TITLE: Relations between language input and the semantic structure of lexical terms in the acquisition of lexical meaning.
AUTHOR: Miguel Ángel Galeote Moreno.
AFFILIATION: Málaga University (Spain).
The purpose of this work is to present the results of an exploratory study done in order to clarify some unexplained results obtained in a previous one, (Galeote, 1994, 1996) about the acquisition of the meaning of several terms related to spacial representation, specifically dimensional adjectives. First, we think it necessary to review some theoretical and empirical issues related to these adjectives and to the results in our study regarding their acquisition.
The structure of the lexical meaning of dimensional adjectives (“grande, alto, largo, ancho, grueso, profundo” -big, tall, long, wide, thick, deep- and their correspondent antonyms) has been the object of an intense theoretical and empirical controversy during the last two decades. Traditionally, a componential structure was presupposed for the meaning of these adjectives within a clasical view of meaning (Carey, 1982). According to such a view, the meanings of the words can be broken down into features or minimal components of meaning. However, the enormous amount of contradictions observed in the empirical arena forced researchers to turn away from this structure and to abandon the study of this lexical field altogether.
We decided to look for a different approach, taking into consideration that the contradictions found in the empirical arena called for a reconsideration instead of an abandonment of the theme. A complete review of the psycholinguistic research of these terms, enabled us to identify a series of theoretical and empirical inconsistencies that questioned the validity of the research done in this area. Of all these inconsistencies, the most crucial one was the inadequate semantic description of the dimensional adjectives, based on the interpretation of E.V. Clark (1972, 1973a) of the analysis done by Bierwisch (1967) for dimensional adjectives in German. For example, while E.V. Clark offers the same semantic description for the corresponding adjectives to the dimensions of width, thickness and depth, considered by her the most complex, Bierwisch offers semantic definitions totally different for each one of them. Something similar happens with the rest of the terms (see Galeote, 1994). As a result, the order of complexity which E.V. Clark proposed for these adjectives, according to her Semantic Feature Hypothesis (E.V. Clark, 1973a), is very different to the result offered in Bierwisch’s original analysis.
This fact, and others that would be impossible to address here (see Galeote, 1994 for a complete review), lead us to develop an alternative semantic description of the dimensional adjectives that would take into account their different meanings according to their different uses in comparing objects. Such a semantic description was done taking into account the semantic analysis of different authors (H. Clark, 1973; Corrales, 1977; Goede, 1989; Greimas, 1970; Lang, 1989; Lyons, 1980; Teller, 1969; etc.).
The next step, was to empirically verify the process of acquisition of these adjectives based on the semantic description we developed (Galeote, 1994). In order to predict the order of acquisition of these terms, we took on a new concept of semantic complexity. This concept of complexity, different from the one described in the classical research (i.e. E.V. Clark, 1973a), took into consideration (1) the inherent complexity of the different semantic features that make up the meaning of these adjectives (for example, in virtue of the perceptive prominence of the verticality, the semantic feature ‘vertical dimension’, defining the adjective “alto” -high/tall- is less complex than the semantic feature ‘horizontal-frontal dimension of objects with a front’, present in the definition of some of the uses of the adjective “ancho” -wide-, because in this case, apart from distinguishing between the horizontal-frontal and horizontal-lateral dimensions of the objects, it is necessary to grasp the criteria to determine the front of a particular object) and, (2) the structural relationships that these features maintain with each other (the semantic features that make up the meaning of dimensional adjectives, do not follow any lineal order, but establish a series of dominant relations between them in function of the characteristics of the objects to which they are applied).
All this, allowed us to identify a very different order of difficulty of these adjectives from the one used in the previous studies. In short, based on this order of dificulty, the following order of acquisition was predicted (Galeote, 1994): size > height-length > thickness > width > depth (the last not being included in our research). With respect to the positive (big, tall, long, etc.) vs. the negative ones (little, short, etc.) terms of these spatial dimensions, we predicted that they would be acquired at the same time as there would be no differences in semantic complexity between them.
