In an article in this journal, Bao (Reference Zhiming2005) proposes a constraint on functional transfer that he claims accounts for features of colloquial Singapore English (and other language contact varieties) better than the congruence constraint proposed by Siegel (Reference Siegel1999) and subsequently developed in later works (e.g. Siegel Reference Siegel2003, Reference Siegel2008a). More specifically, Bao argues that the requirement of surface syntactic similarity for transfer is too strong. His analysis uses Mandarin to exemplify the Chinese substrate languages that were the source of transfer, following the view that there is a universal Chinese grammar (Chao Reference Chao1968: 13). However, the present article shows that Bao's claim is unjustified because the actual source of transfer was a variety of Chinese that differs significantly from Mandarin in the area of grammar he examined.

Functional transfer is a kind of cross-linguistic influence occurring in individual speech that can lead to contact-induced language change in a community. In this kind of transfer, the functions of a grammatical morpheme from a source language are applied to a word in the recipient language that normally does not have a grammatical function (Siegel Reference Siegel2003, Reference Siegel2008a). For example, the Chinese substrate languages spoken in Singapore – Hokkein and related Southern Min dialects, Cantonese and Mandarin – all have a postverbal marker of completive or perfect aspect (often called perfective (pft) in Sinitic studies). In Mandarin, this is le (Bao Reference Zhiming1995: 185):Footnote ²

1. (1) ta qu niuyue le.3sg go New York pft‘He went to New York.’

In the variety of Hokkien spoken in Singapore and Malaysia, the perfect marker is liau (Platt & Weber Reference Platt and Weber1980: 66):

1. (2) Gún tháùke tńg chhù liaú.our boss return home pft‘Our boss has returned home.’

In Singapore English, these functions have been transferred to the English adverb already, which has become a completive or perfect aspect marker, as in the following examples:

1. (3)

   1. (a) I only went there once or twice already. (Platt & Weber Reference Platt and Weber1980: 66)

   2. (b) I work about four months already. (Bao Reference Zhiming1995: 182)

Other postverbal aspect markers, such as continuous, are also found in all the Chinese substrate languages (Chappell Reference Chappell1989, Reference Chappell1992), but there is no evidence of their functions being transferred to English words and incorporated into Singapore English. This can be accounted for by a proposed constraint on transfer (Siegel Reference Siegel1999, Reference Siegel2003, Reference Siegel2008a). For a feature to be transferred, there must be somewhere to transfer to (Andersen Reference Andersen, Gass and Selinker1983) – a morpheme in the RL (recipient language; here, English) that can be assigned the functions of a grammatical morpheme or structure in the SL (source language; here, one of the Chinese substrate languages). In addition to being perceptually salient and having a function or meaning related to that of the corresponding SL feature, this RL morpheme must occur in the same surface syntactic position, at least in some contexts. The absence of such structural congruence will constrain transfer.

In the case of continuous aspect, for example, English has a perceptually salient morpheme with a meaning related to that of the Chinese markers – i.e. keep as in keep running. But this occurs before rather than after the verb. Since the English and Chinese forms are not syntactically congruent, individual transfer did not occur, and so there is no postverbal continuous marker in Singapore English. This congruence constraint also accounts for the absence of various substrate language features in Melanesian Pidgin (Siegel Reference Siegel1999), Tayo (a French-lexified creole spoken in New Caledonia) (Siegel, Sandeman & Corne Reference Siegel, Sandeman, Corne and Siegel2000), Australian Kriol (Munro Reference Munro2004) and Hawai‘i Creole (Siegel Reference Siegel and Michaelis2008b).

However, Bao (Reference Zhiming2005) argues that in Singapore English the properties of the Chinese marker of experiential (exp) aspect guo (meaning ‘happened at least once in the past’) have been transferred to the English word ever – as in the following:

1. (4)

   1. (a) I ever tried this type of fruit before.

      ‘I have tried this type of fruit before.’

   2. (b) Last time I ever bought something on sale.

      ‘I have bought something on sale in the past.’ (Bao Reference Zhiming2005: exx. (17a, c))

But the experiential aspect marker follows the verb in Chinese whereas ever precedes the verb in English – for example (244):

1. (5) wŏ chī guo Ribĕn fànI eat expJapan food‘I've eaten Japanese food (before).’ (Bao Reference Zhiming2005: ex. (18a))

Therefore, according to the congruence constraint outlined above, transfer would not have been expected.

Bao (Reference Zhiming2005: 258) proposes a constraint that is not as strict as congruence, and would allow for such transfer. First, his view is that an entire grammatical subsystem transfers, not just particular items from the system. So, for example, he argues that the entire Chinese aspect system was transferred to Singapore English, thus accounting for already as the perfect marker ever as the experiential marker, and other preverbal aspectual markers. The main constraint is the ‘lexifier filter’ – that the lexifier ‘acts as a filter, sifting out those categories of the transferred subsystem for which its grammar cannot provide straightforward morphosyntactic exponence’ (pages 237–238). Thus, the Chinese aspect markers that have equivalents in Singapore English are the ones that can be expressed by similar morphosyntactic means in both Chinese and English – that is, with preverbal or postverbal functional morphemes. The Chinese aspect markers that do not have equivalents in Singapore English are those for which the lexifier has no similar morphosyntactic means of expression. For example, tentative aspect is indicated by reduplication in Chinese, but reduplication is not a productive morphological process in English, so this aspect is not found in Singapore English (page 262). Bao says (page 260) that the lexifier filter constraint is ‘weaker than congruence’ in that it allows for transfer of the V+guo experiential aspect marking as ever+V, and therefore it better accounts for the data.

