Annotated bibliography

The commentary to a new transliteration and translation of EA 82 contains important grammatical observations on Canaanite Akkadian, including the pleonastic use of the negative in coordination; feminine singular verb agreement with a masculine plural subject; use of the permansive as a perfect; and use of the waw consecutive.

– Giorgio Buccellati, 1994

Suggests the possibility of reading the signs PA, WI as “fa, fi,” on the basis of correspondences in writing, etymology [hardly acceptable] and Greek transcriptions.

– Giorgio Buccellati, 1994

The patterns used to form ordinal numbers are parūs (e.g., šalultu) and paris (šalšu), of which the first one was used for numbers referring to time division. In addition, Akkadian also uses the afformatives -ī (šamanū) and -ay (šalšay). The last portion of the article contains comments to the section on the numerals in DELITZSCH 1889 Grammatik (Delitzsch 1889 Grammatik).

– Giorgio Buccellati, 1994

A survey of data with the general conclusion that Alalakh Akkadian is substantially a form of Mitanni Akk. (i. e., as in Nuzi), though somewhat more archaic and more influenced by NWSem; the peculiarities of Mittani Akk. seem to derive from development within Babylonian rather than from influence of Assyrian (even though geographically closer). Some of the traits of Alalakh (and Mittani) are the confusion of voiced and voiceless emphatic stops, accusative suffixes like -šunū instead of -šunūti, use of the generalizing suffix -me, more frequent use of u and -ma u, frequent omission of the subjunctive normal use of the preterite (over the perfect) to express past tense.

– Giorgio Buccellati, 1994

This article offers a list and an analysis of orthographic and phonetic differences between the Akkadian found in the Alalakh texts and standard Akkadian, with reference to Middle Babylonian and Nuzi dialects. Differences are categorized by noun, pronoun, verb, and by preposition, adverb, conjuction. The language can be designated as Mittani Akkadian. It is similar to the Nuzi dialect in the confusion of voiced, voiceless, and emphatic stops; non-assimilation of s suffixes; occasional confusion of silibants; the writing -šunū for -šunūti; confusion of pronoun gender; blunders in verb inflection; use of the generalizing suffix -me; and general lack of knowledge in the area of syntax. The Alalakh grammar is generally more archaic than Nuzi and has been subject to more West Semitic influence. However, it is identical to Old Babylonian in its use of mimation; frequent preservation of waw; the writings u-ul i-il-la-ak; and the non-contraction of -i-a. Mittani Akkadian shows more parallels with Middle Babylonian in the use of m for w; a becoming e in closed syllables; st becoming lt; loss of difference between accusative/dative; and in the use of ina muḫḫi instead of eli.

– John L. Hayes, 1974

A summary in English appeared under the title “Prepositional Phrases as Attributes of Noun in Akkadian,” RAI 11 (1964) 112-16.

In some cases (pp. 398-400) prepositional clauses are clearly and immediately governed by nouns (i.e., they are adnominal adjuncts or attributes), namely (1) when there is no predicate, e.g., tuppi PN ana PN2 as title of a letter; (2) after a noun like “road” and elsewhere with local determination, e.g., girrū ištu GN or immerātim ina GN; (3) when the prepositional clause is either a complement of company (e.g., awīlum qādum alpīšu) or a partitive complement (e.g., ina mārīšu ištēn).

In other cases (pp. 400-2) it is ambiguous whether the clause is adverbial or adnominal, especially with finals (e.g., ṭuhham ana ukullē alpī lilqū “let them take bran for the feeding of the oxen” or “as oxen fodder”) and with locals (e.g., epšētūšu lemnēti ina GN ašmē “I heard in Nineveh about his evil deeds” or “about his evil deeds in Nineveh”).

To resolve ambiguities, adnominal complements are introduced by ša, e.g., aššum šēm ša ana ukullē bītim ša tašpuram “concerning the wheat for the feeding of the household about which I wrote.”

– Giorgio Buccellati, 1994

The main criterion for the internal classification of the Semitic verbal system is the polarity between stative and fientive verbs, even though there are no Semitic languages in which such a differentiation is perfectly operative. Akkadian is the language most affected by it, Hebrew and Arabic show considerable traces, while in Aramaic and Ethiopic the distinction is hardly apparent. The languages just mentioned are dealt with separately in five chapters. The one on Akkadian, like the other ones, consists almost exclusively of lists of verbal forms, arranged according to vocalization (present-preterite for Akkadian) and meaning (always for the B stem).

The main conclusions (ch. VII and VIII) are: (1) transitive fientive verbs are primarily of the types iparras : iprus ( = qatala : yaqtulu) and iparris : ipris ( = qatala : yaqtilu); (2) no clear vocalic class, and no correlation between Akkadian and West Semitic, exist for intransitive-fientive verbs; (3) stative verbs occur with a vocalism a-a, but especially with i-i (pp. 153-68); (4) the Akkadian present iparras is original, but somewhat later than the preterite; (5) the ending -u is a subjunctive from the earliest Semitic stages; (6) adjectives are primary with stative roots, prefixed finite forms are primary with fientive roots (pp. 192-201).

– Giorgio Buccellati, 1994

A detailed list of occurrences, leading up to the conclusion that the marker -u is used when the corresponding indicative ends with the third radical, and the marker -ni when “the indicative exhibits an affix after the third radical.” “The forms which show both -u and -ni are relatively rare and must be explained as a later evolution” (27). The article ends with comparative remarks with the Babylonian and Middle Assyrian subjunctive, of which the Babylonian would be the most ambiguous, and the Middle Assyrian the clearest.

– Giorgio Buccellati, 1994

The view that patterns with prefix na- belong with the passive (niphal) conjugation is refuted on the basis that many roots with such patterns are intransitive. In fact, they have the same functions as patterns with prefix ma-, namely the expression of relationship of place, time and instrument. The explanation is that na- is dissimilated from ma- when the first or second syllable after ma- contains a labial: this rule is peculiar to Akkadian, and similar forms in Syriac, Hebrew and Arabic are rare and possibly loanwords.

– Giorgio Buccellati, 1994

The Akkadian “present” ikašVd is historically the result of two distinct Proto-Semitic forms. (1) When the final vowel is the same as that of the preterite ikšVd, it is related to the original imperfect, and is therefore parallel to Ethiopic janeger (p. 383f.). (2) When the final vowel is a, it is derived from the Proto-Semitic perfect: the occasional lengthening of the middle radical is due to stress (p. 380, n.1), and prefixation, instead of suffixation, is due to analogy with the first type of present, i.e., the one related to Ethiopic (p. 385). The documentation (pp. 379-83) is aimed at showing that the vocalization of present/ preterite in Akk. is the same as that of perfect/ imperfect in other Semitic languages. By way of introduction (p. 375f.), the view that the Semitic perfect is represented by Akk. permansive is refuted on grounds that the permansive has different functions and that it can be formed from substantives.

– Giorgio Buccellati, 1994

Still the classic work about nominal formation, first published in 1889. Besides the very considerable amount of material collected (including Akkadian, for which the proofs of Delitzsch’s grammar were used), the book is notable for its theoretical approach, whose two ground rules are: (1) derivation of all nominal forms (excepting those which we now call “primary nouns,” see pp. 1-9), from either the “perfect” or the “imperfect” of the verb, where one should note that the introduction of the imperfect-class is one of Barth’s innovations; (2) systematic correlations among all main types of Semitic languages, with rigorous application of phonological rules.

– Giorgio Buccellati, 1994

Arabic (a)ta, Hebrew and Syriac (a)t have an adverbial function. The basic underlying form -ta is connected with the pronominal element tu, ti. The same pronominal element occurs also in “Assyrian” where however the bi-forms šu, ši are also attested. “Assyrian” iš (more seldom aš) is to be connected with the pronominal elements šu, ši, and therefore ultimately with the adverbial ending ta of West Semitic, with which it shares a common adverbial function. Note also the extended forms -āniš (e.g., abūb-āniš) in “Assyrian,” and -anniš (e.g., qador-anniš) in Hebrew.

– Giorgio Buccellati, 1994

The expression of the type PN mer-šu ša PN, frequent in Late Babylonian and Aramaic, is already attested in Akkadian texts from OB Susa, Nuzi, and Boghaz-koy. It may have come into Akkadian and Aramaic through the influence of unknown languages scattered from Elam to Asia Minor.

– Giorgio Buccellati, 1994

A neo-grammatical approach, with two main conclusions: (1) the original verbal form is the “imperfect” (yaqtul), simply meaning “action”; the reasons are the similarity with the imperative, and the great differentiation in the vocalism; (2) the “perfect” (qatala) is secondary, and derives from a combination of the nomen agentis and a personal pronoun. In the section on Akkadian (pp. 10-23) the point is made (following Barth and Brockelmann) that qatala is to be found in Akkadian not as kašad [the stative], but rather as ikašad [the present, without duplication of the middle radical].

– Giorgio Buccellati, 1994

Empahsizes the reality, all-pervasiveness and strong continuity of the triphonetic root (“Dreilautigkeit”), especially for the verbal system. The main exception consists of some words like kalb [the primary nouns] which cannot be considered dualistically [i. e., they do not result from interdigitation of root and pattern].

