1Macao Creole Portuguese (henceforth MCP) is now largely extinct (Holm 1989, 296; Batalha 1990, 55-56) and a relatively under researched variety.
2The aims of the present paper are twofold: to analyze the types, meanings and functions of reduplication in MCP, and to identify its origins.
3The corpus of MCP includes riddles (Pereira 1900c and 1900f; Barreiros 1943–1944), proverbs (Patoá de Macau e proverbios 2010), poems, short stories and drama scripts (Pereira 1899a, 1899c, 1900g, 1901b; Ferreira 1967, 1994 and 1996), songs (Pereira 1900a and 1901a), letters (Coelho 1881; Pereira 1899b, 1899c, 1900b, 1900d, 1900e, 1901b; Bruning 2007), memoirs (Coelho 2013), and internet discussion lists (Língu Maquista 2009–2012; Como Tá Vai? 2012). The attestations cover the period between the beginning of the 19th century and 2013. Additional data are taken from descriptions of MCP (de Silva 1915; Arana-Ward 1977; Batalha 1990) and from glossaries and dictionaries (Batalha 1977; Fernandes & Baxter 2004).
4The examples include the date of the attestation (if known). The sources are mentioned between brackets. All examples appear in the orthography or system of transcription used in the sources. The length of quotations has been kept to a reasonable minimum. Key portions in the quotations are highlighted in bold. All quotations are accompanied by glosses and by their translation.
5The paper is organized as follows. Section 2 discusses reduplication in MCP. It first looks at partial reduplication, with a focus on the phonological constraints on the bases and on the shape of the reduplicant. Next, total reduplication is examined, in terms of the word classes that can serve as bases and of the associated meanings. Also illustrated are instances of word class change triggered by total reduplication. The section ends with a summary of reduplication in MCP. Section 3 is concerned with the origin of reduplication in MCP. The findings are summarized in section 4.
6In morphological theory, reduplication is a “formal” or “morphological operation” (Haspelmath 2002, 24; Booij 2005, 35-36). While total reduplication is a sub-type of compounding, partial reduplication can be regarded as “a kind of affixation” (Booij 2005, 36). However, unlike in affixation proper, the phonological content of the reduplicative affix depends on the phonological composition of the base (Booij 2005, 36). Following standard assumptions in morphological analysis, the reduplicant is taken to be an abstract affix (red) which triggers the phonological operation of copying, with the copy subsequently attached to the base. As shown below, in MCP the reduplicant (i.e. the copy) always precedes the base to which it is attached, in other words, it is a reduplicative prefix.
- 1 The following abbreviations are used in the glosses: 1= first person; 2= second person; 3= third p (...)
7Partial reduplication is recorded throughout the history of MCP. Consider the following examples1:
|
(1) |
a. |
já |
cavá |
pra |
nho~nhonha |
|
|
|
perf |
finish |
for |
red~woman |
|
|
|
|
|
‘finished for the women’ 1865 (Coelho 1881) |
|
|
|
b. |
nho~nhonha |
bixigósa |
|
|
|
|
red~woman |
chicken.pox.marked |
|
|
|
|
|
‘women with chiken pox marks’ 1880 (Pereira 1899a, 56) |
|
(2) |
a. |
Mas |
olá |
minha |
nhu~nhum |
|
|
|
but |
look |
poss.1sg |
red~man |
|
|
|
|
|
‘But look, my lords’ 1888 (Perreira 1899c, 192) |
|
|
|
|
|
‘men’ (Ferreira 1967, Glossary) |
8The meaning expressed by partial reduplication of nominal bases is that of plurality.
9Ansaldo & Matthews (2004, 11) claim that partial reduplication in MCP is “restricted to high frequency [nouns] such as “men” and “women””. Two remarks are in order here. Firstly, the only nouns which can undergo partial reduplication appear to be nhonha ‘woman’ and nhum ‘man’; no other noun is ever attested in a partially reduplicated form. Secondly, there is one other form derived via partial reduplication, not from a nominal base, but from an adverb, cedo ‘early’, from which cecedo ‘very early’ obtains:
|
(3) |
|
|
ce~ce |
domanhâ |
|
|
|
|
red-early |
morning |
|
|
|
|
|
‘very early in the morning’ 2009 (Língu maquista 2009–2010) |
|
10As can be seen, the partially reduplicated adverb acquires an intensifying meaning.
11Consider next the issue of the phonological content of the reduplicative prefix. According to Jayasuriya (2003, 186), in MCP “[partial] reduplication […] targets […] the first syllable”. This is expressed by Jayasuriya (2003, 186) in the formula reproduced below:
12Ansaldo & Matthews (2004, 10) also state that in “the partial reduplication of nouns […] only the first syllable is reduplicated”. However, the analyses of Jayasuriya (2003) and of Ansaldo & Matthews (2004) are both faulty. The three instances of partially reduplicated forms are repeated below for expository reasons:
|
(5) |
a. |
[ɲo.ɲa] |
→ |
[ɲo.ɲo.ɲa] |
|
|
b. |
[ɲuŋ] |
→ |
[ɲu.ɲuŋ] |
|
|
c. |
[se.du] |
→ |
[se.se.du] |
- 2 In the sense of Haspelmath (2002, 24): in prereduplication the reduplicant precedes the base.
13The relevant example is the form in (5b). As can be seen, the partially reduplicated form [ɲu.ɲuŋ] preserves only the consonant in the onset and the vowel in the nucleus of the monosyllabic base [ɲuŋ]; the consonant in the coda does not surface in the resulting form. Therefore, the phonological shape of the reduplicant is not that of the first syllable of the base, as claimed by Jayasuriya (2003, 186) and by Ansaldo & Matthews (2004, 10), but CV. Partial reduplication is an instance of prereduplication2.
14Summing up, partial reduplication does not appear to have been productive in MCP. It appears to occur only with nominal and adverbial bases. With nouns it marks plurality, while with adverbs it conveys an intensifying meaning. Partial reduplication is subject to constraints on both the base and on the reduplicant: (i) it is limited to consonant-initial disyllabic bases; (ii) the phonological shape of the reduplicant is CV.
