Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Semantic options and complex functions: A recursive view of choice

Fontaine, Lise Margaret 2013. Semantic options and complex functions: A recursive view of choice. In: Fontaine, Lise Margaret, Bartlett, Thomas Alexander Marks and O'Grady, Gerard Nigel eds. Systemic functional linguistics: exploring choice, Cambridge: Cambridge University Press, pp. 95-114.

Full text not available from this repository.


At the heart of Systemic Functional Linguistics (SFL) is a model of language as a semiotic system. As speakers of a language, we want to make meaning and as hearers of a language we want to make meaning out of what we hear. The semiotic system for language according to Halliday is modelled by a network of meaning potential (1978:40-41): “The network consists very simply of a set of interrelated systems … a semantic network is a representation of semantic options or choices in meaning.” According to Fawcett (1988:3), this representation of choice “lies at the heart of systemic linguistics”. SFL has shown itself to be extremely attractive for many applications. For many, the attraction is the centrality of the notion of ‘choice’. There is no doubt that the representation of choice in a semiotic system is the main strong point of SFL. However, it is precisely ‘choice’ itself that causes the most misunderstanding. The way in which this core concept is interpreted, both within SFL and for those wanting to adopt its framework (e.g. computational linguists), is very important since it is the concept that lies at the heart of SFL. For example, Winograd had hoped to avoid using this term in his own work because he felt it was inappropriate to think of ‘choice’ as active choosing (1983:289). As he pointed out, the classification inherent in the networks “is being imposed by an observer”. However, since, as he concluded, “the term choice is standard in discussions of systemic grammar”, there was no way to avoid using it and the term was adopted in his computational approach “in spite of its connotation of intentionality" (ibid). Halliday has never claimed that ‘choice’ means choosing. However there is a lack of clarity about the term in most SFL literature. In earlier writing, Halliday states clearly that “the semantic system is a network of meaning potential. The network consists very simply of a set of interrelated systems ... a system is a set of options, a set of possibilities ... Hence, a semantic network is a representation of semantic options, or choices in meaning” (Halliday, 1978:40-41). The meaning potential held within the network represents the semantic choices available in the language. Butler (2002:65-66) discusses the systematic ambiguity of the use of the term ‘meaning potential’, which is integral to the understanding of ‘choice’, calling it a “slippery” term having at least two interpretations. Tucker (1998:41) writes very clearly that: “a system network displays the choices in meaning which can be made, and which lead to a possible realization at the level of form. What they do not do, and cannot do, is provide the basis for these choices”. However it would be possible to infer from this that choices are made and this could lead one to think that choice is an active process. There are two issues underlying the problems highlighted above. One is the interpretation of the system networks and the other is in the interpretation of the term choice. Language is best viewed as a process rather than an object (Fawcett, 1993: 626). Lamb makes a similar point: “the important thing to realize is that (the word) is not an object. The word that comes out of a speaker’s mouth comes out because a particular portion of network gets activated” (Lamb, 1999: 62). This is true for all utterances, even though we often treat text as object (product) in text analysis. This distinction between process and product is extremely important but raises problems of its own. The difficulty is as Andersen describes, “that the model describes wholes as emergent entities, generated by the activities of the parts, but does not specify the relationship between the new macro behaviour and the micro behaviours that created it” (2002:39). Trying to account for the interaction of the parts in a language system is incredibly challenging and it requires a view of language as a complex system. Although few linguists explicitly state so, language is a complex system (Lemke, 1993; Andersen, 2002; Smith, Brighton, Kirby, 2003; Halliday and Martin, 1993) and some go further in considering it as a complex adaptive system (Steels, 2000; Briscoe, 1998). The nature of complex systems is such that recursion is necessarily a central notion. In this chapter, ‘choice’ will be explored with respect to two different views. The first concerns the system network where ‘choice’ is seen as representing the available (semantic) options; this is a classification problem. The second concerns the (complex) language system where ‘choice’ is seen as a means to select options; this is a recursion problem. The main aim of this chapter is to explain how ‘choice’ can be (re)defined as a simple term and a complex function.

Item Type: Book Section
Date Type: Publication
Status: Published
Schools: English, Communication and Philosophy
Subjects: P Language and Literature > P Philology. Linguistics
Publisher: Cambridge University Press
ISBN: 9781107036963
Related URLs:
Last Modified: 04 Jun 2017 04:51

Actions (repository staff only)

Edit Item Edit Item