Chapter 5 of Naphtali Meshel’s The “Grammar” of Sacrifice focuses on “praxemics,” the physical actions of sacrificial ritual. Praxemes are constituted by six elements: an “atomact,” an elementary physical act or gesture; an actor; an object that is acted up; a target for the action and often for the object or material that is acted upon; a location; and a time-frame, indicating when in a sequence of acts a praxeme is performed (130). For instance, a priest (agent) might daub (atomact) blood (object) on the horns of the altar (target) after slaying a hattat offering (time-frame). Not all the instructions of “P” specify all six elements of a praxeme, and many actions that are necessary to the praxeme are left out (to daub on the altar, the priest must walk to the altar, raise his hand, extend his finger, etc.).

With praxemes, Meshel enters well-trodden territory. When others have worked out linguistic models of sacrifice, they have paid particular attention to the actions of the ritual, less on the specified animals or the associated grain offerings and libations. This is bad method, Meshel thinks: “the situation in the study of ritual grammar today is somewhat analogous to a situation in which linguists are aware of one category, for example phonetics, but do not acknowledge any other categories, thus attempting to compose grammars of natural languages on the narrow basis of phonetics. Such attempts would lead to grammars with very low explanatory power” (133). Indeed. But I will argue below that Meshel’s own grammatical analysis leaves us in something like this position.

Meshel’s sixfold analysis of praxemes serves him well in trying to ferret out the logic of the sacrificial system. For example, he notes that saraf (burn) and hiqtir (turn to smoke) might involve identical atomacts: In both cases, an agent puts some part of an animal in fire and burns it. But the atomact is not the only relevant consideration in assessing praxemes. Though the two verbs describe identical actions, they are “always praxemically distinct,” mainly because of the different locations. Hiqtir is always linked with the altar as its target while saraf is never connected with the altar, and “the former may have the function of making smoke as a pleasing aroma to YHWH, whereas the latter is a means of disposal” (142).

Another example: Blood is applied to various objects in different ways. Blood can be tossed, poured, squeezed/drained, daubed, and flicked onto different portions of the altar (145), and certain modes of applications are used exclusively with specific portions of the altar. “In the sacrificial system of P, the object upon which blood is poured is always the base of the altar. . . . ‘Pouring’ applies only to the rest of the blood of a quadruped offered as a purification offering after some of its blood has been applied elsewhere” (149). (It would be enlightening to follow this through to the New Testament; what is the significance of Jesus’ statement that the blood of the covenant He offers at the table is “poured”? By Levitical standards, that seems to be a specific identification of His death, and the Supper, with the sin offering. In what sense is Jesus’ blood poured out “at the base” [cf. the location of the souls of martyrs in Revelation 6].)

Intriguingly, Meshel argues that the phrase “on the altar around” does not indicate that blood is placed on the “sides” of the altar but rather that it is placed on the upper surface. He observes that “when P wishes to communicate the application of blood to the walls of the altar, it states so explicitly, by means of the phrase . . . ‘upon  the walls of the altar,’ Lev 1:15; 5:9” (150). If true, this would mean that the blood of a flock or herd ascension offering is placed on a different portion of the altar than the blood of a bird (cf. Leviticus 1:5, 11, 15).

As in other sections of his book, Meshel notes that there are far more possible combinations of acts, objects, targets, agents, locations, and time-frames than appear in the Levitical system (Meshel lists 8 attested praxemes involving blood, 164-5). Some exclusions may be explained by practical considerations: It’s really tough to toss blood onto the horns of an altar. With their expertise at slings, Benjamites might be able to do it; Levites, not so much. But practical considerations are not always in play: “There is no practical reason why one could not daub blood at the base or walls of the altar, toss blood onto its base, flick blood onto its horns, or drain a bird’s blood onto its upper surface. . . It thus emerges that a large number of possible combinations have not been employed, either because they were deemed inadmissible or, perhaps, because the needs of the sacrificial system envisaged in P are not so varies as to require more combinations than those evidenced in the text” (165). Meshel assembles rabbinic evidence that some of these combinations were later considered legitimate: flocking blood on the upper surface or the horns, tossing blood at the base, pouring blood down the walls (166-70).

The rabbinic fits Meshel’s thesis that the grammar of the Levitical system is generative, how how might one explain the strict restrictiveness of “P”? Meshel observes that “one may proceed from the most stringent rule” and conclude that only the combinations in Leviticus are grammatical, but he doesn’t think this kind of approach will qualify as a grammar, since it would be “a simple list of praxemes” and “not generative” (172).

For my money, the questions Meshel raises at the end of the chapter put a question mark over his enterprise – both about the generative character of the grammar and about the use of “grammar” as a category to explain sacrifice in general. The fact that we have atomact “daub” and target “laver” implies for Meshel that it would be ritually grammatical to daub blood on the laver. The action of daubing blood on the laver can be generated by the grammatical principles of “P.” As Meshel formulates Rule r1: Defining A as the set of possible actions and T as the set of targets, “any member of set A may combine with any member t of set T” (164).

This is certainly a way to run a grammar. Any member of the set N (nouns) can combine with any member of set V (verbs) to form a sentence. But, as Meshel is at pains to demonstrate early in his book, what is grammatical is not identical to what is sensical. Grammatical but apparently nonsensical statements can be funny, and can even be lovely and poetic. But to know what works we need to have more than grammar. When we have formulated a grammar, we then need to consider semantics before we can be considered skillful users of the language. That is, perhaps there are reasons of meaning that prevent Leviticus from instructing priests to pour blood on the walls of the altar. Perhaps what flicking means, and what the horns represent, exclude “flicking blood on the horns.” It may be grammatical, but nonsensical, perhaps blasphemous.

Meshel’s last chapter is on “Meaning,” so in my last installment of this review I will examine how, or whether, Meshel answers my questions.

Peter J. Leithart is President of Theopolis.