Body part tally counting and Bible translation in Papua New Guinea and Irian Jaya.

Pre-publication version, do not quote; for quotable version see: http://tbt.sagepub.com/content/49/4.toc

1998c Body part tally counting and Bible translation in Papua New Guinea and Irian Jaya. The Bible Translator (Practical Papers), vol. 49, no.4, 409-415.

LOURENS DE VRIES

BODYPART TALLY COUNTING AND BIBLE TRANSLATION IN NEW GUINEA

The author is a translation advisor with the United Bible Societies for projects in Indonesia, holds the Chair of Bible translation at the Free University, Amsterdam, the Netherlands and is a senior researcher in the Leiden University Irian Jaya Studies project.

Introduction

The incredible linguistic diversity of New Guinea is also reflected in the rich variety of systems of numerals in New Guinea languages (Laycock 1975, De Vries 1994). Since New Guinea languages easily borrow numeral systems, from national languages, local languages and trade languages, it is not uncommon to find two or three numeral systems coexisting in one language. Bruce (1984: 102) for example writes about Alamblak of Papua New Guinea: 'Three numeral systems coexist in Alamblak: a borrowed tally system, the borrowed New Guinea Pidgin system, and a mixed binary/quinary/vigesimal system (primitive numbers of one, two, five, and twenty)'.

In this paper I concentrate on one type of numeral systems of New Guinea, the bodypart tally system. These bodypart tally systems represent one the most common types of traditional counting systems of New Guinea. At the same time, these systems create the most problems for Bible translators since the highest number in these systems very often is not higher than 29. Using borrowed numerals from national languages such as Indonesian or Tok Pisin, creates its own problems, for example the problem of how to combine the borrowed numerals and the bodypart numerals. First, I describe bodypart tally systems from the perspective of anthropological linguistics. Then I mention some of the ways Bible translations of New Guinea handle bodypart tally numerals.

Bodypart tally systems

In bodypart tally systems, counting starts on the little finger of the left hand (number 1) which the speaker touches or bends with the middle finger and/or index finger of the right hand. The nouns denoting these bodyparts also function as numerals. Via ringfinger (2), middle finger (3) and indexfinger (4) the speaker reaches the thumb (5) after which he or she[1] climbs the left arm, usually starting with the wrist (6), then lower arm (7), elbow (the inside is touched, (8)), upper arm (9) and shoulder (10). After the points on the hand and the points on the arm, points on the head (for example neck, ear, eye, nose and crown of the head) are used. There is much variation in the points used by bodypart tally systems. For example, Kombai uses just two points on the head and proceeds after the shoulder (10) to the left ear (11) and reaches the turning point in the crown of the head (12) after which the ear on the other side (13) is touched, followed by the shoul­der of the other side (14) and so forth until the little finger of the right hand is reached, the highest number of Kombai (23) (De Vries 1993). Korowai uses three points on the head (neck (11); ear (12); crown of the head (13)) and accordingly ends up with 25 when the little finger of the right hand is reached (Van Enk & De Vries 1997). Wambon uses four points on the head: neck (11), ear (12), eye (13) and nose (14) and ends up in 27 (De Vries and Wiersma 1992)[2]. Kombai, Korowai and Wambon are all of the same Awyu family spoken in south-east Irian Jaya. The counting systems of the Awyu family reflect the diversity of New Guinea: within this family one finds three types of traditional counting systems; first, Wambon, Kombai and Korowai (pure bodypart tally systems), second, Mandobo (a mixed bodypart tally system), third, Pisa, Sjiagha-Yenimu and Aghu (counting on hands and feet[3]).

When after the turning point counting continues on the right side of the body, Wambon, Kombai and Korowai all use a prefix meaning 'other side' like very many other New Guinea languages. Although these languages know the distinction between 'left' and 'right', they do not use it in the counting system. The Kombai words for left (wa) and right (uu) are usually found modifying the noun i 'hand' (i uu 'right hand' and i wa 'left hand'). The expressions for 'left hand' and 'right hand' may be used as locative expressions ('he sat at my right hand'). In the counting system the distinction be­tween 'right' and 'left' is in fact not relevant: these (symmetrical) systems work equally well if one starts on the right side but since most people are righthanded, the right hand is the active hand in the bending and touching gestures.

