- 1 The present study is a shorter summary version of a longer and more detailed paper titled ‘Changes (...)
1One of the most colourful grammatical features associated with the contact dialect of Singapore English, colloquially known as ‘Singlish’, is that of the perfective aspect marker, already, which has been studied since the late twentieth century, in works such as Tongue (1979), and Platt, Weber & Ho (1984). To the outsider, the use of this feature appears somewhat redundant, as in standard English it is normally only used to express contrastive adverbial meanings of completion before an anticipated time reference in the context. The use of this marker in Singlish, however, is often simply one of completion alone, without any additional nuances of the anticipation of the event it marks, as in the following:
|
(1)
|
My father pass away already
|
(Platt et al 1984:71), in which the bare verb form appears to replace the morphology of past tense marking. In addition, it was observed to co-occur with negation to express meanings of discontinuity of a past state, as seen in Platt et al (1984), in which it conveys the meaning ‘any more’:
|
(2)
|
I don’t drink coffee already.
|
(1984: 71). In cases such as (2) there is no justification for assuming that already stands in place of a tense marker in Singapore English. In fact, already can be found occurring alongside tense marking in Singlish, in the following examples from the International Corpus of English Singapore:
|
(3)
|
A. How much ha
ICE-SIN:S1A-006#56:1:
|
|
B. I forgot already ha
ICE-SIN:S1A-006#57:1:A
|
2Given such co-occurrences, it is clear that the categorization of tense and aspect in Colloquial Singapore English requires further elaboration: Singapore English has been described by Maurer (2013) as possessing both tense and aspect, although on the surface, it appears to be a strictly aspect-based contact variety. Since the early studies of Platt and his colleagues, there has been a proliferation of studies on already in Singapore English, though much less research has been done on the use of tense marking. However, one of the earliest pioneering studies on past tense marking in Singapore English, Ho and Platt (1993), had concluded that tense marking was variably present in Singapore English and that the variation was one factor illustrating the presence of a contact continuum comparable to that of a decreolising contact continuum as discussed in Bickerton (1981). Their study revealed that the presence of past tense was determined firstly by the lexical aspect of the verb and secondly by the grammatical aspect of the sentence in which the verb occurred (notwithstanding phonological factors such as the avoidance of final consonant clusters created using regular past tense forms, such as walked, etc).
3However, Ho and Platt’s (1993) study failed to adequately account for the co-existence of already appearing alongside the variable tense-marking system, which requires a comprehensive explanation in terms of the distribution of functions. Their study described already as an anterior marker and a perfective form, but did not consider the possibility of functional overlap with the past tense, which is still variable in Singlish. The present study aims to investigate the present-day division of labour between already and the past tense, in order to determine to what extent the aspect marker is competing with tense marking in the dialect, and to propose the hypothesis that already’s present-day distribution in Singlish, now nearly 30 years after Ho and Platt’s study, is being affected by the gradual encroachment of past tense marking on part of the functional territory of already. It is suggested in the present study that such a situation represents what could only be described as a situation of partial renovation in the grammaticalisation of past tense, which, diachronically, is said to emerge out of perfective aspect, and, simultaneously, a case of gradual convergence of Singlish towards the (standard) lexifier, English.
4The first section will provide some general background on the contact dialect frequently referred to as ‘Singlish’, and the language contact situation in Singapore. The second section presents some earlier studies on already and the past tense in Singapore English, and proposes a system of classification based on the grammatical category of iamitives; section 3 describes the grammaticalisation of past tense from iamitives in various creoles and in Singapore English. Section 4 describes the results of a survey of the functional distribution of already in two corpora produced ten years apart from each other. Section 5 discusses the results and offers a diachronic viewpoint on the data, illustrating the gradual decrease in completive functions from the early 20th century until today, and proposes the hypothesis of renewal of such functions by the standard past tense. Section 6 concludes the study.
- 2 « Population Trends 2018 », the Dept. Of Statistics, Ministry of Trade and Industry, Republic of Si (...)
5English has been used in Singapore since the time of British colonization in the 19th century (1819); however, it is only recently, since independence in 1965, that it has become an official language. According to earlier, 20th century studies, the variety known as Singlish or Colloquial Singapore English (CSE) emerged out of the playgrounds of the English-medium schools from the 1930s on (Gupta 1991, 1994), though more recently, Bao (2015), using 19th century historical manuscripts, attributes its source to a possible pidgin-like origin, arguing that few Singaporean children were educated at English-medium schools prior to independence, and that Singlish cannot be such a recent addition to the contact language pool. The national language of Singapore today remains Malay, and dialects of Malay such as Baba Malay (the dialect of the Peranakan or Babas, the descendants of mixed marriages between Malays and Chinese) and Bazaar Malay (a lingua franca until the 1970s, according to Lim 2007), were in active use during the colonial period. Today, the main lingua franca is English, which unites the four major ethnic groups of the population of more than 4 million: the Chinese (representing around 74.3% of the population, the Malays (about 13.4%), the Indians (9%) and ‘others’ (3.2%).2 The original, most important substrate languages included Hokkien (Fujian Chinese), Cantonese, Teochew (Chao Zhou), Hakka (Kejia), and Hainanese, though Ho & Platt (1993) estimate that there were at least ten Chinese dialects in use during the colonial era. The use of what were known (almost pejoratively) as Chinese ‘dialects’ was actively discouraged from around 1979, with the launching of the Speak Mandarin Campaign following the rise of progressive economic development in China (see Newman 1988), and has been shown to be successfully implemented, with the use of the southern Chinese languages now being reduced considerably. However, the influence of the many substrate languages has continued to be observed in Singlish, and still continues to be an important factor in its usage today.
