Tok Pisin, a dialect of Melanesian Pidgin, is currently the most widely spoken language in Papua New Guinea and also one of three designated national languages. It has an estimated 3 to 5 million speakers. Most of these speak it as a second or auxiliary language, but there is now a considerable population of first language speakers (up to 500,000). Sister dialects are spoken in neighbouring countries in the south-west Pacific, namely Pijin in the Solomon Islands and Bislama in Vanuatu (see Meyerhoff).
The island of New Guinea, although inhabited by a human population for several millennia, was one of the last places to be colonized by European explorers. Impenetrable jungle and daunting topography as well as hostile encounters with inhabitants meant that at the end of the 19th century, large areas remained unexplored. By this time the map showed the island divided into three regions. The western half was part of the Dutch East Indies (now the Indonesian Province of West Papua), while the eastern half was divided by a line running down the central mountain spine. The southern half constituted British New Guinea (later Papua), while the northern half of the island, together with the Admiralty Islands to the north and Bismarck Archipelago islands to the north-east made up German New Guinea. After the First World War, German New Guinea became the UN Trust Territory of New Guinea, while Papua became an external territory of Australia. The two were administered jointly as an Australian colony until independence in 1975. Papua New Guinea is one of the most acutely multi-lingual societies on earth, with over 850 distinct languages identified by SIL International’s Ethnologue.
The precursor of Tok Pisin arose in the context of trade in the Central Pacific in the 19th century. An early Pacific Pidgin English was in use from the early 19th century on trading ships, mainly based in Sydney, seeking whale products, sandal wood, bêche de mer and other commodities. An existing pidgin spoken between settlers and Aboriginal people in New South Wales is likely to have provided the basis for many of the early Pacific features. A distinctly Melanesian variety began to crystallize with the start of the Pacific labour trade in 1863, when Melanesians were recruited (and in some cases kidnapped) to become labourers on plantations in Queensland (Australia). The earliest Queensland labourers were mainly from the New Hebrides (now Vanuatu) and the Solomon Islands. In 1883 and 1884, some also came from the Bismarck Archipelago – islands to the north-east of mainland New Guinea, but many more from this region went to plantations in Samoa from 1879 to 1912. Both the Bismarck Archipelago and Samoa were at that time under German rule. Labourers from the other countries as well started going to Samoa in 1878, and many of these had already worked in Queensland. Thus, early Melanesian Pidgin was transported to Samoa. However, after 1885, labourers from the New Hebrides or Solomons were no longer recruited for Samoa, and early Melanesian Pidgin began to diverge into two slightly different varieties – one spoken in Queensland and one in Samoa.
When the labourers from Samoa were returned to their place of origin in the Bismarck Archipelago, especially New Britain and New Ireland, the Pidgin had an immediate relevance in the intensely multilingual situation, and spread as a lingua franca. The returned labourers also became part of the large internal labour force that worked on the plantations of German New Guinea. At this time, the Pidgin expanded and further stabilized. More words were incorporated from German and from indigenous languages, especially those of the Patpatar-Tolai group, and substrate reinforcement of structures compatible with these languages is likely to have taken place (Siegel 2008). This was the beginning of modern Tok Pisin.
As mobility within German and then Australian New Guinea grew, so did the currency of Tok Pisin. In the first decades of the 20th century, it was increasingly used as a lingua franca by missionaries and patrol officers in the northern part of the New Guinea mainland and in the Islands. Later it became used in newspapers and radio broadcasting. As its use was extended into these new areas, it changed linguistically to become more complex, for example, acquiring more vocabulary and morphology.
Contact with the Highlands of New Guinea did not occur until the 1930s, and the Tok Pisin currently spoken in this region displays a number of distinctive characteristics. The southern half of the mainland, known as Papua, had an alternative lingua franca, Hiri Motu (or “Police Motu”), so the expansion of Tok Pisin into this area was much more restricted until recent years.
