The complicated morphology of the Semitic binyanim

Writing my section on morphology, I spent some time thinking about the nature of the binyanim in the Semitic languages. Namely, to what extent are the binyanim derivational and to what extent are they inflectional?

In short, derivation is the process by which the addition of an affix to a base morpheme creates a new word, either by giving a new meaning or shifting the lexical category. For instance, in English the suffix -er can be applied to a verb to create a noun meaning “one who does X”. Thus paint + -er > painter, “one who paints”. Another example is the use of the suffix -ize to change a noun into a verb, item + ize > itemize.

In contrast, in inflection an affix is used to to indicate grammatical information about the base morpheme, often called the stem. The type of information marked includes gender, number, case, tense, aspect, etc. Thus the plural marker -s in English is inflectional, not derivational.

In biblical Hebrew, we usually learn the binyanim in the section on verbs so that we memorize them in the verbal paradigms together with person, number, tense/aspect, strong and weak forms, etc. Because we associate them so strongly with verbs, I think most students consider the binyanim to be inflectional. In fact, when we parse a verb we give binyan along with all the other inflectional information. However, the binyanim are really derivational in nature. That is, they are part of the system in Semitic to build new words. This can be seen partly by the fact that generally Hebrew verbs do not appear in all of the possible binyanim.

The difference between drinking and causing someone else to drink has to do with the meaning of the verb, not its grammatical category. Dr Kaufman stressed this in class and urged us not to translate the hifil as “to cause X”, but rather to find an appropriate English verb in order to stress the lexical nature of the binyan. Of course, there is not always an appropriate verb. “To feed” is the causative of “to eat”, but what is the causative of “to drink”?

Arabic has several stem forms beyond Hebrew. Most interesting is probably the IX stem which is used with stative verbs denoting the inherent possession of qualities, such as color or physical defect, in order to express “to become X”. Thus, for instance, if the root ḥmr = ‘red’, iḥmarra = “to become red.” The rare XI stem has the same function, but is only limited to colors. This form lengthens the medial vowel of the IX form, thus iḥmārra.

Now, here is where inflection and derivation begin to blur. Is “to become red” really a new lexical meaning, or is it an inflection of aspect? That is, “to become” is really ingressive, a category of aktionsart or, as some call it, “lexical aspect”. With certain types of verbs in certain syntactic contexts, the binyanim may give Semitic languages an option for expressing aktionsart by varying the lexical nature of the verb. For example, in biblical Hebrew, the niphal may also have an ingressive sense when describing the subject as coming into a particular state.

This is still derivational in nature, but over time the process of grammaticalization (simply put, shifting a word/form that primarily has a lexical function to a grammatical one) may shift a binyan from a lexical derivation to an inflection of tense/aspect/mood. There are two possible examples from Akkadian.

The first example is the perfect iptarras in Akkadian. This form had always puzzled me since it is identical to the Gt form. However, it is common for a passive form to be grammaticalized into a perfective inflection. Thus, it makes sense that the Gt passive has been incorporated into the Akkadian TMA system as a perfect.

The second, and more complicated, is the iparras present-future form, which Akkadian shares only with Ethiopic and other modern South Semitic languages (yənäggər). The normal present-future form in Semitic is derived from *yaqtulu. Note the doubled middle radical in the Akkadian form, similar to the D-stem or piel. Could these forms be unrelated innovations in Akkadian and South Semitic which have grammaticalized the lexical meaning of multiplicity/repetition into a tense form? That is, repetition first shifted to progressive aspect before finally becoming a present-future tense?

Explore posts in the same categories: Aspect, Semitic Verbal System

11 Comments on “The complicated morphology of the Semitic binyanim

  1. Carl Pace Says:

    What I have found troubling in this discussion, Peter, is the origin of both the present-future in Akkadian and Ge’ez (as you have discussed) and the infixed -tan- stems of Akkadian (which you did not mention). I certainly find the idea of a grammaticalization of the D stem form with its lexical multiplicity to a durative present form, which Kaufman argues, appealing. What remains unexplained for me is why we seem to see these present-future forms with doubling of middle radicals in the Afro-Asiatic sphere, e.g., Tuareg /ilu:kku/ ‘he detests’ vs. /ilka/ ‘he detested’ or Kabyle /yesekkef/ ‘he drinks’ (Andrzej Zaborski discusses this and provides these and other examples in his article “Tense, Aspect and Mood Categories of Proto-Semitic,” in Current Issues in the Analysis of Semtic Grammar and Lexicon I [ed. L. Edzard, J. Retso; Wiesbaden: Harrasowitz, 2005], esp. 15-18). Should we seek the grammaticalization of the multipicative to a durative in an earlier phase of linguistic development? Can you remind me of what Kaufman said about how Akkadian and Ge’ez happened upon their identical but separate innovations, i. e., what motivated the transformation in each case?
    Even though the durative/present-future and the perfect in Akkadian seem to be adequately explained by Kaufman, what is one to make of the -tan- forms, and their absence in the other Semitic languages? Any thoughts?

  2. Jay Says:

    Great post, Pete. More importantly, good questions. I think if linguistics is to become more beneficial to students of Hebrew / Semitics, we need more people like yourself writing “intros” and then applying those intros to questions about the Semitic languages.

