From: "H.M. Hubey" <HubeyH@MAIL.MONTCLAIR.EDU>

More belaboring the obvious.
Turkic Languages, Johanson and Csato.

p. 42

Various thematic stems can also combine with copulative verbs such as er
(be) and bol (become) e.g.
Old Turkic -(V)r bol (become doing), signalling the
transition to an intraterminal state (see pp. 43-4),
Turkish -mIsh ol- (become having done) signalling
the transition into a post-terminal state (seep. 44)

p. 43

With intraterminal items such as presents and imperfects, the event is
envisaged within its limits, that is after its beginning and before
its end. Some are more focal, putting a narrower focus on what is
currently going on at the orientation point, sometimes in the sense of
English progressives, e.g.

Uzbek kelayatir "(S)he is just coming",
Noghay Barayetir "(S)he is just going"...
Turkish okumaktaydim,
Kirghiz oquudamIn" (I am reading).

...

(Mine: Obviously Turkish okumaktaydIm is "okumakta idim" where the
"idim" is from *erdim". Uzbek kelayetir is *keley edir", and this
"edir" would have been "erdir".)

There are corresponding past items, more or less focal perfects such as
Turkmen Iyyaerdim "I was just eating",
Azerbaijanian AlIrdI (<alIr idi).... Kumuk Bara edim "I was going, went" .
(Mine: In Kumuk it is still crystal clear.
In Old Turkic it would have been "bara erdim".)

p. 45

Turkic languages also possess indirective categories, certain kinds
of evidential items used to qualify the experience of the event spoken
about. The expression of this epistemic modification varies accross
languages. Post-terminals such as -mIsh, and -(V)ptIr tend to get
indirective interpretations as 'subjective pasts' e.g.
Turkish Ali gelmish "Ali has [reportedly, apparently, obviously] come'.

Besides these deverbal past tense devices, there are also
tense-indifferent indirective copula particles of the type imish, eken
which combine with nominal stems, e.g.
Turkish imish/(y)mIsh, e.g. Ali geliyormush
"Ali is/was [reportedly, apparently, obviously] coming".

The enclitic copula particles, developed from forms of obsolete verb
er- (be), have already been mentioned. Past markers of the type
edi "was" located the thematic perspectives temporally
e.g. Turkish gelmishti "(S)he had come".
The role of indirective copula particles such as emish, and eken have
just been mentioned.

p. 61

Relative clauses, nominal actor clauses, refer to participant entities
to which actions are ascribed, and are based on participles. They can
be used as the head of a nominal phrase e.g. Turkish gelen
'the one coming/having come'; Kirgiz aythqanI 'what (s)he has said'.
(Mine: notice the intrusive K. Recall chapxan vs chapan. Either the K
was introduced or was lost. Still looking for the smoking gun, but
leaning heavily towards its intrusion.)

The participial markers used,
e.g. Turkish -(y)an, -DIK, -(y)AcAK, -mIsh, -mIsh olan,
Uzbek -Gan, -(a)yatgan, ....
Tuvan -(V)r, -GAn, are relative subjunctors comparable to
English relative pronouns and adverbs. They also signal various
aspecto-temporal meanings.

There are thus post-terminal participles as in
Kazakh kelgen kishi,
Chuvash kilne sin "a/the person who has arrived',
Tuvan ushqan iyash 'a/the fallen tree'

Some subjunctors have a more general terminal use e.g.
Turkish and Azerbaijanian markers -(y)An and -DIK.
Predicates such as gOren, and gOrdUGU can thus be rendered
as 'who sees/saw/has seen' and 'whom [someone] sees/saw/has seen'.
[Now some ticklers from Old Turkic by Marcel Erdal]

p. 148

Interrogative adverbs are formed form the root *qa e.g. qanta "where",
Orkhon Turkic qantan,
Uyghur qantIran "from where", qancha "how much, how far, where to", qanI
"where", the latter used in rhetorical questions.
[Mine: compare these to Latin again. Also compare forms like
qanday=qalay. The obvious answer is
*qanthadh > qanday, *qanthadh > *qanlay> qalay.
There are others which are also easily derivable e.g.
*qandha > qandha > qayra/I . Notice the -nth-]

p. 144

The reflexive pronoun kentU 'own, one's self' takes, number, possessive
and case suffixes. The interrogative qanyu >qayu "which' is formed
from a root *qa. [Mine: see directly above. Also note that
Kbal has kesi,
Turkish kendi.
That again implies exactly what I wrote above.Note the -nt-]

p 146

The form in -mIsh also combines with the copula erUr,
negated ermez e.g. AltI yUz
tUmen yIl ertmish erUr "Six million years have passed".
There are also analytical forms consisting of the
aorist, -mIsh, -dI, -daCI, or -GA(y) followed
by forms of er 'be': erdi for the anteriority, ergey for prospectivity,
and ermish for inferential past.
Uyghur bol 'become' instead of er- marks the action as transformative.

p. 152

With nominal predicates,
er 'be' expresses 'being' and bol 'becoming', bar 'there is' denotes
existence and yoq 'there is not 'non-existence. The predicate often
includes a pronoun. There are also sentences without a copula.
The copulative elemetns ol and erUr are frequently used. 

The conditional erser serves as a conditional junctor, e.g. Ud
ertUrUrler erser 'if they let time pass',
bolmadIlar erser 'if they have not become'.
[Mine: I have to point out the obvious again.
Turkish olmadIlar ise "if they have not become".
The middle-r is lost and the final r is lost, and the inital b is
lost in the other word.

In Kbal bolmadIla ese; the final-r all lost. Now something even more
interesting. Look at the word ertUrUrler. That root shows up in
Old Turkic in many places as o"d.
In Kbal o"t. e.g o"tdu".

Turkish has confused the crowing of roosters (natural time clocks)
with the passing of time (natural enough if you look at other languages)
and now uses the word to mean "singing of birds" whereas
the "passing" especially with respect to time is still used in
other Turkic languages. Furthermore this word o"dh also shows up
being connected with "sky, moon, sun" as it should be. I wrote
about these before many times and in various languages including
Turkic very recently.

Conclusions:

Very obvious.

1. bar (to exist), bol (to become) [ol in Turkish] and er (be) are
all from the same root. It should be obvious by now.

2. The root obviously was of the type *bVth, which gave rise to bol,
and the change *bVth>*bVdh then gave rise to bar. Further loss of the
initial b gave rise to *eth or *edh. Thus the particles such as
ese (erse), eken (edhken), idi (edi<erdi or *edhdi) are quite easily
explicable in terms of both semantics and phonetics. Hence it looks
like we have constructed a verb that goes beyond *PIE.
Was Tokharian a centum [kentum] language? Obviously, I checked a
long time ago :-)

H.M. Hubey wrote:

> One more time to belabor the obvious.
> Really striking similarities exist (AFAIK) between Sanskrit and the
> classical European languages like ancient Greek or Latin. For example,
> Sanskrit has an optative mode and an aorist tense - like Greek, but
> unlike most other European languages (I don't know about Farsi). Then,
> I once found a Sanskrit conjugation table giving some forms of the
> verb "ni", "to lead" - the present has the forms "nayami", "nayasi",
> "nayati" (...) "nayamah", "nayatha", "nayanti". This reminds me
> stronlgy of Latin "eram", "eras", "erat", "eramus", "eratis", "erant".
> These forms aren't present forms, I know, but the principle is the same.
>

> Change Latin forms
> eram > erma
> eras > ersa
> erat > erta

> Is it clear now?

> H.M. Hubey wrote: