Normative Data for American Sign Language - Journal of Deaf

both the grammatical complexity and organizationalprinciples common to the spoken languages of theworld (Klima & Bellugi 1979; Poizner, Klima, & Bel-lugi, 1987). For example, unlike English, ASL is mor-phologically complex and has been compared in typol-ogy to polysynthetic spoken languages (Bellugi &Klima, 1982). Unfortunately, for the majority of deafchildren, access to either spoken or signed language isextremely limited. Of the approximately 500,000 deafpeople who reside in the United States who havedifficulty utilizing spoken language, 90% to 94% ofthem were born to hearing parents who do not knowsign language, yet the primary language of the Deafpopulation is ASL. Thus, unlike children who acquirethe language of their parents, most deaf children donot initially have access to a signed language. Giventhis atypical language learning environment, deaf chil-dren of hearing families are clearly at risk for languagedelay; the need for early language assessment andnormative data is critical. Currently, relatively little isknown about early lexical and grammatical develop-ment in deaf children of deaf parents, who are acquir-ing ASL as their first language. In response to thisneed, we have developed the MacArthur Commu-nicative Development Inventory for American SignLanguage (ASL-CDI), a parental report based on theEnglish CDI (Fenson et al., 1994) to assess early devel-opment of ASL. This article reports on the develop-ment of this form and preliminary results from the datacollection. We discuss both cross-sectional and longi-tudinal data.

To learn more about normal language development in deafchildren, we have developed the MacArthur CommunicativeDevelopment Inventory for American Sign Language (ASL-CDI), a parent report that measures early sign production.The ASL-CDI is an inventory of sign glosses organized intosemantic categories targeted to assess sign language skills inchildren ages 8 to 36 months. The ASL-CDI uses a recogni-tion format in which parents check off signs that their childproduces. The form has demonstrated excellent reliabilityand validity. To date, normative data have been collectedfrom 69 deaf children with deaf parents who are learning signlanguage as a first language. We discuss the development ofthe ASl-CDI and preliminary cross-sectional and longitudi-nal findings from this early data collection with particular fo-cus on parallels with spoken language acquisition. We alsodiscuss the acquisition of first signs, negation, wh-questions,and fingerspelling with developmental patterns providedbased on age, as well as vocabulary size.

American Sign Language (ASL) is the visual-gesturallanguage used by the Deaf population in the UnitedStates. Passed down from one generation of AmericanDeaf to the next, it is an independent linguistic system,not derived from any spoken language. ASL exhibits

A portion of these data was presented at the 1993 American Speech-Language-Hearing Association Annual Convention and the 1995 Ameri-can Education Research Association. We thank the children and familieswho participated in this research. We also thank Kate Provine and manyundergraduate research assistants who collected the pilot data for thisproject. We appreciate the comments by reviewers of an earlier version ofthis manuscript. This work was supported by NIDCD R29 DC00539 andthe John D. and Catherine T. MacArthur Foundation Research Networkon Early Childhood Transitions. Correspondence should be sent to DianeAnderson, Institute of Human Development, University of California,Berkeley, Berkeley, CA 94720-1950 (e-mail: [email protected]).

� 2002 Oxford University Press

Empirical Articles

The MacArthur Communicative Development Inventory:

Normative Data for American Sign Language

Diane AndersonUniversity of California, Berkeley

Judy ReillySan Diego State University

Current Issues in Language Acquisition ofDeaf Children

Currently, there are no forms available for the evalua-tion of deaf children’s lexical acquisition or early signlanguage development (but see Maller, Singleton, Su-palla, & Wix, 1999, for an instrument appropriate fordeaf children ages 6 to 12 years). The lack of availableresources creates two challenges. First, educators facegreat difficulty in evaluating the language developmentof their deaf students. Second, without the ability tocompare the language competencies among differentchildren, researchers are unable to adequately addressa number of highly debated theoretical issues withindevelopmental psychology, such as the relationship be-tween cognitive abilities and language development.

In our review of the literature, we encountered alimited number of studies discussing the early sign lan-guage development of deaf children (Bonvillian & Fol-ven, 1987; Caselli, 1983; Ellenberger, Moores, & Hoff-meister, 1975; McIntire, 1977; Meier & Newport, 1990;Newport & Meier, 1986; Orlansky & Bonvillian, 1985;Padden & Le Master, 1985; Petitto & Marentette, 1991;Schlesinger & Meadow, 1972). Despite the merits ofthese early studies, their usefulness as a foundation forcurrent research in sign language acquisition is unfor-tunately limited by methodological problems that com-monly occur in research with low-incidence popula-tions. Specifically, many of the early studies in signlanguage acquisition involved single individuals orsmall samples, making it difficult to generalize fromthose reports to the deaf population as a whole. In ad-dition, many of the early works on sign language acqui-sition included both deaf and hearing children in thestudies, confounding monolingual with bilingual lan-guage learning. Furthermore, many studies did notconsider the hearing status of the parents and includedmany deaf children born to hearing parents in theirsamples. Deaf children born to hearing parents are of-ten not exposed to sign language until well beyondtheir third birthdays. By including this atypical lan-guage learning population in studies with deaf childrenof deaf parents, researchers cannot accurately obtain anormal language acquisition profile. In sum, previousstudies have suffered from small sample sizes and het-

84 Journal of Deaf Studies and Deaf Education 7:2 Spring 2002

erogeneous groups with respect to hearing status andlanguage input.

In response to the need to characterize normal lan-guage development in deaf children, while also consid-ering the problems of earlier studies, we have developedthe ASL-CDI. This tool will address both research andclinical needs by providing a standardized form to as-sess deaf children’s linguistic development.

History of the Communicative DevelopmentInventory

The MacArthur Communicative Development Inven-tory (CDI) for English (Fenson et al., 1994) is a paren-tal report language assessment tool that uses a recogni-tion format; parents check the words that their childproduces. There are two forms, one for infants from 8to 16 months and one for toddlers 16 to 30 months.The toddler version of the CDI is designed to assessvocabulary production and early grammar. The CDI isnot intended to be an atlas of all the words a child pro-duces. Rather, it is designed to sample a child’s lan-guage, providing an index of the words the child uses.The data from children at the younger ages may actu-ally represent all words known, but the form is in-tended to sample the child’s vocabulary production sothat it can be compared to normative data from otherchildren the same age.

Although a common criticism of parental reportdata is that parents lack specialized training in as-sessing language, the CDI format has proven to be ex-tremely reliable (internal consistency, r � .96; test-retest, r � .95). Because the CDI uses a recognitionformat, with the words already printed on the page, theperson completing the CDI does not have to rely solelyon recall to identify the words a child produces. TheCDI also has high validity (concurrent validity rangefrom .40 to .83; predictive validity range from .60 to.80). Because of its cost effectiveness, ease of adminis-tration, high reliability, and validity, the English CDIhas been adapted to numerous languages includingSpanish (Jackson-Maldonado, Thal, Bates, March-man, & Gutierrez-Clellen, 1993), Italian (Caselli et al.,1995), Japanese (Ogura, Yamashita, Murase, & Dale,1993) and Swedish (Eriksson & Berglund, 1999), with

der, but its rich and multilayered morphology allowsmuch flexibility, often producing OSV (using gram-matical facial expression to topicalize the object) or asingle inflected lexical V with the S and O encoded inspace (Newport & Meier, 1986). The effect of this ty-pological characteristic is that much of the input todeaf children places the verb in a salient position, ei-ther verb initial, or verb final. Given these varying in-puts, one can test the degree to which language struc-ture mediates the composition of early vocabulary bycomparing the relative proportions of nominals topredicates. Do nouns outnumber verbs initially in ASLas they do in English (Bates et al., 1994; Fenson et al.,1994) and Italian (Caselli et al., 1995)? Or is the com-position of toddlers’ vocabularies significantly influ-enced by the nature and typology of the parent lan-guage as seen in Korean (Gopnik & Choi, 1995) andJapanese (Clancy, 1986; Ogura, Yamashita, Murase, &Mahieu, 1999)? If the composition of children’s vocab-ularies is similar across English and ASL (which differgreatly in their structures and typology), one might ar-gue that more general cognitive mechanisms play a sig-nificant role in the types of words children acquire. If,on the other hand, early lexicons reflect the structuralcharacteristics of particular languages, that would sug-gest that the language, rather than general toddler cog-nitive abilities, is driving early lexical acquisition. Wecan begin to address these issues with the data tofollow.

Finally, because most deaf children have hearingparents and are exposed to sign language after infancy,and some not until late childhood or adolescence, theyoften acquire their primary language at different agesand stages of cognitive development. Thus, the acqui-sition of sign language by different populations (e.g.,deaf children born to hearing parents vs. deaf childrenborn to deaf parents) provides a naturalistic experi-ment for the critical period hypothesis for languagelearning (Lenneberg, 1967). When a primary languageis acquired at different ages, to what degree will theinitial stages of language development be comparable?We know from studies on adult late learners of ASLthat different aspects of language are more vulnerableto late acquisition than others (Emmorey, Bellugi,Friederici, & Horn, 1995; Newport, 1991). Related is-

MacArthur Inventory 85

versions for Chinese, German and other languages inprogress.

Theoretical Issues Specific to Signed Languages

With respect to early acquisition, ASL presents an in-triguing contrast to spoken language. The fact that itis conveyed in the visual-gestural modality has led todiscussions concerning age of onset of first signs andthe suggestion that ASL might be acquired earlier thanspoken languages (see Newport & Meier, 1986, for athoughtful discussion). Because sign is conveyed in thevisual-gestural modality and because infants (bothhearing and deaf) use communicative gestures prelin-guistically, in the initial stages of language develop-ment, children learning signed languages may have adistinct advantage over children learning spoken lan-guage, particularly in the acquisition of first “words.”Additionally, although signs have undergone processesof grammaticization, there is a pantomimic or iconicaspect to ASL, and one might suppose that childrenwould exploit this iconicity, thus entering the languagesystem earlier and faster than their hearing counter-parts.