Bodypart tally systems tend to use only bodyparts of the upper half of the human body but there are interesting exceptions. The Yupna men of the Finisterre mountains of Papua New Guinea included navel (30), left testicle (31), right testicle (32) and penis (33) in their counting system (Wassmann 1995)[4].

A typical bodypart tally system: the Korowai of Irian Jaya

To illlustrate this general outline of bodypart tally systems, I will discuss the Korowai system in some detail. Korowai is spoken in the area between the upper Becking and Eilanden rivers of south-east Irian Jaya, Indonesia. Since 1980 some Korowai clans had contacts with Dutch missionaries of the Mission of the Reformed Churches but until today most Korowai still live isolatedly in their swampy rainforests. The Korowai number system is a pure bodypart system used in conjunction with a very rudimentary binary system comprising two non-bodypart based numerals for 'one' and 'two'. In the early 1980s, when I studied the Korowai systems as it functioned in their daily life, the Korowai did not know any other system of counting. First, I present the Korowai bodypart system in (1):

(1) The Korowai tally system

Korowai numeral corresponding bodypart

1 senan little finger

2 senanafül ring finger

3 pinggu(lu)p middle finger

4 wayafül index finger

5 wayo thumb

6 gédun wrist

7 lafol lower arm

8 bonggup elbow

9 labul upper arm

10 main shoulder

11 khomofekholol neck

12 khotokhal ear

13 khabéan head

When khabéan 'head'/'thirteen' is reached, counting on the righthand side of the body begins, with mén 'other side':

14 mén-khotokhal ear on the other side

15 mén-khomofekholol neck on the other side

16 mé-main the other shoulder

17 mén-tabul the other upper arm

18 mé-mbonggup the other elbow

19 mén-tafol the other lower arm

20 mé-nggédun the other wrist

21 mé-wayo the other thumb

22 mé-wayafül the other index finger

23 mén-pinggu(lu)p the other middle finger

24 mén-senanafül the other ring finger

25 mén-senan the other little finger

When these numerals are used as modifying elements in noun phrases, the noun anop 'amount' cliticizes obligatorily to the numerators signifying that the numerators are used attributively (ATTR):

(2) pinggup

middle.finger

'three'

gol pinggu-anop

pig middle.finger-ATTR

'three pigs'

gol wayafül-anop

pig index.finger-ATTR

'four pigs'

The element anop cannot be added to senan 'little finger/one' and senanafül 'ring finger/two' to form modifying numerals:

(3) *gol senan-anop

pig one-ATTR

'one pig'

Instead, lidop 'one' and pol 'two' are used as modifiers in noun phrases:

(4) gol lidop

pig one

'one pig'

gol pol

pig two

'two pigs'

Lidop and pol are the only two non-bodypart derived numerals of Korowai. The word pol 'a pair/two' is used in combination with a word meaning 'just' to denote the concept of 'few':

(5) gol pol-tanukh

pig pair-just

'(a) few pigs'

There are metaphorical idioms derived from lidop 'one' and pol 'two' by reduplication. Compare:

(6) mahüon lido-lidop

story one-one

'a straightforward story'

(7) ye fimelon po-po-telo

he intestines two-two-COP.3SG.NF

'He is in doubt.'

The bodypart tally systems which I studied, those of isolated Awyu societies like the Korowai in south-east Irian Jaya, function especially in two kinds of contexts. In the first place, in the context of institutionalized payments in shell-money and goods (bridal payments, compensation payments), in the second place in the context of sagogrub- and pigfestivals that also have a market function (cf. Venema 1990, Schoorl 1957, Boelaars 1970).