– Giorgio Buccellati, 1994

The reflexive t element may become a lexical element as in tabālu from wabālu, “to carry”, or dagālu from glw, Arabic iǧtalā, “to look at”.

– Giorgio Buccellati, 1994

Discusses the elements used in the formation of stems (“voices”) in Semitic, with ample role given to Akkadian, see esp., about n (p. 392), š (p. 395 f.), t (p. 398), tn (p. 403-6) and length (p. 410-12). Of particular interest is the analysis of the š stem as utilizing the same element of the personal pronoun ši: thereby “the prefix to the verbal stem is an objective pronoun” and “the action of the verb is naturally thrown from the speaker on to another absent person represented by the prefix; so, in Assyrian, altum ‘I concealed”, shapel ušaktum [sic] ‘I (did that) one concealed, i.e, “I caused to conceal”” (p. 395); in W Sem. we find a hiphil instead of a shaphel because the personal pronom is hu (p. 396) [see SPEISER 1936 “Formatives” (Speiser 1936 Formatives)]. The main conclusion is in favor of an original biradical root.

– Giorgio Buccellati, 1994

Gives first an outline of all forms of the personal pronoun, independent and suffix, including also the dative; then an outline of the Agau and Galla pronouns (pp. 72-77), stressing the resemblances (in line with an early Hamito-Semitic comparativism); finally an outline of “Akkadian” (i.e., Sumerian, thus on pp. 65, 81) pronouns, as deduced especially from a grammatical tablet published in conjunction with the article (78-88). [Cf. BEZOLD 1919s Verbalsuffixformen (Bezold 1910 Verbalsuffixformen)].

– Giorgio Buccellati, 1994

After the completion of major text publications and of the preliminary work of decipherment, the time is ripe for a comprehensive grammar, which the author says he will begin to publish by 1887. The prolegomena deal especially with the nature of the documentation and with phonological change.

– Giorgio Buccellati, 1994

The discovery of a set of dative pronominal suffixes by BERTIN 1885 “Pronouns” (Bertin 1885 Pronouns) has gone unnoticed, but the abundance of such forms in the newly discovered OB texts proves beyond doubt the validity of his insight. A careful survey of the data is given on pp. 4-14, with exemplification divided by period and dialect, and adding the independent personal pronouns to the suffixes. Main conclusions: final -m in the dative forms begins to disappear at the time of Hammurapi; from about 1500 B.C. on a single form is used instead of two as in the older periods, namely: naši (for niati, niašim), kunušu (for kunuti kunušim), ka (for ka, kum), šunuti (for šunušim), šinatu (for šinati, šinašim). On this basis, several dozens of texts from Assurbanipals’s library are considered as copies of texts dated either before or after 1500 B.C.

– Giorgio Buccellati, 1994

Applies to Semitic a theory elaborated by M. Bloomfield for Indo-European, namely that “words belonging semantically to the same class have a strong tendency to influence one another morphologically, producing new forms which are a blend of several more original forms, as e.g., Latin sinexter, a blend of sinister and dexter” (35); this process is called “congeneric assimilation.” After describing various formal means by which new roots can be derived (prefixation, consonantalization of vowels, etc.) the author gives examples from the main Semitic languages, in three groups. (1) Addition, e.g., Akk, šelibu vs. Hebr. šûʿāl “fox”: b added in Akk. because it occurs in other animal names such as kalbu. (2) Modification: Akk. nēšu vs. Hebr. layš: l changed to n in Akk, under the influence of nimru “panther,” našu “eagle,” nadru “wild, fierce.” (3) Related groups of roots, e.g., Akk. zumbu “fly” zibū, zizānu, zunzunnu, zirzirru “locust,” zuqaqipu “scorpion,” etc. With the application of the principle of congeneric assimilation one can avoid having to posit special phonetic laws for sound changes.

– Giorgio Buccellati, 1994

Short contributions of which the following are of interest for Akkadian: 1. Middle weak roots were originally biconsonantal with short vowel, successively lengthened when accented in open syllable; 2. Afformative elements of reflexive conjugation are objective pronominal particles, those of causative are subjective (nominative) or agential; 13. The personal pronoun of the first pers. sing. at the beginning of inscriptions is not appositional (vs. Poebel) because in some South Semitic inscriptions the pronoun is not followed by any resumptive; 14. Poebel’s etymology of qâtu “hand” from lqḥ “to take” cannot be accepted.

– Giorgio Buccellati, 1994

The change a > e next to gutturals is difficult to explain from an articulatory point of view because a is more homogeneous than e with those consonants. Hence it must be postulated that the fricatives, laryngeals, and velar Aleph 3-5 “shifted to the palatal position before disappearing,” a phenomenon similar to the change from Latin carum to French cher “flesh.” On the basis of this assumption the author goes on to explain a number of Akkadian examples.

– Giorgio Buccellati, 1994

Only the first section includes observations about Akkadian. The emphatic consonants constitute “a series which partakes of the characteristics of two place categories at the same time,” the added element being a back oral “tension” (212-15).

– Giorgio Buccellati, 1994

Hebrew forms of the type ynsr are explained as representing a strong conjugation, against the suggestion by Rössler 1961 (“Tempusform”) [Rossler 1961 Unerkannte Tempusform] to interpret them as the indicative of a special prefix tense like Akkadian inaššar.

– Giorgio Buccellati, 1994

After a few remarks on phonology (št used alongside lt, dropping of mimation and consequent hypercorrection in the use of CVm signs in word final position) Bloch analyzes in detail the verb, preposition and conjunction. The most important features which Bloch singles out as typically MB are subjunctive of the 3 sg. fem. permansive (parsatu), occasional assimilation of e/i in verbal forms with i in last syllable (lilebbiš next to lilabbiš, lack of assimilation in the type iddimšu, lack of vocalic syncope in pret. B I/w (ūšibū as regular type), in last weak verbs the final vowel is regularly short and contraction (irdē for irdiam) is also regular, the use of ki as a conjunction and as an emphatic particle instead of -ma. Of general interest for Akkadian grammar is especially the discussion about lenghtening of middle radical in the present (pp. 311-13, 329 f.), and of last radical in verbs middle weak (333 f.).

– Giorgio Buccellati, 1994

The grammar (pp. 9-76) is based especially on DELITZSCH (Delitzsch 1889 Grammatik), and has no independent value.

– Giorgio Buccellati, 1994

After a detailed history of the problem of triconsonantal and biconsonantal roots, a lexical analysis of the Semitic words containing gl, kl and ql is given, with the conclusion that they must all be reduced to a set of basic meanings closely connected with one another: “to resound” “to roll—be round” “to move.” There follows a discussion of the processes by which actual words are differentiated by enlargement from the biconsonantal roots: reduplication, lengthening of middle vowel or of second consonant, root-formative consonants (“Wurzelformantien”: s, h,˒, n, m, t). The last section attributes psychological functions to acoustic realizations so that e.g., voiced velars are taken to indicate weight or slow motion (p. 62); on the basis of this criterion further lexical categorizations are given of words from the roots g/k/ql and b/pr, and the conclusion is reached that words with voiced consonants are closer to the original meaning of the root (p. 64, 66). In the concluding chapter the author stresses the historical reality of the biconsonantal root.

– Giorgio Buccellati, 1994

A brief statistical analysis of sentences in the Taylor Prism, divided as declarative (the most frequent), relative, negative, imperative, cohortative. Dealing with word order, notes a rather high number (about 80) of sentences with object after predicate, as compared with 228 of predicate in last position: emphasis is not a sufficient explanation, and he suggests as reasons for the inversion the desire for chiastic effect with short sentences (18 cases) and for a sense of completion when a short clause concludes a long paragraph (24 cases). On p. 307 he suggests that the negative particle lā was probably used only in subordinate clauses. Finally it is noted (p. 308) that -ma is more common a conjunction than u.

– Giorgio Buccellati, 1994

A statistical study on word order in Assyrian royal inscriptions from the Middle Assyrian period down to Adad-Nirari III; the data are given first separately for each king, then tabulated for all kings together. Out of a total of 2249 sentences, only 75 show an inverted order with the object occurring after the predicate, 36 of these being found in only one king (Aššurnaṣirpal). In the text of the article, the cases of inverted order are studied more closely, with a fine sense for style; the main reasons for inversion are chiastic order, occurrence at the end of a paragraph, formulaic expressions, desire for close connection or euphonic structure, and emphasis. (The continuation of the article, promised at the end, unfortunately never appeared).

– Giorgio Buccellati, 1994

By analogy with the endings -ū/-ā for the masculine and feminine plural of the adjective in its predicative function [stative], a new ending -ūt- for the masculine plural of the attributive adjective was introduced next to the already existing ending -āt- of the feminine plural.

– Giorgio Buccellati, 1994

The majority of scholars, following Nöldeke, consider the two endings -at and -t of the feminine as equally proto-semitic. Brockelmann defends instead the view that -t is a secondary development phonologically conditioned. He documents his thesis by analyzing in all Semitic languages the patterns where -t, instead of -at, occurs, namely, qatal, maqtal, patterns with long first vowel and those with long vowel in a syllable before feminine ending; ample room is given to Akkadian.

