Re: Is Basque IE?

Yahoo's latest negative improvement fails to reproduce the text of the earlier post.  This is intended to be a reply to message #71289.

*****

 

I do not recognize the advantage of restricting sources to such a small number of books.  Avoiding complexity means evading reality.  I have not seen the original paper or the critical responses, but only the excerpts from academia.edu labeled v01a and v01h.  I will try to keep my own criticism straightforward and brief.

 

Excerpt v01a, p. 5:  Without consulting many more sources, how can you know that your six chosen sources provide 90% of the information you need?  No individual author is infallible.  Restricting yourself to few sources gives you bibliographic tunnel vision.

 

P. 7:  In my opinion, the resemblance between Basque and IE personal pronouns is best explained by assuming that Basque is a Nostratic language, related to PIE but not derived from it, superposed on a Cantabro-Pyrenaean substrate having only one nasal phoneme.  This phoneme was usually reflected as *[n], but when two examples occurred in successive syllables, the first was dissimilated to *[m].  Hence the pronoun *mi 'I, me' (surviving as such in archaic Etruscan, but in PIE incorporated into the verbal system) became Proto-Vasconic *ni, but the noun *menti 'hill, mountain' retained *m...n because the C-P substrate allowed this sequence, and Basque has regularly _mendi_.  (Note the Tarraconian place-name Mentissa/Mentesa/Mentosa.  The Nostratic root is likely identical with PIE *men- 'to project' in Latin _mons_ etc.  Or the place-name is simply of IE origin and no long-range inference is valid here.)  Thus I disagree with Michelena and others about */m/ in Proto-Vasconic.  Michelena derives Bq. _mehe_ 'thin' from *mene, this in turn from *bene.  But other Bq. words like _behe_ 'bottom' and _behi_ 'cow' show no such development *ben- > *men-, so I think *men- was a permissible sequence.

 

I consider Bq. _zu_ (originally plural, 'ye') to be cognate (through Nostratic) with PIE *tu(:) 'thou'.  As *-i marked singular personal pronouns, so *-u marked plural (and PIE *tu: can be explained by monosyllabic lengthening under the accent, without a laryngeal).  At an early stage of PIE (or even before PIE as such), the plural became used as a polite singular (like French _vous_), and eventually a new second-person plural came into use.  But this did not happen in Basque until historical times.  Your scenario is highly implausible because in it, only the branch of IE which became Basque (apparently centum, based on your claim about *k^) retained *tu as a plural until recently, while all others independently made it a singular quite long ago.

 

P. 8:  I must agree that the etymologies by Gorrochategui and Lakarra given in the table are forced and unconvincing.  However, this does not mean that any alternative is necessarily preferable.  Some words simply cannot be plausibly analyzed on the basis of other words with the material which we have available.

 

Excerpt v01h, pp. 20-32, etymologies by number:  1.  Bq. _aho_ 'mouth, face' probably continues ancient *ano (cf. _ahate_ 'duck' < Lat. acc. _anatem_, etc.), so it would go better with Lat. _a:nus_, and I will refrain from the obvious jokes.

 

2.  Bq. _aitzin_ 'frontside; before', _aitzi_ (postp.) 'against', etc., probably are of Celtic origin as suggested in v01a, p. 13.

 

3.  Bq. _aita_ 'father' could be another Celtic loan or an independent creation, depending on what you think of Jakobson's "mama-papa words".

 

5.  Bq. _ar'_ 'male' (with r' denoting the hard rhotic in composition, thus _arra_ with the article), if it came from *anar, should have given *ahar in northern dialects (with soft -r).  If I compared _ar'_ with anything in IE, it would be the masculizing element in OHG _chataro_, NHG _Kater_ 'tomcat' (and a few other masc. animates), though I cannot say which direction this went.

