¹ b^egib^eon b^ekar YHWH has been reconstructed on the basis of v. 9, which reads wayyoqi^cum bāhār lipně YHWH.

² The Treaty between Israel and the Gibeonites, BA 27 (1964), 99f For other examples of this type of treaty curse, see his Common Trends in Curses of the Near Eastern Treaties… ZAW 75 (1963), 158–63, 168–70Google Scholar.

³ Doctrines of Causality in Hittite and Biblical Historiography. Parallel, A, VT 5 (1955). 1–12Google Scholar.

⁴ F. C. Fensham, op. cit. Three curses can be singled out here: natural calamities shall overtake the transgressor (drought, famine, etc.); the progeny of the transgressor shall be obliterated; and the corpse of the transgressor will be exposed. All three of these curses are involved in 2 Sam. 21:1–14. That many of the prophetic curses are derived from treaties has been persuasively argued by Hillers, D. R., Treaty Curses and the Old Testament Prophets (Rome, 1964)Google Scholar; and there appears to be no reason to refuse to apply similar reasoning to events such as that contained in 2 Sam. 21:1–14.

⁵ This Hittite parallel is the Plague Prayer. Cf. Keilschrifturkunden aus Boghazkoi (Berlin, 1921–1944)Google Scholar. Cf. ANET, p. 395.

⁶ David's Monarchy and the Gibeonite Claim: 2 Sam. 21:1–14, PEQ (1955), 167Google Scholar. However, he suggests the wrong kind of ritual.

⁷ A. S. Kapelrud (King and Fertility. A Discussion of 2 Sam. 21:1–14, Interpretationes ad Vetus Testamentum Pertinentes Sigmundo Mowinckel Septuagenario Missae [Norway, 1955], 112–22Google Scholar; King David and the Sons of Saul, Supplement to Numen IV [1959], 294–301Google Scholar) centers his discussion on the question of famine and the intimate connection (in Israel as elsewhere in the Ancient Near East) between king and fertility. He explains the slaying of Saul's sons as part of an ancient fertility rite that satisfies David's political and religious aspirations. H. Cazelles, op. cit., interprets this passage similarly. The covenantal interpretation, which we have adopted, is perhaps more plausible, since there is no evidence, as far as I can tell, that the agricultural rites of Canaan included human sacrifice to reenact Mot's death.

⁸ Cf. Fitzmyer, J. A., The Aramaic Inscriptions of Sefire I and II, JAOS 81 (1961), 207Google Scholar, for a discussion of this passage and its reconstruction.

⁹ Some of my remarks on this pericope have been based on the analysis presented by Michael Coogan in the Hebrew 200 Seminar, Harvard, Dec. 19, 1967. His evidence for the covenantal aspects of this passage include reference to the verb smd, the interpretation of this incident by Hosea, and other valuable points. It may be appropriate here to quote M. COOGAN on the covenantal aspects of smd: “Van Zyl suggests that the use of smd here is to be compared with its use as a legal term in several Akkadian texts from Ugarit, where there is a formulaic use of the root to express the transfer of property (The Moabites [Leiden: E. J. Brill, 1960], 123, n. 6). We can on the basis of several texts reconstruct the formula: ina “samsi umi^mi X sa-mi-id ana PN (adddriti), ‘in broad daylight’or sometimes ‘in the presence of the king’] X [a vineyard, house, land, etc.] is attached to PN (forever)’ (see Nougayrol, Jean, Textes Akkadiens et Hourites des Archives Est, Ouest et Centrales, PRU III [ = MRS VI] [Paris: Imprimerie Nationale, 1955]Google Scholar, Texts 15.37 [P. 35]! 15–182 [p. 35], 16.207 [p. 109], 15.136 [p. 121], 16.154 tP- 127L etc-; and NOUGAYROL'S commentary on the formula “Resultats des Actes,” 22Sf.). The property under consideration is transferred for several reasons (sale, royal gift, etc.). This legal terminology from Ugarit suggests that in our passage we are dealing with a legal idiom in the Northwest Semitic milieu, and that its occurrence in Nm. 25 can be interpreted to mean that Israel, formerly the property (e.g., s'gulldh, Ex. 19:5, etc.) of Yahweh, has now in a formal way attached itself to Baal Peor…The verb smd as used in Nm. 25 has therefore as its primary connotation attachment by covenant, with the possible secondary sense of sexual attachment.”

Cf. also the reconstruction of this event by M. NOTH, Israelitische Stamme zwischen Ammon und Moab, ZAW 60 (N.F. 19) (1944), 11–57 (quoted by M. Coogan)Google Scholar.

¹⁰ See footnotes 2 and 3.

