From: Vincent DeCaen (decaen@epas.utoronto.ca)
Date: Thu Oct 05 1995 - 08:34:43 EDT
re aorist as unmarked aspect, etc.
Prof John Wevers (emeritus) here at U of Toronto subscribes to the
view, and we've had several "interesting" ;-) exchanges on the
question. I don't think the facts support it, nor do my theoretical
investigations in tense-aspect support it.
I'd like to "float" some ideas too. but first I want to make sure
we're talking about the same things.
---------------------------------------------------
let's define "aspect" as the internal temporal contour of an "event
structure" using a time line T of points t. Two points at time index
i and j are selected such that i<j. the event begins at ti, continues
over a definite span, and terminates at tj, followed by a steady state
(= stative-resultative).
the aspects are defined on this internal time line:
-inceptive x=i
-durative/progressive i<x<j
-terminative x=j
-"perfect" or better, "anterior" j<x
-perfective/punctiliar x=(i through/including j)
these are good enough. we could quibble whether the anterior includes
j, and how to "conflate" the time line for perfective. I don't
believe languages grammaticalize x=i or x=j, so that leaves us, not
coincidentally, with three possible aspects. not every language
possesses a grammatical anterior, and they get by very well, thankyou.
notice also that there is nothing corresponding to an unmarked or
natural aspect in this scheme.
--------------------------------------------------------------
terms:
STRICT COMPOSITIONALITY
simply that the semantics should be derived straightforwardly from the
morphosyntax or "form" rather than independent stipulation (on the
traditional view). e.g., rather than defining independently an
anterior for English, we notice that the auxiliary have is inflected
for tense and that the participle has the extra bit -en/-ed. we
should be able to assign each bit a little semantic content, and by
strict composition arrive at the semantics of "anterior".
DEFAULT(ING)
*not* synonymous with unmarked ("unmarked" where required content is
not offered; for me = "underspecified"). a default is clearly
specified, but can be overridden (so it's not clear whether it's
semantics or pragmatics). in English, the default aspect is
perfective, so the simple past tense is perfective (unless
overridden), as is the simple present, or so I would claim. In such
systems, a separate and obligatory progressive is required (it turns
out most systems outside of Europe are perfective defaulters; major
exception, Athapaskan family).
---------------------------------------------------
1. "completed" "complete"
I think this is where the confusion starts. let us define these away
as "completed" = anterior/perfect and "complete" = perfective/punct.
1B.
"completed in the past"
confuses the obvious aspectual implicature of the past tense with the
aspectual reading. let us make sure "completed" isn't in fact past tense.
2. subjunctive is a "tense" in the sense of INFL projecting to a
maximal IP (for all those syntax fans: if you're not, just ignore
this). "form before meaning," is my motto.
3. how many aspects?!? formally there are 4-5, based on formal criteria
and the number of columns in the traditional paradigms. the question
is whether synchronically we can divide out "future stem" from "aorist
stem" (I don't think so).
__________________________________________________
some floating:
1. the aspectual value of the verb is only one of several factors
determining the overall tense-aspect reading: nature of subject,
nature of object, adverbials, lexical value of verb stem, pragmatics,
etc, etc. so on this view it is possible to get perfective readings
of an "imperfect"; and similarly, non-punctiliar readings of the
perfective "aorist". remember COMPOSITIONALITY over the whole clause.
2. formally, looking at the *productive* processes, we see that the bare
stem is generally associated with the durative, and that extra
stuff/processes is associated with the other aspects. it is
reasonable to assume that this correlates with the configuration of
the system.
3. let us define Greek's aspectual default as "nonperfective", so that
the simple forms are durative/progressive. the nondurative is
therefore obligatorally encoded, in this case by the perfective
("aorist stem"). as is generally the case, the anterior/perfect is an
add-on orthogonal to the basic contrast.
3B. in such systems, tack on the present inflection to the perfective
stem, and presto! "future tenses". hmmmm. do we want to separate out
"future" and "aorist" stems on a strictly compositional approach??
4. the gnomic reading of Greek perfectives is no different from that
of Russian which works the way I described Greek, nor from the English
simple tenses which overlap in many ways (although in many ways, you
get mirror-image phenomena). this is why we translate, eg., proverbs
with the simple present in -s (ie. nonpast perfective).
---------------------------------------------------
this has been "floating" and not in any way definitive. these views
are driven by a theory for tense-aspect in Universal Grammar. if any
of you are "into UG" we could follow this up offlist in a more
technical way.
ciao
---------------------------------------------------
Vincent DeCaen decaen@epas.utoronto.ca
Near Eastern Studies, University of Toronto
Religion and Culture, Wilfrid Laurier University
"The wise ones of Agarttha study all holy languages
in order to arrive at the universal language,
which is Vattan." Eco, Foucault's Pendulum
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:28 EDT