[b-hebrew] Verb Inflection & Tense

Rolf Furuli furuli at online.no
Tue Oct 4 09:53:19 EDT 2005

Dear Kevin, 

I have no knowledge of Polynesian languages.  At the University of Oslo there is a weekly seminar where we read Sumerian texts. Sumerian is a language family of itself with no other known relatives, although some traits may superficially seem to be similar with Basque or Tibetan. On the basis of your explanations it seems to me that there is a similarity with the seemingly chaotic situation of the Polynesian languages and that of Sumerian. For example, the new grammar of D. O. Edzard is very differend from the standard grammer of M-L. Thomsen. My experience from several years of study is that tense clearly is not grammaticalized in Sumerian.  And because the constructions of the verbs and their clauses are far more complex than for instance Hebrew or Akkadian, and the language is dead, I see now way to know whether aspect is grammaticalized in Sumerian.  But the grammars claim that aspect is there.

However, the Polynesian languages have living informants, and when these informants describe actions, they do so from a certain vantage point (the deictic center).  I am quite sure that my model can be applied to the Polenesian languages in a systematic, scientific way, even if the situation seems to be chaotic. One result that should be rather easily obtained is whether tense is grammaticalized.  When no verb form, or verb form together with particles, systematically refer to the past or to the future, we can draw the conclusion that tense is not grammaticalized.

To ascertain whether aspect is grammticalized or not may be harder.  But I use three different parameters in order to find and define aspect: the quality of focus, the angle of focus, and the breadth of focus of the intersection of event time by reference time.  I would guess that the application of these parameters to Polynesian languages would at least give some results, because it would be difficult to communicate in a language whose verbs can have any meaning.  Interestingly, I claim that imperfective and perfective verbs in classical Hebrew can signal both incomplete and completed situations (or bonded and unbounded situations, if this terminology is preferred). If this is true, it means that the two Hebrew aspects are similar in several respects. Nontheless, there are clear differences as well, and on the basis of these we can say that aspect is grammaticalized in Hebrew.  So, I suspect that the seemingly chaotic state of Polynesian verbs has some order after all, and that this can be demonstrated by an extensive analysis of thousands of verbs and work with informants.

Best regards

Rolf Furuli
University of Oslo

  ----- Original Message ----- 
  From: Kevin Riley 
  To: Rolf Furuli 
  Sent: Tuesday, October 04, 2005 3:43 AM
  Subject: Re: [b-hebrew] Verb Inflection & Tense

          Comments below.

        -------Original Message-------

        From: Rolf Furuli
        Date: 10/03/05 17:07:20
        To: b-hebrew at lists.ibiblio.org
        Subject: Re: [b-hebrew] Verb Inflection & Tense

        Dear Kevin,

        Your points below regarding Polynesian and Austronesian are informative.
        But there is one problem, which I comment on below.

        > Myself, I don't see major difference between completed aspect ('having
        > studied') and the past tense. I cannot imagine how scholars could
        > establish
        > that Polynesians mean aspect, not tense in this case. For example,
        > consider
        > a hypothetical Polynesian who studied but did not graduate. If Tagalog has
        > aspects, that Polynesian should use imperfect in describing his studies.
        > My
        > guess, he would use perfect - which would therefore be not perfective
        > aspect, but past tense.

        I agree that scholars would have great problems in distinguishing between
        "completed aspect" and "past tense" in the situation you mention. But the
        reason for the problem is not that this distinction cannot be made in a
        living or a dead language, but rather that assumptions that need not be
        correct are used. Thus, the problem may be theoretical rather than

        A colleague of mine who has spent many years among native speakers on
        islands in the Pacific ocean, and who is an expert of the native languages,
        said to me two weeks ago: "It is high time that we abandon the traditional
        definitions of aspects with their Slavic origin in the study of the Pacific
        languages. These definitions simply prevent progress in the language

        It is very important to realize that while tense is a universal term, which
        means that its nature can be understood even in languages that do not have
        grammticalized tenses, aspectual definitions are not universal. So the
        problem of distinguishing between "completed aspect" and "tense," as
        mentioned above, is the application of the English aspectual model to the
        languages of the Polynesians.

        There is a simple model by which tense and aspect can be distinguished in
        any language without looking at the foreign language in the light of one`s
        own language:

        1) Use a corpus with a huge number of verbs (more than ten thousand verbs is
        2) Find out whether verbs with the same morphology systematically refer to
        the future or to the past (smaller explainable exceptions are accepted). If
        not, tense is not grammticalized in that language.
        3) Find out whether verbs with the same morphology systematically refer to
        completed or uncompleted events (smaller explainable exceptions are
        accepted). If not, aspects in the traditional sense is not grammaticalized
        in that language.

        I have made these tests for classical Hebrew (my corpus had 79,574 verbs),
        and my finds are negative. Neither tense nor aspect (with the definitions
        completed/incomplete or complete/incomplete) are grammacalized in classical
        Hebrew. This is the falsification part of the approach based on the
        hypothetic deductive method. But there is a positive part as well.

        4) Use the parameters "event time," "reference time" and "deictic center".
        Find out whether reference time intersects event time in a systematically
        different way in verbs with one morphology in contrast with verbs with
        another morphology. If that is the case, the language probably has aspects,
        and their nature must be defined on the basis of the nature of this
        intersection of event time by reference time *in that language*.

        In a living language this model is easy to apply, because we have
        informants. In a dead language, the situation is more difficult, and only in
        a few clear-cut cases (hundreds rather than thousands) can the test of the
        intersection of event time by reference time be applied. In the other cases
        we can only use the cruder test of the relationship between event time and
        the deictic center.

