² ‘The “Epistolary Perfect” in Hebrew letters’ Biblische Notizen [BN] 22 1983 34–40.Google Scholar

³ For example Y. Aharoni in his excellent publication of the Hebrew ostraca from Tel Arad (ktwbwt'rd Jerusalem 1975) thrice vocalized epistolary perfect forms of the verb šlḥ ‘to send’ as participles (Arad 16:1; 21:1; 40:2) and restored one of these forms effaced at the end (40:2) as a participle (šlḥ[m]) rather than as a perfect (šlḥ[m]). Though the knowledge that the Aramaic attestations of the idiom in question had perfect verbal forms was sufficient to correct these particular forms (cf. Pardee Ugarit-Forschungen [UF] 10 1978 310–11 318–19 323–4) an entry in the Hebrew grammars regarding proper epistolary usage such as those that one finds in the classical grammars (see quotes below) would have made Aharoni's task a simpler one (and it would have been possible to avoid repeating Aharoni's error in the English edition; see Dion's, P.-E. review of the English edition Journal of the American Oriental Society ]JAOS] 103 1983 471–471).Google Scholar Such an entry was difficult to write on the basis of Biblical Hebrew alone (a rapid perusal of the letter fragments preserved in the Hebrew Bible yielded only two clear epistolary perfects šlḥty in 2 Kings 5:6 and 2 Chronicles 2:12) but is possible now on the basis of the relatively large number of extrabiblical Hebrew letters. An example from Akkadian: In spite of the many ‘epistolary’ perfects and preterites in other Mari letters Jean, C.-F. restored the last line of Archives Royales de Mari [ARM] II (Paris 1950)Google Scholar text 106 as a present-future verbal form rather than as a preterite. The text with the last line properly restored is:

This restoration (in place of the editor's ụ-ša-[ra-aš-šu]) and translation are based on the assumption that the letter and the caged lion were sent on the same boat.

⁴ Pragmatics Cambridge 1983 74–74.Google Scholar

⁵ Greenough, J. B.Howard, A. A.Kittredge, G.L.B.L., D'Ooge (ed) Allen and Greenough's new Latin grammar for schools and colleges (1983 repr. of 1903 edition) New york301–301.Google Scholar Note that in the second example pridie is an epistolary adverb and corresponds in English usage to ‘I answer all your letters today’. Of equal interest is the statement in C.G. Zumpt's Latin grammar: It is a peculiarity of the epistolary style in Latin that the writer transfers himself to the time at which the letter is read by the person to whom it is addressed; and hence the writer speaks of actions and conditions in the same terms as he would use if he were present at the moment the letter is received. In consequence of this he frequently used the imperfect and perfect where in English we should use present; e.g. Haec scribebam media nocte I write this at midnight (or scripsi haec media nocte when the action is to be described as a completed one and not as going on at the time); Novi nihil nunc erat apud nos siquidem certa tibi afferri vis there are no news here: Quae ad eam diem quum haec scribebam audiveramus inanis rumor videbatur. Dicebant tamen &c. what we have heard till the moment I write this &c. As these preterites are only formal they may be joined with the adverbs nunc etiamnunc instead of which real preterites would require tunc and etiamtum. Comp. Cic. ad Alt. v. 16 4; xvi. 3 6; ad Quint. Frat. iii. 1 2. But this peculiarity is very frequently not observed. (A Grammar of the Latin Language 9^th ed. Eng. transl. by L., Schmitz corrected and enlarged by C., Anthon 3^d ed. New York 1851 §503 [pp. 345–46]) A through-going study of the classical epistolary conventions would be of use to Semitists: just how ‘frequently’ is the peculiarity not observed and to what extent was it considered ‘optional’ (Lakoff)? From the paucity of exceptions to the verbal epistolary conventions in the ancient Semitic texts we have consulted it would appear to us that these conventions may not have been considered ‘optional’ and they were perhaps therefore more firmly anchored in the Semitic verbal systems than were the corresponding forms in the classical languages.Google Scholar

⁶ Cambridge (Mass.) 1963 § 1942 (p. 433).

⁷ Language 46 1970 838–49 quoted from p. 847. Lakoff here uses ‘performative’ in an expanded sense of the term according to which every sentence contains either an implicit or explicit performative verb in the present case ‘(I hereby declare that) I have written to you’. This usage of ‘performative’ is beyond the limits set by us for this term (see discussion in the second part of this paper).

⁸ See Lakoffs other examples and her conclusions p. 848.

⁹ For citations of the texts on which these conclusions are based see the article cited in n.2 above.

