Saturday, October 22, 2011

Swarming in Hebrew

For about the past year, I've been preoccupied with the question of "argument alternations" or "diathesis alternations" in Biblical Hebrew, just because I think they're interesting and also because I think they've been neglected in the lexicography and grammar of Biblical Hebrew. I've been trying both to understand them in a general way, and also to identify them in Hebrew.

Very briefly: What I mean by "argument alternation" is an alternation in the syntactic or semantic arguments of a verb that leave the form of the verb unchanged. For instance in English we can say "John broke the window" and also "the window broke." The object of the first sentence has become the subject of the second sentence, but the verb is unchanged. Another example is "Water filled the tank" and "The tank filled with water." The object of sentence 1 becomes the subject of sentence 2, while the subject of sentence 1 is placed in a prepositional phrase in sentence 2, but there is no change to the verb filled. Such phenomena have been the object of intense scrutiny, for instance in Beth Levin's English Verb Classes and Alternations (1993).

It seems to me that in Biblical Hebrew, alternations like this are rarer than in English, because typically in Hebrew when the semantic arguments change their syntactic mapping from, say, Patient/Object to Patient/Subject (as in the "broke" example), the verb stem or binyan also changes, so that, for instance, shavar (Qal) "he broke (something", would change to nishbar (Niphal), "(something) broke."

However, this is not always the case, and there is actually a fair number of alternating verbs in Biblical Hebrew. One small group that I have just run across has to do with the verbs meaning swarm or teem. For English, verbs of this type were studied at length in an article by Maurice Solkoff, "Bees are swarming in the garden" (1983). Briefly, verbs of this type (and it is larger than just the verb swarm and its synonyms), display an alternation in which the semantic Agent may switch syntactic Subject slots with the Location word, so that one may say "Bees are swarming in the garden" (agent subject) or "The garden is swarming with bees" (location subject). This type of alternation occurs, it seems, only with intransitive verbs.

Interestingly, the root שׁרץ in Biblical Hebrew apparently displays just such an alternation. The agent-subject type is attested; e.g. Gen 7:21 ha-sheretz ha-shoretz al ha-aretz, "the swarm that swarms on the earth"; or, e.g., Gen 9:7 "swarm (shirtzu) in the earth and multiply in it." But the location-subject is also attested: Gen 1:20 "let the waters swarm (yishretzu) with a swarm of creatures"; or, Exod 7:28 "the river will swarm (ve-sharatz) with frogs." A concordance will turn up all the examples, but if you want to track them down the references are Gen 1:20-21; 7:21; 8:17; 9:7; Exod 1:7; 7:28; Lev 11:29, 41-43, 46; Ezek 47:9; Ps 105:30. The synonymous verb רמש displays the same behavior, although the location-subject examples are few: Gen 9:2, Lev 20:25, both tirmos ha-adamah, "(all the creatures that) the soil teems with."

What is not immediately clear to me is whether the location-subject type has the same entailments in Hebrew as it does in English. It has been observed that in English the location-subject type has a "holistic" effect, and entails that the Location is more affected, more filled up, than with the agent-subject type; "the garden is swarming with bees" entails that every part of the garden has bees swarming in it, while "bees are swarming in the garden" does not have the same entailment. But in Hebrew I'm not sure that the waters in Gen 1:20 (location subject) will be more full of swarming creatures than the earth will be in the apparently similar Gen 9:7 (agent subject). This needs to be looked at in greater depth.

It is also not clear to me (yet) whether the core of the Hebrew verb has to do with motion (crawl, creep) or with numerical increase (abound, teem). It would be nice if the agent types would line up with the motion idea and the location types with the idea of increase, but they don't.

In fact, it is possible that this is another type of alternation altogether: the location-subject types do not have the look of intransitives at all. I've translated them that way (as do most English Bibles), but the complements in fact could be construed as accusative direct objects: not "the river swarmed with frogs" but "the river swarmed frogs" (i.e. produced them in swarms); "let the waters swarm a swarm of creatures," etc. This is in fact the way the LXX construes these instances. This, then, would be an example of the so-called Causative-Inchoative alternation, consisting of an intransitive verb and a transitive alternation meaning "to cause to (intransitive meaning)." In this case, the agent-subject would be intransitive and the location-subject would be transitive.

So we have, as it were, alternative alternations. Which is it? I'll have to think about it some more, but I'd be happy to receive the views of others.

BIBLIOGRAPHY: See also David Dowty, The Semantic Assymetry of Argument Alternations (and why it matters) (2001), at his personal webpage; Maurice Salkoff, "Bees Are Swarming in the Garden," Language 59 (1983): 288-346.