Re: Questions about Hungarian

On 5 May 2004 Rob Haden <magwich78@...> wrote:

> Is it possible that the so-called pronominal accusative is really a
> reduced form of the ablative in -tA?
  One of the greatest and recurrent porblematics of the Uralic language
is the distinction of the indefinite and the definite direct object
(cf. accusative vs. partitive object, indefinite vs. definite
conjugation).

  Therefore I think that the pronominal accusative derived rather from
deictic pronoun *tV that could be an emphatic apposition.


> I fail to see how a deverbal noun-forming suffix can become a 1sg
> personal indefinite verbal suffix.
  Cf. also Finnish 3rd person markers that have the same antecedents as
present pasticiple (i.e. a deverbal noun-forming suffix in words like
elävä 'living being, animal'). Morover, -sz- is both an improductive
present marker in words like lesz 'to become, will be', játsz(ik) 'to
play' and a 2sg person marker (e.g. mész 'thou go'), as well as suffix
present participle in Zyrien /lokt1s/ 'coming; kommend', Votiak /potis/
'leaving; abgehend' and for gerund in Vogul /u:nl@s/ 'sitting; Sitzung,
Sitzen'.
  Hungarian past derived from a past participle constuction _without_
copula: you used participle as a predicate without copula, such nominal
predicates are common in FU languages. In this case a verbal noun may
be reinterpreted as a finit verb form and it's ending may be considered
as a personal marker. Especially when a new, just developing
conjugation "looks for" new distinctive markers.
  Note that the Hungarian suffix -t is not just a past particple
marker, present marker but also a "true" (nowadays improductive)
deverbal noun-forming suffix in words like élet 'life; Leben', állat
'animal'. There're no tight boundary between these cathegories.


> Also, Estonian minek does not derive from minema, but both derive from
> the root/stem mine-.  Same with Mordvin, I bet.
  I think you should be a bit more benign when you read my text. I gave
the lexematic form of the verbs (i.e. the form in which they can be
looked up in a vocabulary) as it's usual. But for the indentification
of the verbal stem, I marked the morphematic boundary by "." (cf.
mine.ma, er'a.ms) as it's usual.

  Your objection is not a substantial argument against may statement
that in the above cases suffix *-kk is not just a deverbal noun-forming
suffix, but a suffix for verbal nominals.

  May I propose that our discussion would be focussed in the future on
the mertis and not on the other's notation style, terminology etc?


> According to Sammallahti, Finnish underwent the reverse process: first
> /ď/ merged with /i/, and then medial unstressed /i/ became /e/.
  I think neither me or you will resolve this issue. But I'm still
wonder how could an unstressed vowel in posttonic (i.e. just after a
stressed) syllable became _opener_, i.e. more articulated. I think
Sammallahti supported this fairly problematic phonetic process with a
number of unquestionable examples, I am right?

  Morever, if there was a medial unstressed /i/ > /e/, why do we have
such words in Finnish as etsin 'I search', why do we have such markers
as -i- (imperfect), -isi- (conditional) etc.?


> Perhaps the suffix had become -p and then, due to a change in
> phonotactics (namely, stops cannot exist in word-final position), it
> became -pi.  I still fail to see how final /a/ can become /i/ in some
> cases and remain intact in others.
  And I fail to see why this -a/ä disappeared in this case and remaines
intact in others? You'd like to change an occasional process to another
occasional process. But you suppose the existance of a phonotactically
invalid situation for a while. If this could be really invalid, it
wouldn't occur even for a while: the final vowel would remain.

  How many examples do you know when a full word-final vowel
disappeared deus ex machina? There's a number of cases showing that
there were intermadiate forms during the disappearance. And these
intermadiate forms tend to be less articulated, first become closer
later become reduced (see Mordvin, Hungarian and even French,
Portuguese). These final sounds -a/ä began to disappear and the first
phase was their closing. Finals as -u and -y form an improductive group
in Finnish,therfore they became the most common closed vowel that is
harmonically neutral, i.e. -i.

  And why did they begin this process? It's a tendency in some
dialects, cf. Karelian (Aunus dialect) akku ~ Finnish akka, K. jalgu ~
F. jalka, K. jumalu ~ F. jumala, K. kylmy ~ F. kylmä, K. seinü ~ F.
seinä. The same, but more complete process took place in the case of
Estonian and resulted in the loss of word-final vowels. Dialects
interact and literary languages are dialectal chimeras to some extent.


> According to Ante Aikio et al., Decsy's reconstructions are full of
> errors.
  I suppose Décsy says the very opposite. But my source is not Décsy
but Bereczki, Rédei and Hajdú.

  May I ask you to enumerate the FU languages that (1) have no
harmonically neutral vowel at all, (2) show clear evidences of vowel
opening in posttonic syllables?  The list of these languages will be
long, if there was no neutral vowel in PFU, and the medial syllables
contained close vowels. But if the list will be short, there're not
enough _practical_ arguments for the PFU medial /i, 1/.

> I thought X-SAMPA for /ä/ was /2/.
  X-SAMPA /2/ is for ö.


> I don't know how Sammallahti reconstructs it.  The Hungarian suffixes
> you mention appear to have come from earlier *-nejk and *-nejl,
> respectively.
  From a PFU optative (potential) suffix *-ni/n1 we would expect
*-ni.j.k/n1.j.k, *-ni.j.l/n1.j.l, cf. nyíl 'arrow' < PFU */n'3rk-/ (or
*/n'1rk-/ in case of Sammallahti's inventory) ~ */n'o:rk/-. In these
words Hungarian preserves the velar state of this /i/, e.g. plural
nyil.a.k 'arrows', not +nyil.ek. Therefore from *-ni.j.k/n1.j.k we
would expect harmonic -ník~nék/nák, -níl~nél/nál.
  A contradiction occurs here in Hungarian if we start from FPU
*-ni/n1.