¹⁰ Professon John Wansbrough has called our attention to a usage of the Arabic perfect which may be comparable to the older Semitic epistolary perfect used in documents such as treaties. It is described as follows in Wright's grammar; ‘An act the occurrence of which is so certain that it may be described as having already taken place. This use prevails in promises treaties bargains &c. ⃛’ W., WrightA grammar of the Arabic language translated from the German of Caspari and edited with numerous additions and corrections revised by W.R., Smith and M.J. de, Goeje third ed. Cambridge 1955 II § 1(e) (p. 2–2).Google Scholar Note that Wright maintains this usage as separate from what now would be called the ‘performative’ perfect: ‘An act which is just completed at the moment and by the very act of speaking’ (§ 1 [d] p. 1). Wright does not however specifically mention letters in this section. When our colleague J. Stetkevych was asked for an example from Classical Arabic epistolary usage he found the following text: (standard greeting formulae then) faqad ba'aṯtu 'ilaykum ‘abdan min'ibâdi allahi⃛ fasmàû lahu wa’ ṭîû'amrahu ‘I hereby send to you one of the servants of Allah⃛so listen to him and obey him’ (AḯmadZakî, ṧafwat ed. Jamharat rasâ'il al-'Arab I Cairo 1971 letter 504 from ‘Alî to the people of Egypt ]p. 480–480]). A study of epistolary conventions in Classical Arabic would be welcome.Google Scholar

¹¹ With the collaboration of S.D., SperlingJ.D., Whitehead AND P.-E., DionHandbook of ancient Hebrew letters Chico 1982.Google Scholar

¹² RS = field number of the Missin de Ras Sharma. KTU = M., DietrichO., LoretzJ., SanmartinDie keilalphabetischen Texte aus Ugarit KevelaerNeukirchen-Vluyn 1976.Google Scholar

¹³ This is of course the perspective and the formulation of the standard Akkadian letter which Ugaritic shares with Akkadian and which is rare in later West Semitic epistolary formulae. On the basis of the Akkadian epistolary formulary which is in many respects similar to that of Ugaritic one would expect rgm to be an imperative for Akkadian uses the messenger formula qibima ‘say’. There is at least one Ugaritic letter that confirms this conclusion: in the one letter clearly from a female rgm is not marked for feminine gender as one would expect if the form were perfect or participial (RS 17. 435⁺:2 = KTU 2. 36; cf. Pardee's re-edition in AFO 29–30 1983–84 321–9). On the other hand there are examples of even the rgm/qabû formula being stated as epistolary perfects/preterites rather than as ‘present’ imperatives. The syntax of one Ugaritic letter seems to indicate an exceptional use of rgm as a perfect: C., VirolleaudLe Palais Royal d' Ugarit II(= PRU II) Paris 1957Google Scholar text 10(/RS 16.264 = KTU 2.26) begins tḥm rgm mlk which it appears should be translated ‘Message: the king has said’ (thought it is not impossible that rgm is a noun: ‘Message: word of the king’). This formula is as nearly as we can determine from unvocalized texts unique at Ugarit and if verbal it has parallels in the sixteen cases of Akkadian preterite forms of the verb qabû in the Amarna letters (see below n.35) and in the one Phoenician letter extant today which unites the messenger formula expressed by means of the verb 'mr with the epistolary perfect of the ame verb: 'mr Γḥty 'ršt 'mr 'ḥtk bš: ‘Say to my sister: (Thus) says your sister Basu’ (see text and bibliography in Handbook [cited above n.11] 165–8 cf. also p. 121). This feature of Phoenician may now have been located in an early dialect of Hebrew as well for several scholars believe that the jar inscriptions from Kuntillet Ajrud which contain the 'mr⃛'mr formula are scribal practice texts with epistolary formulae. See most recently J.C., Greenfield ‘Notes on the Phoenician Letter from Saqqare’ Orientalia n.s. 53 1984 242–4.Google Scholar Though he may be correct these non-Ugaritic usages of perfect/preterite forms at least call into question J.-L. Cunchillo's conclusion regarding PRU II 10: 1–2 that the unique formulary indicates an archival rather than an epistolary context for this document (‘Un billet du roi d' Ugarit (KTU 2: 26)’ Materiali lessicali ed epigraphici I Rome 1982 25–9 esp. 26–26).

¹⁴ The form is termed ‘epistolary’ because the sender is representing himself as having fallen at the time of writing the letter. For the performative aspect of this formula. see below discussion of performativity. On the etymology of ql see W. von, Soden ‘Kleine Beiträge zum Ugaritischen und Hebräischen’ in Hebräische Wortforschung: Festschrift zum 80. Geburtstag von Walter Baumgartner (Supplements to Vetus Testamentum 16) 1967 291–300 esp. pp.295–6Google Scholar and references cited in Pardee's note on this verb in Bibliotheca Orientalis 34 1977 4. For the ‘Prostration formula’ in Ugaritic and in the Akkadian texts from Ras Shamra see S.E., Loewenstamm ‘Prostration from Afar in Ugaritic Accadian and Hebrew’ Bulletin of the American Schools of Oriental Research 188 1967 41–3;Google ScholarO., Kaiser ‘Zum Formular der in Ugarit gefundenen Briefe’ Zeitschrift der deutschen Palāstina-Vereins 86 1970 10–23 esp. pp. 21–2;Google ScholarF.B., Knutson in Ras Shamra parallels III Rome 1975 421–2;Google Scholar and A.L., Kristensen ‘Ugaritic epistolary formulas: a comparative study of the Ugaritic spistolary formulas in the context of the contemporary Akkadian formulas in the letters from Ugarit and Amarna’ UF 9 1977 143–58 esp. pp. 147–50 157 A comparison with standard Akkadian and sub-standard Amarna Akkadian will also be found here belwo; see examples VI-IX.Google Scholar