For the Mandobo, an Awyu group living east of the Upper-Digul close to the border with Papua New Guinea, Boelaars (1970: 185) gives a pricelist, in Mandobo tagèt[5] (little kauri-shells), for pig-related transactions during a pigfestival:

(8)

it=right foreleg, 2 tagèt

it=left foreleg, 2 tagèt

ngébén tirambon=neck+head+upper half of back, 2 tagèt

kikemambuman=lower half of back, 4 tagèt

kondok=left hindleg, 2 tagèt

kondok=right hindleg, 2 tagèt

wobogan=neck+breastpiece, 1 tagèt

mènggop=belly, 2 tagèt

omunop=intestines, 2 tagèt

one whole pig=19 tagèt

Of course, during the festivals anything of value can be purchased by barter or paying kauri-shells. Since persons of different Awyu tribes meet each other during the festivals, magic formulas of other tribes in the area can often be bought during the festivals. The prices for stone axes, magic formulas, nose- and eardecorations, birds of paradise etcetera can easily be handled within the limits of the bodypart counting systems.

The bridal payments and compensation payments for deaths are a topic for heated and prolonged negotiations between representatives of patriclans or patrilineages. Boelaars (1970: 157) gives the following list of items that constituted the original Mandobo bridal payment (amot or rankandün):

(9)

30 tagèt (shells), 5 or 6 mbotnon (string of little beads)

2 tenggat (nose decoration of stone)

2 warok (nose decoration made of birdsclaw)

1 janop (part of a white shell that is tied to the string of kauri-shells. It is a symbol of the female genitals and is demanded upon first payment.)

1 mbandit (bow)

1 enop (big netbag)

1 jang (small netbag)

5 kurèn (stone axes)

1 irik (string of dogteeth)

1 aomban (string of pigteeth)

Counting of kauri-shells, pigteeth, dogteeth and valued objects such as stone axes in these contexts of festivals and bridal/compensation-payments takes place with the objects to be counted positioned in between the parties involved in the exchange. The objects are picked up one by one and carefully inspected. Counting is matching of these visible objects with the bodyparts used as tallies.

Two points emerge from considering the place of counting in Awyu societies; first, that the shell-money prices of goods and services rarely exceeds 30 shells. According to my Wambon, Kombai and Korowai informants bridal payments (Mandobo, 24 to 30 shells) and pigs (Mandobo, 19 shells) are the two most expensive things in a man's life. Thus the highest numbers in the bodypart systems of the Awyu area are easily high enough to satisfy the basic counting needs of these societies. The second point is that counting in the great majority of cases involves concrete visible objects that are touched and matched with the bodyparts that are also touched.

The conventional touching gestures are an integral, obligatory part of the use of bodypart nouns as numerals. Since Bible translations are a written genre, the gestures which always accompany the use of bodypart numerals in spoken genres, are absent. The gestures are important for bodypart tally systems since they disambiguate between the secondary, numeral meaning and the primary meaning of bodypart words. If speakers refer to their wrist, they will never use the conventional touching gesture accompanying the use of 'wrist' in its meaning 'six'. Of course, in written genres the context will usually give some help in finding out whether the bodypart words are used as numerals or not. Also the attributive suffixes which in many Papuan languages go with the use of bodypart numerals as modifiers in noun phrases, help to identify the numeral use.

When some of the isolated Kombai and Korowai clans in the north came into regular contact with the mission in the 1980, their counting system came to be used in wholly new contexts, for example to denote the seven days of the week and the twelve months of the year, concepts introduced by missionaries and government personnel. Kombai, Korowai and Wambon used the same strategy to adapt the counting system to the new needs. They formed compound nouns with the noun for 'sun;day' as the modified noun and the bodypart/number nouns from 1 to 6 as modifying nouns. Thus Monday= little finger-day, Tuesday=ring finger-day, etcetera. For Sunday, the languages take the Indonesian word for worship sembahyang, adapt it phonologically and use it in a compound noun with the word for 'day' (for example Kombai rembayana-rei 'Sunday').