15Total reduplication of nominal bases is richly attested throughout the history of MCP:
|
(6) |
a. |
dá |
bença |
pra |
criança |
criança |
|
|
|
give |
blessing |
for |
child |
child |
|
|
|
|
|
‘give blessings to the children’ 1865 (Coelho 1881) |
|
|
|
b. |
nos ôtro |
mulér- |
mulér |
|
|
|
|
1pl |
woman |
woman |
|
|
|
|
|
‘we, women’ 1885 (Pereira 1899a, 59) |
|
|
|
c. |
pędra- |
pędra |
|
|
|
|
stone |
stone |
|
|
|
|
|
‘stones’ 1959 (Thompson 1961, 293) |
|
|
|
d. |
ovo- |
ovo |
pa |
come |
|
|
|
|
egg |
egg |
for |
eat |
|
|
|
|
|
‘eggs to eat’ 2010 (Língu maquista) |
|
|
|
e. |
ilotro- |
sua |
filo- |
filo |
|
|
|
|
3pl |
poss |
son |
son |
|
|
|
|
|
‘their sons’ 2012 (Como Tá Vai) |
16As shown by, among others, de Silva (1915, 163), Arana-Ward (1977, 28, 46), Jayasuriya (2003, 187), and Gaião (2010, 22), total reduplication serves to optionally mark plurality. Given that plural marking is not compulsory, totally reduplicated forms do not normally co-occur with other means of expressing plurality, such as cardinal numerals, other quantifiers, etc. Occasionally, however, the co-occurrence of totally reduplicated nouns and cardinal numerals or other quantifiers is attested, as in the following example:
|
(7) |
a. |
com |
três |
outro |
nhona |
nhona |
|
|
|
with |
three |
other |
woman |
woman |
|
|
|
|
|
‘with three other women’ 2009 (Língu maquista) |
|
|
|
b. |
tudu |
portuguez |
portuguez […] |
fazê |
guéra |
|
|
|
|
all |
Portuguese |
Portuguese |
make |
war |
|
|
|
|
|
‘all the Portuguese were fighting’ 1898 (Bruning 2007) |
|
|
|
c. |
tudo |
laia- |
laiadi |
genti |
|
|
|
|
all |
kind |
kind |
of |
people |
|
|
|
|
|
‘all kinds of people’ (Coelho 2013) |
17Ansaldo and Matthews (2004, 11) write that, rather than plurality, total reduplication expresses the sense of ‘various’. In support of this conclusion they adduce the following example
|
(8) |
|
pintad |
laia- |
laia |
côr |
|
|
|
painted |
kind |
kind |
colour |
|
|
|
|
‘painted in various colours’ (Ferreira 1996, 291) |
18However, closer inspection shows that it is not total reduplication, but the meaning of laia ‘kind, category, type’ (Fernandes & Baxter 2004, 97), which triggers the sense of ‘various’.
19Note, finally, that none of the previous works on MCP mentions the occurrence of distributive reduplication of nouns, which is exemplified below:
|
(9) |
a. |
Espinho |
chuchú |
pê |
/ |
Sangui |
góta |
góta |
|
|
|
thorn |
prickle |
leg |
|
blood |
drop |
drop |
|
|
|
|
|
‘The thorn pricked my leg / Blood flew out drop by drop’ (Pereira 1901a, 706) |
|
|
|
b. |
core |
casa- |
casa |
vendê |
merenda |
|
|
|
|
run |
house |
house |
sell |
food |
|
|
|
|
|
‘[he] ran from house to house to sell food’ (Ferreira 1967, texts, 85) |
20In MCP, then, total reduplication of nominal bases has a pluralizing or a distributive meaning. Pluralizing total reduplication appears to be a still frequent pattern, contra Arana-Ward (1977, 46-47, 68) and Batalha (1990, 57).
21Total reduplication of adjectives is well documented in MCP:
|
(10) |
a. |
pra |
fazê |
palacio |
grande |
grande |
pra |
official |
|
|
|
for |
make |
palace |
big |
big |
for |
official |
|
|
|
|
|
|
‘to build a very big palace for the officials’ 1869 (Coelho 1881) |
|
|
|
|
b. |
gente |
curto- |
curto |
|
|
|
|
|
people |
short |
short |
|
|
|
|
|
|
‘very short people 1899 (Pereira 1900a, 241) |
|
|
|
|
c. |
assi |
vêlo- |
vêlo |
|
|
|
|
|
so |
old |
old |
|
|
|
|
|
|
‘so very old’ (Ferreira 1967, texts, 85) |
|
|
|
|
d. |
cása |
alto- |
alto |
qui |
chôma |
prédio |
|
|
|
|
|
house |
high |
high |
rel |
call |
building |
|
|
|
|
|
|
‘big houses which [they] call buildings’ (Coelho 2013) |
22The meaning conveyed by adjectival reduplication is emphasis (Jayasuriya 2003, 187) or that of intensifying effect (Ansaldo & Matthews 2004, 11), similar to that of a superlative. To this, Ansaldo & Matthews (2004, 11) add a pluralizing meaning, since, on their view, “reduplicated adjectives are also found with plural noun phrases”. In other words, “adjectival reduplication can indicate plurality on the noun it modifies” (Ansaldo 2009: 168). This claim is illustrated with the following two examples ((25) and (26) in Ansaldo & Matthews 2004, 11-12; see also Ansaldo 2009, 168-169):
|
(11) |
a. |
Maior |
parte |
são |
gente |
grande- |
grande. |
|
|
|
greater |
part |
are |
people |
big |
big |
|
|
|
|
|
|
‘It’s mostly important people.’ 1869 (Coelho 1881) |
|
|
|
|
b. |
pa |
tudo |
nôsso |
amigo |
china- |
china |
|
|
|
|
|
for |
all |
our |
friend |
Chinese |
Chinese |
|
|
|
|
|
|
‘for all our Chinese friends’ (Ferreira 1996, 213) |
23However, in (11a) grande-grande actually means ‘very big’, and the translation of (11b) should read ‘for all our friends, the Chinese’.
24To conclude, total reduplication of adjectival bases is productive in MCP. The meaning expressed by adjectival reduplication is that of emphasis, intensification or augmentation.
