INFANT BEHAVIORAND DEVElOPMENT17,
19-129 (1994)
PIERREA. HALLÉ BÉNÉDICTE DE BOYSSON-BARDŒS Laboratoire de Psychologie Expérimentale. Paris This experiment examineswhether Il-month-old and 12-month-oldinfants are able to recognize familiar words in a situation yielding no extralinguistic cues.Two experimentswere run to compare infants' interest for familiar words. chosen in the early productive vocabulary of young infants. against f".lrewords infrequent in French usage.80th experimentsuseda preferenceparadigm in which prcferencewas indexed by attention span.Lists of familiar woro.'Iwere auditorily presentedto each child in the absenceof any possible referent object. A preferencefor familiar words was round to be very consistent in 12-month-oldsand just emerging in Il-month-oids. These results were interpreted as revealing the existenceof a developing receptive lexicon by 11 months of age.
languageacquisition receptivelexicon ward recognition familiar words phoneticcomplexily
tion when it is considered that production of intended meaning by means of words requires at least partial understanding of these words. Indeed, a sizeable body of studies based on observations from naturalistic settings, diaries, or parental reports have consistently reported a substantial lag of production behind comprehension. Goldin-Meadow, Seligman, and Gelman (1976) found a developmental shift at about 2 years from a "receptive stage" to a "productive stage" where production and comprehensionlexicons moved into alignment. ln a longitudinal study of children from 9 to 18 months, Benedict (1979) found that comprehension development is ahead of production development by several months; she located the onset of word comprehension at around 9 months and the onset of word production at around 12 months. The two studies mentioned also indicate that words produced with meaning are, as is logical, understood. Snyder, Bates, and Bretherton's (1981) data, based on The authors gratefully acknowledge support from a intensive maternaI interviews, globally supgrant from le Ministère de la Recherche et de la ported Benedict's (1979) estimation: 13Technologie (1990), from la Fondation de la Recherche month-olds had a receptive lexicon of about 45 Médicale (1989), and from the JapaneseSociety for the words and a productive lexicon of about Il Promotion of Scienceto P.A.H. (1989-1990). Gratitude is woras. Clark aridHechf (1983) have suggested expressedto the infants and parents who participated in this study. to Catherine Marlot and to Hugo Berroyer who that observational studies, diary studies, or helped as researchassistantson the project. parental report studies tend to overestimate Correspondenceand requests for reprints should be comprehension capacities in young children. sent to Pierre A. HaIlé. CNRS Paris. Laboratoire de To put it more properly, those studies do Dot Psychologie Expérimentale.28 rue Serpente.75006 Paris. permit us to separate the contributions to comFrance. The emergence of a lexicon is of paramount importance in the process of language acquisition. Developing a lexicon, however primitive, entails coding and storing word-sounds in some way for comprehension and perhaps in another way for production. When are coding and storing word-sounds flTst evidenced in infants? This is the question that motivates this experiment. During the last 20 yeaTSor so, contributions in the field of early word comprehension have generally pointed to the asymmetries between production and comprehension processes.Huttenlocher (1974) stressed that "[receptive language] involves recognition of words and reca// of objects, acts, and relations for which they stand, whereas [productive language] involves recognition of objects, acts, and relations and reca// of the words that stand for them" (p. 335). ln that light, word comprehension must logically be ahead of word produc-
119
120
Hailé and Boyssan-Bardies
prehension of linguistic and nonlinguistic cues. This is because children can rely on nonlinguistic knowledge in many situations to display "correct" responses.They also may respond to some features only of verbal commands disregarding other potentially contrastive features: Huttenlocher (1974) reported the example of Wendy, Il months, stopping to rip out pagesof a book when told"No!" by ber mo~therbut also when told "Yes!" with the same forbidding intonation. Indeed, in the studies mentioned and others, researchershave always strived to avoid misinterpretation of children's responses to speech, in particular, testing them contrastively as in the "no versus yes" example. Still, it seems that young children recognize sound patterns within sentencesin conjunction with specific contexts: They may store words as sound sequencesplus situation (Menyuk & Menn, 1979). Experimental settings, on the other band, may suffer the drawback of not eliciting children's responsive behavior as easily as naturalistic settings do. Do we get more reliable data from experimental studies? Experimental procedures that have been widely used involved a verbal command on one band and a visual presentation of referent objects (or events) on the other band. Thomas, Campos, Shucard,Ramsay, and Shucard (1981) tested 11- and 13-month-olds on their comprehension of reportedly unknown versus known abject words. ln each trial, infants were presented a word (known, unknown, or nonsense) while being shawn an array of four abjects among which were two referent objects for the known and unknown words. Thirteen-montholds 100kedproportionally longer to the referent of the known ward when that ward was spoken than when the nonsenseword was spoken. This result was not obtained for unknown words nor was it for either known or unknown words in the 11-month-old group (Behrend, 1987). Hence, a developmental shift seems to occur between Il and 13 months: Children become clearly able to recognize known words and match them with referent obj~cts. Oyiatt.