An enduring issue in developmental psycholinguis-tics has been the role of cognition in the process of lan-guage acquisition (see Bates, Bretherton, & Snyder,1988; Piatelli-Palmarini for the debate between Piagetand Chomsky, 1980; Slobin, 1973, 1985). For example,some theorists propose that cognitive abilities precedethe onset of language and first words (Goodwyn &Acredolo, 1993; Gopnik & Meltzoff, 1985; Piaget,1952). Thus, if first signs consistently appear earlierthan first words, this would suggest that underlyingcognitive abilities for language are in place sooner thanpreviously thought. This would also indicate that thedevelopment of speech mechanisms restricts the age atwhich first words are used, rather than the underlyingcognitive abilities. In a similar vein, another interestingissue to explore in ASL involves the composition of theearly vocabularies of children learning ASL. In En-glish, word order is rather rigid (SVO) and subjects area required sentential component for English; thus, theinput to children acquiring English is filled with nomi-nal initial utterances. ASL also utilizes a SVO sign or-

sues pertain to the pattern and sequence of acquisition.Will late learners of ASL as a first language (e.g., deafchildren born to hearing parents) use similar strategiesand reflect a similar developmental sequence? Whatwill this tell us about the process of language acquisi-tion? Although the current data does not address theselatter questions, future data will be able to do so.

Method

Development of ASL-CDI

To adapt the English version of the toddler CDI tomake it linguistically and culturally appropriate fordeaf children, we made a number of changes. First,based on available data from our own lab from studieson early sign language vocabulary development and inconsultation with Deaf colleagues, we deleted words orcategories inappropriate for an ASL version of theCDI. For example, the Animal Sounds category wasomitted, as this is not a pertinent area to assess in deafchildren who are not exposed to these noises. In addi-tion, signs that reflected Deaf culture and experiences(such as TTY/TDD)1 were added, whereas words con-sidered ASL equivalent forms, such as horse/pony,were modified to include only the “formal” word (i.e.,HORSE)2 because the words are represented by thesame sign in ASL. In some cases, signs of virtuallyidentical form (e.g., EAT and FOOD, SIT andCHAIR) but different grammatical function (nounversus verb) were also modified to only include onesign.3 These changes affected five signs (EAT/FOOD,SIT/CHAIR, DRINK/DRINK, SLIDE/SLIDE,SWING/SWING) and in all cases, the sign was en-tered as a verb in the Action Signs category. Althoughthe English CDI has retained both the noun and verbforms of certain words, the determination to limit theASL-CDI signs was based on the pilot data wherebyparents consistently endorsed only the verb form of thesign even when presented with both options. Severalother signs could fall across categories such as BI-CYCLE (the noun, “bicycle,” and the action, “bicyc-ling”) and SHOWER (“the shower” vs. “showering”).However, the majority of ASL signs on this checklistfall clearly into only one grammatical category.

Finally, significant changes were made to the gram-


matical portion of the ASL-CDI. As ASL is a visual-gestural language with a multilayered grammaticalstructure, grammatical aspects cannot be easily dem-onstrated in a written form, as can be done with spokenlanguages. Thus, the grammar section was limited toinquiries about a child’s linguistic competence includ-ing questions about language use in the home, abilityto fingerspell, and the child’s longest sign productions.

Pilot data were collected from 30 children (16 girlsand 14 boys), ages 10 to 36 months, from 11 states. Allof the children were deaf, as were both of their parents.Ninety-seven percent of the families reported usingASL in their homes; 3% reported using Pidgin SignedEnglish (PSE). Sixty-three percent had deaf maternalgrandparents and 47% had deaf paternal grandparents.For each completed form, parents were paid $5.00.

Data from these forms were then analyzed. Signsused extremely infrequently (by 10% or less of thechildren) were deleted. Signs were added to the ASL-CDI if at least 10% of the parents spontaneously wrotein that their child was using them. These cut-off pointswere arbitrarily defined. The semantic category ofBody Parts was eliminated because pointing to thebody part rather than a formal sign is generally usedto identify body parts. Based on these changes, thecurrent version of the ASL-CDI, which consists of537 signs in 20 semantic categories, was created. The20 categories are Animals, Vehicles, Toys, Food andDrink, Clothing, Small Household Items, Furnitureand Rooms, Outside Things, Places to Go, People,Games and Routines, Action Words, Descriptive Signs,Signs about Time, Pronouns, Question Signs, Preposi-tions and Locations, Quantifiers, Helping Verbs, andConnecting Signs. All questions about early grammati-cal development were retained. See the appendix forthe current version of the ASL-CDI.

Normative Data Sampling

To begin the normative data collection, letters describ-ing our project were sent to schools that reported usingeither ASL or sign with speech as their primary com-municative mode, as identified in the American Annalsof the Deaf Reference Manuals (Craig & Craig, 1993;Stuckless & Corson, 1994). Of the 632 schools con-tacted, over 100 have replied to our letter. Although the

signs produced at Time 2 were not included in thisanalysis. Because test-retest reliability involves admin-istering the same test to the same child on two differentoccasions and then correlating the two scores, newsigns produced at Time 2 cannot be included in theanalysis as their presence would artificially inflate thereliability score. Overall, the ASL-CDI demonstratedhigh test-retest reliability exceeding an r of .91 (indi-vidual scores ranged from .82 to 1.00). This reliabilityrate is comparable with that reported for the Englishversion of the CDI (r of .95) and suggests that parentsare extremely consistent in their reporting of theirchild’s sign language production.

External validity refers to the extent that parentsare actually checking signs that their children produce.To assess external validity, we videotaped 10 children(age range from 17 to 34 months with reasonable distri-

Table 1 Number of children included at each month inthe cross-sectional norming sample

Age Number

8 months ............................................................. 19 months ............................................................. 010 months ........................................................... 411 months ........................................................... 212 months ........................................................... 113 months ........................................................... 314 months ........................................................... 115 months ........................................................... 316 months ........................................................... 217 months ........................................................... 218 months ........................................................... 319 months ........................................................... 220 months ........................................................... 021 months ........................................................... 322 months ........................................................... 223 months ........................................................... 524 months ........................................................... 425 months ........................................................... 126 months ........................................................... 027 months ........................................................... 528 months ........................................................... 329 months ........................................................... 330 months ........................................................... 331 months ........................................................... 432 months ........................................................... 633 months ........................................................... 234 months ........................................................... 335 months ........................................................... 136 months ........................................................... 0


majority of the responders did not have deaf childrenwith deaf parents acquiring ASL, or children in the ap-propriate age range, virtually all indicated that they be-lieved the project was important and offered to helpwith future data collection on deaf children of hearingparents.

Currently, data from 110 ASL-CDIs (69 children:34 longitudinal) have been compiled from the ASL-CDI (see Table 1 and Table 2). Our normative sampleincludes 32 girls and 37 boys who range in age from 8to 35 months. These children reside in 16 differentstates in the United States. All of the children are deaf,as are both of their parents. Of these children, 66 (96%)of their families reported using ASL in their homes;one reported using a combination of total communica-tion with ASL; one family reported using the generic“sign,” while one family did not report. Forty-two per-cent have maternal deaf grandparents, and 43% havepaternal deaf grandparents; 35% of the families haveno deaf grandparents. A primary consideration for in-clusion in this study was that both the child and herparent be deaf because that assumes a typical languagelearning environment. Because parents were asked toreport what language they used in the home (ratherthan to choose among predefined language categories),families were not excluded if they did not name “ASL”as their home language. The presence of a deaf house-hold where sign language is the primary form of com-munication was sufficient to allow inclusion and we be-lieve that it is highly likely that ASL is the languageused in the home.

Results

Reliability and Validity

Test-retest reliability was established for the ASL-CDIusing 25% of the normative data (n � 16). ASL-CDIswere initially collected and then readministered 5 or 7months later (age range: 11 to 29 months at first test-ing), with the same parent completing the form at eachdata point. The total number of signs produced at eachdata point was tallied. Then, the correlation betweenthe data points was made using only the signs pro-duced at both times and the signs produced at Time 1but were not checked at Time 2 (missing signs). New

bution among this range) on the same day that parentscompleted ASL-CDIs for them.4 Videotaping sessionsranged from 1 hour to 2.5 hours and captured periodsof free play as well as structured tasks designed to elicitlanguage production. We then compared transcrip-tions of the videotapes with the signs that parentschecked on the forms, giving particular attention tothose signs that the child produced but the parent didnot endorse on the checklist (e.g., signs that the childdoes in fact know but the parent did not report). Welimited our analysis to only those signs that werein common between the videotaping session and the

Table 2 Longitudinal data: Age at testing (vocabularysize)

Child Time 1 Time 2 Time 3 Time 4 Time 5

1 8 (2) 18 (18) 25 (35) 30 (122) 36 (166)2 10 (2) 13 (4) 23 (116)3 10 (3) 13 (59) 19 (231) 25 (319) 31 (499)4 10 (3) 24 (104)5 10 (3) 30 (272)6 11 (17) 14 (75) 18 (147) 30 (344)7 12 (88) 19 (202) 26 (376)8 13 (47) 29 (260)9 13 (7) 30 (179) 36 (210)10 14 (107) 26 (374)11 15 (102) 21 (239) 28 (409)12 15 (30) 23 (136)13 15 (27) 24 (261) 31 (439)14 15 (64) 27 (171) 35 (361)15 16 (65) 33 (472)16 18 (138) 23 (454)17 18 (83) 24 (211) 30 (349) 35 (475)18 18 (199) 26 (454) 36 (519)19 19 (295) 26 (320) 33 (488)20 18 (39) 34 (277)21 21 (115) 28 (327)22 22 (345) 29 (438)23 22 (162) 34 (303)24 23 (138) 27 (198)25 23 (83) 31 (138)26 23 (86) 32 (256)27 24 (348) 29 (391) 33 (476)28 24 (201) 35 (328)29 24 (133) 36 (364)30 25 (279) 34 (410)31 27 (284) 32 (337)32 27 (269) 32 (322)33 27 (205) 34 (316)34 29 (280) 35 (399)


ASL-CDI. A ratio between these two scores (numberof signs produced by the child during videotaping andendorsed by the parent on the ASL-CDI compared tothe total number of signs produced by the child thatthe parent could have either endorsed or not on theASL-CDI) yielded an external validity score of .87,ranging from .71 to 1.00. This suggests that parentsaccurately report the signs that their children pro-duced as measured by the ASL-CDI.

