Re: polysynthetic languages

On Saturday, September 20, 2003, at 05:22 , Nik Taylor wrote:

> John Cowan wrote:
>>
>>
[snip]
>> Oh really?  Which five morphemes merged to make Latin -o: represent
>> 1sg pres indic act?  Even if you want to say that the last three are
>> zero.
Hang on!  -o: occurs as 1st person singular active marker in the _future_
tense of 1st & 2nd conj. verbs, and in the _future perfect_ of all active
(i.e.
not passive or deponent verbs).  So it does not denote present tense.

Certainly -o:, as a 1st sing marker, is suffixed only to indicative forms;
  but the converse is not true, i.e. if it's indicative and active the 1st
pers,
singular must end in -o:.

Many active indicative tenses form the 1st singulars with -m (all imperfect
indicatives, all pluperfect indicatives, and the future indicatives of 3rd
& 4th
conjugations). But, just to complicate things, we find that all subjunctive
tenses use -m.

If the morpheme -m is used to denote 1st sing. for all subjunctive and some
indicative tenses, we cannot, I think, include 'mood indicator' as one of
the
meanings fused into -m.  'Mood' must be indicated in some other way.

If mood is denoted in some other way (and it is), then we should no more
bundle it in with -o: than we do with -m.  All we can say is that in the
indicative mood, some tenses form 1st sing. with -o: and others with -m;
nor is the distribution the same form all verbs.  Two very common verbs,
e.g. form their present indicatives with -m
sum "I am", possum "I can",
and so does one less common irregular, 'inquam' "I say".

We know historically that the Latin passive forms are a secondary formation
within Latin itself and are not derived from IE.  Therefore it is
misleading
IMO to talk about some 'indicative marker' being fused in -o:   There was
no
passive to distinguish the indicative from.  When the passive was
developed,
it developed a new set of subject morpheme endings.  The passive suffixes
are
marked for voice, the active ones unmarked.

The only morphemes fused are two: first person + singular

> Well, okay, so they're presumably not always historically separate
> morphemes (altho, I don't actually know much about IE historical
> linguistics),
My knowledge is a bit rusty and, probably, now behind modern theories;
but my recollection is that -o: is from an earlier -o: + -m.   But it
ain't as simple as that.  PIE, IIRC, had a system of thematic and athematic
verbs and a set of subject endings for 'primary' (i.e. non-past) tenses and
another for secondary (past or remote) forms, thus there were four sets of
each.  Latin has simplified and recast all this.  So the original separate
'proto-morphemes' of PIE or pre-PIE will not always be directly relevant
to the way the morpheme fusion of PIE survives in Latin.

> but at any rate, whether historically separate or always a
> single morpheme isn't the important part.  The important part is that
> we're dealing with affixes.
Indeed - but I think you are basically right in that the flexions of
languages
like Latin did develop diachronically from more obviously agglutinating
structures.

> Whether a language indicates "1st person
> present indicative active" with, say, 4 distinct morphemes or a single
> morpheme is less important than whether those morphemes can occur
> independently of the verb.  You can't say just "o:" in Latin, for
> example, but in English you can say "I do" (which is the closest English
> equivalent of that meaning).
and 'ydw' in welsh   :)
--------------------------------------------------------------------------------------------------
To return to the 'mood indicator' I referred to above.

Several years ago I wrote a Prolog program to parse Latin verbs, and it
successfully parsed all verbs, _including irregulars_, with the sole
exception
of the 'deponent verbs' (those with passive forms but active meanings);
obviously, with a bit of fine tuning it would've handled those as well.  So
I got to know the Latin verb pretty well (actually, I already knew it
fairly
well even before then).

Basically, a Latin verb can be analyzed into three main parts:
the base, e.g. curr- (infectum)
the 'tense sign', e.g. -e:ba- (imperfect indicative)
the personal ending, e.g. -m (1st singular)

Thus 'curre:bam' = I was running

The absence of the tense sign, i.e. zero tense sign, marks the present
indicative (active & passive) and also the perfect indicative (active
only).

This, I hasten to add, is a  simplification of the actual system.  But
the analysis of the formative morphemes in a flexional language does
require considerable care.  An agglutinating lang makes analysis
much easier  :)

Ray
===============================================
ray.brown@freeuk.com    (home)
raymond.brown@kingston-college.ac.uk   (work)
http://home.freeuk.com/ray.brown
===============================================

Replies