single wo rds from connected "speech. " These models
and methods in th e srudy of language develo pmenr. Early studies in the modern era
demonstrate how dramatic differences in outpur may emerge from small differences in
of language research assumed that unive rsal as pects of acquisitio n were the proper ob­
rhe learner (i.e., network) a nd/or in rhe inp ut (Elman , 1990; Redingron & Chater,
ject o f study and that phenomena that varied across children were trivial. Thus, small
1998) . C on nec tionisr modeLI also o ffer a potentially valuable merhodological rool for
sample srudies were the norm, and no anempt was made ro select children who rep­
explo ring the exrent ro which a single learning mechanism can account for the kinds
resented the range of possible appro aches ro acqu isitio n. We see now that individual
of vari abi li ty we have desc ribed within and across domains (e.g., the lexicon vs. syn­
diffe rences can tell us about the processes by which children ex uact information from
t'LX) rhat have been characrerized as disrincr language modules. The facr of individual
the linguistic interactions in which they panicipate. Assessing the extent and type of
differences in semanrics, morphology, and synrax and the continuity observed from
indi vidual difFerences helps us construct a theoty of how children learn , rather than just
first words to early se nrences argue against a str ictly modular explanation of language
a description of what they know. Mo reover, emphasis on the srudy of English learners
developmenr.
has limited the data we repon and the theories we const('ucr. English is a language with
Interesr in children's lise of phrasal speech is especially imporranr in view of the
relatively impoverished morphology and littl e opportuniry ro exploit word-order vari­
fact rhat a sign ificant proporri on of everyday adulr speech may consist of phrases s{Qred
ation in simple sentences. Thus, the range of normal var tJ tion in approaches ro lan­
and retr ieved as a whole. These form ulae co nsi st of idioms (e. g., kick the bucket), col­
guage may be limtted by the characte ristics of English, and we are in danger of thinking
locatio ns (e.g. , sheer/pure coincidence) , adj uncts (e.g., by and large), $entence fra mes
that characteristics of English learne rs (e .g., a tendency to srarr with no uns) are in fact
(e.g., please pass the x), and standard si tuational urrerances (e.g., can I help you') (Becker,
universal for all language learn ers.
314 Chapter Eight Individual Differences Suggested Projects 315

1975; Naninger & DeCarrico, 1992; Pawley & Snyder, 1983). Wray (1998) argues We are leli: with me quesrion of where me differences originate. It has been sug­
rhat formulae perform important sociolinguistic functions that ease the burden of con­ gested mat children's varying approaches ro language and orner cognitive problem areas
strucri ng utterances from scratch each rime we have some ming ro say, especial ly in the reflect basic temperamental differences-for example, in risk-taking tendencies~r in
course ofcommunicatively predictable siruations. To date, syntactic theory has focused asymmetries in how information is processed. Such hypOtheses await further, m ore in­
almost entirely on rule-generated construcrions and has neglecred ro address the terdisciplinary investigarions ro tesr mem adequately. H ardy-Brown (1983), for ex­
processes by which children acquire, srore, and produce me full range of communica­ ample, suggests tha t we employ research designs from me fiel d of behavioral genetics
tively effective utterances. ro disentangle the effects ofherediry and enyironment on individual differences in rate
Methods for collecting data on language acquisition have been developed on the and sryle of acquisition. These merhodolo~es would include adoption studies, which
assumption that all children go about it much the same way. Thus, many analyses of assess the cognitive and linguistic abilities of bo th birth and adop tive parents and com­
children's spontaneous speech are based on utterances elicited in the contex t of play pare these measures to the child's developing linguistic skills. Meanwhile, the fact of
with roys, usually a set of novel toys provided by the experimenter. Children who rely individual differences has implications not only for theory, but for research a nd edllca­
on imitation and routine co ntexts as sources of their utterances may be relatively dis­ tional practice as well. We can apply what we know about individual differences to
advantaged in this novel situation and need ro be observed during more familiar daily ame nd and improve ou r methods of collecting language data, intervening with children
events. Similarly, phrasal speech may be more consistently elicited during interactions at risk for language del ay or deviance, an d teaching reading and foreign la nguages.
with an older sibling. Finally, the recognition that there are many ways to learn a language and that
lr is also important to note that different culrures vary in the degree ro which they normally developing children may differ from one another in how they accomplish the
encourage and support va rious child tendencies. The highly referential child is ap pto­ task should help us to th ink more crea tively abOUt therapy, intervention, and edllcation.
priate and rewarding ro a middle-class American mother w ho sees naming as a sensi­ A sin gle therapeutic or educational method is unlikely to work for all child ren, and the
ble and intelli ge nt way ro use language, but nOt to a Kaluli mother, who would view failure of one method does not imply [hat success is impossible. Th'e delayed or lan­
naming as "talking ro no purpose" (Schieffelin, 1986). The skill of imitation may be guage-handicapped child, like the normally developing ch ild , m ay exploit or avoid im­
relatively littl e valued by American mothers, b ut it is crucial for children whose care­ ita ti on, may search cautiously for rules or recklessly try our utrerances, may be mo re
give rs instruct th em to repeat modeled utterances as a way both ro learn and ro func­ easily involved in social games, or may demand a referent ial vocabulary. All of these
tion socially, as do Kaluli (Schi effd in, 1986). Kwara'ae (Watso n-G egeo & Gegeo, preferences are compatible with successful language acquisiti on, and all can be utilized
1986), and Basotho (De muth , 1986) mothers. The e:<istence of culwral variation in by therapists, teachers, and parents.
the language learning environment has been proposed by many as an argument against
a strong environmental influence on language acquisition; we see va ri ation instead as
a fact that must be undersrood, in the light of information from sruciies of individual Key Words
differences, as evidence that children have many mech'1I1isms avai lable for th e acq ui­
sition of language that are differentially exploited in diFferent cul rural and linguistic agglurinated pronominal straregy

