Yet another areal/GENERALITIES

Michael Witzel witzel at FAS.HARVARD.EDU
Wed Mar 3 20:46:46 UTC 1999

Now that all grammatical categories have been nicely muddled up in the
discussion of a lot of Japanese and some Sanskrit, it is perhaps better to
return to this aspect (sorry, side) of the question.  Since details are too
numerous, I dividide this into two messages.

In colloquial parlance, we tend to confuse tenses with aspects, and aspect
with  'type of action' of verbs  (Aktionsart). These are *three* different
things.  Sorry, I was too brief last time; I assumed more general
understanding of the verb system.

In detail: Many languages have an ASPECT system,  not (our English type)
TENSES, such as present, past, future, perfect etc.

Those which do *not* have a *grammaticalized* aspect system (such as
English, Hindi, modern Japanese) have tenses (but they *can* have
Aktionsarten -- I don't think there is a general Engl. expression; see
further below) .
Other, e.g., 'Aspect' languages,  do not need tenses, since they can make
the same distinctions by  using aspect or Aktionsart functions and forms.

Aspect languages usually distinguish the GRAMMATICAL CATEGORIES
imperfective ('not completed' action) and perfective ('completed' action).
Obviously, some verbs primarily fit only one  aspect:  you can sit, eat, be
'forever' (durative action --> imperfective aspect ), but you cannot arrive
'forever' (terminative action --> perfective aspect );  in Vedic, we also
have inchoative action > perfective aspect ('I left', which again you
cannot do 'forever'). And there are other, more specialized cases ('to

In Vedic these two CATEGORIES still are fairly well distinct:
*imperfective aspect   =  expressed by the GRAMMATICAL  FORM present stem,
*perfective aspect   =    aorist stem  (cf. below, DETAILS)

Many if not most Vedic verbs (see below) follow either category, as per
their meaning (sit :: arrive/leave), by using the bare ROOT as present ::
aorist  (thus, root present, Panini's no. 2) or root aorist.
The clearest case is of course :  as-ti  (imperfective: 'he exists',
theoretically, forever) :: a-bhU-t 'he became' (in Vedic; not 'he was!', as
mostly found in class.Skt.); -- cf. the English echo with:    is :: be,
been).  Here, Skt. (and similarly Latin, Germanic) even resorts to two
different verbs to make the distinction, and both forms are directly from
the root.  -- Or cf. the cases I used last time  a- ga-ccha-t ::  a-gan(t).
(< IE  *e-gwm-sk'e-t :: *e-gwem-t).

The problem arises when you want to use these verbs in the other aspect.
Thus,  agan(t) 'he made a step, went forth' :: a-ga-ccha-t 'continued to go
forth, went '. Note that in this case the present stem (with  past tense
agacchat) is specially characterized by the Indo-European inchoative -sk'e-
> Skt. -ccha- (yacchati, icchati, prchati < *prk'-sk'eti > posco, forschen
etc.)  You need -ccha- to express continuing action. gam-a-ti is not a
Vedic Present (but subjunctive aorist "he will go", a small difference!)

In the same way, you need a special suffix (s-Aor. for example)  in verbs
which fall into the action types (Aktionsart) such as durative:   stau-ti
'he praises' ('forever')  :: a-stau-S-It  'he finished praising (just now)'
<<NB: the 'just now' (sub-)function is not IE but a Vedic development, see
below ) .  >>

The IE aspect system is clear if one compares other IE languages (see, e.g.
gam : ga-cch, in Skt., Avestan, Greek, Armenian, Latin, etc., see the
recent compendium : H. Rix (et. al.), Lexikon der indogermanischen Verben.
Die Wurzeln und ihre Primaerstammbildungen. Wiesbaden 1998.
600 pp. of examples leave no doubt about this basic distinction.

That not all verbs in Vedic follow the  IE  root aor . versus  specially
designated pres. stem,  or, conversely, root present :: specially
designated aorist  distinction clearly any more, is no great  wonder: quite
a time span separates IE from Vedic.
As in all living languages, we see development: in the case of gam, we have
some 5 diff. aorist formations; (for details see K.Hoffmann, J. Narten, T.
Goto, C. Werba etc.); similarly, we have different present stems for about
80 verbs, often with a difference in MEANING, for example  bibharti 'bring'
:: bharati  'carry'! Here we can even speculate on the 'meaning' (gramm.
function) of the reduplication: continue to carry, and you will bring

The continuants / remnants of this aspect division (Pres. ::  Aor. stems)
are clear enough in the various IE languages in order to support an
original PIE aspect division. It is quite another question as to what was
going on in pre-PIE, or 'early' IE.  Our reconstructed IE is not of one
level, -- just as you have Rgvedic (with injunctives) and Upanisadic (no
new living injunctives, new 'double' subjunctives) and Classical/Epic  (no
subjunctives, injunctives used), within a time span of c. 1000 years.  A
reconstruction of the  "pre-Vedic"  based  on such items would yield a
confusing system.

I hope this clarifies the terms & limits of the present discussion.


Michael Witzel
        witzel at
Wales Professor of Sanskrit

Dept. of Sanskrit & Indian Studies,
Harvard University                    
2 Divinity Avenue                               (Electronic Journal of
Vedic Studies)
Cambridge MA 02138, USA                 (Harvard Oriental Ser., Opera Minora)

phone: 1- 617 - 495 3295 (voice & messages), 496 8570, fax 617 - 496 8571
my direct line (also for messages) :  617- 496 2990
home page:

More information about the INDOLOGY mailing list