After the independence of Papua New Guinea in 1975, Tok Pisin was designated a national language (alongside English and Hiri Motu), and it has adopted an increasingly important role in internal communication. The adoption of Hiri Motu as a national language was largely for political reasons in the face of a Papuan secession movement, but Tok Pisin has made considerable inroads in areas of the Southern region (formerly Papua) at the expense of Hiri Motu. This is especially so in the National Capital District (Port Moresby), where Tok Pisin is the main medium of communication at an informal level. Tok Pisin is also increasing its role in formal situations; for example, parliamentary debates are predominantly in this language. Like many other contact languages, Tok Pisin has received its share of abuse and has been ridiculed as a corrupted or degraded form of English, especially by expatriates in the territory in earlier years, and some of these attitudes may still persist. However, countering this is a pride in the language as one that belongs to the people of Papua New Guinea.
In spite of this, English remains the official language of education. However, a total reform of the education system in Papua New Guinea began in the 1990s. This changed the six years of primary schooling in the medium of English to three years of Elementary School followed by six years of Primary School. The language of instruction and initial literacy in Elementary School is now chosen by the community; English is introduced in the second or third year of Elementary School and becomes the medium of instruction in Primary School. Although exact figures are not available, many communities, especially in urban areas and certain rural areas, such as the Sepik, have chosen Tok Pisin for their schools.
Information is based on Mihalic (1971), Laycock (1985) and Smith (2005a). The basic system has five vowels as indicated in Table 1.
Table 1. Vowels |
|||
front |
central |
back |
|
close |
i |
u |
|
mid |
e |
o |
|
open |
a |
There are 18 consonants in an idealized core inventory (see Table 2). Special orthographic symbols are given in angle brackets.
Table 2. Consonants |
|||||||
bilabial |
labio-dental |
alveolar |
palatal |
velar |
glottal |
||
plosive |
voiceless |
p |
t |
k |
|||
voiced |
b |
d |
g |
||||
nasal |
m |
n |
ŋ <ng> |
||||
trill |
r |
||||||
fricative |
voiceless |
s |
h |
||||
voiced |
v |
||||||
lateral |
l |
||||||
glide |
w |
j <y> |
The outline given below is necessarily an idealization. There is considerable variability, and a great deal of research remains to be done. Variation is mainly from two directions. Firstly, the Tok Pisin of second language speakers is influenced to a greater or lesser degree by the phonology of their first language. For example, intervocalic prenasalization of voiced stops is widespread in many of the languages of the Sepik and Madang areas so far described, and this feature may appear in the Tok Pisin spoken of this region (Laycock 1985). Similarly, voiceless stops may be realized medially as fricatives in many Highlands languages, and this feature may be preserved in the Tok Pisin spoken in the Highlands region (Smith 2005a). In a complex multilingual language situation it is difficult to make generalizations, but there is also widespread variation in the production of /l/-/r/, /s/-/t/, /p/-/f/ and voiced/voiceless stops. Some non-standard realizations may be the source of ridicule as “bush” features. Secondly, those fluent in both Tok Pisin and English may make some use of English phonemes, especially in disambiguating homophones. As Laycock (1985) has observed, Tok Pisin speakers familiar with English have potentially the whole of the English phoneme inventory at their disposal, and forms such as bitʃ ‘beach’ and bidz ‘beads’ could be used in place of the formerly homophonous bis.
Intonation and stress patterns have not been comprehensively studied. There is some indication that stress may be contrastive in pairs of words such as ‘nating ‘nothing’ and na’ting ‘I think, probably’, or ‘palai ‘lizard’ and pa’lai ‘fly’. In the Tok Pisin of first language speakers and fluent second language speakers who use the language as a primary vehicle of communication, considerable reduction of stressed syllables can be observed (Smith 2002). Wurm (1985) described a number of intonation patterns in rural second language speakers in the Eastern Highlands some 50 years ago, while Faraclas (1989) looks at stress reduction in East Sepik, noting that sex differences are apparent in creolized varieties, with females tending towards English stress patterns more than males. The rising intonation of discourse markers, especially nau and pinis is a feature of many narratives.
Some attempts have been made to standardize the orthography of Tok Pisin, but the existence of competing systems from the government and different missions has led to some confusion. The most influential is probably that introduced by Frank Mihalic of the SVD Catholic mission, whose orthography in his pioneering dictionary (1971) is widely used. This is especially so since the style was adopted by the mission’s publishing arm Word Publishing, which produces a popular weekly newspaper Wantok Niuspepa.
