From: João Simões Lopes Filho
----- Original Message -----
From: Glen Gordon <glengordon01@...>
Sent: Friday, August 25, 2000 6:42 PM
Subject: Re: [tied] Proto-Steppe Numerals
> The theory has evolved. Take note that I now consider Altaic, Gilyak and
> Korean as three distinct branches of an AltaicGilyak grouping. Thus Gilyak
> is on a par with Korean since Gilyak shows the characteristic lenition of
> voiceless stops. As well, Altaic I is to be considered AltaicGilyak and
> Altaic II is to be considered Altaic proper.
> Also, there are improved reconstructions now for the Steppe words for
> "seven", "eight" and "nine":
> 6 *ru
> 7 *rara
> 8 *munra
> 9 *nura/nukura
> Now on to the discussion:
> >But if you compare, say, the numerals of PIE and Proto-Quechua, you'll
> > >find even more spectacularly similar words for '5' and '6'; and if you
> > >take Proto-Austronesian and PIE, you'll find the correspondence
> >the words for '2' and '3' in both families quite literally too >good to
> This is such a deceptive arguement. The geographical proximity between
> Uralic and Mongolic can hardly be compared to the other pairs you
> this only serves to make their relationship seem more likely. This
> with common grammatical particles, vocabulary and grammar shared between
> Uralic and Mongolic that show regular sound correspondances, besides just
> "three" and "four", makes the link even more persuasive. Many sources even
> admit to the probability that Altaic and Uralic are related (unlike your
> and Quechua/Austronesian pseudohypothesis that only serves to confuse the
> issue here)
> Thus, when we have a pair of numerals that correlate perfectly between two
> language groups already suspected of being related, we have to sit up and
> take note.
> >Your explanation of the relation between '3' and '9' in Japanese in
> > >ingenious and I have no difficulty accepting it.
> It's not my idea. I nicked it from something I read a while back on
> Classical Japanese. :)
> >However, some problems remain: what you gain by solving this problem >is
> >rough match between Japanese, Mongolic and Uralic, with >Samoyedic,
> >Tungusic, Korean, Nivkh, Yukaghir, Chukchi->Kamchatkan and Eskimo-Aleut
> >showing apparently unrelated forms and the >alleged Indo-Tyrrhenian
> >giving you headaches, as you yourself >admit. As the Proto-Japanese
> >is somewhat conjectural, >everything really hangs on the agreement
> >Mongolic gur- and >Uralic kol(me) -- this is in fact the REAL basis for
> >your kWul(mu).
> In fact, the Mongolic word would be *qur- or *kur- since Middle Mongolian
> seems to have a variant /qurban/. I've seen *gur- reconstructed before,
> based of course on Mongolian and Japanese (and possibly other languages
> I forget) but Japanese can't tell us anything about the voicing of the
> initial stop and no doubt Modern Khalkha /gu"rav/ was relied upon.
> >By "reaching down", i.e. using forms not from proto-languages but from
> > >their arbitrarily chosen daughters you depart from the normal
> >procedure of the comparative method and greatly increase the >probability
> >of a chance match, as in Greenberg's mass comparison (especially when you
> >work with very short roots such as *ra or *na).
> The theory has evolved. Only "six" can now be said to be a "short" root.
> Secondly, I don't arbitrarily choose daughters and "reach down" - you're
> creating myths. The attestations that I select are ones that show up in
> than one language family (hence its hardly arbitrary). They also show
> regular sound changes which I'm beginning to understand more and more.
> The Steppe word *rara is a product of *ra- "seven" plus the suffix *-ra
> in higher numerals. It was already mistaken as a reduplicative form when
> AltaicGilyak changed it to *NaNa (N = velar nasal) since it should have
> become **Nara. This change of *r- to *N- occured via a uvular *R which
> appears to be a product of areal influence with other SinoDene languages,
> but I digress. According to zompist.com, the Nivkh word for "seven" is
> /ngamg-/ but I suspect strongly that this is a typo for an actual
> >From AltaicGilyak *NaNa, we get Altaic *nana (Japanese nanatsu, Tungus
> nadan). Afterall, Altaic has no velar nasals (despite Starostin's far
> loonier reconstructions of numerals like his *Nu "three" based on Khalkha
> gu"rav, Japanese mi- and Turkic *otur???).