Consider the Kombai system for the days of the week:

(10) The Kombai names for the days of the week

raga‑n‑e-rei

little.finger‑TR‑CONN-day

'Monday'

raga‑ragu‑n‑e-rei

ringfinger‑TR‑CONN-day

'Tuesday'

woro‑ragu‑n‑e-rei

middle.finger‑TR‑CONN-day

'Wednesday'

woro‑n‑e-rei

index‑TR‑CONN-day

'Thursday'

abalof‑e-rei

thumb‑CONN-day

'Friday'

go‑n‑e-rei

wrist‑TR‑CONN-day

'Saturday'

rembaya-n-a-rei

worship-TR-CONN-day

'Sunday'

In the following Wambon examples (11) the bodypart numerals are used in expressions for the months of the year:

(11) sanov-o-wakhot

little.finger-CONN-moon

'January'

sanopkuniv-o-wakhot

ring.finger-CONN-moon

'February'

Interestingly, when bodypart numerals are used in these expressions for the days of the week or the month of the year, the touching gestures always accompany the use of those expressions. Thus whenever gonerei 'Saturday' is used, the speaker touches his left wrist.

After the highest point (one time round from little finger of left hand to little finger of right hand), some languages seem to extend the system by adding a word like 'again' or 'bring forth'. For example, some Korowai informants claimed that, when twentyfive is reached, the highest point after one round, one can continue by prefixing laifu- 'produce' to the bodypart/numbers and start all over again with the little finger of the left hand and climb to the turning point:

(12)

laifu-senan produce-little finger/26

laifu-senanafül produce-ring.finger/27

laifu-khabéan produce-head/38

This would extend the bodypart system of Korowai to 38. However, I have never heard such extended numerals used outside sessions with informants.

Bodypart tally systems and Bible translations

Whereas bodypart tally systems easily and spontaneously adapt to function in some new contexts like the days of the week or the months of the year, the high and complex numbers found in the Bible present serious problems for societies employing bodypart tally systems. In order to express numbers higher than the highest number, there are basically two solutions. First, using borrowed numerals in combination with bodypart numerals. Second, expanding the bodypart tally system beyond its natural limits by using periphrastic numeral expressions based on multiples of 'rounds' plus traditional bodypart numbers.

As far as borrowing is concerned, bodypart tally systems offer a number of logical 'break-off' points after which borrowed numerals could take over:

(a) the turning point, sometimes called the central point, that is the point on the head after which the counting starts its descent via the other side of the body; in the Korowai system khabéan 'head'/'thirteen' is the turning point after which counting on the righthand side of the body begins, with mén 'other side' (mén-khotokhal 'ear on the other side'/'fourteen');

(b) the highest point, that is the number associated with the little finger of the right hand, after one 'round', in the Korowai system mén-senan 'the other little finger/twentyfive';

(c) the extended highest point after an extra round while prefixing a word meaning 'return' or 'again', in Korowai that would be laifu-khabéan 'produce-head/'thirty-eight'

The South Yali Bible translation of Irian Jaya exemplifies the first option: the South Yali translation uses borrowed Indonesian numerals for numbers higher than fourteen, the crown of the head, which is the turning point of the South Yali tally system. South Yali is a language of the Great Dani family of Irian Jaya, Indonesia.

Kombai, also of Irian Jaya, exemplifies the second option, introducing Indonesian borrowed numerals after the highest point, after one round. In Kombai, 23 is the highest number after one round. Consider the following example from the Kombai translation of the Gospel of Mark (chapter 8 verse 19):

(13)

Nu roti ambalo-khu mofena lefa

I bread thumb-ATTR that take.SS

famo lima-ribu khof-o ri

cut.SS five-thousand man-CONN for

'...I took the five pieces of bread and cut bread for five thousand men..'

For 5000, the Indonesian numeral limaribu is used whereas for 5 the Kombai bodypart expression ambalokhu is used. I have not (yet) seen translations employing the 'return' prefixes to extend the tally system after one round (see the Korowai example (12) above) and switch to borrowed numerals after one extra round.

