Written Language & Literacy - Volume 5, Issue 1, 2002

The studies reported in this volume of WL&L (5: 1–2, 2002) all derive from a joint project entitled “Developing literacy in different contexts and in different languages”, funded by the Spencer Foundation, Chicago. The study encompasses seven languages — Dutch, English, French, Hebrew, Icelandic, Spanish, and Swedish — for which data were collected in Europe, Israel, and the US by graduate research assistants in education, linguistics, and psychology, under the supervision of a project director in each country — each of whom is listed as a first or “lead” author in the articles which follow the introduction to this collection.

The present study presents contrastive analyses of task-oriented spoken and written discourse in terms of lexical diversity, lexical density, and word length. In an age-matched within-language comparison (Swedish), written discourse consistently scored higher on these measures. It is suggested that the same type of differences will hold for any language, because of the difference between speech and writing in processing constraints. The absolute scores, however, can vary substantially for reasons of language typology. An extended, cross-linguistic analysis (English, Hebrew, Icelandic, Swedish), focusing on word length, was made to substantiate that claim. Further, cross-age-group comparisons of lexical quanta indicated a dynamic interaction between speech and writing in development. Spoken discourse eventually comes to “learn” from the development of writing.

This paper examines how choice of subject NP types and structures changes in the development of text construction, and the extent of variation in the developmental patterns which are produced in speech and in writing. The population for this study consisted of 80 participants — 40 grade-school children and 40 university-level adults — with 20 participants in each of four languages: Dutch, Hebrew, English, and Spanish. The database for each language-specific analysis consisted of 40 grade-school texts and 40 adult texts. In each group, half were spoken texts and half written, half were narratives and half expository texts: altogether 320 texts. All subject NP slots in each text were counted and classified by category of realization (zero, pronoun, or lexical), by pronoun type (personal vs. impersonal), and by lexical complexity (terminal NPs governing a single lexical noun vs. non-terminal NPs governing more than one lexical noun). In general, the written expositions of adults are the preferred site for lexical subjects and for non-terminal subjects. Among both children and adults, narratives contain more personal subject pronouns, and expository texts contain more impersonal pronouns. Several cross-linguistic differences emerged (mainly between Spanish and the other three languages), reflecting differences in the syntactic, inflectional, and pronominal patternings of the target languages.

This paper forms a bridge between the article on noun phrase patterning by Ravid et al. 2002 and that on passive voice constructions by Jisa et al. 2002. The study reports on a cross-linguistic, developmental study of verbal structures and verb types used in two genres of written discourse: personal narratives and expository texts. The study is aimed at (a) establishing the profile of linguistic features that characterize and differentiate these two genres; (b) identifying the developmental changes beyond middle childhood that lead to the proficient use of a full repertoire of verbal structures in the construction of both types of text; and (c) providing fresh empirical evidence for cross-linguistic similarities and differences in the linguistic devices used for Genre differentiation. The paper begins to address these issues by considering quantitative aspects of Genre differentiation in four age-groups (grade-school children, junior high school, high school, and adults) and in five languages (Dutch, French, Hebrew, Icelandic, and Spanish). We expected narratives and expository texts to be characterized by contrasting distribution of the categories that we analysed — verb tense, aspect, mood, voice, and person — across the age-groups and languages under study. To test this prediction, all verbs in our sample were analysed using common coding procedures in all five languages, followed by a statistical analysis of the frequency distribution of each coded category (as our dependent variables) across Age and Genre in each of the languages.