Nouns are generally invariable, without any grammatical inflections. There is no grammatical gender, although biological gender can be indicated by adding man ‘male’ or meri ‘female’ after the noun, for example, pikinini man ‘a male child’, pikinini meri, ‘a female child’. Plural is indicated by the particle ol, which is identical in form to the third person plural pronoun, placed before the noun.
(1) man ‘man’
ol man ‘men’
There is no definite article, but the word wanpela ‘one’ may act as an indefinite article. The demonstrative dispela ‘this’ is in common use.
Possession is indicated by the word bilong placed between possessum and possessor:
(2) mama bilong dok ‘the dog’s mother’
puripuri bilong ol Sepik ‘Sepik people’s magic’
It appears that the use of the pluralizing ol, previously somewhat optional in some contexts, is becoming increasingly obligatory in first language speech. The particle ol is also increasingly used with demonstratives with plural reference as dispela ol ‘these’, sometimes severely reduced in form in rapid speech to forms such as slol.
One or two nouns in Tok Pisin include the unanalyzed plural suffix from either English (anis ‘ant’ from English ants) or German (binen ‘bee’ from German Bienen ‘bees’). In addition, the sporadic and irregular use of the English plural suffix -s has been noted for a long time. However, there has been little systematization, and it is only more recently that speakers familiar with English have begun to adopt -s plurals more consistently. Smith’s (2002) study of first language speakers shows increasing use of obligatory marking of plurals with -s among some speakers, often with retention of the redundant ol marker as well. This is especially true with more recently adopted lexis such as bois ‘boys’ and gels ‘girls.’ A few examples of -s pluralization on words of non-English origin may be heard, but they are very uncommon. The nouns most commonly taking the -s suffix in Smith’s corpus were, in order, boi ‘boy’, fren ‘friend’, perent ‘parent’, wik ‘week’, gel ‘girl’ stiudent ‘student’ and ticha ‘teacher’, i.e. mostly recent additions to the lexis, and heard frequently in bilingual educational contexts.
Most adjectives take the adjectival suffix -pela. Such adjectives are placed attributively before the noun they refer to although they can also be used in predicate position:
(3) bikpela haus ‘a big house’
haus i bikpela ‘the house is big’
There is a smaller group of adjectives which do not take the suffix and are placed after the noun:
(4) han kais ‘left hand’
man nogut ‘a bad man’
A notable feature of the pronominal paradigm in Tok Pisin is the existence of inclusive and exclusive forms of the first person plural pronouns, with respect to the person being addressed (see Table 3). Thus inclusive forms indicate ‘we including you’ and exclusive forms ‘we but not you’.
(5) haus bilong mipela ‘house belonging to us
(not including you)’
haus bilong yumi ‘house belonging to us
(including you)’
Table 3. Pronouns |
||
subject/object |
||
1sg |
mi |
|
2sg |
yu |
|
3sg |
em |
|
1du excl |
mitupela |
|
1du incl |
yumitupela |
|
1pl excl |
mipela |
|
1pl incl |
yumi |
|
2du |
yutupela |
|
2pl |
yupela |
|
3pl |
ol |
Also in widespread use are dual forms (mitupela, yumitupela) and occasionally trial forms also (mitripela, yumitripela).
The existence of these structures has often been invoked as an argument for the importance of substrate influence, as direct parallels to this pattern can readily be found in substrate languages such as Kuanua (currently spoken around Rabaul in East New Britain). As with nouns, possession is indicated by the use of bilong before the pronoun, e.g.:
(7) haus bilong mi ‘my house’
wari bilong yu ‘your problem’
em bilong mi ‘It is mine.’
An alternative form of the third person pronoun em is the suffix -en. This is attached only to the preposition long or to the possessive bilong, which become longen and bilongen respectively:
As noted below in §9, longen has more recently come to have a role in relative clause delimitation in Highlands Tok Pisin, probably by analogy with other languages spoken in the area.
The division of sentences into subject and predicate by the particle i (usually referred to as the predicate marker) has received widespread comment, but so far the exact nature and range of its functions has not been adequately determined.
According to Keesing (1988: 133) the particle, derived from the third person singular English pronoun he, adopted new functions according to substrate patterns of speakers in the south-west and central Pacific, who reanalyzed it as a subject-referencing pronoun and then as a marker separating the subject from the predicate. It does appear that substrate patterns are still influential in determining the degree to which i is used in different regions of Papua New Guinea today, as it is very much more frequent in the New Guinea Islands, where there is substrate reinforcement, than in the Highlands, where there generally is not (Smith 2002: 117).
