Lee, Nala H: A Grammar of Modern Baba Malay

1 Typological summary

The author provides an overview of the typological properties of Baba Malay in chapter 1: eight vowels /i e ɛ u o ɔ ə a/ (a high number when compared to many other Malay varieties) and 19 consonants (p. 33). As for morphology, there is almost no affixation (pp. 4–6, 91, 107–110, 234–238). Some original Malay affixes are integrated in verb roots but they are not productive; for instance, there is no meaning contrast between nangis and menangis ‘to cry’ (pp. 112–114). In other Malay varieties, men- has several functions, such as an active or progressive marker. The only affixes used productively are ter- ‘accidental, movement’, and the transitivizing -kan. There is also some productive reduplication (deadjectival adverbs, tentative aspect but almost no noun reduplication for plural) (pp. 60–62). Basic word order is AVO and SV (p. 6–7, 197–198). Noun phrase order is more flexible: ADJ N beside N ADJ, GEN N beside N GEN, and there are prenominal and postnominal relative clauses (pp. 142–157). The language has prepositions rather than postpositions (pp. 127–135).

2 The speakers

Chinese settlements in Malacca are reported from the 15th century. The speakers of Baba Malay are descendants of (male) traders from the Southern Chinese mainland – speaking Min dialects like Hokkien – who came to Malacca and developed ties with indigenous women of the islands, who were not only Malay speakers but also speakers of Batak languages, Balinese, and Javanese. As a result, the cultural heritage of Baba Malay people combines Chinese and Malay traditions, dress, and religion; to these, British traditions were also added.

The number of speakers is decreasing. Many speakers are shifting to English, a fact which is discernible through the large number of English loans in the examples in the grammar.

3 The names of the language: Peranakan, Baba Malay

The name of the population and language is derived from the Malay word anak ‘child’. The form consists of a middle voice prefix per-/ber, the word anak ‘child’, yielding ber-anak ‘to give birth’. This with the nominalizer -an can also indicate ‘womb’, or it could mean ‘locally born’. According to the author, the term is said to be similar in meaning to the term Crioulo/Criollo ‘creole’ in Portuguese/Spanish.

The term Baba Malay comes from the term baba, a respectful reference term for males, and it is of unclear etymology, perhaps from Hokkien, perhaps from Persian, as has been claimed (Khoo quoted in Lee p. 21), the latter being unlikely. Douglas (1873: 8) gives the word as Amoy Chinese with the meaning ‘halfcaste Chinese from the straits’, whereas Pijnappel (1875: 39) in his Malay dictionary considers it a Chinese loan with the meaning ‘Sir’ (Malay equivalent: tuan). Thus, the term baba or bābā refers to males of (partly) Chinese origin, but the ultimate origin has not been verified.

4 Data gathered

The bulk of the data in the grammar were collected during fieldwork between 2012 and 2014, mostly from four elderly speakers (70–84 years old). Each example in the book is attributed to a named speaker.^[1] Five texts have been included; these represent naturally-occurring speech from a conversation and narratives. Elicited data were also used. Some written poems are also included in the book, but not used in the grammatical analysis because of their stylized nature.

5 Synopsis of chapters

The book starts with an introductory chapter (pp. 1–15) about the background of the language and the main speakers consulted for the work. It also contains a brief typological sketch of the language, and a guide to the grammar, explaining technicalities like the build-up of examples, abbreviations and transcription conventions. The second chapter (pp. 16–32) deals with the two speech communities in Malacca (Malaysia) and Singapore, their history, traditional culture, and their roots in China and the Malay world. The names of the language and the community are discussed, as well as the endangered state of the language with only a few thousand speakers, almost all of advanced age. The community has taken the initiative to preserve the language, however. The third chapter (pp. 33–90) deals with phonetics and phonology. The phonemes are discussed in detail, illustrated with spectrograms and lists of minimal pairs. Phonotactics is discussed more briefly on two pages, notably syllable shapes: V, CV, CCV, VC, CVC, CCVC, CVCC. Resyllabification, which happens when an affix is added, is also discussed. Reduplicated forms, both lexicalized and productive (to mark plurality, intensity, and mitigation), are also discussed here. The chapter further deals with some synchronic phonological processes, such as consonant deletion and vowel assimilation.