Vocabulary Production

Sign Production by Toddlers: Cross-Sectional Data

The youngest child reported as producing signs is 8months old, with a vocabulary of two signs (MILK,BATH). Four 10-month old children are each reportedas using two or three signs, all nouns with one iconicverb (CLAP). An 11-month-old child has a vocabularyof 17 signs, which also consists entirely of nouns. It isnot until the age of 12 months that a child is reportedas having a more diverse vocabulary consisting of a va-riety of lexical items such as nouns, verbs, quantifiers,and adjectives. By 34 months, a child was reported ashaving a vocabulary of 518 signs including all 20 se-mantic categories. Among the three 34-month-oldchildren included in this sample, vocabulary sizesrange from 264 to 518 signs, with a mean of 402 signs.

To provide a more descriptive overview of the de-velopment in children’s sign language between 8 and 35months, we clustered the children into five approxi-mately 6-month age groups (see Table 3). The follow-ing table provides information about the mean, me-dian, and ranges of their vocabularies in the five agegroups. Similar to other reports, the median number ofsigns rises significantly over the 27-month period (i.e.,8–35 months), reflecting a productive vocabulary al-most 150 times as large by 3 years of age as comparedwith the vocabulary size at 8 months.

To investigate developmental patterns, we con-ducted a regression analysis using age as the indepen-dent variable and vocabulary size as the dependentvariable. As expected, there was a positive correlation(r � .72) between age and vocabulary size, indicatingthat vocabulary size increases with age, F(1,67) � 175.97,p � .05. Figure 1 illustrates the vocabulary develop-

Developmental Changes in Vocabulary Composition:Cross-Sectional Data

To address changes in vocabulary composition, we ini-tially conducted a regression analysis with age as theindependent variable and noun:predicate ratio as thedependent variable. Children who did not produce anypredicates were not included in this sample as their ratioscore was undefined. Common nouns included the fol-lowing categories from the ASL-CDI: Animal Names,Clothing, Furniture and Rooms, People, Food andDrinks, Places to Go, Outside Items, Small HouseholdItems, Toys, and Vehicles. The combined commonnouns category contains 280 items comprising 52% ofthe total checklist. Predicates were defined as the sumof three categories from the ASL-CDI: Actions Signs,Helping Verbs, and Descriptive Signs (163 possibleitems, totaling 30% of the checklist).5 The noun:predi-cate ratio for the entire checklist (e.g., if all items wereendorsed) is 1.72. There was a significant negative cor-relation between age and the ratio of nouns to predi-cates as measured by the ASL-CDI, F(1,63) � 10.18, p �

.05 (see Figure 2). Specifically, after the age of 2 years,children begin to acquire more verbs and adjectivesinto their language, resulting in the observed decreaseof the proportion of nouns to predicates. (As noted ear-lier, a handful of signs on this checklist could poten-tially have classified as either a noun or a predicate, andfuture research should consider more carefully theclear classification of a sign as either a noun or a predi-cate to gain clarity on the development of this rela-tionship.)

However, there is considerable variability in ourdata. That is, the vocabulary sizes of the three 21-month old children included in the normative samplerange from 105 to 214 signs. There is a 22-month oldchild with a reported vocabulary of 348, while older


ment observed from the ASL-CDI. Recall that theASL-CDI provides an index of true vocabulary size;thus, a child’s actual vocabulary may be considerablylarger than reported here.

From 8 to 11 months of age, most children are pro-ducing little meaningful language (although they are allproducing some signs). This early onset of languageproduction by deaf children has been noted frequentlyin the literature (Bonvillian, Orlansky, & Novak, 1983;Orlansky & Bonvillian, 1984, 1985; Petitto, 1988) andraises interesting questions about the underlying na-ture of language development. However, hearing chil-dren produce communicative gestures at young ages(Acredelo & Goodwyn, 1988; Petitto, 1988, 1992; Volt-erra & Caselli, 1985). The average number of commu-nicative gestures for an 8-month-old child is 10 (Fen-son et al., 1994). So, it is quite possible that bothhearing and deaf children produce communicative ges-tures at young ages, but that only deaf children aregiven credit for having produced a lexical item.

By 12 to 17 months of age, the productive vocabu-laries are considerably larger. These findings are con-sistent with the long-reported claim that expressivelanguage begins primarily around 12 months of age(Fenson et al., 1994; Gesell, 1925; Lenneberg, 1967;McCarthy, 1954). More interesting is that the earlyproductive vocabularies of these children exceed thosereported for English-speaking children (Bates et al.,1994; Fenson et al., 1994). However, by 18–23 months,the median scores and ranges for both ASL and En-glish are comparable. This finding is consistent withmany reports that suggest that deaf children learningASL demonstrate an earlier onset in language produc-tion but no such advantage is observed by the end ofthe second year (Meier & Newport, 1990; Orlansky &Bonvillian, 1985; Prinz & Prinz, 1981).

Table 3 Descriptive statistics of children’s productive sign vocabularies

Mean n Median n Range of productiveAge group signs produced signs produced sign vocabularly

8–11 months (n � 7) 8 3 2–1712–17 months (n � 12) 61 62 7–10718–23 months (n � 15) 149 138 39–34824–29 months (n � 16) 252 261 102–41730–35 months (n � 19) 380 411 249–518

0

100

200

300

400

500

600

7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Age in Months

Voca

bula

ry S

ize

children have only 154 or 205 signs. Thus, it is possiblethat children with different vocabulary sizes are at dif-ferent developmental stages, regardless of their age,which may have resulted in a confound between devel-opmental and compositional variance. Thus, subse-quent analyses used vocabulary size as a continuousvariable or divided the children into six developmentallevels (e.g., by vocabulary size) rather than by chrono-logical age. Table 4 shows the groupings that were used.

To address the possible confound of developmen-tal and compositional variance, we conducted a regres-sion analysis using vocabulary size as the independentvariable and noun:predicate ratio as the dependent

Figure 2 Noun:predicate ratio by age.

Age in Months

Figure 1 Development of productive vocabulary from 8 to 30 months.


variable. There was a significant negative correlationbetween vocabulary size and the ratio of nouns topredicates in the vocabulary, F(1,63) � 12.48, p � .001;r � �.41. As vocabulary size increased, the proportionof nouns to predicates decreased. It is clear that nomi-nals dominate children’s early lexicons, and then predi-cates begin to take on a more substantial role as vocab-ulary size increases. This profile of “nominalization topredication” (Bates et al., 1994) has been observed inother languages as well, including English, Italian, andSpanish (Bates et al., 1994; Caselli et al., 1995; Jackson-Maldonado et al., 1993). The significant shift in vocab-ulary composition appeared to occur when children

if growth in language development occurred evenlyacross all vocabulary types (e.g., growth in nouns wascomparable to that of predicates). An investigation ofFigure 4 reveals that growth is quite different acrossthe three vocabulary types. Nouns, predicates, andclosed-class items follow distinctly different develop-mental paths.

The pattern for common nouns is an inverted-Ureflecting the predominant role that nouns play in theearly stages of lexical acquisition. Although neither thelinear nor quadratic trends were statistically signifi-cant, as can be seen in Figure 4, there is a clear peak inthe proportion of nouns in the child’s vocabulary whenthe child’s vocabulary reaches 51–200 words and subse-quently drops off. During the acquisition of the first200 signs, nouns occupy the greatest proportion of thetotal vocabulary, which is consistent with the patternobtained for the lexical acquisition of English-speakingchildren. Moreover, this pattern is consistent with thefinding that the initial stages of lexical acquisition aredominated by learning the names of common, everydayitems. The abundant use of nouns peaks at a mean of63% and begins to drop off considerably after the vo-cabulary size of 101–200 signs is attained. Eventually,the mean drops to 51%, which is just below the 52%figure that common nouns occupy within the checklistas a whole (i.e., the checklist ceiling).

The developmental pattern for predicates is quitedifferent than that of nouns (see Figure 4). The pro-


attained a vocabulary size of approximately 200 signs,as noted by the decreased variability in scores (see Fig-ure 3).

Following the analyses by Bates et al. (1994), wesummarized the developmental trends in vocabularycomposition for three variables: percent commonnouns, percent predicates, and percent closed-classitems (also called function items). Common nouns andpredicates are defined above. Closed-class items in-cluded the ASL-CDI categories of Connectors, Prepo-sitions, Pronouns, Quantitative Signs, and QuestionSigns (53 items, 10% of the checklist).

The horizontal dashed lines in Figure 4 representthe absolute percentage that each vocabulary type com-prises within the entire checklist (i.e., 52% nouns, 30%predicates, and 10% closed class). These lines reflectthe checklist ceiling and are the percentage of scoresthat the children would obtain if their parents endorsedall of the items on the checklist. These lines also rep-resent the percentages that we would expect to see

Table 4 Descriptives of vocabulary groups

Vocabulary range n Age range

0–50 12 8–19 months51–100 10 12–23 months101–200 13 14–29 months201–300 15 19–34 months301–400 7 22–35 months400� 12 27–34 months

Figure 3 Noun:predicate ratio by vocabulary size.