– Giorgio Buccellati, 1994

The formation of nouns is often affected phonologically by other nouns which are close in meaning and phonological shape. This occurs (1) for word pairs, p.10 ff, e.g., Arabic Qābīl - Hābīl, instead of Qāyn - Hābīl; (2) for nominal patterns, p. 14 ff. e.g., Aramaic ʾeštyāmā from Akkadian šatammu under the influence of gayyānā, p. 23; (3) for the analysis of roots, p. 24 ff. The latter is the only category in which Akkadian is explicitly considered (p. 35 f.), and with the general remark that the influence of rhyme in word formation is generally scarce in Akkadian. Among the examples given is magāru, which the author connects with Hebr. māgan, “to give as a gift,” Aram. maggān, “gratis,” explaining the final r through phonological influence of Akkad maḫāru. [See already, though without Akkadian documentation, S. FRAENKEL, “Zum sporadischen Laut Wandel in den semitischen Sprachen,” BA 3 (1898), pp. 60-86.]

– Giorgio Buccellati, 1994

The “stative” is not a tense of the verb, but rather a nominal sentence with predicate in the predicative state and the subject as a pronominal suffix in the nominative. Arguments: (1) since the stative occurs with all nouns, it would have to be considered a denominative tense (tenses proper cannot be derived from nouns); (2) the stative does not take the modal (i.e., verbal) endings of the ventive in cases where it would be phonologically possible; (3) the stative occurs regularly whenever the predicate of a nominal sentence is not followed by a qualification or complement. The last point is based on the observation that no nominal sentence of the type anāku šarrum “I am (the) king” occurs in Akkadian proper: šarr-āku or anāku šarrum-ma must be used instead. The predicate occurs in the normal or construct state when followed by a qualification or complement: anāku šarrum damqum “I am a good king” or atta šarr-ī “you are my king.” Thus an analysis of the stative belongs to syntax, since šarr-āku and anāku šarrum damqum are strictly equivalent, i.e., they are both nominal sentences. Also, šarr-āku must be correlated in function to the West Semitic nominal sentence (ʾanōkī melek), not the West Semitic perfect (mālaktī).

– Giorgio Buccellati, 1994

PDF available here.

– Giorgio Buccellati, 2016

The aim of this volume is «to provide a description of Babylonian which may serve both as a systematic theoretical statement of the structure of the language, and as a guide towards a better understanding of the textual record» (p. VIII). The Grammar analyzes mostly Old Babylonian in a synchronic way, but it also quotes examples from later periods when Old Babylonian instances are lacking; it is also enriched by several sections which refer to the historical background of specific phenomena.

[The volume to which this website represents the companion.]

PDF available here

Review by N.J.C. Kouwenberg in Bi. Or. 55/1-2 (1998)

– Stefania Ermidoro, 2020

G. Buccellati offers in this paper an overview on Akkadian and Amorite phonological system.

PDF available here.

See also abstract.

– Marco De Pietri, 2023

The fundamental volume of G. Buccellati about Mesopotamian and Israelite religion.

See also the companion website.

– Giorgio Buccellati, 2016

G. Buccellati’s volume about the development of politics in Ancient Mesopotamia.

See also the companion website.

– Giorgio Buccellati, 2016

In A Critique of Archaeological Reason, Giorgio Buccellati presents a theory of excavation that aims at clarifying the nature of archaeology and its impact on contemporary thought. Integrating epistemological issues with methods of data collection and the role and impact of digital technology on archaeological work, the book explores digital data in order to comprehend its role in shaping meaning and understanding in archaeological excavation.

Wider description available here. See also the companion website.

– Giorgio Buccellati, 2017

G. Buccellati’s volume about literature in Ancient Mesopotamia.

See also the companion website.

– Giorgio Buccellati, 2017

The reference book of Giorgio Buccellati about Hebrew and Mesopotamian spirituality.

– Marco De Pietri, 2023

An interesting and lucidly written article, in which the author defends, esp. against Barth, the derivation of the WSem. perfect from the Akk. permansive, on the grounds of formal correspondences not only in the B-stem (p. 204 f.), but especially in the derived stems (p/ 206 f.). The permansive is the combination of an adjective with stative meaning plus a pronominal suffix, or a nominal sentence which, if the pronoun is suffixed, “becomes a verb” (p. 205); if the predication of the state expressed in the adjective is limited to particular circumstances, it may become active (p. 207). A review of the main usages of the permansive. Apart from the rare active usage, the permansive is normally used to express timeless state, whereas the preterite is used to denote “the emergence of an event into being out of a preceding condition,” hence the greater frequency of the preterite in narratives where “events involving action are naturally pictured as coming into being out of a fixed point in time” (p. 209). The rest of the article, dealing primarily with Hebrew, objects to the use of the notions of perfect and imperfect, instead of those here suggested of condition and action, and suggests to adopt (for Hebrew) the terminology of “aorist” for perfect and “subsequent” for imperfect.

– Giorgio Buccellati, 1994

After emphasizing the methodological difficulties in the dialectal classification of Semitic languages, and after giving reasons for considering South-Arabic as a linguistic unity (the unity of Akk. is assumed), the author considers the main points which have been made in favor of considering the two languages as a common dialectal group: the pronominal suffix of the third person in š, the free combination of derived verbal stems, the prefix š of the causative, the imperfect or Akk. present of the type iparras, plus some lexical items. The conclusion is that the similarities are more likely the result of archaisms dating to the time before the separation and preserved independently, rather than common innovations after the separation. Hence Akk. and South-Arabic should not be considered as constituting a dialectal group.

– Giorgio Buccellati, 1994

A review of Bravmann 1938 (“Hilfsvokal,” Le Monde Oriental 32 (1938), pp. 1-102), accepting some of his applications and criticizing others. Also criticized is Bravmann’s recourse to stress to explain the origin of the phenomenon.

– Giorgio Buccellati, 1994

A clarification of the concept of pattern, with a study of its alterations as due especially to phonological change. The author points out various phonological encroachments upon the morphological system (“le triomphe de la phonétique sur la grammaire,” p. 83), whereby the unity of the pattern as carrier of meaning is broken, and the pattern is split into combinatory variants (see esp. p. 79). Most examples are from Arabic and Northwest Semitic.

– Giorgio Buccellati, 1994

A fundamental article dealing with the basic categories of internal inflection. First, the reality of the Semitic root is defended (pp. 119-21): the root is an abstraction, but of the same kind as other current types such as phoneme, prefix, etc.; more precisely, the root consists of implicit elements made explicit by associative analysis. A differentiation is made between verbal and substantival root. The second part (pp. 123-4) analyzes the pattern (“schème”) which is the form or structure of the word (to be differentiated from the “theme” which is the word without inflectional indices.) Patterns belong to grammar, roots belongs both to the lexicon in their individual signification and to the grammar in their system, general form and usage.

– Giorgio Buccellati, 1994

Dealing primarily with living languages, the author considers the general agreement on the subject as unwarranted, and reexamines the basic criteria for the distribution into classes according to place and mode of articulation with new observations especially for apicals (= dentals), sibilants and emphatics.

– Giorgio Buccellati, 1994

The personal pronouns in Hamito-Semitic occur in two main sets: subject and genitive- object (rather than separate and suffix). The two sets are different in the stems, but parallel in the patterns used. Various points of detail are made, especially for Cushitic and Egyptian, in the attempt to foster a historical understanding of the Akkadian system. Some of the conclusions relevant for Akkadian: vocalic uniformity in the set of subject prefixes (ʾaprus, taprus, etc.) goes back to common Hamito-Semitic (p. 10); n as a plural marker in the set of subject pronouns (nīnu, attunu, etc.) also goes back to common Ham.-Sem. (23, 35); t used for the subject set, and k for the genitive-object are the result of a special development in Semitic (33); the dative marker š, peculiar to Akkadian within Semitic, is also attested in Cushitic (36). Verbal roots are built on the pattern CCVC, nominal roots on CVCVC: the latter, more “static” in semantic value, are probably original over the verbal roots, which are “dynamic.” The vocalism of the last syllable has semantic value: u stresses the subject agent and is appropriate for intransitives; a expresses the connection between subject and object, and is appropriate for transitives; i is ambivalent (like the genitive, which can be either subjective or objective). As for the tenses, Proto-Sem. had a two aspect system consisting of imperative (iprus) and perfective (paris), to which Akkadian added present (iparras) and perfect (iptaras). The “ventive” mood is understood, on the basis of Cushitic, as expressing emphasis or special interest on the part of the speaker. The permansive is close to the nominal conjugation of Cushitic, Hamitic and Berber, and is therefore common Ham.- Sem.; in Akkadian it provides a mechanism for activating nouns by giving them verbal dynamism, while remaining at the same time independent of and outside the proper system of tenses. The conjugation system of Akkadian is closest to Proto-Sem. The description of the semantic value of the stems is based on the notion of various degrees of verbal dynamism (for this concept see pp. 43, 47, 73, 92, 124ff.). (See EDZARD 1964; VON SODEN 1965).