contex ts . analytic style referential

code oriented
rote/holistic sryle

expressive
synthetic

Summary message oriented


transfo rmational syntax

nominal strategy

It is important ro reiterate that even though individual differences in styles Ot strate­


gies for bnguage acquisition are striking, the d ifferences obse rved may reflect prefer­
ences or tendencies rathe r than dichoromies. Child ren who are classified as highly
imitative p roduce many nonimitated utterances. Pronominal chil dren produce some
nominal word combinations. Reftrential chil d ren a re not incapable of socially expm­
Suggested Projects
sive speech. Language acquisition is a remat kably buffered process with a high rate of
success; clearly, most ch ildren control many different strategies and mechanisms that 1. Analyze the speech of one child with two different ad ults, one familiar and o ne Lln­
contribute ro language development. familiar, in a variery of si tuations, ro determine how much incrachild variabiliry
316 Chapter Eight Individ"al Differe1lces References 317

occurs on percentage of imitative unerances, use of pronouns, and use of object­ Bates, E., Marchman, V, T hal , D ., Fenson, L., Dale. P., Reznick,]. S., Reilly,]., & Hartung.
versus social-oriented unerances. J.(1993). Developmenral and stylisric variarion in rhe composirion of early vocabulary. jour­
nal ofChild Language, 21,85-123.
2. Find parents of three children aged fourteen to eighteen months. Administer the
Becker, ] . (1975). The phrasallo:icon. BBN Reporr no. 308 1, AI Repon no. 28. Cambridge,
MacArthur Co mmunicative Development Inventory to the parents, and interview MA: Bolr, Baranek, & Newman.
them about their impressions of their children's language learning style, to deter­
BelJugi, U. (1967). The acquisition of negation. Unpublished docroral disse([Jrion, H arvard
mine, for example, if the children sta rted to talk using a lot of jargon babbling or Un iversity. ,
with clear words, if th ey were cautious comprehenders or reckless producers with
Benedicr, H . (1979). Early lexical developme~ r: Co mprehension and production. journal of
lo ts of errors, or if they were highly or minimally imitative. Use the look-up tables Child Language, 6, 183-200.
in Bates and colleagues (1993) to see where the children score on referentiality. If
Berman. R. A. (19 85). T he acquisirion of Hebrew. In D. I. Slobin (Ed.), The cross linguistic study
time permits, tape and transcribe about twenty minutes of parent-child interac­
a/language acquisition, Vol. I: The data. Hillsdale, N]: Erlbaum.
tion (enough to get about 100 child unerances) and anal)'2e the ch ild talk for the
Bloom, l. (1993). The transition from infancy to language: Acquiring the power ofexpression.
dimensions of individual differences discussed in this chapter.
Cambridge, UK: Cambrid ge Un iversity Press.
3. Record the interactions of three different care giver-child pairs (children should be Bloom, l., Lighrbown, P., & Hood, L. (1975). Srrucrure and variarion in child language. Mono­
about the same age andlo r vocabulary level). Code parental speech for its com­ graphs ofthe Sociery for Research in Child Development, 40.