¹¹ The problem here about certain occurrences of the root yq^o in classical Hebrew is occasioned by the possible presence of a related root, nq^o “be severed, alienated, estranged,” in biblical Hebrew (as indicated by the massoretic pointing; cf. BDB, p. 668b, s.v., where this root is said to occur in Ez. 23:18.22.28). Besides the hifil/hofal occurrences of yq^c in Nm. 25:04 and 2 Sam. 21:06.09.13, yq^c is listed by BDB (p. 429, s.v.) as occurring in the qal in Ez. 23:17.18, Jer. 6:08, and Gn. 32:26. One may accept BDB's assertion of two roots appearing in the qal in biblical Hebrew (yq^c and nq^c), or one may repoint some of the seven questionable verb forms as either all from the root nq^c or all from the root yq^e:

Of the three possibilities listed above, the MT series, which assumes two roots, ap-pears to be the least likely. These seven occurrences are more likely all from yq^c or all from nq^c, as Ez. 23:18 indicates:

watt^ogal taznitteha watt^cgal 'et 'erwdtdh wtq^c napšl mē^calehd ka'āser nq^ch

napsi me^cal 'dhotdh

When she carried on her harlotry so openly and flaunted her nakedness, my nepes wtq^c from her, as my nepes nq^ch from her sister.

It seems unlikely that this one verse would contain two separate but related roots referring to the same act or state of separation. We would prefer the third possibility, i.e., that all occurrences reflect the root yq^c, meaning “to separate (intransitive).” The pre-exilic spelling probably would have included the waw of the nijal perfect, i.e., nwq^ch, but this waw could have dropped out in later transmission.

With regard to the causative stem of yq' in biblical Hebrew, it has been suggested that we have two more possible occurrences. In 1 Sam. 31:10 we read what happened to Saul's body and armor after his death:

wayydsimu ‘et keldw bet 'astdrot

w^ocet g^ewiyydto tdq^ccu b^ehomat bet sān

They put his armor in the temple of Ashtaroth and they fastened his body to the wall of Beth Shan.

Rudolph Kittel suggested in the apparatus of Biblica Hebraica that one might perhaps read hoqi^cu for taq^ecu. In I Chron. 10:10 we have the parallel to this inci-dent:

wayydsimu ‘et keldw bet ‘elohehem

w^ecet gulgalto taq^ecu bet dagon

And they put his armor in the temple of their gods, and fastened his head in the temple of Dagon.

In this passage, J. BEGRICH similarly suggests an emendation to hoqi^c We would hesitate to accept these proposed emendations since the usage of tq^e meaning “to thrust into, fasten, impale” is a well attested and natural one in biblical Hebrew. Cf., e.g., Is. 22:23.25. Moreover, as S. R. Driver has pointed out (Notes on the Hebrew Text and the Topography of the Books of Samuel [Oxford, 1913], 230f.)Google Scholar, hwqy^c in its certain occurrences in biblical Hebrew refers to an act of execution (upon living bodies, therefore), and thus it may be doubted whether it is safe to restore it here.

¹² On the root waqa^ca cf. Lane, E. W., Arabic-English Lexicon (Frederick Ungar; New York, 1956)Google Scholar, Supplement to vols. 7 and 8, 3057f., s.v. waqa^oa.

Concerning naqa^ca, see Kazimirski, A. De Biberstedj, Dictionnaire arabefrancais (Paris), vol. 2, 1329Google Scholar, 1330. My description of the Arabic custom of naqva is drawn from Smith, W. Robertson, The Religion of the Semites (London, 1927), 641Google Scholar, 491. As this author points out, one can compare this custom with Samuel's execution of Agag to fulfill the herem (1 Sam. 15:33), in which the verb sasap or sasa^o, “to hew to pieces,” is used. Although the evidence is not sufficient, is it not possible that Samuel carries out the herem in this manner because of Agag's transgression of a covenant between the Israelites and the Amalekites?

¹³ Lexicon syriacum (Helix Saxonum, 1928), 601b, s.vGoogle Scholar.

¹⁴ See, for example, Archives royales de Mari (hereafter Armt), Vol. II, 37.11.6. In this letter there is also a reference to another ritual involving a puppy and lettuce, but whether this ritual involved the killing of the puppy is not certainly known. On this last point see G. Mendenhaix, Puppy and Lettuce in Northwest Semitic Covenant Making, BASOR 133(1954), 26–30Google Scholar. Mendenhall discusses its similarity to the passover ritual of Ancient Israel, where, for example, in the J account of Ex. 12:21–23 there is the slaughter of an animal and the use of a plant.

For an analysis of the technical phrase haiaram qatdlum (bent X u Y), cf. M. Noth, Old Testament Covenant-Making in the light of a Text from Mari, The Laws in the Pentateuch and Other Studies (translation of Gesammelte Studien zum Alten Testament [Munich, 1960]), 108–17Google Scholar; , Munn-Rankin, Diplomacy in Western Asia, Iraq 18(1956), 88–91Google Scholar; Dossnsr, G., Les Archives Epistolaires du Palais de Mari, Syria 19, 108fGoogle Scholar. This last article discusses a Mari letter in which Zimri-lim was advised to “kill the ass of friendship” with Idamaraz.

