3

I noticed that the first three digit words for most Austronesian languages are awfully close to P.I.E. I speak Tagalog and at first, I had thought that the words for one two and three had been taken from Spanish, (Isa, Dalawa, Tatlo) but when I went to look at the Austronesian Languages page on Wikipedia, I was baffled to see that they had been so close to P.I.E. Why is this so?

Marco Ruben Abuyuan Llanes
  • 141
  • 4
  • 4
    From Wikipedia: Proto-Austronesian: əsa/isa, duSa, təlu. Proto-Indo-European: Hoi-no-/Hoi-wo-/Hoi-k(ʷ)o-, d(u)wo-, trei-/tri-. They don't seem very similar to me. The only point similarity is the d and t, but no other numerals share similar initial consonants. – curiousdannii Sep 11 '16 at 14:38
  • 1
    @curiousdannii your list makes them even closer. A Nostratic connection? The lost tribes of Sulawesi found in Hallstatt culture? – Mitch Sep 12 '16 at 17:14

2 Answers2

4

The word for "two" is dua in Malay/Indonesian and duo in Latin. This is a classic example of how words in two unrelated languages turn out looking the same, by pure coincidence.

fdb
  • 24,134
  • 1
  • 35
  • 70
  • That's very odd. – Marco Ruben Abuyuan Llanes Sep 11 '16 at 15:20
  • 13
    No, it is not odd. The statistical probability of random correspondences within a relatively small battery of phonemes is actually fairly large. – fdb Sep 11 '16 at 15:33
  • 8
    @MarcoRubenAbuyuanLlanes think of it this way: if each language has tens of thousands of words, then statistically some will resemble each other. what's more, high-frequency words tend to be short, and the shorter they are, the more likely will coincidences be. – melissa_boiko Sep 11 '16 at 21:21
  • 4
    http://www.zompist.com/chance.htm – fdb Sep 11 '16 at 21:31
  • 6
    Two is most assuredly not odd. :) – Robert Columbia Feb 17 '17 at 14:43
  • 2
    The thing with coincidence, as with rainfall (viz. cado "I fall", accident, katastrophe, etc.) is that people recognize patterns and they suck at statistics so that in result they don't know any better than to attribute happenstance like flood to the gods of coincidence, and poor answers to the eggspertiese of iranianists and celtcysts. just great – vectory Mar 03 '22 at 14:51
0

Linguists have noticed this similarity, and it is not just a one-off case like another answer seems to suggest. Quoting from Shrikant Talageri's blog:

Isidore Dyen, in a paper presented in 1966 and published in 1970, makes out a case showing the similarities between many basic words reconstructed in the proto-Indo-European and proto-Austronesian languages, including such basic words as the first four numerals, many of the personal pronouns, and the words for "water" and "land". And Dyen points out that "the number of comparisons could be increased at least slightly, perhaps even substantially, without a severe loss of quality" (DYEN 1970:439).

Also:

• a) The very first four numerals: Proto-Indo-European (*sem, *dwōu/*dwai, *tri, *qwetwor) and Proto- Austronesian (*esa, *dewha, *telu, *pati/*epati).

• Compare Tocharian sas/se 'one', Romanian patru 'four', Welsh pedwar 'four' and Malay sa/satu 'one', epat 'four'. [Malay dua 'two' and tiga 'three' require no comparison].

• b) Personal Pronouns: I, we, you, he/she/it, (demonstr.) this/he: PIE *eĝh, *ṅsme, *yu, *eyo/*eya, *to/*eno. PA *aku, Tagalog ka-mi, Tagalog ka-yo, PA *ia, *itu/inu.

• c) "Land" and "water": PIE *wer, *ters. PA *wair and *darat (Sanskrit vāri and dharā).

[EDIT: vāri (वारि) = water and dharā (धरा) = land. Clarifying this because the quoted text gives the opposite (and wrong) impression.]

The author of the blog subscribes to the out-of-India theory (OIT), in which India is the homeland of PIE. That makes any contact between PIE and Proto-Austronesian a reasonable consequence.

cobra
  • 101
  • 1
  • 2
    does Shrikant Talageri have any linguistic qualifications? If not his blog is not evidence that linguists have noticed. Regardless, the other examples cited are poor evidence, similar pronouns are much more widely spread than just these two families and so cannot be evidence of an especially close relationship between the families. PIE has no reliably reconstructible word for "land" as different branches reflect different etyma, and no such form for water exists – Tristan Feb 28 '22 at 13:16
  • 2
    most damningly, correspondences appear to be inconsistent with each other (does PIE t correspond to PA t or d? Does PIE r correspond to PA r or l? etc); unless a conditioning environment for a split is identified, or he can identify a plausible phoneme that merges one way in PIE and another in PA (noting that PIE & PA both have both r & l, and both t & d), this situation cannot have been reached by regular sound change and must be rejected in the comparative method – Tristan Feb 28 '22 at 13:19
  • 1
    @Tristan As I quoted, the blog mentions that it was linguist Isidore Dyen who found those similarities. As long as that is true, isn't that enough evidence that linguists have noticed? Talageri's linguistic qualifications are immaterial for this specific part at least. – cobra Feb 28 '22 at 14:02
  • 4
    ah yes sorry, you're right, I missed that part. His claims don't stand up to much scrutiny from the comparative method though, but then Dyen was a supporter of lexicostatistics, which is a different method, and one which has only limited support amongst the broader linguistic community – Tristan Feb 28 '22 at 15:39
  • 1
    @Tristan The comparative method "don't stand up to much scrutiny", as a matter of irony, seeing linguists widely agree that there is no rigurous method for reconstructing semantics and that scholarly opinion should be honoured on those counts. PIE *ters- "dry" is widely cited. Interpreting that to mean "land" should be understandable. – vectory Mar 03 '22 at 15:16
  • 3
    @vectory the quote appeared to suggest ters as the word for water. It being "dry" is indeed widely cited. I can find no source for wer meaning either water, or land, or anything especially similar to either – Tristan Mar 03 '22 at 15:22