¹⁵ See also examples II and III below and Kristensen UF 9 1977 150–1. The greeting formula with šulmu (lū šulmu) is the standard Middle Babylonian form and is in usage at Ras Sharma e.g. J., NougayrolLe Palais Royal d' Ugarit III(= PRU III Paris 1955) p. 5–5Google Scholar RS.15. 14: 5–8 (this and the other Akkadian texts cited in this study are cited according to the various editors' transliterations and translations): For another example see text V below. For an example of the lū stative greeting formula in Amarna see text X. For Old Babylonian usage see text XV and n.46.

¹⁶ One relatively complete example of the Akkadian nominal šlm status-formulae from Ras Shamra may be given: J., NougayrolLe Palais Royal d' Ugarit VI[= PRU VI] Paris 1970 text 20:3´–5´Google Scholar: Other examples of the formulae in various forms and in varying stages of completeness are: PRU III pp. 13–14 RS 16. 111: 7–9; pp. 15–16 RS 15. 33: 13–15; J., NougayrolLe Palais Royal d' Ugarit iv (= PRU iv) Paris 1956 pp. 196–7 RS 17. 78: 6–7; pp. 221–3 RS 17. 383: 6–9; pp. 223–5 RS 17. 422: 8–10;Google ScholarJ., Nougayrol et at. Ugaritica v Paris 1968 texts 33: 3´-4´Google Scholar; 44: 7–9; 48: 6–7; 49: 19–20; 54:6–8; 57: 9–14´; PRU vi 1: 7 (cited below as text VI); 9: 4–7; 14: 6–8; 15: 6–9. Note Nougayrol's switch from an interpretation of minummé as an interrogative in his earlier publications to that as an indefinite pronoun in PRU vi. This is in line with its description by the Chicago Assyrian dictionary (Chicago 1956–) as an indefinite pronoun (M II 97).

¹⁷ See for example F.M.T., BöhlDie Sprache der Amarnabriefe mit besonderer Berücksichtigung der Kanaanismen Leipzig 1909;Google ScholarW.L., Moran A syntactical study of the dialect of Byblos as reflected in the Amarna Tablets Dissertation Johns Hopkins 1950; and the many studies in more recent years by A.F., Rainey e.g. ‘Reflections on the suffix conjugation in West Semitized Amarna Tablets’ UF 5 1973 235–62; ‘Morphology and the prefix-tenses of West Semitized El-'Amarna Tablets’ UF 7 1975 395–426.Google Scholar

¹⁸ See further discussion below at example XV.

¹⁹ For bibliography on and examples of the ‘disclosure formula‘ see the Handbook (ref. above n. 11) p. 127 and BN 22 1983 36.

²⁰ As suggested in the article cited in note I probably restore š ilt cognate accusative of' iš al at the beginning of this line.

²¹ KTU 5.9 IV 1–3 tḥm ḏmn l‘iṯtl ’ Message of Whomever to ‘Iṯatilli’.

²² Longer greeting formulae are known in letters from inferior to superior e.g. RIH 78/3 + 30 (Bordreuil, P.Caquot, A. ‘Les textes en cunéiformes alphabétiques découverts en 1978 á Ibn Hani ’ Syria 57 1980 343–73 esp. 356–8CrossRefGoogle Scholar). Longer greeting formulae between equals are also known from Akkadian letters from Ras Shamra e.g. Ugaritica v 55: 4–12; 57: 2´–8´.

²³ The term applied to this text is W. G. E. Watson's (personal communication).

²⁴ The reading ymm in both PRU II and KTU is incorrect. What appears to be a second m is in fact a ṭ in the partial alphabet which was written on the right edge of the tablet. The end of line 3 wraps around the right edge of the tablet. When the scribe was writing the alphabet on the right edge he carelessly wrote ṭ over the m of ym then seeing the mess he had made he wrote another ṭ below the first. The first ṭ is indeed a sign made up of three wedges and not the two-wedged sign copied by Virolleaud as his second m (PRU II 39).

²⁵ On ‘ desires’ and ‘ requests ’ as an epistolary topos see the brief section in the Handbook (reference above n. 11) based on a suggestion of P.-E. Dion (§ 7. 13 pp. 181–2).

