In this chapter I will sketch work on information structure pre Halliday. This work will include mention of Aristotle, Weil and Prague School work e.g. Mathesius and explain his three part definition of theme. I will also briefly mention the work of Firbas and Daneš (discussed in detail in chapter 3) in order to contextualise what Davidse calls Halliday’s separating theme. The separation of the theme and information systems is a distinguishing feature of SFL and the chapter will describe how the systems fit into the textual metafunction and into the wider SFL architecture.

Chapter 2 provides a critical in depth examination of Hallidayan Theme. It first reviews work which influenced Theme and shows not only how the work of Prague School scholars influenced and motivated Halliday but also why Halliday separated Theme and Information. I will review apparent inconsistencies in definitions of Theme which emerge from attempting to conflate Mathesius’ three aspects of theme into Theme and Information. Theme is as Gómez González (2001) argued, identified both by place and matter metaphors. In order to rectify the divergence I examine the cognate notion of topic in order to motivate an analysis which codes for both Theme and topic. Topic unlike Theme is argued to be “the matter of current concern” whereas Theme is shown to be an orienting device which relates the clause to its context. It is shown that a clause must have a Theme but that topic is potentially optional though in the unmarked case Theme and topic will be conflated.

Chapter 3 focuses on the system of Information. It details the importance of tonality and tonicity in the speaker’s projection of which lexical items are Given and which are New. Close examination of tonality choices illustrates that while tone groups/information units are often coterminous with clauses, it is more accurate to say that they also frequently correspond to clausal cores and sub-clausal units. Hence there will be more tone groups/information units than clauses in a text. This fact entails that New elements may occur at any point within a clause. While there is strong evidence equating tone groups with information units and the tonic syllable with Focus, the informational status of pre-tonic lexical items is far less clear. One of the reasons for this it that the concept of recoverability is itself on close examination not entirely clear.

Chapter 4 provides a more expansive exploration of recoverability by examining various approaches including Firbas’ Functional Sentence Perspective, Krika’s distinction between common ground management and common ground content, Prince’s theory of information structure, Gundel and Hedberg’s Givenness hierarchy and Lambrecht’s functional theory of information structure. The discussion centres on whether information structure is binary or ternary with inferable items neither Given or New. I also review differences between hearer/speaker given and new and text Given and New, After reviewing the evidence the issue remains unresolved though the preponderance of evidence suggests that information structure is binary with the primary cut being between Given and New elements; though there may well be more delicate distinctions between elements which are present and those which are inferable. The chapter concludes by arguing for a firm distinction between recoverable and identifiable elements though it notes that both can be grouped together under a superordinate category of retrievable and that the relation between the lexicogrammatical and prosodic realisation of identifiable and recoverable lexical items is a mutually re-enforcing redundancy. Thus, it is shown that topical items are predicted to be Given which as we have already seen are predicted to be Thematic though when it suits the speaker’s purpose they can be separated. It will be noted that topic, Theme and Given appear to be identical with Mathesius’ three aspects of theme which I discussed in chapter 2.

Chapter 5 considers IS from a different perspective; namely predictability between classes of lexical items where an earlier item prospects a later one. In the chapter I outline a series of chaining rules which are supplemented by a number of formal devices and argue that they are sufficient to account for the dynamic and incremental flow of wordings and the chaining of tone groups in English as speakers achieve target state. The second part of the chapter reviews the concept of the mutuality of information and argues that speakers do not share information but rather signal their and their interlocutors’ unequal access to information/knowledge on a moment by moment basis.

Chapter 6 investigates the proposed model of IS by coding real speech and analysing how the four levels of the IS model combine to enable the speakers to weave together lexicogrammatical and prosodic choices into a coherent and cohesive text.