Volume 11 1902 > Volume 11, No. 1 > The Polynesian numerals, one, five, ten, by John Fraser, p 1-8
                                                                                             Previous | Next   

- 1
The Journal of the Polynesian Society.
VOL. XI. 1902.
Part II.—The Numeral “Five.”

THE first thing that I have to say on this head I will write down as Proposition IV:—The Polynesian word for ‘five’ is lima, which also means the ‘hand’; it is therefore probable that language-names for ‘five’ are founded on words for ‘hand.’ And that is not merely probable, but certain; and it is also so natural that no proof of this Proposition is needed. I will here mention only three facts:—the Turkish word for ‘fifty,’ that is, five tens, is éli, and that is formed from el, ‘hand,’ il-ik, ‘hand’; the Dravidian dialects of India have kei, ‘hand,’ and ei-du (for kei-du), ‘five’; the Indo-Persian atilang and laman both mean ‘five’ and also ‘hand,’ from roots ati and lam, as below. Observe here that the Dravidians form the word for ‘five’ from ‘hand,’ not ‘hand’ from ‘five.’

Proposition V.—Words for ‘hand,’ in several languages, show that the meaning of their base is “to hold,” ‘to grasp.’ The hand is the ‘grasper’; that also is a natural meaning. In science, the monkeys are classified as quadrumana, because they use their four extremities as hands for grasping. In Scotland, hadd means ‘to hold’; even a child will understand you if you say “Hadd it fast; don't let it fall.” It is a stronger word than ‘have’; have simply implies possession - 2 after acquiring; hadd means ‘to grasp firmly and to hold in continued possession.’ And so English legal documents conveying property use a double term, ‘to have and to hold.’ The original force of the verb to have in English has been much weakened by the use of have and had (hav-ed) as auxiliaries, but hadd is not used in that way in Scotch.

Language, which, in the process of forming words, has stereotyped many of the original ideas of man, conveys this notion of ‘grasping’ in the name for ‘hand.’ Hand itself is only a strengthened form of hadd; the Sanskrit1 has har-ana, ‘taking,’ ‘seizing,’ and, as a noun, ‘the hand,’ from the verb hri (har), ‘to seize, to take, to steal.’ From this root comes the Greek άρ-π-άζω ‘I take, I steal,’ and possibly also the noun χεìρ, ‘the hand.’ If you should ever be among the courtly Hovas of Madagascar and there wish to be considered polite, you must not use the common word tanana for ‘hand’; you should say fandray, ‘ the taker.’

It is clear, then, that the hand is that with which we ‘take or seize’ anything.

Proposition VI.—In the Aryan languages there is a simple root ga, ka, which means ‘to take, to seize.’ As usual, this monosyllable may be amplified by adding on to it single consonants or single syllables. Thus arise the root-forms gam, gab, kap, kat; for m and b on oriental lips are almost the same in sound. Examples here are:—Pâli gahi, ‘taking’; Latin kap-io (capio), ‘I take’; English grab, where the rough r is thrown in to intensify the sound and the meaning; Malay chap-ei, ‘seize.’

Proposition VII.—The consonant of this root ga, gam, etc., may be changed, according to established rules; for example:—

1. The guttural g, k may be softened into a sibilant: as, Sk. sam, ‘with,’ which is akin to Lat. cum; Sk. çat-a, 100, Gr. έκατύν; Lat. can-is, ‘a dog,’ Sk. çran; Sk. paç-u, Lat. pec-u, ‘cattle.’

2. The g may be softened into h; as, Sk. ha (a particle), for the Vedic gha, Gr. γέ; Sk. hams-a, ‘goose,’ Lat. ans-er, Germ, gans, Eng. gand-er.

3. The g may be softened into y; as, Sk. yam-a, ‘twin,’ Lat. gem-inus.

4. The g may become the dental d; as, Gr. γΗ˜, ‘earth,’ whence, -mētēr, “Mother-earth.” Our children say dood for good.

- 3

5. The g having been changed into d, that d may become a liquid l or r: as, Sk. daha (in compounds), 10, Prâkrit raha; Lat. lingua (for dingua), Eng. tongue.

