Journal of Hebrew Scriptures - Volume 9 (2009) - Review

A student seeking to set the Hebrew Bible into its ancient Near Eastern context will certainly have to delve deep into the treasures of Ugaritic literature. Faced with the question of how best to introduce a beginner to Ugaritic the authors of this book decided to provide a historical and cultural background (Chapter 1, pp. 1-30) before teaching the language. Even so, “language” is given pride of place in its subtitle and it is on that that the book is mainly focused. They opt for an inductive method of teaching, first from a selection of prose (school texts, letters, administrative and legal documents: Chapters 2-5, pp. 31-116) and then from a selection of verse (Chapter 6, pp.117-148). A concise summary of Ugaritic grammar is provided in Chapter 7 (pp. 149-179). The book concludes with a Glossary (Chapter 8, pp. 180-209), a classified bibliography of suggested supplementary reading (Chapter 9, pp. 210-221) and a very selective index (pp. 222-226), in which a few English proper names and some Ugaritic and Akkadian lexemes are all amalgamated.

Pre-publication readers have lavished fulsome praise on this book. The back-cover includes the phrase “very welcome addition” (from Mark Smith) and the statement “many aspects… make it a must-have” (from Brent A. Strawn). The last two paragraphs on that cover are reproduced as the last two paragraphs on p. i, but with an error in the last line, one of the first indications that the final stages of production were sloppy.

On-line reviews of the book were the first to appear. The first was in JHS in 2008
(http://www.arts.ualberta.ca/JHS/reviews/review290.htm), more or less as soon as the book was published. It was fairly short and in it Charles Halton appropriately applauded the authors for identifying a paedagogic gap waiting to be filled. He noted, but almost en passant, that one or two details needed to be modified, and this elicited an immediate and friendly response from one of the authors dated 20 January 2008: “Thanks for the review Charles. The ‘Primer’ was the culmination of 10+ years of teaching Ugaritic by Joel and myself. If anyone catches any typos/corrections, please send them along to me. At some point, I plan to do a 2nd edition with corrections and updates. Best regards, Bill Schniedewind” (http://awilum.com/?p=491). I am grateful to the editors of JHS for allowing me the opportunity to make this supplementary assessment.

In another on-line review, Robert D. Holmstedt was also generally positive (http://www.bookreviews.org/bookdetail.asp?TitleId=5978&CodePage=5978 or http://prophetess.lstc.edu/~rklein/Doc10/primer.pdf), but he identified errors the nature of which suggested neither authors nor publisher had checked the manuscript with the diligence it deserved. Some are simply untidy but others are of material significance.

The first printed notice was by O. Loretz, who concisely but approvingly wrote “Dem Primer ist Erfolg zu wünschen!”[1] I subsequently wrote a similarly concise but more critical paragraph for the Book List of the Society for Old Testament Study (2008, pp. 236-237). At the time of writing these are the only published assessments I know of, though I have been told that others are waiting in the queues. My further reading of the book has revealed more serious deficiencies which are likely to cause difficulty to a student who works through the book as intended, chapter by chapter. They will need to be considered when preparing the promised second edition.

Frequent typographical mistakes mar Chapter 1. Although some can be corrected easily they dilute one's confidence in the text. There are also matters of inconsistency, the responsibility for which the publisher's reader should share with the author. The format of “Gilgamesh Epic” (p. 9) and of “Gilgamesh epic” (p.10) should be the same; the distance from Ras Shamra to Minet el-Beida is expressed metrically (p. 8, §1.2, line 4: 1 kilometer), but to Ras Ibn Hani imperially (p. 9, line 4: 3 miles), and elsewhere, more convincingly, both styles are found (p. 12 : a medium sized state covering 1,240 square miles [2000 km2]). Careless phraseology can raise questions that are more than cosmetic. Some are reasonably easy to resolve, such as “olives” instead of “olive groves” (p. 8 line 5), but “began under the direction of Claude Schaeffer and his successors in 1929” is an ambiguous way of describing work that was begun under Schaeffer and later continued under his successors (p. 8, §1.2, line 1f.).

Geography emerges as a weak point. To say that Ugarit is located “on the northern coast of the eastern Mediterranean” (p. 5) is misleading. It suggests a site in southern Turkey whereas Ras Shamra is on the northern part of the eastern Mediterranean coast. A map (Fig. 1.1) clarifies the question but it has not been drawn well. White is used for water and shadow for land, so that most of the place names appear in black against a shadowed background. It would have been better to use white for land and shadow for water. The next map (Fig. 1.2) is worse, for the place names are so badly distorted that some are hardly legible. It is also differently orientated and includes Kadesh, Byblos and even the eastern extremities of Cyprus within the kingdom of Ugarit. The third map (Fig. 1.3), a sketch map of the region immediately around Ugarit, is much clearer, though it appears to be a direct copy of the handiwork of Marguerite Yon.[2]  It is conceded that the plan of the tell which shows the find spots of tablets is “based on Yon and O. Pedersen,” but those words should not be taken to refer to a jointly authored book, for there is no such book. The only citation for Yon in the Bibliography is The City of Ugarit at Tell Ras Shamra, but neither there nor in the earlier French edition (Yon 1997) is any such plan to be found. Nowhere in the Bibliography is Pedersén cited, but in Archives and Libraries in the ancient Near East[3] he acknowledges that he had modified for his purpose a plan in Contenson.[4]

A student has good reason to be confused when an unsubstantiated argument is compounded with inconsistency and error. The scribes responsible for the texts of Ugaritic epic literature are said to have enjoyed royal sponsorship (p. 10-11). The evidence for the suggestion is that the scribe Ilimilku “was supported by the patronage of king Niqmaddu.” This is based on translating t‘y.nqmd, a phrase in the colophon of Ilimilku, as “from the patronage of Niqmaddu” (but see G. del Olmo Lete, “Ug. t‘, t‘y, t‘t: nombre divino y acción cultual,”).[5] That translation depends on deriving the admittedly difficult word t‘y from t‘ “offering.” The semantic link between the two words is not immediately obvious, yet no further explanation is given. On the contrary, an associated footnote supports an alternative translation, taking t‘y as a gentilic. If this alternative explanation is to be rejected, arguments for doing so should be given. If it is accepted (or accepted as possible), the original proposition becomes untenable (or uncertain). In the translation of the colophon the name Ilimilku is abruptly changed to Ilimalku (p. 11), so the reader is at a loss to know whether this is a misprint or an alternative vocalisation. A teacher may well, but a student may well not, recognise that the transliteration b‘.trmn in the colophon is an obvious error for b‘l.trmn “lord of trmn.” Similarly, rendering that place-name in English as THRMN before broaching the subject of Ugaritic phonemes will seem inelegant to an expert and frustrating for a novice. Here was an ideal opportunity to refer to the comparable Akkadian name šarrumanu.[6]

On p. 13 the transcription of the name nqmd, previously rendered as Niqmaddu, is abruptly changed to Niqumaddu. The reference is to the name as it appears in hieroglyphic script on a fragment of an alabaster jar. But only the excavation number is given (RS 15.239). A student will be hard pressed independently to find the photograph and drawing of the text first published by C.F.A. Schaeffer.[7] The important concordance of excavation numbers and publication data compiled by Bordreuil and Pardee in the series Ras Shamra-Ougarit[8] is never referred to. Schaeffer transliterated the hieroglyphic signs letter by letter as nyk3m‘dy and, writing in French, transcribed the name as Nikamédi. The transliteration given here, nyk3šm‘dy (p. 13, end of paragraph 3) is obviously a careless mistake. The transcription Niqumaddu compounds that mistake, interpreting the sign for y variously as i and u, and allowing the sign for 3 also to represent u. Such an error and such an inexactitude means a student's confidence in the reliability of this primer will struggle to survive. It is incumbent on teachers to quote sources accurately and to explain any changes that are introduced. Drawing attention to points of general interest before points of detail is also desirable. This example of unskilled calligraphy could well be the work of an Ugaritic scribe, much more familiar with cuneiform than hieroglyphic, trying his hand at decorating a beautiful Egyptian alabaster vase with the name of his king participating in an Egyptian ritual. Whether it is a standard offering ceremony or his marriage with an Egyptian princess must be left open.

In furtherance of the aim to teach the language through actual texts, the first document to be introduced in Chapter 2 is a well-preserved linear abécédé (KTU 5.6). No hand-copy accompanies the photograph of the tablet and sometimes imagination is needed to identify details in the cuneiform font with those on the inscription itself. The solid cuneiform font is reminiscent of something lead typesetters would have used in the past, rather than one with open wedge-heads now in common use. The photograph fails to show the last sign on the first line. The second tablet, a damaged tabular abécédé (KTU 5.14) with correlated syllabic cuneiform signs, is purported to give “an indication of the names of the letters” (2.5.2, p. 36). A more suitable term would be “pronunciations” rather than “names.” Whether these first documents do fulfill the authors' aim of replicating the ancient methods of Ugaritic teachers is questionable and the first steps taken by a student of this primer are not necessarily the same as those first taken by a pupil in antiquity. It is far from certain that a novice scribe was taught first to write out letters in order before writing actual words. But even if it were certain, the ancient pupil would certainly not have modified this order of letters which “approximates the order of the Hebrew alphabet with some additions” (pp. 35-36) to that of “modern dictionaries” which “follow the order of the Hebrew alphabet plus additions” (p. 38). In fact recently published works, such as del Olmo Lete, [9] Tropper,[10] and Halayqa, [11] all adopt a romanised order, similar to that followed by modern Akkadian dictionaries.

Chapter 3 covers seven letters, in which epistolary style, vocabulary and grammar are illustrated. On several occasions cross-references are provided to Chapter 7, where a more structured description of grammar is presented. The shortest of the letters (KTU 2:12) is given first and used to show that its format corresponds unit for unit to a typical Akkadian letter (3.1). Why the lexical and grammatical analysis of this letter is postponed (see 3.5), when the points it illustrates could as easily have been discussed without delay, is strange. If that letter had been analysed directly after explaining its format, the exhaustive commentaries for the two letters (KTU 2.10 and 2.11) which are given precedence would have been considerably simplified. When it comes to analysing KTU 2.12 the only phrase needing explanation is šb‘d w.šb‘id, “seven times and seven times.” The simple statement that the termination -'id “seems to incorporate a vowel letter” deserves amplification,[12] and it is frustrating to follow the cross-reference to Chapter 7 to the paragraph on multiplicatives (7.5.3), and find no mention of šb‘id there.