municative intent (e.g., descriptions versus behavioral directives as used by Pine Bowerman , M. (1973). Early syntaClic d.-velopmmt: A cross-linguistic study with special referena
(1994). Do the parents differ on these dimensions' to Finnish. London: Camb ridge Uni versity Press.
Bowerman, M. (1976). Semanr ic facrors in rhe acqu isirion of rules for word use and se nten ce
consrrucrion. [n D. Morehead & A. Morehead (Ed s.). Directions in normal and deficient cl'ild
Suggested Readings la Nguage. Bal ri more: University Park Press.
Bares, E. , Brerhercon, I., & Snyder, L. (1988) . Fromfi", 1V0rds to grammar: Individual diffir­ Braine, M. D. S. (1963). The onrogeny of English phrase srr ucrure:'The firsr phase. Language.
mees and diJSociablc mechanisms. Cambrid ge, UK: Ca mbridge University Press.
39,1- 13.
Lieven, E. V. M. ( 19%). Va riar ion in a crossli nguistic co ntexL In D. Slobi n (Ed .), The crosslin­
Brani gan , G . (1977, Seprember 30). If rI,is kid is in the one-word period, how come hes saying

guistic mtdy ofl'llIguagc acquisition. Volume 5: Exp'I!Iding rile con".\·ts. Hillsdale. NJ: Erlbaum .
whole ,..nterzces' Paper presenred ar rhe 2nd Ann ual Bosro n Univers ity Conference on Language

Deveiopm enr, Bosron.

Ne lson. K. (198 1). Indiv id ual d ifferences in langu:lge devciopmenr: Im pli c:\[io ns fo r develop­
mcnr and langu:lge. Devdopmmtill hychology, 11, 170-1 87.
Brerherron, I. , McNew, S., Snyder, l., & Bares, E. (J 983). Individual differences ar 20 monrh"

Perers, A. (1983). The II"its of language ({cqll/jition. Cunbridge. UK : Camb ridge University
Analyric and holisric srraregies in hnguage Jcquisirion.joumal ofChild Language, 10,293-320.

Press.
Brown, R. (1973). A first language. Cambridge. MA: Harvard Uni versity Press.

Sho re, C. 1'v1. (1995). fn dividual diffirences in language developm""t. Thou sa nd Oaks, CA: Sage.
Buss is , A. M., ChirrenJen , E. A. , Amarel, M., & Klausner, E. ( 1984) . Inquiry into meaning: An

irwwigation oflearning to read Hil lsdale . NJ: ErIbaum.

Ca maioni , l., & Lon gobardi, E. (1995). Nawre and srability of indi viduJI differences in ea rl y

References lexical developmenr of Irali an speaking children. Fim Langl/age. 15, 203 - 218.

Adamson, l. B.. & Tom :lsello,. M. (1984). An "exp ressive" chil d's langu :lge develo pmen t. In­ Camaioni, l. , & Longobardi . E. (200 I). Noun versus verb emphasis in Irabn morher-ro-child

jam Behavior ({nd Development. 1, 4. speech. journal afChild Langl/age, 28. 773- 785 .

Bassano, D. (2000). Early develop mcn r of nouns and verbs in French: Exploring rhe inrerfuce Cazden , C. (! 968). The acquis irion of noun and ve rb intl ecrioll s. Child Development, 39,

berween lex icon and gram mar. jOllrnal ofChild Language, 21, 521 -559.
433-438.

Bassano, D .. Maillochon, I., & Em e, E. (1998). Developmenral cha nges and v:triabi liry in rhe
Childers, J. B. & Tomasello. M. (200 I). The role of pronouns in young chi ld ren's acq ui sirio n
early lex icon: A study of French child ren's nawra lisric producrioll s. journfll ofChild Langl/age,
of rhe English rransirive co nsrrucrion . Developmental Psy"hology, 31,739-748.

