1 The first 50 words: what are their constraints?. In: 24 th IALP Congress, Amsterdam 23-27 august, 1998. Program and Abstract Book. Amsterdam: Nijmegen University, 1998, 202. EXAMINING THE FIRST WORDS CONSTRAINTS* Leonor Scliar-Cabral (UFSC/CNPq), Brazil 1. Introduction This paper is particularly addressed to one issue: what are the constraints that govern the form of the children s first lexical items despite the facilities proportioned by Child Directed Speech (CDS)? If the se constraints are determined by the child s biopsychological and linguistic maturity, such as the absence of the internalized phonological and morphosyntactic subsystems of his/her native language, and consequently, by the absence of respective phonoarticulatory schemes which command their gestures, those items will show some common properties: 1) global items for each reference, representing the child s immediate needs; 2) absence of syntactic markers, both free and bound morphemes; 3) predominance of one word at a time ; 4) phonetic fluctuation of the same item; 5) private character of the child s variety; 6) two convergent inputs for the first items: the surrounding utterances to which the child is exposed and his/her productions at the babbling stage. The examples were taken from data video-taped by Giovanni Secco and Elena Nicoladis, collected twice a week, from two children, one a Brazilian Portuguese speaker, from 13 up to 15 months, and the other, an American bilingual English/Portuguese speaker, from 12 up to 14 months. Elena Nicoladis transcribed the English productions and Giovanni Secco the Brazilian Portuguese ones. 2. The first child s utterances Despite epistemological divergences which emphasize one factor or the other, most language acquisition researchers accept the interplay of three determinants of language development in humans: the biopsychological (more recently called innate guided learning, Gould and Marler 1987); the maturational and the
2 environmental (which pays attention either to the input characteristics or to the interactions between the child and the others). * A first version was presented at the IALP Child Language Committee, Amsterdam, 1998. Equal weight is attributed to the three determinants by the author, although in this paper I will focus on the maturational one. 2.1Global items for each reference representing the child s immediate needs 2.1.1 Phonetic fluctuation Extensive research on child s earlier phonetic discrimination empirically proves the preferences towards the acoustic parameters of his/her native language at the price of ignoring the acoustic cues of other sounds to which they were sensitive at birth: these studies appear to indicate a systematic decline in sensitivity to nonnative contrasts between 6 and 12 months of age, Jusczyk 1997, 80-1). However, it does not mean that the child at this age has already systematized the phonetic features of his/her native language, nor the phonotactic rules or the phonetically conditioned allophones, and it certainly does not mean that the child control the vocal gestures to produce them. What is certain is that the child is discovering the principle that a chain of sounds may be linked to a reference, the purpose of which is to meet his/her immediate needs. Three main characteristics arise from the comparison between the Brazilian and the American bilingual child arguing in favor of global (gestalts) schemes commanding the production of the first items: phonetic fluctuation of the same item, a high degree of assimilation in the same item and syllabic reduplication. The last two characteristics share mutual properties and will be discussed together. The first one demonstrates that the contrast among phonological units is not firmly established as the examples bellow show: Brazilian child Adult s Form Meaning American bilingual child Meaning English 1. [mej i u]~[ bi 1. [ n n ~[ n n:]
3 ~[ mi u] /bej i u/ kissy ~[ n n banana 2. [nene ne]~[nenene ne] /ne ne/ doll or flower 2. [ t i::]~[ti::]~[di::] cheese 3. [ patu]~[ tatu] / patu/ duck 3. [ m m ~[m Mama 4. [a lo]~[o eo e] /a lo/ Hello 4. [ ]~[ t]~[ t]~[ t] shut Observe the fluctuation of the same item particularly on the area of [+ or nasal] (item 1); place of articulation (item 3) in the Brazilian child. In English, besides the same fluctuation shown in Portuguese, such as [+ or nasal] (items 1 and 3); place of articulation and [+ or voice] (item 2), two remarkable aspects may be noticed: 1) the fluctuations of vowels (1, 3 and 4) and the child s fluctuation of word and syllable structures (1 and 4): item 4 is dramatic, showing the child s attempts. Examples 2 and 4 disconfirm the hypothesis that the preferred initial word structure is the trochaic. Again, the fact that in experimental settings, the