The first letter to be textually analysed (KTU 2.10) is written on a clay tablet on which the introductory unit is marked off from what follows with a long horizontal line. The significance of this line is carefully noted (p. 51, para. 2, and again on p. 52, para. 2), but the line has been omitted from the copy of the tablet as printed, even though a footnote (p. 46, note 6) suggests it should have been reproduced. Placing such a strong emphasis in the grammatical analyses on the conjectural restoration of vowels leads to encumbering paradigms with more details than are absolutely necessary. The long paradigm showing the several different ways of vocalising nouns (Fig. 3.3) obscures the fact that the consonantal form of the stem takes only two inflections, -t and -m. Similarly the paradigm for imperative forms (Fig. 3.4) overlooks the fact that there are no consonantal inflections for number or gender. There is a column in Fig. 3.3 listing consonantal forms (in part hypothetical), but not in Fig. 3.4.

The notes on vocalisation seldom leave room for alternative views. The first word of KTU 2.13, tḥm, is vocalised as taḥmu, with the explanation that “in our view, a final /-u/ marks this as vocative” (p. 47). But Chapter 7 cites only two examples of vocatives with final aleph (7.4.6, p. 159) and both end in -'i. Applying the term vocative to an inanimate noun should be further clarified, especially as a few lines later it is described and translated as a bound form. The vocative in Ugaritic is fully covered in Tropper. [13] The possibility of an alternative parsing, taking it as the object of the imperative rgm, even though this seems easier on the syntax, is summarily dismissed. The restrictions syntax places on parsing are completely ignored when saying that the verb in the greeting yšlm.lk (p. 52) could be taken as either G or D. To argue against G is hard since D-forms take a direct object.

Administrative texts, by far the most frequent type to have been found at Ras Shamra, are the subject of Chapter 4. Because the problems they present generally involve lexemes and interpretation rather than grammar many teachers prefer to leave them for advanced classes. The four that have been selected here (KTU 4.143; 4.43; 4.266; 4.709) certainly do not present any serious grammatical problems and could easily have been annotated before the more complex grammar of the letters covered in Chapter 3. But they do have problems of interpretation which would have been much easier to discuss if some outline translation had been provided. How the titles they are given were chosen is not immediately clear. KTU 4.143 is described as “an agricultural record,” which fails to draw attention to its focus on the production of the zt (“olive”) centred on the gt (translated as “royal estate”). Even though gt is shown to be cognate with a Hebrew noun meaning “winepress,” the link between a place where fruit is pressed ( gt) and fruit that is to be pressed (zt) is not mentioned. The only meaning for zt in the Glossary is “olive,” which is likely to lead a student to translate line 2 as “250 olives.” Recording the numbers of olive trees was a much more likely task in hand. The “record of tribute” in 4.43 begins with the word tlt, but a student could be excused for thinking that that word was a numeral since there is no reference in the notes or in the Glossary to the homograph tlt, “copper or bronze” (see del Olmo Lete [14], s.v tlt (V), and Tropper,[15] s.v. tlt4). Similarly the word miḫd, “harbour” in (4.266: 5) is passed over in silence, but this is the only clue that the document is concerned with “maritime commerce.” Turning to the Glossary for guidance in translating the first phrase of this document, ym.ḥdt, leads to “new day” instead of “day of the new moon.” Using paired angled-brackets for the restoration of yr<ḫ> in both transliteration and cuneiform departs from convention.

Numerals permeate administrative texts but the help in the notes and in Chapter 7 for understanding them is sometimes less than adequate or confusing. According to 7.5 (p. 164) numerals may be “bound to or in apposition to the noun numbered” but the examples cited later in Chapter 7 include none that mirrors ḫmšm.l.mitm. In that same paragraph it is said that numbers “may be written logographically especially in administrative texts,” but no logogram occurs in any of the administrative texts selected here. So perhaps it was an afterthought to add the remark that “numbers usually are spelled out even in administrative texts.” The cardinals from 3-10 are said to exhibit “the commonly attested Semitic polarity” (7.5.1) yet these texts include ḫmš.kkrm (4.709:4), šb‘.kkr (4.709:1) and tmn.kkrm (4.43:5), with no suggestion that kkr is a feminine noun. The patterns of concord in administrative texts do not always follow the patterns found in literary texts so it is necessary to expand the relevant paragraph in Chapter 7 to show the implications of these differences. It is not clear why there is no entry for mat or mit in the Glossary, or why tlt is glossed as “third” when the cited cognates all mean “three.” A simple note on kbd, following the numeral 154 (4.43:5, not line 4), to show that the word reflects Akkadian kibittu “total,” would have been more apposite than what is found in the Glossary. There “total” is taken as a derived meaning from kbd, “liver,” which is said to be cognate with gabīdu (an Amarna dialect form of kabattu). With the meaning “total” it is said to be cognate with Mari kabittum (for which read kibittum). The word kbd then occurs twice with a different meaning at the end of the next document (4.43: 6, 7). Intelligent students, when left with no note for guidance, may infer that bright Ugaritic scribes were ignorant of some of the universal laws of arithmetic.