Chapter 4 (pp. 91–141) deals with parts of speech. The language has nouns, verbs, adjectives, adverbs, determiners (a category that includes demonstratives, numerals and quantifiers, hence the label ‘noun modifiers’), prepositions, conjunctions and discourse elements (interjections and particles). Chapter 5 (pp. 142–229) covers the bulk of the book in a chapter called ‘syntax’, covering NPs, VPs and APs at length, but with less on sentential syntax or clause combining. The final chapter (6) deals with differences between the speech of the two communities (pp. 230–247). A number of useful appendices follow: a Swadesh 100-list, a list of 71 kinship terms, and a list of temporal expressions (pp. 248–257). The texts are a retelling of a Pear Story (for which see Chafe 1980), three stories based on Aesop’s fables, a conversation of the author with an elder, and ten pantuns (traditional Malay verse) composed by an elder (pp. 257–342). This is followed by a Baba Malay-English lexicon of almost 1,100 headwords, with information on pronunciation, denotation, etymology, and variant forms (pp. 342–357). An English reversal index follows (pp. 358–373). The book ends with a list of tables (pp. 375–376), a list of figures (p. 376), a list of references (pp. 379–384), and an index of grammatical terms (pp. 385–389).

Readers of this review may wonder: where’s morphology? As said above, there is very little morphology in the language, and the few morphological processes of the language are dealt with in the parts-of-speech section (reduplication, the transitivizing suffix -kan, and the accidental marker ter- , which is roughly a non-control marker). The nominalizing suffix -an is dealt with in the noun section, and reduplication in several sections and subsections.

With the above synopsis in mind, let me briefly discuss some further properties of the language. These will show that many grammatical properties have developed fairly recently, as morphemes often have both a lexical and grammatical meaning. Reflexives are formed with the particle sendiri ‘self’ or ‘own’ (adjective). Reciprocals are formed with satu sama satu ‘one with one’, satu sama lain ‘one with other’ or sama sama ‘with with’. Several interrogative pronouns contain the element apa ‘what’. Aspect and tense are indicated with preverbal particles: pernah for experiential perfect (‘ever’), belom for future (‘not yet’), nanti for future (‘later’), ada for progressive, habitual and perfective, baru for recent perfect (‘just’), and sudah for perfective (‘already’). Passives are formed with the verbs kena ‘subjected to’ and kasi ‘to give’. The language also has some serial verb constructions, mostly directional (‘he descend enter’). Sentence structure may not always be AVO, but also Topic-Comment. Indirect and direct speech are unmarked and only differ in the use of pronouns.

6 Surprising properties

I like to think that each language has at least one feature that is more or less unique among the languages of the world. There seem to be fewer of those in Baba Malay than in other languages but there are some. I discuss some of these “unexpected” properties here.

The main meaning of ada in standard Malay is ‘there is’, i.e., locative/existential, and in some varieties it also acts as a possessive verb. In Baba Malay, it has a locative/existential meaning (pp. 115–119), but it has also developed into an aspect marker, with two quite different aspectual meanings: perfective and progressive-habitual. These are two meanings that seem to contradict one another. How can we know that dia ada beli apple ^[2] (3SG ada buy apple) means ‘she bought an apple’ or ‘she is buying an apple’? Unfortunately, Lee (pp. 170–176) does not provide an explanation or a developmental pathway, along the lines of those exemplified in Kouteva et al. (2019).

Kasi originally meant ‘to give’, but in Baba Malay (and other varieties of Malay) it has developed additional meanings as a grammaticalized passive and causative (pp. 164–170). This seems contradictory: passives decrease the number of arguments, and causatives increase them. But it is a pattern found in many languages, including Greenlandic, where the same morpheme is used for causative and passive (Haspelmath 1990).

Another surprising property is the freedom of ordering in the noun phrase (pp. 142–157), but not in the verb phrase or whole sentences. I do not recall any other language where adjectives, demonstratives, possessors (genitives), relative clauses and (some?) quantifiers may be placed both before and after the noun. Impressionistically, the order of elements in the noun phrase is cross-linguistically more fixed than the order of sentential elements, but Baba Malay seems to be an exception.

Most tense-mood-aspect meanings are expressed by preverbal particles in Baba Malay, rather than morphological markers (pp. 170–176). These can be compared to some extent with Germanic and Romance preverbal auxiliaries, which have developed as compensation for the loss of Latin and Proto-Germanic verbal morphology. In pidgins, aspect is not grammaticalized, and mood and especially tense are indicated by adverbial markers with meanings like ‘later’ or ‘before’ (Parkvall and Bakker 2013: 42). Whereas pidgins use adverbs, creoles always (but not only) have preverbal particles for tense-mood-aspect. Baba Malay is creole-like in having preverbal markers like ada ‘progressive, habitual, perfective’, sudah ‘perfective’, baru ‘recent past’. Future tense, however, is expressed with adverbial markers nanti ‘later’ and belom ‘not yet’, a strategy that is more reminiscent of pidgins. These can appear in preverbal position, but also sentence-initially.

