Excerpt
Table of Contents
1. Introduction
2. An Overview on the Topic
2.1 Definition of Lexical Categories
2.2 Radford’s Theory
2.2.1 The Four Stages in Language Acquisition
2.2.2 Categorization and Projection in Language Acquisition
3. Data
3.1 A Child’s Noun-System
3.2 A Child’s Verb-System
3.3 A Child’s Preposition-System
3.4 A Child’s Adjective-System
3.5 The Lack of Functional Elements in Early Child English
3.6 Summary
4. Other Opinions and Comments
5. Conclusion
6. References
Affirmation
I hereby confirm that this paper was written by myself and that all direct and indirect quotations from other sources have been documented appropriately.
1. Introduction
In terms of Universal Grammar, our language is made up out of grammatical categories, namely lexical categories and functional categories (compare 1997 Radford: 29-60). What are grammatical categories? When little babies enter our world - do they carry categories within them? What are their first words? Do they belong to a certain category and is the child aware of that? How do children’s first word-combinations look like? Are there similarities to the adults’ language?
This paper suggests answers to these questions. Since every language has a more or less different grammar, the focus stays on the English language. This makes it possible to go into detail. Moreover, the concern lies in early child English up to the age of about two years. The overall claim is that children up to that age only produce words and word combinations belonging to thematic or lexical classes. This is also Radford’s thesis presented in his book Syntactic Theory and the Acquisition of Syntax (1990).
To be able to understand what lexical categories are, the following chapter provides a definition of grammatical categories. Afterwards, Radford’s theory will be described. In the next section, examples of children up to the age of about two years are given and analyzed concerning the occurrence of lexical categories. Other opinions will be presented and discussed in the following section. The paper closes with a conclusion.
2. An Overview on the Topic
2.1 Definition of Lexical Categories
Lexical categories and functional categories build up the two grammatical categories each language consists of. Each word can be assigned to one of the two groups. To define what is meant by lexical categories it is therefore necessary to explain functional categories, too. In the following, a brief description of which elements belong to which category and major differences between the two will be given.
Lexical categories consist of nouns, verbs, adjectives, and prepositions (compare Cook, Newson 1988: 136 and Radford 1990). If we look into a lexicon,we will find that these four classes build up the majority of our vocabulary (Cook, Newson 1988: 136). Haegeman and Guéron count adverbs as a fifth lexical category but also mention that this is a problem because adverbs like carefully and long, for example, have the properties of adjectives (1999: 58). Radford, who plays an important role in this paper, does not list adverbs as a class of lexical categories (compare 1990), therefore and to avoid problems, we will consider only the first four groups to be lexical categories. (Nouns, verbs, adjective, and prepositions will not be defined here, because of the assumption, that a reader of a research paper on linguistics is familiar with the according definitions.)
Haegeman and Guéron state that functional categories are determiners, consisting of articles (the, this, that, etc.) and demonstrative pronouns (my, your, his, etc.), and auxiliaries, such as will, may, and must (1999: 60-62). Radford adds the category complementizers, which consists of words like that, if, for, and whether (1990: 49). They introduce a complement clause, like that in “I know [ that the sun is shining]” (Radford 1990: 49). The bracketed part of the sentence is the complement clause (compare Radford 1990: 49). Radford, Cook, and Newson add tense and agreement inflections like the past tense -ed or the plural -s and the infinitival to and consider them, together with the auxiliaries, to make up one category, thereby referring to Chomsky (compare Cook, Newson 1988: 148 and Radford 1990: 49-50). This category is then called inflection. To sum up, functional categories consist of determiners, complementizers, and inflections.
The main difference is supposedly, that lexical categories are open and functional categories are closed classes. This means that lexical category words are impossible to be counted and new words can easily be invented. In contrast, it does not take long to list all functional category elements, they are fixed, and it is impossible to create new functional elements. (Compare Radford 1990, Cook and Newson 1988, Haegeman and Guéron 1999)
Further differences are that functional elements, in opposition to lexical ones, are, in general, phonologically and morphologically dependant, select only one complement and no arguments, and they cannot be separated from their complement. Functional elements are not descriptive, do not belong to a theme, or cannot be connected to the real world. They serve in a grammatical way and only occur in combination with parameters, whereas lexical elements do not. (Compare Radford 1990, Cook and Newson 1988)
2.2 Radford’s Theory
Radford claims that the development of grammar in young children’s language starts by building up grammatical categories as those explained above (1990: 20). How and when grammatical categories are developed in early child English, according to him, is what this chapter’s concern lies upon.