²⁶ Three possible examples from contexts too difficult for certain classification: RS 15. 98 = PRU II 21: 1 yblt RS 22. 03: 2 ṯlẖt (P. Bordreuil ‘ Cunéiformes alphabétiques non canoniques I). La tablette alphabétique senestroverse RS 22. 03’ Syria 58 1981 301–10) and RS 17. 435+: 6 štn[t] (D. Pardee ‘ The letter of Puduhepa: the text ’ AfO 29–30 1983–4 321–9 esp. 321 325 327).

²⁷ We do not pretend that the Ugaritic and Akkadian verbal systems are alike; simply that one can observe the stylistic conventions used in each language and in terms of the verbal system of each language say something about what one would expect in the other language; e.g. the Akkadian preterite will frequently correspond to the perfect (qatala) in Ugaritic prose and the correspondence of Ugaritic qlt with Akkadian amqut for example is thus an expected one (see examples VI-IX below). From this we can extrapolate that if Akkadian uses preterite (and perfect) verbal forms in a letter for the sending of that very letter Ugaritic would probably use perfect forms. This argument is taken one step further below where ‘Canaanite’ perfect forms which replace expected Akkadian preterite and perfect forms are cited.

²⁸ See below n.42.

²⁹ On the interrelationship of the preterite and perfect see below n.42 and conclusion.

³⁰ amqul (maqdtu) is the normal verb of the prostration formula and all forms cited here are from that root. (We would question P. Artzi's analysis of the much rarer form imqut as a first person form: ‘Some unrecognized Syrian Amarna letters (EA 260 317 318)’ Journal of Near Eastern Studies [JNES] 27 1968 163–71CrossRefGoogle Scholar esp. p. 165 n. 17 and p. 167. Is this not a third person form representative of the West Semitic tendency to shift back and forth between first and third persons when referring to oneself and between second and third persons when referring to one's interlocutor? as is iqbi see above n. 13 and below n. 35). On the alternate form ušẖeẖin see Nougayrol Ugaritica v p. 135 n. I and W. von Soden Akkadisches Handworterbuch Wiesbaden 1965–81 1263.Google Scholar

³¹ For the rare form imput see preceding note.

³² J.A., KnudtzonDie El-Amarna-Tafeln Leipzig 1915.Google Scholar

³³ cf. A. F. Rainey ‘E1-‘Amarna notes’ UF 6 1974 295–312 esp. 305. N. Na'aman uses the forms maqtāti/jmaqtili as one argument in favour of Amarna letters 63–5 and 282–4 being from the hand of a single scribe and from Palestine (EA 138: 4 and 336: 5 are not included in the argument): ‘The origin and historical background of several Amarna letters’ UF 11 1979 673–84.

³⁴ These occur both in letters between the ‘great kings’ (e.g. EA 19: 71 al-la-par 85 ul-te-b[il] [Tusratta of Mitanni to Amenophis III]) and in those of minor Canaanite rulers to the Egyptian pharaoh (see indexes in Knudtzon [reference n. 32]note especially the six occurrences of istapar [plus several more wholly restored] used in letters from Rib-Addi of Byblos to replace qab00FB;: EA 108: I; 116: 1; 119: 1; 121: 1; 122: 1; 123: 1). On the Ras Shamra usages of šapšru and šūbulu see below n. 42. For examples from another Syrian source see Kāmid il-Lōz 69: 277: 4 and 69: 279: 4 uš-te-bi-la-ku ‘I hereby send to you⃛’ (Edzard, D. O. ‘Die Tontafeln von Kāmid el-Lōz’ in Edzard, D. O.Hachmann, R.P., MaibergerG., MansfeldKamid el-Loz-Kumidi: Schriftdokumente aus Kamid el-Loz (Saarbrūcker Beitršge zur Altertumskunde 7) Bonn 1970 55–62).Google Scholar

³⁵ EA 68: 2; 74: 1; 76: 1; 78: I; 79: 1; 81: I; 83: 1; 88: 1; 89: 1; 92: 1; 105: 1; 106: 1; 107: 1; so fully restored in 70: 1; 75: I; 91: 1; 109: 1; 137: 1 (all from Rib-Addi of Byblos); 260: 2; 317: 3; 318: 4 (writer's name not given in 260; 317 and 318 are from Dagan-takala who was of Syrian origin according to Artzi JNES 27 1968 163–71the form iqbi is one of Artzi's arguments for grouping EA 260 with the other two see p. 165). See also E., SalonenDie Gruβ- und Hōflichkeitsformeln in babylonisch-assyrischen Briefen (Studia Orientalia edidit Societas Orientalis Fennica xxxvm) Helsinki 1967 62–3.Google Scholar

³⁶ uššuru in EA 34: 52 and 120: 41 (ušširti) cf. also the Akkadian perfect forms of this root in example X; šapāru once in EA 256: 31 (šaparti); we have found no examples of qabû as the Canaanite epistolary perfect.

