Re: The map of typological features


"Darkstar" <darkstar100@xxxxxxxx> wrote ...
Douglas G. Kilday wrote:
"Darkstar" <darkstar100@xxxxxxxx> wrote ...
Brian M. Scott wrote:
Darkstar <darkstar100@xxxxxxxx> wrote:

[...]
Are they [reduplications] unique to Greek or can you find
something
similar
in many IE languages? Probably not.

On the contrary, it's an inheritance from PIE. It's present
in Sanskrit, and there are at least traces in Latin, Gothic,
and Old Irish, to name just the ones that I'm sure of. 'In
so far as they have a perfect, reduplication is the general
rule in the historical [IE] languages.' (Szemerényi,
Introduction to Indo-European Linguistics, §9.4.3)

The reduplications of perfect in PIE played a minor role parallel to
that of the Ablaut and personal endings. In Latin they are
rudimentory.
It's a unique and rare example, while, say, in many African languages
they are still the main tool to form plurals. Besides, in Greek,
reduplications are often unstable in vowels ("didomi",
"dedoka").

The Latin perfect <fefelli:> could not have been inherited from PIE or
Proto-Italic.

Why not?

Brian already explained why not for PIE. For Proto-Italic, we may note that
Osco-Umbrian does have intervocalic -f- from PIE *bh, *dh, and *gWh in
intervocalic position, but Latin has -b-, -d-, and -v-. Even if late
Proto-Italic had -f- from PIE *bh, this became -b- in Latin. We may also
note that Latin <Tiberius> is borrowed from archaic Etruscan <Thihvarie>
(with <hv> = /f/), so the formation of the Old Latin perfect *fefallei is
not likely to antedate the mid-7th cent. BCE; otherwise it would have become
*feballei, Lat. *febelli:.

The only way to explain the -f- is to posit the formation of
*fefallei in Old Latin. That means the reduplication, far from being
rudimentary, was still productive then.

Greek <dido:mi> and <dedo:ka> do not illustrate "unstable reduplication"
at
all. They are regular reflexes of two distinct types of reduplication
inherited from PIE. Perfect stems with the reduplication Ce- have
already
been discussed. Present stems with the reduplication Ci- are typically
associated with verb-roots originally ending in laryngeals. They are
frequent in Greek, and are also found in Latin (<bibo:> from *peH3-, cf.
<po:tus>; <gigno:> from *g^enH3-; <sero:> (earlier *siso:) from *seH1-)
and
Sanskrit (<pibati>, root <pa:->; <jighrati>, root <ghra:->).

My error: <gigno:> is from *g^enH2-; it goes with <gna:tus> not <gno:tus>,
of course.

Any other convincing examples of clear-cut reduplications coming from
laryngeals?

The reduplications of the form Ci- don't "come from laryngeals", but most of
the well-known ones are formed on roots ending in laryngeals in PIE. Other
examples are Greek <tithe:mi> from *dheH1-; Latin <sisto:>, Greek <histe:mi>
from *steH2-; Grk. <hie:mi> for *yiye:mi from *yeH1-; Grk. <pi(m)ple:mi>
from *pleH1-. The non-laryngeal root *sed- has Latin <si:do:> for *sisdo:
(and Umbrian 3sg. impv. <sistu> for *sisdeto:d, Lat. <si:dito:>). In most
of these the simple root has a stative sense, while the Ci-form is an
imperfective causative. The predominance of root-final laryngeals may
indicate that these represent reflexive (or mediopassive or the like)
suffixes in an early stage of PIE, but *sed- shows that such suffixation was
not the only way to produce statives.

In fact, I don't see any obvious reduplications here. (Ce- in one case,
and Ci- in the other, and even one imaginary proto-form *siso). I have
too many questions.

The initial consonant is reduplicated, not the entire stem.

Could "sero" and "gigno" be just epenthetic developments from *seo and
*gino or direct renderings of laryngeals via different phonemes?

Very implausible. If you want to argue that, you need to find other
examples that illustrate a clear pattern of epenthetic /s/ and /g/ under
specified conditions, or whatever it is you mean by "direct renderings of
laryngeals via different phonemes" under specified conditions.

