Using Dictionaries (was Re: Greek question (night?))

[Robert Whiting]

On Tue, 16 Mar 1999 X99Lynx at aol.com wrote:

> In a message dated 3/15/99 5:22:43 AM, Bob Whiting wrote:
>
> <<In many instances the nominative singular is the
> linguistically least-marked form.  Many times features of the
> root that are important for comparative work are suppressed in
> the nominative singular because of the phonotactics of the
> language, but will still be found in the stem (the form to which
> other case endings are added). >>
>
> I think I should explain that I was well aware that languages that have
> dropped case systems and case endings generally don't use the nominative
> as the surviving form.  (The example I had in front of me was Latin
> "infans"(nom)  > Old French "enfes" (nom.) but "infantem" (accusative)>
> Mod French "enfant.")

This is a different principle, but is part of the explanation of why
part of the stem that is not found in the nom. sing. of the parent
language may turn up in the daughter languages.

> If you look at my original post, however, I was stumbling towards a
> slightly different question.

Yes, and I must admit that I was not responding to your question,
but taking the opportunity to make a different but not entirely
unrelated point.  But I must also admit that part of the reason
that I didn't respond to your question was that I found it difficult
to understand what was being asked.

As I remember your question, it was "Does it matter that /t/ does not
appear in most of the nom. singulars? ... Why is the /t/ not regarded as
just a fairly common stem that sometimes emerges in root and sometimes
doesn't?"

The answer to the first part is no, as long as we know why it doesn't
appear.  To be able to answer the second part, I need to know how you are
using "stem"  and "root."  This problem is not necessarily your fault, but
'stems' from the fact that different linguists use the terms in various
ways.

> What I asked was why the nominative singular (in those languages in
> which they occur) was not in considered in reconstruction.

The answer is that they are.  The fact that the segment in question
has disappeared in the nominative singular does not affect the
reconstruction as long as the reason is known.  The problem is that
most dictionaries don't give the reason for the disappearance.  But
if you don't see it in the dictionary entry you don't have any right
to assume that it wasn't (isn't) there.

> The answer that both you and our moderator have given is that they are
> I guess "truncated" forms, not revealing or mutating the stems. (Part
> of the "phonotactics" - interesting word.)

"Truncated" is not really the right term.  The disappearance of /t/ in
this position in both Greek /nuks/ and Latin /noks/ is simply the result
of a phonotactical rule:

     A dental stop + s is assimilated [note assimilated from ad+similated]
     to ss, which is further simplified to s after a consonant, long
     vowel, or diphthong, and when final in both Greek and Latin.
         C. D. Buck, _Comparative Grammar of Greek and Latin_, p. 145

So the reconstructed forms of the Greek and Latin nominatives are
simply */nukt-s/ and */nokt-s/, respectively, with */ts/ > */ss/ > /s/
by a normal phonological rule.  This makes more sense than assuming
that the nominative is the "basic" form and the genitive et cetera
have some miscellaneous garbage thrown it that has nothing to do with
the root.

> As you say, the nominative is "linguistically the least marked form."

Not "is" but "in many instances is."  There is no rule that says it has
to be.  There is rather a tendency for languages to use the least-marked
form for the most frequently used form.  This is a simple rule of economy.

> But that was of course the reason I choose the nominative singular,
> precisely because it was the least marked form.  I don't think its that
> hard to see why one might naturally go to the form stripped even of
> stems to get to the elemental form of the word.  Not that one should,
> but that one might.

Again, your use of "stems" here is not transparent.

> After all, to say that only the "ablative" form survived is to suggest
> that the word now carries the additional grammatical baggage of the
> ablative.

Sorry, but huh?

> Or, to go further, the notion that some stems might be nothing more than
> phonetically dictated walls, separating sounds in the roots that cannot
> be adjacent to the sounds of regular case endings.  They could even
> survive as such in daughter languages where the endings no longer
> exists.  Or that the stem served no other function than to signal the
> user that a particular word followed a peculiar or dialectical
> case-ending system.

Ah, are you using "stem" in the sense of "root consonant" or "root
augment/extension"?

> In that light, I hope the question why the nominatives were not
> considered in the reconstruction of the word doesn't sound that
> ludicrous.
>
> I wasn't for a moment doubting the reconstruction (of "night.")  And the
> explanation of word-ending rules is good enough for me.  (But I wonder
> why e.g. Latin couldn't have solved the problem the way it did in
> forming the rather regular-looking adjective - nocturnus.)

The phonotactical rules that produce <nox> are not operative for
<nocturnus>.  It's not a question of "solving the problem"; it's a matter
of what happens in a language when certain sounds (or classes of sounds)
come in contact.  The answer wasn't worked out by a committee or an
academy.  It is a matter of a tradeoff between ease of articulation
and the level of morphological differentiation needed to disambiguate
meaning that the users of the language resolve with even thinking about
it.

> So I'll take a chance and ask the question anyway - can a stem ever be
> seen as something like a vestigal case ending in reconstructing PIE -
> not part of the original word, but a compounded form that produced a
> universal "stem" in the daughter languages?

If by "stem" you mean "root augment" this is a very complex question that
has been discussed extensively without reaching any particular conclusion.
There is an extensive literature on this, not only for PIE but for other
language families as well.  The problem is that "root extensions" don't
behave mathematically.

> <<Thus using a dictionary without being aware of the phonotactical rules
> of the language is a recipe for disaster when doing comparative work.>>
>
> Well actually the dictionary did give me some nominative singulars -
> which is what my question was about.  And as far as the languages go,
> I'm still not bad at Greek or Latin, though I've forgotten a bit.  And
> (but only if you insist of course) I can give a number of instances
> where prominent native speakers of both tongues expressed the notion
> that the nominative was "the true form of words."

Prominent native speakers of Greek and Latin knew more about Greek and
Latin than any of us ever will.  But they didn't know much about
historical linguistics.  Have a read through the Cratylus to get an
idea of the level of historical linguistics of the time.

> By the way probably just as serious a problem with dictionaries appears
> to be on the "semantic" side.  3000 year old words are often defined
> with a precision that - if you know your Homer, so to speak - can be as
> perilous as assuming that the nominative is worth talking about.

When you get right down to it, nobody really knew his Homer except Homer.
Contrary to popular opinion, dictionaries do not define words (Academies
do that, or try to) but only record usage (or at least they are supposed
to).  As A. Leo Oppenheim (for many years Editor-in-Chief of the Chicago
Assyrian Dictionary) was often heard to mutter "I know what it says, but
what does it *mean*?"

Bob Whiting
whiting at cc.helsinki.fi