– Giorgio Buccellati, 1994

Against the traditional bipartition East vs. West Semitic, the latter comprising Arabic and “Abyssinian” as a unitary group (South Semitic), the author isolates South Arabic within South Semitic, and points to its relationships with Akkadian in the following areas: prepositions, “tenses” and stems of the verb. Akk. Present is archaic (and iqattal is the originary form rather than iqatal, p. 735), and so is the suffix conjugation, a conclusion which he supports with “Hamitic” evidence. The broken plurals, of which Akkadian shows only incipient traits were developed in South Arabic under the influence of Arabic. Akkadian and South Arabic represent the oldest stage of Semitic. The conclusion correlates the results of linguistic analysis with data of physical anthropology.

– Giorgio Buccellati, 1994

1. The ending of the accusative is originally -ã, i.e., a with nasalization of -an/m). The same nasals in a demonstrative sense are found in particles (e.g., Akkadian anna “yes,” umma for direct speech), in the marker for determination (mimation in a prehistoric stage of Akkadian) in the prefixes of nominal patterns (which appear as frozen genitive constructions, esp. participles in mu- and nomina loci in ma- meaning “that-of-[the verbal concept],” as well as in suffixes (gentilics like Aššurāʾaʾu “the-one-of-Aššur,”) ordinals, abstracts, and elsewhere. 2. The element s, phonologically connected with h/ʾ, is also demonstrative in nature and is found in interjections, personal pronouns, determinative pronouns, the terminative -iš, adjectival patterns šaqtal/ʾaqtal, and the conditional conjunction šumma. 3. The element t/d has demonstrative functions in the Arabic determinative pronoun, the feminine, pronouns, prepositions (itti), and conjunctions. 4. The element k/g has demonstrative functions in pronouns, prepositions, and conjunctions (ki). The same elements are also found substantivized (e.g., š in šumu “that-of-someone,” “name”) and verbalized (in the stem affixes). The exemplification is from Hamito-Semitic and Sumerian (!). [Methodologically, the article is indicative of an exaggerated tendency to reduce everything to a few proto-forms, with a cavalier attitude toward phonological and other rules.]

– Giorgio Buccellati, 1994

Various correlations are introduced to show that the case endings are all of demonstrative-pronominal nature. The Akkadian plural u/i is explained as contracted from -āʾi, being the unnasalized form of -ān- (p. 24).

– Giorgio Buccellati, 1994

After a critical review of earlier literature, especially Bauer, in which the general failure to extend the analysis outside of Semitic is emphasized, the main facts of the Hamitic verbal system are summarized: a is used for immediate present, i for proximity, u for distance, all of them derived from local categories. Two main points are then stressed: (1) The purpose of the vowel was not to differentiate tenses, but rather fiens (u, with variant i) from factum (a). Both categories had originally prefix and suffix conjugations, of which eventually the first became standardized for action, the second for state. (2) Psychological categories help to explain the change in position of the afformatives. (See also ZDMG 83 (1929), pp. 90-1.)

– Giorgio Buccellati, 1994

Intensity is the basic function of the D stem, and is explained historically by deriving the morpheme from incomplete reduplication: qatal and qat’al (dissimilation) and qattal; it expresses plurality of subject or of object, longer duration of the action, greater degree of effectiveness, greater degree of concern about something. The latter leads to the use of the stem in a denominative and declarative sense, and through this to the causative sense (“concern about the actualization of a verbal concept”). In the š stem, š/s or h/ʾ (etymologically identical) are “deictic elements referring to distance,” so that the stem expresses direction and has an inchoative meaning (labāru “to age,” šulburu “to begin to age”). Therefore the occasional overlaps in meaning between D and Š are from completely different starting points.

– Giorgio Buccellati, 1994

A clear and concise survey by the dean of comparative hamito-semitic studies, placing Akkadian within the context of its linguistic family (pp. 99-104).

– Giorgio Buccellati, 1994

An analysis of verbs of the type labisa/yalbasu, for which the Akkadian equivalents are given as roots with root vowel a (ṣabit/iṣbat). They are considered as internal deponents, or as “adherent” verbs, in that they place an emphasis on the effect of the process on the agent (thus e.g., for Arabic šariba “to drink” the process is seen from the viewpoint of the stomach, rather than of the drink).

– Giorgio Buccellati, 1994

After renewing the argument in favor of the linguistic reality of the root, the question is asked as to whether an affixed form is dimorphemic, say ma + qtal rather than maqtal. The author’s answer is that there is no dimorphism in the case of prefixes (pp. 76-80): a preformant is not a morheme by itself, but rather the prefixed pattern as a whole is a discontinuous morpheme. Suffixes, on the other hand, are mostly to be analyzed as separate morphemes (pp. 80-92). Hence there is a limit to the use of internal inflextion in forming and inflecting words, since suffixation is a different procedure. The author’s examples are from Arabic and Ethiopic, but his theory applies equally well to Akkadian, and has been followed in this grammar.

– Giorgio Buccellati, 1994

Revised translation of Latin notes published for the students in 1911. After general remarks about phonetic laws, describes the Akkadian word stress and gives a comparative list (in Akk. and Hebr.) of abstract nominal patterns from the strong verb and of attested patterns from roots first-nun.

– Giorgio Buccellati, 1994

The notion of aspect is connected not with the tenses, but with the nature of the root. The verbs of state express durative condition and occur primarily in the permansive. The verbs of action express one of three aspects—momentary or punctual (“to die”), iterative or punctative (“to tremble”) and durative or linear (“to pull”); they occur in the preterite for the past, and in the present for present and future. The verbs of state become verbs of action (as ingressive) when used in the preterite or present (“to become great”).

– Giorgio Buccellati, 1994

The first classical formulation of Akkadian grammar, more comprehensive and detailed than any of the earlier syntheses, with a broad documentary basis and a serious effort at internal coherence and systematization. Truly a masterpiece for its time, it retained a unique value as a reference work for over 60 years, until VON SODEN 1952 Grundriss, and is still today worthy of careful reading. Besides the traditional tripartite description of the language in phonology (pp. 72-127), morphology (pp. 128-322) and syntax (pp. 323-64), it contains an overview of previous research and a section on writing. The approach is of course influenced by the grammar of the other Semitic languages, but comparative considerations are kept to a minimum, and the language is mostly seen from within its own system. Only seldom does comparativism lead to erroneous results (e.g., the consideration of t- as the regular prefix of 3rd sing. fem. in the indicative, p. 249, 251); otherwise, odd phenomena are recorded as such, without attempts at leveling data to make them square with preconceived schemes (e.g., for our ventive, p. 134, 254, and our perfect, p. 239). The main feature of the book is that, even though intended as an introductory manual (see esp. p. 80*), it contains a considerable amount of original discussion of grammatical problems, and an impressive array of documentary evidence; see esp. the sections on Sumerian (pp. 69-71, 195-97), the vowel e (pp. 74-79), stress (pp. 120-127), primitive nouns (pp. 146-50), biconsonantal roots (pp. 143- 57, 313 f.), verb forms of state and action conceived as “Existenzweisen” (pp. 234-49), the nature of final weak radical (pp. 297-302), and root vowel (196 verbs are listed for strong roots and roots I/n, I/ʾ, pp. 261- 65, 273 f., 280 f.). Good documentation is given even for relatively rare forms, such as the ŠD stem (p. 233 f.) and then for every type of root, strong and weak.

– Giorgio Buccellati, 1994

In a date formula from Ešnunna i-ḫa-ZU stands for iḫḫaṣu = imḫaṣu and not for ʾīḫuzu, though it may also be a scribal mistake for i-<im>-ḫa-ṣu.

– Giorgio Buccellati, 1994

In NA “Our lord” is rendered, in nominative and accusative, by bēlīni (the same form is also found, as an Akkadogram, in Hittite), which is probably a plural. Other such forms were then derived by analogy: ṭēmī-ni, dullī-ni, zittū-ni.

– Giorgio Buccellati, 1994

After emphasizing the need of dialectal-descriptive studies, gives a list of the nouns (with brief Akkadian context, translation and references) occurring in the OB texts of Susa; the nouns are divided according to the nominal patterns as given by VON SODEN 1952 Grundriss.

– Giorgio Buccellati, 1994

An analytic presentation of the conjunctions found in NB letters, a corpus of about 1100 letters; 53 different conjunctions are listed in alphabetical order (most of them are compounded on the basis of 9 or 10 basic conjunctions), carefully indicating for each the notional value, the consecutio temporum, and occasionally the word order. An exhaustive documentation is given for each conjunction.

[I have counted a total of 993 examples in Dietrich’s documentation. It is important to note that 733 of these, i.e., almost 75%, are with ki or a compound based on ki; in the second place of frequency are adi and ultu, or compounds, with 68 and 61 examples respectively, i.e., almost 15% together. In other words, in spite of the typological wealth, the productive conjunctions are very few, and ki is by far the most common; in turn ki, is especially common in a temporal function, with 415 examples, i.e., about 55% of all ki occurrences. Structurally one may also note that, of the elements entering in composition, bīt, lā, and muḫḫi occur in final position only, adī and ultu occur alone or initial, adi, ki and ša occur alone, initial or final.]