6While Ho and Platt (1993) referred to Singapore English as on a ‘contact continuum’, and discussed the notion of style-shifting according to the speech setting, the interlocutor, and the formality of the speech event, the sociolinguistic situation of Singapore English has often been described as diglossic (see Ferguson 1959). Gupta’s original (1991, 1994) diglossic references described Singapore Colloquial English (now relabelled Colloquial Singapore English), as the L(ow), and Singapore Standard English as the H(igh), a sub-variety which according to Gupta, varied little from any other international standard. The two diglossic poles were later revised by Alsagoff (2010) in terms of a Cultural Orientation Model (COM) with the functional roles of a diglossic situation represented by a global sub-variety as the H and a local sub-variety as the L, a description more reflective of the present-day options available to the speakers. However, unlike many diglossic situations, the H and L in Singapore are both constantly linked in an inescapable contact situation and the contact situation is thus relatively complex: although the substrate languages may not have had so much of a direct influence on the English spoken in the region today, the effect of having four official languages (Mandarin Chinese, English, Malay and Tamil) does have influence, as well as the contact between the standard sub-variety and the form known as CSE or Singlish. As a result, it is often quite difficult to determine where borrowings come from (e.g., the frequently discussed discourse particle lah has been said to have come from either Malay or Hokkien, according to Lim 2007), and code-switching between Singlish and the standard, as well as between Mandarin and Singlish, is becoming increasingly reader2021-02-17T19:37:00REAcommon (personal observation). Other recent studies, such as Leimgruber (2009) and Deuber, Leimgruber and Sand (2018) used the model of social indexing (Eckert 2008) to describe the way in which Singaporeans use elements from the contact language pool to self-identify and express their allegiances to a particular social group.
7Bao (2015), arguing for a more continuous feature pool in the development of Singapore English, refers to the morphosyntactic consistency between the southern Chinese languages and Mandarin, thus ruling out any loss of substrate influence in the current situation in which Mandarin is expected to have replaced the southern Chinese languages as a contact influence. At the same time, he discusses the constant presence of Singapore Standard English (SSE) in the contact feature pool, a situation which accords with the endogenous situation described by Chaudenson (2001) in which the developing contact variety is continuously in contact with the lexifier. Bao refers metaphorically to a “gene pool which is continuously replenished from the same linguistic stock” (2015: 31) from the time of colonization. It is with this observation in mind that the present study considers Singlish as demonstrating a convergence towards the lexifier (the local standard variety in the present study), in which the encroaching presence of past tense in the standard variety is in competition with the older, substrate-influenced means of marking aspectual functions using already. Bao (2010) had previously discussed the lexifier as having a restraining influence on substrate-dominated functions, such as in the marking of the passive using the co-verb give in a serial-verb construction (e.g. He give me scold ‘He scolded me’ (I was scolded by him) (Bao 2015: 233), described as an unsuccessful transfer due to its incompatibility with the surface syntax of the standard variety. Bao’s analysis pointed to the role of a ‘Lexifier Filter’, and it will be shown later that this ‘filter’ is pivotal in explaining the changes taking place in the use of already in Singapore English.
8Following Platt et al (1984) and other descriptive accounts of the use of already in the late 20th century, the earliest comprehensive account of the aspect marker was Bao (1995), whose study attested to the replication of all the functions found in Mandarin Chinese le. Where Platt et al (1984) and others of that time had referred to the use of already as marking completive meanings, Bao (1995) described it as a perfective marker, marking not only completive meanings but also the functions of an inchoative marker, something which had been observed by Platt et al (1984) (see (2)), but not given a formal description at the time. Bao’s study was further developed in his 2005 study, in which the suggestion of a formal interaction between the substrate and the lexifier was proposed, the transfer from substratum languages entailing an entire system, not just isolated features, and involving the mediation of the lexifier. The interaction was outlined in the following manner:
|
(4)
|
a. SYSTEM TRANSFER (SYSTRANS)
Substratum transfer involves an entire grammatical subsystem.
b. LEXIFIER FILTER (LEXFILTER)
Morphosyntactic exponence of the transferred system conforms to the (surface) structural requirements of the lexical-source language.
|
9The mediation of the Lexifier Filter required that the transfer from the substrate should comply with the morphosyntactic constraints of the lexifier, so that, for example, already could not appear verb-finally in Singlish, even though it was possible in Chinese (Bao 2015: 218):
|
(5)
|
I wash my hand already
|
|
(Mandarin Chinese: wǒ xǐ-le shǒu
*I wash already my hand
|
10Only if the verb was the final element in the clause could already appear verb finally; i.e. when it was clause-final at the same time.
