Agent-less indirect adjectival verb forms in Egyptian and Arabic: the case of jrrw n.f. and maf ul lahu, "for whom one acts."

• Author: Depuydt, Leo

Wolfgang Schenkel Tubingensi sexagenario

Past discussions of those peculiar Egyptian verb forms that are sufficiently defined by the three terms "agent-less," "indirect," and "adjectival" have focused on three issues: their location in the verbal system, their history (and that of the affiliated relative verb forms and passive participles), and the resemblance they bear to certain usages of the participle in Arabic, usages often represented by generic examples involving passive participles such as al-mar atu l-maqtulu abu-ha "the woman whose father has been killed" or ar-rajulu l-maqtulatu ummu-hu "the man whose mother has been killed." The present paper focuses on the third problem, the parallelism with Arabic, and any relevance it may have to the first and the second.(1) But before addressing the third problem in section 4, some background will be provided in sections 1 to 3. In sections 1 and 2, the Egyptian and Arabic phenomena are surveyed separately. In section 3, past discussions on the similarity between the two are discussed.

1. AGENT-LESS INDIRECT ADJECTIVAL VERB FORMS IN OLD, MIDDLE, AND LATE EGYPTIAN

   1.1. Definitions of "Agent-less," "Indirect," and "Adjectival"

   One of the hurdles obstructing the beginning student's path to an adequate knowledge of earlier Egyptian is a proper understanding of the nature of verb forms such as pssw (masc. sing.) and psst (fem. sing.) in Coffin Texts (de Buck 1935-61, henceforth CT) VI 284c, e, and g-h, coffin B1Bo (cf. Verhoeven 1984: 124).

   jw.f rh rn n psjjt tw psst jm.s . . . jw.f rh rn n wjjt tw psst jm.s . . . jw.f rh rn n nbt pw pssw jm.f

   "He knows the name of that psjtt-pot in which one cooks . . .;

   he knows the name of that wjjt-pot in which one cooks . . .;

   he knows the name of that nbt-pot in which one cooks."

   The verb forms pssw and psst exhibit three features that sufficiently characterize them as instances of the distinct area of the earlier Egyptian verbal system that forms the topic of the present paper.(2)

   First, they display a changeable gender and number ending, which is zero in the masculine singular. Since variable gender and number endings are perhaps most characteristic of the adjective as a word type,(3) pssw and psst may be conveniently described as adjectival. Indeed, like adjectives, pssw and psst modify substantives. Verb forms exhibiting gender and number endings - or just number endings, as in the plural future participle of the type stp.tj.sn "they who will choose" - are very characteristic of earlier Egyptian. The endings refer to entities - persons, places, or things - involved in the events expressed by the verb forms. Adjectival verb forms disappeared entirely from Egyptian in the later stages of the language.

   Second, pssw and psst do not specify who does the cooking. This justifies denoting pssw and psst as agentless.(4) Incidentally, nor do the two verb forms specify what is being cooked. But absence or presence of what is being cooked is not a criterion for delineating the verb forms at hand. For example, what is being cooked does receive mention in pst drw.s "whose ends one has cooked" (PT 1485b; cf. Verhoeven 1984: 130).

   Third, the entities denoted by the gender and number endings of pssw and psst, namely the various pots, neither perform nor undergo the cooking. If they did, the verb forms would be called direct. However, the pots are what one cooks in. Because the gender and number endings refer neither to the actor or agent nor to the undergoer of the event of cooking, pssw and psst will be called indirect.

   The indirectness of pssw and psst is remarkable from the point of view of English. Traditionally, pssw and psst are classified as passive participles. Although hieroglyphic writing fails to express many morphological distinctions because vowels are not written, I do not doubt that pssw and psst are, in form, identical to passive participles. This interpretation is supported empirically by the ending w in pssw, an ending that is typical of that verb form. It is difficult to see what else w could signify. In English - and in most if not all other languages exhibiting them - passive participles are as a rule direct. Therefore, if pssw and psst were translated in accordance with how passive participles generally function outside Egyptian, the pots would be what is cooked, and not what is cooked in. In fact, passive participles derived from ps(y) "cook" do occur with direct function, as in jwf ps "cooked meat" in Sinuhe B88 and pst nbt "all that is cooked" in Sinuhe B92 (cf. Verhoeven 1984: 115-17). Direct function appears normal if Egyptian is considered from the point of view of English. Indirect function is remarkable and typically Egyptian.