6. The g may become p, b, simply or through the influence of an intrusive r or w; as, Beloochi gwath, Persian bád; Bel. gist, Pers. bist; Gr. κω˜ς and πω˜ς; Gr. λν´κ-oς, Lat. lup-us.

7. The guttural g or k may become a palatal; as, from root kap, Malay chap-ei, ‘to seize,’ as above.

8. The gh, that is, g aspirated, is dropped when it comes between two vowels, or it becomes v; as, Sk. lagh-u, ‘light,’ Lat. lex-is; so also bh; as, Irish gabh-ar, ‘goat,’ pronounced like gau-ar, Lat. cap-er.

9. The labials b and p, when aspirated, are sounded like f or r; and m, when aspirated, becomes v, and then y; as, Sk. kām-a, ‘(hot) desire’; Dravidian kây, ‘to be hot, to burn.’

10. The k may become t, a process which is going on in Polynesia.

11. In eastern languages, especially in Polynesia, the initial g or k is very often dropped.

Proposition VIII.—From this root ka, ga, are formed many words in the languages of Europe, Asia, and Occania, meaning ‘to take hold of,’ the ‘hand.’

I give some examples of each of these two meanings:—

(A.) To seize, grasp, etc. (B.) Hand.
1. kapi-o (Lat.), I take (capio) 1. kai (Weddahs of Ceylon)
2. kap-o (Maori), to seize 2. kei—for kamh, kav, kay? (Dravidian)
3. kappa - kapa (New Britain), to handle 3. keka (Polynesian of Paumotu)
4. kube-ta (Fiji), to lay hold of 4. kha (Naga tribes of India)
5. ga-hi (Pali), taking 5. kappi (ancient Akkadian)
6. ga-hanam (Pali), holding 6. kan-im (New Caledonia)
7. gan-hāti (Pali), take, capture 7. khut-i (Caucasian)
8. gâng-am (Malay), to grasp in the hand 8. ga-tu (old Assyrian)
  9. ge-tō (Torresian)
  10. got (Vayu of Nepal), hand and arm
and, by Prop. VII.—2, 11  
9. 'abh-ati (Pali), to take, seize 11. gub-blem (Vayu), hand (blem=‘flat’)
10. 'ati, 'adi (Aneityum), to take, to take off 12. gabh-asti (Sk.), arm
11. 'ave (Samoan), to take  
12. 'abia (Motu), to take and, by Prop. VII.—2
  13. 'ati-lang (Panjab)
and, by Prop. VII.—7  
13. chap-ei (Malay), to grasp in the arms and, by Prop. VII.—2
  14. hat; hath (Nepal)
  15. hatth-o (Pali)
- 4
and, by Prop. VII.—2 16. hath (Maithili)
14. hop-u (Maori), to seize 17. hath-na; át (Kashmir)
  18. hasta (Sk.),
and, by Prop. VII.—5  
15. labh, lambh (Sk.), obtain, get and, by Prop. VII.—7
16. lâ (Sk.), to take 19. chak; yak (Naga tribes)
17. lâbha (Sk.), acquisition  
18. leb (Maithili), to take and, by Prop. VII.—1
19. leh (Aneityum), to take 20. çam-an (Sk.)
20. lau (Fiji), handle  
21. lamb-ano (Greek), I take and, by Prop. VII.—5
22. lab-ē (Greek), a handle 21. lamh (Gadhelic)
23. labh (Beloochi), obtaining 22. lá (Nepal)
  23. lam-an (Kashmir)
and, by Prop. VII.—6 24. lag (Magar of Nepal)
24. pam (Tukiok Is.), to hold 25. las (Pukshto)
25. pam-pam-ina (Tukiok Is.), to handle 26. a-ring, ‘my hand’ (Nager to the north of Gilgit, in the eastern Hindu Kush)
26. pa-lum (N. Brit.), to handle  
27. pa-ksh (Sk.), to seize  
  Other examples are:—
Other examples are:— 27. yam-an (Hebrew), right hand
28. grah; gras (Sk.), to seize 28. immi (old Assyrian), right hand
29. sham-ala (Arabic), to grasp  
30. ādā (for kada?) (Pali), to take 29. ni-kman (Aneityum)
31. gamu-ta (Fiji), to take hold of 30. lima (Polynesian)
32. juw-it (Java), to nip, pinch 31. thab-ung (Iaian of Loyalty Islands)
33. chub-it (Malagasy), to nip  
34. a-ngkub (Malagasy), pincers 32. lima-na (New Britain)
35. kham-as (Hebrew), to grasp 33. no-kob-en (Erromanga)
36. 'um-i-ki (Hawaii), to pinch 34. rim-ata (Tasmanian)
37. khab-as (Arabic), to take, grasp 35. 'a'a-o(=kakao)(Samoan), the hand of chiefs
  36. gav-é (Santo)
  37. sham-aleh (Gipsies of Egypt)
  38. ada (New Guinea)