As in many languages of the region, politeness distinctions are crucial. Two styles are distinguished: kasar ‘coarse’ and halus ‘refined’ registers (pp. 67–68, 89). Remarkably, these are distinguished via some regular phonological patterns.

A final surprising feature is that clause combining seems to be rare. The grammar does not discuss many subordinate clauses. Conjunctions are mentioned (pp. 135–138, 209–213) but not always illustrated. Are such temporal, causal etc. clauses rarer or more absent than in other languages, or did the speakers use them less in the fieldwork sessions for some reason? That is not clear. Direct and indirect speech are both unmarked. Relative clauses, however, are discussed in Chapter 5 (pp. 193–194) and elsewhere. Postnominal relative clauses are indicated with nang/yang (N nang S), and prenominal clauses are marked with punya/nia (S punya N ….). The latter cannot be used for possessives, presumably because nia/punya is also a possessive marker (Malay punya both means ‘to possess’ and ‘belonging to’).

7 Chinese influence

In a language that is spoken by people of partial Chinese descent, and who continue many cultural traditions of their Chinese ancestors, one question is where we find more recent Mandarin influence in the Singaporean variety, and Southern Chinese in both. Aye (2021) discusses the Chinese lexical contribution to Bazaar Malay and Baba Malay, and Lee (2018) elaborates on grammatical differences. As for the lexicon, Hokkien influence is strong in the kinship system – 55 of the 71 kinship terms derive from Chinese (Aye 2021: 139; Lee pp. 250–254) – as well as in terms for household activities, cooking, spirituality and connected calendric expressions, and clothing. These deal with the more intimate family environment. This may suggest that the Chinese male ancestors and their ancestry continued to be influential in the home. Chinese influence seems also much more extensive than the influences of substrate languages in creole languages, which never exceed 4 % of the total lexicon, and which are usually absent in everyday vocabulary (Bartens and Baker 2012; Mufwene and Condon 1993). In Lee’s wordlist of close to 1,100 words, 16 % are of Hokkien origin, many more than in creoles. Even some of the personal pronouns in Baba Malay are from Southern Chinese (Hokkien), probably stimulated by the avoidance of pronouns by Malay speakers – pronouns that are, by the way, also used in Djakarta Malay and some other contact varieties of Malay.

In the grammatical realm, one can think of the use of a plural marker maɲak (manyak) ‘many’ before the noun, whereas other varieties of Malay use reduplication (p. 62). Not many examples are present in the examples or texts, however. Whereas Malay marks possession through juxtaposition POSSESSED POSSESSOR, Baba Malay uses POSSESSOR CONNECTOR POSSESSED, copied from Hokkien. Yet, grammatical traits of Hokkien are not widespread, and, as discussed above, some of the peculiarities of Baba Malay are shared by other contact varieties of Malay, or pidgins and creoles, and hence not necessarily attributable to Hokkien influence.

8 Contact history

Adelaar et al. (1996) mention the following characteristics of Pidgin-Derived Malay varieties, which are believed to be derived from a pidgin called Bazaar Malay:

1. possessive constructions consisting of possessor + *puña + possessed item;

2. plural pronouns derived from singular pronouns + *orang ‘human being’;

3. retention of *tər- and *bər- as the only productive original Malayic affixes;

4. *ada, the Malay existential marker, indicating progressive aspect;

5. reduced forms of the demonstratives *ini and *itu preceding a noun and functioning as determiners;

6. the use of a reduced form of *pərgi ‘to go’, as a verb as well as a preposition meaning ’towards’;

7. causative constructions consisting of the auxiliaries *kasi/*bəri ‘to give’ or *bikin/*buat ‘to make’ + the head verb;

8. the use of *sama or another word as a multifunctional preposition (also for direct and indirect objects).