– Giorgio Buccellati, 1994

The element -t- is “inversative” in that it serves to express the contrast to a basic form: in the noun, it expresses the anti-masculine gender, i.e., the feminine; in the verb, the change of direction of the change, hence it is not only reflexive, as commonly assumed, but also passive and especially durative. Some of the argumentation is based on von Soden (though no Akkadian documentation is given), but see VON SODEN 1965 “Sprachvergleichung”

– Giorgio Buccellati, 1994

1) Distinguishes three main types of nouns: those which are originally nominal and are not necessarily triradical (nominal root), those which are from roots nominal and verbal at the same time (e.g., bēlu “lord” “to rule”); and the nominal deverbalia which are derived from originally verbal roots by means of “innere Nominalbildung” or of “akkreszive Elemente” (internal and external inflection). For various reasons, descriptive and comparative (pp. 233-35) the verb is considered primary over the noun, and more precisely in the basic stem. 2) Criticizes in detail the reasons adduced by others in favor of the priority of the aorist (= preterite) over the perfect (= stative): the last vowel of the aorist is neither more varied (esp. in Akkadian), nor more regular than in the perfect (p. 251 ff.); there is nothing to prove that prefixation is a frozen and therefore older morphological process (p. 258 f.); there is no general basis to the claim that expression of the present antecedes that of the past (p. 258 f.); the “Hamitic” languages do not prove the priority of the aorist (p. 259 ff.); the aorist cannot be derived from the imperative, esp. in Akkadian (p. 262 ff.). Instead, the perfect is considered primary, because nominal forms may be more easily derived from it, and because Egyptian has no aorist (p. 272 ff.). 3) Verbal roots (as they appear in the primary form of the perfect) may be divided in three categories: originary (e.g., from the previous two elements גאר “to end, boundary” and בוא “to go”); derivative (e.g., Akk. qerēbu “to approach” from the previous two elements), and determinative (the weak verbs, which add determinative elements, to originary verbs). Triradicalism is thus a secondary phenomenon, which was accepted more systematically in Semitic than in Hamitic. The final section is about onomatopoetic verbs, with pa

– Giorgio Buccellati, 1994

The feminine marker t was originally deictic and was then “extended” to distinguish other things that required to be distinguished, such as, precisely, gender; in other words the marker t “is not an original element distinguishing the female sex or feminine gender” (p. 90f.). Similarly, the masculine form of the numeral is original, and it was originally used with nouns of either gender. When the numeral proceded the noun to which it referred, it was in the construct state and was taken as a collective (marked by the t of the feminine); if the noun was also a collective, the sequence of two collectives was felt undesirable, and the numeral placed in the non-collective, i.e., masculine form. The “disagreement” was then normalized to all cases, even when no collective was present. The central part of the article contains documentation from the main Semitic dialects, Akkadian being treated on pp. 92-5.

– Giorgio Buccellati, 1994

In personal names including a feminine divine name and applying to a male individual, the agreement of the predicate is with the individual rather than with the divine name. This rule applies regularly when the predicate is nominal (^mNingal-gāmil); with a verbal predicate (Ištar-idinnam) it is regular only in OB and NB/LB. This emphasizes the fact that PN’s were considered as compounds, so that the gender characterization is meant to apply to the name as a whole.

– Giorgio Buccellati, 1994

The morphological structure of the stem system is presented on the basis of the syntactical structure, previously established by the author in the same article. The system is based on the interrelationship of minimal pairs (syntactically), the members of which are defined both by functional and morphological opposition. Morphologically each pair is characterized by the presence or absence of specified markers (e.g., prefix n = passive (stem N), prefix = active (stem B).
The only exception is the correlation infix t + long middle radical = factitive passive (stem Dt). These markers are isolated without recourse to the notion of internal inflection; rather they are defined on the basis of affixation, with a uniform vowel whenever needed.

– Giorgio Buccellati, 1994

Synonyms or words with similar meaning occur in regular sequences, the criteria for the arrangement of the words being that words which are shorter in terms of number and/or quantity of syllables occur in initial position, while words which are longer occur in final position, e.g., uṣṣu ~ mulmullu. A unit of two words may also be used in place of a single long word, e.g., narū ~ šumu šaṭru, surrāti ~ lā kēnāti,. The principle applies also for sentence structure, so that within the same sentence long clauses follow shorter ones, e.g., appul aqqur ina išāti ašrup. Applications of the theory follow, with respect to lexical texts (pp. 25-34) and conceptual categories (pp. 36-49).

– Giorgio Buccellati, 1994

Remarks on the most important patterns. Note especially: fiʿl is for substantives from transitive roots (e.g., pilšum), fuʿl is for substantives from adjectival roots (pušqum), faʿl for (1) primary nouns (i.e., concrete objects, qarnum), (2) substantivized adjectives (aplum), and (3) very few substantives from verbal roots (paṭru); the adjective of the Š stem is a true elative as in Arabic.

– Giorgio Buccellati, 1994

Only formal criteria have been used so far for the arrangement of patterns. This article is meant to emphasize the semasiological evolution in order to bring out more clearly the fundamental meaning of the patterns. Two main types of evolution are observed. (1) Use of substantives as adjectives (e.g., rugum “distance > far”) caused by ellipsis, predicative and adverbial use of the substantive apposition. (2) Use of noun of object for persons, i.e., personification (e.g., nurbum “proximity > near relative”; see esp. pp. 126-33 for Akkadian).

– Giorgio Buccellati, 1994

Essentially etymological in character, it isolates various particles as constitutive elements of nouns and esp. of adverbs. For Akkadian: ša in aššum, šumu, ištu (44: 185-87); ta in itti (45: 48 f.) but not in the adverbial -iš which is of Sumerian origin (pp. 135-37); n in ana (p. 144 f.); la in ultu, lū (pp. 207-9); Egyptian (!)‛ in ina (46:46-50).

– Giorgio Buccellati, 1994

Ugaritic evidence and modern literature (especially GOETZE) on the possibility of a formation yaqattal in Northwest Semitic is examined, and a negative conclusion is reached. If Akkadian iqattal is in fact to be linked with South-Semitic forms, one must assume either that yaqattal, though originally common Semitic, was lost in Northwest Semitic, or (and more likely, according to Fenton) that yaqattal was “created only in a part of the Semitic (Hamito-Semitic) area and never spread over the entire area.”

– Giorgio Buccellati, 1994

Mahrûm-ma as a locative adverbial, or mahrûm-ma as an accusative-adverbial, means “instantly” in the texts from Mari; emphatic -ma serves in this case to form an absolute superlative.

– Giorgio Buccellati, 1994

Differently from common opinion, the origin of the notion of a tripartite division noun, verb and particle is derived from Aristotle’s Poetics, chapters 20-22. After a well-documented and instructive survey of the material pertaining to the Syro-Arabic transmission of the Poetics, the author gives an interesting analysis of the main grammatical terms in Aristotle, and their equivalents as used by Arab grammarians.

– Giorgio Buccellati, 1994

Common Semitic had originally two verbal forms, characterized by number of syllables and, secondarily, by stress: a short form yáqtul for completed action, and a long form yaqtúlu for incomplete action. The coexistence of both forms is proved by the fact that various traces in the historical languages show that the two forms are essentially indifferent with respect to mood; the data may be summarized graphically as follows, with parenthesis for infrequent attestation:

Akkadian ikaššad and Geez yaqáttal are both secondary: the first is “un nom d’agent conjugué avec préfixe” (p. 50), while the second is a substitute for the long form yaqtála which, due to phonological development, had become less distinctive with respect to the short form yáqtel (p. 58). For Akkadian note especially p. 48, where Saraw’s view concerning the consonantal lengthening of the type ikunnū is tentatively accepted.

– Giorgio Buccellati, 1994

– Giorgio Buccellati, 1994

See below Fleisch 1961 Traité.

– Giorgio Buccellati, 1994

Stresses the internal unity of the sentence, which in the case of Arabic is marked by three facts: strong phonological unity among words of the same sentence (ligature of word cluster initial with preceding word, crasis and sandhi), designation by Arabic grammarians of the clauses as specific units, pause as formal marker of sentence final.

– Giorgio Buccellati, 1994

Though neither the Esquisse (which contains a simpler statement of the author’s grammatical theory) nor the Traité (where the theory is further developed and a wealth of material added from ancient Arab grammarians) deal in any way with Akkadian, both volumes are mentioned here because they provide an excellent application of some basic principles followed in my own grammar. This is true especially for internal inflection, which includes both lexical derivation (e.g., nominal patterns) and grammatical alternation (e.g., broken plural), but excludes the pronouns. Concerning the affixed patterns, the author follows the thsory which derives them from originally independent morphemes, whereby ’ā (as preformative of the pattern ’af‛alu) is a “particule interjactive du language affectif,” or ma a relative particle. One should not fail to appreciate the delightful French style, especially of the Esquisse.

– Giorgio Buccellati, 1994

On the basis of general linguistic theories (esp. Benveniste) the author interprets (pp. 251-57) the relative phrase as an apposition characterized by determination, i.e., the rel. phrase behaves like a determined noun in apposition. After a survey of the relative phrase in Arabic and Sumerian (pp. 257-67), the author criticizes von Soden’s conception of the Akkadian relative phrase as taking the place of a syntactic genitive and maintains that it is instead an apposition (pp. 269-75). In a phrase like bīt īpušu the construct state is not meant to introduce a genitive, but rather to express determination (pp. 274-75). Against Kienast 1960 (“Punktualthema”), the author maintains that the ending -u of the Akkadian subjunctive is not that of a Proto-Semitic modal ending (see esp. pp. 248-49) but rather that yaprusu is the Proto-Semitic form to express unaccomplished time (pp. 275-82, esp. p. 281 f.), thereby rejecting that the form iparras may be Proto-Semitic.