25According to Ansaldo and Matthews (2004, 12), “reduplication of an adverb” occurs “to serve as a preverbal adverbial”. Indeed, reduplicated adverbs frequently occur in pre-verbal position, throughout the history of MCP:
|
(12) |
a. |
cêdo, |
cêdo |
já |
tem |
na |
rua |
|
|
|
early |
early |
already |
have |
in |
street |
|
|
|
|
|
|
‘[they] went out in the streets very early’ 1865 (Coelho 1881) |
|
|
|
|
|
|
sentado |
omsong- |
omsong |
na |
minha |
tope |
|
|
|
|
|
|
seated |
alone |
alone |
in |
poss.1sg |
room |
|
|
|
|
|
|
‘sitting all alone in my room’ 1899 (Pereira 1901b, 780) |
|
|
|
|
|
|
vagar- |
vagar |
andá |
|
|
|
|
|
|
slowly |
slowly |
walk |
|
|
|
|
|
|
‘walk very slowly’ (Ferreira 1967, Glossary) |
|
|
|
|
|
|
azinha- |
azinha |
sai |
di |
loja |
|
|
|
|
|
|
quickly |
quickly |
go.out |
of |
shop |
|
|
|
|
|
|
‘[he] left the shop very quickly’ 2009 (Língu maquista) |
|
|
|
|
|
|
mánso- |
mánso |
ta |
encurtâ |
unga |
vestido |
|
|
|
|
|
|
quietly |
quietly |
prog |
shorten |
indef |
dress |
|
|
|
|
|
|
‘[she] was very quietly shortening a dress’ 2009 (Língu maquista) |
26However, reduplicated adverbs also occur in post-verbal position, as demonstrated by the following examples, dating from various periods in the history of MCP:
|
(13) |
a. |
olá |
azinha- |
azinha |
pra |
tudo |
este |
ancúsa |
|
|
|
look |
quickly |
quickly |
for |
all |
dem |
thing |
|
|
|
|
|
|
‘take a very quick look at all these things’ 1887 (Pereira 1900b, 261) |
|
|
|
|
b. |
fuzí |
tudo |
azinha- |
azinha |
pra |
Hongkong |
|
|
|
|
|
run |
all |
quickly |
quickly |
for |
Hong Kong |
|
|
|
|
|
|
‘[they] all ran very quickly to Hong Kong’ 1888 (Pereira 1899c, 193) |
|
|
|
|
c. |
falá |
manso- |
manso |
|
|
|
|
|
speak |
quietly |
quietly |
|
|
|
|
|
|
‘speak in a very soft voice’ 1888 (Pereira 1899c, 193) |
|
|
|
|
d. |
andá |
manso- |
manso |
|
|
|
|
|
walk |
quietly |
quietly |
|
|
|
|
|
|
‘go very quietly’ (Pereira 1901a, 704) |
|
|
|
|
e. |
falá |
mánso- |
mánso |
|
|
|
|
|
speak |
quietly |
quietly |
|
|
|
|
|
|
‘to speak in a soft voice’ (Fernandes & Baxter 2004, 69) |
|
|
|
|
f. |
gritá |
forti- |
forti |
|
|
|
|
|
shout |
loudly |
loudly |
|
|
|
|
|
|
‘shout very loudly’ 2009 (Língu maquista) |
|
|
|
|
g. |
tempo |
ta |
vo |
âazinha- |
azinha |
|
|
|
|
|
time |
prog |
fly |
quickly |
quickly |
|
|
|
|
|
|
‘time flies very quickly’ 2009 (Língu maquista) |
27Note, first, that one and the same reduplicated adverb may occur either in pre-verbal or in post-verbal position: compare the position of azinha-azinha in (12d) with (13a), (13b) and (13g), and the position of manso-manso in (12e) with (13c), (13d) and (13e) respectively. Moreover, that reduplicated adverbs are allowed in either position is also demonstrated by the occurrence of intra-speaker variation:
|
(14) |
a. |
cedo- |
cedo |
sai |
di |
casa |
|
|
|
early |
early |
go.out |
of |
house |
- 3 Also in Ansaldo and Matthews (2004, 13), example (32), and in Ansaldo (2009, 169), example (11).
|
|
|
|
|
|
‘leave home very early’ (Ferreira 1996, 313)3 |
|
|
|
|
b. |
vai |
Portugal |
cedo- |
cedo |
|
|
|
|
|
go |
Portugal |
early |
early |
|
|
|
|
|
|
‘leave very early for Portugal’ (Ferreira 1994, 58) |
28In light of the above, it can be concluded that total reduplication of adverbs is productive in MCP. Reduplicated adverbs actually occur both in pre-verbal and post-verbal position, contra Ansaldo and Matthews (2004). Finally, as with adjectives, total reduplication of adverbs expresses an intensifying meaning.
29Jayasuriya (2003, 186) claims that “verb reduplication is not attested in the […] MAC [= Macao Creole Portuguese] sources”. Total reduplication of verbs is not mentioned by Gaião (2007) either. On the other hand, it is illustrated and discussed by Ansaldo and Matthews (2004, 14-15). As shown by the examples below, verbal reduplication is recorded, but less frequently, throughout the history of MCP:
|
(15) |
a. |
capí- |
capí |
aza |
|
|
|
squeeze |
squeeze |
wing |
|
|
|
|
|
|
‘opening and closing its wings’ (Pereira 1901a, 704) |
|
|
|
|
|
|
‘to drip’ (Ferreira 1967, Glossary) |
- 4 Fernandes & Baxter (2004, 145) also list the variant aronça-aronça ‘to wander, to stroll’.
|
|
|
|
|
|
ronça- |
ronça4 |
|
|
|
|
|
|
wander wander |
|
|
|
|
|
|
‘to stroll’ (Ferreira 1967, Glossary) |
|
|
|
|
|
|
torcê- |
torcê |
su |
corpo |
|
|
|
|
|
|
twist twist poss body |
|
|
|
|
|
|
‘twist his body’ 2009 (Língu maquista) |
|
|
|
|
|
|
pê |
mâm |
tremê- |
tremê |
|
|
|
|
|
|
leg |
hand |
tremble |
tremble |
|
|
|
|
|
|
‘[his] legs and hands were trembling’ 2009 (Língu maquista) |
- 5 Example (46) in Ansaldo & Matthews (2004, 15).
30Ansaldo & Matthews (2004, 15) write that verb reduplication is associated with the meaning of “repeated action”, and that “repeated verbs are typically used in verb phrase sequences describing simultaneous actions”. The latter claim, based on a single example5 however, is not corroborated by any other evidence. Furthermore, example (15c) shows that total reduplication of a verb can also express continuity. It appears, then, that the meaning conveyed by total reduplication depends on the aspectual class to which the verb belongs.