Later this one was added:

1. loss of the typically Austronesian symmetrical voice (Adelaar 2004)

Almost all of these are attested in Baba Malay. Possessives are formed with puña (often reduced to nia or mia): POSSESSOR mia POSSESSED. First person plural is kita as in several Malay varieties, but the others are lu-orang and dia-orang (orang ‘people’). Not only is ter- productive, but also the transitive suffix -kan is. Ada as an aspect marker was discussed above. It is less clear whether ni and tu are definite articles or just short forms for the demonstratives ini and itu, but several examples suggest indeed that they are less deictic, e.g., a sentence like anjing tu gonggong ‘that dog barked’. (pp. 93, 112). If it is a demonstrative, it can only be an answer to a question, or a choice of one out of several dogs. Therefore it seems to function as a determiner, referring to an earlier mentioned dog. Elsewhere, tu mia [< punya] anjing gets a clearly deictic translation ‘that dog there’. In Baba Malay, we find both pegi and pi for ‘to go’, the latter more frequent. Kasi ‘to give’ is also used as a causative marker. Sama is used as an instrumental preposition, a comitative, a coordinating conjunction between noun phrases, part of a reciprocal construction, and an adjective meaning ‘same’. In short, Baba Malay should be classified among the pidgin-derived varieties of Malay.

9 Assessment

I profoundly like the grammar. It is well written, clear, and most answers to typological questions can be found in the book. There are 823 numbered example sentences (some repeated, and indicated as such), and those illustrate the claims made in the text.

However, after finishing reading the book, I have a number of questions. First, the language does not seem to have many complexities. Is the language indeed one that is close to the least complex extreme of the range of languages of the world? Or is the language influenced by obsolescence due to the lack of regular use by interviewees? The examples in the book are often extremely short, as if they are cut from longer sentences, and with very simple meanings. Likewise, the texts sometimes give the impression of a hesitant speaker, where English fills many lexical gaps (e.g., friend, from, meat very fresh, used to it, pancake, share). One gets the impression that speakers are less versed in the language that is clearly their mother tongue but probably no longer their daily or dominant language.

Another potential drawback is that there are some issues one cannot test because of the lack of examples, despite the quantity of example sentences. For instance, have the demonstratives developed into definite articles? This is a common process, and almost universal in creolization, and potentially a diagnostic feature here of creolization. In most creoles, reduced forms of the demonstratives function like markers of specificity: specific-known for what is derived from demonstratives, and specific-unknown for what is derived from the numeral ‘one’ (Bickerton 1984; Holm and Patrick 2007). In lexifiers of creoles, definite and indefinite articles have a discourse function: has the referent been mentioned before? One cannot deduce whether it is one or the other just based on isolated sentences, and the possibility is not discussed. In creoles, these “articles” have a reference to the specificity of the nouns. Unfortunately, the texts cannot resolve this. Yet, whenever short forms are used, translations with ‘the’ rather than ‘this/that’ seem more natural (compare short forms on p. 258 exes. (1), (3), p. 259 (6), p. 261 (22), p. 273 (119) with long forms in p. 262 (29), p. 264 (48), p. 269 (85)).

When one deals with the negation of noun phrases, one would expect examples like ‘I see no trees’ or ‘no cars were driving around’. In the section dealing with this type of construction, five examples are given of bukan N, all translated with ‘not X’. But when would one produce an utterance like ‘not onions’ in English? The lack of context makes it impossible to understand the process of constituent negation on the basis of the examples given.

Being interested in clause combining, I was somewhat disappointed with the section on “adverbial clauses”. I had expected complex temporal, locative and manner clauses, but they were all sentences with an adverbial expression of the type “last week”. Luckily the short section on subordinating conjunctions contains a few examples with pasair ‘because’, bila ‘when’ and sunngu.pun ‘although’. Complex clauses are strikingly rare in the material.

Whereas English words are fairly frequent in the examples and in the text (e.g. pear, apple, January, library, aunty, spelled as in English), these are absent in the Baba Malay lexicon in the back of the book. The author probably does not consider them as part of the language, even though some (like apple) have no indigenous equivalent in the material. Only around half a dozen words listed as English are included in the lexicon, perhaps only those that are phonologically integrated. These are: ala ‘exclamation of regret’, alamak ‘exclamation of dismay’, ano ‘anonymous’, bég ‘beg’, berus ‘brush’, botol ‘bottle’, gostan ‘reverse’, kék ‘cake’, minit ‘minute’, numbor ‘number’, paya ‘papaya’, sup ‘soup’. The etymon of not all of these is obvious, e.g. alamak, and some may be Arabic (ten other Arabic loans are listed).

10 What kind of contact language is Baba Malay: creole, intertwined language or something else?

For typologists, the origin of the language is perhaps less crucial a topic than the grammatical properties. However, for those interested in contact language typology or in social typology and its linguistic consequences (e.g., Trudgill 2011), Lee’s grammar is certainly an important one. Here I will discuss the language from a contact and social typology point of view. These views represent my thinking, as Lee only briefly touches upon the language’s typological connection with other contact languages.