The notes on the three selected Legal Texts (KTU 3.3; 3.4; 3.9) in Chapter 5 are not without problems. KTU 3.3 begins with the heading spr ‘rbnm, “record of guarantors” but crucial observations have been missed in the notes on ‘rbn. That the noun comes from ‘-r-b can be taken for granted, but the noun is not listed as such in the Glossary, leaving the reader to work out how “guarantor” can be derived from “to enter.” That it is to be derived from a verb cognate with Hebrew ‘-r-b, “to stand as surety” is given only later (line 2). The word supplied to justify the vocalisation is urrubānu, “guarantors,” which occurs only once (RS 16.287: 7, PRU III, p. 37) as ú-ru-ba-nu, to be transcribed urubānu (see AHw). It is usually taken as an Akkadian plural, which could have been reflected in the Ugaritic suffix -m. Since no clue to explaining the suffix -n is given, a search has to be made in 7.4.8.10. There the only citation to be found is adn, a noun in which -n is better understood as a radical than a suffix.

In order to show that this Ugaritic text corresponds to similar Akkadian texts from Ras Shamra the first lines of RS 15.81 (also to be found on p. 37 of PRU III) are transcribed and translated (Fig. 5.1). But if the Akkadian text referred to earlier (RS 16.287) had been chosen instead, it would have provided the context for the use of urubānu and the phrase nabātišunu, “their fleeing” (line 8), would have supplied a closer parallel to b.'bth, “his fleeing.” [16] Negligently PRU page numbers are given only for one of these texts.

The phrase ḥwt.tth is said to be parallel to Akkadian ina māti šanīti (ana is the preposition used in RS 15.81) without providing in the notes or in the Glossary sufficient information to identify tt as a feminine formation of the ordinal tn, “two,” or to explain the suffix -h. The ordinal and cardinal meanings of tn should be separated, and “different” be given as a secondary meaning of the ordinal, or even (see Tropper)[17] be listed as a separate lexeme.

The next section of text is analysed line by line, but this is inadequate since the sense units are longer than the lines. In lines 5-6, w.mnm.šalm / dt.tknn, “and whatever it is that the investigators establish,” the relative particle (dt) has been transposed from its expected position, immediately following its antecedent (mnm), to one immediately preceding the verb of the relative clause. This is not clear from the translation as given, “and all the investigators, whatever they might establish,” where mnm is apparently rendered twice, with “all” and “whatever,” and the relative particle is omitted. The possibility that the verbal suffix -n embodies a resumptive object suffix also deserved to be noted. In the second part of the sentence (‘l.‘rbnm / hnhmt / tknn, “against these guarantors they shall establish it,” lines 7-9) the indirect object of the verbal form tknn is explicitly and emphatically identified in clause initial position. The suffix -n embodies the direct object. Some students could feel disheartened when faced with what may appear as convoluted syntax in an Ugaritic sentence with hardly any help on how to define what is happening or how to compose a fluent translation.

The last sentence in the introduction to the text mentions that it ends with the names of witnesses, but the horizontal line on the tablet separating this final section is not reproduced in the transliteration. None of the names or the patronyms of the witnesses are listed as such in the Glossary. Some elements are listed but the reader is expected to work out whether rb in ilrb is “rain” or “numerous, great, chief” and whether yn in ilyn is really “wine.” Searching for mtn and kb (‘bdkb) will prove vain. For the final gentilic gn‘ym (line 13) neither in the notes or in the Glossary is any reference made to the town of Gana.[18]

There are a number of difficulties with the notes on the next text (KTU 3.4) which can most conveniently be listed as they occur.

1. The “near demonstrative” pronoun in ymhnd (line 1) is said to have been declined to agree with ym, but in 7.3.5 (p. 157), where the term deictic is used, it is said to be not declinable. Transcriptions of the personal names in this text are usually given but in the corresponding translations they are often simply transliterated; upper-case letters are used as far as line 7 and then from line 8, without comment, transcriptions are used. The name agdn, is said to exhibit the suffix -n but no cross reference to (or from) ‘rbn in the heading of the previous letter (KTU 3.3) is given.
2. The spelling of the word aḫh (line 4), where it is the object of a verb, is contrasted with aḫyh in KTU 1.12:ii:50. But there may be some significance in the fact that it appears there in the phrase šr aḫyh, “prince of his brothers,” a genitive plural.
3. The verb tttbn (line 17) is parsed as š-stem from t-w-b, but there is no comment on the assimilation *št > tt in Chapter 7. To translate ksp as money can be criticised as an anachronism, but to say that the two words ksp.iwrkl, “the silver of iwrkl” form a double accusative misuses that term.
4. The explanation of wtb (line 19) as an infinitive is probably the best that can be done, given that any suggestion that the waw-conversive may have been emergent is far too speculative. But to say that it functions “quite adequately” when there is a clear change of verbal subject from the earlier indicative is far too optimistic.

