Phaleran 'screeves' [was Re: Agglutinating -> inflecting]

Quoting Peter Bleackley <Peter.Bleackley@...>:

> Has anyone evolved an inflecting conlang from an agglutinating one? If so,
> how did you go about it?
This is as good a time as any to review Phaleran verb inflection.
Phaleran has three screeves: prospective, progressive, perfective.
(I borrow this term screeve from Georgian, since a screeve is a
uniform morpheme class whose members will not necessarily be tenses
or aspects, but may have qualities of both.)  These are
historically, and sometimes synchronically, agglutinations of the
tense/aspect marker plus person markers and irrealis mode (if
present).  Frequently, these combinations violate various principles
of syllable structure and adjacency constraints, and are thus
simplified in some way:

              Pros (*l)   Prog (*s)   Perf (*n)
1 sg (*t)      -lt-          -st-        -nd-
1 pl-in (*kh)  -lkh-         -skh-       -nkh-
1 pl-ex (*þ)   -lþ-          -sf-        -n-

2 sg (*š)/S/   -lš-          -šš-        -ntš-
2 pl (*l)      -l-           -sl-        -ndr-

3 sg-prox (*s) -ls-          -ss-        -nt-
3 sg-obv (*n)  -ln-          -sn-        -nn-
3 pl (*m)      -lm-          -sm-        -mm-

These here are straightforwardly the conflation of some tense/aspect
marker plus the reduced form of bound pronouns.  It gets trickier
though in the irrealis paradigms, which are considerably more
inflectional when combined with preceding tense/aspect and person:

              Pros (*l)   Prog (*s)   Perf (*n)
1 sg (*t)      -lta-         -sta-       -nta-
1 pl-in (*kh)  -lkta-        -skha-      -nkta-
1 pl-ex (*þ)   -lta-,        -sta-       -ntþa-
               -lsta-

2 sg (*š)/S/   -lšta-        -šta-       -ntša-
2 pl (*l)      -lta-         -V:lta-     -ntla-

3 sg-prox (*s) -lsta-        -V:sta-     -ntha-
3 sg-obv (*n)  -ltna-        -V:nta-     -nta-
3 pl (*m)      -V:lta-       -V:mta-     -ntma-

(These markings are placed immediately after thematic markers
(originally derivational transitivity marking) and before
evidentials:

  tha-ri-lsta-n
  be-INTR-3SgProxIr-Quot
  "...that he (prox) be X")

There are two general facts that bring us to these facts.
First, Phaleran allows glides, liquids, nasals, /s/ and /š/
to be codas, but not stops or other fricatives, and no
complex codas are allowed.  There are much looser conditions
on what may be a valid onset:  in principle, any singleton
consonant may be an onset, and in clusters no rise in sonority
is required; glottalized segments may not be clustered with
other segments, and other clusters must be harmonic in VOT
(both aspirated, or both unaspirated). Thus, sequences like
-l-t-ta- or -s-kh-ta- which create illicit clusters may be
solved by deletion with compensatory lengthening or, in some
other cases (l-n-ta-) by metathesis (> -ltna-).

The second salient fact is that the above system, the standard one,
represents dialect mixing, where some dialects favor metathesis
of consonants to avoid deletion as in prospective irrealis,
while others favor deletion of illicit sequence and lengthen the
preceding vowel, as in the progressive irrealis.  In practice,
leveling tends to occur in the direction of the latter possibility,
by deletion with compensatory lengthening, among nonstandard
dialects of the Twolyeo region.

 =========================================================================
Thomas Wier            "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637