31In sum, verb reduplication is less frequent compared to that of nouns, adjectives and adverbs (see also Ansaldo & Matthews 2004, 14). The meanings conveyed by verbal reduplication are repeated action and continuity.
32Total reduplication of numerals is only mentioned by Ansaldo & Matthews (2004, 15-14). It is relatively rare, but examples can be found in texts dating from various periods:
|
(16) |
a. |
pra |
comprá |
unga |
unga |
|
|
|
ancusa |
for |
buy |
one |
one |
thing |
|
|
|
|
|
|
‘to buy things one by one’ 1888 (Pereira 1899b, 125) |
|
|
|
|
b. |
Unga- |
unga |
ta |
falá |
|
|
|
|
|
one |
one |
prog |
speak |
|
|
|
|
|
|
‘[They] are speaking one by one.’ (Ferreira 1996, 72) |
- 6 Example (50) in Ansaldo & Matthews (2004, 16).
33Ansaldo & Matthews’s (2004, 16) conclusions regarding the “preverbal placement of the reduplicated phrase” are contradicted by both (16a) and one of their own examples6, which demonstrate that reduplicated numerals also occur post-verbally:
|
(17) |
|
|
Êle |
largá |
pê |
pulá |
dôs- |
dôs |
degrau. |
|
|
|
|
3sg |
leave |
leg |
jump |
two |
two |
step |
|
|
|
|
|
|
‘He left jumping two steps at a time.’ (Ferreira 1996, 86) |
34As already shown by Ansaldo & Matthews (2004, 15), “reduplicated numerals have a distributive sense”.
35Summarizing, numeral reduplication appears to be less productive. The totally reduplicated form of a numeral may occur either in pre-verbal or in post-verbal position, and is associated with a distributive meaning.
36None of the previous studies of reduplication in MCP mentions the occurrence of total reduplication of quantifiers. This is not attested in any 19th century and early 20th century texts.
37One quantifier which can serve as base for total reduplication is tánto ‘an indeterminate quantity’. Its totally reduplicated counterpart tánto-tánto means ‘so much, so many’ (Fernandes & Baxter 2004, 159). Consider the following examples:
|
(18) |
a. |
têm |
tánto- |
tánto |
ancuza |
|
|
|
have |
much |
much |
thing |
|
|
|
|
|
|
‘have so many things’ (Ferreira 1967, texts, 85) |
|
|
|
|
|
|
tanto- |
tanto |
felicidad |
|
|
|
|
|
|
much |
much |
happiness |
|
|
|
|
|
|
‘so much happiness’ 2009 (Língu maquista) |
|
|
|
|
|
|
já |
olá |
tanto- |
tanto |
lindo |
fotografias |
|
|
|
|
|
|
perf |
look |
much |
much |
nice |
picture |
|
|
|
|
|
|
‘[I] saw so many nice pictures’ 2009 (Língu maquista) |
38Another quantifier which can undergo total reduplication is unchinho ‘very little’. The form obtaining via total reduplication, unchinho-unchinho, means ‘very little, very few’:
|
(19) |
|
|
unchinho- |
unchinho |
ora |
|
|
|
|
very.little |
very.little |
time |
|
|
|
|
|
|
‘from time to time’ (Fernandes & Baxter 2004, 165) |
39Quantifier reduplication is not productive, and, as shown above, it expresses an intensifying meaning.
40Total reduplication may also target interjections, another fact that has gone unnoticed in previous work on reduplication in MCP. Reduplication of interjections is not attested in 19th century and early 20th century texts, and it appears to be confined to just one interjection, amen ‘yes’:
|
|
|
|
|
|
‘yes-man’ (Fernandes & Baxter 2004, 7) |
41Previous work on reduplication in MCP has failed to notice that total reduplication can also be a word class changing operation, i.e. the totally reduplicated form belongs to a different syntactic category than its base. This use of total reduplication is not widespread, but it applies to several types of bases.
42Total reduplication of a nominal base yields an adverb (21) or a verb (22):
|
(21) |
|
|
sópa- |
sópa |
mulado |
|
|
|
|
soup |
soup |
wet |
|
|
|
|
|
|
‘drenched’ (Fernandes &Baxter 2004, 154) |
|
(22) |
|
|
andâ |
coté- |
coté |
|
|
|
|
walk |
sprain |
sprain |
|
|
|
|
|
|
‘to limp, to hobble’ (Fernandes & Baxter 2004, 53) |
43A totally reduplicated adjective turns into an adverb:
|
(23) |
|
|
calado- |
calado |
|
|
|
|
quiet |
quiet |
|
|
|
|
|
|
‘stealthily’ (Ferreira 1967, Glossary) |
44Verbs can also undergo change of word class via total reduplication. The resulting form is a noun, in (24), or an adverb, as in (25a, b):
|
(24) |
|
|
corta- |
corta |
|
|
|
|
cut |
cut |
|
|
|
|
|
|
‘kite competition to try to cut an opponent’s string’ (Fernandes & Baxter 2004, 52) |
|
(25) |
a. |
Tem |
tantu |
dinheiro |
pra |
sentá |
sentá |
vai |
faze |
quatru |
|
|
|
have |
much |
money |
for |
sit |
sit |
go |
make |
four |
|
|
|
|
|
|
dia |
di |
festa? |
|
|
|
|
|
|
day |
of |
holiday |
|
|
|
|
|
|
‘Do you have that much money to have no more, no less a four-day celebration?’ 1898 (Bruning 2007) |
|
|
|
|
b. |
virâ- |
virâ |
vêm |
|
|
|
|
|
turn |
turn |
come |
|
|
|
|
|
|
‘to come in sometime’ (Fernandes & Baxter 2004, 170) |
45Finally, an interjection turns into a noun:
|
|
|
|
|
|
‘yes-man’ (Fernandes & Baxter 2004, 7) |
46Reduplication in MCP has been shown to apply to more bases than hitherto mentioned in the literature (Jayasuriya 2003, Ansaldo & Matthews 2004, Gaião 2007). As seen in 2.1, partial reduplication also applies to adverbs, and total reduplication also targets quantifiers and interjections, as shown in 2.2.6 and 2.2.7. Furthermore, the data analyzed in 2.2 demonstrate that total reduplication is associated with more meanings and functions, such as continuity in the case of verbs, and change of word class with nouns, adjectives, verbs and interjections.