11Bao (2005) classified three functions for Singapore English already: the completive, marking a complete event occurring before the moment of speaking or before a reference point in the discourse, the inceptive, marking the beginning of an action, and the inchoative, marking a change of state in the present. These functions are represented in the following example, which is 3-ways ambiguous between all three functions:
|
(6)
|
Miss Lim eat cake already
a. ‘Miss Lim ate/has eaten the cake’ (completive)
b. ‘Miss Lim has started to eat/is about to eat the cake’ (inceptive)
c. ‘Miss Lim eats cake now (and she did not before) (inchoative)
|
12In (6c) it can be seen that the inchoative not only refers to a change of state, but also the change of a habitual activity (this can be seen in the negative to indicate a discontinuity of a past state as in (2), a use which could never occur in standard varieties of English). In addition, completive uses can refer to the discourse function of bounding off a subordinate clause to indicate temporal sequencing:
|
(7)
|
After you wax the car already, I pay you
|
|
(Bao 2005: 239n).
|
- 3 According to Heine and Kuteva’s (2003; 2005) model, the R-language in such cases would be both Sing (...)
13Other studies which followed Bao (2005) referred to the process of contact grammaticalization to describe the use of the aspect marker in Singlish; notable amongst these was Matthews and Yip (2009) who described already in Singapore English as a case of Ordinary Contact-Induced Grammaticalization, using the model of Heine and Kuteva (2003, 2005); i.e. a process in which a functional model language (M) uses material from a replica language (R), to replicate the function, drawing on universal strategies of grammaticalization.3 Universal strategies in the case of already are reflected in traditional typological surveys, and as can be seen below, the use of grammatical morphemes derived from lexical sources meaning ‘already’ or ‘finish’ is widespread in the Southeast Asian region. Regardless of which substrate language is the model, then, the substrate functional needs in either Chinese languages or Malay have proven to be instrumental in triggering the grammaticalization process: Szeto, Matthews & Yip (2019) also demonstrate the significance of substrate factors, and find the presence of already in the bilingual Cantonese-English acquisition stages of young children, even though it disappears later and is never found amongst adult speakers. The application of a grammaticalization approach lends a new perspective to the descriptive accounts of already in Singlish, and raises the question of how it is accommodated into other grammaticalization accounts of perfect or perfective markers universally.
14The role of grammaticalization was also taken into account in Ziegeler & Lee (2019) a study comparing the extent of constraints observed from the source meanings of already in Singlish and Colloquial Malaysian English (CME), constraints which could be inhibiting the generalization of already in referring to accidental events. It was found that although the Chinese and Malay substrate languages could readily use their completive aspect markers (le/liao in Chinese, and sudah/dah in Malay) with expressions describing accidental events, this usage was frequently rejected in Singlish and CME, particularly amongst older speakers. Thus, it was hypothesized that it was lexical retention or persistence from English, rather than the substrate, that was imposing these constraints on its use, since English already is also infelicitous when co-occurring with expressions of accidental events.
15Other studies, which settled for a substrate-based explanation instead, included Li and Siemund (2020), a diachronic study using a large corpus of data compiled from the Singapore National Archives Oral History Interviews (detailed below). Their study also found that numerous other post-colonial Englishes used already as an aspect marker, thus attesting to its universality. Li and Siemund referred to calquing from Mandarin Chinese le to explain its presence in both contemporary Singlish and the Singlish spoken in the early post-colonial period, and mentioned in addition the process of imposition (Winford 2003) in which the speaker’s most dominant language in a contact situation is most instrumental in transferring functions to the less dominant language. The explanation of imposition was also used by Teo (2019), using narrative data obtained from interviewing 20 Singaporean participants (12 Chinese and 8 Malays) to discuss the same process as the reason for the higher frequency of inchoatives or non-completive functions of already amongst Singaporean Chinese than amongst Singaporean Malays. Although the process of imposition is likely to play a role, it is hypothesized in the present study that there is a more motivating factor of convergence towards the lexifier, and the ever-increasing shift to past tense across the dialect creates the appearance of a more frequent use of inchoatives overall.
- 4 See Dahl (1985) for more details on the distinctions between perfects and perfectives. Bybee and Da (...)
16As noted above, the tendency in late 20th century studies was to describe already as a marker of perfect or perfective aspect, often with little discrimination of the ways in which each of these terms is used.4 More recently, the World Atlas of Linguistic Structure (Haspelmath, Dryer et al (2008) revealed 21 forms derived from lexical sources meaning ‘already’ or ‘finish’ in the Southeast Asian and West African language regions, suggesting the presence of an areal feature, possibly spread by contact. Dahl (2006) and Olsson (2013) isolated this grammatical type (or gram) under the label of iamitives, a category also found in the Portuguese-based creoles of India and Sri Lanka (Kuteva et al 2019, Clements 2011) with functions similar to the perfect, but also expressing meanings of current relevance and transitions to new situations (Dahl and Wälchli 2016). Dahl and Wälchli (2016) and Dahl (2019) increased the number of languages with iamitives to 370, using a Multidimensional Scaling Model survey, and other examples derived from Spanish ja ‘already’ had been found in the L2 Spanish of a Chinese immigrant living in Spain (Clements 2003: 57). Clearly, it would seem that Singlish already is a typical universal strategy for grammaticalizing the same functions of iamitives as found in other languages.