¹⁵ Cf., for example, ARMT I, 37, lines 19–21:

aš-šum na-pi-is-ti- [su]

la-pa-tim awil És-nun-n [a(ki)]

is-pu-ra-am

Concerning your treaty, the master of Eshnunna has written me.

Other examples of the technical use of this phrase are II, 62, 9–10; II, 77, 10–11; II, 77, 7'–8'; II, 77, 5, 12; XIII, 147, 6ff., 26, 27.

¹⁶ Arch, d'hist. du droit orient. 5, 133ft. Quoted by Mccarthy, D., Treaty and Covenant (Rome, 1963), 55Google Scholar.

¹⁷ Diplomacy in Asia, Western, Iraq 18(1956), 91Google Scholar.

¹⁸ Cf. ARMT, Vol. II, 237: “Les parties contractantes mettent ainsi leur vie en jeu; le parjure s'expose a perir par la strangulation dont il a ebauche la geste au moment du serment.”

¹⁹ Cf. , Munn-Rankdj, Diplomacy in Western Asia, Iraq 18(1956), 90Google Scholar.

²⁰ Enuma Elis, VI, 98:

They [the great gods] pronounced among themselves a curse,

Swearing by water and oil to place life in jeopardy (ii-lap-pi-tu napsdti).

²¹ Cf. Fitzmyer, J., The Aramaic Inscriptions of Sefire I and II, JAOS 81, 178–222Google Scholar; The Aramaic Suzereign Treaty from Sefire in the Museum of Beirut, CBQ 20, 444ffGoogle Scholar.

²² Fitzmyer, J., JAOS 81(1961), 190Google Scholar.

²⁸ Cf. E. F. Weidner, Der Staatsvertrag ASsurniraris VI. von Assyrien mit Mati'ilu von Bit-Agusi, Arckiv fur Orientjorschung, Band VIII (1932), 17–34Google Scholar. The transliteration and translation of the quoted passage are found on pp. 18–19. We have, for convenience, utilized here the English translation of Mccarthy, D., Treaty and Covenant, 195Google Scholar, which faithfully represents Weidner's German translation.

²⁴ Cf. the text, translation, and notes on this treaty, presented by Wiseman, D. J., JCS 12 (1958), 124ffGoogle Scholar.

²⁵ Cf. Wiseman, D. J., The Vassal Treaties of Esarhaddon (London, 1958)Google Scholar.

²⁶ Cf., for example, the discussion of L. Koehler, Problems in the Study of the Language of the Old Testament, JSS 1 (1956), 6–7Google Scholar, where he proposes that b'rit comes from the root brh (written br" in 2 Sam. 12:17), “to eat invalid food, or the food of mourners.” B'rit in this interpretation means a food that if eaten by two people at the same time brings them into fellowship with each other. The suggestion is highly implausible. W. F. ALBRIGHT, on the other hand (The Hebrew Expression for “Making a Covenant” in pre-Israelite document, BASOR 121 [1951], 22, n. 6), states that the word b'rit can scarcely be separated from Akkadian biritu, “bond, fetter,” the exact derivation of which is obscure. Also M. Noth, Old Testament Covenant Making…, The Laws of the Pentateuch and Other Studies (Philadelphia, 1966 [English Translation]), 112Google Scholar, denies the Hebrew brh etymology. This rejection of Koehler's etymology is shared also by Kraus, H. J., Worship in Israel ([English translation] Richmond, Va., 1962), 119, n. 78Google Scholar.

²⁷ Cf. Noth, M., Das System der zwolf Stamme Israels (Stuttgart, 1930), 100ffGoogle Scholar.

²⁸ One may wonder whether this statement: “…this shall be done to his oxen,” is not a later softening down of an original “…this shall be done to him.”

²⁹ An objection to our bringing these two passages into the picture might be that the verb used is nth not krt, btr or hwgy^c. However, the importance of these two examples lies not in the actual verb used but in the ritual rationale behind the action. An examination of treaty terminology in our sources shows that a variety of verbs was used to describe the cutting up of the animal used in the covenant ceremony, not all of which became frozen in specialized usages. For example, in Gn. 15–10 we have bdtar battawek. In Jeremiah 34:18 we have karat Hsnayim. In Akkadian we find tabafyu, qatdlu, etc. The objection would perhaps be relevant on the assumption that karat or hwqy^c or other technical words could be the only verbs used in Israelite treaty contexts to describe the ritual ceremony of slaying an animal. Such an assumption appears unnecessary. One may indeed wonder why the likely karat is not actually used instead of ndtah, but its absence does not seem to be a valid argument against a covenantal interpretation of these two passages, just as the use of bdtar does not argue against a covenant ceremony being described in Gn. 15. Only some verbs formed frozen covenantal formulae, and in Israel karat predominated. But in no case did the specialized use of a root in a technical covenantal phrase make it the exclusive vehicle for describing the cuttingup ceremony of covenant ratification.

I should like to thank professor Frank Moore Cross, Jr. for his valuable suggestions in the preparation of this article.