0

2

4

6

8

10

12

0 50 100 150 200 250 300 350 400 450 500 550

Vocabulary Size

No

un

:Pre

dic

ate

Rat

io

portion of predicates shows a slow, steady, linear in-crease as the vocabulary size increases; this lineargrowth was statistically significant, F(1,67) � 34.48, p �

.05. In children with vocabularies between 1–50 signs,predicates occupy only 17% of their total vocabulary.However, by the time the vocabulary size has grown to400� signs, predicates occupy 33% of the total vocab-ulary, which is slightly above the 30% checklist ceilingfor predicates. When compared to English-speakingchildren, the patterns of proportion of predicate usageappear similar. Notably, however, the proportion ofpredicates in the vocabulary of these deaf children isnearly twice that of their hearing counterparts whoare learning English. Although both the English andthe ASL checklist have comparable total predicate ceil-ings (24.4% and 30%, respectively), the percentage ofpredicates in the vocabulary of the deaf children isconsistently higher than those of the English-speakingchildren (see Table 5). This difference may reflect gram-matical differences between the languages and thefrequency of verb initial or verb final utterances in ASL.

Finally, the developmental pattern for the propor-tion of closed class signs was analyzed (see Figure 4).This pattern, similar to that for predicates, reflects arelatively steady, linear increase across the vocabularyrange; this growth is also statistically significant,F(1,67) � 29.06, p � .05. Unlike the other categories ofnouns and predicates whose relative usage appears tolevel off by 400 signs, the use of closed class signs ap-

Figure 4 Vocabulary composition: Proportion of nouns, predicates and closed class signs ateach vocabulary level. Flat dashed lines represent the percentage that each vocabulary type occu-pies within the checklist as a whole.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0–50 51–100 101–200 201–300 301–400 401+

Per

cen

t

Nouns

PredicatesClosedClass

Vocabulary Size


Table 5 Comparison of proportion of predicates (Englishdata extracted from Bates et al., 1994)

English English ASL

0–50 .07 .1751–100 .10 .22101–200 .14 .23201–300 .17 .30301–400 .21 .33400� .24 .33

pears to continue to rise during this period. This samepattern has been observed in English-speaking chil-dren, for whom 400 words was viewed as a “take-off ”point for closed class words (Bates et al., 1994).

In sum, with respect to the lexical composition ofearly sign vocabularies, both linear and nonlineartrends were observed. Nouns dominate the early vo-cabularies relative to other categories, but their pre-dominant presence levels off around 200 signs. Com-pared to nouns, predicates occupy a much smallerproportion of the total vocabulary, but escalate slowlyover time. Notably, deaf children appear to have agreater proportion of predicates in their early lexiconthan English-speaking children. Finally, closed classitems show a relatively strong linear trend until the vo-cabulary size reaches 400 signs, at which time growthwithin this category appears to accelerate. Our longitu-dinal data add some detail to the cross-sectional dataand confirm a number of the developmental patterns.

2000). Although this type of vocabulary burst seemshighly unlikely when observing our current data, basedon the linear trajectories observed, future investiga-tions with larger numbers of children tested at closerintervals will help clarify this point.

With respect to changes in vocabulary compositionsuch as noun:predicate ratio, proportions of nouns,proportion of predicates, and proportion of closed-class items, we observed similar patterns to those re-ported in the cross-sectional data. With a vocabularysize of less than 200, children demonstrated a relativelyhigh noun:predicate ratio, reflecting the abundance ofnouns in their early vocabulary. Once a vocabulary sizeof 200 is attained, there is a dramatic shift downward inthe ratio as more predicates enter a child’s vocabulary.There is one interesting pattern subtly apparent in thecross-sectional data that is more pronounced in thelongitudinal data. At a vocabulary size of about 300 to400 signs, children’s noun:predicate ratio drops con-siderably below the checklist ratio of 1.72. This shiftsuggests that during this period of time, children areacquiring predicates at a much more rapid rate thanthey are acquiring nouns. Thus, it is possible that a“verb burst” may be occurring during that time. Oursample size is too small to confirm this finding, but itwarrants additional research.

The pattern of proportion of nouns, proportions of


Figure 5 Vocabulary development trajectories for longitudinal data(for 13 children with more than three data points).

0

100

200

300

400

500

600

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Voca

bula

ry S

ize

Age in Months

Sign Production by Toddlers: Changes in Vocabulary:Longitudinal Data

Quantitative analyses on the development on children’svocabularies were not conducted. Obviously, as thechildren age, their vocabulary size will increase. More-over, the varying lengths of time between testings makecomparisons difficult and may possibly obscure anactual vocabulary spurt. However, the children’s vocab-ulary sizes at each data point are shown in Table 2. Fig-ure 5 shows the development of the children’s vocabu-laries over time; only those 13 children who had threedata points or more were included. Though we recog-nize that we do not have month-to-month data fromany child, we can still visually observe patterns in thedata. The most striking element of the graph to us isthat vocabulary development is remarkably linear formost children. This point is of interest because of theoften reported vocabulary burst observed in English-speaking children, sometimes as large as 75 words inone month (Bloom, 1973; Dromi, 1987; Goldfield &Reznick, 1990; Reznick & Goldfield, 1992). Recentmonthly longitudinal data have shown that nearly ev-ery child learning English demonstrates a vocabularyburst at some point in the first 3 years, although thetiming and size of this burst with respect to the child’sage or vocabulary size varies tremendously (Goodman,

predicates, and proportion of closed class items in thechildren’s vocabularies looked virtually identical to thecross-sectional data (see Figure 4). When their vocabu-lary was between 50–100, children showed the highestproportion of nouns. Over time, most of the childrendemonstrated steady increases in the proportion ofpredicates. Several children demonstrated a peak inproportion of predicates when their vocabulary sizeranged from 150–250 signs, which was followed by adrop in proportion of predicates as their vocabularysize grew. Again, these data suggest a more rapid ac-quisition of predicates during a limited period of time.For closed-class items, development was linear but rel-atively flat.

Early Vocabulary Development

First Words: Cross-Sectional Data

In the previous sections, we have talked about the num-ber of words in children’s lexicons and the composi-tion of these early vocabularies, but not their content.A recent topic of interest has been the degree to whichchildren’s first words are similar or different acrosslanguages or cultures. The knotty problem of whatconstitutes a word, idiosyncratic first words, and dif-ferent cultural practices, including input differences,have all been raised in this discussion. Although theCDI data do not represent a solution to these very realproblems, they do provide interesting data on whatparents report their children’s first words to be and anopportunity to look at the actual content of children’searly lexicons. Due to space limitations, we have madean arbitrary decision to present the first 35 words fromour data set and compare the list to that obtained fromdata collected from the English CDI (Fenson et al.,1994). Criterion for inclusion for both English andASL are that at least 50% of the sample must use thegiven word/sign by 18 months of age. That is, at least50% of the sample by 18 months of age used each ofthese words (n � 788 for English and n � 20 for ASL).Table 6 displays the list of the first 35 words/signs.

When we look across the lists of first words andsigns, we see strong similarities in the content of theseearly lexicons. As we noted, nouns, especially namesfor people, animals, and things to eat, far outstrip thenumber of verbs or predicates. The significant people,


especially Mommy and Daddy, appear early in bothlanguages. Animal names are also a common theme,but there is a clear reflection of modality: early lexiconsfrom English include animal sounds (generally used toname the animals), but ASL includes only animalnames. In fact, the range of animal names in the earlysign vocabulary is somewhat broader than in English,wherein both the animal name and its sound arelearned early, as in dog and woof, resulting in a semanticoverlap among these first words.

Additional common topics are things to eat, greet-ings, and clothing. An interesting lexically driven dif-

Table 6 First 35 words or signs to emerge in English orASL (English list extracted from Fenson et al., 1994)

English ASL

Daddy DADDYMommy MOMMYBye BYEBall BALLBaby BABYNo NOShoe SHOEBottle MILKCookie COOKIEKitty CATDog DOGUh oh MOREEye EAT/FOODNose DRINKBird BIRDCracker CRACKERBanana BANANAJuice DIAPERHi GRANDMABaabaa RABBITMoo CLAPOuch FINISH (all-done)Woof COWYumyum HATBalloon HORSEBook BOOKBath NAME SIGNS (including

child’s own name)Duck DUCKPeekaboo CRYNite-nite BATH/WASHCar FISHCheese TREEVroom KISSKeys SLEEPApple RAIN

quence refer to concepts or ideas that are less abstract(e.g., names for things and people) than words acquiredlater in the sequence (e.g., the manner, the causality,the time frame). In addition, linguistic constraints alsoappear to play a role in the sequence (Bloom et al.,1982). Specifically, more complex wh-forms requiremore complex, descriptive verbs, whereas simpler wh-forms could easily be used with the copula or simpleverb forms such as go and do.

Wh-Forms: Longitudinal Data

Our longitudinal sample allows a more detailed look atthe emergence of these signs within an individualchild’s vocabulary. Using the children for whom therewere at least three longitudinal data points, we investi-gated the order in which question signs entered the vo-cabulary. In general, question signs did not emergeuntil a child attained a vocabulary of 100 signs. With-out exception, WHERE was the first or one of the firstquestion signs to be acquired. The age at whichWHERE was reported by virtually every parent was 18months. The vocabulary size at which children ap-peared to have the question sign WHERE was between150 and 200 signs. The signs WHAT, WHO, andWHICH emerged next between 24 and 30 months ofage, with almost all the children producing those signsby age 30 months. The vocabulary range for the acqui-sition of these signs was from 250 to 300 signs. Finally,FOR-FOR and WHY emerge between 30 and 35months of age, although several parents report youngerchildren to be producing these signs. HOW and DO-DO emerge the latest of all these signs and were notreported to be produced by any child younger that 32months. Children had attained a vocabulary size ofgreater than 350 signs before these last four questionsigns emerged.