Instead of the cuneiform font used elsewhere a photograph is provided for the obverse and the reverse of the last of the legal texts (KTU 3.9). Had photographs been provided for some of the other texts the observation about the “crudely-made signs” on this tablet would have had greater significance. There are no photographs of the edges of the tablet, so as it is presented, the last lines of each side are hardly or not at all legible. The horizontal lines drawn below lines 1 and 4 have not been recorded in the transliteration. But these are minor criticisms compared to the negligent way in which the two or three occurrences of an aberrant w are explained. It is first suggested that in the word btw, “his house” (line 4), w is a mistake, either to be deleted as a dittograph or emended, then (though uncertainly) that it constitutes an elision of h, and finally that it may reflect Phoenician orthography. A student could reasonably expect all three suggestions to be explained further. The first word in line 6, which the photograph clearly shows to have been written lwm, is wrongly transliterated as lkm and translated “for you” without further comment. The second word in that line, which has clearly been written wm, is similarly wrongly transliterated, as km. In the notes it is transcribed rather curiously but without comment as wa*[him]-ma. In view of the awkwardness of the script and the difficulty of interpreting w in btw, a reasonable assumption would be that in all three instances the w is to be read as h. To find the reasons for this apparent lapse will probably not be possible, but the fact that the scribe has often written w elsewhere on the tablet with the first two horizontals foreshortened may be a clue. It could well be that these three examples of w with two full-length horizontals (and with the expected middle-long horizontal written as two short ones) show the peculiar way in which this scribe with his admittedly rough hand wrote h. These are the only occurrences of h on this tablet.

The units of text again seldom coincide with the length of the lines. It would have been much better to set out the notes according to the speech units rather than the lines, especially as there are problems in identifying the subjects of the verbs and the beginnings of direct speech. These problems are not covered at all in the notes. Preferring “I placed” to “he placed” for št (line 5) can be supported by the fact that the next verb is obviously first person (agrškm “I will expel you,” line 6). But since the first verb governs lwm (line 6), to be transliterated as lhm, “for them,” rather than lkm, “for you” (see above), direct speech more easily begins with the conditional sentence introduced with hm (not km, see above). The translation of lines 5-7 will then follow the pattern: “He set a storeroom for them. ‘If I expel you from my house I shall pay fifty (shekels of ) silver.’”

Some remarks in the notes seem out of place, such as in the overview of the structure of the text, where the first line is tautologous and should be deleted. Others are asides in the notes about the subject matter. These belong more properly to the introductory paragraphs, such as the material from Amos 6:4-7 to support the idea that feasting and music was sine qua non for a mrzḥ. Others are concerned with etymological questions of vocabulary which could be dealt with more succinctly in the Glossary, such as ydd from n-d-d, “to wander.” To say it is cognate with Biblical Hebrew n-d-d needs some qualification, for there “to flee” is a more secure (though here inappropriate) meaning. The possibility that a homonymous root occurs in this passage to give the meaning “to stand up against” should have been mentioned. This suggestion is found sub voce in Tropper,[19] a source all students of Ugaritic need to consult (alongside Pardee's extensive comments on the book),[20] even if some of Tropper's ideas do not meet their teachers' (or Pardee's) approval. General untidiness is revealed in the way that the lines on the obverse of the tablet are numbered separately from those on the reverse in the transliteration but consecutively in the notes. There is no excuse for using so many different styles within so few pages for denoting conjugations. The variations G suffixed, G suff, G prefix, D Prefixed and D pref all occur on pp. 114-116. One remark could be understood as an unjustifiable slur on academic colleagues. The word ibsn is translated anachronistically and with no etymological or textual support as “pub” or “barroom” and then followed by the comment, “This seems reasonable, though it does cut against the general tendency for modern scholars to assume that unfamiliar words, places, and artifacts are religious in nature.” But modern scholars have not usually suggested a religious term for ibsn: del Olmo Lete uses “almacén,” translated by Watson as “warehouse,” Tropper uses “Vorratskammer,” and Pardee uses “drinking-club.”

The seven poetic passages in Chapter 6 are annotated much more summarily than the prose in earlier chapters, to the extent that the last four have no notes at all (KTU 1.100 [Snake Bite Text]; 1.2 iv [Yam and Ba‘al]; 1.19 [Aqhat]; 1.23 [the Birth of the Goodly Gods]), and no cuneiform text is given for the last three. The notes on the first three texts (KTU 1.114 (El's marziḥu); 1.5 vi:11-25 (Mourning from the Baal [in contrast to “Ba‘al” for 1.2 iv] Cycle); 1.14 i-ii (Keret [it is surprising to find the name vocalised thus in this book] Epic) vary in quality and usefulness. Some glosses on the vocabulary simply repeat information in the Glossary (e.g. hdm “footstool” 5 vi:13), but for some reason others are treated differently in the Glossary or are omitted from it.

lexeme	glossary	notes	reference
atnt	missing	urine	1.114:21
ul	power	freeman or strength	1.14 ii:35
‘db	to prepare, set	offer	1.114:4
‘mr	earth (?)	heap of corn	1.5 vi:14
yšq	missing	to pour	1.5 vi:14

Some verbs are parsed but others are not, and the Glossary does not always confirm the meaning of the verbal stem occurring in the text, for example yštql “he enters” (1.114:17) is said to be Gt, but š-q-l is listed only as G. Similarly, q-y-l “to fall,” is listed only as G, which is inadequate information to explain the alternative parsing, “št of the root ql indicating some sort of falling motion, ‘falls over himself’.”