(A.)—To grasp.—No. 1. From capio comes cap-istrum, ‘a handle,’ with which compare Nos. 3, 20, 22, 25, 26.—No. 3. New Britain is one of the islands (Melanesian) of the Bismarck Archipelago, east of New Guinea.—No. 5. Pâli is one of the Prâkrit or original and common languages of India.—No. 18. Maithili is one of the dialects of the Behar district in Bengal.—No. 19. Aneityum is the most southerly island of the New Hebrides (Melanesian).—No. 23. Tukiok Island is a small island close to New Britain. In the eastern Pacific there is another Duke-of-York island; so I call this one Tukiok, the native pronunciation of its name.

- 5

(B.)—Hand.—No. 1. The Weddahs are the aborigines of Caylon.—No. 2. The Dravidian languages (Tamil, Telugu, etc.) are mostly in the Madras Presidency of India. They are non-Aryan and pre-Aryan.—No. 3. The Paumotu group of islands is far east in the Pacific. Although Polynesian, its dialect has many striking peculiarities, as if from a strong admixture of previous inhabitants.—No. 4. The Naga tribes (Turanian) are on the N.E. frontier of India.—No. 7. Caucasian; khuti is from one of the many languages spoken in the Caucasus region.—No. 10. Torresians are the Papuan natives of the islands in Torres' Straits between New Guinea and Australia.—No. 11. The Vayus are non-Aryan a borigines in the Central Himalaya region, subject to Nepal.—No. 21. The language of the Scottish High-landers.—No. 31. Iaian is the native name for the language of Uvéa Is. there. Nengonese is the language of Maré Is. of the same group.—No. 33. Erromanga, one of the southern New Hebrides.—No. 36. Santo, a large island, most northerly of the New Hebrides.—No. 38. The ada, ‘hand,’ of the Koiari tribe on the southern coast of New Guinea, may be the Pāli ādā, ‘to take.’

Proposition VIII.—Lima, the Oceanic word for ‘five,’ which also means ‘hand,’ comes from the root ga, ka, and is connected with Aryan, Semitic, and non-Aryan words for ‘hand.’

This Proposition implies that the mode of counting from five upwards has been by hands—the quinary system. I must give some proof of this statement. Hence the three sections that follow.

I.—Five is ‘Hand.’

(a.) The Roman numerals VI., VII., VIII., are founded on V., which is ‘five.’ Now V. is a picture of the side view of the thumb and the forefinger when the hand is held up in a natural way; VI. is these two and one finger more, VII. two fingers more, VIII. three fingers more, which exhausts the hand, when the thumb and the forefinger of that hand have been used for ‘five’; therefore, the symbol for ‘nine’ is not based on V., but is IX., that is, X. or ‘two hands’ with one finger less, as will be shown further on.

(b.) The numeral signs of the old Mayas of Central America correspond with the Roman system; for the numbers under five, they used simply dots; for five a bar, for six a bar and one dot, and so on; for ten, two bars; for nineteen, three bars and four dots; for twenty, four bars.

