Momentary-Durative

[Jens Elmegaard Rasmussen]

To "petegray" and List,

Thank you, Peter, for a challenging reply. I'll try and do it justice.

On Wed, 14 Jul 1999, petegray wrote:

> [...]
> Yes - but you have made the assumption that this root always shows the nasal
> in the present.   This is not the case.
> (a) Albanian shows two forms, one with, and one without the nasal.
> (b) Old Irish shows no nasal.
> (c) Greek a-tuzomai shows no nasal.
> (d) If Pokorny is right to link the root with stud- then we have no nasal in
> Latin studeo, nor in the various reflexes of it in Germanic.

Sure, verbs can form other derivatives too. Gmc. *staut-i/a- must be an
old intensive (*stu-sto'wd-); studeo is a stative in *-eH1-; atu'zomai 'am
frightened' has a velar in atukhthei's and does not look related; Alb.
shtyj, -yn is no evidence since the nj-present is productive (if the old
form is shtyenj it is hard to understand as related at all). But Old
Irish -tuit 'falls' is *tudeti and identical with tudati. So maybe for
this verb the stem restructuring should be pushed back in to the
prehistory of the protolanguage to give us PIE variant presents - though
palallel development of course must be very common.

>>  There are so many instances of the _same_ verbal root turning up
>> with a nasal present in IE branches that have nothing else in common ...

> And there are very many instances of roots showing nasal presents in one IE
> language, and not in another, or indeed appearing in both forms in the same
> language, as Latin cumbo and cubo, or Greek leipo: and limpano:

Right, and in such cases it is advisable to assume a set of nasal present
and root aorist, since they are so often combined, and are retained
together with so many verbs. Thus Gk. leip-/limp- are inded both presents,
but *ley{kw}-e/o- is a perfectly normal IE root-aorist subjunctive which,
due the closely allied function of that category, often came to compete
with the inherited present.

>>. Is it a coincidence
>> that *k^lew- 'hear' forms a nasal present in Indo-Iranian and Celtic?

> Yes.   It has no nasal in Greek or Latin or some others. (and so on with the
> rest of your list)

Why should it have, if the individual languages went crazy and levelled
almost all differences between the aspect stems?

> I cannot agree with you that certain verbal roots always had nasal presents
> in all IE languages.   This is simply factually untrue.   But without that
> assumption, your argument collapses.

Not "in all IE languages", only in the IE protolanguage, which apparently
had a general (unilateral) solidarity between nasal presents and root
aorists. From this common point of departure the languages developed all
sorts of differences (even closely related lgg. differ in this respect,
oh yes).

>>. Does that not indicate
>> that the assigment of the nasal-infix structure as the present of certain
>> roots was fully lexicalized in the protolanguage?

> No, for the reasons just stated.   There are enough counter-examples to show
> that whatever the conditions were for the "selection" of nasal, full grade
> or suffixed present, they are now as unrecoverable as they are within Latin,
> Greek, or Sanskrit, where very little difference of meaning can be
> consistently shown.

>> Oh no, it works only in the direction that a nasal present has a root
>> aorist beside it in practically all cases.

> Not a surprise, if you ignore all the langauges where this is not the case.
> Likewise, not a surprise, since root aorists were an old form.   The same
> could be said of full grade presents, or suffixed presents.   So the
> statement really is without significant meaning.

No, thematic presents and root-presents typically take the s-aorist (or
suppletion).

>> Also the reduplicated presents
>> and the y-presents generally form root aorists.

> Yes - so your argument (that a root aorist implies a nasal present) is
> really rather weak.   Yet without this argument, you cannot derive tuda'ti
> the way you do.

No again, a root aorist _is implied by_ a nasal present. And since that is
found, there is very good reason to believe that there was a root aorist
in IE. And then, surprise, that offers all the basis you need for a smooth
and natural derivation of the tudati type - smooth at least in comparison
with many stories of verbal stem-formation that can be followed in
attested languages. The story also explains what is perhaps the most
important point of all, which I did not mention in my posting: the
punctual aspect value observed by Renou in the use of the tudati present
type (Louis Renou: Le type ve'dique tuda'ti. Me'langes Vendryes, Paris
1925, 310-16, a classical and oft-quoted philological paper written
completely without comparative or even diachronic bias). - In fact, I am
only saying that tuda-, which is structured like vida-, may have been an
aorist stem just like the latter.

