When describing a variety such as Singapore English, the first question that arises is: which Singapore English? The Singlish used by the lesser-educated hawker is quite unlike the polished Standard English employed by a minister when addressing parliament, and not just in the choice of vocabulary: pronunciation and grammar differ, too. And yet both have to be considered instantiations of Singapore English, as they are, obviously, English, and used by Singaporeans. This variability is omnipresent, as I hope to make explicit in the following description.

Phonology

Vowels

The monophthongs of SgE are generally given as in Table 3.1 (see e.g. Bao 1998 or Deterding 2007a). This inventory shows the absence of phonemic length distinction (i.e. beat and bit are pronounced identically); this absence seems generalised in the speech community. There is, however, conditioning by phonological environment: vowels are ‘relatively short, except in open syllables’ (Bao 1998: 156). Thus beat and bit are both /bit/, but bee, being /bi/, is realised as [biː].

SgE has five diphthongs (face and goat are monophthongised to /e/ and /o/ respectively) which are phonologically identical to RP's, but phonetically much narrower, the two centring diphthongs particularly so. Deterding (2007a: 26-7)mentions that the diphthong /uə/ distributes differently in SgE from how it does in RP. In the words poor, tour, and sure the vowel is usually a diphthong [uə], whereas in pure and cure it tends to be [jɔ]. This, Deterding notes, is ‘the inverse of the common pattern in modern RP’ (2007a: 26).

The preceding chapters have highlighted the problems of analysing the interaction between Singlish and Standard English as diglossic, with similar problems encountered in the continuum approach. Furthermore, the data show that the usual variety-based approach to this variation is flawed, or at least unsatisfactory in its explanatory power. This chapter recapitulates these issues, and thematises the use of the ‘variety’ as a structural unit in fields such as sociolinguistics, contact linguistics, and World Englishes.

Codes, varieties, and code-switching

So far, terms such as code, variety, and subvariety have been used as near-synonyms when talking about SgE. Matthews (2007) offers definitions for each, saying that a variety is ‘any form of language seen as systematically distinct from others’, and that a code is ‘any language or distinct variety of a language, whether or not it is actually thought of as a code in any illuminating sense’. This may need a bit of clarification. In sociolinguistics, variety is a term that is used widely, and came about initially to avoid the loaded terms accent, dialect, and language. The problems in identifying language boundaries along a dialect continuum, for instance, are well known. It is, therefore, convenient to use variety not just for every dialect making up the continuum, but also for the superimposed standardised ones. The same holds true for code, but works at an even more abstract level: it can be applied not just to varieties of language as we normally understand it, but also to any means of communication, such as, for example, the Morse code (Matthews 2007: 62).

Early models and diglossia

The continuum hypothesis

In his seminal 1975 article, John T. Platt introduces two important concepts into the way in which SgE is viewed. Firstly, he applies the concept of a post-creole continuum, championed by DeCamp (1971), to the SgE speech community. His is a much more dynamic model than Ferguson and Stewart's formulae. Secondly, he introduces the term creoloid to describe a particular type of contact variety. I will first examine the concept of the post-creole continuum, before explaining the reasons for Platt's choice of the term creoloid.

The Jamaican continuum

Much of Platt's inspiration for the post-creole continuum analysis comes from DeCamp's work on Jamaican English, which considers ‘post-creoles’, i.e. creoles that ‘gradually merge with the corresponding standard language [that is, their lexifier]’ (1971: 349). This is not always the case; similar situations may result in quite different outcomes: Haitian Creole French, for instance, has continued co-existing with Standard French in a diglossic situation without any sign of merger (Ferguson 1959; DeCamp 1971: 351). DeCamp (1971: 350) defines a post-creole situation as one where he says there is ‘no sharp cleavage between creole and standard’. Rather, ‘there is a linguistic continuum, a continuous spectrum of speech varieties ranging from … “bush talk” or “broken language” … to the educated standard’ (De-Camp 1971: 350). This continuum, made up of its extremes, the basilect and the acrolect, and the intervening mesolects, presents a seamless succession of subvarieties: it is in fact so fine-grained that it cannot be subdivided into discrete lects.

As the previous chapters have shown, the variation in Singapore English between what has, traditionally, been called ‘Standard English’ on the one hand and ‘Singlish’ on the other hand is less than straightforward. The diglossia approach, presented in Section 2.1.3, took a domain-based distribution of two relatively clearly identified varieties as its premise, in which the choice of one variety over the other was motivated, essentially, by the nature of the exchange: university lectures, all but the least formal writing, public speaking, school and church settings, all called for the high (H) variety, Standard English, whereas everyday conversations, peer-group interaction in the playground, and generally all informal speech was carried out in the low (L) variety, Singlish. That this view of variation in Singapore English is unsatisfactory was subsequently explained at length, drawing mostly on data showing the problematic interplay of features ostensibly ‘belonging’ to both H and L in the same utterance.

Alternatives were found in Alsagoff's cultural orientation model (see Section 2.2.1), a view of the Singlish–Standard variation that takes into account the potential social meanings encoded in ‘switches’ from H to L and/or vice versa. Still, the view there was largely one that took a supposed ‘variety’ of English (either Standard or Singlish) as its unit of analysis (a point further addressed in Chapter 6), and was also largely concerned with cultural orientation (thence the name of the model), rather than with an overarching theory that could explain all of the social meanings potentially indexed through the various features used.