Our blog has moved!

You should be automatically redirected in 5 seconds. If not, visit
and update your bookmarks.

Sunday, February 10, 2008

Arabic and Aspect and Tense

In a discussion on another blog, www.AncientHebrewPoetry.typepad.com , John Hobbins asked me about tense and aspect in Arabic, with an implicit question on its relevance to biblical Hebrew tense-aspect,

"Does [Östen] Dahl represent a consensus point of view when he interprets the classical Arabic verbal system as at root aspectual? Or is there a continuing debate in that field that mirrors the one in ours?"

Well, it depends on what one asks, and whom one asks.

If you are asking "at root", then that is a historical, etymological question.

[[As an aside:

(on the web see Larisa Avram , Generative Linguistics and Child Language Acquistion, an open-access e-book, http://ebooks.unibuc.ro/filologie/avram/ Her chapter on tense-aspect is short and may be very helpful for European Semitists to get beyond their 'either tense or aspect' mentality)

Child development and creole studies would both point to the probability that a verb form in Semitic was 'aspectual' at some proto-language stage, because aspect is theoretically the first paramater to be morphologized within a language. But if that morphology fixes itself at a binary stage, then that morphology will be used for a whole 'TAM' within the theoretical language. [So Bickerton, see below, and to some degree Avram.] By the time we reach Akkadian we already have a ikattab//iktub dichotomy where the [ktub] root looks fairly time oriented to some, or at least is considered the predecessor for the Hebrew vayyiqtol that is considered by most to be fairly time oriented. If ktub had become a temporal root with perfectivity, then we may not know what the proto-Semitic imperfective aspectual root looked like, or in any case it is irrelevant to West Semitic. The ktub root was reinterpreted (or was proto-Semitic following Moscati and evidence of imperatives) as an imperfective aspectual root within binary West Semitic systems, as evidenced in Arabic. We also have a problem of limited data for the the second-fourth millenia BCE, and we have the iconoclastic material of 1m BCE Hebrew where ktub is both temporal past a.k.a. perfective in vayyiqtol and future-imperfective aspectual in yiqtol. But I don't think you wanted to ask "root" questions of this nature. And I'm not sure I want to make final pronouncements on Proto-Semitic.]]

If you are simply asking about tense and aspect, then again, it depends whom you ask.

If you ask the first native Arabists, the answer is that Arabic includes time within itself, as is evidenced by their naming kataba "al-maaDi" and yaktubu "al-mustaqbal."

al maaDi is past, al mustqbal is future.

In the 19th century this was turned on its head by Europeanean Arabists, as evidenced by Wright, §77a (p. 51)

"A Semitic Perfect or Imperfect has, in and of itself, no reference to the temporal relations of the speaker (thinker or writer). … The Arab grammarians themselves have not, however, succeeded in keeping this important point distinctly in view, but have given an undue importance to the idea of time, in connection with the verbal forms, by their division of it into the past (al-maaDi), the present (al-Haal or al-HaaDir), and the future (al-mustaqbal), the first of which they assign to the Perfect and the other two to the Imperfect."

However, the famous Indoeuropeanist, Jerzy Kurylowicz, Studies in Semitic Grammar and Metrics, 1972, was unimpressed with the basic claim to 'aspectuality' in Semitic. He had a good grasp about what aspect was about, controlling more than one Slavic language. (book is currently in boxes, I don't have access. http://www.worldcatlibraries.org/wcpa/top3mset/803958)

Why the discrepancy between mother-tongue speakers and among linguists? The Arabic yaktubu can be used very much like the biblical Hebrew participle. For marking imperfectivity in the past one finds kana yaktubu in parallel to First Temple biblical Hebrew haya kotev. The prefix form in Arabic was used to mark imperfectivity. But it was also sensitive to time and expressive of time.

As mentioned above, Arabic was a binary verbal system, basically dividing a past//non-past on the time axis, a perfective//imperfective on the aspect axis, and having a fairly complex modal axis including a subjunctive prefix verb with 'a', a short prefix perfective jussive/pasts, and a suffix verb for many conditionals, and prefix verbs with a rhetorical -n-.

Now binary verb systems are notorious for being multi-dimensional. In fact Derek Bickerton, a linguist famous for creole studies, claimed that most linguists working in the field of verbal syntax were too often irresponsible and arbitrary in their definitions of tense, aspect and mood. The 'meaning' of a verb depended more on how many slices of the pie were involved and on internal oppositions, rather than on specific 'time', 'aspect', and 'mood' features. A binary verb system is thus inherently prone to produce 'tense-aspects', something that is neither a pure aspect nor a pure tense. See the chapter cited above on the web book, for an overview of how tense-aspect fusion can be handled in a neutral manner.

This proto-typical binary tense-aspect-mood dichotomy is what we see in Arabic and explains why mother-tongue Arabs used time in their definitions over a millenium ago, and why some linguists like Dahl might want to put Arabic in their 'aspect' category. So yes, Arabic is problematic and boils down to definitions and whether or not 'tense' and 'aspect' are kept semantically clear, whether 'tense-aspect' is allowed in the definitions, or whether 'aspects' are redefined in order to mark time, too.

While noting the similarity to the biblical Hebrew tense-aspect-mood dichotomy in yiqtol~qatal, it is also helpful to note the differences. Hebrew marked "imperfectivity and past" by using haya + participle. Imperfectivity alone could be signalled by both yiqtol and veqatal, as the classic passage in Gen 29:1-3 shows. Arabic was using yaktubu for the participle uses and did not have a sequential system.

Especially important for your other question about Hebrew participles is to note that Arabic can use yaktubu for actual present tenses while biblical Hebrew already incorporated the participle.

Finally, I wouldn't want to imply that native nomenclature is always correct. Arabic explanations of the "lam yaktub" structure he did not write wander all over the board, because it is an isolated historical relic from the same system that produced the bH vayyiqtol. But to throw out their perception that time is included in their basic verb would be equivalent to throwing out English complaints when some says, "Look, the English future is really a volitional, since it uses 'will'." That mistakes etymology for semantics. English has a future, and it marks it with the volitional lexeme 'will', not the strangest thing in the world. For that matter, someone could claim that the modern Greek is "really" a volitional in drag, since it uses a remnant from θελειν 'wish, want' in order to mark the modern future perfective θα γραψω 'I will write' and the modern Greek future imperfective θα γραφω 'I will be writing'.

All of this goes to suggest that Arabic includes time inside of its binary TenseAspectMood, and that etymology is not always a good indicator of semantics. I would say the same of biblical Hebrew.


John Hobbins said...

That's very helpful. I understand the background of the TAM debate on my blog much better now.

John Hobbins

evepheso said...

Sorry to be commenting on this post so late, but I didn't think to leave a comment here until recently when I remembered what you wrote about the English future.

A non-etymological argument can (and has) been made about the English will being volitional rather than a future tense:


The evidence is syntactic distribution, not etymology. Note, this is the analyis from the Cambridge Grammar of the English Language

Randall Buth said...

Thank you, evepheso, for an informative link. In the comments of that link, you will find that Östen Dahl calls English 'will' a "wanna-be tense". If you say that it fits in a syntax machine in the same way as modals, then it is a "future" modal. But it marks futurity, not volitionality. E.g., "He will come even though he doesn't want to."

So my English still has a future tense. And that is an etymological development of the verb 'will'.

And now for a 'qal vaHomer' argument. If analysts are capable of confusing the descriptions in a language where mothertongue speakers know better, how much easier time will they have scrambling a system without mothertongue speakers?

evepheso said...


I merely was pointing you toward the discussion itself, not arguing, challenging, or even disagreeing you.