(c.) The islanders of the Andaman group in the Bay of Bengal are a very primitive people. Their numeration resembles the Australian; for they say ‘one,’ ‘two,’ and no more; all beyond is ‘several, many, numerous, innumerable.’ When they wish to indicate other numbers than ‘one,’ ‘two,’—the number ‘nine’ for instance,—they tap the nose with the fingers of either hand in succession, beginning with the little finger, saying ūbatūl (‘one’), īkpôr (‘two’), and for each of the other fingers of that hand, ankā, ‘and this’ then four fingers of the next hand are held up together, the thumb being turned down; that is ‘nine’; at the same moment they say ardūru, ‘all.’ For ‘ten,’ the one hand is used as for ‘five’; but, before the two hands for ‘ten’ - 6 are held up, they are brought together at the sound of the word ardūru, ‘all.’ They do not use the toes in counting higher numbers. The ‘first’ in a row of objects is called ô-kôtap, the second tô-koyo-to.

(d.) In the Admiralty Islands (Northern Melanesia), sangop is the word for ‘ten’; but, when using it, the natives hold out the two hands with the fingers pointing forwards, bring the two open hands together, and clap them once. They use subtraction to express ‘eight’ and ‘nine’—anda sip, 9; anda huap, 8; sip is ‘one’ and huap is ‘two.’

These examples are sufficient, I think, to show that ‘five’ is ‘a hand.’

II.—‘Lima’ is from ‘Ka.’

Next, my Proposition requires me to show that lima comes from the root ka, ga. That is an easy task. Thus:—

Root ka, ka-m
1. çam-a, i.e., kam-a, ‘hand’ 2. ni-k(a)m-a-n, ‘hand’ 3. dim-y (Prop. VII.—4), i.e., dim-é or dim-a, ‘five’ 4. lim-a (Prop. VII.—5), ‘hand’ and ‘five’

(1.) is Sanskrit; (2.) is from Aneityum, of the New Hebrides, where the ni is a common demonstrative prefix; (3.) is from Madagascar; and (4.) is the Oceanic word in question.

III.—Aryan and other cognates of ‘Lima.’

Next, Lima is cognate, in origin, to some Aryan, Semitic, and non-Aryan words for ‘five’ and ‘hand,’ all of which come from the same primal root ka.

(a.) Aryan. (b.) Semitic.
1. †ka-ka; cf. Prop. VII.—(B.), 3 and 35 1. Old Assyrian, (y)im-mi, ‘right hand’
2. †kam-ka 2. Hebrew, yam-an, ‘right hand’
3. †Gr. (old) κεγ-κε Hebrew, hham-esh, ‘five’
4. Lat., quin-que 3. Arabic, khams, ‘five’
5. Gadhelic, co(mh)-ig, cu-ig Arabic, sham-aleh, ‘hand’
6. Gr., πέμ-πε, πέντε  
7. Sk., pan-cha (c.) Non-Aryan.
8. Old Pers., pean-che, pan-dj 1. Akkadian, kap-pi, ‘hands’
9. Pre-Roman tribes, pum-p 2. Dravidian, kei, ‘hand’; (k)ei-du, ‘five’
10. German, fü-f; Goth., fim-f  
11. English, five 3. Santo, gav-é (gamh-e), ‘hand’
All these words mean ‘five’ 4. Aneityum, ni-kman, ni-jman, ‘hand’
† These are obsolete in the Aryan tongues. 5. Vayu, gub-blem, ‘hand’
  6. Tasmanian, rim-ata, ‘hand’
- 7

Strange as it may seem, this evidence shows that the Gaelic lamh of Scotland and the Polynesian lim-a, ‘hand, five,’ are of the same origin, but with many intermediate links to connect them.

Proposition IX.—Lima is Melanesian, Polynesian, Micronesian, Papuan, and Indonesian in its distribution.

(A.) Melanesia.

(a.) For ‘hand,’ these are the variations found in the New Hebrides;—Lima, rima, jima, juma, ma, nel-limi, lime-gi; irregular words for ‘hand’ are gav-é, no-kob-en; raga; rari, vera, fera, rer.

(b.) For ‘five’:—Lima (alone, or with i-, e-, a-, ki-, mo-, prefixed), rima, lim; also tara-ligma, gai-lima, kadi-lum, kari-lum, suk-rim, all of which five words mean ‘one hand.’