2.2.1 The Four Stages in Language Acquisition
According to Radford, a child undergoes four main stages while learning a language (1990: 20). The first one is the prelinguistic stage, in which the child stays from birth up to the age of one year (Radford 1990: 20). Here the child utters no words (Radford 1990: 20) but some sort of “babbling” (Radford 1990: 21). This stage occurs before the child’s first words (Radford 1990: 20). If we listen to a child in the prelinguistic stage we will hear sound patterns like “/ma/ , /wa/ , /ba/,” having nothing to do with words but with the phonology of the language the child is confronted with (Radford 1990: 21).
At the age of 12 to 18 months, the child is in the single-word stage (Radford 1990: 20). It utters isolated single words (Radford 1990: 20). These are the child’s “first lexical items” (Radford 1990: 21). They are simple combinations of a word and its meaning (Radford 1990: 21). If the child mis-segments the language it hears, this will lead to mis-lexicalization, that means, the child will form wrong combinations of word and meaning (compare Radford 1990: 21). According to Radford, grammar is not present at this point in the child’s language (Radford 1990: 21-22). The words are acategorial at this stage, which means, out of the child’s view, they do not belong to grammatical categories yet (Radford 1990: 22). Radford gives an example of a tourist that is given an expression to say good-bye in another language. He argues that the tourist knows how to use the expression but does not know what it really means. He cannot divide it up into its grammatical units (Radford 1990: 22-23). Objects get names in the single-word stage; an example of a child’s expression to that age would be the single word “[a]pple” (Radford 1990: 20-21).
Stage three in the language-learning process of young children is the early multi-word stage, according to Radford (1990: 20). This stage occurs at the age of one and a half and lasts until the age of two years (Radford 1990: 20). Here, the child combines two to four words to form a sense-making unit (Radford 1990: 20). A child at the third stage might say something like “Jane want apple,” where we can find a clear subject-verb-object order (compare Radford 1990: 21).
The last stage Radford mentions is the later multi-word stage, which appears at the age of two years and lasts until the age of two and a half years (1990: 20). A child in this stage puts up to five or seven words together to form a meaningful unit (Radford 1990: 20). An example could be “I want to have an apple” (Radford 1990: 21). Radford stresses the use of functional elements like the pronoun I, the infinitival particle to, and the determiner a, in the example-sentence (1990: 21).
2.2.2 Categorization and Projection in Language Acquisition
Categorization, according to Radford, has its start at the age of one and a half years, namely at the beginning of the early multi-word stage (Radford 1990: 21). As mentioned above, in the single-word stage words are acategorial to the child (Radford 1990: 22).
To find out that this claim is true, one should look at an adult’s language and compare it to the one of the child (Radford 1990: 23). The first thing to take into consideration should be the morphological evidence, e.g. the use of inflections (Radford 1990: 24). We need to find out whether the child attaches inflections productively, selectively, contrastively, and appropriately or not (Radford 1990: 24). Radford mentions two things that need attention here: One is the acquisition of the use of inflections and the other is mastery of the use of inflections. Acquisition is reached, when inflections are only attached to appropriate category word-stems, while mastery is achieved, when inflections are always and only attached to the appropriate category word-stems (Radford 1990: 24). Radford claims, that acquisition already shows that categorization has taken place (1990: 24).
The next aspect in finding out about categorization in early child language is to look for syntactic evidence (Radford 1990: 25). Direct evidence would come from distribution (Radford 1990: 25-26). We should look at the child’s speech and find out whether word-classes are combined appropriately or not (Radford 1990: 25- 26). Indirect but also distributional evidence can be found by tests, like presenting incomplete sentences that have to be completed by the child or by asking wh- questions (who-, where-, and what-questions), where the child has to give answers that are of a phrasal nature rather than a clausal one, and we should see whether the responses are semantically and categorically correct (Radford 1990: 26-30).
Another sign for the presence of categorization can come from the use of redundancy rules, that is, for example, the fact that words like comb and brush function as well as nouns as as verbs. Once a child is aware of this, categorical overextension can take place. In the considered case, this would appear when a child turns a noun into a verb like in “Mummy trousers me,” said by a 27 months old child. (Radford 1990: 37)
Another rule the child learns is that noun, verb, and prepositional phrases are formed within a cross-categorical symmetry. In English, we combine a noun with a constituent to build up a noun phrase, a verb and a constituent to make up a verb phrase, and so on. (Radford 1990: 44)
Radford’s proof for categorization is given by Allison’s spontaneous speech transcripts at the age of 16, 19, 20, and 22 months, material from Lois Bloom in One Word at a Time (1973) (compare Radford 1990: 31). Radford claims that there are four major categories in Allison’s speech: nouns, verbs, prepositions, and adjectives (1990: 31). There is no categorization at the age of 16 and of 19 months, but clear evidence of categorization at the age of 20 and of 22 months (Radford 1990: 31).
