Issues in modelling the textual metafunction

Editors’ introduction One important aspect of the nature of texts is the way in which the flow of information through the text is managed. The writer or speaker typically aims to signal to the addressee the status of each ‘chunk’ of information that is given — how it fits in with the rest of the information around it in terms of its degree of newness or familiarity to the addressee. In this paper Fries focuses on two concepts related to this issue: participant entry and information structure.

Participant entry concerns the way in which entities are referred to in the text: whether an entity is to be taken as identifiable by the addressee at that point in the text or not. Using Martin’s (1992) terminology, Fries distinguishes between presenting and presuming reference. If a nominal group has presenting reference, it signals that the addressee is not expected to be able to identify the entity referred to from their own knowledge. Presenting reference typically involves indefiniteness (‘a’, ‘some’, etc.). Presuming reference, on the other hand, signals that the addressee is expected to identify the entity, and typically involves definiteness (‘the’, etc.). The two following utterances exemplify the difference:

There’s a car parked outside [presenting: ‘you don’t know the car I’m referring to’]

Do you need the car today? [presuming: ‘you know the car I’m referring to — ours’]

Information structure concerns the ‘pulse’ of information in a stretch of text. Speakers signal through intonation how the information is to be segmented into chunks (see also Thompson & Thompson in this volume), and they also signal what is newsworthy in each chunk and what is to be taken as already known to the listener or as not needing as much attention. The New in each chunk is centred around the tonic syllable (the syllable which is most prominent), and typically comes at or near the end of the chunk. The Given has less prominence and typically comes at the beginning of the chunk. Since Given and New are signalled by intonation, they are usually associated with spoken language. However, Fries argues that the concepts can be usefully applied to written text as well, and he describes how this can be done.

There is clearly a relationship between participant entry and information structure, in that presuming reference and Given both concern something that the addressee already ‘knows’, while presenting reference and New both concern something that the addressee does not ‘know’; and in many cases, they are indeed matched. Fries demonstrates, however, that they need to be distinguished because the matching is far from straightforward: presuming reference can occur in New and presenting reference can occur in Given.

It might be expected that participant entry and information structure would depend primarily on the addressee’s actual state of knowledge: that if the addressee cannot identify an entity the speaker will use presenting reference, and if the addressee does not know some information the speaker will present it as New. Again, Fries shows that this is too simple a picture. He argues that speakers always have the choice of how to present information. Even when they are aware that an entity is identifiable to their addressees they may use presenting reference, as if it were not. Similarly, even when they are aware that information is obvious to their addressees they may present it as New. Through a range of examples including extracts from a biography, conversation, poetry, a research presentation and advertisements, Fries explores the kinds of factors which can lead speakers to make such choices (his analysis of written advertisements forms an interesting comparison with Thompson & Thompson’s analysis of spoken advertisements). His central argument is that we cannot understand information flow in text if we assume that it is determined by external ‘facts’: instead, we must see it as constructed through motivated language choices. This in particular ties his paper in with the rest of this volume: texts have a life of their own and, like any other living organism, they need to be studied not just in terms of their parts (clauses, etc.) but as wholes that are influenced by their environment (the world around them) but are not determined by it.