The separation of subject and predicate is illustrated by the following:
However, i is also used in other contexts where the term predicate marker is not appropriate, such as preceding the words go ‘go’ and kam ‘come’ after a verb to indicate direction or sustained action, as shown in the following extract, where i is glossed as PM for convenience:
To illustrate the degree of variability in the use of i, two extended extracts from Smith (2002), first from the New Guinea Islands (see ex. 12) and secondly from the Highlands (see ex. 13), are presented. In (12), i appears in just about every position it could be found in, while example (13) has no cases out of several possible positions, which are marked with [Ø].
(12) Na taim mipla i kam long Utu mipla [Ø] bin [Ø] stap na long moning lo Sarere mipla i bin go bek long taun. [Ø] Go raun long taun na mi lukim olsem taun blong Kevieng i no bikpla tumas we long Madang, we long Madang i bik, taun i bikpla na planti olsem planti man i sae raun long taun na i gat planti ol stua na ol ka i wok long ron long rot.
‘And when we came to Utu we stayed (the night) and in the morning on Saturday we went back to town. Going round the town I saw that Kavieng town is not very big, whereas Madang is big, the town is big and lots of people walk around town and there are lots of stores and cars running on the road.’ [Male, 16, New Ireland]
(13) Wanpela Fraide mi - mi wantem bikpela brata blo mi mipela [Ø] pinisim skul na mipela [Ø] go l’ aus na [Ø] stap nau mipela [Ø] ting osem [Ø] nogat wantla problem ba [Ø] kamap osem na mipela femli olgeta mipela [Ø] go [Ø] stap lo aus, t(a)sol lo nait, tu ouklok wantla man [Ø] kam na [Ø] tok osem “nogat ol [Ø] kilim Malipu Balakau,” na mipela [Ø] fret nau mipela [Ø] pekim olgeta samting, mipela [Ø] go putim l’ aus blo wan(pe)la masta man nau mipela [Ø] go slip lo wantla apatment.
‘One Friday my big brother and I finished school and we went home and at home we didn’t think that any problem would come up, so we just stayed at home, but in the night at 2 o’clock a man came and said they had killed Malipu Balakau and we were afraid and we packed all our things and went and put them in a European’s house and we went to sleep in an apartment.’ (Female, 14, Western Highlands)
One distinctive feature of Tok Pisin is the use of a suffix -im (< English him) on certain verbs to indicate transitivity:
Certain transitive verbs do not take the suffix, notably gat ‘to have’:
Many verbs have a transitive form with -im and an intransitive form without:
The tense-modality-aspect (TMA) markers given in (18) are used in Tok Pisin in the order shown here in the verb phrase after the predicate marker (adapted from Sankoff 1991: 64). Items that are given in the same column cannot co-occur with each other.
ken i kam
A. The future marker bai. This occurs before the predicate marker, and so is somewhat different from the TMA markers which follow. Derived from the English adverbial phrase bye and bye, bai earlier appeared in the form baimbai and was typically sentence initial. Sankoff & Laberge in a seminal paper (1973) noted the grammaticalization of the expression, involving a move closer to the verb, phonological reduction and a loss of stress. There is a correlation between the position of bai relative to different pronouns used with it: First and second person pronouns follow bai in the majority of cases, while third person pronouns most often come before bai.
Although bai is most often described as a future tense marker, it also has an irrealis function.