Emotion Signs: Cross-Sectional Data

In the development of emotion signs, signs about phys-ical states (e.g., SLEEPY, HUNGRY, THIRSTY)emerge earliest at around 15 months or within the first100 signs. By 24 months, virtually every child is re-ported to be producing signs for physical states. Thesefindings are consistent with our earlier naturalistic data(Reilly, McIntire, & Bellugi, 1990). In addition, this is


ference is names for body parts. Whereas these appearin English, they are not part of the ASL early vocabu-lary for the previously mentioned reason that individ-ual lexical signs for most body parts do not exist. Theyare signaled by points on the signer’s body to the ap-propriate body part.

Although predicates are few in these very early vo-cabularies, ASL has several more than English. How-ever, many of those that do occur in ASL (e.g., SLEEP,CLAP) have a strong iconic or gestural form, makingit difficult to determine if these early signs are gesturesor signs. These iconic verbs may well be present in asimilar form in the vocabularies of young hearing chil-dren as well.

Early vocabularies also include social phrases androutines, such as words or signs for no, for leave-taking(BYE : bye-bye), and for quieting children, sh! Perhaps,the most salient feature of these vocabularies is the veryhigh degree of similarity in their content, reflectingthe shared common functions of early words and theshared interests and concerns of toddlers as well as theshared culture.

Wh-Forms: Cross-Sectional Data

Within ASL, there has also been a growing interestwith respect to the emergence of various sign cate-gories, specifically question signs (e.g., WHERE,WHAT) and emotion signs. An investigation of ourdata reveals that question signs first emerge around 12months of age. The signs WHERE and WHAT are thefirst to appear; by age 21 months, virtually every childin our sample has either WHERE or WHAT (or both)in their vocabulary. WHERE and WHAT are followedby WHO, WHICH, FOR-FOR (“what is it for?”)around 24 months. HOW, WHY, and DO-DO (“whatare you doing/what does it do?”) emerge last around 30months. This sequence has been observed in English-speaking children as well as second-language learnersof English. In English, what, where, and who emerge be-tween 22 and 27 months, followed by why, how, when,and which acquired later, around 30 months (Bloom,Merkin, & Wootten, 1982; Fenson et al., 1994). Thepattern of acquisition has most frequently been ex-plained as resulting from constraints on cognitive de-velopment (Ervin-Tripp, 1970; Fahey, 1942; Tyack &Ingram, 1977). That is, words acquired early in the se-

comparable to development in English, wherein 50%of the children are reported to be expressing physicalstates by 23 months. Signs that denote feelings (e.g.,SAD, HAPPY, SCARED, ANGRY) begin to emergearound 18 to 20 months, or when children acquire avocabulary size of 100 to 200 signs. CRY is the earliestreported emotionally based sign, as it is a salient signfor children because of the obvious physical behaviorassociated with crying. Signs such as SCARED andSAD also occur commonly in the vocabularies of 20-month-old children. Of the emotion signs available onthe ASL-CDI, ANGRY is the last to emerge at around30 months. This development of signs for emotion alsomaps nicely onto the development in English, whereinby about 24 months, 50% of children are producingwords for emotion. In English, cry also emerges first(around 22 months), followed by sad, happy, and scared,with mad being acquired last at around 29 months ofage (Fenson et al., 1994).

Cognitive Verbs: Cross-Sectional Data

We looked at the emergence of three cognitive verbs:WANT, LIKE, and THINK. WANT consistentlyemerged first at around 18 months of age or when achild’s vocabulary reached 200 signs. LIKE emergednext at around 24 months of age or a vocabulary size of350 signs. THINK was last to emerge at between 30and 36 months with a vocabulary over 450 signs. Whilealmost every older child was producing WANT andLIKE, only about half was producing the signTHINK. Again, the acquisition timeline of these cog-nitive verbs is consistent with that observed in English.

Table 7 Comparisons of developmental sequences for negation

Devel. seq. based on previously Devel. seq. based on parentalreported cross-sectional data Devel. seq. based on parental report (age in month when(age in month of earliest report (age in month of approx. 50% of childrenproductiona earliest reported production demonstrate sign

NO (18) NO (12) NO (18)DON’T-WANT (19) NONE (12) NONE (24)NONE (21) DON’T-WANT (18) DON’T-WANT (24)CAN’T (24) CAN’T (19) DON’T-LIKE (26)DON’T-LIKE (24) DON’T-LIKE (21) DON’T-KNOW (30)NOT (31) NOT (22) NOT-YET (30)DON’T-KNOW (31) DON’T-KNOW (25) CAN’T (32)NOT-YET (36) NOT-YET (26) NOT (32)

aAnderson and Reilly, 1998.


Wanna emerges earliest in English (around 23 months)followed by like at 25 months, with think produced last,beyond the age of 30 months (Fenson et al., 1994).

Negation: Longitudinal Data

We have previously reported cross-sectional data ad-dressing the acquisition of negation in ASL (Ander-son & Reilly, 1998). Those data were obtained throughnaturalistic observation and semistructured videotapedinteractions with deaf children, and the order of acqui-sition of negative, manual signs is presented in Table 7.With the current data obtained through parental re-port, we are able to confirm the developmental timelinefor manual negation. As noted earlier, the sign NO isone of the first signs that children produce and repre-sents the earliest form of manual negation. Of the 27longitudinal children who were reported to be produc-ing signs for negation, every one produced NO first.The next signs to be produced consistently areDON’T-WANT and NONE, which were common inthe vocabularies of children ages 18 to 24 months (vo-cabulary size between 150 and 250 signs). Between24 and 30 months, DON’T-LIKE, DON’T-KNOW,and NOT-YET emerge (vocabulary size consistentlyranged from 300 to 400 signs). The last to appear arethe signs CAN’T and NOT, which are not seen consis-tently until children are older than 30 months (and thevocabulary size reaches at least 350 signs).

If we compare the first two columns of Table 7, theacquisition sequences obtained previously through ourcross-sectional data and currently through parental re-port are remarkably similar. The only significant differ-

there are individual letters (Akamatsu, 1985; Hirsh-Pasek, 1987; Padden, 1991). A similar phenomenon isnoted in the vocabularies of hearing children who learnmorpheme combinations or amalgams such as “I’ll-getit” in response to a ringing telephone long beforethey learn the individual words of I, will, get, and it.

As in other analyses, age did not appear to be a fac-tor in predicting whether children were fingerspellingor not. Children as young as 13 and 15 months werenoted to be fingerspelling words and 19- and 21-month-old children fingerspelling letters, whereasseveral 23- to 27-month-old children were reported tobe doing neither. Rather, vocabulary size does seem tobe an important factor. With one exception, none of the14 children with vocabulary sizes under 100 signs wasfingerspelling. The one child who had a vocabulary of86 signs was reported to be fingerspelling words butnot letters. The opposite pattern is observed for chil-dren with vocabulary sizes above 100 signs (n � 47).With two exceptions, all of the children were makingattempts to fingerspell words, letters, or both. The twochildren who were not had vocabulary sizes of 105 and116. By the time a child’s vocabulary reaches 400 signs,he or she is consistently attempting to fingerspell bothwords and letters.

Fingerspelling: Longitudinal Data

Our longitudinal data provide some additional infor-mation about the development of fingerspelling. Of the32 children sampled longitudinally, 17 of the childrenwere reported to either be (1) fingerspelling both wordsand letters at all time points or (2) not fingerspelling atone time point and then fingerspelling both words andletters at the subsequent time point. The remaining15 children had parents who differentially endorsedtheir fingerspelling behavior over time. With remark-able consistency, parents noted that their child couldfingerspell words at least several months before theycould fingerspell letters. Specifically, 12 parents re-ported that their child was fingerspelling words at onetime point and began fingerspelling letters at the sec-ond time point. In addition, three parents reportedtheir child was not fingerspelling at all at one timepoint, but at a subsequent time point, their child wasfingerspelling words but not letters. No parent ever re-ported that their child produced fingerspelled letters


ence is the earlier production of signs as reported by theparents. Such a difference would be expected, given thedifferent methods of data collection. The data in thefirst column were obtained by reviewing videotaped in-teractions with the child that only sampled their vocab-ulary (e.g., they had to produce a negative sign duringthe videotaped interaction in order to be included).The data in the second column obtained through pa-rental report reflects a more broad range of the child’svocabulary. Notably, by parental report, every sign ap-pears to be acquired earlier than our previously re-ported data, although the sequence is nearly identical.

If we evaluate the last two columns of Table 7(which compare the first occurrence of the sign versusthe age at which approximately 50% of the children areproducing the sign), the developmental sequences varyin only one significant way. There is a notable shift ofthe signs CAN’T and NOT to the end of the sequence.Thus, although children as young as 19 and 22 monthswere observed producing the signs CAN’T and NOT,this appears to be precocious production based on theentire longitudinal data set.

Development of Other Communicative Abilities

Fingerspelling: Cross-Sectional Data

Our investigation into the acquisition of grammar fo-cused on two components: fingerspelling and wordcombinations. Recall that these data are collected byasking the parent whether the child performs a particu-lar grammatical form or signal; not all parents re-sponded to all the grammatical questions. Our explora-tion of early grammar revealed some interestingpatterns. First, parents were asked whether their chil-dren were fingerspelling words or letters. For mostchildren, parents responded that they were either fin-gerspelling words and letters or they were doing nei-ther. However, for 17 children, parents differentiallyendorsed what their child was fingerspelling. Interest-ingly, for 16 of the 17 children, their parents noted thattheir child was only fingerspelling words but was notyet fingerspelling individual letters. This finding hasbeen noted by many parents and researchers of deafchildren, who point out that children simply learn thefingerspelling handshape combination (e.g., R-I-C-E)as a whole, single lexical unit before they discover that

before words. Such a pattern lends strong support tothe finding that deaf children learn handshape config-urations for fingerspelled words before they recognizethe correspondence between handshape configurationsand the individual alphabetic characters.

