Grammatical Violations Research Articles

Learning artificial grammars with competitive chunking.

Abstract : When exposed to a regular stimulus field, for instance generated by an artificial grammar, subjects unintentionally learn to respond efficiently to the underlying structure: Miller (1958) reports that subjects memorize letter strings generated by an artificial grammar faster than randomly generated strings. Reber (1967) reports that, following rote memorization of exemplar sentences, subjects efficiently discriminate grammatical from non-grammatical strings. We explored the hypothesis that the learning process is chunking and that grammatical knowledge is implicitly encoded in a hierarchical network of chunks. Grammatical judgments are then based on the degree to which integrated representations of strings can be built using those chunks. We trained subjects on exemplar sentences while inducing them to form specific chunks. Their grammatical knowledge was then tested with a discrimination task. We found that subjects were less sensitive to grammatical violations that preserved their chunks than to violations that did not. We derived the theory of competitive chunking (CC) and found that is successfully reproduces, via computer simulations, both Miller's experimental results and our own. Keywords: Unintentional learning, Artificial grammars, Chunking, Perception(Psychology).

The ability of normal and language-impaired children to produce grammatical corrections

The ability of normal and language-impaired children to correct grammatical violations of word order was examined. Ten language-impaired and 10 linguistically normal subjects were sampled from the following age levels: 6, 7, 8, 9, and 10 years, resulting in a total of 100 subjects. Using a procedure designed to focus attention on language form, each of the children performed a task requiring the correction of 10 sentences containing word order errors. Normal 6-, 7-, and 8-year-olds performed significantly better than their language-impaired age-matched peers. In addition, the performance of language-impaired 9-, and 10-year-olds was superior to that of the younger impaired groups. In the normal groups, the only age level differences were produced by the 6-year-olds, who performed significantly more poorly than two of the older (8- and 10-year-old) groups.

Children's capacity to correct grammatical violations in sentences.

ABSTRACTPrevious studies of children's knowledge of grammatical well-formedness have typically involved eliciting acceptability judgements from children. In the present study, children were expected to employ their tacit knowledge of the rules of grammar to correct sentences containing grammatical rule violations. Two groups of children (mean ages of 5;5 and 6;4) were presented with ungrammatical sentences resulting from either word-order changes or morpheme deletions. The results indicated that both groups of children performed well when presented with ungrammatical sentences containing morpheme deletions; but the performance of both groups was lower for the word-order violations, with the difference between the two types being greater for the 5-year-olds. It is argued that these results reflect differences in the extent to which the two types of rule violations rendered the original sentences meaningless.

Effect of Syntactic and Semantic Violations on Sentence Intelligibility

A technique for measuring minor differences in the intelligibility of words in context was developed and used to measure the effect of violations of syntactic and semantic structure on intelligibility. Hypothetical analyses of English were drawn from transformational grammar and from analytic philosophy. These hypotheses about linguistic competence were converted to hypotheses about which one of a pair of sentences containing the same test word was more acceptable or “grammatical.” The average signal-to-noise ratio (S/N) at which the test word could just be recognized was found for each sentence. The difference between the S/N of the “good” and “bad” sentences in a pair was compared with the results from a questionnaire that asked for differential judgements of acceptability. In general, semantic violations did not affect intelligibility while grammatical violations did. Results are categorized according to the type of syntactic and semantic violation.