> The fact that Gilyak, Tungus, Japanese (and possibly Mongolic Khalkha
> /doluun/) are the only ones preserving "seven" is hardly a shocker either.
> FinnoUgric languages as well as Turkic languages manage to overwrite their
> numeral with the ubiquitous Semitoid borrowings. Of course, we know that
> *septm is an early loanword as is Etruscan /semph/ which, too, has this
> pesky omnivorous numeral.
> Steppe *rara and the daughter languages that represent it are conscious
> deliberate choices. One would be hard pressed to find any other numeral
> with as wide an attestation and regularity as *rara.
> >Frankly, I still can't see why Samoyedic should be less likely than
> >Finno-Ugric to have retained Proto-Uralic forms. [...] To give a
> >example, Tungusic is your only witness for '6' (with some
> >feeble support from Japanese).
> I'll retract the statement and rethink this. Mind you, aside from
> *ru is retained in Gilyak /ngax-/ and Korean yo-sot.
> >Your admission that there is NO ATTESTATION AT ALL for the >reconstructed
> >numeral '8' is staggering. [...] As a matter of fact, >you abandon the
> >comparative method entirely in this case and simply >project your
> >form onto Proto-Steppe by "inverted >reconstruction". [...] as far as I
> >see, the corresponding >Sumerian, Dravidian and Afroasiatic numerals (if
> >reconstructable at
> >all) can hardly be cognates. From what evidence and by what esoteric
> > >method did you distil your Nostratic reconstruction then? The answer
> >the same as before: you project a still more ancient (Dene->Caucasian)
> >onto Nostratic -- ALL THE TIME WITHOUT A SHADOW OF >EVIDENCE.
> Du calme, monsieur. I knew that this numeral would ire you. First the
> logical assumption is that there must have been some word for "eight" in
> ProtoSteppe if so many of the others appear to be reconstructable. Thus I
> provided the likeliest reconstruction that I could come up with at the
> Given that you haven't even asked for the "shadow of evidence" but rather
> simply went straight for an emotional conclusion makes it appear that you
> are more interested in biased sentiment rather than rational thought. :(
> There is lots of evidence regarding DeneCaucasian *mnrit in both of its
> large branches:
> NostratoDene (s-Group):
> Nostratic *munri
> SinoDene *bRy@:t (via heavy stress on second syllable)
> SinoTibetan *bryat (Cantonese baat).
> VascoCaucasic (t-Group):
> Basque bederatzi "nine" (semantic shift due to Semitoid loanwords)
> NEC *bunLi (Chechen baarx)
> Hurrian miri-
> Along with *sulrit "seven" (Nostratic *(u)ra), it's clear that these two
> words were counting away from ten, since DC *sul is also attested and
> "three" (Nostratic *hul). It all even shows the regular disappearance of
> *s in Nostratic that I love so much (also *u "I (abs)", ultimately from
> DeneCaucasian *si).
> Finally, in Nostratic itself, outside of Steppe, *munri might be attested
> Semitic, Egyptian and Dravidian. And yes, you're right, Nivkh /minr/ is an
> attestation of Steppe *munra. I lost track of the Nivkh numerals in my
> >If you relax it ad hoc (as you do for the numerals '3' and '4' in your
> > >response), you sabotage your own project. In the case at hand you make
> > >things far worse by relaxing semantic constraints at the same time, as
> > >when you allow the '5' word for to mean '4' or '6' (even with so much
> > >latitude you can't get an unambiguos reconstruction for '5' and have
> >be content with two different forms in free variation).
> The theory has adapted. The term for five is now securely *kit:u which
> regularly changes to Altaic *ïtï through vowel harmony rules and softening
> of the voiceless stop *k. The Uralic phonetic deviation is caused by an
> early association between the pairs "one"/"two" and "five"/"six".