Morphological evidence for this claim, at the age of 16 and 19 months, is that there is no occurrence of inflections in Allison’s speech to that time (Radford 1990: 31). Syntactic evidence comes directly from the fact that there are no word combinations at the age of 16 months and only a couple of combinations at the age of 19 months (Radford 1990: 31-32). Indirect syntactic evidence comes from wh-questions, because there is only one example of a completion sentence (Radford 1990: 32). The wh-questions show that categorization has not taken place at the age of 16 and 19 months, because the answers by Allison belong to the wrong categories (Radford 1990: 32-33).
At the age of 20 and 22 months, the number of nouns, verbs, prepositions, and adjectives has increased in Allison’s speech, according to Radford (1990: 33). Proof for the beginning of categorization comes morphologically from the use of inflections at the age of 20 months (Radford 1990: 33). Direct syntactic evidence for categorization comes from the fact that noun and verb phrases can be found in Allison’s speech transcript at the age of 20 months and increase in the transcript at the age of 22 months (Radford 1990: 33-34). Indirect syntactic evidence is given through wh-questions which are now correctly answered, considering categorical correctness (Radford 1990: 34).
Four different categories occur in Allison’s speech; these are nouns, verbs, adjectives, and prepositions (Radford 1990: 34). Radford claims that there is an awareness of the morphological and syntactic characteristic of these categories at the same time (1990: 34-35). His overall result is that categorization must have started by the age of 20 months in Allison’s case (1990: 35).
Radford’s thesis is that “all constituents in early child English will have a purely lexical categorical status, so that children at that stage will make no productive use of functional constituents at all” (1990: 47). Throughout his book he wants to give empirical proof on this (1990: 47). One example of how to find evidence lies in imitative speech: Reproduction can only be correct in those cases in that the according morphosyntax has been acquired (compare Radford 1990: 54). Radford states that only lexical category items are reproduced and that functional ones are omitted at first (1990: 55).
There are three stages concerning categorization in the language acquisition- process, according to Radford (1990: 48). The first one of these is the precategorical stage. It appears during the single-word stage, namely at the age of one year until the age of one and a half years (Radford 1990: 48). With more or less than 20 months of age, the child enters the lexical stage, where it uses lexical word categories together with their phrasal projections (Radford 1990: 48). Only after having gone through these two stages, the child will enter the functional stage at about the age of two years (Radford 1990: 48). At this stage the child will begin to use functional word categories and will build phrasal projections with them (Radford 1990: 48).
To understand what phrasal projections are, one needs to know what X-bar syntax is. Radford gives a brief explanation of it on page 50 (1990): The overall point is that word categories are projected into phrasal categories. This happens symmetrically and is possible for all categories, which leads to the term cross- categorical projection symmetry. A word, labelled X, out of any category is named head of a phrase, if it is projected into an X-phrase by adding modifiers. This happens in the following way: The word X plus one or more complements make up an X’ = X-(single)-bar construction. X-bar plus an adjunct make up another X- bar. X-bar plus an appropriate specifier finally make up an X’’ = X-double-bar = XP = full X-phrase. In English we have a general phrase structure looking like the following, concerning X-bar syntax:
[X’’xspecifierx[X’xadjunctx[X’x[Xxhead]xcomplement(s)]]].
By no means does a phrase have to consist of a specifier, an adjunct and a complement. The only obligatory part of a phrase is its head. When analyzing single phrases, X is replaced by N if the head is a noun, by V if it is a verb, by A if it is an adjective, and by P if it is a preposition. What is said in this paragraph is a very simplified definition of X-bar syntax, but it is sufficient for our purposes. Examples are given in the next chapter of this paper.
Categorization and projection, according to Radford, develop at the same time in early child English (1990: 57). He claims that this is inherent and like an instinct (1990: 57). What children need to do first is to set configurationality and linearity parameters (Radford 1990: 58-59). These are different in every language, and children have to set them according to the language they are exposed to (Radford 1990: 58-61). Radford claims that this already happens when the children are very young (1990: 59, 61). The English language is strongly configurational, because its word order is relatively fixed (Radford 1990: 59). It is a head-first language that means that the head of a phrase always precedes its complement (Radford 1990: 59). More over it is an adjunct- and specifier-first language, that means that the head of an English phrase always follows its adjunct and specifier (see the X-bar structure of an English phrase above). Since this paper only considers early child English, the different word-order and head positions of other languages will not be mentioned here.
Setting the parameters is not very difficult, since they are only choices of first or last and they are fixed within one language according to Universal Grammar (Radford 1990: 61). Once the parameters are set, Radford supposes that it should be an easy task to project word categories into phrases (1990: 61).
[...]