17However, as already had been previously observed to carry the functions of a perfective marker (and sometimes, a perfect marker), it may be questioned why such markers need an additional label to distinguish them from perfectives. The lexical source alone does not provide the justification for isolating a new gram-type. The relations between iamitives and perfect(ives) may be seen in Olsson (2013), in which 7 main semantic and pragmatic characteristics are seen to be associated with this gram-type, crosslinguistically. These are listed in Table 1, showing overlap with the English perfect and cross-categorized with the sub-categories of already in Bao’s (2005) study: inchoatives (I), inceptives (ICP), and completives (C).
Table 1. Characteristics of iamitives, as detailed in Olsson (2013) and showing overlap with Bao’s (2005) sub-categories of already, and the English perfect aspect.
|
Characteristics of iamitives (Olsson 2013)
|
Bao (2005)
|
Overlap with the English perfect aspect
|
|
a. Co-occurrence with stative predicates, rendering inchoative senses.
|
(I)
|
|
|
b. Co-occurrence with dynamic predicates, rendering either the completion or inception of an action.
|
(C)
(ICP)
|
Completive functions only
|
|
c. Occurrence with imminent or prospective future, relating the present state to some future event
|
I
|
|
|
d. Expressing progress so far (associated with accumulated events or actions)
|
C
|
Possible in British English, with adverbs such as yet, so far etc.
|
|
e. Expressing discontinuity of a previously-holding state (= ‘no longer’).
|
I
|
|
|
f. Co-occurring with adverbial meanings of ‘at last’
|
C
|
Optional co-occurrence of the perfect
|
|
g. Sequentiality in discourse – one event happening before another.
|
C
|
Possible in some temporal/conditional clauses.
|
18From Table 1, it can be seen that the class of iamitive aspect is partially overlapping with the class of perfects in English, and so only some of its functions can be seen as relevant to that category. The additional functions which justify the isolation of this gram-type are those of the inchoatives and inceptives, in rows (a), (c), and (e) on the table.
19Given the partial overlap of functional characteristics between iamitives and the English perfect, the question is whether iamitives are at an earlier stage, grammaticalizing to become a perfect, or at a later stage, grammaticalizing on further to develop into a different category. Although the shared characteristics in Table 1 are only relevant to the English perfect, Kuteva et al (2019) place the iamitive prior to the development of a perfect on a crosslinguistic grammaticalization path. Bybee et al (1994: 105) mapped the progress of verbs across 76 languages from sources meaning ‘finish’ or expressing direction, roughly as follows (with anterior substituted for ‘perfect’):
|
(8)
|
(finish/directional verbs) > completive > anterior > perfective/simple past
|
20Although they did not include already as a lexical source for perfects, verbs meaning ‘finish’ were categorized together with ‘already’ sources in Haspelmath, Dryer et al (2008) (see above). In Bybee et al’s (1994) study, the target grammaticalization function of a perfective was past tense. Until then they had no actual data supporting this prediction, but they assumed that past tenses would naturally evolve out of perfectives because they were seen to be older than perfectives in the languages studied. It may well be the case for many perfectives (and indeed, Bisang (2004) has observed tense-like functions for Mandarin Chinese le in some syntactic contexts).
21Bybee et al (1994) describe inchoatives as a function emerging when a perfective marker is extended to apply to a stative predicate. However, this could possibly be described as taking place when the current relevance inferences associated with a perfect aspect are over-analyzed as current results (of completed changes of state). One typical example is from Olsson (2013: 18):
- 5 The following abbreviations apply throughout: SG/s: singular, NEG: negative, ANT: anterior, PF/PFV: (...)
|
(9)
|
Jakarta Indonesian
|
|
Kamu
|
tidak
|
bias
|
memakan-nya.
|
Itu sudah
|
busuk.
|
|
2SG
|
NEG
|
can
|
eat-3
|
it
|
rotten5
|
|
‘You can’t eat this. It is rotten.’
|
22In (9), the aspect marker sudah (‘already’) also found in Malay, indicates a state change from not being rotten to being rotten, thus shifting the focus from the event changing the state to the outcome of the state change. This is described by Olsson (2013) as a focus on a ‘new situation’, and, it will be seen in the data below, that some of the iamitives described as ‘new situations’ have an almost predictive sense and could never be associated with past events, but only with the relevance of the current situation to the immediate future (category (c) in Table 1). In this way, already and its counterparts could well be described as markers of prospective, as well as retrospective aspect, in a semantic generalization with stative predicates that may be described as shifting from:
|
(10)
|
completion > resulting state > future consequence
|
each phase in (10) being inferentially derived from the previous one. Such a shift rules out any form of development to past tense functions – such functions may be taken care of with non-stative predicates, rendering completive functions, as in (8). The use of a completive marker with a stative predicate, though, is only a metaphorical extension of its literal meaning, since states can never be completed, and thus by inference will be seen as a change of state, from which there may be future consequences. However, the possibility of past tense functions in the grammaticalization of iamitives has been observed in creoles, as discussed below.
