Jeff Jones writes:
 > Hi Tim, comments and questions below
 > On Tue, 25 Jun 2002 00:13:43 +0100, Tim May <[log in to unmask]> wrote:

[note that I've snipped various stuff all over the place]

 > >ICED aspect
 > >
 > > If this infix is applied, the word describes a change in the state of
 > > the patient. It may be inceptive (entering the state), cessative
 > > (leaving the state), evolutive (becoming more in the state), or
 > > devolutive (becoming less in the state).
 > I expect to borrow your terminology for my revised system.
 > Would you say that evolutive and devolutive apply only to scale-type
 > qualities (like hot, fast, heavy, large) and to quantities rather than to,
 > say, things or actions?
For the most part, certainly.  Other uses can be imagined, but they'd
be ambiguous, and probably avoided in formal usage.  I tend to
consider conjugations without any obvious literal meaning to be of
possible poetic use ;).

For some qualities, the cessative and devolutive are identical to a
negative formation of the inceptive and evolutive, but not for all.
Morneau makes some interesting points about this here:

Incidentally, I'm wondering whether there's a better terminology (not
for each aspect itself (the terms for which I got from Rick Harrison's
article on aspect here:

except for devolutive, which is an obvious counterpart to evolutive)
but for the groups PPH and ICED, which seem to fall together
naturally).  I'm not aware of any general terms for these.

 > I have this problem too. It seems that there are really a large number of
 > core case roles, but no more than 3 for a given type of word. It's hard to
 > decide which get lumped together as a single case.

Yes.  It's tempting, especially when you start playing around with
voice, to abandon the idea of case roles with any kind of common
semantics at all, and just treat each root like a function with a
number of arguments the meaning of which is defined individually by
each function.  But if you take that to its logical conclusion, you
get lojban place structure, which while very fine in its way is not
something I aspire to emulate.

 > 'Yemls has something like:
 > A-case (or ergative) for Actions and remote cAuses,
 > P-case (or absolutive) for Patients, reciPients, Perceivers,
 > C-case for Clarifiers, Co-patients, etC.
 > For 'Yemls, syntactic expediency sometimes comes into play also.
 > I'm sure there are languages which (A) have no ditransitive _forms_, so it
 > should be possible to (B) eliminate ditransitive roots. The drawback of (A)
 > is that using essentially multiple separate verbs is inelegant, especially
 > for marking core case roles. Of course, you're only concerned with (B)
 > here. What you might consider is, instead of thinking _only_ in terms of
 > deriving ditransitives from transitives or transitives from intransitives,
 > also think about deriving transitives from basically ditransitive roots (or
 > intransitives from transitives) by deleting arguments. This also relates to
 > grammatical voice, I think.

Certainly you can do such things with voice, and derive say "merchant"
"to sell" "to buy" "to be sold" from a ditransitive root "to sell X
Y", and some of the forms may be incapable of taking all the
arguments.  I'm already assuming I'll do something along these lines
with transitives, and any ditransitives I end up with.

 > > Anyway, the use of the causative allows many transitive verbs to be
 > > reduced to intransitives, and some ditransitives can be reduced to
 > > transitives ("to teach" from "to know").
 > This is the part I didn't understand (examples snipped). Wouldn't a
 > causative affix _add_ an argument? Oh, wait a second. I interpreted
 > "reduced" differently from what you intended. You mean that a transitive
 > verb (with its own root) can be _replaced_ by an intransitive verb +
 > causative. Let me go back and rewrite some of my comments above ....

Yes, that's what I meant.  I was aware when I wrote it that it could
probably be taken two ways.