On 01 May 2004 Trebor Jung <[log in to unmask]> wrote:

> Why do vowels get lengthened when suffixes are added (fa 'tree' -> fa'k
> 'trees')?

  The vowel was originally long but in a previous period of language
history every word-final long á's and é's shortened. However, before
suffixes, they were not in word-final position, therefore they remained

> What are the vowel harmony rules?

  In terms of harmony rules, vowels can be back ("velar"), front
("palatal") and neutral: /a/, /o/, /u/ are always velar, /2/ and /y/
and always palatal, while /i/ and /e/ are neutral (but sometimes the
behave as palatal ones). Neutral vowels may came from a previous /1/
or from an /aj/, /oj/, /uj/ diphtongue.

  Suffixes can be homomorphic, dimorphic or trimorphic. The harmony
doesn't affect the homomorphs. Dimorphs have a variant to be added to
velar words and another to be added to palatal ones. Trimorphs has a
velar variant and two palatal ones: the one is added to illabial, the
other to the labial stems. (The labiality of the stem depends on the
labiality of the last stem vowel [no matter is it palatal or not].)

  The harmonic group (velar or palatal) of a word is coded in its last
non-neutral vowel, e.g. "balek" 'dupe' is velar word because "e" is
neutral, therefore its last non-neutral vowel is the back "a". Only the
last element is taken into consideration in case of counpound words.

  But there's a number of special subgroups:
- About 60 stems have neutral vowels only but they belong to the velar
  harmonic group (i.e. "híd" 'bridge', "cél" 'aim, target').
- In type "tiszta" 'clean, clear', "ritka" 'rare' etc. final "-a" is
  dropped before some suffixes, but the velar harmony is retained, e.g.
  "tiszt-a" > "tiszt-ul" 'to became clean(er)/clear(er)';
- In a number of mixed words, final "e" behaves palatal, e.g. "október-
  ben" 'in October';
- A number of mixed words having final "e" can belong to both
  harmonic groups, e.g. "konkrét-an/en";
- Word "férfi" is generally a velar (because it's compound "fér[j].fi"
  'husband+boy/son') word but before certain suffixes it can be treated
  as palatal, too.

> Why have enni and inni collapsed 2pl and 3pl (esznek, isznak)? (Altho
> that "collapsing" might just be an illusion of my notes.) What are
> some other /n/ > /s/ verbs?

  They form a small subgroup: lesz, tesz, vesz, hisz, visz, eszik,
iszik. However, I don't understand what you mean by "collapsing". For
details cf. my posting on 28 Apr 2004 07:11:06 +0100 with the subject
"Re: NATLANG: Humagrian glossary".

> Are there any other bipersonal suffixes like -lek?

  No, this is the only one. Cf. my posting on 26 Apr 2004 10:02:08
+0200 with the subject "Re: Hungarian tense, aspect, mood...".

> P.S. How do the modern languages compare with PFU? What was PFU's
> phonemic inventory? How many cases did it have?

  Hmmm. It's the topic of several university semesters... The
reconstructed vocabulary of Proto-Uralic (PU) and PFU is about 1000-
1200 roots. In present-day Hungarian the frequency of the continuants
of these roots is 54% (probably a plenty of the 18% of the uncertain
etymologies, and some of the 15% of unknown etymologies belongs to PFU
herity). It's a quite big ratio as compared to e.g. IE herity of the
Slavic languages. The grammatical structure is still very homogeous
throughout the modern FU (and Uralis) languages. However PFU was
dissolved about 4000 year ago and this time gap made considerable
differences between the PFU branches.
  A vocabulary concordance example: The cardinals '1-6' and '100' are
common in FU languages (except Hungarian that changed '1'; '100' is
Aryan borrowing), but the structure of '8-9' is same in fact that the
elements don't compare: 'two-minus-ten, one-minus-ten'.

  PFU's reconstructed phonemes were the folowing (* denotes uncertain
ones, they were probably dialectal allophones):
 Consonants: /p/, /pp/ /t/, /tt/, /k/, /kk/, /m/, /n/, /n'/, /N/, /tS/,
/ts'/, /w/, */D/, */D'/, /j/, */G/, /s/, /S/, /s'/, /l/, /l'/, /r/.
 First syllable vowels: /a/, /o/, /u/, /E/, /e/, /i/, */3/, maybe */y/.
 Second syllable vowels: /a/, /E/, /e/.
Vowels had no quantity but there was vowel harmony (/e/ in second
syllable was neutral, /a/, /o/, /u/, */3/ were back vowels, the rest
was front one).

  Roots had structure (C)VC(C)V(CV) -- a few of the were trisyllabic,
they were early derivations -- plus grammatical words (pronouns,
particles) with structure (C)V.

  Cases: nominative: zero suffix, Accusative I: /m/, *Accusative II:
/t/, Genitive: /n/, Locative I: /na/ ~ /nE/, *Locative II: /t/,
Ablative: /ta/ ~ /tE/, Lative-prolative I: /k/, Lative-prolative II:
/j/, Lative-dative: /n'/.    (Prolative: along something)
  Those marked by an asterisk are uncertain. Variants marked by I and
II resp. are dialectal, synchronic or morhosyntactical variations (e.g.
Bereczki thinks that separate accusative marker were used on nouns-
adjectives [type I] and pronouns [type II]) Accusative and genitive
were marked only on definite words, indefinite ones had zero suffix.

> What sort of changes did Finnish, Hungarian, Estonian etc. go through
> from PFU?

  I think this can't be aswered in a single posting. Those who read
Hungarian can consult with the abridged topics of the seminar "Finno-
Ugric elements of the Hungarian language" at the Budapest University
  But if you have more specific questions, I'll try to answer them.