The implication of some notes cannot be appreciated when no translation is supplied or the translation given is insufficient. The word ‘d is translated “until, up to” (1.100:3), which more or less agrees with what is stated in the Glossary, but there is no definition for the following word šb‘. The Glossary gives no guidance about whether to take it as an infinitive from š-b-‘, or as the noun “satisfaction,” or as the ordinal or cardinal of the numeral “seven.” Instead of grouping “satisfaction” together with “seventh” it would have been better to group all forms of šb‘ semantically related to “seven” as a separate lemma. Similarly, it is claimed that yks (1.5 vi:16), from k-s-y, “ to cover” (perhaps D-stem, but the hesitation suggested by the question mark in the note is not sustained in the Glossary), can be taken as passive or active. Without alternative translations the relative merits of the two suggestions cannot be assessed. This verb occurs in the sentence lpš.yks / mizrtm, usually translated along the lines of “for clothing he puts on sackcloth.”[21] How to transform this into a passive expression is not immediately obvious. The authors may have intended to use “medial” instead of “passive” and to translate “he covers himself,” but for that a much fuller explanation is required. Translations would also be helpful when doubts are raised about the conjugation of a verb. Ambiguity is noted for ytb (5 vi:12) but no guidance is given about whether it represents the suffix conjugation, the preterite, the participle or the infinitive (as advocated by Smith). Solving such problems is not made easier by the brusque presentation of the declension of weak verbs in Chapter 7 (pp. 176-177).

Grammatical observations can be inexact. To describe ṣb'i (14 2:33) as a superlative suggests it is not a noun but an adjective. It is, of course, a noun, repeated in a bound phrase to impart the quality of excellence. The verb wyṣi (14 ii:32) is explained as “jussive with i marking ø vowel.” The i may represent a vowel followed by a glottal stop, but that is not the same as representing a “ø vowel.” This particular form was not included with the other final aleph verbs in 7.6.18.3 (p. 177). There the closest analogy is tṣi, but this is said to reflect taqtil, not taqtul, the paradigm form cited in 7.6.11 (p. 172).

Inconsistencies in format persist, e.g. “3cdual” (114:18) and “third common dual” (114:9). Text has been omitted, e.g. the subheading “column i” omitted on p. 129. Some misprints are irritating, e.g. y.lmn (114:8), corresponding to ylmn in the cuneiform, but others are misleading. Virollaud's initial hand copy of what we now know as KTU 14 (as reproduced in Herdner)[22] showed enough slight scuffing at i:22 for what was probably to be read as ḫtkh to look more like ḫtkp. Whether the scribe failed to write the third wedge of the h or the scuffing made it illegible, there is general agreement that the preferred reading is ḫtkh. The transliteration given here is ḫtk<h>, which corresponds exactly to the cuneiform text as given, where angled brackets are also used. This format can be criticised on two counts. On the one hand, angled brackets are not expected in a cuneiform copy but only in a transliterated text, where a measure of editorial interpretation is necessary. On the other hand, they indicate that there is no space on the tablet for the bracketed element, which is clearly not the situation here, whether that element were h or p. Furthermore, the headword for the note, ḫtkp, does not correspond exactly to the transliteration or to the probable reading.

The didactic approach adopted by the authors necessarily means that salient points of Ugaritic grammar occur more or less at random in the texts selected. Chapter 7 provided an opportunity for all of them to be set in order and where necessary to be described more fully in a less piecemeal fashion. Here was a good place to add supplementary observations, dispensed with earlier because of space. But some of these observations deserve more precise phraseology. In the phonology section it is not quite correct to say that supplementing the abécédé with two aleph signs “is an early application of matres lectionis” (p. 149). That term means to use an existing consonant also to represent a vowel. The additional signs seem to show an attempt by scribes, who when writing cuneiform would have been aware of the importance of vowel notation, to indicate the vowel associated with the glottal stop. A similar supplementation is found when the alphabet established for West Semitic was adapted to write Greek. Extra signs were appended to indicate other consonants and, in a complementary step, unwanted graphemes were utilised to indicate Greek vowels.