        The basic problem in studies of Hebrew verbs is that we apply an aspectual
        model construed on the basis of aspects in English, Russian, or other
        languages, instead of testing the possible aspects in Hebrew in their own
        right. For example, many students of Hebrew use as a premise in their study
        that WAYYIQTOL, which, for the most part portray past, completed actions,
        either MUST be past tense or express the perfective aspect. Any claim that
        WAYYIQTOL is imperfective is unacceptable and will be rejected. However,
        in the Phoenician Karatepe inscriptions, the infinitive absolute plays the
        same role as WAYYIQTOL in Hebrew. There are 16 infinitive absolutes with
        prefixed WAW and 5 without prefixed WAW that describe past, completed
        events. No one would claim that the infinitive absolute has an intrinsic
        past tense or perfectivity. Yet, it functions as if that was the case! The
        lesson we can learn is that we cannot know the intrinsic meaning of verb
        forms by looking at them from the outside, i.e., by looking at their
        functions. The reason is that the functions of verbs in most cases are
        pragmatically conditioned, i.e., the reason for the choice of verb is the
        context. We therefore need to look for clauses that are so clear-cut that
        the particular functions of the verb with a great deal of certainty can be
        said to come from the nature of the verb alone. I have done this test in
        classical Hebrew, and my conclusion is that WAYYIQTOL, WEYIQTOL, and YIQTOL
        are imperfective, i.e., the intersection of event time by reference time in
        a few hundred clear-cut cases is systematically different (and qualifes for
        the definition "imperfective") from the intersection in QATAL and WEQATAL.

        But please note: In order to test these conclusions one has to get rid of
        the straightjacked of the traditional aspectual definitions. One must start
        afresh with the fundamental parameters "reference time," "event time," and
        "deictic center" and try hard to conduct the study with as little prejudice
        as possible.

        This is where Polynesian languages [among others] present difficulties.
        They do not have a simple Perfect(ive)/Imperfect(ive) distinction, nor a
        tense distinction.   'Ka' and 'kua' are tenseless as they can be used to
        refer to past, present and future.   'Ka' has no aspect, as it makes no
        claims at all about whether the action is completed or incomplete, ongoing,
        or any other condition except that it is a state or action.  'Kua' has
        aspect, as it always refers to what are viewed as completed actions or
        states completely entered in to, but there is no 'opposite' that indicates
        the imperfect in the same way.  It usually refers to fairly recent past, but
        not always  certainly not often enough to attribute 'past time' as one of
        its attributes.  'E...ana' is an aspect as it refers to ongoing or
        incomplete actions.  It is made up of 'e' which has the meaning 'non-past'
        and 'ana' meaning 'imperfect'.  'e..ana' *can* refer to the past despite the
        'e'.  Used alone, usually only in story-telling - 'ana' [sometimes 'ka ..
        ana' in some dialects] refers to past time, 'e' to future [never present]
        time.  I think 'ana' is used mainly with the passive, but may be wrong on
        that.  You cannot use 'i ... ana' to refer to past imperfect actions as in
        English "I was walking" as 'i' is used to mark time without aspect.  In
        eastern dialects 'e... ana' is usually replace by 'kei [present position] te
        [def. art.]...' and that has a past form 'i [past position] te ...', but
        then you have a nominal [or pseudo-verbal as it is usually called]
        construction.  Despite 'kei' meaning 'present position' it can have future

        Ka ... = beginning of new action - 'inceptive' ( but translating as "began
        to ..." is wrong most of the time) [... = position of 'verb']
        Kua ... = complete action or completed state - perfect/ive [but also implies
        recent past if used of the past - in Hawai'ian it is 'ua and is inceptive
        (replacing 'ka') as often as perfect/ive]
        E ... ana = imperfect/ive
        I ... = past action
        E ... = future action
        ... ana = past action

        Ka is the most commonly used, and has neither tense nor aspect
        Kua cannot always be replaced by 'e ... ana' to make the change from perfect
        to imperfect.  Sometimes 'i' is necessary when the imperfect [e... ana] is
        cancelled, and 'ka' must be used if the perfect aspect [kua] is to be
        cancelled.  There is no way of indicating present tense as opposed to past
        or future.  And ... ana seems to be used only with passive forms, not active
        [although my limited knowledge may be misleading me on that].

        How do you make sense out of this - i.e. present it as a system??  No form
        of verb contrasts with another, but all allow different aspects of the
        action/state to be emphasised.  There is not a perfect/imperfect distinction
          as you cannot use the imperfect in all the situations you can use the
        perfect or vice versa.  Can you speak of a tense distinction when there is
        no 'present' - although there is a past and in some situations a future -
        and the conditions under which 'i' and 'e' can be used overlap only in a
        minor way?  I am not sure you can talk of states in terms of tense at all.
        While all the Polynesian languages differ greatly in the particles they use
        and there is no way of equating the use of any particle in one with a
        particular particle in another, all share this confusing charateristic of
        not being able to oppose one form to another in the way that you can
        past/present/future in tense based languages or perfect/imperfect in aspect
        based languages.  Perhaps neither tense not aspect are the basis for the
        system of forms.  I am not sure applying your tests help to any greater
        degree than the traditional definitions in explaining how the system works.
        When you consider that the passive is the preferred form of the verb, that
        adjectives and nouns, verbs and adverbs are often distinguished only when a
        particle is added, and that the use of the articles is still a mystery even
        to native speakers [i.e they know when to use which form, but have no idea
        of why apart from 'that's just the way it has always been'], perhaps your
        colleague's comment about abandoning traditional European definitions needs
        to be applied more widely than just to aspect.  The Pacific languages
        [Austronesian, Indo-Pacific and Australian, each in their own ways] have not
        been kind to European linguistic theories - perhaps that may partly even the
        score where we lost out in other areas :)

        Kevin Riley


More information about the b-hebrew mailing list