>> [in] PIE, each verb had mostly picked one form for its present. The massive
>> concord among the IE languages in this respect

> There is no such concord.  Even a single language shows variety of
> formations of the present, and the variety across the languages is
> considerable.

Yes, but easily explained as secondary, or not relevant, if the forms
adduced are not from the primary verb, or if they belong to types that got
productive in the separate lives of the individual languages. When that is
subtracted, it is quite strinking how well the old core of the different
branches get to resemble each other. The impression that IE verbal stem
formation just mutates erratically and that no common system is
recoverable is not compatible with current knowledge.

> But it would be good to have firm data on this - anyone got
> three weeks to spare going through the text books?

So you haven't done that? It's quite easy now: LIV (Lexikon
Indogermanischer Verben, by Helmut Rix & al., Wiesbaden 1998) offers all
you need, including reconstructions and developmental tracings.

>> - while the even greater discord is easily
>> explained by continued normal language change.

> No.   Normal language change would not turn a nasal present into a
> reduplicated one or vice versa -

Why not - if they were both equally expected as the companion of a root
aorist? Is that not the way analogy works? Let me give you two
illustrative case stories. The IE status of a reduplicated present from
the two verbs *sed- 'sit down' and *steH2- 'stand up' cannot be questioned
(there is *si-sd-e/o- in Ital., Gk., IIr.; and *sti-steH2- or rather its
old subj. *sti-stH2-e/o- in IIr., Gk., Ital. and Celt.), nor can a root
aorist be avoided for the latter; I accept it for both, of course, but
*sed- as aor. is not so well-preserved. Now, the paradigms *si'-sd-e-ti,
*se'd-t and *sti'-stH2-e-ti *sta'H2-t thus restored had a funny fate in
Balto-Slavic and Germanic: BSl. changed *sisd-e/o- into *sind-e/o-, and
Gmc. back-formed *st at 2-n-t-e/o- from the aor. *staH2-t (drawing the -t- to
the stem, i.e. apparently after an adjustment of *staH2-t to "*staH2t-e"
with the productive ending of the perfect). Why would they do such a
thing? There can be only one answer: because the reduplicated present type
was going out of fashion, while the nasal present was still productive and
still went well with root aorists.

>  -, nor would it turn a -sk suffixed one into a nasal present, etc, etc.

That's right, these do not belong together, and I never said they did.
However, in a few odd cases the whims of chance did end up combining them,
as when Hitt. nasal verbs form durative derivatives with -sk- like all
other verbs do, or Toch. makes itself a prs. formant -na"sk- (-na:sk-),
thereby giving inherited nasal presents a productive present-stem marking.
And if there are both a set "root-aor. + nas.-prs." and an inchoative
"s-aor. + sk-prs.", and the functional difference was lost, you may end up
finding a nasal present combined with an s-aorist, or an sk-prs. with a
root aorist. In some cases one may even suspect that this had happened
before the dissolution of the protolanguage; a prominent case is Gk.
gigno:'sko: with root-aor. e'gno:n, but Gothic kunnai{th} combining nasal
infix with the vocalism (*-n-e:-, not *-n-o:-) of the s-aor. that also
gave the stem of Gmc. *kne:-(j)i/a- 'know', retained with the -s- in Hitt.
ganest 'recognized' (*g^ne:'H3-s-t). In a principled analysis, this
reflects two IE verbs, (1) *g^n.-n¢-H3-ti, aor. *g^no'H3-t; (2)
*g^n.H3-sk^e'-ti, aor. *g^ne:'H3-s-t. The latter set is directly continued
as a synchronic paradigm by Arm. {cv}ana{cv}'em, aor. caneay as seen by
Jasanoff.

The most rewarding experience during the time I have been watching
Indo-European Studies has been to see the protolanguage come alive and
assume an increasingly well-established structure, with archaisms and
productive patterns, just like the living languages we know. I react in
defense of the field when I see somebody turning the clock back to a stage
we have left behind.

Jens