This chapter examines reference in planned, deliberated language production. Using evidence from examples of institutional writing (writing text in academic, professional, and social settings), the effects of the written mode of production on referential choice, including evidence from keystroke logging methods which give us insight into revisions during the writing process. In comparison with spontaneous discourse, we demonstrate how the relatively low frequency of pronominal reference that is typical of spontaneous discourse is replaced with more complex referential expressions. We also consider whether referential choice is a feature of genre as we examine the referential profiles of different text types. The chapter concludes that referring expressions in prepared discourse tend to be more frequently information rich and consequently, the modification and qualification zones of the noun phrase are more active in this type of discourse.

Human history has created a large variety of sign systems for communication. These systems were developed at different times for different purposes. While oral language has developed as part of human biological evolution, written texts, realistic pictures, maps, and graphs are cultural inventions. Human oral language might have originated from gestures supplemented by sound patterns. It is a biological anchored feature of the human species, as manifested in somatic, perceptual, and neurological pre-adaptations. Early writing systems used iconic ideograms which were gradually transformed into symbols. This made production and discrimination easier but increased the required amount of learning. Further development led to writing systems using phonograms plus orthographic ideograms. Realistic pictures are older than writing systems. They represent content by similarity but also show allegories of social relationships. Maps are realistic pictures of a geographic area facing the problem of how to present a curved earth surface on a two-dimensional surface. Graphs are visuo-spatial objects representing a subject matter based on analogy due to inherent common structural properties.

The information-structural categories of focus and topic are examined with respect to the constructions in which they can feature, including fronting, dislocation, subject inversion, and presentational sentences. The role and effects of illocutionary force distinctions (e.g., declarative, interrogative, exclamative) and different predication types (thetic vs non-thetic predications) are also taken into consideration. Despite the many similarities, (micro)variation in this area proves quite considerable. In relation to Romance comparative data, this chapter shows that this variation can be accounted for by pragmatic-discourse factors and structural licensing principles that are indeed related to information structure.

Most linguists regard writing as an uninteresting and somewhat artificial add-on to spoken language. However, the availability of the written form influences language structure in profound ways at both the individual level (mental grammar) and the social level (language as a conventional system). The lexical and grammatical differences between spoken and written language are partly attributable to different contexts of use and partly to the fact that the written form provides a ‘processing crutch’ which enables speakers to produce more complex structures than they would otherwise be able to produce; structures which are initially learned with the aid of writing may then find their way into the speech of highly literate speakers. In the long run, the cumulative effects of these individual changes lead to changes at the level of the community language. Thus, languages with a long tradition of literacy, such as English, differ in substantial ways from oral languages.

This chapter reports on a poetic inquiry with English language learners in the Japanese English as a Foreign Language (EFL) university context. It focuses on the use of haiku – a Japanese poem containing seventeen syllables in a three-line 5–7–5 syllable pattern with the usage of a seasonal reference and a cutting word – and analyses features of spoken language in a corpus of English language haiku poetry written by Japanese second language (L2) learners. The chapter begins by reviewing previous studies of the use of haiku in L2 contexts. It then describes a quantitative, corpus-based study which involved the analysis of textual and linguistic features of English language haiku writing. The data, consisting of a total of 2,017 haiku poems written by 204 first-year engineering students at a Japanese public university, were submitted to statistical analyses. The results illustrate some specific features of English language haiku produced by Japanese L2 writers: haiku poetry is a short, descriptive text which presents each writer’s emotional reactions to his or her daily life, and it also includes such spoken language features as the twelve verbs most frequently used in spoken discourse (Biber and Conrad ), evaluative and emotive adjectives, contractions and vague language. This study suggests that the task of composing haiku in English can play an important role in L2 learning in terms of raising learners’ awareness of typical spoken forms in the target language.