This order of difficulty was corroborated by the results of our experimental work (Galeote, 1994), in which we examined the comprehension of these adjectives by children between 2;9 to 6;0 years of age. The only unexpected result of this study was that children from the younger age group (2;9 to 3;6 years, mean 2;11) had a higher knowledge regarding positive adjectives belonging to the dimensions “altura” (height) and “longitud” (length) (“alto” -tall- and “largo” -long-) than negative ones (“bajo” and “corto” -both of them short in English-).
To explain these results, we considered the two most common explanations proposed in previous studies in which similar results were obtained: (1) children’s perceptive preferences towards bigger objects (E.V. Clark, 1973b; Richards, 1979) and (2) the frequency of use of dimensional adjectives. The first explanation, which claims that since children have a preference towards bigger objects, they would naturally choose the objects representing the positive adjectives (the tall vs. the short palm tree, the long vs. the short table, the big vs. the small house, etc.) and thus make more errors with negative adjectives, was not confirmed in Galeote (1994).
The second explanation refers to the frequency of use of these adjectives using the normative study done by Kucera and Francis (1967) for the English language and states that the frequency of use of these adjectives is a better predictor of acquisition than their semantic complexity (Blewitt, 1982; Carey, 1982; Ingram, 1989; etc). Our results ruled out this explanation. Thus, the correlation obtained between the frequency of use and order of acquisition of these adjectives, did not appear to provide a definite explanation[1]. In spite of this, the higher frequency of use of positive adjectives as compared to the negative ones in the normative studies, made us think about the possibility that this difference could have an effect on the performance of 3 year old children with the positive adjectives belonging to the dimensions of “altura” (height) and “longitud” (length).
In order to check this possibility, we decided to adopt a different strategy, one that would take into account the way in which adults talk to children, instead of focusing on the frequency of use in normative studies. Several studies indicate that the way adults talk to young children has certain peculiarities such that these frequencies are very different from the ones reflected in normative studies which have been taken as a reference point. One of the characteristics of the language that adults use with children is its simplification, which would be reflected in the use of shorter phrases, less subordinate clauses, etc. (Ingram, 1989; Kaye, 1980; Snow, 1977; etc.). In the field of dimensional adjectives, this simplification could be reflected by a generalized use of the adjectives “grande” (big) and “pequeño” (little) (Ravn & Gelman, 1984). This has been confirmed by Robb and Lord (1981) in parents of children aged 2 and 3. Our experiment was therefore designed to verify whether there was a differential behaviour among adults in terms of the use of positive vs. negative adjectives in child directed vs. adult directed speech. We thus formulated the following hypothesis: the better performance of the children with the positive terms belonging to the dimensions of “altura” (heigh/tallness) and “longitud” (length), compared to the negative terms, was due to their higher frequency of use in child directed adult speech and not to their lower semantic complexity.
Method
Subjects.
Our study was done with 20 adults (13 women and 7 men), parents of children between the age 3 and 4. In order to avoid bias, none of the children had been involved in the comprehension experiment. The socio-cultural level of these adults was similar to the adults participating in our previous study and they were also living in the same area. The sociocultural level of these adults was average medium-low. The men had an intermediate school education, and were required to posess some qualification in their jobs. The majority of women (10) were housewives. All of the women had at least primary school education. The participants were between 28 and 40 years of age, with a mean of 32 years. All of them lived in the town of Málaga (Spain).
Materials.
Ten pairs of objects, 5 for each dimension studied, were used. For the dimension of “altura” (height) the following pairs of objects were used: houses, tables, palm trees, glasses and pictures. For the dimension of “longitud” (length) the following pairs were used: tables, wooden blocks, pencils, tubes and roads. The wooden blocks and the tubes were real materials, the rest were graphic representations of the actual objects. To ensure the application of the adjectives, the objects differed significantly in the dimension of reference. That is, the extension of the dimension of the object that represented the negative pole was 2/3 of the extension of the same object representing the positive adjective (i.e., the “bajo” -short- object of the pair of “palm trees” was 2/3 of the “alto” -tall- object in the dimension of “altura”). The testing materials used for this study were exactly the same as the ones used in the previous study in order to compare results.