 

6.  Bq. _argi_ 'bright; light (n).' is almost certainly a Celtic loan.

 

7.  The protoform *anets 'breath' should have given *ahets, not _hats_, in northern dialects.

 

9.  Bq. _barre_ 'laughter'.  Apart from dubious semantics here, your very peculiar soundlaw *sw- > *b- is not well grounded, and your high frequency of alternative etymologies shows that the critics are right.  You have too many ways of allegedly deriving a word.

 

10.  Your explanation of Bq. _bederatzi_ 'nine' involves ad-hoc voicing of -t- in _eta_ 'and'.  If this is from "allegro speech" or a similar arbitrary makeshift, why can you provide no other examples?

 

13.  Bq. _bero_ 'hot; heat' < *belo (cf. _bele_ 'raven' < *belle(?), _beltz_ 'black', Aquit. pers. name Belexco, etc.), which you assign to PIE *swer-, does not provide convincing evidence of *sw-

> *b-.

 

15.  The by-form _biga_ 'two' could have been extracted from _bigarren_ 'second', where -ga- is part of the ordinal suffix.  Assigning it to a nt. pl., with the usual _bi_ as fem. pl., is purely arbitrary and does not provide convincing evidence of *dw- > *b- in Basque.

 

I could go on with each entry, but it is getting late.  I note that most of your numbered soundlaws are actually sets of similar but independent soundlaws.  I counted 143 of these.  Combined with your arbitrary treatment of *h, metathesis, voicing, contraction, and the like, you have practically as many soundlaws as you have lexemes, i.e. 201, if not more.  Your semantics can be forced (e.g. #30 'twist' > 'turn' > 'turn back' > 'come back' > 'come') or downright implausible (e.g. #33 '(soft) internal organ' > '(hard) bone').  Serious oversights characterize the following:

 

40.  Bq. _gizon_ 'man' has a combining form _giza-_ (e.g. _gizakote_ 'big hefty fellow', _gizarte_ 'society', _gizatzar'_ 'big man, giant, brute, cad') which you choose to ignore.  Very likely _gizon_ continues an ancient compound (represented by the pers. name Giso:n or Kiso:n in Greek letters) of the simplex _giza_ and _on_ 'good', hence 'good man, bonhomme'.  There is no principled way of deriving _on_ from Lat. _bonus_ or a Romance reflex of it.  Old Latin _dvenos_ shows that the Lat. vocalism is not primary, so your comment under #71 is unmotivated.  The alleged reduction *bo- > *o- is contradicted by _bost_, _bota_, _botz_, etc.

 

77.  Bq. _su_ 'fire' was originally *sut as we see from _sutondo_ 'proximity to fire'.  The divinity Sutugius (CIL 13:164) was perhaps a god of the hearth.

 

78.  It would be astonishing if any speakers replaced their simple noun for 'nose' with 'having smell inside'.  Moreover the alleged protoform *h1ens-h3odoros contains an *-s- not found (for example) in Grk. _énudros_ 'having water inside, holding water'.

 

79.  Bq. _txori_, _xori_, diminutives of _zori_ 'fate, luck', orig. 'bird', if they resemble PIE *ster-, *stor- 'type of bird', simply got lucky.

 

80.  Bq. _ur_ 'water' (with soft -r) has a combining form _un-_, an interesting and important fact which you choose to ignore.

 

96.  Bq. _ama_ 'mother' (according to Trask the first Bq. word with /m/, so he must have had the ability to remember past lives, like Alice Cooper) can be derived from Vulgar Latin *amma, of P-Italic origin (cf. Oscan _Ammaí_ 'to the Mother'), if not simply a Jakobsonian creation.

 

In sum, I find no merit in your theory.  You need to consult a much wider array of sources, stop ignoring details of word-formation, and incorporate material from place-names and personal names (which often yield ancient forms of lexemes unavailable otherwise).  Of course, since you have obviously invested considerable time and personal reputation into this theory, I do not expect my criticism (or that of others) to have any effect.

 

Douglas G. Kilday