Due to the variation in time between testings, it isnot possible to say how much time elapses between theonset of fingerspelling words and the onset of fin-gerspelling letters. Four months is the shortest dura-tion of time between these two behaviors that we haveavailable in our data. However, fingerspelling of letterslikely occurs relatively soon after the fingerspelling ofwords occurs. With respect to the vocabulary size ofthe children when fingerspelling words or letters oc-curs, most children were fingerspelling words by about100 signs, although several parents noted that theirchild was fingerspelling words with a vocabulary of lessthan 50 signs. Fingerspelling of both letters and wordsoccurred around 300 signs (earlier than the 400 signsreported for the cross-sectional sample).

Multisign Combinations: Cross-Sectional Data

Parents were also asked about whether their child wascombining signs and to provide examples of theirchild’s three longest sentences. Not surprisingly, agewas not a good predictor of whether children werecombining signs. Rather, vocabulary size again ap-peared to be the best indicator. Of the 11 children withvocabulary sizes under 50 signs, only 2 were reportedto be combining signs. Those two children had eachonly produced one exemplar of sign combination(“WHERE BLANKET” and “ME DRINK”). For the14 children with vocabularies ranging between 50 and110, 7 were reported to be combining signs while 7were not yet doing so. Notably, all of the examples pro-vided by the parents were two-word combinations suchas “MOMMY EAT,” “DADDY WORK,” “MORECRACKER,” “WANT DRINK,” and “DOLLSLEEP.” Beyond a vocabulary size of 115, every childwas reported to be combining signs. Moreover, nearlyevery parent provided at least one exemplar of multisigncombinations. When compared to English norms,these data are quite comparable (Fenson et al., 1994).In English, when the vocabulary size is less than 100,very few children are reported to be combining wordsregularly. However, once the vocabulary size reaches


101–200 words, approximately 75% of parents reportthat their child produces multiword utterances withsome regularity.

Future Directions

Two aspects of the ASL-CDI that are clearly differentfrom CDIs in other languages are the lack of a compre-hension form and the absence of a significant grammarsection. On the English forms, the parent is asked toendorse words that their child understands in additionto the words the child produces. The grammar sectionon the English forms has several parts. One sectionasks the parent to endorse other words their child mayproduce such as children, bought, breaked, and mouses inan attempt to understand how the child is producingplurals, past tense forms, and so on. In addition, theparent is presented with two sentences such as “Daddycar” and “Daddy’s car” and asked to choose whichsounds most like the child’s speech.

For the ASL-CDI, a comprehension form wouldbe relatively easy to develop and administer. However,an evaluation of grammatical development is muchmore complex. As ASL is a visual-gestural languagethat engages multiple channels simultaneously (e.g.,eyebrows, mouth, torso, and hands), a significant por-tion of the grammar occurs in a nonmanual form,which has no easy written translation. For example,ASL relies on a multilayered organization in which theverb stem, such as GIVE, and its aspectual and inflec-tional markers (e.g., habitual and distributive) co-occurin space (see Klima & Bellugi, 1979). These nonmanualgrammatical forms are evident at all levels of the gram-mar: phonology, morphology, and syntax. Because themultilayered nature of ASL does not lend itself to awritten format, we believe that a video format for agrammatical section might be more successful. That is,several sentences could be presented visually that con-trast on one variable (e.g., one including a nonmanualgrammatical marking and one without). Then, parentscould identify which one most closely reflects theirchild’s signing.

Although the data presented in this article repre-sent vocabulary acquisition from a large sample of deafchildren with deaf parents, there was considerable vari-ability in vocabulary size across the age range. Thus,we are not yet able to delineate the parameters of “aver-

way to children who are learning spoken languages.The sequence and developmental timeline for signs foremotion, cognitive verbs, and wh-forms map compar-ably onto those in English. However, some notable dif-ferences did emerge.

First, deaf children in this sample were producingsigns at the earliest age sampled, at 8 months of age.This is considerably earlier than children learning En-glish, who generally do not produce their first wordsuntil age 12 months. Notably, the content of the firstwords was comparable between English and ASL. Thisapparently precocious onset of language in deaf chil-dren suggests that the underlying cognitive abilities forlanguage are in place earlier than previously thoughtand that it is the development of speech mechanismsthat restricts the age at which first words are produced.

Second, our normative data revealed that the con-tent of the vocabulary of the youngest children in thesample is composed entirely of nouns, similar to find-ings from hearing children learning spoken English.The first 35 signs in both ASL and English are remark-ably similar and generally involve things to eat, greet-ings, and articles of clothing. However, children learn-ing ASL demonstrate more verbs in their early lexicon,a pattern that remains throughout the first 3 years. Theincreased use of verbs may well be related to the struc-ture of ASL, which often places verbs in the initial po-sition of an utterance, thus making it quite salient forthe receiver.

In concluding, we hope that this normative datawill serve a significant role in both the research andclinical realm. The ASL-CDI provides a standardizedtool with which to compare children’s language compe-tence. Normative data obtained from the inventory willassist researchers addressing basic theoretical ques-tions about language and language learning and clini-cians and educators who work with young deaf chil-dren. We can develop a better understanding of therelationship between language and other cognitive abil-ities with the use of the ASL-CDI as a measure oflanguage performance. As the normative data set con-tinues to grow, parameters for average and atyp-ical development can be defined. Such guidelines willserve a critical role for educators and parents ofdeaf children.


age” or “atypical” vocabulary production. However, asthe normative sample continues to grow, a long-termgoal of this project is to provide such information. Wehope that the ASL-CDI can be used as a screening toolfor identifying language delays in deaf children. In par-ticular, deaf children of hearing parents who are at riskfor language delay can be identified early and interven-tions established. Once an intervention is in place, theASL-CDI can be used to monitor progress and assesschange and improvement over time. In addition, theASL-CDI can be used as a guideline by hearing par-ents who are learning sign language to emphasize thosesigns that their deaf child is likely to acquire first.

In addition to defining the parameters of atypicallanguage development, we anticipate that the ASL-CDI will lend itself to an investigation of other issuesin language development. For example, the ASL-CDIcould be used as a tool to investigate vocabulary onset,vocabulary spurts, morphological development, birthorder effects, and parental style. In addition, becausethe ASL-CDI can provide critical standardized infor-mation about early sign language development, re-searchers will finally have a tool with which to appro-priately match young deaf children for language level.Once language variables are controlled for, the effect ofother factors can be better clarified, that is, behaviorproblems, cognitive abilities, and so on. In sum, wehope the ASL-CDI will be useful in matching childrenon language level in order to address other importantclinical and theoretical issues in development.

Conclusions

In this article, we have presented information on thedevelopment of the MacArthur Communicative Devel-opment Inventory for American Sign Language (ASL-CDI) and discussed the early normative data. Overall,the development of the ASL-CDI has been successful.Preliminary analyses have demonstrated high reliabil-ity and validity for the form. Responses from both theDeaf community and educators of the deaf have indi-cated that the ASL-CDI is easy to complete and willprovide valuable information about early sign devel-opment.

Results from these normative data indicate that onthe whole, deaf children who are acquiring ASL in anatural environment develop sign language in a similar

Appendix

The MacArthur Communicative Developmental Inventoryfor American Sign Language for Children 8 to 36 Months

Copyright 1992 All Rights ReservedFor information, contact Diane Anderson and Judy Reilly, Department of Psychology, San Diego State University,San Diego, CA 92120.(For simplicity, in this appendix we have omitted the check boxes that would normally appear beside each word.)

Family InformationChild’s Name Today’s dateParent’s Name Child’s BirthdateAddress Male Female

AgeTelephone

Is your child deaf? Yes No When did he/she become deaf?What was the cause?Any deaf brothers or sisters? Yes No If yes, how old?Is mother deaf? Yes No Are mother’s parents deaf? Yes NoIs father deaf? Yes No Are father’s parents deaf? Yes NoAny other deaf relatives?What language does your child use at home?Which hand does your child prefer for signing? L R For eating? L RFor rolling or throwing a ball? L R

Early UnderstandingBefore children begin to sign, they understand what signs mean. They respond to people signing familiar signsand phrases. Below are some examples of phrases. Does your child do any of these?1. Respond to his/her name sign? Yes No2. Respond to NO or FINISH by stopping what he/she is doing? Yes No3. Look for mommy/daddy if you sign WHERE MOMMY/DADDY? Yes No

In the list below, make an X by the sign phrases that your child understands.BE-CAREFUL GET-UP SPIT IT OUTBRING-ME GIVE-TO MOMMY TIME GO BYE-BYECHANGE DIAPER GIVE-ME HUG THROW BALLCOME-HERE GIVE-ME KISS YOU HUNGRY?DADDY/MOMMY HOME GOOD GIRL/BOY YOU TIRED/SLEEPY?DON’T TOUCH/NO TOUCH OPEN YOUR MOUTH YOU-WANT GO?FINISH SIT-DOWN YOU-WANT MORE?

Vocabulary ChecklistPlease look through the list of signs below. If your child uses a sign, put an X in the box. Your child may not yetuse the “correct” adult sign for a word, but may use a “baby” sign. Even if your child uses a baby sign please putan X in the box. Don’t worry if your child only knows a few signs now because this is a collection of signs frommany children at different ages.