  3. Peter Bekins Says:

    Carl, do you really know anything about Tuareg? I thought it was a Volkswagen.

    It is a good point, what is Zaborski’s conclusion? Or does he just list the forms? The reason Kaufman saw them as separate innovations was because he felt the form survived way too long in Ge’ez to truly reflect an ancient Hamito-Semitic development. If you remember he considers roughly 1000 years to be the shelf-life of a verbal form, notwithstanding other linguistic forces. If Ge’ez was an earlier language, then the fact that the form survived on the two peripheral ends of Semitic would be good evidence that it developed at an older stage of the language. In contrast, the wide distribution of the *yaqtulu form suggests that it is the more original form. Of course who knows about some of this proto stuff.

    He didn’t explain the forces that caused the innovation in the respective languages, I think he was just generally in a grammaticalization mood.

    As for -tan-, this may have been one of the places where he saw a Sumerian influence on the Akkadian system. That is, Sumerian uses a wide array of affixes to specify all the grammatical information, the relationships between participants, and parameters such as location, direction, etc.

  4. Carl Pace Says:

    Of course I don’t know anything about Tuareg (or Kabyle); I was just culling an example from Zaborski’s argument, which is of course more his than mine.

    Volkswagen…funny. I suppose the ad must go something like “Everyone’s favorite Afroasiatic dialect is now the cutting edge of German technology…”

    I see what you mean about the idea of “shelf life”, though that puts us in a place of dependence on another theory. I certainly have aligned myself with Kaufman’s views, but as you know, it is not easy for any of us to enter into contest about such issues with those who don’t agree with Kaufman. It is at times like this that I wish he would put more into writing, so that we could better see the structure of his thinking and the evidence that he would bring to bear for his argument.

    The issue of the origin of the iparras-type present-future is certainly a contested issue. Zaborski is, as you could guess, not the only one who believes that there was a proto-semitic form /yVqattVl/, thought there are many varieties of this argument. Alice Faber (“Genetic Subgrouping of the Semitic Languages” in Hezron, The Semitic Languages) believes that the present-future yaqtulu was not a Proto-Semitic form, but rather an innovation that can be used to define the so-called “Central Semitic” family of dialects (a classification that Kaufman, among others, does not believe in). Zaborski sees the /yVqattVl/ form as Proto-Semitic, but he also leaves some room for the existence in Proto-Semitic of the form /yaqtulu/, which he calles “Imperfective” (how very Slavic of him) and which he sees as lying behind the subordinating iprus-u form in Akkadian. In fact, he suggests that the development of /yaqtulu/ into a subordinating form in some of the Semitic languages follows a known pattern of linguistic development.

    As for a conclusion, Zaborski just deals with a variety of forms and tries to trace their development backwards toward Proto-Semtic, but never comes to an overall conclusion. I hope all of this helps to contextualize my previous comment better.

  5. David Kummerow Says:

    Causative of “to drink” may sometimes be able to be rendered thus:

    1a. The cattle drank at the river.
    1b. He watered (ie, “caused to drink”) the cattle at the river.

  6. Peter Bekins Says:

    Yeah, but for some reason in my personal idiolect I reserve “to water” for plants. I never liked the sound of sentences like 1b. On the other hand, that’s another good example of how denominatives tend to become causatives.

  7. David Kummerow Says:

    I was wondering about that when I posted. I wonder what the extent of the acceptability of (1b) might be? Maybe it is just Aussie English, or even rural Aussie English (given that I grew up in rural Australia.

    (1b) for me is acceptable for inanimate objects (as for you with “plants”) but would seem to be extended to animates as well, as in (1b). Human referents are unacceptable. That is, I cannot say “I watered my son from a cup”.

  8. Peter Bekins Says:

    I think it is an issue of frequency. My dictionary says that “to water” can be used transitively with plants or animals. We urbanites don’t water very many animals, and I’ve never heard anyone say “I need to water the dog.” I wonder if that is because we consider dogs and cats a step closer to people than other animals?

  9. Sue Blue Says:

    How about throwing the following concepts into the Pot? Proto-Semitic letters were originally pictographs– concrete, action-based letters placed together to signify a particular function. I find it helpful to imagine that I’m an ancient Hebrew person who has just named an action. Naturally, subsequent modifications to the pictograph, to specify tense, mood, verb to noun migration,etc., does not eliminate the original intent of the pictograph. The changes just morph actions into various beautiful positions and we thus all become more abstract, but lovelier!! It’s a simple aesthetic revolution, originating in concrete but potentially brilliant minds. When we analyze morphology, please let’s work on understanding the ancient Hebrew culture.

  10. Ahab Says:

    You ask: [Q]”To feed” is the causative of “to eat”, but what is the causative of “to drink”?[/Q]

    How about “lave,” in Mendelssohn’s Elijah: “Thanks be to God; He laveth the thristy land” ?

  11. Peter Says:

    Well, I think ‘lave’ is technically to wash or bathe, so this may be a mixed metaphor that works because the land is inanimate. Pouring water onto the land satisfies its ‘thirst’, but I don’t think it would mean ‘to cause to drink’ if you said that ‘God laveth his flock’, etc. Of course, ‘lave’ isn’t really part of my everyday vocabulary so I guess it is possible.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: