Language norms

Some of the early approaches to norms considered ‘norm’ as a mediating concept between the language system and speech (Coseriu 1970; Hjelmslev 1942). The norm, as proposed by Coseriu (1970), restricts possibilities that the system of a language permits, and speech, then, is the realization of the norm. In other words, the norm is a restriction on the possible patterns and structures that are compatible with a language system (Bartsch 1982: 52). In contrast, Prague school linguists focused their attention on the function of norms and the difference between norms and their codification (Hartung 1977: 51). Havránek (1964), for instance, emphasises that all linguistic communities have their own linguistic norm, whether codified or not. This means a shift from the norm of a (standard) language to the norm of a dialect, of a sociolect, of a register and so on. A distinction is also made between the norms of written and spoken languages. The Prague school linguists thus brought the concept of norm to the level of linguistic communities by pointing out that dialects had norms of their own, albeit not codified ones (see Švejcer 1987).

However, the above approaches have been criticised (Bartsch 1982: 53, 55; see also Bamgbose 1987) for their focus on ‘the norm’ of a language (as in Coseriu 1970) or ‘the norm’ of a dialect (as in the Prague school), which hides the possibility that conflicting norms could exist within a linguistic community. The distinction made by the Prague school between norms and their codification is, nevertheless, an important step towards a focus on norms relevant to linguistic communities (or communities of practice; see section 2.2.2) rather than on norms of the standard language (see e.g. Schwarz 1977: 73;

Švejcer 1987) – a central distinction for the purposes of this study.

Now, in order to arrive at a definition of language norms, let us compare three different approaches to language norms: Bamgbose (1987, 1998), Bartsch (1982; 1987) and Piippo (2012). All these approaches distinguish between norms relevant to linguistic communities and those of the standard language, and all of the approaches consider norms to be variable in scope and agree that different norms may exist within the same community. However, the norm concept is defined differently in each approach, and thus a closer look at each approach is in order.

15

I start with Bamgbose (1987: 105, see also 1998) who defines a language norm as “a standard language form or practice that serves as a reference point for other language forms or practices”, and then proceeds to distinguishing three norm types: code norm, feature norm and behavioural norm. He defines the norm types thus:

(i) Code norm: A standard variety of a language or a language selected from a group of languages and allocated for official or national purposes.

(ii) Feature norm: Any typical property of spoken or written language at whatever level (e.g. phonetic, phonological, morphological, syntactic, orthographic, etc.) and the rules that go with its production or use.

(iii) Behavioural norm: The set of conventions that go with speaking including expected patterns of behaviour while interacting with others, the mode of interpreting what is said, and attitudes in general to others’ manner of speaking.

(Bamgbose 1987: 105) The definition of a language norm as a standard language form or practice implies that a language norm is a codified norm, although the differentiation between the three types of norms means that different norms may exist within a community, and some norms may, for instance, cut across communities and thus have a wider scope than other norms (Bamgbose 1987: 111). However, codification is central to Bamgbose’s (1998: 5) approach in that he argues that for a variety to become a point of reference for usage and acceptance, it needs to be codified.¹²

Bamgbose (1998) does not go into much detail in terms of the relationship between the different types of norms, but considering his discussion of the codification of different varieties of English, it appears that a code norm refers to a standard variety (such as Standard British English), a behavioural norm refers to what is appropriate usage in interaction, and a feature norm determines acceptable linguistic form. Bamgbose’s (1998) point, then, appears to be that behavioural norms should guide the creation of feature norms in order for a variety to become a point of reference for usage and acceptability – and thus a language norm. While I fully agree with this starting point for the codification of varieties, for my purposes, I need a definition of language norms

12 Seidlhofer (2009a, 2011) takes up Bamgbose’s (1987, 1998) differentiation between feature and behavioural norms in discussing ELF, and sees particular relevance in behavioural norms. See chapter 3 for a discussion on this.

16

according to which language norms do not necessarily have to mean standard language forms or practices.

Bartsch’s (1987) theoretical model of norms of language provides some answers.

Bartsch (1987: xii) defines norms as “the social reality of correctness notions”, with correctness notions considered to be necessary for the recognisability and interpretability of linguistic expressions. In Bartsch’s (1987: 172) model, acceptability is not necessarily identical with linguistic correctness or grammaticality of the utterance (irrespective of the standard compared to), which means that ungrammatical expressions may very well be constructed acceptable in the course of interaction. For Bartsch (1987: 172, 213), acceptability means correctness with respect to the highest norm of communication, that is, achieving understanding, whereas correctness is achieved by compliance with established linguistic norms. This distinction between acceptability and correctness makes it possible to see interaction as a possible site for norm negotiation: acceptable usage that deviates from correct usage can, when it recurs, lead to change of language norms.

However, in the model, correct usage is still seen in relation to established (codified) norms, a system that is already there, waiting to be used, rather than something that is maintained and reconstructed in interaction (see Piippo 2012: 110). While I draw on Bartsch (1987) and distinguish between acceptability and correctness in my conceptualisation of language norms, I see both acceptability and correctness as maintained and reconstructed by speakers in interaction.

The third definition of language norms I want to take a closer look at is the one given in Piippo (2012). For Piippo (2012), norms are empirical phenomena, and importantly in her theory, established (codified) norms are not an assumed yardstick.

Piippo (2012: 27) defines language norms as “concepts of appropriate, expected and meaningful conduct”, by which she means that:

[language norms] are representations that contain the knowledge of a certain linguistic element’s social range as well as its social domain. In other words, norms are knowledge about semiotic signs and their social meaning potentials. This includes knowledge about by whom and in what type of situations the sign could be appropriately and meaningfully used. (Piippo 2012: 232–233)

This understanding of norms builds on the notion of appropriateness of an utterance or expression to a specific situation. Piippo (2012: 29) rejects earlier definitions of norms that rely on the notion of correctness (e.g. Bartsch 1987) in order to, on the one hand, emphasise the context-bound nature of normativity and on the other, avoid the

17

connotations of prescription and grammatical well-formedness connected with the notion. However, on the whole, this makes the definition too broad for the purposes of this study: appropriateness deals with questions of socially acceptable behaviour, such as the social acceptability of swearing (see Warren 2006: 128–129), whereas my focus is on the construction of acceptable language. This kind of acceptability is what speakers achieve through the process of language regulation in interaction (section 5.1.2): by intervening in one’s own or one another’s language, but also by letting, for instance, unconventional forms pass. When I talk about language norms, then, I refer to representations of acceptable, rather than appropriate, linguistic conduct. What speakers construct as acceptable linguistic conduct is further seen to exert different degrees of

“normative force” (Bartsch 1987: 166) – understood in this study as a sense of

‘oughtness’, that certain kind of conduct is expected of members of a group (see Piippo 2012: 40). This means that we can expect different normative force for the linguistic conduct that speakers construct as acceptable compared to the conduct they construct as correct.