AnimalsIf your child uses a sign check the box.ALLIGATOR DEER HORSE SPIDERANIMAL DOG LION SQUIRRELBEAR DONKEY MONKEY TIGERBEE DUCK MOUSE TURKEYBIRD/CHICKEN ELEPHANT OWL TURTLEBUG/ANT FISH PIG WORFBUNNY/RABBIT FOX ROOSTER ZEBRABUTTERFLY FROG SHEEPCAT GIRAFFE SKUNKCOW HEN SNAKE

ClothingIf your child uses a sign check the box.BELT EARRING PANTS SWEATERBIB GLASSES SANDAL SWIMSUITBOOT GLOVES SCARF UNDERPANTSBUTTON HAT SHIRT WRISTWATCHCOAT/JACKET HEARING AID SHOE ZIPPERDIAPER NECKLACE SNOWSUITDRESS/CLOTHES PAJAMAS SOCK

Games and RoutinesIf your child uses a sign check the box.BASEBALL BYE LET’S-SEE SHH/SHUSHBASKETBALL CALL(TDD) LUNCH SHOPPINGBATH DINNER NAP STORYBIRTHDAY FOOTBALL NO THANK-YOUBOWLING GOLF PEE/BM WASH-FACEBREAKFAST HI PEEKABOO WASH-HANDSBRUSH-HAIR I-LOVE-YOU PLEASE YES

Action SignsIf your child uses a sign check the box.AWAY FEED LOVE SLIDEBITE FIND MAKE SMILEBLOW FINISH OPEN SPILLBREAK FIX PAINT STANDBRING/CARRY GET/TAKE PICK/CHOOSE STAYBUILD GET-A-SHOT PLAY STOPBUY GIVE POUR SWIMCATCH GO PULL SWINGCHASE HAVE PUSH TAKE-PILLCHAT HEAR PUT TALKCLAP HELP READ TASTECLEAN HIDE RIDE-IN TEACHCLIMB HIT RUN TEAR/RIPCLOSE HOLD SAY TELL

COOK HUG SEARCH-FOR THROWCRY HURRY SEE TOUCHCUT JUMP SHAKE WAITDANCE KICK SHARE WALKDRAW KISS SHOW WASHDRINK KNOCK SIGN WATCHDROP LEARN SIT WORKEAT LICK SKATEENTER LIKE SLEEPFALL LOOK THINK

Furniture and RoomsIf your child uses a sign check the box.BASEMENT DOOR MIRROR TABLEBATHROOM/TOILET DRAWER REFRIGERATOR TVBATHTUB DRYER ROCKING CHAIR WASH MACHINEBED GARAGE ROOM WINDOWBEDROOM KITCHEN SHOWERCLOSET LIVING ROOM STAIRS

ToysIf your child uses a sign check the box.BALL BOOK GAME PUZZLEBALLOON BUBBLES PENCIL ROBOTBAT DOLL PRESENT/GIFT TOY

PeopleIf your child uses a sign check the box.AUNT COUSIN GRANDPA PEOPLEBABY COWBOY INDIAN POLICEBABYSITTER DADDY MAILMAN SANTA CLAUSBOY DOCTOR MAN SISTERBROTHER FIREMAN MOMMY TEACHERCHILD/KID FRIEND MONSTER UNCLEChild’s Name GIRL NAME SIGNS WOMANCLOWN GRANDMA NURSE

Food and DrinkIf your child uses a sign check the box.APPLE CORN LOLLIPOP SAUCEBAGEL CRACKER MEAT SODA/POPBANANA CREAM CHEESE MILK SOUPBEANS DONUT NUT SPAGHETTIBREAD EGG ORANGE STRAWBERRYBUTTER FRENCH FRIES PANCAKE SYRUPCAKE GRAPES PEANUT BUTTER TEACANDY GREEN BEANS PIZZA TOASTCARROT GUM POPSICLE TOMATOCEREAL HAMBURGER POTATO VANILLACHEESE HOT DOG POTATO CHIPS VITAMINSCHOCOLATE ICE PUMPKIN WAFFLE



COFFEE ICE CREAM SALAD WATERCOKE JELLY SALT WATERMELONCOOKIE JUICE SANDWICH

Connecting SignsIf your child uses a sign check the box.AND BECAUSE BUT IF/SUPPOSE

Prepositions and LocationsIf your child uses a sign check the box.ABOVE FRONT ON UPBACK HERE OUT WITHBEHIND9 INSIDE/IN OVERDOWN NEXT-TO TOFOR OFF UNDER

Outside ThingsIf your child uses a sign check the box.BACKYARD/YARD HOSE ROCK STREETCLOUD LADDER SHOVEL SUNFLAG LAWN MOWER SNOW TREEFLOWER LIGHTNING SNOWMANFOREST MOON SPRINKLERGRASS RAIN STAR

PronounsIf your child uses a sign check the box.HE/SHE/IT MYSELF THEY YOURSELFHIS/HERS OUR WEME THAT/THESE YOUMINE/MY THEIR YOUR

Places To GoIf your child uses a sign check the box.BEACH FARM MOVIE RESTAURANTCAMPING GAS STATION OUTSIDE SCHOOLCHURCH HOME PARTY STORECIRCUS HOUSE PICNIC TOWNCOUNTRY McDONALD’S PLAYGROUND ZOO

QuantifiersIf your child uses a sign check the boxA LOT/MUCH ANY NOT SAMEALL ORE NUMBERS SOMEGONE/ALL-GONE NONE OTHER

Question SignsIf your child uses a sign check the boxDO-DO HOW WHERE WHOFOR-FOR WHAT WHICH WHY

Small Household ItemsIf your child uses a sign check the boxBAND-AID COMB MOP SOAPBASKET9 FORK NAPKIN SPOON

BLANKET GLASS NEWSPAPER TAPEBOTTLE HAMMER PAPER TDD/TTYBOWL KEYS PICTURE TELEPHONEBOX KNIFE PLANT TISSUE/KLEENEXBROOM LETTER PLATE TOOTHBRUSHBRUSH LIGHT PURSE TOWELCAMERA MEDICINE RADIO TRASH/GARBAGECENT/COIN MONEY SCISSORS VACUUM

VehiclesIf your child uses a sign check the box.AIRPLANE B OAT FIRETRUCK SLEDAMBULANCE BUS HELICOPTER TRAINBICYCLE CAR/DRIVE MOTORCYCLE TRUCK

Signs About TimeIf your child uses a sign check the box.AFTER MORNING SOON TOMORROWBEFORE NIGHT TIME TONIGHTDAY NOT-YET TODAY YESTERDAYLATER NOW

Helping VerbsIf your child uses a sign check the box.CAN DON’T-LIKE NEED-TO/MUSTCAN’T DON’T-WANT WANTDON’T-KNOW LET-ME-SEE WILL

Descriptive SignsIf your child uses a sign check the box.ANGRY EMPTY LITTLE SLEEPYBAD FAST LONG SLOWBETTER FINE NEW SOFTBIG FIRST NICE STICKYBLACK FULL NOISY STUCKBLUE FUNNY OLD THIRSTYBROWN GOOD PINK TIREDCAREFUL GREEN POOR UPSETCOLD HAPPY PRETTY WETCRAZY HARD PURPLE WHITECUTE HEAVY QUIET WINDYDARK HIGH RED YELLOWDEAF HOT SAD YUCKYDELICIOUS HUNGRY SCAREDDIRTY HURT 9 SICKDRY LAST SILLYDoes your child try to fingerspell words? Yes NoDoes your child fingerspell letters? Yes NoDoes your baby combine signs to make sentences? Yes NoIf yes, please write examples of your child’s three longest ASL sentences:


velopment. In J. Kyle & B. Woll (Eds.), Language in sign (pp.116–125). London: Croom Helm.

Caselli, M. C. (1983). Communication to language: Deaf chil-dren’s and hearing children’s development compared. SignLanguage Studies, 39, 113–144.

Caselli, M., Bates, E., Casadio, P., Fenson, J., Fenson, L., Sand-erl, L., & Weir, J. (1995). A cross-linguistic study of earlylexical development. Cognitive Development, 10, 159–199.

Clancy, P. (1986). The acquisition of Japanese. [Monograph]Crosslinguistic Study of Language Acquisition, 4, 373–524.

Craig, W. N., & Craig, H. B. (Eds.). (1993). Educational pro-grams for deaf students. American Annals of the Deaf, 138,101–162.

Dromi, E. (1987). Early lexical development. New York: Cam-bridge University Press.

Ellenberger, R., Moores, D., & Hoffmeister, R. (1975). Earlystages in the acquisition of negation by a deaf child of deaf parents(Report No. 94). Washington, DC: Department of Health,Education and Welfare.

Emmorey, K., Bellugi, U., Friederici, A., & Horn, P. (1995).Effects of age of acquisition on grammatical sensitivity: Evi-dence from on-line and off-line tasks. Applied Psycholinguis-tics, 16, 1–23.

Eriksson, M., & Berglund, E. (1999). Swedish early communi-cative development inventories: Words and gestures. FirstLanguage, 19, 55–90.

Ervin-Tripp, S. (1970). Discourse agreement: How children an-swer questions. In J. R. Hayes (Ed.), Cognition and the devel-opment of language. New York: Wiley.

Fahey, G. (1942). The questioning activity of children. Journalof Genetic Psychology, 60, 337–357.

Fenson, L., Dale, P., Reznick, S., Bates, E., Thal, D., & Pethick,S. (1994). Variability in early communicative development.[Monograph] Society for Research in Child Development, 59,1–189.

Folven, R., & Bonvillian, J. (1991). The transition from nonrefer-ential to referential language in children acquiring AmericanSign Language. Developmental Psychology, 27, 806–816.

Gesell, A. (1925). The mental growth of the preschool child: A psy-chological outline of normal development from birth to the sixthyear, including a system of developmental diagnosis. New York:Macmillan.