25,493-53 1.
Choi, S. (2000). C3Cegiver inpur in En glis h and Ko rean: Use of nouns and verbs in book- read­

Bares, E., Brctherron, l., & Snyder. l. (1988). From first words to grammar: fndivid"al diffir­
ing and roy-play COflre xrs. joumal ofChild Language, 21, 69- 96.

en"es and dissociable mechan isms. C ,mbridge, UK : Ca mbridge Unive rsiry Press.
Chomsky. N. (1957). Syntactic scraCilires. The Hague: Mouron.
..........-~

318 Chapter Eight Individual Differences References 319

Goldfoeld, B. A. (1990) . Pointing, naming, and talk about objects: Referential behavior in chil­
Cipriani, P., Chilosi, A. M., Borrari, P., & Poli, P. (1990, July). Some data on tramitional phe­
nomena in the acquisition ofItalian. Paper presemed a[ the 5[n Imernaci onal Congress for the dren and mo [hers. First Language, 10, 231-242.
Study of C hild Language, BudapesL . Goldfield, B. A. ( 1993). Noun bias in maternal speech ro one-year-olds. journal o/Child Lan­
Clancy, P. M. (1985). The acquisition ofJapanese. In D. I. Slobin (Ed .). The cro;s-linguistic study guage, 20, 85-99.
o/langllage acquisition, Vol. 1: The data. Hillsdale, NJ : Erlbaum.
Goldfield , B. A. (1998). Why nouns before verbs? The view f[om pragmatics. Proceeding> ofrhe
Clask, E. V. (1985). The acquisi[ion of Romance, with special reference to French. In D. J. Slobin
Bouon Univmity Conference on Language Developmenr. Somerville, MA: Cascadilla Press.
(Ed .), The cro"lingllistic study oflanguage acquisition, Vol. 1: The data. Hillsdale, NJ: Erlbaum .
Goldfield, B. A. (2000). Nouns before verbs in,comp rehension vs. producrion: The view from
Della Coree, M., Benedict, H. , & Klein, D. (1983). The relationship of pragmatic dimensions of
pragmatics. journal o/Child Language, 27, 50It-520.
mothers' speech to the referemial-expressive distinc[ion. journal ofChild Language, 10,35-44.
GoldfIeld, B. A., & Reznick, J. S. (1990). Early lexical acquisi[ion: Rate, content, and ,he vo­
Demu[h, K. (1986). Pro mp[ing routines in [he language socialization of Baso[ho children. In
cabulary spurr. journal o/Child Language, 17, 17 1-183.
B. Schieffelin & E. Ochs (Eds.), Language socialization aero" cultum. New York: Camb ridge
Goldfield, B. A., & Reznick, J. S. (1996). Measuring the vocabulary spurr: A reply to Mervis
Universiry Press.
and Benrand. journal 0/ Child Language, 23, 241-246.
Dore, J. (1974). A pragmatic description of early language developmenL journal o/Psycholin­
Golinkoff, R. M., Mervis, C. B., & Hirsh-Pasek, K. (1994). Early o bject labels: The case for
guiuic Research, 4, 343-35 1.
lexical principles. journal ofChild Language, 21, 125-155.
Echols, C. H. (1993). A percep[ualiy based model of children's eariieS[ produc[ions. Cognition,
Gopnik, A. , & Choi, S. ( 1991) . Do linguis[ic differences lead ro cognitive differences' A cross­
46,245-296.
linguis[ ic s[ud y of semantic and cogniti \'e developmen[. Firu LangZiage, 11, 199-215.
Elman,J. L. (1990). Findingmucwrein [ime. Cogr,itiveScience, 14, 179-211.
Gopnik, A., & Choi , S. (I 995). Names, relacional words, and cognitive developmenr in Eng­
Elsen, H . ( 1996). Two routes to language: Sryliscic variation in one ch ild. First Language, 16,
lish and Korea n speakers: Nouns ace not always learned befote ve rbs. [n M. Tomasello & W. E.
Merriman (Eds.), Beyond names for ching>. Hillsdale, NJ: Erlbaum.
141-158.