Why Praesens only, why not other forms if laryngeals belonged to the root?

I don't know, but I have outlined a somewhat plausible guess above. The
Praesens represents an old infectum, if we believe that most of the
surviving tense-marking was originally aspect-marking.

Maybe "bibo", "pibati" are developed from infant talk in PIE, like
universal "mama", "papa", "bebe"?

Those aren't universal (see Georgian for example). In any case 'drink'
isn't a mama-papa word in the Jakobsonian sense.

Reduplication was also used in PIE noun-formation; an example with
reflexes
in Greek, Sanskrit, Italic, and Germanic is *kWe-kWlo- 'circle, wheel,
people dwelling round'. It's a good time for you to stop trying to
shoehorn
PIE into a preconceived morphotypology.

[...]

ASFAIR, tones in Greek followed the position of the
dynamic stress. Correct me if I'm forgetting something.

Until the end of the classical period the Greek accent was
described in terms of pitch, not stress. The shift from a
pitch accent to a stress accent seems to have taken place
with the development of Koine; see, e.g., Geoffrey Horrocks'
survey of the history of the language.

Okay, but for the 40th time. Was pitch in classical Greek completely
independent or did it depend on the position of the syllabic stress?
Was it phonological and how many minimal pairs do you know? Someone
enlighten me.

Any introductory textbook will give you the "rules of Greek accents".
One minimal pair is <dêmos> 'common people' vs. <de:mós> 'fat'.

So on which syllable does the accent fall? Could that (under any
circumstamnces) be prounced as <dE:mos> and <de:mOs>/<demOs>?

In the 'people'-word on the first syllable; in the 'fat'-word on the second.
The vowel written as Eta was probably [E:] or [Æ:]. Its length is known
from scansion. Its quality as an open vowel is suggested by comments in
Plato's Cratylus to the effect that Eta and Alpha were the "biggest" sounds.

Many verbs have
pairs, and some have triplets, of forms distinguished only by accent, as
any
introductory textbook will tell you.

By accent or by pitch? If Greek had nothing but pitch accent where the
pitch always goes with the accent, that would be equivalent of not
having any pitch at all.

I don't understand what you mean. The scholarly consensus is that most
dialects of ancient Greek (Boeotian may be an exception) had pitch-accent,
and so if forms (for concreteness, those in Attic) are "distinguished only
by accent", most of us regard that as equivalent to "distinguished by
pitch".

Neither, I fully understand the difference between "pitch" and "tone",
and "stress" and "accent". In Russian, there's no difference between
these words.

I'm sorry if I made matters more confusing. The important thing here is the
reconstruction of accent in late PIE, and I think most scholars would
consider pitch much more plausible than stress. For the accent in earlier
stages of PIE (and other matters pertaining to the history of PIE) one is
less likely to find consensus.



.

Relevant Pages

  • Re: The map of typological features
    ... it's an inheritance from PIE. ... That means the reduplication, far from being ... Until the end of the classical period the Greek accent was ... described in terms of pitch, ...
    (sci.lang)
  • Re: The map of typological features
    ... it's an inheritance from PIE. ... In Latin they are rudimentory. ... That means the reduplication, far from being ... Until the end of the classical period the Greek accent was ...
    (sci.lang)
  • Re: The monumental stupidity of PIE theorists further illustrated
    ... chimerical "Laryngeals" in it. ... Saussure, even before Hittite was discovered, his reconstructed PIE ... had laryngeals since IE languages seemed to have reflexes of them. ... presentation of establishment PIE theoy (restricted to Sanskirt Greek ...
    (sci.lang)
  • Re: open letter to the Google company, on the value of the scientific groups
    ... not considered terribly interesting by students of PIE. ... reflexes of vanished "laryngeals" just about everywhere. ... state of affairs, even just in respect of the velars, which ...
    (sci.lang)
  • Re: The monumental stupidity of PIE theorists further illustrated
    ... chimerical "Laryngeals" in it. ... Saussure, even before Hittite was discovered, his reconstructed PIE ... had laryngeals since IE languages seemed to have reflexes of them. ... presentation of establishment PIE theoy (restricted to Sanskirt Greek ...
    (sci.lang)