Instead, some translations extend the bodypart system by employing periphrastic numeral phrases, the second solution. For example, Angguruk Yali has 27 as the highest number after one round and for higher numbers periphrastic numeral phrases are used in the Angguruk Yali Bible, based on the Yali word teng 'round'. For example, the number 30 is 'one teng (round=27) and ringfinger (=3)'. A number like 70 is rendered as 'three times teng'. Since that is not an exact equivalent (three times 27 not being 70), the Indonesian numeral is added in brackets and in italics. In 2 Samuel 6 verse 2, the Angguruk Yali translation has teng teng angge (tiga puluh ribu)', literally 'so many rounds (thirty thousand)'; in 2 Sam 15 verse 11 two hundred is 'eigth teng (dua ratus)', literally eight times round (two hundred)'. The idea is that mentioning 'eight times round' helps the Yali readers to conceptualize the number involved, which is then exactly specified by the Indonesian loans within brackets.

A similar strategy is followed in the Ngalum translation but in Ngalum, an Ok family language of the eastern highlands of Irian Jaya, the exact number is given in digits introduced by the Ngalum word nentep 'as'. Thus Ngalum has in Matthew 13 verse 8 for sixty times: 'deng lao topa bangup nentep, 60,..'= 'round two add wrist like, 60,..' (one round (deng) is 27 in Ngalum, so two (lao 'two') rounds is 54, plus (topa 'add') bangup 'wrist/six' makes 60).

Since bodypart tally systems in most cases are rarely used for numbers higher than one round, it is questionable to what extent native speakers can conceptualize the higher numbers denoted by periphrastic numeral phrases. It has often struck me that even the translators themselves make mistakes when backtranslating the periphrastic numeral expressions into Indonesian whereas they have no problem backtranslating the bodypart numerals that fall within the range of the first round up and down the arms. Stylistically, the periphrastic extensions of the bodypart tally system creates quite heavy and marked numeral expressions, especially when a digital or Indonesian number specification between brackets is added.

In New Guinea, when minority languages start to interact with national languages, numerals and discourse conjunctions belong to the very first categories to be borrowed. In Irian Jaya, we find discourse conjunctions like terus 'furthermore, and', baru 'and, next' and jadi 'therefore' and Indonesian numerals very soon after initial contact in oral texts recorded by linguists and anthropologists. Whereas the reasons behind the early borrowing of discourse conjunctions are complex, the early borrowing of numerals can be connected to the introduction of Indonesian or Papuan New Guinea money in interior communities.

If a speech community is in the process of borrowing numerals from Indonesian or Tok Pisin, translators just starting the translation of the Bible generally should be advised to use the bodypart tally system within its natural range (one round) and switch to borrowed numerals for the higher numbers. This avoids the stylistic problems of the periphrastic extensions. In most cases, the community will switch to borrowed numerals for the higher numbers sooner or later and it would be unnatural to use borrowed numerals in daily life but periphrastic numeral phrases in the translation of the Bible.

In revisions of New Guinea translations it would be a good idea to lower the 'break-off' point for the use of borrowed numerals if borrowed numerals have become more generally used in the meantime, for example to use bodypart tally numerals just for numbers lower than the turning point as in the South Yali translation.

Whatever solution is chosen, all translations have to cope with the fact that higher numbers will become only easy to understand if such higher numbers play a role in the daily transactions and dealings of a people. When this is not the case, the complex higher numbers found in the Bible will remain difficult to grasp. When performing understanding tests on Bible translations, bad results in the testing of higher numeral expressions should therefore not automatically lead to revisions of translations. If such revisions imply added clarifications, there may be stylistic losses without any conceptual gain.

Abbreviations

ATTR attributive

CONN connective

COP copula

NF non-future

SG singular

SS same subject (switch-reference)

TR transitional sound

References

Boelaars, M.S.C.: Mandobo's tussen de Digoel en de Kao. Assen: Van Gorcum 1970.

Bruce, Les: The Alamblak language of Papua New Guinea (East Sepik). Canberra: The Australian National University 1984.

Enk, Gerrit J. van & Lourens de Vries: The Korowai of Irian Jaya. Their language in its cultural context. Oxford/New York: Oxford University Press 1997.

Franklin, Karl & Joice Franklin: The Kewa counting systems. Workshop Papers of the Summer Institute of Linguistics, New Guinea Branch, 1-15. Ukarumpa: SIL 1961.

Laycock, D.C.: Observations on number systems and semantics. In: Wurm, ed. 1975: 219-233.