Evans, M. A. (1993). Communicacive co mpecence as a dimension of shyness. In K. H. Rubin


Ham pson, J. (1988). Individual differences in sryle of language acquisition in relation to social
& J. S. Asendorpf (Eds .), Social withdrawal, inhibition, and shyness in childhood. Hillsdale, NJ:
ne,works. [n S. Salzinger, J. An((obus, & M . Hammer (Eds .), Social rze<works o/children, ado­
lescents, and college slU.dena. Hi llsdale, ~J: Edbaum .
Erlbaum.

Fenson , L.. Dale, P., Reznick: J. S., Saees, E. , ThaI. D., & Pechick, S. (1994). Variabiliry in early
Hampso n, J., & Nelso n, K. (1993). The relation of matetnallanguage to variation in rare and
communic;t[ive deveiopmenL (Vlonognlphs o/the Society for R"search in Child Development, 59.
sryle of language acquisi,ion. JOl/mal ofCbild Language, 20, 313-342.

Ferguso n. C. A. (1979). Phonology as an individual access sys[em: Some dJta from language
Hard y·l3rown, K. (1983). Universals and individ ual differences: Disentangling tWO approaches
acquisi[ion. [n C. J. Fillmore, D. Kempler, & W. S.-Y. Wang (Eds.), Individual diffirences in
[0[he study of language acquisition . DevelopmenralPsychology, 19, 6 I 0-624.
langllage ability tint/language b,htlvior. New York: Academ ic ['ress.
Harc, B., & Risley, T ( 1995). lvleaningfiti diffirenw in the everyday experience o/young Ameri­
Ferguson. C. A., & Farwell, C. (1975 ). Words an d so unds in early language acq uisition. Lan­
can children. Baltimore: Paul H . Brookes.
grlt/ge, 51, 419-439. Heath , S. B. (1983). Ways wich UJords: Langllllge. lift, and work in comml/nities and citIJsrooms.
Ferrier, L. J. (1975, September). Depmdency and appropriateness in early language development.
Cambridge, UK: Cambridge Universiry Press.
Paper prese,,,,:d It the 3rd Ineernacional Child Language Sym posiu m, London.
Hoff-Ginsberg, E. (1991). Mother-child conversation in different social classes and commu­
Fillmore, L. W. (1979). Individual differences in seco nd language acquisi[ion. [n C. J . Fill­
nicarive settings. Child Development, 62, 782-796.
more, D. K~mpler, & W S. - Y. Wa ng (Eds.), Individual diffirenw in langllage ability and lan­
Hoff-Ginsberg, E. ( 1998) . The relation of birch order and socioeconomic starus ro children's
gllage behllvior. New York: Academi c Press.
language experience and language development. Applied Psycholinglliuies, 19, 603-629.
Furrow, D., & Nelson, K. (1984). Envirollmcmal correlates of individual differences in lan­
Horgan , D . (1981). Rate of language acquisi[ion and noun emphosis. jOllrnal ofPsycholingllis­
guage acqu isi[ion. jourf/fli o/Child Language, 11, 523-534.
tic Research, 10,629- 640.
Goldfield, B. (1982, October 9). Intra-individual variacion: Pauerm ofnominal and pronomi­
Kim, M ., McGregor, K. K.. & Thompson, C. K. (2000). Early lexical develo pment in English
nal combi"fltiom. Paper presenred at [he 7th A.'1nual Boston Universiry Conference on Lan­
a nd Korean-s peaki ng children: Language·general and language-sp ec ific parre",s. Jo urnal of
guage D eveiopmen[, Bosmn.
Child Langlltlge, 21" 225-254
Goldfield, B. (19 85/86). Referential and expressive language: A s[tidy o f cwo mother-child
KJein , H. B. (1978). The relationship b"ween percep","1 )'"aregies and production strategies in
dyads. Fim Langrlflge. 6, I 19-131.
learning rhe pbono!t;gy ofearly lexical ;cem;. Bloomingron: Indiana Universiry Linguistics Club.
Goldfield . B. (19 87). The concributions of child and caregive r to referen rial and expressive lan­
Kuczaj , S. A., & Maratsos, M. P. (1983). The initial verbs of yes- no questions: A differe nt kind
guage. Applied ?'ycholingl/isties, 8, 267-280.
of general grammatical category. DevelopmelllalPsychology, 19,440-443.
r ·~·-
~ - _. _ ,..---",­ _. .