- 7 Jayasuriya (2003) uses the 1978 edition of dos Santos Ferreira (1996).
47In addition to identifying a wider range of bases for reduplication, as well as of meanings and functions associated with it, the analysis has also disconfirmed previous claims (Ansaldo & Matthews 2004) regarding constraints on the placement of totally reduplicated forms of adverbs and of numerals. This has been made possible by the methodology used in the present paper. Previous studies of reduplication in MCP have considered an extremely limited set of data. For instance, all MCP data in Jayasuriya (2003) and in Ansaldo & Matthews (2004) are from a single source (Ferreira 1996)7. José dos Santos Ferreira was one of the last fluent speakers of MCP, and also a writer. In other words, the MCP data analyzed by Jayasuriya (2003) and by Ansaldo & Matthews (2004) reflect the usage of a single speaker, one speech style, and are illustrative of just one historical period, the second half of the 20th century. On the contrary, as specified in section 1, this paper has examined a large body of texts, illustrating a variety of genres and speech styles, produced by different users of MCP, and covering a period of almost two centuries.
48The main findings regarding reduplication in MCP, in light of the data analyzed in 2.1 and 2.2, are set out in the table below:
Table 1
Base |
Type of reduplication |
Output |
Meaning/Function |
Noun |
Partial |
Noun |
Pluralizing |
|
Total |
Noun |
Pluralizing |
|
Total |
Adverb |
Change of word class |
|
Total |
Verb |
Change of word class |
Adjective |
Total |
Adjective |
Intensifying |
|
Total |
Adverb |
Change of word class |
Adverb |
Partial |
Adverb |
Intensifying |
|
Total |
Adverb |
Intensifying |
Verb |
Total |
Verb |
Iterative, continuity |
|
Total |
Noun |
Change of word class |
|
Total |
Adverb |
Change of word class |
Numeral |
Total |
Numeral |
Distributive |
Quantifiers |
Total |
Quantifier |
Intensifying |
Interjections |
Total |
Noun |
Change of word class |
49Summing up, reduplication in MCP produces both semantic and categorial changes. In the former case, reduplication is iconic, i.e. it is functional in nature: it expresses meanings such as plurality in nouns, intensification or augmentation in adjectives and adverbs, iteration or continuity in verbs. In the latter case, reduplication is non-iconic, and, since it triggers change of word class, it is derivational.
50In their attempt at identifying the origin of reduplication in MCP, Ansaldo & Matthews (2004) consider three possible sources: Pidgin Portuguese; the Min (Hokkien) and Cantonese adstrates; the Malay substrate. In what follows the potential relevance of each of these sources is evaluated.
- 8 According to Mühlhäusler (1997, 8-9), creolization can occur with jargons, stabilized pidgins, and (...)
51A variety of Pidgin Portuguese may have been brought from “the Gulf of Guinea colonies” (Ansaldo & Matthews 2004, 17) by Portuguese sailors and African soldiers or as slaves. Indeed, in the second half of the 16th century, Macao’s population did include Africans (Ansaldo & Matthews 2004, 3). However, Pidgin Portuguese cannot be the source of reduplication in MCP. As is well known, Mühlhäusler (1997, 6) identifies the following “qualitatively different stages in the development of a Pidgin”8: jargon; stable pidgin; expanded pidgin; creole. Pidgin Portuguese appears to have been a jargon. In the literature on reduplication in pidgins and creoles it has been shown that the occurrence of productive morphological reduplication correlates with the developmental stage of the variety at issue. Bakker (1995, 33) already noted that reduplication “is rare in pidgins as a productive process”. Bakker & Parkvall (2005, 516) also write that “pidgins are for the most part devoid of reduplication”. With reference to the functions of reduplication, Bakker & Parkvall (2005, 519) state categorically that “reduplication as a grammatical process is virtually absent from pidgins”. Both (Bakker 2003, 44) and Bakker & Parkvall 2005, 514) conclude that reduplication in jargons and stable pidgins is unproductive. Such conclusions are a fortiori true of partial reduplication, which, as seen in 2.1, is attested in MCP.
52According to Ansaldo & Matthews (2004, 2), settlers originally from Fujian, speakers of a dialect of the Min group known as Hokkien, “were allowed to live inside the colony” of Macao and “were therefore in a position to influence the emerging Portuguese creole from the outset”. Consequently, Ansaldo & Matthews (2004, 2-3) argue that “the early presence of the Hokkien community in Macao […] must be considered a potential factor contributing to Sinitic influence”. Later, after the relaxation of the restrictions placed on Chinese presence in Macao, there was a massive influx of Cantonese speakers from the neighbouring areas, gradually leading, among others, to a demographic shift, with the Chinese population outnumbering by far (a 14-to-1 ratio) the Portuguese-origin settlers. In light of the historical evidence, both the Min (Hokkien) and Cantonese dialects are indeed legitimate potential sources of adstratal influence on MCP.
- 9 For a comprehensive grammar of Papía Kristang see Baxter (1988); a brief grammatical sketch of Pap (...)
53Consider finally the Malay substrate. Undoubtedly, MCP has a Malay substrate. This is due to the strong historical relationships between the Portuguese colonies of Macao and Malacca (Holm 1989, 296-297; Ansaldo & Matthews 2004, 4-5). The latter “was by far the most influential colony in Macao” (Ansaldo & Matthews 2004, 4). Moreover, Malay may also have exerted its influence “via trade networks between Macao and the Malay archipelago” (Ansaldo & Matthews 2004, 8). Further evidence is provided by the significant Malay component of the MCP lexicon. Malay is the second most important contributor, after Portuguese (de Azevedo 1984, 44-45; Batalha 1990, 55; Fernandes & Baxter 2004, xxiv): for instance, 194 out of the 1169 lexical items recorded in Batalha (1977), i.e. 16.6%, are of Malay origin. On the other hand, the Malay influence on MCP raises several methodological issues when searching for comparable evidence. Surprisingly, Ansaldo & Matthews (2004) make reference exclusively to Sneddon (1996), a grammar of Standard Indonesian. However, evidence should rather be sought in: colloquial Malay; non-standard varieties of Malay (e.g. Brunei Malay); Malay-lexifier pidgins (e.g. Bazaar Malay, Singapore Indian Malay); Malay-lexifier creoles (e.g. Baba Malay); other Portuguese-lexifier creoles with Malay as their substrate language (e.g. Papía Kristang, Batavia and Tugu Creole Portuguese). In particular, a very likely source of indirect Malay influence is Papía Kristang. As is well known, this creole, heavily influenced by Malay, has a grammatical structure9 extremely similar to that of MCP (see e.g. Batalha 1990, 55; Holm 1989, 291 and 297; Ansaldo & Matthews 2004, 5).