   1.2. Location of Agent-less Indirect Adjectival Verb Forms in the Verbal System

   Two questions are posed: If viewed as a part, how do these forms relate to larger wholes ([section]1.2.1)? If viewed as a whole, how can they be subdivided into smaller parts ([section]1.2.2)?

   1.2.1. The Four Types of Adjectival Verb forms. The four traditional types are as follows:

   TYPE a: active participles, e.g., jrrt "she who does" (Gardiner 1957: [section][section]353-57, 359, 362, 365-68, 370-75, and 390-400);

   TYPE b: passive participles, e.g., jrjjt "what has been done" (Gardiner 1957: [section][section]353-56, 358, 360-62, 369-70, 374, 379, and 390-400);

   TYPE c: relative verb forms, e.g., jrrt.f "what he does" (Gardiner 1957: [section][section]380-87, 389, and 390-400); and

   TYPE d: agent-less indirect forms, traditionally designated either as a special use of the passive participle or as "impersonal" relative verb forms, e.g., jrrw n.f "for whom one acts" (Gardiner 1957: [section][section]371, 376-77, and 388).(5)

   This multiplicity of terms obscures the fact that the adjectival verb forms are a single species whose main types can be classified according to the same overarching criteria. The most obvious candidates as criteria, form and meaning of the verb forms themselves, are not adequate. They are especially inadequate to classify type (d).

   Form is somewhat elusive as a criterion because of the nature of hieroglyphic writing. For example, active and passive participles are very often written alike. As regards (c) and (d), the most probable theory, proposed by Kurt Sethe and Alan H. Gardiner (cf. Schenkel 1975: 26-34), holds that they basically have the form of (b), the passive participle, because they derive from it. But this cannot be positively proven because of the lack of vowels in the hieroglyphic script. Form is therefore not optimal as a criterion.

   As regards meaning, Sethe's and Gardiner's theory holds that the derivation of (c), the relative verb forms, from (b), passive participles, involved a shift in meaning from passive voice in (b) to active voice in (c): (c) jrrt.f "that which he does" is thought to derive from (b) jrrt "that which is done," to which a suffix pronoun is attached. As regards type (d), then, it will be assumed below, according to a concept called "competing analogies," that it is sometimes active, sometimes passive, sometimes neither, and sometimes both. Meaning too is therefore not suitable as a criterion, not only because meaning is in general more elusive than empirical form, but also because type (d) is very often not distinctly either active or passive.

   Instead of form or meaning, two binary criteria that are implied in the terms "agent-less" and "indirect," which have been defined in [section]1.1, will be used to classify all the adjectival verb forms. These two criteria pertain to observable characteristics of entities and circumstances involved in the events expressed by adjectival verb forms. According to the first criterion, I, the agent is (A) or is not (B) expressed.(6) According to the second criterion, II, the gender and number ending of the verb form is (A) direct, that is, refers either to the actor or agent (Aa) or to the undergoer (Ab), or is (B) indirect, that is, does not refer to the actor/agent or undergoer. The resulting table is conveniently distinctive and based on unequivocal criteria. It may even be suitable for clarifying matters to the beginner. The two criteria by which it has been obtained seem straightforward.

   (a) (b) (c) (d) active passive relative agent-less participles participles verb forms indirect forms I A B A B(7) ("agent-less") II Aa A Ab/B(8) B ("indirect") No alternatives will be suggested here for the traditional names of (a), (b), and (c). The terms "agent-less" and "indirect" will be used to denote (d). Like (d), (b), passive participles, are also agent-less (I/B), and (c), relative verb forms, can also be indirect (II/B), but only (d) is always both agent-less and indirect.(9)

   Traditionally, instances of (d) are classified either as passive participles or as relative verb forms. But it is certain that neither term can cover all the instances of (d) as a group. There are instances in which the term "passive participle" is clearly not suitable, as well as instances in which "relative verb form" is not. On the one hand, instances of intransitive verbs (type 3 in [section]1.2.2 below) cannot be passive participles in the full sense of the term (even if they are in all probability passive participles in form as the outcome of an interesting linguistic development). The simple reason is that intransitive verbs by definition cannot be passive in meaning. In intransitive instances, the term "impersonal relative verb form" is therefore to be preferred. On the other hand, a past tense instance such as rdjj n.f wrrt (type 1 in [section]1.2.2) cannot be a relative verb form, because the past relative verb form exhibits the affix n, as in rdt.n.f "what he has given." In these instances, the term "passive participle" seems to be the only option left.(10) In conclusion, neither "passive participle" nor "impersonal relative verb form" can encompass all the instances of (d). But "agent-less indirect," with reference to...