– Giorgio Buccellati, 1994

– Marco De Pietri, 2023

The introduction deals briefly with the notion of dual from the point of view of historical linguistics. The first part reviews the data in the individual languages; Akkadian is treated on pp. 36-46 (based on secondary literature): the dual is productive in Old. Akk., and used also for occasional dual (i.e., for objects not occurring naturally in pairs); from OB on it becomes archaic and restricted practically to natural pairs; in the verb, the dual is only archaic. The second part of the book reviews in detail the secondary literature on the Semitic dual; Akk. appears substantively only on p. 199 f., where it is used to prove the existence of the dual in Proto-Semitic.

– Giorgio Buccellati, 1994

Against Brockelmann, and following Rössler, the author defends the antiquity and importance of Akkadian in comparative Semitics, particularly with regard to the Akk. present as a Proto-Semitic form.

– Giorgio Buccellati, 1994

Primitive (or unmotivated) nouns do not exhibit apophonic alternation in their vocalism, nor can one establish semantic classes of such words based on vocalic differentiations hence the vowel of primitive nouns is part of the lexeme, not of morphology (internal inflection). The same conclusion is suggested for the imperative and prefixed verbal forms of the B stem of verbs of action, as well as for the suffixed verbal forms of the verbs of state.

– Giorgio Buccellati, 1994

An attempt to locate Semitic languages within the framework of a universal typological perspective; special reference is made to Old Babylonian. The starting point is an article by J. H. Greenberg, in which agglutination is defined as mechanical construction of morphemes (rather than as inalterability of the root); agglutination and its counterpart, fusion, may exist in the same language, and the proportion of their occurrence in a language is understood as the index of agglutination. On the basis of his examination of a sample text of 100 words from the Code of Hammurapi, Fronzaroli gives as index for 08 0.01, one of the lowest on the scale (next to Eskimo; the index for English is 0.30): this means that 08 is fusional to a high degree. With respect to derivation and inflection, the situation is different: ‘for the same 08-sample, the index of derivation is 0.59; of inflection 1.40; of prefixation 0.24; of suffixation 0.98; of infixation 0.70. In the last section the author analyzes the notion of “word” in Semitic languages, expecially in such difficult cases as the construct state.

– Giorgio Buccellati, 1994

Critical remarks about the traditional methods of comparative Semitics, followed by a typological study of single representative phenomena and a reassessment of inner-Semitic relationship. The causative and pronominal morpheme š/s is common not only in Akkadian and South-Arabic, but also in “Hamitic”; besides, it is also found to a limited extent in NW Semitic (except Aramaic) and Arabic, whereas the parallel morpheme h is not found outside of the latter two groups; hence š/s (as in Akkadian) is archaic, whereas h is a NW Semitic innovation. Similar is the situation for the Akkadian interrogative pronouns mannum-mīnum, formed of the adeictic pronominal theme m plus the deictic element n, and equally archaic. From a general point of view and with special reference to morphology, Akkadian is the most archaic Semitic language. Unitary linguistic traits among Semitic languages are the result of developments from the second millennium on (neo Semitic); for the archaic phase it is more difficult to presuppose true linguistic unity, particularly between Akkadian and West Semitic. Rather, one should investigate possible isoglosses between single Semitic languages and “Hamitic” groups (very tentatively, a parallel is suggested between Akkadian and Cushitic—both of which exhibit the verb in final position).

– Giorgio Buccellati, 1994

No Akkadian root contains two different emphatic consonants. When such combination would be expected according to comparative Semitics, the corresponding voiceless (only occasionally the voiced consonant) appears. The relative “strength” of the emphatics may be stated as follows: ṣ is always retained, and q is retained over ṭ.

– Giorgio Buccellati, 1994

Theoretical and systematic treatment of the graphic rendering of Akkadian, touching on phonological questions such as the distinction between phonemics and phonetics (pp. 525-26, 531-36) or problems of phonemic distributional classes, especially diphtongs (pp. 539, 541) and initial y (pp. 537-39).

– Giorgio Buccellati, 1994

The t-form occurs in complementary distribution with other forms, hence it must have specific functions; these are described under three main headings. (1) The t-form as a tense in coordinated clauses: it occurs normally after preterites and before the present, thus serving “to link the past to the present,” and denoting “the action which has just been performed and still affects the situation” (p. 312); the best translation is by means of an English perfect. A simple preterite is often found instead of the t-Form when there is a negation (pp. 313-16), with certain verbs like bašū (p. 317), when followed by a dative personal suffix (pp. 318, 332). But the t-form is not properly a tense because the t element is found also in the imperative, prohibitive and precative (p. 319)-— (2 ) The t-form as a relative-tense in subordinate clauses: it functions in a similar way as in coordinate sentences, in that it “refers to the action which, seen from the speaker’s point of view, has just been performed and is still of actual interest” (p. 321). Thus it serves to express the consecutio temporum as a past perfect or a future perfect. — (3) The t-form as an aspect: the t infixed to the present and also to the preterite may have the objective function of modifying the meaning of the root (unlike the tense t-form which defines simply the point in time). The functions are two: reflexive - reciprocal (p. 323 f.) and more frequently separative which “denotes a movement from a fixed point in the direction of an unspecified goal” (p. 324); an extensive documentation arranged alphabetically by verb follows. - In the conclusions, the various functions of the t-form are derived from the basic Semitic function as a reciprocal-reflexive, the productivity of the separative function being justified on the basis of an influnece of the Sumerian prefix ba-. -It may be noted that while no term is given to the tense-function in the title, the term “aorist” is introduced on p. 321 and then used regularly in the article.

– Giorgio Buccellati, 1994

A new phoneme ẓ (as in Arabic ẓā) is postulated for Old Babylonian on the basis of spellings in the Code of Hammurapi and private letters and contracts (though not royal letters) which occur in the following manner: written [zz] for phonemic /zz/ and /ṣṣ/, and written [z] for /ss/ and /ẓẓ/ as well as for /z/.

– Giorgio Buccellati, 1994

While in phonology Nuzu Akk. is close to Middle Babylonian, in other respects it shows substantial differences, e.g.,: the treatment of sibilants in the writing system (derives from Old Akk.); in morphology, the use of -me to form the indefinite relative (mannu-me “whoever”), lack of feminine forms of the verb, suffix -u for the 1 p. plural of the verb (nilqū-ma); in syntax, peculiar rules of governement (direct object in the nominative when before the subject; verb in the plural when direct object is in the plural), and use of -ma for emphasis only. The conclusion is that Nuzu Akk. is a third dialect next to Assyrian and Babylonian.

– Giorgio Buccellati, 1994

A thorough analysis of Ugaritic verbal forms, with the conclusions that (l) the suffixed form qatila is used more like the Akk. permansive than like the WSem perfect, especially because it has a stative connotation and does not funcion as a tense; (2) there are two prefixed forms, just as in Akk., namely yaqtulu for the past, and yaqat(t)alu for the present-future.

– Giorgio Buccellati, 1994

An important study on the syntactic and semantic function of the D stem, in which the common interpretation of the stem as primarily “intensive” is convincingly refuted. In a review of the verbal system as a whole, the difference is emphasized between Akkadian and the other Semitic languages, especially in that the Akkadian stative is not part of the verbal system. The conclusion is then drawn that the D stem is derived from the stative of the B stem (p. 5), and that its primary function is denominative; more spacifically, when derived from an adjective, the D stem is factitive. Just like the stative, the D stem may be divided into 3 categories: (l) durative, with meaning “to put a person or a thing in the condition which the stative indicates,” e.g., ṭubbum “to make good”; (2) perfect, “to make somebody have something” e.g., lubbušum “to clothe”; (3) passive, “to put a person or a thing in the state described by the stative,” e.g., uḫḫuzum “to make something fit.” Other apparent D forms are in fact Bn, e.g., yu-rá-n-qid uraqqid “he danced.”

– Giorgio Buccellati, 1994

Non-mathematical texts belong to two main OB dialects, Northern (Hammurapi court, Dilbat, Sippar) and Southern (Larsa). Mari seems to belong to a third group, whereas too little is known about Ur and Uruk. Several features that distinguish the two dialects are singled out, most of them dealing with orthography: signs with emphatic stops ( /ta/ = TA in the North vs. DA in the South), with samekh (SA, AS vs. ZA, AZ), with labial ( /pi/ = BI vs. PI), with y ( /aya/ = A-IA vs. A-A), marker of contraction or intonation length (used regularly in the North, and extended in the South to cases without contraction or special intonation), phonetic complements (VC vs. CVC). Phonologically, nasalization of long voiced stop begins in the South. Syntactically, addition of the subjective genitive to the infinitives (tuppi ina amārī-ka) is typical of the South. Lexically, a word such as unnedukkū is typically Southern. -On the basis of these criteria, Goetze distinguishes the mathematical texts in groups coming from Larsa, Uruk, and the North.

