On the syntax of discourse particles in Colloquial Singapore English: The case for two syntactically distinct particle groups

To date, there has been relatively little theoretical work on the distributions of discourse particles in Colloquial Singapore English (CSE), with much of the literature favouring pragmatic descriptions of individual particles (Lim, 2007: 447). For instance, lah has been described as a marker of solidarity, emphasis, persuasiveness, amongst other emotive attitudes (see Lim, 2007: 448). While detailed syntactic descriptions exist for common particles (e.g. Bell & Ser, 1983), few analysts have considered the syntactic distributions of CSE particles as a whole. What descriptions exist of CSE particle syntax are limited to the observations that they occur individually (i.e. they do not stack) and in clause-final position (e.g. Wee, 2004: 117). Some exceptions to these rules have been noted, for instance in standalone hor (Gan, 2000: 37) and mid-clausal ah (Gupta, 1992: 43), though no syntactic explanation has been offered. Thus, still little is understood about CSE particle syntax. I have attempted to shed light on this matter by proposing two syntactically distinct groups of CSE particles. Particles in Group I display characteristics stereotypical of CSE particles, namely that they cannot stack and only occur clause-finally. Group II particles display ‘atypical’ syntactic properties, namely that they can stack and elide their CP complement. I note also that mid-clausal occurrence is possible for both particle groups, although these occurrences are marked uses. The analyses and observations proposed in this paper are based on my own intuitions as a native CSE speaker and supported by acceptability judgment data from other native- speaking informants. The particles I use for my analyses are ah, lah, wat, lor, hor, leh, meh and mah, which are considered representative of the CSE particle inventory (e.g. Lim, 2007; Smakman & Wagenaar, 2013). In addition, I use the particle eh, which has not yet been described in the literature but I argue is a member of the CSE particle inventory. I adopt the conventions for representing Mandarin tone, firstly because a more complex representation of intonational contour is unnecessary for my present purpose, and secondly because the particles used with different Mandarin tones elicit salient differences in speaker judgment. There are three pieces of evidence I appeal to in my proposal of two distinct groups. First, Group I particles always occur with a CP complement, while Group II particles can elide their CP complement. Second, compared to Group I particles, Group II particles encode a ‘stronger’ Call on Addressee (CoA; see Beyssade & Marandin, 2006), in that they request more of an addressee than a typical question particle would. Third, Group II particles can stack (both with Group I particles and each other), whereas Group I particles are only licensed individually. The division of CSE particles across two syntactically distinct groups provides insight into the relatively underexamined area of CSE particle syntax. It is hoped this represents the first step towards a more sophisticated syntactic theory of CSE particle distribution.