LINGUIST List 26.1711

Tue Mar 31 2015

Sum: Morphology of tense across languages

Editor for this issue: Anna White <>

Date: 28-Mar-2015
From: Ariel Cohen <>
Subject: Morphology of tense across languages
E-mail this message to a friend

Dear all,
Some time ago, I posted a query entitles “Morphology of tense across languages”:
Dear all,
Does anyone know of a language where present tense is signaled by overt morphology, while past/future is not?
Thank you very much.
I received very helpful replies from the following people:
Martin Haspelmath
Ivan Kapitonov
Bruno Olsson
Steve-Monica Parker
Daniel Ross
Thank you so much!

Here is a brief summary of the replies:

1. Martin Haspelmath helpfully sent me his paper:
Haspelmath, M. (1998). The semantic development of old presents: New futures and subjunctives without grammaticalization. Diachronica, 15(1), 29-62.
In this paper he shows examples from Udmurt and Kannada, where the future is more marked than the present (progressive).

2. Vanya Kaitonov pointed to me a discussion on Linguistics Stock Exchanges
The gist of the discussion is that it is difficult to answer, because the distinction beween tense and aspect is not always clear, and also because there are languages with no clear present tense, just “non-past” or “non-future”.
Most of the discussion is about languages where a non-present tense is not unmarked, but has less “stuff”.
But one possible candidate is Anii, where “the unmarked form isn't a present, it's a past.”.

3. Bruno Olsson suggests a candidate language:
Marind (Papuan language of south New Guinea) is described by Drabbe (1955) as having a present tense prefix Vp- (where V is gender agreement), and bare verb forms giving Perfective Past:
ep-ano-kiparud-at (MASC.PRES-1sg-tie-DUR) ''I (male) am tying''
no-kiparud (1sg-tie) ''I tied''
From p. 39, Drabbe, Petrus. 1955. Spraakkunst van het Marind.

4. Steve Parker suggests “ Michif, a native American language mixed with French.”. He provides an interesting example, from an unknown source:
“[diminikwan] ‘I drink’
[timinikwan] ‘I drank’
In spite of this minimal pair, there is no need to posit a phonemic contrast between /t/ and /d/ in this language.
As a matter of fact, all of the obstruents in verbal forms in this language are voiceless at the underlying level.
The phonetic [d] here is the surface realization of an underlying sequence /n - t/, where /n-/ is a prefix (probably a tense or aspect marker). These two consonants fuse or merge together into the intermediate segment [d], which preserves features of both. This is called coalescence”

5. Daniel Ross agrees that it is “Unlikely due to markedness.” He notes that “one potential example is Arabic, where the past tense is the citation form of the verb and the present involves a different structure.”
But warns that “ this is complicated because:
1. Arabic morphology is not concatenative; the triliteral root system uses three consonants in various templates of vowels (and sometimes other consonants) in addition to, sometimes, affixes. So determining what is ''bare'' or ''uninflected'' is tricky. But just the three consonants (with minimal vowels inserted) is used for the citation form, which is 3SG.PAST.
2. This applies most obviously to uninflected 3SG, where in present tense there is a prefix and in past tense there is not suffix.
3. One marker of this (equal in both tenses) may be that there are prefixes in the present and suffixes in the past (for complicated historical reasons probably to do with differences in VSO and SVO order).
4. It's not clear these are really tenses; they're possibly more properly analyzed as aspects, although drawing a true distinction there is tricky.”

Thanks again to all who responded.

Ariel Cohen
Ben-Gurion University of the Negev

Linguistic Field(s): Morphology

Page Updated: 31-Mar-2015