child shows preference for listening to sequences with consistent rhythmic patterns and a fixed syllable order (Jusczyk, op.cit.: 97), as the findings of Morgan and Safran (1995) proved, does not mean that the child uses the same patterns while trying to produce his/her first words. Instead, what the data demonstrate is that tying the mental representation of a sound chain to meaning represents a tremendous and new difficulty for the child, who attempts to reach this relationship, using different word and syllabic structures, even some, the gestures of which are very difficult, as can be observed in example 4 of the Brazilian child [o eo e] (variant of [a lo]). Observe that example 4 is the child s attempt to imitate what Brazilian adults usually say at the beginning of a phone call: [a lo a lo], an utterance where the [+vocalic] feature is predominant. 1.1.2 Phonetic assimilations and reduplications Phonetic assimilation is more frequent to be found in the Portuguese productions. The following features are mentioned: [+nasal] (item 1 of the Brazilian child, [mej i u] ~[ mi u] both preserving place of articulation); [+coronal] (item 3 [ tatu], preserving manner of articulation [- continuous] and [-voice]); vocalic place of articulation [-low, -high] (item 4 [o eo e]). Examples of the American
4 child s English productions are 1 [ n n and 3 [ m m ], the last one of partial assimilation. Assimilation provokes partial or total reduplication, a process observed in CDS and in the children s own productions. Reduplication is another evidence that it is easier for the child to maintain the same or neighboring global vocal gesture, which is also evidence, that he/she is not combining isolated segments. The following examples are of total reduplications found in both children: Brazilian child, example 2 and American child, examples 1 and 3 (in one token); examples of partial reduplications: Brazilian child, examples 1, 3 and 4; American child, example3 (in one token). 2. Constraints The main constraints which block a child of producing more complex utterances from the beginning are not of peripheral nature, since at birth, he/she discriminates a large range of categorical acoustic parameters and also produces a large range of sounds. They can be better explained if we consider that very different amounts and areas of cortical activation are involved in structured circuits to account for verbal language processing and for the way lexical items, including regular and irregular verbs, are stored (Jaeger et al. 1996). In addition, infants are not equipped from birth with those ready circuits. As Lecours states (1983,176), the process of brain maturation is expressed by numerous chemical, biological and anatomical changes. An evolutionary prospective (Cosmides, Tooby, 1994:64,5) claims that there are specialized circuits for reasoning and regulatory mechanisms which organize the way we interpret experience, build up knowledge and take decisions, being multi-modular: these specialized circuits develop with growth and are not ready at birth. Special attention must be given to the circuits necessary for: 1) the stable internalization of the self opposed to the other(s) and to the object of knowledge, a condition for establishing the persons of discourse and, in consequence, illocution; 2) grasping grammatical meanings which by definition do not have any concrete
5 referential counterpart no matter what their overt manifestation is in different languages: free or bound morphemes and/or word order. The absence of the referred circuits around 12 to 14 months explains why the child is unable to use the persons of discourse and a proposition or sentence: he/she is limited to the space of reference of here and now (hic et nunc). The fact that grammatical morphemes usually are either weak bound segments or clitics only makes the task more difficult. Bibliographic References Cosmides, L. e Tooby, J. Beyond intention and instinct blindness: toward an evolutionarily rigorous cognitive science. Cognition,1994, 50(1-3):41-77. Gould,J.L. and Marler, P.(1987) Learning by instinct. Scientific American 256, 62-73. Jusczyk, P.W. (1997) The discovery of spoken language. Cambridge, London: Bradford and The M.I.T. Press. Jagger, J.J., Lockwood, A. L., Kemmerer, D.L., Van Valin Jr. R.D., Murphy, B.W. and Khalak, H.G. (1996). A positron emissio n tomographic study of regular and irregular verb morphology. Language 72 (3): 451-97. Lecours, A.R.(1983) Cerebral maturation and language acquisition. In AR. Lecours, F. Lhermitte and B. Bryans (eds.) Aphasiology. London, Baillière Tindall, 172-189. Morgan, J.L. and Safran, J.R. (1995) Emerging integration of sequentialand suprasegmental information in preverbal speech segmentation. Child Development 66, 911-936.