54Ansaldo & Matthews (2004, 10) note that “Indonesian has traces of such reduplication”, illustrated below:
|
|
|
|
a. |
te~tamu |
|
|
|
|
|
red~guest (Ansaldo & Matthews 2004, 10) |
|
|
|
|
b. |
te~tua |
|
|
|
|
|
red~old |
|
|
|
|
|
‘elders’ (Ansaldo & Matthews 2004, 10) |
|
|
|
|
c. |
le~laki |
|
|
|
|
|
red~man |
|
|
|
|
|
‘man, male’ (Ansaldo & Matthews 2004, 11) |
55However, this type of partial reduplication differs from the one attested in MCP, both formally and in terms of the meaning expressed. In Indonesian, the portion attached to the base consists of the first consonant of the base and the vowel /ə/ (Dyen 1967, 33; Macdonald & Soenjono 1967, 53; Sneddon 1996, 21). Therefore, the reduplicant is a duplifix (in the sense of Haspelmath 2002, 24), i.e. an element made up of a copied segment – the first consonant of the base – and a fixed segment – the vowel /ə/. Its phonological shape, /Cə/ thus differs from /CV/, the form attested in MCP. Semantically, Indonesian partial reduplication mostly yields the same meaning or a meaning closely related to that of the base (see e.g. Sneddon 1996, 21), whereas in MCP it has a pluralizing meaning (with nouns) or an intensifying meaning (with adverbs).
56On the other hand Papía Kristang and MCP exhibit formal similarities. Papía Kristang also has /CV/ reduplicants, cf. examples (1)-(2):
|
|
|
|
so~soti |
|
|
|
|
red~kind |
|
|
|
|
‘kinds’ (Jayasuriya 2003, 188) |
57Similarities between Papía Kristang and MCP extend to the occurrence of identical forms; compare (3) and the example below:
|
|
|
|
se~sedu |
|
|
|
|
red~early |
|
|
|
|
‘very early’ (Jayasuriya 2003, 189) |
58Also, as in MCP, partial reduplication expresses a pluralizing or an intensifying meaning. However, Papía Kristang partial reduplication differs in several respects: it also yields /C(C)VC/ reduplicants (30a-b); it also applies to vowel-initial bases, yielding VCVC reduplicants (30c); it also applies to bases consisting of more than two syllables (30d):
|
|
|
|
a. |
fem~femi |
|
|
|
|
|
red~woman |
|
|
|
|
|
‘women’ (Jayasuriya 2003, 188) |
|
|
|
|
b. |
kren~krensa |
|
|
|
|
|
red~child |
|
|
|
|
|
‘children’ (Jayasuriya 2003, 187) |
|
|
|
|
c. |
idad~idadi |
|
|
|
|
|
red~age |
|
|
|
|
|
‘elderly (people)’ (Jayasuriya 2003, 188) |
|
|
|
|
d. |
fam~familia |
|
|
|
|
|
red~descendant |
|
|
|
|
|
‘descendants’ (Jayasuriya 2003, 187) |
59To conclude, partial reduplication in MCP resembles to some extent the one attested in Papía Kristang, but is comparatively more restrictive and less frequent.
60Total reduplication of nouns is certainly of Malay origin, but not for the reasons mentioned by Ansaldo & Matthews (2004, 11), who write that “when [reduplication] is used the plurality typically has the sense of ‘various’”, and who conclude that “the semantics of reduplicated plurals support the assumption that Malay influence has played a role”.
61Several remarks are in order here. The sense of ‘various’ rather than that of ‘plural’ is specific rather to Indonesian. Moreover, even in Indonesian, “there is some question as to whether the reduplicated form [of nouns] denotes simple plurality or whether it denotes both plurality and variety” (Macdonald & Soenjono 1967, 55). Also, as noted by Sneddon (1996, 17), “it is sometimes stated that reduplication of nouns indicates variety rather than plurality”, but “Indonesian writers disagree on this question”. As shown in the discussion of example (8) in 2.2.1, in MCP total reduplication of nouns has a pluralizing meaning only. This accords with evidence from varieties of Malay as well as from other Portuguese-lexifier creoles with a Malay substrate. According to Othman & Atmosumarto (1995, 39) in colloquial Malay reduplication of nouns expresses plurality. Similarly, in Bazaar Malay it “indicates indefinite plurality” (Aye 2005, 97). Also, in Singapore Indian Malay nouns are “reduplicated to mark plurality” (Rekha 2007, 54). Moreover, Bazaar Malay (31) and Old Baba Malay (32) share with MCP, cf. example (7a), the characteristic of allowing the co-occurrence of reduplicated nouns and cardinal numerals.