³⁷ The volitive notion is of course provided by the volitive particle lū (cf. W. von, SodenGrundrifi der akkadischen Grammatik [GAG] Rome 1969 §81b [p. 104]). The stative greeting formula is rare in Middle Babylonianwe are assuming that lū šulmu consists of lū + noun and is not a stativized noun; in any case it is an invariable frozen form. On Old Babylonian see n. 46.Google Scholar

³⁸ G., DossinArchives Royales de mari I Paris 1950.Google Scholar

³⁹ See the bibliographical overview by Pardee, and J.T., Glass ‘Literary sources for the history of Palestine and Syria: the Mari Archives’ Biblical Archaeologist [BA] 47 1984 88–99.CrossRefGoogle Scholar

⁴⁰ J.R., KupperArchives Royales de Mari III paris 1950. There is a recent English translation of this letter by J.M., Sasson in BA 47 1984 p. 118 n.8.Google Scholar

⁴¹ See above under example I on šlm in Ugaritic and below discussion of example XV.

⁴² For the sequence preterite to perfect regular in the Mari letters see A., FinetL'accadien des lettres de Mari Brussels Paris 1956 256–7;Google Scholar on the so-called consecutio temporum in Old Babylonian in general see von Soden GAG §80d (p. 104) §156c (p. 209). This feature is apparently lost in the documents of Middle Babylonian origin from Ras Shamra for there the distribution is primarily lexical. Properly ‘epistolary‘ forms of šapāru and šūbulu are virtually always perfects: PRU m pp. 12–13 RS 11. 730: 10; PRU iv p. 188 RS 17. 292: 18; p. 194 RS 17. 385: 8; p. 214 RS 17. 152: 9; p. 294 RS 19. 70: 11; Ugaritica v 22: 27; 26: 17; 28 II 2´ 20´; 37: 23; 41: 8 14; 42: 9; 48: 17; PRU vi 4: 9; 6: 15; 7 A 10 B 8; 8: 16; 14: 10 11; 16: 9'; the only epistolary preterites of which we are aware are ù-še-bi-la-ku-ni in RS 6. 198: 14 (F., Thureau-Dangin ‘Une lettre assyrienne à Ras Shamra’ Syria 16 1935 188–93. written in Assyrian dialect!)Google Scholar and [lu-ù ù–] še-[b]i-la and u-se-b[i-la] in Ugaritica v 57 III 4´ 5´. Epistolary forms of tarādu on other hand are always preterite: PRU III pp. 11–12 RS 15. 18: 12; PRU iv pp. 196–7 RS 17. 78: 9; pp. 217–18 RS 17. 143: 28; pp. 223–5 RS 17. 422: 31; Ugaritica v 43: 8.

⁴³ G., Dossin in C.F.-A., SchaefferUgaritica I Paris 1939. Sasson has also recently translated this letter: BA 47 1984 p. 116 n.2.Google Scholar

⁴⁴ F.R., KrausAltbabylonische Briefe [ABB] v Leiden 1972.Google Scholar

⁴⁵ As this is the first properly epistolary form in this letter the preterite form is acceptable; see above n.42 and below discussion to example XVI.

⁴⁶ Stative forms of šalāmu are regular in greeting formulae in Old Babylonian letters (see Salonen Gruβ-und Hōflichkeitsformeln [reference above n.35] 14–15 with many examples in following pages).

⁴⁷ Judging from Kraus's translation he is taking ak-nu-kam-ma uš-ta-bi-la-ak-kum as constituting hendiadys. This would be an argument in favour of our analysis of the first verb as an epistolary preterite. It is not impossible however that ak-nu-kam-ma is simply a statement of action completed prior to the time of writing (it is the money that is stated to be sealed not the letter). One cannot argue that the first verb form must be preterite according to consecutio temporum (see above n. 42) for there are instances of the first verb being a perfect form (see discussion of example XVI) and other instances of the epistolary sequence preterite⃛ preterite e.g. with these same verbs ARM x (G. Dossin Paris 1967 1978) text 50: 31–3:

There being no indication to the countrary we assume that hte items in question (lock of hair and hem of garment as proofs of ‘prophetic’ authenticity) were sent with this letter.

⁴⁸ The term is Whiting's; it is discussed in detail in the work cited in n. 49 but see already R.M., Whiting ‘The R Stem(s) in Akkadin’ Orientalia n.s. 50 1981 1–39 esp. p. 2 n. 7.Google Scholar

⁴⁹ R.M., WhitingOld Babylonian letters from Tell Asmar Chicago forthcoming (as Assyriological Studies 22) text 20.Google Scholar

⁵⁰ On anumma plus perfect see GAG §80b (p. 104). The coincidence of anumma and perfect was already remarked by A., Goetze in the first full-scale treatment of the infixed -t- perfect: ‘The t-Form of the Old Babylonian Verb’ JAOS 56 1936 297–334Google Scholar esp. p. 309 ‘⃛sentences beginning with anumma contain regularly a t-form’. This usage is not however invariable: just as consecutio temporum is not always observed (see Finet reference above note 42) so there are instances of anumma + preterite e.g. ARM ii 107: 26—7.