Goldfield, R., & Reznick, S. (1990). Early lexical acquisition:Rate, content, and the vocabulary spurt. Journal of ChildLanguage, 17, 171–183.

Goodwyn, S. & Acredolo, L. (1993). Symbolic gesture versusword: Is there a modality advantage for onset of symbol use?Child Development, 64, 688–701.

Gopnik, A., & Choi, S. (1995). Names, relational words, and cog-nitive development in English and Korean speakers: Nounsare not always learned before verbs. In M. Tomasello & W.Merrimen (Eds.), Beyond names for things: Young children’sacquisition of verbs (pp. 63–80). Hillsdale, NJ: Lawrence Erl-baum.

Gopnik, A., & Meltzoff, A. (1985). From people, to plans, to ob-jects: Changes in the meaning of early words and their rela-tion to cognitive development. Journal of Pragmatics, 9,495–512.


Notes

1. Teletypewriter/telecommunication device for the deaf.2. In accordance with sign language transcription conven-

tions outlined by Baker and Cokely (1978), glosses for manualsigns are represented by English words in capital letters.

3. Nouns and verbs are distinguished from each other byfrequency and manner of movement where within the signingspace, nouns are repeated, generally smaller, and restricted inmovement (Launer, 1982; Supalla & Newport, 1978)

4. Three children under 17 months of age were also video-taped, but did not produce any signs during their taping session.As such, they could not be used in the external validity calcu-lation.

5. It is important to note that we are defining “nouns,”“verbs,” and “function words” based on their role in adult lan-guage. We recognize that children may use signs in ways thatvary from adult use.

References

Acredelo, L. & Goodwyn, S. (1988). Symbolic gesturing in nor-mal infants. Child Development, 28, 40–49.

Akamatsu, C. (1985). Fingerspelling formulae: A word is moreor less than the sum of its letters. In W. Stokoe & V. Volterra(Eds.), SLR ’83: Sign language research (pp. 126–132). SilverSpring, MD: Linstok Press.

Anderson, D., & Reilly, J. (1998). The puzzle of negation: Howchildren move from communicative to grammatical nega-tion in ASL. Applied Psycholinguistics, 18, 411–429.

Baker, C., & Cokely, D. (1978). American Sign Language: A teach-er’s resource text on grammar and culture. Silver Spring, MD:T. J. Publishers.

Bates, E., Bretherton, I., & Snyder, L. (1988). From first wordsto grammar: Individual differences and dissociable mechanisms.Cambridge, MA: Cambridge University Press.

Bates, E., Marchman, V., Thal, D., Fenson, L., Dale, P., Rez-nick, J., Reilly, J. & Hartung, J. (1994). Developmental andstylistic variation in the composition of early vocabulary.Journal of Child Language, 21, 85–123.

Bellugi, U., & Klima, E. (1982). The acquisition of three mor-phology systems in American Sign Language. In Papers andReports on Child Language Development (pp. 1–35). Stanford,CA: Stanford University.

Bloom, L. (1973). One word at a time: The use of single word utter-ances before syntax. The Hague: Mouton.

Bloom, L., Merkin, S., & Wootten, J. (1982). Wh-questions:Linguistic factors that contribute to the sequence of acquisi-tion. Child Development, 53, 1084–1092.

Bonvillian, J., & Folven, R. (1987). The onset of signing in youngchildren. In W. Edmondson & F. Karlsson (Eds.), SLR ’87:Papers from the Fourth International Symposium on Sign Lan-guage Research. Hamburg: Signum Press.

Bonvillian, J., Orlansky, M., & Novak, L. (1983). Early sign lan-guage acquisition and its relation to cognitive and motor de-

Hirsh-Pasek, K. (1987). The metalinguistcs of fingerspelling: Analternate way to increase reading vocabulary in congenitallydeaf readers. Reading Research Quarterly, 22, 455–474.

Jackson-Maldonado, D., Thal, D., Bates, E., Marchman, V., &Gutierrez-Clellen, V. (1993). Early lexical development ofSpanish-speaking infants and toddlers. Journal of Child Lan-guage, 20, 523–549.

Klima, E., & Bellugi, U. (1979). The signs of language. Cambridge,MA: Harvard University Press.

Launer, P. (1983). “A plane” is not “to fly”: Acquiring the dis-tinction between related nouns and verbs in American SignLanguage. (Doctoral dissertation, City University, NewYork, 1983). Dissertation Abstracts International, 43, 9-B.

Lenneberg, E. (1967). Biological foundations of language. NewYork: Wiley.

Maller, S. J., Singleton, J. L., Supalla, S. J., & Wix, T. (1999).The development and psychometric properties of the Amer-ican Sign Language Proficiency Assessment (ASL-PA).Journal of Deaf Studies and Deaf Education, 4, 249–269.

McCarthy, M. (1954). Language development in children. In L.Carmichael (Ed.), Manual of child psychology. New York:Wiley.

McIntire, M. L. (1977). The acquisition of American Sign Lan-guage hand configurations. Sign Language Studies, 16,247–266.

Meier, R., & Newport, E. (1990). Out of the hands of babes: Ona possible sign advantage in language acquisition. Language,66, 1–23.

Newport, E. (1991). Contrasting concepts of the critical periodfor language. In: S. Carey & R. Gelman, (Eds.), The epigene-sis of mind: Essays on biology and cognition (pp. 111–130).Hillsdale, NJ: Lawrence Erlbaum.

Newport, E., & Meier, R. (1986). The acquisition of AmericanSign Language. [Monograph] Crosslinguistic Study of Lan-guage Acquisition, 10, 881–938.

Ogura, T., Yamashita, Y., Murase, T., & Dale, P. (1993). Somepreliminary findings from the Japanese Early CommunicativeDevelopment Inventory. Paper presented at the Sixth Inter-national Congress for the Study of Child Language July,Trieste, Italy.

Ogura, T., Yamashita, Y., Murase, T., & Mahieu, A. (1999). Ac-quisition of nouns and verbs in Japanese children. Paper pre-sented at the International Congress for the Study of ChildLanguage, July, San Sebastian, Spain.

Orlansky, M., & Bonvillian, J. (1984). The role of iconicity inearly sign language acquisition. Journal of Speech and Hear-ing Disorders, 49, 287–292.

Orlansky, M., & Bonvillian, J. (1985). Sign language acquisition:Language development in children of deaf parents and im-plications for other populations. Merrill-Palmer Quarterly,31, 127–143.

Padden, C. (1991). The acquisition of fingerspelling by deaf chil-dren. In P. Siple & D. Fischer (Eds.), Theoretical issues in signlanguage research (vol. 2, pp. 191–210). Chicago: Universityof Chicago Press.


Padden, C. A., & Le Master, B. (1985). An alphabet on hand:The acquisition of fingerspelling in deaf children. Sign Lan-guage Studies, 47, 161–172.

Petitto, L. (1988). “Language” in the prelinguistic child. InFrank S. Kessel (Ed.), The development of language and lan-guage researchers: Essays in honor of Roger Brown (pp. 187–221). Hillsdale, NJ: Lawrence Erlbaum.

Petitto, L. (1992). Modularity and constraints in early lexical ac-quisition: Evidence from children’s early language and ges-ture. In Megan R. Gunnar & Michael Maratsos (Eds.),Modularity and constraints in language and cognition (pp. 25–58). Hillsdale, NJ: Lawrence Erlbaum.

Petitto, L. A., & Marentette, P. F. (1991). Babbling in the manualmode: Evidence for the ontongeny of language. Science,251, 1493–1496.

Piatelli-Palmerini, M. (Ed.). (1980). Language and learning: Thedebate between Jean Piaget and Noam Chomsky. Cambridge,MA: Harvard University Press.

Piaget, J. (1952). The origins of intelligence in children. New York:International Universities Press.

Poizner, H., Klima, E., & Bellugi, U. (1987). What the hands re-veal about the brain. Cambridge, MA: MIT Press.

Prinz, P., & Prinz, E. (1981). Acquisition of ASL and spokenEnglish by a hearing child of a deaf mother and a hearingfather: Phase II, early combinatorial patterns. Sign LanguageStudies, 30, 78–88.

Reilly, J., McIntire, M., & Bellugi, U. (1990). Faces: The rela-tionship between language and affect. In V. Volterra & C.Erting (Eds.), From gesture to language in hearing and deafchildren (pp. 9–30). New York: Springer-Verlag.

Reznick, S., & Goldfield, B. (1992). Rapid change in lexical de-velopment in comprehension and production. Develop-mental Psychology, 28, 406–413.

Schlesinger, H. S., & Meadow, K. P. (1972). Sound and sign:Childhood deafness & mental health. Berkeley: University ofCalifornia Press.

Slobin, D. (1973). Cognitive prerequisites for the developmentof grammar. In C. A. Ferjuson & D. I. Slobin (Eds.), Studiesof child language development. New York: Holt, Rinehart &Winston.

Slobin, D. (1985). The cross-inguistic study of language acquisition(vols. 1, 2). Hillsdale, NJ: Lawrence Erlbaum.

Stuckless, E. R., & Corson, H. C. (Eds.). (1994). Educationalprograms for deaf students. American Annals of the Deaf,139, 173–238.

Supalla & Newport, E. (1978). How many seats in a chair? Thederivation of nouns and verbs in American Sign Language.In P. Siple (Ed.), Understanding language through sign lan-guage research (pp. 91–132). New York: Academic Press.

Tyack, D., & Ingram, D. (1977). Children’s production and com-prehension of questions. Journal of Child Language, 4,211–224.

Received February 8, 2001; revisions received July 27, 2001; ac-cepted August 1, 2001

Normative Data for American Sign Language - Journal of Deaf

Documents

Transcript of Normative Data for American Sign Language - Journal of Deaf