|
|
|
|
Ada |
anak- |
anak |
perempuan |
empat. |
|
|
|
|
be |
chid |
child |
female |
four |
|
|
|
|
‘[I] have four daughters.’ (Aye 2005, 98) |
|
|
|
|
tujoh |
dewa |
dewa |
ada |
menari |
|
|
|
|
seven |
deity |
deity |
be |
dance |
|
|
|
|
‘seven deities were dancing’ (end of 19th c.) (Thurgood 1998, 114 |
62In Papía Kristang reduplication of nouns is a pluralizing device (Baxter 1988, 102-103), as “the plural is generally formed by repeating the word” (Anon. n.d.), and in Batavia Creole Portuguese “the plural may be expressed by reduplication of the noun” (Maurer 2011, 22), as shown in (33) and (34) respectively:
|
|
|
|
botang- |
botang |
|
|
|
|
button |
button |
|
|
|
|
‘buttons’ |
|
|
(34) |
Batavia Creole Portuguese |
|
|
|
|
inchidu |
kung |
fula |
fula |
|
|
|
|
filled |
with |
flower |
flower |
|
|
|
|
‘in full bloom’ 1780 (Maurer 2011, 133) |
63Finally, just like MCP, Tugu Creole Portuguese allows the co-occurrence of cardinal numerals or of the quantifier tudu with a reduplicated noun, cf. (7a-b) :
|
(35) |
Tugu Creole Portuguese |
|
|
|
|
a. |
Amiang |
pasa |
dosong |
omi |
omi. |
|
|
|
|
|
tomorrow |
pass |
two |
man |
man |
|
|
|
|
|
|
‘Tomorrow two people will leave.’ 1937 (Maurer 2011, 191) |
|
|
|
|
b. |
sebab |
eli |
ola |
tudu |
sorti |
sorti |
|
|
|
|
|
because |
3sg |
see |
all |
sort |
sort |
|
|
|
|
|
|
‘because he sees everything’ 1891 (Maurer 2011, 302) |
64Ansaldo & Matthews (2004, 11) write that, “since adjective reduplication with intensifying effect is very widespread”, the fact that this is also found in MCP “does not reveal its origins”. Moreover, in support of their claim that in MCP reduplicated adjectives are an alternative to reduplicating nouns to express plurality, which would point to a Malay source, Ansaldo & Matthews (2004, 12) quote Sneddon (1996, 12), according to whom “reduplication of an adjective usually occurs when the nouns it describes is plural” and it shows that “the characteristic indicated by the adjective applies to all objects”. Again, this applies to standard Indonesian, in which, as shown by Macdonald & Soenjono 1967, 57), “this device is not commonly met with” anyway.
65On the other hand, reduplication of adjective typically has an intensifying meaning in Malay, as in MCP. Consider first descriptions of earlier Malay. Marsden (1812, 39) writes that “a kind of superlative is produced by doubling the adjective”. Similarly, Favre (1876, 100) states that “a sense of superlative is given to the adjective by doubling it”. According to Maxwell (1907), “the adjective may […] be intensified by duplication”. The same point is made by Winstedt (1913, 59), who mentions the fact that “the superlative absolute […] may be expressed by […] reduplication”. Consider the following examples:
|
|
|
|
a. |
burong |
kechil- |
kechil |
|
|
|
|
|
bird |
small |
small |
|
|
|
|
|
|
‘very small bird’ (Maxwell 1907, 41) |
|
|
|
|
b. |
jimat- |
jimat |
di- |
nĕgĕri |
asing |
|
|
|
|
|
careful |
careful |
in |
country |
foreign |
|
|
|
|
|
|
‘be very careful in foreign countries’ (Winstedt 1913, 59) |
66In modern (colloquial) Malay, “reduplication of an adjective serves to intensify its meaning”, whereas this “is not common in Indonesian” (Macdonald & Soenjono 1967, 57). Similar views are expressed by Othman & Atmosumarto 1995, 70), who state that “doubling […] stresses the adjective in the superlative form”, and by Sulaiman (2009, 266), who writes that “an adjective is reduplicated to indicate emphasis or intensity”. Reproduced below is an example illustrative of this usage in modern Malay:
|
|
|
|
Semua |
anak |
kucing |
itu |
hitam- |
hitam. |
|
|
|
|
all |
child |
cat |
dem |
black |
black |
|
|
|
|
‘All the kittens are very black.’ (Sulaiman 2009, 266) |
67Consider also the following example from Brunei Malay, a non-standard variety:
|
|
|
|
par- |
anak- |
ku |
ani |
gauk- |
gauk. |
|
|
|
|
person |
child |
1sg |
dem |
naughty |
naughty |
|
|
|
|
‘My children are very naughty.’ (Clynes 2001, 24) |
68The same holds for Bazaar Malay and Singapore Indian Malay. In the former, reduplication “intensifies the meaning of adjectives” (Aye 2005, 100), and the latter resorts to reduplication of the adjective “to intensify or emphasize its meaning” (Rekha 2007, 58), as exemplified in (39) and respectively (40):
|
|
|
|
kayuh |
semua |
besar- |
besar |
|
|
|
|
wood |
all |
big |
big |
|
|
|
|
‘All the woods are very big.’ (Aye 2005, 100) |
|
(40) |
Singapore Indian Malay |
|
|
|
|
besa- |
besa |
femli |
|
|
|
|
big |
big |
family |
|
|
|
|
‘very big family’ (Rekha 2007, 58) |
69Reduplication of adjectives with an intensifying effect is also reported for Papía Kristang and Tugu Creole Portuguese. According to Baxter (1988), adjectives are reduplicated for intensification; Jayasuriya (2003, 187) also writes that reduplication expresses “emphasis on or an augmented degree of the quality”. Similarly, Maurer (2011, 111) states that in Tugu Creole Portuguese reduplication of adjectives is used “for intensification”. This is shown in (41) and respectively (42):
|
|
|
|
Eli |
belu- |
belu |
ta |
bai |
mar. |
|
|
|
|
3sg |
old |
old |
prog |
go |
sea |
|
|
|
|
‘Although he was very old, he was [still] going fishing.’ (Jayasuriya 2003, 187) |
|
(42) |
Tugu Creole Portuguese |
|
|
|
|
dretu |
dretu |
korsang |
|
|
|
|
right |
right |
heart |
|
|
|
|
‘sincere heart’ 1885 (Maurer 2011, 170) |
70To sum up, MCP adjective reduplication with an intensifying meaning can be traced back to Malay, of which it is also a typical feature.