Note the editor's present-tense translation we assume relfection an understanding of ù-śa-bi-lam as what we are noe calling an epistolary preterite.

⁵¹ This example is cited from the transliteration and translation prepared for the Chicago Assyrian dictionary by Lewy, J.. Lines 17–19 may be found in transliteration in K., HeclerGrammatik der KÜltepe-Texte Rome 1968 £76g (p. 127).Google Scholar

⁵² G., ContenauTrente tablettes cappadociennes (Musée du Louvre Département des Antiquités Orientales: Textes cunéiformes vol. 4) Paris 1920.Google Scholar

⁵³ The full context of these lines may be indicated by citing B. Landsberger' translation; note especially ‘hiermit’ in the translation of line 19 (Assyrische Handelskolonien in Kleinasien aus dem dritten Jahrtausend (Der Alte Orient 24/4) Leipzig 1925 21): Brief des Assur-nāda an Assur-İdi: 10 Minem gelāutertes Silber das dem Alahum und mir gemeinsam gehört hat NN. dem Alahum und dem Assuѓ-taklāku geschickt. Ich gab folgenden Auftrag: ‘Die Prozente (nishātum) und die Provision (šadūatum) fügt hinzu und schicket sie unserm Vater und den Vertretern des Alahum!’ Du bist mein Vater Du bist mein Herr für das ganze Silber und Gold das ich Dir hiermit schicke kaufe ein: Hálfte gutes Blei Extraqualitát für die andere Hálfte Stoffe.

⁵⁴ In spite of the use of the term ‘ performative’ in the long quotation from Lakoff given above her name is omitted from the following list for the reasons stated above in n. 7.

⁵⁵ ‘Zu den Grundfragen der Aspekttheorie’ Indogermanische Forschungen [IF] 53 1935 280–300 esp. 291–5. He is reacting to a discussion of A. Debrunner (‘ Sprachwissenschaft und Klassische Philologie’ IF 48 1930 1–25) in which the Greek term actually used is έγρáψaμευ (p. 17).

⁵⁶ GAG § 80c (p. 104).

⁵⁷ ‘Der Koinzidenzfall im Akkadischen’ Zeitschrift der Deutschen Morgenldndischen Gesellschaft Supplementa I Teil 1 1969 148–52.Google Scholar

⁵⁸ Untersuchungen zur Formensprache der babylonischen ‘Gebetsbeschworungen’ (Studia Pohl. Series Maior 5) Rome 1976 192–201.

⁵⁹ ‘Zur Deutung der SI.BI-Klausel in den spätaltbabylonischen Kaufverträgen aus Nordbabylonien’ Die Welt des Orients 9 1977–78 206–12 esp. pp. 208–9 n. 6.

⁶⁰ ABB ix 1981 text 114 note b (p. 75); 151 note c (p. 97).Google Scholar

⁶¹ Reference above n. 11.

⁶² Biblica 65 1984 267. Note that in Pardee's BN article (which appeared between the Handbook and Lawton's review thereof) the term ‘performative’ was used but only applied specifically to the blessing formula (p. 35 n. 8). Lawton's contention that there is ‘nothing specifically “epistolary”’ about the epistolary perfect was already countered in the BN article (viz. the term is descriptive of one usage of the perfect not of an overtly marked morphology or syntax). For the usefulness of the grammatical category for non-speakers of Biblical Hebrew and of Old Babylonian Akkadian the examples cited here above in note 3 should be sufficient indication.

⁶³ Reference above n. 4 pp. 226–83.

⁶⁴ Semantics Cambridge 1977 II 725–45.

⁶⁵ See n. 63.

⁶⁶ Koschmieder is not to be found in the bibliographies of Lyons and Levinson. Contrast German-language treatments e.g. W. Gross Verbform + Funktion. wayyiqtol für die Gegenwart St. Ottilien 1976 esp. pp. 49–50; H. Schweizer Metaphorische Grammatik St. Ottilien 1981.

⁶⁷ ‘Durchkreuzungen von Aspekt- und Tempussystem im Präsens’. Zeitschrift für slavische Philologie 7 1930 341–58. His earlier Zeitbezug und Sprache (1929) is unavailable to us.

⁶⁸ IF 53 1935 288.