The correlation of Ugaritic phonemes with those in other Semitic languages (Table 7.1, p. 151) allows orthography to intrude into the phonology. Hebrew letters are sometimes used to indicate Hebrew, Aramaic and Phoenician phonemes, though Arabic is not used in a corresponding way. Some of the phonemes recorded as blanks in the Akkadian column would not have been blanks if the vagaries of cuneiform writing had been taken more into account. A changed format for the table would make matters clearer. Arabic and Ugaritic have the fullest set of phonemes, so those two columns could be more easily juxtaposed. It would not have been inappropriate to include here a column for Old South Arabian. Similarly, since Hebrew and Phoenician have more similarities with one another than they have with Aramaic, they could also be juxtaposed. Reworking is required to correct the lines that make no sense. Ugaritic b does not equate with Arabic f, nor p with p; Proto-Semitic g never equates with b, and j rather than g would be better for Arabic. The notes state that Proto-Semitic d may be reflected in Ugaritic d or d (p. 152) but no indication of this is given in the table, even though the similar bifocation for Aramaic q and ‘ is indicated.

The evidence to support some statements is weak, as when it is asserted that the three aleph signs always indicate the vowel that follows the glottal stop except when aleph is word final (p. 149). Almost all the words cited to illustrate this assertion begin with aleph, so it is obvious that the vowel follows the consonant. The word rpim, “healers,” vocalised rāpi'īma, is the one example given with non-initial aleph. But since with words that do not end in aleph the plural suffix -īma is written as -m, the sign i in rpim could just as well indicate the vowel that precedes the aleph. No examples are given where 'a is omnivalent or where 'i is word final. It can be difficult to comprehend the patterns in which vowels and consonants are assimilated. Evidence from assimilated forms such as yšu, “he lifts” (p. 150), to show that a geminated consonant could be written with a single letter is undisputed, but references to cognate nouns do not always carry quite the same weight. It is almost certainly correct to say that prt, “cow,” should be transcribed parratu, but this overlooks the fact that the cognate word in Syriac is written parāta, “ewe.” The example given to illustrate vowel elision is tittar˙u > *tiytara˙u (p. 154), where clearly < should have been used instead of >. Such an error is as serious as using + for - in a primer on arithmetic. It should have been picked up at the latest while the book was still in the proof stages.

The notes on morphology are not without problems. The table 7.3.3 lists the pronominal suffixes (in other chapters tables of paradigms were numbered as figures but in Chapter 7 they are numbered as paragraphs). But even though the nominal suffixes are distinguished from the verbal suffixes only with first singular nouns, amalgamating them into one table is messy.

The “determinative-relative” pronoun d (7.3.4) is described first as determinative, i.e. “the one of” (7.3.4.1), and then as relative, i.e. “the one who” (7.3.4.2). Comparable forms in Hebrew, Aramaic and Arabic are cited for the vocalisation, but Akkadian ša, with which the Ugaritic pronoun shares many features of syntax, is not mentioned. Akkadian dictionaries separate ša immediately preceding a noun in the genitive (CAD, s.v. a; AHw., s.v. A) from ša introducing a subordinate clause related to an antecedent (CAD, s.v. b; AHw, s.v. B). Since neither paragraph 7.3.4.1 or 7.3.4.2 cites an example without an antecedent, distinguishing the two pronouns from the evidence adduced presupposes that the meanings of the texts are precisely understood, but in Ugaritic meanings are often elusive. The penultimate sentence in 7.3.4.2 sounds like an inappropriate intrusion and would fit better in 7.3.4.1. In fact the antepenultimate sentence of 7.3.4.1 is remarkably similar (“declined” is changed to “construed”) which suggests that someone has botched a cut-and-paste job.

Sometimes there are apparent contradictions. The proposition that the prefix vowel of the š-stem “might have been u as in Akkadian” (7.6.3, p. 167) is qualified with the statement that “there is some evidence that favors /a/.” The supporting example chosen is yšš'il (yušaš'ilu) “he shall cause to enquire,” which supports the main proposition, not the qualifying statement, and so should be moved back. The passive of the š-stem is then illustrated by the form yttb, which is translated “he shall be seated,” but “he shall be made to sit” would evoke the causative nuance better. The evident regressive assimilation *št > tt is left without any explanatory comment. To illustrate the reflexive idea of the Gt-stem the form yitsp is translated as “he gathered” (7.6.1) which gives little sense of a reflexive. Similarly for the št-stem the form yštḫwy is cited, but neither the principal translation, “he shall ask for life,” or the paraphrase, “greet by prostration,” give a reflexive sense. The root ḥ-w-y is here marked with an asterisk, but in the next paragraph and in some other parts of the book, a square-root symbol is found. On several occasions, probably because of inconsistent typesetting, this symbol seems to have been curiously replaced by a sublinear point, e.g. on p. 129 with n-p-l. In the Glossary roots are usually distinguished from forms with uppercase characters. That comparatively simple typesetting problems were not always solved seems clear from “¨Zrich” (p. 166, note 8).

Some phraseology is opaque and could prove misleading. It is said that within the four basic stems of the verb “there are [sic at] least eight tenses or aspects” and the reader is at a loss to know whether the word “moods” has been accidentally or deliberately omitted. The “Suffix Conjugation” is distinguished from the “Prefix Conjugation” in 7.6.5, but “conjugations” is also used in 7.6.4 for verbal forms with lengthening or reduplication of radicals (the L-form and the R-form). Perhaps “stem” was the word intended for these verbal forms, especially as it is admitted that they are sometimes considered as irregular D-forms.