The problem is that Bao (Reference Zhiming2005) bases his analysis on only one variety of Chinese, Mandarin. And although he says (page 238, footnote 2) that Mandarin is now the dominant Chinese dialect in Singapore, it was hardly spoken in the 1950s, when education in English became more widespread (Lim & Foley Reference Lim, Foley and Lim2004: 4) and Singapore English had already begun to develop. In 1957, only 0.1 percent of the population spoke Mandarin as their first language, as opposed to 53.2 percent speaking Southern Min varieties (30.0% Hokkien, 17.0% Teochew, 5.2% Hainanese, 1.0% Foochow) and 17 percent speaking Cantonese (Kuo Reference Kuo, Afendras and Kuo1980: 41), as well as 11.5 percent speaking Malay. Widespread promotion of Mandarin as the lingua franca among the Chinese did not begin until after independence in 1968 (Bao Reference Zhiming2005: 238, footnote 2), and accelerated with the Speak Mandarin Campaign, which began in 1979 (Bokhorst-Heng Reference Bokhorst-Heng, Harris and Rampton2003). The first published grammatical description of Singapore English appeared in 1974 (Tongue Reference Tongue1974). Thus, it seems clear that when Singapore English was developing, the dominant variety of Chinese was not Mandarin, but Hokkien and closely related Southern Min varieties, which were spoken by more than three quarters of the Chinese population.

It would seem logical then to examine the influence on Singapore English of the Min dialects rather than Mandarin. However, Bao takes the view, following Chao (1968: 13), that all the varieties share a ‘universal Chinese grammar’. This is despite arguments to the contrary presented by Ansaldo (Reference Ansaldo1999) and Matthews & Yip (Reference Matthews, Yip and Chappell2001). With this regard, Bao states (page 238, footnote 2) that ‘[t]heir arguments, however, do not affect the aspectual system’.

This view is contradicted, however, by the work of Chappell (Reference Chappell1989, Reference Chappell1992). She demonstrates that the preferential strategy for marking aspect in Southern Min is with preverbal adverbs or auxiliaries as opposed to the use of suffixes in other varieties of Chinese. With regard to experiential aspect, Southern Min most often uses a preverbal marker băt or băk, unlike Mandarin and Cantonese, which always use a postverbal marker, guo and kwo respectively. While some Southern Min dialects also have a postverbal experiential marker kè or kuè due to the influence of Mandarin, the use of preverbal băt is preferred. And although Mandarin has a preverbal adverb ceng which is similar in meaning to băt, unlike băt it is not used on its own to indicate experiential aspect. Chappell (Reference Chappell1989: 17) gives the following example:

1. (6) Guă băt k'iI expgo‘I've been (there before)’

In later work, Chappell (Reference Chappell and Chappell2001) analyses preverbal băt in Southern Min as an evidential marker.

Ho & Wong (Reference Lian and Wong2001: 84) also mention this preverbal marker in Hokkien, and note that it ‘provides an exact semantic and grammatical fit with ever in SgE [Singapore English]’. This includes the use of the marker in affirmative sentences, unlike existential ever in other varieties English (see Israel Reference Israel1998). Ho & Wong (Reference Lian and Wong2001: 85) provide the following example (glossing the experiential marker as ever):

1. (7) Goa bat khi Jitpun.I ever go Japan ‘I've been to Japan (before).’

Further evidence that ever is more likely a result of transfer from Hokkien rather than Mandarin is found in affirmative responses to yes/no questions in Singapore English that include ever – for example (Ho & Wong Reference Lian and Wong2001: 80):

1. (8)

   Q: You ever go Japan (or not)?

   A: (Yes), ever.

1. (9)

   Q: Your husband ever bring home fish to eat or not?

   A: Ever.

This matches what is done in Hokkien (Ho & Wong 2001: 85):

1. (10)

   Q: Li bat khi Jitpun bo?you ever go Japan not

   A: Bat.ever

This contrasts to responses in which the postverbal experiential aspect marker (exp) is used, which require the verb as well as the marker – for example in Mandarin (Ho & Wong Reference Lian and Wong2001: 84):

1. (11)

   Q: Ni qu guo Riben meiyou?you go expJapan not

   A: Qu guo.go exp

Furthermore, in Malay the preverbal marker pernah also has properties similar to those of ever in Singapore English. This can be seen in the following examples from Ho & Wong (2001: 85), in which they gloss pernah as ever, as they do with Hokkien bat in examples 7 and 10 above.

1. (12) Saya pernah pergi ke Jepun.I ever go to Japan‘I've been to Japan (before).’

1. (13)

   Q: Awak pernah-kah pergi ke Jepun?you ever.qp go to Jepun

   A: Pernah.ever

Thus, transfer from Malay could have complemented or reinforced that from the Southern Min varieties of Chinese.

In conclusion, the evidence presented here demonstrates that the properties of the preverbal aspect marker ever in Singapore English were very likely the result of transfer from the experiential (or evidential) aspect markers in the substrate languages. Since these markers, like ever, occur preverbally in substrate languages spoken by nearly 65 percent of the population (53.2% Southern Min varieties, 11.5% Malay), there is no reason to believe that transfer occurred from Mandarin, which has a postverbal marker, especially when it was spoken by only 0.1 percent of the population. Therefore, there is no reason to dismiss the congruence constraint as being too strong, and no need to propose an alternative weaker constraint.