no future (finis)

From: Vincent DeCaen (
Date: Tue Oct 24 1995 - 22:43:02 EDT

I'm kind'a hoping we can end this thread soon, but since I've gotten
flooded with questions re "future tense" I think it's only fair to
respond a little.

first, I'm not sure what's wrong with either formalism or theoretical
elegance. I do admit a heavy emphasis on form because of a
methodological principle: strict compositionality (vs atomism). put
simply, we must take form seriously, and assign meaning to the bits so
that when put them together we get the correct semantics.

e.g., what tense is "I would have been being stupid (had I answered
all the questions)" ? atomism would say a "conditional perfect
progressive" or some such gobbledy-gook. compositionality would say,
"past tense of will"; and stem have, and the participle been, and the
participle being, which when put together gives the correct semantics.
my claim is that when we call the Greek thingy "future" we're in the
atomistic camp; and I'm interested in what the compositional camp
would do with it--that's all, period.

the other problem is consistency of form to meaning. and I'm sorry,
but you're not going to get this except maybe in Esperanto. if I say
that the English plural is the stem plus /-s/, some young'un will
point out that women obtains and not womans. well, that 's quaint.
should we give up the analysis of the regular plural on the basis of
less productive morphology? the problem in Greek is that sort of thing
is so characteristic of the verbal morphology. so because I was using
the sigmatic aorist for shorthand is no reason to through out the baby
with the bath water.

1. no, the analysis was not driven by the "binary fallacy"; though it
is curious how far binarism as a heuristic can get you, and how much
semantics is intuitively binary.

2. making explicit the model I'm assuming. well, since I'm just toying
with ideas, we shouldn't take it too seriously, but I assume contra
Olsen that there is a distinction between perfective and "perfect" or
anterior (semantically, the latter is much more marked; even formally,
it appears with more stuff, or quirkier stuff).

                Non-Past Past
                [ ] [PAST]

Imperfective "Present" "Imperfect"
[ ] [ , ] [PAST, ]

Perfective "Future" "Aorist"

Anterior "Perfect" "Pluperfect"
[ANT] [ , ANT] [PAST, ANT]

now, it is a nice result to me that formally we get the "stems" lining
up, and the augments lining up; and it looks conceptually
clean/elegant. but hey, aesthetics is a personal thing.

3. is there not one speaker of Russian or cognates, or Hungarian or
Georgian or any of these systems that use (unconstroversially) the
combination of nonpast + perfective = future?? I mean, the cogency of
the model stems from my work on Russian, Polish, Hungarian and Mofu
Gudur; so what more can I do than refer you to these systems to see
the semantics for this proposal?

4. there are some predictions in moving to the nonpast + perfective.
one is that there are other uses of the "future" other than future.
since Greek is not my specialty, I won't venture out here; but I do
remember anomalous "futures" that fit the semantic bill.

5. the diachronics is of course a problem. two problems actually. a)
looking at Sanskrit we see two stems, e.g., sigmatic stems: -s- vs. a
"future" -s-y-. I can see how it could be argued that the two
collapsed. on the other hand, history is opaque to language learners.
b) there is the constrast lu-s-o-nt-os vs lu-s-a-nt-os, future part
and aorist participle respectively. here I would argue that the future
is expected with the "present" or regular endings. I would venture
that the presence of -a- is the source of additional past (is there
not something in Hittite -Ha- ? memory fails now).

6. I've done a lot of surveying of languages, and the only candidates
for bona fide "future" is the inflected Romance infinitives, and
systems with "metrical tense" like Bantu systems or Inuktitut
(Eskimo). I'm not interested in taking apart counterexamples here.
It's clear historically that the clitic of *habere is on the end of
infinitives, and formally the analysis still holds. I could regale you
with data from Portuguese and others that still have material showing
up between infinitive and clitic auxiliary--some other time.

generally, the metrical systems are subject to decomposition.

the Bantu systems are interesting because of the apparent
one-TMA-particle-per-verb rule. here the key is that stacking of TMA
particles requires auxiliary supports (morphological requirements I
guess). when you look at the additional data, you can separate out T,
M, A and in that order. I'm most familiar with Swahili, and I'm still
working on the so-called future tense (probably irrealis mood).

anyway, the burden should be on those supporting the future tense (the
world apparently does just fine with just past/nonpast, thank you very
much). there're some good books, but the only one that comes to mind
is Fleischman's 1982 "The future in thought and language". she's
dealing with Romance mostly.

Conclusion: points suggesting this route (give it to a grad student to
work out, I'm working on Semitic syntax right now).

a) cross-linguistic lack of bona fide future not subject to decomposition
b) synchronic symmetry: stem and aspect
c) doing without one major semantic feature [+/-Future] that doesn't
do much otherwise and is in complementary distribution with perfective
in just the nonpast.
d) Slavic, etc, nonpast + perfective = future.

sorry I'm not more forthcoming: it's been a long couple 'a days. sigh.

Vincent DeCaen

Near Eastern Studies, University of Toronto
Religion and Culture, Wilfrid Laurier University

I really do not know that anything has ever been
more exciting than diagraming sentences.
                                 --Gertrude Stein

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:30 EDT