– Giorgio Buccellati, 1994

Important observations on general linguistic method as applied to Akkadian, and on the following specific points: nature of the ŠD stem, existence of an -n- infix stem, non-quadriconsonantal nature of šukēnu and šupēlu.

– Giorgio Buccellati, 1994

“It is a universally recognized fact that Akkadian does not tolerate the sequence of two or more short open syllables in the middle of a word.” This rule does however admit some exceptions, which may be grouped in two main categories, (l) Sumerian loan words with double consonant closing the word in Akkadian, such as abarakkum, igigallu; this exception may perhaps be due to a special stress pattern (p. 235, n. 5). (2) Words with a in second syllable, immediately followed by r, as in išarum “fire.” A list of such words is appended with references from OB texts: išarum “fire,” zikarum “male,” šikarum “beer,” ašariš “there,” eperūm “earth,” eperu “provisions,” ebertum “territory beyond (a river etc.),” ašaret “ten,” and as Sum. loanwords: igarum “wall,” amarum “pile of bricks,” šigarum “bolt,” ugarum “uncultivated part of the territory of a city,” diparum “torch,” Šubarūm “Subarian” (pp. 234-35). Cases like wagrum, instead of the expected wagarum are to be explained as formed analogically on adjectives of the common pattern paris: qamirum (p. 237). Special cases: la-bi-rum is to be explained as lābirum (p. 236, n. 4); nakrum is based on two patterns, paris and paras (p. 237).

– Giorgio Buccellati, 1994

An important article, with good documentation concerning (a) the phonemic shape of cardinals, ordinals and fractions and (b) their syntactic rules of agreement and word order. As for (b) the author shows that fractions are construed as substantives (i.e., normally in the construct state), ordinals as adjectives (i.e., in complete agreement with, and in a position following the head noun, except for nouns of time, which are preceded by the numeral: ina rebūtim šattim “in the fourth year”), while special rules obtain with the cardinals, as follows: disagreement of state between numeral and head noun, disagreement of gender (except with numerals 1 and 2), agreement of number, and position indifferent (i.e., numeral either before or after head noun); as for the state, the numeral in the absolute state is indeterminate, but in the normal state it is determinate, e.g., kibrāt erbētim “the four quadrants,” erbē ubanātim “four fingers.”

– Giorgio Buccellati, 1994

After a short history of the problem a list of all OB examples of the plural in -ānū known to the author is given. This special plural form is taken to stress individuality, in that it refers either to “some” of the units counted or to all the units taken “individually.” His examples are 33 in number, and include the following words: il-, šarr-, bēl-, simm- “spice,” āl-, ekall-, dūr-, mūt-, murṣ-, rugummāʾ-, tīr-, širḫ-, mišl-. The corresponding logographic marker is DIDLI which also stresses individuality. The second main portion at the article is comparative in nature, and shows that the suffix -ān of Semitic is used to (l) create concrete out of abstract nouns, (2) mark an individual as opposed to a class, (3) ascribe an adjectival quality to a specific individual.

– Giorgio Buccellati, 1994

After a tabulation of 39 pairs of preterite/present N with final vowels u/u or i/i, the conclusion is drawn that the vocalization of the N stem and that of B are identical except for two cases, which are emphasized in the following scheme: B i/i. u/u. a/a. u/a N u/u u/u i/a i/a. The innovative element is the i of the preterite, a fact which implies that the non-innovative element (i.e., the present) must be traced back to Primitive Semitic.

– Giorgio Buccellati, 1994

The particle ša is not “generalizing” because it serves to individualize, i.e., to add a more precise and definite specification to the head of the construction. This is apparent especially where ša serves for annexion (of which Goetze states the distinctive conditions vs. the use of the construct state) and in the expression of the type ša nadānim which is not indefinite (“something to pay with”) but rigorously definite (“the amount necessary for full payment”). Clauses with ša without immediate antecedent normally precede all other constituentes in the sentence (ša kaspam inaddinu še’am ileqqe) or else, at least in the Code, the pronoun used is šu instead of ša. The conclusive definition of ša given by Goetze is: “a particle of reference which ascribes a specification to a definite person or a definite thing” (p. 79).

– Giorgio Buccellati, 1994

Noting that ths safest criteria for determining vocalic length in Akkadian are syncope and Assyrian vowel harmony (both of which occur only with short vowels), the author studies seven words that exhibit some difficulty as to their vocalic quantity.

– Giorgio Buccellati, 1994

Depending on the treatment accorded to sibilants, various dialects are distinguished in OB, all of which can be grouped in two major categories: the first one, at the margins of the OB area (Southern OB, Mari, Susa), retains a hitherto unrecognized proto Semitic sibilant designated as s_x, while the second or central group (Northern OB, Upper Mesopotamia) represents an element of change in that sₓ coalesces with PSem. sₓ. Similar differentiations may be observed in the treatment of the pronominal suffixes of third person after words containing sibilants or dentals. These conclusions are based on a detailed treatment of the writing peculiarities in the various dialects, summarized in a detailed chart at the beginning. The new consonant s_x posited by Goetze has not been indicated in my phonemic inventory for, besides being in any case a phenomenon of limited attestation geographically and lexically (only 8 words are affected) the whole problem is in need of more discussion (for OAkk. see Gelb 1961, MAD II2, pp. 34-40; 209).

– Giorgio Buccellati, 1994

Collection of data on phonology and morphology, with interesting remarks of a statistical nature (e.g., the lexicon includes less than 150 words); recurrent peculiarities of orthography indicate that there were family traditions in the writing habits of the scribes.

– Giorgio Buccellati, 1994

A listing of the numerals occurring in syllabic writing.

– Giorgio Buccellati, 1994

A listing of forms of personal pronouns (indep., suffixes, ramānu), demonstrative, determinative, relative, indefinite, distributive, and number (e.g., gabbu “all”) pronouns. A common characteristic trait is the displacement of feminine by masculine forms, or, more broadly, that there is “a general absence of fixed grammatical forms.”

– Giorgio Buccellati, 1994

A list of nouns including adjectives, participles and a select group of infinitives is given in alphabetical order. By way of introduction, there are some remarks on writing and phonology (pp. 2-14), and on morphology (pp. 14-20).

– Giorgio Buccellati, 1994

A listing of data based on over 1000 texts, with emphasis on morphology. It summarizes and brings up to date earlier studies by the author on the same subject. Some of the traits of the dialect are masculine often used for the feminine, loss of mimation, incorrect use of case endings, large use of the t-preterite [perfect], but not of the t-stem (normal preterite is regular in negative statement; in positive statement preterite and perfect are used indifferently, according to scribal choice), occasional confusion of persons and numbers in verbal forms, and interchange of subject and object in syntax but only with the suffix -šunūti (e.g., iddinaššunūti “they gave him”). These characteristics are of Babylonian, not Assyrian origin; the spoken language was East Hurrian, which comes through especially in loanwords (see a long section on this, pp. 51-63).

– Giorgio Buccellati, 1994

A study of consonantal incompatibilities in the structure of verbal roots, with special emphasis on Arabic. The main conclusions are that homorganic (including identical) consonants are excluded from positions I and II of the same root, homorganic (but not identical) consonants are excluded from positions II and III, homorganic (including identical) consonants are not frequent in positions I and III. For his purposes, the author establishes four sections of homorganic consonants: back consonants (laryngals, pharyngals, velars), liquids, front consonants (dentals and sibilants), labials.

– Giorgio Buccellati, 1994

(1) Analysis of the Akkadian writing system shows that the middle radical was geminated in the present; objections to the contrary are unfounded and derived from biased a-priori considerations. Gemination in the present occurs also in Ethiopic (yanäggar), Cananite (Amarna tillakūna), Berber (“habitual” igattal), Bedauye (akantib) and in a sense ancient Egyptian (imperfective partiple mrr vs. perfective participle mr). (2) Another distinguishing trait between present and preterite found, outside of Akk., in Berber, Bedauye, some Cushitic and Chad languages, and perhaps South Arabic, is the contrast a (present) / i (preterite) in the last syllable of the pattern. (3) Negative command is formed by means of the present in Akk., Berber, Bedauye, Cushitic. (4) Prefixed or infixed t is used to form the present in Berber and Bedauye and a related tense for recently completed action (“perfect”) in Akk. — All this indicates that the present is a Proto-Afro-Asiatic feature.

In JAOS 73 (1953), pp. 164-66 W. LESLAU wrote a rejoinder entitled “The Imperfect in South-East Semitic,” to show that the Proto-Ethiopic present (based on South-Ethiopic) was not geminated. Immediately following, on p. 167 f., GREENBERG replied, elaborating his reasons for Proto-Ethiopic *yəqättəl.

– Giorgio Buccellati, 1994

The vowel a occurs frequently as plural marker in all branches of the Afro-asiatic group: the vowel replaces 0% or other vowels in ’ the singular, and it tends to occur between the penultimate and the last consonant of the root, as with the Hebrew segolates *malk/*malak-īm “king/s.” In Akkadian such, a is found in the reduplication plurals alak-tu/alkak-ātu “way/s,” šam-u/šamam-ū “sky/s,” m-ū/ mām-ū “water/s”; such reduplications are common in Cushitic e.g., Somali dab/dabab???b “fire/s.”