Baba Malay is an intriguing language for those who (like myself) work on the linguistic consequences of social events in the past. Interrupted and partial language transmission is the best explanation for the structural properties of pidgins and creoles. The latter share the lexicon with another language whereas morphosyntactic structures are largely innovated through grammaticalization – and the areas of innovation appear to show many similarities across creoles (e.g. Bakker et al. 2011, 2017; Kortmann and Szmrecsanyi 2009). Structures were lost in a process of pidginization, and new structures had to be introduced to fill grammatical gaps. Baba Malay does show many of the types of innovations found in creoles in general: preverbal particles indicating tense-mood-aspect, an existential derived from the verb ‘to have’, preverbal negation, an emerging determiner system possibly expressing specificity, an analytic plural marker meaning ‘many’. Above, I discussed traits that are assumed to be derived from a pidginized form of Malay, or expansions of it. Is Baba Malay then a creole, or a mixed language of Malay and Chinese, i.e., a so-called intertwined language (Bakker and Mous 1994, Bakker 2017)? In Malay dialectology, Baba Malay belongs to a set of varieties called Vehicular Malay (Paauw 2008), or Pidgin-Derived Malay by Adelaar et al. (1996). These varieties are spoken by populations who presumably shifted to Malay and abandoned their original language. We have seen that most of the characteristics that have been postulated for pidgin-Malay-derived varieties are also found in Baba Malay. However, Baba Malay also has different properties. There are also many traits that can be linked with Hokkien and with Malay. Some have described it as a Malay dialect with heavy borrowing from Hokkien Chinese (Tan 1980). Others have described it as a creole language (Ansaldo and Matthews 1999), i.e., a language that continued the lexicon of one language, here Malay, but the grammatical system was not transmitted from the lexifier or copied from another language.

Others suggest it could be what is now called a mixed or intertwined language (Shellabear 1913). Intertwined languages sometimes develop in hybrid populations, where the grammatical system of the women is combined with the lexicon of the male ancestors (Bakker 1997, 2017). Indeed, the speakers of Baba Malay have Malay and Chinese ancestry. This type of mixture is indeed encountered with other Peranakan groups in Indonesia, including those with Dutch male ancestors (De Gruiter 1990; Dreyfuss and Oka 1979; Van Rheeden 1994; Wolff 1983). In that case, according to the model, the lexicon of Baba Malay should be Chinese (males) and the grammatical system Malay (females), but, if anything, Baba Malay would show the opposite. It is known, however, that Chinese traders quickly shifted to colloquial Malay. Its lexicon, as listed in the glossary in the grammar, is some 85 % Malay; the rest is mostly Chinese or English. All other cases of intertwining involve languages with some level of morphology, and then the type is easy to identify (Bakker 2017, 2020) as affixes from one language are combined with roots from another. However, with an analytic “grammar” language like Hokkien, it is harder to arrive at an “intertwined language” verdict, as affixes are lacking. There are some potential Hokkien grammatical influences in Baba Malay (“including reduplication for the tentative aspect and relexified structures for the perfective, progressive, experiential perfect, and habitual functions originating from Hokkien possessive verb u”; p. 2), but otherwise the grammatical system is much closer to other varieties of Malay than to Hokkien. In short, it is not an obvious Hokkien-Malay intertwined language.

There is a rich literature about Chinese writing in Malay from Indonesia (Hoogervorst 2017; Salmon 1981), but the literary tradition is rarely an inspiration to linguistic researchers (but see Aye 2021 for Baba Malay). If indeed a variety of colloquial Malay was the ancestral language of the immigrant Chinese, Baba Malay could have been an intertwined language with Hokkien, with recent English and Mandarin Chinese influence. In other words, maybe the language was an intertwined language in the past. An early study of the language in fact concluded that “[t]he Baba idiom is Chinese rather than Malay” (Shellabear 1913: 58). Many of the early migrant Chinese males in the Malay-speaking regions had adopted colloquial Malay as their vernacular language at an early stage, and perhaps adjusted it to a Sinitic analytic structural type.

The jury on the “contact language type of Baba Malay” issue is still out.

11 Conclusions

I thoroughly enjoyed reading this grammar, which provides an accessible and much needed grammatical description of this endangered language. The critical remarks about some aspects of the description should not be taken as a negative view of the book. All grammars contain gaps and aspects where users would like more information. As suggested above, typologists and contact language specialists will also find this grammar an interesting read.