(c.) For ‘hand,’ Fiji says lig-a, that is, lim nasalized, and New Britain has lima-na.

(B.) Polynesia.

(a.) In Fornander's list of Oceanic numerals, there are fifteen examples of words for ‘five’ from the islands occupied by the brown Polynesians. Of these words 5 are lima, 6 are rima, 1 is nima, 1 is ima, 1 is ringa, and 1 (Paumotu) is aggoka.

(b.) In his list also there are twenty-two words for ‘five’ from the islands of the East Indian Archipelago. Of these 10 are lima, 3 are rima, 1 is nima, 2 are limo, 1 is lima-k, 1 is lema, 1 is lep-lim, 1 is lumi, 1 is run-toha, 1 is ma-toha; the last two appear to me to mean ‘one hand.’

(C.) Micronesia.

(a.) Lima in all Micronesia.

(b.) Although Formosa and the Philippines are only adjacent to Micronesia, yet a census of their words for ‘five’ may be taken here. I have a list of thirty-two localities in these islands. Of these 15 have lima, 2 have tima, 5 rima, 3 rima with a prefix (da-, tu-), 1 hrima, 1 lima-ngan, 1 nimo, 1 magal, 1 hasub, 1 rassoum, 1 laleup.

(D.) Papuan.

(a.) Of eighteen districts on the south coast of New Guinea (British territory), 1 has lima with the prefix ko, 1 has limi, 4 have ima, 1 has nima-na, 1 has nim with the prefix kwei, 1 has ima'a; the other words seem to have no connection with this root.

(b.) The Torresians have no words for ‘five,’ but the Australians at Cape York, opposite to them, say ungatua, ‘the whole’ (sc. hand) for ‘five.’

(E.) Indonesia.

This term, as I use it, includes the whole area of the East Indian Archipelago. For this area I have Wallace's list of words from 59 localities, and Crawfurd's from 14. Of words for ‘hand,’ Wallace - 8 gives the following varieties, viz., tangan, lima, rima, niman, lima-mo, fahan, gia, ala, kanin, mot: additional words from Crawfurd are hasta, chas, langan, which three are evidently from the root kam. Hasta is a Sanskrit word for ‘hand,’ and, although my Sanskrit Dictionary gives a different account of it, I take it to be for ha(m)s-ta, from kam: chas is for kams, and lang-an is from kam nasalized.

In Wallace's list, lima appears seventeen times for ‘five’; variations are rima, nima, lim, en-lima, de-lima, lep-lim, lima-nu. In Crawfurd's fourteen, lima appears 6 times: other forms are limo, la-lima, lime, lema, rim, each once, dimé once; other words for ‘five’ are gangsal, poncho (Sk.), luwi. Crawfurd cannot explain the derivation of gangsal. I take it to be for gams-al from our root gam.

In fine, from all these considerations and arguments, I am clearly of opinion that the Polynesian word for ‘five’ is the ‘hand’ as the ‘grasper,’ and that it comes from a primitive root ka, kam.

While this is going through the Press, an important confirmation of the basis of my arguments has come under my eye. Philologists are sometimes asked, Where is this root of which you make so much? Does it exist anywhere? Those who read these pages will see that it has many guises and disguises in many languages, yet it is clear that the root gam, kam did once exist. But I can produce proof that it exists now. In a recently published dictionary of the language of the island of Efaté, New Hebrides, I find (page 88) these entries:—“kam, ‘native tongs,’ literally, that which seizes, grasps; (dialects)—kau, as a noun, ‘tongs,’ as a verb, ‘to grasp with the hand’; kam-kam, ‘scissors’; kam-ut-ia, ‘to grasp with the fingers’; kam-i, ‘to seize, grip, take with the fingers’.”

(To be continued.)

1  It should be understood everwhere that I quote the Sunskrit merely as a very ancient tongue, which, by means of its oral and written literature, has preserved many of the oldest forms of Aryan speech in a proximately pure condition. The richness of its vocabulary helps one also to see how easily words were at first made from simple elements. I hold likewise that the Aryan, Semitic, and Turanian families of language are closely related in their elemental stock of words.