320 Chapter Eight Individual Differences References 321

Lawrence. V W .• & Shipley. E. F. (1996). Parenral speech ro middle and working class children
Ninio, A., & Bruner,]. (J 978). The achievement and antecedents of labeling. journal ofChild
from two racial groups in rhree settings. Applied Psycholinguistics. 17. 233-255.
Language, 5, 1-1 5.
Leonard. L.. Schwam, R., Folger, M .. Newhoff. M ., & Wilcox, M. (1979). Chi ldren's imira­
Olsen-Fulero, L (1982). Sryle and srabiJiry in momer conversarional behavior: A srudy of in­
rions of lexical irems. Child Development, 59, 19-27.
dividual differences. journal ofChild Language, 3, 543-564.
Lieven. E. V M. (19 80). Language develapment in young children. Unpublished docroral dis­
Pawley, A., & Snyder, F. H. ( 1983). Two puzzles for linguisric meory: Narivelike selection and
sertarion, Cambridge Universiry.
narivelike fluency. In J. C. Richards & R. W. Schmidr (Eds.), Language and communication.
New York: Longman.
Lieven, E. V M. (1989). The linguisric implicarions of early and sysremaric variarion in child

language development. Proceedings ofthe Berkeley Linguistic Sociery, 25, 203-214.


Perers, A. M. (1977). Language learning strar4ies: D oes the whole equal me sum of rhe pam'
Language, 53, 560-573.
Lieven. E. V M., Pine,]. M., & Baldwin, C. (1997). Lexi cally based learning and early gram­

marical developmenr. journal ofChild Language, 24, 187- 2 19.


Perers, A. M. (1983). The units oflanguage acquisition. Cambridge, UK: Cambridge Universiry
Press.
Lieven, E. V M., Pine,]. M., & Barnes, H. D. (1992). Individ ual differences in early vocabu­

lary development: Redefining rhe referenrial-expressive disrincrion. journal ofChild Language,


Perers, A. M. (2001). Filler syllables: Whar is rh ei r srarus in emerging grammar' journal of
Child Language, 28. 229- 242.
13,287-3 10.

Macken, M. (1978). Permi([ed comp lexiry in phonological development: One child's acquisi­
Perers, A. M., & Menn. L. (1993). False stam and filler syllables: Ways ro learn grammatical
morphemes. Language, 69, 742-777.
rion of Spanish consonants. Lingua, 44,219-253.

MacWhinney, B. (1985). Hungarian language acquisirion as an exemplificarion of a general


Pine, J. M. (1992a) . The funcrional basis of referentialiry: Evidence from children's spontaneous
model of language developmenr. In D. l. Slobin (Ed.), The crosslinguistic study oflanguage ac­
speech. First Lang"age, 12.39-55.
quisition, Vol. 2: Theoretical issues. Hillsdale, NJ: Erlbaum.
Pine. J. M. (i992b). How referential are "referential" children' Relarionships between marec­
MJrkman , E. M . (1989). Categorization and naming in children: Problems ofinduetion. Cam­
nal-report and observa rional measures of vocabulary composirion and usage. journal ofChild
Language, 19. 75-86 .
bridge, MA: MIT/Bradford.

Mas ur, E. F. (1982). Morhers' responses ro infJnrs' objecr- relared gesrures: Influences on lexi­
Pine, }. M. (1994). Environmental correlares of variarion in lexical Style: Inreracrional sryle and
rhe srtuc(Ure of rhe inpu t. Applied Psycholinguistics, 15, 355- 370.
cal developme nt. joumfll ofChild Language, 9, 23-30.

Mervis, C. B., Golinkoff. R. M., & Be rrrJnd, J. (1994). Two-year-olds re Jdily learn mulriple
Pine, J. M. ( 1995). Va riarion in vocabu lary developmenr as a funcrion ofbirrh order. Child De­
velopment, 66, 272-281.
bbels for rhe same basic-level caregory. Child Development, 65. 1163-1177.