23While already in Singapore English is unlikely to progress on to become a tense marker, its counterparts have been observed in some Portuguese-based creoles to express such meanings, as in the following example, from Kristang, a creole spoken by no more than 2500-3000 speakers living in Malacca, Malaysia, and Singapore (according to Hancock 2009). One example comes from Baxter (1988):
|
(11)
|
Kristang
|
|
|
|
|
eli
|
ja
|
bai
|
mar
|
(onti anoti)
|
|
3s
|
PF
|
go
|
sea
|
yesterday night
|
|
'He went fishing (last night)'. (Baxter 1988:119)
|
24The perfective marker (PF), ja, in Portuguese is also derived from an adverb meaning ‘already’, and, given its co-occurrence within a specific time-frame, it could not be translated here as anything other than a past tense. Baxter also adds that it cannot be glossed using sudah ‘already’ in Malay:
|
(12)
|
Malay
|
|
dia
|
pergi
|
laut
|
(tadi malam)
|
|
3s
|
go
|
sea
|
ago night
|
|
'He went to sea last night' (Baxter 1988: 101)
|
25Other data from Bidau Creole Portuguese (Baxter and Cardosa 2017) reveal the use of ja glossed as perfective with stative predicates, but past when the predicate is non-stative. In Korlai Creole Portuguese (spoken on the west coast of India), double marking appears, with ja reanalysed as a past tense marker alongside inflectional marking of past tense, as in the following:
|
(13)
|
Korlai Creole Portuguese
|
|
Lorenz
|
ja
|
Kato
|
|
Lorenz
|
PAST
|
Sang
|
|
‘Lorenz sang’ (Clements 2006 : 87)
|
26In (13) ja appears redundantly as the affix -o on kato is already a form of past tense marking found in Marathi, an adstrate language of the region. Clements notes, however, that ja is now rapidly disappearing in such contexts, leaving only the affixal form of marking past tense, and Clements (2003) and Clements and Koontz-Garboden (2002) note that Korlai and Daman (another Portuguese creole spoken close to Korlai in India) have a much greater number of morphological affixes than creoles such as Kristang in Malaysia. The reason, they suggest, is the more continuous presence of the lexifier, European Portuguese, in the region (for Daman it lasted from 1559 until 1961, and for Korlai 1594 until 1740). This was not repeated in Malaysia, where the Portuguese lexifier of Kristang was found only infrequently after the early 17th century, and although Baxter (2018) notes a possible intermittent diglossic presence of Portuguese in Malacca until the 19th century, it was not sufficient to influence the development of more complex morphological means of marking past tense, with ja surviving longer in Kristang than in Korlai. Thus, although we have no evidence for already being used as a past tense marker in Singlish, other iamitives have been translated as past tenses in various creoles.
27Little research has been undertaken to date on the use of past tense marking in Singlish, and Alsagoff and Ho (1998: 137) had claimed that past tense was not marked morphologically in the dialect. However, as noted above, Ho & Platt (1993) offer an exhaustive study of the level of past tense usage in Singapore English in the late 20th century. The factors controlling its use, according to Ho & Platt, range from the lexical aspect of the verb (whether punctual, stative or non-punctual), factors of pronunciation when the regular past tense suffix -ed forms a cluster which the consonant immediately preceding it (e.g. walked) – such factors result in a reduction of the consonants regardless of the morphological status they represent. Other factors included the education level of the speaker, and the grammatical aspect of the sentence, with non-punctual aspect referring to the expression of habitual or iterative action or progressive aspect. In their study, they found that punctual actions were marked in general more than statives, (56.2% as against 36.9%) and statives more frequently than non-punctuals (36.9% as against 14.7%). Notably, past habitual actions were frequently described using the present tense, as they referred to non-punctual events. Ho and Platt’s (1993) study was one of the first studies to reveal that use of past tense in Singlish was variably controlled by a range of factors, both semantic and phonological, as well as sociolinguistic.
28In a later study, Ziegeler (2012) found that, while past habituals in Singlish were often left unmarked for tense, other observations of past tense marking often seemed to reflect what seemed like a bizarre mix of uses. For example:
|
(14)
|
(Original Post: Sorethroat, Any remedy to relieve/cure it?)
|
|
yest 23th June. . . sore throat severe again and slight fever. . .ate Chinese Med now. . .
cos western med cant help me. . .
‘Yesterday, 23rd June, [I had a] sore throat again and a slight temperature. [I] take Chinese medicine now, because Western medicine can’t help me.’
http://www.flowerpod.com.sg/forums/index.php?showtopic=24344&st=40 (24/06/08)
|
- 6 Observed personally by Amelyn Thompson.
|
(15)
|
… they graduated next year [talking about some friends]6
|
- 7 The discourse marker used in this example, la, is often used to emphasize or reinforce an assertion (...)
|
(16)
|
Prunes good enough la. I just ate it as a snack and i lao sai:lol7
|
|
‘Prunes are good enough. I just eat them as a snack and I get diarrhoea: ‘laugh out loud’.’
http://www.flowerpod.com.sg/forums/index.php?showtopic=1387 (24/06/08)
|
|
(17)
|
Have forwarded your message to Lilyn who will be contacting you as I no longer worked in the clinic. Have a blessed day.
|
29Some cases involved the use of past tense where the present perfect would be generally preferred:
|
(18)
|
That lemon tree was in the ground for about two months [it was still in the ground] (Ziegeler 1995)
|
30What all these examples have in common is that the use of past tense can be readily explained with reference to the characteristics of iamitives in Table 1 above: in (14) it is the inception of an action (b), in (15) it is (c), that of imminent future, in (16) it is (g), the marking of sequentiality in discourse, in (17) the past tense marks the discontinuity of a previously-holding state (‘no longer’), characteristic (e). (18) appears to demonstrate the characteristic listed in (d), of progress so far. It is hypothesised that in such examples, the past tense is replacing the use of already in order to avoid using a typically Singlish (and hence, stigmatized) form; in other words, although already is not becoming a past tense marker in Singlish, it might be considered as yielding place to the past tense in certain non-standard functions, which have been seen as hypercorrections in the examples above. The following study will attempt to determine if this is the case, and if it is an indication that the standard past tense is taking over some of already’s functions in Singapore English.