71According to Ansaldo & Matthews (2004, 13), “adverbials (reduplicated or not) are postverbal in […] Malay”, and cannot, therefore, account for their pre-verbal placement in MCP. Since reduplicated adverbs in pre-verbal position are found both in Cantonese and, more productively, in Min dialects, Ansaldo & Matthews (2004, 13) go as far as to claim that this is “the clearest evidence for Sinitic influence in Macanese reduplication”. An identical point is made by Ansaldo (2009, 169-170). However, the premises on which this conclusion is based are not valid. As seen in 2.2.3, both pre-verbal and post-verbal reduplicated adverbs are recorded throughout the history of MCP. Furthermore, while adverbs frequently occur in post-verbal position in Malay, their pre-verbal placement is also well documented. Note, incidentally, that this also applies to Indonesian. The following examples from Sneddon (1996), the grammar of Standard Indonesian to which Ansaldo & Matthews (2004) constantly refer, demonstrate that both reduplicated (43a-b) and non-reduplicated (43c) adverbs can also occur pre-verbally:
|
|
|
|
a. |
Penyakit |
itu |
betul- |
betul |
dapat |
di- |
sembuhkan. |
|
|
|
|
|
disease |
dem |
real |
real |
can |
pass |
cure |
|
|
|
|
|
|
‘This disease can really be cured.’ (Sneddon 1996, 210) |
|
|
|
|
b. |
Dia |
diam- |
diam |
meninggalkan |
istana. |
|
|
|
|
|
3sg |
quiet |
quiet |
leave |
palace |
|
|
|
|
|
|
‘He quietly left the palace.’ (Sneddon 1996, 210) |
|
|
|
|
c. |
Dengan |
cepat |
Micheal |
lari |
ke toko. |
|
|
|
|
|
with |
fast |
Michael |
run |
to |
shop |
|
|
|
|
|
|
‘Quickly, Michael ran to the shop.’ (Sneddon 1996, 256) |
72The same also holds for Brunei Malay and Bazaar Malay, as illustrated in (44) and (45) respectively:
|
|
|
|
Saudagar |
ani |
batah- |
batah |
jatuh |
sakit. |
|
|
|
|
merchant |
dem |
long.time |
long.time |
fall |
sick |
|
|
|
|
‘Eventually, the merchant fell sick.’ (Clynes 2001, 27) |
|
|
|
|
plan- |
plan |
jalan |
balek |
|
|
|
|
slowly |
slowly |
walk |
return |
|
|
|
|
‘[they] very slowly walked back home’ (Aye 2005, 100) |
73A Malay origin, then, is compatible with both the pre- and the post-verbal placement of reduplicated adverbs in MCP. On the contrary, Min and/or Cantonese influence would only account for the preverbal reduplicated adverbs of MCP.
74Ansaldo & Matthews (2004, 15) identify “close parallels” between MCP and Malay and Cantonese. However, the case for Cantonese rests on the claim that in MCP reduplicated verbs are typically associated with simultaneous actions, which, as seen in 2.2.4, cannot be substantiated. On the other hand, the meanings expressed by reduplicated verbs in MCP, i.e. repeated action or continuity, closely match those reported for Malay, e.g. “repetition, frequency and continuity in action” (Winstedt 1913, 102):
|
|
|
|
‘blowing continually’ (Winstedt 1913: 102) |
75Similar meanings are identified in the literature on Malay-lexifier pidgins. In Bazaar Malay, reduplication of verbs “adds the iterative or durative aspect to the meaning” (Aye 2005, 101), as in (47); in Singapore Indian Malay, it is used “when an action is repeated” (Rekha 2007, 57), as in (48).
|
|
|
|
ada |
orang |
jahat |
lawan- |
lawan |
|
|
|
|
be |
person |
bad |
attack |
attack |
|
|
|
|
‘there were bad people fighting all the time’ (Aye 2005, 101) |
|
(48) |
Singapore Indian Malay |
|
|
|
|
ini |
jantan |
biseng- |
biseng |
|
|
|
|
dem |
man |
scold |
scold |
|
|
|
|
‘the man kept scolding’ (Rekha 2007: 58) |
76Consider also Papía Kristang and Batavia and Tugu Creole Portuguese. In the former, verbs are reduplicated for iteration (Baxter 1988); reduplication of verbs conveys the meaning of “continuous or discontinuous repetition” (Jayasuriya 2003, 186), see (49); and in the latter it appears to express repetition or continuity, as in (50a-b).
|
|
|
|
|
|
yo |
lembrá |
lembrá |
|
|
|
|
|
|
1sg |
remember |
remember |
|
|
|
|
|
|
‘as far as I can recall’ (Jayasuriya 2003, 186) |
|
(50) |
Batavia Creole Portuguese |
|
|
|
|
a. |
bringka |
bringka |
olu |
|
|
|
|
|
play |
play |
eye |
|
|
|
|
|
|
‘twinkle with the eye’ 1780 (Maurer 2011, 220) |
|
|
|
|
b. |
bira bira korpu turn turn body |
|
|
|
|
|
|
‘lounge, stroll about’ 1780 (Maurer 2011, 279) |
77Therefore, in the case of verbal reduplication too, Malay is the plausible source for its meanings in MCP.
78As with reduplicated adverbs, Ansaldo & Matthews (2004, 16) conclude that in the case of numerals too “the preverbal placement of the reduplicated phrase follows the Chinese pattern”. Similarly, Ansaldo (2009, 170) writes that “the pre-verbal position reveals more direct Sinitic influence”. However, such claims run into difficulties. Recall from 2.2.5 that in MCP reduplicated numerals are not restricted to the pre-verbal position, but also occur post-verbally. Obviously, the post-verbal placement cannot be traced back to a Chinese pattern, but it accords, instead, with that in Malay. Furthermore, in Malay (and Indonesian) reduplicated numerals are not found only in post-verbal position, as suggested by Ansaldo & Matthews (2004, 16) and by Ansaldo (2009, 170). As shown below, reduplicated numerals with a distributive meaning can also be placed pre-verbally:
|
|
|
|
a. |
satu |
satu |
tunas |
muda |
bersemi |
|
|
|
|
|
one |
one |
shoot |
young |
sprout |
|
|
|
|
|
‘one by one, the new shoots sprout’ (Iwan Fals 1994) |
|
|
|
|
b. |
satu |
satu |
daun |
jatuh |
ke |
bumi |
|
|
|
|
|
one |
one |
leaf |
fall |
to |
earth |
|
|
|
|
|
‘one by one, leaves fall on the ground’ (Iwan Fals 1994) |
79It follows that the pre- and post-verbal position of reduplicated numerals with a distributive meaning in both MCP and Malay points to the latter as the likely source.
80Reduplication in MCP applies to more bases and has more meanings and/or functions than stated in previous work.
81There appears to be no evidence for Pidgin Portuguese, Cantonese or Min (Hokkien) influence on the reduplication patterns attested in MCP or on the meanings expressed by reduplicated forms.
82Reduplication in MCP, possibly via Papía Kristang, is ultimately of Malay origin. All types of reduplication attested in MCP are found in colloquial Malay, non-standard varieties of Malay, Malay-lexifier pidgins and creoles or Portuguese-lexifier creoles with a Malay substrate as well. Also, MCP reduplicated forms occur in the same position and express identical or similar meanings as in these varieties.