Miller, \Y/., & Ervin , S. (1964). The development of gramma r in child IanguJge. In U. Bell ugi
Pine, }. M .. & Lieven. E. V M. (1 990). Referenrial sryle ar rhirteen monrhs: Why age-defined

& R. Brown (Eds.). Th e acquisirion of language (pp. 9-34) . Monographs of,he Sociery for Re­
cross-secrion:d measures are inappropriare for rhe srudy of straregy differences in early language

development. journal ofChild Language, 17, 625-631.

uarch in Child Development, 29 (Serial No. 92) .

Na rringer, J. R., & DeCJrrico. J. S. (1992). Lexical phrases and It'nguage teaching. Oxford: Ox­
Pine. }. M .• & Lieven, E. V M. (1993). Rea nalyzing rore-Iearned phrases: Individual diffe rences

in rhe transirion ro mulri-word speer.h. journal ofChild Language, 20, 551-571.

ford Universiry Press.

Nelson, K. (1973). Strucrure and strategy in leaflling ro ralk. Monographs oftheSocieryfor Re­
Plunkerr. K. (19 93). Lexical segmen rarion and vocabulary growrh in early language acquisirion.

journal ofChild Language, 20, 43-60.

search in Child Development. 38.

Nelson. K. (1975). The nominal shifr in semanti c-synracric development. Cognitive l's),cholagy.
Ramer, A. L ( 1976). Synracric sryles in emerging language. journal ofChild Langriage, 3, 49-62.

7,46 1-479.
Red ingron. M., & Charer, N . (J 998). Con necrionisr and srarisrical approaches ro language ac­

quisirion: A distriburional perspecrive. Langullge and Cognitive Processes. 13, 129- 191.

Nelson. K. ( 1981) . IndividuJI diffe rences in language development: Implicl(io ns for develop­

menr and language. Developmental Psychology, 17, 170-1 87.


Reznick . J. S .. & Goldfield. B. A. (1992). Rapid change in lexical developmenr in compre­

hension and production. Developmental Psychology, 28, 406-413 .


Nelson , K. E. .• Baker, N .. Denninger. M .• BonvilliJn. J.. & KapLIn, B. (1985) . Cookie verSUS
do-ir-again: Imirarive-referenr ial and personal -soc ial-syn racric-iniriarin g language sryles in
Rispoli , M. (1989). Encounters wirh Japanese verbs: Caregiver sentences and rhe categorizario n

young chi ldren. Linguistics, 23, 433-454.


of transirive and inrransirive acrion ve rbs. FirfC Language, 9, 57-80.

Nelson. K. E. , & Bon villian. J. D. (1972). Concep rs and words in rhe 18-molHh-0Id: Acquir­
Rollins, P. R.. Snow. C. E., & Willert. J. B. ( 1996) . Predicrors of MLU: Seman ric and mOr­

ing concepr names under conrrolled condirions. Cogllition, 2, 435-450.


phological developmenrs. First Langriage, 16, 243- 259.

N inio. A. (1980). Pictllre book reading in morher-infanr dyads belonging ro tWO subgroups in
Rosenblarr, D. (1977) . Developmental trends in infanr play. In B. Tizard & D. Harvey (Eds.) ,

Israel. Child Development. 51, 587- 590.

Biology ofplay. London : William Heineman n Medical Books.

322 Chapter Eight In dividual Differences References 323

Scarpa, E. (19 90). Prosodic strategies fo r che produccion oFlong utterances. Unpublished man­ Tomasello, M. (1992) . First verbr: A cas, rtudy of early grammatical development. Cambridge,
uscript, University of Campinas, Brazi l.
UK: Cambridge University Press.
Sch ieffel in , B. (1986). How Kaluli children learn what to ray, what to do, and how to feel. New
To masello, M., & Farrar, M. J. (1986). Joinr .ttenri on and ea rly language. Child Develllpmtmt,
York: Cambridge University Press.
51. 1454- 1463 .
Shore, C, Dixon, w., & Bauer, P. (1995) . Measures ofli nguiscic and non-linguiscic knowledge To masello, M ., & Kruge r, A. C (1992). Joinr attenrion on act ions: Acquiring ve rbs in oscen­
of objects in the seco nd year. First Language, 15, 189-202.
sive and non oscensi ve CO ntexts. journal ofChild Language, 19,311-333.