- 8 The National University of Singapore Faculty Staff research Support Scheme (2008) is gratefully ack (...)
31In order to investigate the possibility that the past tense is replacing already in Singlish, the aspect marker was first searched in two recent corpora which were compiled more than ten years apart from each other. The International Corpus of English Singapore (ICE-SG) is a small corpus of around 1 million words, constructed in the early 1990s (according to Hiramoto 2015), and containing spoken as well as written data. The first 120 files were used, as they contain the most colloquial spoken data in which the form is more likely to appear; already is not a feature of the standard sub-variety (the written part of the corpus contains mainly Singapore Standard English). The first 120 files of the ICE-SG amount to around 217,000 words, according to Deuber et al (2018). The second corpus, the Flowerpod Corpus, of over 700,000 words, was compiled in 2008-9, although it contains material dating back to reader2021-02-17T19:51:00REA2005;8 thus, there is at least a ten-year time period between the two corpora, enabling a diachronic perspective on the data. The Flowerpod Corpus has the advantage of being comprised solely of internet blog conversations, thus producing a more informal, unedited style than the ICE-SG offers, and retaining also the anonymity of its contributors, although the results show that already was not particularly sensitive to the formality of the discourse.
32Each token extracted was classified as either Completive, Inchoative, or Inceptive, the sub-categories of Bao (2005), and checked against the criteria in Table 1, from Olsson (2013). Excluded were tokens which were unambiguous as standard English examples, although there were three tokens in the ICE-SG corpus which were unresolvedly ambiguous between Standard English and Singapore English – these were also excluded. Pre-verbal uses were excluded for this reason, as it was not possible to determine whether they were being used standardly – the survey was therefore based on clause-final or sentence-final examples (and even in standard English, already may occur in sentence-final position sometimes; however, the sentence or clause-final uses most often corresponded to inchoative meanings, which are not amongst the functions of standard English already). The first survey produced the following results:
Table 2. Frequencies of various functions of Singlish already in the ICE-SG Corpus and the Flowerpod Corpus. (χ2 = 11.29, p = < 0.003)
|
Completive
|
|
Inchoative
|
|
Inceptive
|
|
Totals
|
|
N
|
%
|
N
|
%
|
N
|
%
|
|
|
ICE-SG
|
55
|
42.63
|
64
|
49.16
|
10
|
7.75
|
129
|
|
Flower-pod
|
44
|
31.88
|
92
|
66.66
|
2
|
1.44
|
138
|
|
Totals
|
99
|
36.3
|
156
|
58.42
|
12
|
4.49
|
267
|
33Interestingly, there were no creative spelling variants of the aspect marker in the blogs and the only available form used the standard spelling, already, although some features of text-talk spellings appeared elsewhere. The following are examples of each category:
- 9 One and nuh, also lor and hor in (19b) are borrowed discourse particles in Singapore English; see W (...)
|
(19)
|
Completives/anteriors
|
|
a. Sequentiality in discourse (Type (g) in Table 1)
All these not cooked yet <ICE-SIN:S1A-007#10:1:B>
You eat already you can die one nuh <ICE-SIN:S1A-007#11:1:B>
‘If you eat them, you can die’9
|
|
b. Completion of an action with dynamic predicates (Type (b in Table 1)
Flowerpod Corpus: Posted by leadtin 12-10-2008 11:07 PM
1130 showing some luxury pens lor. 200k USD fountains.
then the $20 muji alu pen hor... Is made of light and thin aluminium one. i dropped and got deep deep dent already.
|
|
c. Progress so far (Type (d) in Table 1)
Flowerpod Corpus: Posted by diamondsinated Jul 24 2005, 04:02 PM
i wanna get the mascara base too.. but out of stock at bugis and fareast branch when i went to look for it last week.. anyone saw it in the stores already?
|
|
(20)
|
Inchoatives
|
|
a. Co-occurrence with stative predicates (Type (a) in Table 1)
She continued making noise ya and then he asked her are you cold
cool already <ICE-SIN:S1A-036#167:1:B>
|
|
b. Occurrence with imminent/prospective future (Type (c) in Table 1)
Flowerpod Corpus: Posted by qianz Dec 25 2005, 03:02 AM
have finished a large bottle of manuka honey umf 15. but doesn't seem to have any effects on me. still suffer breakouts. shall stop already cos it's so expensive.
|
|
c. Discontinuity of a previous-holding state (= ‘no longer’) (Type (e) in Table 1)
Flowerpod Corpus: Posted by miss_Bronze Apr 28 2006, 12:40 PM
ya lo dear.. i have a feeling if we wash it, it will become slighly [sic.] rougher and it will no longer be as soft and fluffy already.. hmm.. the most get new one at $10 lo [‘the most you can do is to get a new one for $10’]
|
|
(21)
|
Inceptives (Type (b) in Table 1)
|
|
a. Has it started yet <ICE-SIN:S1A-019#7:1:B>
It's started already <ICE-SIN:S1A-019#8:1:B>
|
34As seen in Table 2, there is a difference of 10.75% between the two corpora in the frequency of completives, and 17.5% in the frequency of inchoatives, according to these data, and if the time periods of the two corpora are taken into account, this represents a 10.75% decrease in the use of completives and a 17.5% increase in the use of inchoatives. The frequency of inceptives remains below 10% for both corpora, which is less informative in terms of their diachronic frequency across the two corpora, although in the later period they appear to be less frequently used, at only 1.44% of the total spread of functions.