Procedures.
In order to identify the possible differences in adult language depending upon to whom language was directed, the task was divided in two parts, following a similar procedure to the one developed by Anglin (1977) in a study about the acquisition of hierarchically related terms. In the first part, adults had to produce the terms as if they would be talking to a child (not present), and during the second part they had to produce the terms talking to an adult (experimenter). Like in the experiment done with children, the order of presentation of the objects was altered for each pair, for each presentation of pairs and, for the tasks.
Results and discussion
Results of this study are shown in table 1. The subject’s responses were classified into three categories: elicitation of the appropriate term (“alto-bajo” -tall-short-), elicitation of general terms (“grande-pequeño” -big-little) and errors (for example, a dimensional error using “alto” -tall- to mean “largo” -long- or complete mis-use of the word, for example, calling a long object “gordo” -fat-).
Table 1
Percentage of appropiated, general and wrong terms produced for the adult subjects as a function of who is being spoken to.
As can be seen in table 1, results confirmed in the first place, the existence of clear differences in the type of language used, as a function of to whom it was directed. If language was directed to an adult, the subjects used almost exclusively appropiate terms; if language was used to adress a child, a significant production of general terms is observed, confirming the simplification of language by adults suggested by previous authors (Ravn & Gelman, 1984 and Robb & Lord, 1981). The virtual absence of errors indicate that adults correctly identified the relevant dimension of reference.
However, the most interesting aspect of these results is the distribution of frequencies in relation to the positive and negative terms of these dimensions when language was used to address a child. While positive terms received a higher percentage of appropriate terms in relation to the percentage of general terms used, the opposite pattern is observed with negative terms. Consequently, this data suggests that children would hear appropriate positive terms more frequently while negative terms would be heard with less frequency, being replaced by the adult with the term “pequeño” (little). Young children could get confused by the use of the term “pequeño” (little) instead of “bajo-corto” (short) learning the incorrect contrast “alto-largo / pequeño” (tall-long / little) and delaying the learning of the correct (“alto / bajo” and “largo / corto”). In other words, children could learn the positive terms “alto-largo” before their correspondent antonyms “bajo-corto”. In fact, during the comprehension exam and when confronted with the statement “dame el x más alto-largo” (give me the taller-longer x), many children frequently said “éste (this one), [taking the appropriate object], éste es el más ‘alto -largo’ y éste es el más ‘pequeño’ (chico)” (this is the taller one and this is the smaller one).
These results suggest that the type of lexicon used by adults to address a child influences the pattern of acquisition encountered in 3 years old children concerning their better knowledge of positive adjectives in the dimensions of height and length. However, this effect is rapidly overcome and children acquire the correct meaning of these adjectives, as well as the corresponding negatives (“bajo” and “corto” -short-), by the following age group (3;9) (Galeote, 1994, 1996). Thus, as the effect of speech directed to children is restricted to the adjectives “alto” (tall) and “largo” (long), and only for the youngest age group, the inherent semantic complexity of the dimensional adjectives, based on the semantic description we developed, seems to be a better factor in determining their acquisition than environmental language.
These results forced us to inquire the reason why environmental linguistic input could have a stronger influence on younger age groups. A possible factor could be the limitations in linguistic resources of children at this age, and their dependence on the linguistic model provided by adults which could have repercussions on the acquistion process, including that of meaning (Benelli, 1983; Ingram, 1989; Peraita, 1988; etc.). This possibility seems to be supported by the results of several experimental studies (Nelson and Bonvillian, 1973; Nelson, 1973; Della Corta, Benedict, and Klein, 1983; etc.). In addition, a number of studies on categorization have corroborated the influence of parent’s language (how parents label/describe objects for children) on the conceptual knowledge and delimitation of a category in small children (Callanam, 1985, 1990; Shipley, Kuhn, & Maden, 1983; etc.). A determining factor could be the role that parents in providing children a credible source of information regarding the content and limits of a category (Shipley, 1989).