⁶⁹ Zur Bestimmung der Funktionen grammatischer Kalegorien (Abh. Bayerischen Ak. der Wiss. Phil.-hist. Abt. N.F. 25) 1945 esp. pp. 6 28 29. This article is referred to and the term ‘Effektivus’ is used again in Die noetischen Grundlagen der Syntax (Sitz. der Bayerischen Ak. der Wiss. Phil.-hist. Kl. Jg. 1951 Ht. 4 esp. pp. 27–8.

⁷⁰ That the description of concurrent acts may not be termed ‘performative’ according to more modern presentations is clear from Levinson's example 'I now beat the eggs till fluffy’. He terms this ‘simply a report of a concurrent action’ (= Koschmieder’s ‘Berichtsprasens’!) and goes on to state that ‘we shall need other criteria [than the first person verbal form] as well if we are to isolate performatives alone’ (Pragmatics [reference above n. 4] p. 232).

⁷¹ ‘Other minds’ in Logic and reality (Aristotelian Society Supplementary Vol. 20) 1946 148–87 quotation from p. 174.

⁷² Edited from lecture notes by J. O. Urmson and M. Sbisa Cambridge MA (second ed.) 1962 1975.

⁷³ La Philosophic analytique Paris 1962 271–304 quotation from p. 271; an English translation (from the French publication not taken from Austin's [presumably English] original) by G. J. Warnock appeared in Philosphy and ordinary language ed. C. E. Caton Urbana II. 1963 22–54 and was reprinted in The philosophy of language ed. J. R. Searle Oxford 1971 13–22.

⁷⁴ Lectures IX–XII (pp. 109–64).

⁷⁵ ‘De la subjectivité dans le langage’ Journal de Psychologie 55 1958 257–65 quotation from p. 265; reprinted in E. Benveniste Problémes de linguistique générate I Paris 1966 258–66.Google Scholar

⁷⁶ ‘La philosophic analytique et le langage’ Les études philosophiques 1963 3–11 terms defined on pp. 7–8; reprinted in Problémes (reference preceding note) pp. 267–76.

⁷⁷ According to the introduction to How to do things with words (p. vi of the second edition cited in n. 72) the idea was first hatched in Austin's mind in 1939.

⁷⁸ References above nn. 57 and 58.

⁷⁹ Rather than to refer to the ‘constatation’ as accomplished with the sending of the letter it would be more precise to say that the gap between the dictation and the reading of the letter is not taken into account in the formulation of the letter since it is the reader's tense/aspect perspective that is adopted. Compare the epistolary time-line in Wilke's note (reference above note 59): there is no reference in letters we have consulted to the time-gap between the second and third arrows in that time-line. Thureau-Dangin in his commentary on the form ú-še-bi-la-ku-ni cited above in n. 42 (RS 6. 198: 14) makes the following observation: ‘La lettre n'étant en principe que le texte du discours que le messager doit tenir á son arrivée le moment oii elle a été écrite et expédiée est envisagé au passé’ (Syria 16 [193] 192). He thus describes the epistolary perspective from the point of view of the reading of the letter not of the sending of the letter—see above our comment on the Ugaritic letter cited as example I. (For similar statements with regard to Mari letters see Thureau-Dangin RA 38 [1941] p. 41 n. 6 and Orientalia n.s. 12 [1943] p. 112 n. 2). More recently D. Arnaud has explicitly compared Akkadian usage with classical usage:‘… le rédacteur de la lettre se transporte idéalement au moment oú le destinataire l'aura en main. C'est l'imparfait épistolaire de nos grammaires latine’ (Syria 59 1982 p. 104 n. 26). We must note however that the letter to which Arnaud's comment makes reference does not contain an epistolary perfect/preterite if the text is analysed according to the classical definition: táš-pu-ra in line 18 for example refers to the letter to which the present author is responding and would correspond therefore to the pluperfect in Latin usage. Even more recently Arnaud has referred to another verbal form as a ‘passé épistolaire’ (Syria 61 1984 p. 16 n. 2: [as]-tap-pár-ma in Ras Ibn Hani 81/4: 12); this example does fit the classical definition for it refers to persons sent with the letter itself.

⁸⁰ Mayer first cites Heimpel and Guidi then sums up their position as follows: Die Verfasser zeigen anhand von Beispielen aus altassyrischen und altbabylonischen Briefen und Urkunden dass zu dieser Zeit und in diesen Textgattungen fü den KF [Koinzidenzfall] das Práteritum verwendet wird (demgegeniiber konstatiere das Perfekt in der Briefformel ‘aštapram’ eine mit der Absendung des Briefes vollendete Handlung: ‘ich habe dir nunmehr geschrieben’ bzw. ‘wegen der Sache X schreibe ich nunmehr’).