> IndoTyrrhenian *kWetWe (expected **ketWe) would naturally have been
> by the surrounding labiovelar enriched numerals *kWel "three" (Etr. ci)
> *penkWe "five", as well as "four"'s own final labial *tW. Why, it would be
> miracle if it ever came out **ketWe under these circumstances!
> >Your protoword for '5' (*kit:u/*kWut:u) develops the meaning '4' in >one
> >branch (Indo-Tyrrhenian) and '6' in another (Uralic). Oritur
> >quaestio, dominus Glen: is there a Steppe branch in which it means >just
> >what it's supposed to mean? Your answer is "Altaic" -- but in >actual
> >you only mention Old Japanese. Let's look at the words >for '5' elsewhere
> >in Altaic: Mongolic *tabun, Tunguzic *tuNga, Korean >tasôt, Old Turkic
> >No good.
> What are you talking about Piotr?? The words are indeed attested in Middle
> Mongolian ta-bun (cf. qur-ban, do"r-ban, etc!!). It is also present in
> Tungusic! Korean has /ta-sot/ for *i"ti" just as it has /yo-sot/ for *ru,
> both with the -sot termination! Of course, Turkic has nothing to do with
> And if you think I'm lying about "five", Gilyak ALSO shows this numeral
> along with surrounding correlations in one long four-numeral streak!
> 4 n@- *nil
> 5 tho- *ïtï
> 6 ngax- *Nu
> 7 ngang- *NaN
> 8 minr *munro
> Let's shake our heads a little and succumb to the reality of genetic
> relationship, 'kay?
> >How about extending the search? If we look farther afield (and I've
> > >checked Nivkh, Samoyed, Yukaghir, Eskimo-Aleutian and
> >Chukchi->Kamchatkan, lest you should think I'm lazy),
> Apparently, you're lazy cuz Nivkh slipped from under you :)
> >I understand your private methodology well enough to see that it's
> > >self-defeating. By saying that it's better to reconstruct ANYTHING
> > >rather than nothing at all you ignore the distinction between a
> >proposal and unfounded speculation.
> Piotr, aside from being highly bias in favor of the mainstream status quo
> complete ignorance and agnostism regarding prehistoric stages of
> reconstruction which only works against advancing knowledge, you also
> misunderstand that this theory isn't the end of things but a tentative
> proposal, as I keep saying. It will adapt as it already has and because of
> this, your statements are only founded for the current stage of this
> which has already changed somewhat... so keep arguing :)
> It truely IS better to reconstruct anything rather than crying
> everytime someone attempts to address a question that needs to be
> Here is an answer. If it is not to your satisfaction then by all means,
> alter it.
> >This is a misunderstanding of scientific methodology. The null
> >must always be as conservative as possible in order to
> >exclude unwarranted inferences. Thus, you can't start by ASSUMING that
> >Steppe families are related and what remains to discovered is just >HOW
> >they are related.
> The Steppe hypothesis is not much of an assumption. I validly judge
> relatedness based on the commonalities of grammatical systems. The Steppe
> languages all exhibit certain specific characteristics that distinguish
> from other languages of the Old/New World.
> >Such a proposition could never be falsified, for it cannot be proved
> >two languages are unrelated: the evidence of their genetic >relationship
> >may have been obliterated by historical change. The >initial null
> >hypothesis must assume their NON-relatedness --only then >can we hope to
> >make some progress by trying to falsify it.
> ...but how do we falsify NON-relatedness, Piotr? That seems frankly like a
> fruitless game in itself.
> >You can replace it with a new null hypothesis only if you manage to
> >demonstrate that all or some of the languages being compared show
> > >systematic resemblances which exceed non-genetic "background
> >(due to borrowing, chance, etc.).
> This already has been provided time and time again by many others before
> via commonalities in grammar and vocabulary as well as set sound
> God, I hope I responded to all the major points, phew...
> - gLeN
> Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com