Even though the simplification of language to which we referred, explains the higher percentage of general responses obtained when the language is directed to children, the differential behaviour of adults in relation to negative and positive adjectives, when semantic complexities between them has been ruled out, needs to be explained. One possible explanation is that the differences are an effect of the norm of use of the terms corresponding to the dimension “altura” (height) when referring to people in Spanish language (Corrales, 1977). According to this author, “grande” (big) and “pequeño, chico” (little, wee), seem to be applied frequently instead of “alto-bajo” (tall-short), to quantify the dimension of “altura” (heigth). This phenomena, however, does not affect both adjectives in an equal way, but follows a trend toward the positive adjective as obtained in our study. That is, according to Corrales, “pequeño” would be used more frequently to refer to “bajo” than “grande” to refer to “alto”. Even though Corrales refers to the dimension “altura” specifically, it is probable that according to the result of our exploratory study, the same thing happens to the dimension “longitud”. This effect would be restricted to informal registers, specifically familiar, of the spoken language (Corrales, 1977). It is precisely in this environment in which young children develop. This phenomena could also be reinforced by the fact that, when speaking to children, and for affective reasons, the use of diminutives is often observed in addition to the use of the adjective “pequeño” to address a great number of objects in the child’s surroundings.
Conclusions
Although it is necessary to continue investigating, the results of our exploratory study indicate the existence of clear differences between the language used by adults to address a child vs. the one used to address another adult. Even though the differences seemed not to provide a definite model to explain in a general way our results (they were restricted to only the adjectives “alto” -tall- and “largo” -longitud-, and only for the youngest age group), we can not rule out that this is not the case in the lexical terms appertaining to other semantic fields. That is, the order of acquisition of some lexical terms could be better explained by the linguistic model used by the adult with the child than by the structure or semantic complexity of the terms.
In this sense, it is necessary to consider the influence of the linguistic input in the process of language acquistion, and in particular, the influence of the linguistic model the adult offers to the child. This experimental strategy could be useful in order to determine to what extent the course of linguistic and conceptual development is innately programmed or is flexible and responsive to the particular properties of the environment (Huttenlocher, Smiley, & Ratner, 1983).
References
Anglin, J.M. (1977). Word, Object and Conceptual Development. New York: Norton.
Benelli, B. (1983). Effects of language acquisition on the development of children's knowledge. Cahiers de Psychologie Cognitive, 3, 2, 203-221.
Bierwisch, M. (1967). Some semantic universals of german adjectivals. Foundations of Language, 3, 1-36.
Blewitt, P. (1982). Word meaning acquisition in young children: a review of theory and research. In H. Reese & L. Lipsitt (Eds.), Advances in child development and behaviour (Vol. 17). New York: Academic Press, pp. 140-195.
Callanam, M.A. (1985). How parents label objects for young children: the role of imput in the acquisition of category hierarchies. Child Development, 56, 508-523.
Callanan, M.A. (1990). Parents’ descriptions of objects: potential data for children’s inferences about category principles. Cognitive Development, 5, 101-122.
Carey, S. (1982). Semantic development-State of the art. In E. Wanner & L. Gleitman (Eds.), Language Acquisition: The state of the art. Cambridge: University Press.
Clark, E.V. (1972). On the child’s acquisition of antonyms in two semantic fields. Journal of Verbal Learning and V. Behav., 11, 750‑758.
Clark, E.V. (1973a). What’s in a word? On the child’s acquisition of semantics in his first languaje. In T.E. Moore (ed.), Cognitive development and the acquisition of language. New York: Academic Press.
Clark, E.V. (1973b). Non linguistic strategies and the acquisition of word meanings. Cognition, 2, 161-182.