B. Past tense marker bin. This is derived from the English been and also comes before the verb, but after the predicate marker. It should be noted that many verbs with past time reference are unmarked, and the marker bin might thus appear to be useful only for making time reference clear or for past in past situations. However, its use is frequently redundant:
A. Completed action – pinis. The marker pinis appears to have competed in earlier times as a past tense marker, but is now placed after the verb to indicate completion:
B. laik ‘about to’. The marker laik has a lexical meaning ‘like (to)’ but is also used aspectually before the verb to indicate something that is shortly about to happen:
In some cases is it difficult to disambiguate the aspectual and lexical meanings:
C. Habitual action – save (sa). The word save, from Portuguese saber ‘know’, is well known in pidgins and creoles worldwide and as with other languages, it has a lexical meaning ‘know’ in Tok Pisin. However, the word has also been grammaticalized as an aspect particle indicating habitual action:
The full form is seldom heard by fluent speakers, and the reduced form sa is more usual:
Two other aspectual constructions, wok long before the verb and i stap after the verb have rather similar functions in indicating progressive or continuing action:
Occasionally, both pre-verbal wok long and post-verbal i stap may be used together:
Three modality markers mas ‘must’, ken and inap ‘can’ are in common use. The use of mas can be epistemic as well as implying obligation, as seen respectively in the following two examples:
The functions of ken and inap are respectively permissive and abilitative:
However, inap can also be used in a permissive sense, as the following example shows:
Table 4. Tense-aspect-mood markers |
||
form and position |
function |
etymon |
(i) bin V |
past tense |
been |
bai (i) V |
future/irrealis |
by and by |
V pinis |
completive |
finish |
(i) sa(ve) V |
habitual |
Portuguese saber |
(i) wok lo(ng) V |
progressive |
work along |
V i stap |
progressive |
stop |
(i) mas V |
obligative |
must |
(i) ken V |
permissive |
can |
(i) (i)nap V |
abilitative |
enough |
Tok Pisin has only four simple prepositions: long ‘to, at, in’; bilong ‘of, from’; wantaim ‘with’; olsem ‘like’. There are also complex prepositions with long, such as antap long ‘on top of’ and insait long ‘inside of’.
Word order is SVO. There may be occasional deviation from this order for stylistic effect, but this is rare.
Yes-no questions are distinguished from statements by a rising intonation. In content questions, the three main question words are wanem ‘what’ (< what name), husat ‘who’ (< who’s that) and we ‘where’. Other question words are compounds formed with wanem: wanem taim ‘when’, bilong wanem ‘why’ and olsem wanem ‘how’.
Coordinating conjunctions are na ‘and’, o ‘or’ and tasol ‘but’. Subordinating conjunctions for adverbial clauses include bipo ‘before’, bihain ‘after’, long wanem or bikos ‘because’, long taim ‘when’ and sapos ‘if’.
A considerable amount has been written on relative clause formation in Tok Pisin. A more detailed account appears in Smith (2002), from where the following examples are mainly taken. The simplest relativization structure typical of earlier stages of the language is simple juxtaposition, often in conjunction with distinctive intonation contours:
More often, the relative clause is introduced by a personal pronoun:
More recently the interrogatives husat ‘who’ and we ‘where’ have extended their function to relative pronouns, first noticed in written texts translated from English (Siegel 1985).
Sankoff & Brown (1976) describe a novel method of relative clause delimitation arising from bracketing with the deictic and anaphoric particle ia and cite the following example:
However, it seems that this structure is not as neat and tidy as it may at first appear, and in many cases alternative interpretations of deictic or anaphoric functions of ia are possible, as discussed by Bradshaw (2007).
Finally, especially in the Highlands areas, the expression longen ‘to him’ (occasionally long em) is implicated in the delimitation of relative clauses, although its exact function is not entirely clear:
Although English is the lexifier language of Tok Pisin, many words of English origin have different meanings, for example spak (from spark) ‘drunk’, and puspus (from push) ‘sexual intercourse’. There are also many examples of semantic extension: kilim (from kill) ‘hit’ or ‘beat’ as well as ‘kill’; pisin (from pigeon) ‘bird in general’, and gras (from grass) means not only ‘grass’ but also ‘hair’, ‘fur’ and ‘feathers’. Tok Pisin also has words from other languages. Some examples are given in Table 5.
Table 5. Some Tok Pisin words of non-English origin |
||
origin |
Tok Pisin word |
meaning |
New Guinea Islands languages (such as Tolai) |
lapun kumul muruk palai kiau |
‘old (person)’ ‘bird of paradise’ ‘cassowary’ ‘lizard, gecko’ ‘egg’ |
Samoan |
malolo laplap |
‘rest’ ‘loin cloth, sarong’ |
Malay |
binatang lombo sayor |
‘insect’ ‘chilli’ ‘leafy vegetable’ |
German |
gumi beten raus bros |
‘rubber, tube’ ‘pray’ ‘get out!’ ‘chest’ |
Portuguese |
pikinini save |
‘child’ ‘know’ |