The words in the glossary are arranged according to the alphabetic order suggested by Gordon.[23] This is not quite the same as the order an ancient scribe would follow for an abécédé, which has been described as “not suitable for alphabetizing in a work of modern scholarship.”[24] The authors of this book, who wished to follow the spirit of Ugaritic paedagogy, have had to be slightly less purist. In this they follow the example of Bordreuil and Pardee who used it for their Glossaire, but integrated sign 1 with 2 and 3 and sign 20 with 21.[25] Today's standard dictionary (del Olmo Lete)[26] as well as the short lexicon designed for student use (Tropper)[27] prefer to take Akkadian dictionaries as their model. It can be commended as practical and an implicit recognition that Ugaritic represents an element of cuneiform literacy.

The decision to amalgamate English proper names with Ugaritic lexemes in the Index results not only in untidiness but unevenness. A sample check suggests that not nearly enough attention has been paid to compiling it. Some important items are missing, such as the disputed translation of the homonym zbl at 1.14:ii, 45. The problem was noted (with a misprint) on p. 131, and it was one of sufficient importance to have been mentioned in the index. On the other hand we find three references to mlkny (a hypothetical dual form of the root mlk, but curiously not italicised) which are only incidentally concerned with that lexeme. The first reference (p. 56) is to a paradigm table for the “G Suffix Conjugation” of the strong verb (Fig. 3.7) and the second (p. 72, presumably an error for p. 73) to an identical table (Fig. 3.11, an unnecessary duplication) pertaining to the form qlny which is derived not from a strong but from a “midweak root.” The third reference (p. 171, presumably an error for p. 169) is to the same table in a slightly modified format (7.6.7). It seems strange in an index to cite a lexeme, one that is not actually attested, to illustrate a less common grammatical form instead of identifying the form that deserves to be indexed. The fact that ttpl is listed incorrectly in bold font, exactly as it occurs correctly on p. 129, suggests that there has been some kind of automatic copying and pasting from text to index. In fact the three page references for ttpl are all wrong (125, 128 and 131 seem to be errors for 123, 126 and 129). In itself this is frustrating but discovering that the first of them leads simply to some cuneiform script, the second to the corresponding transliteration and only the third to an explanation of the verb form, makes one's frustration explode into exasperation. There are so many wrong page references in the Index that, if automatic software was used to generate them, it must have been misapplied or seriously bugged. If the Index is essentially the result of human effort then that effort was not exerted on the final page proofs.

It has been most disappointing to find so many difficulties with a book that had such a well judged aim. Perhaps it was not sufficiently focused. Several successful surveys of Ugaritic published in the 1980s paid much less attention to language.[28] Of the most recently published grammars Bordreuil and Pardee[29] does include a historical and cultural survey, but Tropper[30] follows the tradition of earlier grammars[31] by essentially restricting itself to questions of language. If the historical and cultural overview had been briefer more attention could have been paid to the language without increasing the size or cost.

This primer expected students first to understand essentially straightforward prose directly extracted from original texts before confronting more ambiguous passages of verse. It had the potential to become a standard classroom text. Unfortunately it fails to achieve that status because it is replete with errors and obscurities. But even without these, questions remain about the paedagogical presuppositions of the authors. The most successful results of applying the inductive method to language teaching involve living languages, including Modern Hebrew and Modern Arabic. More limited progress has been claimed when applying the method to written languages, in particular to Classical Latin and Greek but also to Biblical Hebrew. But with its relatively tiny text corpus and only minimal traces of vowels, Ugaritic challenges the method to the extreme. Not only because of this, but also because no serious student would expect to or would be expected to begin learning Ugaritic before obtaining competence in a mainstream Semitic language, Ugaritic is suited to a more formal approach. That explains the success of books such as those of Sivan and Tropper, where a concise description of the language is followed by an annotated anthology of texts.

It is also of fundamental importance that teachers first draw a student's attention to features that are evident and describe the best attempts to explain those features, including their own preferred explanations, before proceeding to speculate about possibilities that are not actually attested. Similarly students must first familiarise themselves with facts before assessing the strengths and weaknesses of rival explanations and advancing their own ideas. Given that Ugaritic has now firmly taken its place in the Semitic family and the phonetic value of all 30 graphemes is more or less accurately established, any attempts to understand the language depend first on recognising cognate lexemes. Once the possible links have been established, understanding depends on integrating the grammatical information inherent in the consonantal text and accommodating that to the demands of syntax. It is only at that stage that speculation about vocalisation is appropriate.

Here priority is always given to vocalisation in such a way that for most of the time students will be denied the opportunity of juggling lexical and grammatical ambiguities inherent in the consonantal text with whatever competence they have acquired from an ability in another sphere of Semitic studies. It should have been possible to take a few passages and show that different translations could be produced from different vocalisations. Such an exercise would elicit from the mind of a student of Biblical Hebrew the challenge of vocalising unpointed texts. The revocalisation of Ugaritic has been given priority in the research work of scholars, presupposing as it does rigorous investigation into the comparative Semitic grammar. Even so, it is essentially speculative and can be regarded as secondary for a beginner. This is not to say that it should be ignored in a primer, but it should not be put in prime position.

M.E.J. Richardson, Leiden University