35The data may also be compared with the results of Li and Siemund (2020) whose corpus was extracted from the National Archives of Singapore Oral History Interviews. The speakers in their data were born between 1899 and 1958, with an average year of birth being 1924, and were interviewed between 1979 and 2008; thus providing a valuable diachronic source for the study of the emerging dialect, reflecting most closely the Singlish of the early 20th century. In Table 3, the proportions of completive vs. non-completive already in the ICE-SG and the Flowerpod corpora are compared with Li and Siemund’s data from the Oral History Interviews; this is also compared with the frequencies of completive vs. non-completive already obtained from Teo (2019) based on interviews of 12 Singaporean Chinese and 8 Singaporean Malay speakers:
Table 3. Distribution of completives vs. non-completive tokens of already across 4 separate studies, showing chronological increase of non-completives relative to completives (χ2 = 133.74, p << 0.00001)
|
Completives
|
Non-completives
|
|
Li & Siemund (2020) OHI
|
228 = 71.9%
|
89 = 28.07%
|
|
ICE Sing.
|
55 = 42.63%
|
74 = 56.91%
|
|
Flowerpod
|
44 = 31.88%
|
94 = 68.1%
|
|
Teo (2019) (both groups)
|
80 = 28%
|
205 = 71.9%
|
36Clearly, there is a steady rise shown in the proportions of non-completives to completive functions, something that Teo (2019) in his study had attributed to the mother tongue languages of each ethnic group surveyed (Chinese and Malay Singaporeans, though he did not survey Indian Singaporeans). Teo had found a higher proportion of non-completives with Chinese Singaporeans than with those of Malay ethnicity. However, even in his study, the overall frequency of non-completives to completives within the Malay Singaporean group was still higher. For Li and Siemund’s study, it could be argued that the corpus was composed mainly of narrative, thus lending itself to a greater preference for completive uses of already in describing past events. However, Teo (2019) also used the genre of narrative to derive his data, and the ratio of non-completives to completives was still found to be higher. Figure 2 illustrates the distribution of these ratios across the four time periods represented in the corpus compilation, shown on the horizontal axis: Li and Siemund’s (2020) OHI, representing the early 20th century (1), ICE-SG, for the period ending 1995 (2), the Flowerpod, mainly between 2008-9 (3), and Teo (2019) (4), the most recent.
Fig. 2. Distribution of completive vs. non-completive already in Singlish, across 4 ascending time periods according to corpora studied.
37As noted earlier, Kuteva et al (2019) place iamitives derived from source meanings of ‘already’ or ‘finish’ in their catalogue of crosslinguistic grammaticalization strategies for perfects and perfectives, suggesting that the pathway may eventually lead to the development of past tense functions for such grams. This has been illustrated, to some extent, in the case of Kristang, in section 3., though it is not illustrated in the case of Singlish already. On the other hand, we see the hypercorrect uses found in section 3.1, which give the impression of past tense being used non-standardly where an iamitive would have been expected, but was not used. It may be questioned why this is so.
38From the data shown above, the clear predominance of inchoatives or non-completive functions seen in more recent years in Figure 2 and Tables 2-3 are hypothesized to be due to the relative reduction in the use of completives, rather than the sudden rise of the use of non-completives per se. The figures may only be a slight indication of what could be revealed with larger amounts of data, but there is a consistent and steady decrease in completives relative to non-completives, which is also showing in Teo (2019). At the same time, there is no indication that already may be grammaticalizing onwards to become a past tense marker, as predicted for iamitives sharing the grammaticalization trajectory of perfects and perfectives, crosslinguistically. This is hypothesized to be due to the increase of past tense marking in the functions normally occupied by completive already, which would be unlikely to progress on to become a past tense marker due to their co-existence with past tense already established in the environment in the standard sub-variety. The operation of Bao’s (2005, 2010, 2015) Lexifier Filter, in this way, would not permit the use of already to act as a past tense marker when past tense was incompatible with the features of the English adverb, and standard past tense morphology was already available in the contact language pool to serve the functions instead.
39Thus, on the surface, it may appear as though already has reached the end of its grammaticalization cycle, and is giving way to another morphosyntactic element to inherit the stages which would otherwise be predicted of already itself. The process by which such take-overs occur is referred to in grammaticalization studies as renovation, or renewal (Lehmann 2015, Hopper & Traugott 2003), a means of providing continuity of function when an older form becomes too eroded or opaque to be able to continue expressing that function. Quite often the older form and the newer form may co-exist side-by-side in a situation of ‘layering’ (Hopper 1991) for some time before the older form eventually disappears in a process known as ‘loss’ (Hopper & Traugott 2003: 172). Hopper & Traugott (2003) provide the example of the French passé simple being renovated by the passé composée historically (e.g. elle s’évanouit ‘she fainted’, renewed by elle s’est évanouie ‘she fainted’), in which it is a more periphrastic form that replaces the former synthetic form expressing past tense by affixation. Renovation has been discussed in the contact grammaticalization literature (see Heine & Kuteva 2005), but with little reference to the presence of a lexifier as a source for the renewal. The process does not necessarily refer to a post-grammaticalization stage of a grammaticalizing form, but rather to the perpetuation of a grammaticalizing function through different formal means.