Simonsen, H. G. (1990). Child phonology: System and variation in three Norwegian children and
Tomasello, M ., & Todd , J. (1983). Joint attenri on and lexical acquisi tion style. Fim Language,

one Samoan child. Doctoral dissertacion , Departmenc ofLinguiscics and Philosophy, University
4, 197-212. f

of Oslo.
Urwin, C (1978). The de velopment of communication berween bli nd infants and rheir par­
Sio bin , D. I. (1968). Early grammatical develllpment in [eva al languager, with rpecial attention
enrs. In A. Lock (Ed.), Action, gerture, and rymbol. New York: Academ ic Press.

to Soviet mearch (Working Paper No. 11). Berkeley: University of Californ ia, Language-Be­
Vi hman, M. M. (1993) . Variable paths co early word production. journal of Phonecicr, 21,

havior Research Laboratory.


6 1-82.

Slobin, D. 1. (19 85). The crorr-linguistic rwdy of language acquisition, Vol. 1. Hillsdale, NJ: Watson-Gegeo, K, & Gegeo, D. (1986). Calling-out and repea ri ng routines in Kwa ra'ae chil­

Erlbaum.
dren's language socialization. In B. Sch ieffelin & E. Ochs (Eds.), Language rocializarion acroff

Siob in , D . I. (1992) . The crorrlinguirtic rcudy of language acquisition, Vol. 3. Hillsdale, NJ:
clIlmrer. New York: Cambridge Unive rsity Press.

Erlbaum.
Wolf, D., & Gardnet, H. (1979) . Style ,n d sequence in symbolic play. In M. Franklin & N.
Slobin , D. I. (1997). The crorr/inguistic study of language acquirition, Vol. 5: Expanding the Smi th (Eds.). Early rymbolizarion. Hillsdo.le, NJ: Erlbaum.

contexts. Hillsdo.le, NJ: Erlbaum.


W ray, A. (1998). Proco language as a holistic system for social inreraccion. Langllage and Com­
Sm irh, N. (19 73). The acquisition ofphonology: A case smdy. London : Cambridge University
munication, 18, 47-67.

Press.

SmoC1:ynsb, M. (1985). The acquisition of Polish. In D. l. Siobin (Ed.), Thecrorrlinguirtic rcudy

oflanguage acqllirition, Volume 1. Hillsdale, NJ: Erlbaum.

Starr, S. ( 1975). The relatio nship of single words to two-word sen tences. Child Development,

46.701 - 708 .

Ste rn, C, & Ste rn , W. (1928). Die kindcrrpmche. Leipzi g, Germany.

Scokes, W. T, & Holden , S. (1980, Octobe r). Individual patterns in early Lmgl/age development:

Is there a one-word period' Paper presented at che 5th Annual Boston Universicy Conference on
Language Deve lopme nt , Boston .

Tanouye. E. K. (1979). The acquis it ion of verbs in Japan ese child ren . Papar and Reports on

Child Langllage Development, 17 (Department of Linguistics, Stanford Universi ty), 49-56.

Ta rdif, T (1996) . Nouns are not always learned before verbs: Evidence from Mandarin speak­
ers' early vocabularies. Developmental Prychology, 32, 492-504.

Tardif. T .. Gelman , S. A., & Xu, F. ( 1999). Putting the "noun bias" in contexc: A co mparison

of English Jnd Ma ndarin. Child Developmen t. 70, 620-635 .

Tardif, T . Shatz, M., & Naigles, L. ( 1997). Caregive r speech and children's use of nouns versuS

ve rbs: A com pa riso n of Engl ish, Italian. and Mandarin . journal of Child Language. 24,

535-565

Thai, D .. & Bates, E. (\ 988). ReLzrionrhips between language and cognition: Evidence fro m lin­

guistically precocious children. Paper prese nted to the Annual Convention of che American

Speech-language- Hearing Association, Boston.

ThcakstO n, A. L., Lieven, E. M., Pine, J. M ., & Rowland , C F. (200 2). Going, go ing, gone:
The acquisi tion of the ve rb "go." jOl/mal ofChild Language, 29, 783- 8 11.

You might also like