⁸¹ Because of this clear evidence for the partial interchangeability of the perfect and preterite forms to express epistolary acts in Akkadian it appears to us that the appearance of one or the other of the two forms must have depended on the respective nature of each of the forms (see above at text XVI). Any general explanation of epistolary verbal expressions which is based on the semantic characteristics of one form to the exclusion of the other is therefore ruled out by the very distribution of the forms. We refer particularly to Wilke's explanation (reference above n. 59) which is based on the Heimpel/Guidi distinction (preterite = performative perfect = our ‘epistolary’) and according to which the perfects in epistolary usage are examples of futurum exaclum. Not only is this explanation ruled out by the appearance of both the perfect and the preterite in properly epistolary usage but it appears to us the weight of the comparative evidence from Greek Latin and Slavic is in favour of the explanation according to which the writer adopts the perspective of the reader. Note that Thureau-Dangin's examples to which we have made reference in n. 79 included both preterite and perfect forms. Be it added in passing that Wilke's claim that his explanation of the perfect forms is buttressed by the regular occurrence of a volitive form after the perfect (‘Bei den bisherigen Behandlungen dieser Erscheinung ist m.E. nicht I genügend beachtet worden daB auf diese Perfekta (stets mit Ventiv) ganz überwiegend asyndetisch I eine Wunsch- oder Befehlsform folgt’) is severely weakened by the many cases in which there is no volitive form following the perfect. This is incontrovertibly the case when a letter ends with an epistolary perfect as in ARM III 28: 14; 52: 14; 59: 21; 62: 18 (to choose examples only from that i volume; cf also Stol's example ABB IX 151: 28–30).

⁸² It is perhaps worth pointing out that the perfect/pluperfect distribution in Latin cannot occur in West Semitic because (among other things) there is only one perfective form in that language; and that the Akkadian perfect/preterite forms have functions quite different from the mentioned Latin forms (see above passim esp. at example XVI and nn. 42 45 47 50).

⁸³ Reference above n. 4.

⁸⁴ Of Mayer's West Semitic examples there are several debatable cases e.g. Biblical Aramaic s*lahnā wehôda'nā (Ezra 4: 14) are examples of the epistolary perfect not of Koinzidenzfall (p. 190 reference above n. 58). On p. 191 Mayer analyses the Ugaritic prostration formula as Koinzidenzfall; on our disagreement with this see here below. On the other hand his other example from Ugaritic prose ‘ankn rgmt (PRU v 8: 6) may be correct for the regular epistolary formula with rgm precedes in lines 1–5. Unfortunately the context is not certain and no analysis of line 6 is certain. (This tablet was unavailable for collation when Pardee was in Syria).

⁸⁵ In the original draft of this paper; on ql see already BN 22 1983 39–40.

⁸⁶ The stative of bašú is rare and the present-future form is used in contexts wherein statives of other verbs are found (GAG § 78b [p. 102]).

⁸⁷ GAG §77d(p. 101).

⁸⁸ P., JoüonGrammaire de I'hébreu biblique Rome 1923 §41b (p. 95).Google Scholar

⁸⁹ Zeitschrift für Assyriologie 62 1972 124.

⁹⁰ UF 6 1974 305.

⁹¹ How to do things with words (reference above n. 72) p. 81 cf. p. 85.

⁹² BN 22 1983 40 note 27 and here above n. 84.

⁹³ Grundriss der vergleichenden Grammatik der semitischen Sprachen II Berlin 1913 77ca p. 153). This section is not cited by Mayer in his several references to Brockelmann (pp. 187–8 reference above n. 58). This knowledge was certainly behind J. Friedrich's description of the phenomenon in Hittite (Hethitisches Elementarbuch (second ed.) Part I Heidelberg 1960 §262a [p. 139]; cf. already the first edition 1940 §272a [p. 79]) in the following terms: ‘Im Briefstil kann sich der Absender auf den Standpunkt des Empfangers stellen und statt des Prasens das Prateritum gebrauchen’ since he continues: ‘Derselbe Gebrauch nicht nur im Lateinischen sondern auch im Akkadischen’ (with references to the comments of Thureau-Dangin cited here above in n. 79). While the observations of Thureau-Dangin were thus still known to Hittitologists as late as 1960 they have as far as we have been able to determine disappeared from Assyriological lore (cf. Arnaud's statements given above in n. 79 made without reference to Thureau-Dangin). Had they not the greater part of the present article would have been unnecessary.

⁹⁴ Koschmieder cited the Hebrew

‘ich segne ihn hiermit’ in his first article on Koinzidenzfall (Zeitschrift für slavische Philologie 7 1930 354) and did so in all subsequent articles on the topic.

⁹⁵ Koschmieder linked the two from the very beginning as is evident from the titles of the articles cited in nn. 67–9.

⁹⁶ It is probably because both epistolary conventions and performatives are expressed by perfectives that Koschmieder once (only!) confused the two categories (see above and n. 55 for reference). On aspect in Biblical Hebrew see most recently R. Bartelmus HYH Bedeutung und Funktion eines hebräischen ‘Allerweltswortes’ St. Ottilien 1982 and Pardee's review in Catholic Biblical Quarterly 47 1985 107–10.