40In the case of Singlish already, the renovation by the past tense follows after what seems like a long period of layering, observed in Ho and Platt’s (1993) account of past tense marking in Singlish. The layering situation exists in a contact feature pool in which there is continued competition between the substrate languages, the adstrate languages, and the continued presence of a standard lexifier placing controls on what may eventually grammaticalize in the dialect: already would be inconsistent with the lexifier grammar, and it is for this reason that the adverb does not occur directly after the verb as its counterpart could in Chinese languages. Moreover, the sociolinguistic situation, which discourages the use of the non-standard Singlish sub-variety, also has a role to play in the promotion of standard forms.
41Thomason and Kaufman (1988) regarded decreolization as a form of convergence towards the structure of the lexifier, and although Singlish is not typically considered a classical creole, the situation is comparable: seen from the point of view of Singlish, the renovation of already by the past tense in the colloquial sub-variety may be regarded in the same way as a type of convergence towards the lexifier. Seen in terms of grammaticalization processes, though, with past tense marking as a target function, it is a case of structural renovation across contact that is encroaching on and replacing the completive functions of already. The inchoative functions are screened out by the Lexifier Filter of Bao (2005, 2010, 2015) since they could never be compatible with the morphosyntax of English grammar, and so they remain ostensibly as the most frequently-used function of already in the present-day data, representing a kind of ‘selective’ renovation: only some of the functions of already are usurped by the English past tense. However, as shown in section 3.1, even then there is some overlap where past tense is used non-standardly in functions of already not selected for renovation.
42That the continued presence of the lexifier in the Singaporean English contact pool is instrumental in promoting this renovation is clear from the comparisons with other creoles in section 3; where Portuguese-based creoles with more continuous contact with their lexifiers spoken in India were seen to develop affixal morphology much faster than those which had been isolated from their lexifiers for longer periods, such as Kristang in Malaysia. The contact situation in Singapore, which could be described as a contact diglossia, ensures that both the colloquial and the standard sub-variety are in constant contact with one another and have been since colonization, as noted by Bao (2015).
43Since it is hypothesized that it is past tense which is renewing the functions of the completives, one might also expect to see a concomitant increase in the use of past tense forms over the same periods shown, and in the ideal situation, an attempt to quantify such uses. However, given that past tense morphology is marked on open class items, this would present an overwhelming task to accomplish within the scope of the present study, with results that might not be justifiable, in any case, without taking into consideration the factors underlying the use of past tense which may not always equate with the factors underlying the use of completives. Therefore, a small sample of cases was taken in which already appeared most frequently (twice or more) with the same verb, in order to compare the frequency with which these verbs appeared in past tense or past participle form with or without already. (The verbs selected were from the same portion of the ICE-SG corpus as used in the rest of the survey.) It was found that the past tense forms most frequently used with already: bought, change(d), finish(ed), and forgot(ten), were up to 27 times more likely to appear as bare past tense forms without already than with it, and even with a verb such as finish, which is virtually synonymous with the functions conveyed by completive already, the bare past form was likely to appear without already at least 4 times more frequently. Thus, even with the lexical past tense verbs most favoured by already, the overall preference for the bare past tense without already was more preferred.
44The present study supports the earlier study by Teo (2019) in showing that there is a gradual narrowing of functional range exhibited by the iamitive aspect marker already in today’s Singlish, with the decline of completive functions leaving mainly non-completives to be marked obligatorily by already. The relative reduction of completive functions is explained by a selective renovation by the past tense, and a convergence towards the standard English lexifier grammar, motivated and assisted by the stigmatization of Singlish in the national language policy of Singapore. However, this type of convergence may also be considered as a form of renewal of an iamitive function in contact grammaticalization, since there is evidence from some Portuguese creoles that similar functions of the adverb ja (meaning ‘already’) can progress on to become past tense forms. In the Singapore case, the continued presence of the lexifier in the contact environment (also discussed in Ziegeler 2015), acts as both a motivating force and a controlling force on the grammaticalization of the iamitive, in an endogenous situation in which the constant interaction with the Lexifier Filter ensures that renovation by past tense is only selective and therefore there is limited convergence towards the standard sub-variety with which it co-exists.
45Further research in this area might consider whether the functions of inchoatives and inceptives, being incompatible with the lexifier grammar, remain indefinitely in the dialect to mark it as a distinctive, identifying feature, or are replaced eventually by other adverbial forms more reflective of standard usage. Ideally, it would also be necessary to investigate whether the same tendencies may appear in the other contact English situations noted in Li and Siemund (2020), such as Malaysian English (which has not experienced such prolonged exposure to standard English as has Singlish), Krio (Sierra Leone Creole) or Cape Flats English (S. Africa).
Flux RSS