Schoorl, J.W.: Kultuur en kultuurveranderingen in het Muju-gebied. Den Haag: Voorhoeve 1957.

Venema, Henk: Sagogrub festival. Irian, Bulletin of Irian Jaya, vol. XVII, 38-63, 1990.

Vries, Lourens de, & Robinia Wiersma: The morphology of Wambon of the Irian Jaya Upper-Digul area. Leiden: KITLV Press 1992.

Vries, Lourens de: Forms and Functions in Kombai, an Awyu language of Irian Jaya. Canber­ra: Australian National University 1993.

Vries, Lourens de: Numeral systems of the Awyu language family of Irian Jaya. Journal of the Royal Institute of Linguistics and Anthropology, vol. 150-III, 540-567, Leiden 1994.

Vries, Lourens de: Notes on the morphology of the Inanwatan language. NUSA, Linguistic Studies of Indonesian and other languages of Indonesia, vol.40, 97-127. 1996

Wassermann, J.: The final requiem for the omniscient informant? An interdisciplinary approach to everyday cognition. Culture and Psychology, vol.1, 167-201. London and New Delhi 1995.

Wurm, S. A. (ed.): New Guinea area languages and language study, vol.1: Papuan languages and the New Guinea linguistic scene. Canberra: The Australian National University 1975.

[1]The Alamblak tally system (by now most probably no longer in use) differentiated between male amd female speakers: it had a men's tally system and a women's tally system in which the female breasts were incorporated. The men's system was of the usual type starting with the little finger and going up the arm via shoulder (10), neck muscle (11), side of neck (12), ear (13) and eye (14) to the nose, the highest point (15), after which it went down again with a qualifier meaning 'other side' (other side eye=16) (Bruce 1984).

[2] The Kewa of Papua New Guinea (Franklin and Franklin 1961) seem to have the most elaborate bodypart tally system: they count to 47 by employing a maximum number of points on hand, arm and head. After the five fingers, the Kewa continue with back of the hand (6), palm of the hand (7), wrist (8), lower arm (9), radius bone (10), ulna bone (11), inside of elbow (12), lower part of upper arm (13), upper part of upper arm (14), side of shoulder (15), shoulder (16), collarbone (17), neck (18), jaw (19), ear (20), temple (21), eye (22), corner of the eye (23), nose (24). The nose is the Kewa turning point after which they descent the right side of the body finally reaching the right little finger (47!).

[3]In hands-and-feet systems, counting starts with the fingers of the left hand, and then continues with the fingers of the right hand, with 'both hands' denoting 10. After 'both hands' these systems continue with the toes until 20 is reached (usually the noun 'body' or 'dead body' representing 20, two bodies=40, one body and one hand=25 etcetera). Hands-and-feet systems come in all sorts, many of them have primitive numerals for 1, 2, 5 and 20. In theory, languages with these systems can express much higher numbers with their binary/quinary/vigesimal primitives but in practice higher numbers require very long co-ordinate numeral phrases which are so difficult to process that speakers normally switch to the shorter, borrowed numerals, for example 131 would be expressed as newo-gagua ga iragiro nagia ga iragiro newowa sugeri nagia in the Inanwatan language of the Bird's Head area of Irian Jaya, Indonesia (De Vries 1996), literally 'one hand bodies (five bodies=100) and one body (20) and both hands (10) and one (=131)'. I have no data on Bible translations of New Guinea using such long, co-ordinate numeral phrases of hands-and-feet systems.

[4]The intriguing question how Yupna women count is left unanswered by Wasserman (1995). The Yupna system is exceptional in many ways: it combines a hands-and-feet system (until 20) with a bodypart tally system employing bodyparts of the lower half of the body.

[5]The cognates of Mandobo tagét 'kauri-shell' in Kombai (rakhe) and Wambon (takhet) form the basis of the verb for 'to buy/to sell' which is derived in Awyu languages by verbalization of the noun denoting kauri-shells with the help of support-verbs. For example Wambon takhimo- 'to buy/to sell' and Kombai rakhimo- 'to buy/to sell'.