Clark, H. (1973). Space, time, semantics and the child. In T.E. Moore (ed.), Cognitive development and the acquisition of language. New York: Academic Press, pp. 27-63.
Corrales, C. (1977). El campo semántico «dimensión» en español. Santa Cruz de Tenerife: Ed. Aula de Cultura de Tenerife.
Della Corta, M., Benedict, R., & Klein, D. (1983). The relationship of pragmatic dimensions of mother’s speech to the referential-expressive distinction. Journal of Child Language, 10, 35-43.
Galeote, M.A. (1994). La Adquisición de los Adjetivos Dimensionales y la Estructura del Significado Léxico (Doctoral Dissertation). Madrid: Universidad Nacional de Educación a Distancia.
Galeote, M.A. (1996). La Adquisición del significado de los Adjetivos Dimensionales: Revisión crítica y nuevas consideraciones. In M. Pérez Pereira (Ed.) (1996), Estudios sobre la adquisición del castellano, catalán, eusquera y gallego. Santiago de Compostela: Servicio de Publicacións e Intercambio Científico Campus Universitario Sur.
Goede, K. (1989). Language acquisition and development of children’s «bigger» and «more» judgements. In M. Bierwisch & E. Lang (Eds.), Dimensional adjectives: Grammatical structure and conceptual interpretation. New York: Springer-Verlag., pp. 419-432.
Greimas, A.J. (1970). La semántica estructural. Madrid: Gredos.
Huttenlocher, J., Smiley, P., & Ratner, H. (1983). What do word meanings reveal about conceptual development?. In T.B. Selier & W. Wannenmacher, Concept development and the developmnet of word meaning. Berlin: Springer-Verlag.
Ingram, D. (1989). First Language Acquisition: method, descripction and explanation. Cambridge: Cambridge University Press.
Juilland, A., & Chang-Rodríguez, E. (1964). Frequency Dictionary of Spanish Words. Londres: Mouton & Co.
Kaye, K. (1980). Why don’t we talk baby talk to babies. Journal of Child Language, 7, 489-507.
Kucera, H., & Francis, W.N. (1967). Computational analysis of present-day American English. Providence: Brown University Press.
Lang, E. (1989). The semantics of dimensional designation of spatial objects. In M. Bierwisch & E. Lang (eds.), Dimensional Adjectives: gramatical structure and conceptual interpretation. pp. 263-418. New York: Springer-Verlag.
Lyons, J. (1980). Semántica. Barcelona: Ed. Teide, S.A. (2th Edition) (original: Semantics. Cambridge: Cambridge University Press, 1977).
Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society of Research in Child Development, 38, nº 149.
Nelson, K.E., & Bonvillian, J. (1973). Concepts and words in the 18-month-old: acquiring concept names under controlled conditions. Cognition, 2, 435-450.
Peraita, H. (1988). Adquisición del Lenguaje. Madrid: Cuadernos de la UNED.
Ravn, K.E., & Gelman, S.A. (1984). Rule usage in childre’s understanding of “big” and “little”. Child Development, 55, 2141-2150.
Robb, M., & Lord, C. (1981). Early uses of «big» and «little» by mothers and children. Papers and Reports on Child Language Development (Stanford University), 20, 108-115.
Richards, M.M. (1979). Sorting out what’s in a word from what’s not: evaluating Clark’s semantic features acquisition theory. Journal of Experimental Child Psychology, 27, 1‑47.
Shipley, E.F. (1983). Categories, hierarchies and induction. Manuscript no published (cited by Callanam, 90).
Shipley, E.F., Kuhn, I.F., & Madden, E.C. (1989). Mothers’ use of superordinate category terms. Journal of Child Language, 10, 571-588.
Snow, C.E. (1977). The development of conversation between mothers and babies. Journal of Child Language, 4, 1-22.
Teller, P. (1969). Some discusion and extension of Manfred Bierwisch’s work on german adjectivals. Foundations of Language, 5, 185-217.