– Giorgio Buccellati, 1994

On the basis of data from all branches of the Afro-Asiatic family a set of suffixes n/t/n is isolated, of which n signals the masculine singular, t the feminine sing., and n the plural. In Akkadian traces of this set are found in the numeral “one” ištēn/ištēt (p. 319), and in the plural (masculine) -ān (p. 321).

– Giorgio Buccellati, 1994

A rejoinder to SPEISER 1947 “šumma,” with the etymological conclusion that the element šum is related to West Semitic conditional particles.

– Giorgio Buccellati, 1994

After a survey of previous literature (pp. 246-59), the position is taken that the form paris, because of its origin as a nominal clause, subsumes both “members of the opposition stative: fiens, or être: devenir” (pp. 260-62). Specific examples (pp. 264-73) are then interpreted to show that paris forms may express “inchoative” action (e.g., ištu wašbāku “since I settled down” ) “transitive” action (e.g., šeʾam maḫrāku “I have received grain”) and movement or change of status (e.g., šumma waṣīat “if she has gone out”). In addition, since paris is also a verbal form, it subsumes the opposition accompli: inaccompli as well, which is expressed otherwise by the pair iprus : iparras. With such function, paris results identical to West-Semitic qtl. The secord part of the study adduces examples to show (l) that iprus: iparras are to be differentiated only according to aspect, not tense (hence Haldar denies the existence of a perfect with t, p. 46), and (2) that the correlation accompli: inaccompli obtains also for the iprus: iparras of verbs of condition without necessarily being ingressive, e.g., ul iballuṭ “he shall not remain alive,” p . 36. See for the whole study the sharp criticism by von Soden 1964 (“Zu Haldar”).

– Giorgio Buccellati, 1994

Contains a bibliographical review (pp. 268-71) of the writings about the assumed special relationship between Akkadian and South-Arabic, concluding that the two did not constitute a single linguistic area.

– Giorgio Buccellati, 1994

Though this article does not deal with Akkadian, it is quoted here as one of the few examples of structural description of an ancient Semitic language. The author chooses the language of Jerusalem at about 600 B.C., and follows “a formal method, which asks only what forms exist and in what combinations.” After a statement of the sources and of the major historical/phonological changes before and after 600 B.C., there is a compact description of the phonemes and their distribution, the morphemes and their distribution, and syntax (“successive inclusive constructions”).

– Giorgio Buccellati, 1994

The present iqátal is the oldest Semitic verbal form, not a later derivative from the preterite, and it is preserved only in Akkadian and Ethiopic (the documentation is exclusively taken from Ethiopic). The perfect of the other Semitic languages is derived from the Akkadian permansive, because the latter is more flexible in formation and vocalization than the former.

– Giorgio Buccellati, 1994

A list of Akkadian words, with Hebrew cognates, exhibiting the vowels a, i, u, and ā, ī, ū.

– Giorgio Buccellati, 1994

Against suggestions to the contrary, Haupt maintains that e has phonemic status, on the following grounds: (l) Existence of minimal pairs contrasting i and e, e.g., ri-e-mu “grace” vs. ri-i-mu “wild ox.” (2) Loanwords from Akkadian into other languages, e.g., be-e-lu “lord” as bēl in Hebrew down to bēlus in Latin; 13 such loanwords are listed. (3) Comparative considerations: in other Semitic languages a changes to e under influence of an i and as the result of the “quiescing” of a “guttural.” In the final summary, the rules about Akk. e are clearly stated: a in conjunction with the quiescing of ʾ3-5 becomes regularly e, sometimes also with ʾ1-2; exceptions to the first part of the rule are alību, annu, agalu, adi, atūdu, agrabu and āribu. The extension of the rule to vowels not in contact with the laryngeal is stated by saying that a becomes e under the influence of a preceding or following e.

– Giorgio Buccellati, 1994

“Assyrian” [i.e., SB, NB, LB, NA] does not preserve anywhere Proto-Semitic w. What is sometime understood as w is in fact a labial [i.e., labio-dental] spirant v, and corresponds to original m, postvocalic b or intersonantic aleph, but not to original w. Instead, original w becomes aleph when initial or intersonantic, while the diphthong aw becomes ô , written ü. Most of the documentation, based on comparative considerations with other Semitic languages and on alternate forms of writing in cuneiform, concerns: m>v disappearance of m, and ʾ > m if intersonantic (pp. 264-77). The author concludes (277-85) by refuting the analysis given by others (esp. Zimmern) of possible examples in Assyrian of Proto-Semitic w.

– Giorgio Buccellati, 1994

Brief survey of the corpus of texts then available. Linguistic features common to Assyrian and Ethiopic, e.g., absence of an article, present tense, ending -āni for the plural, ending ‑ku for the first person sing of the permansive, vocalism i in the last syllable of the pret. D, emphatic and copulative ‑ma, lexical items; as a result, Ethiopic is considered closest to Assyrian of all Semitic languages. Special peculiarities of Assyrian in distinction from other Sem. languages include the loss of w, y, h, ś, the different uses of the tenses of the verb, presence of a sibilant in the causative and the pronom. suffixes of third person, the tn stem, the realization napras with a labial in the root, and lexical items. The article ends with some remarks on phonology and on the continuity of the literary language.

– Giorgio Buccellati, 1994

Special emphasis on the need of studying articulatory phonetics.

– Giorgio Buccellati, 1994

Criticizes points of detail in Barth 1887 “Nominalpräfix na” showing especially that nouns like nēberu are also of pattern mapras, not mapris, and also adding other examples of the assimilation ma- > na-. The article ends with the criticism of another article by Barth concerning final weak verbs.

– Giorgio Buccellati, 1994

A continuation of Haupt 1890 “Nominalpräfix na” with more observations of detail concerning nouns with prefixed m- or n-, including also a more complete list of such nouns.

– Giorgio Buccellati, 1994

A coincidental (KOSCHMIEDER), or perhaps better a performative (BENVENISTE) verb is one which serves to express identity of speaking and action, as in uškaʾʾin ana ṭuppim ša ālim “I bow to the tablet(-ordinance) of the city”: here action and statement of the action are simultaneous (and performed by the same subject). In Akkadian (as in Hebrew and Arabic, p. 152) the past tense is used, i.e., the preterite, as in the example just quoted; in such cases then a correct translation must be in the present. The perfect may at first seem to serve the same purpose, but actually that is not the case: thus, in the letters, aštapram refers to an action which is completed with the sending of the letter.

– Giorgio Buccellati, 1994

Even though it does not deal with Akkadian per se, it is quoted here because it contains (pp. 180-188) an instructive review of theories concerning the gender incongruence between numerals and corresponding head nouns, and a new formulation of the theory of polarity to explain the same phenomenon (pp. 188-93, 196). Polarity of genders is assumed to have been a principle governing the entire system of plural formation of Proto-Semitic (i.e., feminine markers were used in the plural of nouns which were masculine in the singular); the incongruence of numerals is but a survival of this system: a singular masculine noun became feminine in the plural, and the numeral agreed with it by being also in the feminine; at a later stage, the plural of the noun was changed into a masculine, but not so for the numerals which remained in the feminine. “This meant that instead of a positive agreement with the co-occurring plural (the gender of which underwent changes), the original corollary, [i.e.,] the co-existent but originally irrelevant negative agreement [=discord] with the unterlying singular noun, became the only workable device”(p. 196).

– Giorgio Buccellati, 1994

A strong case is made for a Proto-Semitic opposition between *yáqtul as prefix-preterite (= Akkadian preterite iprus) and *yaqtúl for the jussice (= Akkadian desiderative l-iprus), based solely on stress. In general, this theory accounts well for the radical difference in meaning between various forms of prefix-perfect on the one hand and jussive on the other. In Akkadian, it accounts well for the phonological difference between precative *lu-iprús liprus and asseverative lū íprus, which remains unchanged because “the/initial stress of the verb created a juncture that prevented contraction” (p. 5). Rightly, the author does not claim that this stress opposition was preserved in historical Akkadian, in which the pair may in fact have become íprus and líprus. In the conclusion a new historical development of the Semitic languages is presented; the system at the origin of this development is considered as having three verbal forms: “an imperfect yVqat(t)Vl(u) [corresponding to Akkadian present], a perfect *y’VqtVl [= Akk. preterite] and a jussive *yVqt’Vl [= Akk. precative]” (p. 21).

– Giorgio Buccellati, 1994

The first comprehensive account of the verbal system, with a section (1:389-93) on writing. The preterite B (“Aorist of Qal”) is described at some length as being the most common form in the available texts: two paradigms give the inflection for person and for the ventive mood (the latter is called “augment of locomotion” with verbs of motion, and “augment of regimen” with transitives followed by a pronominal suffix or a noun as the object); in addition it is stated that the verb can take two enclitics, a subjunctive in the form -u after consonant and -ni after vowel, and a copulative in the form -va [i.e., -ma], which is adversative when the first sentence is positive and the second negative; he also notes that the conjunction between nouns is u as a separate word (1:383-88). The “Aorist” is used for past action (2:141-43). The main forms from weak roots are noted on p. 149 f. Other tenses are the preterperfect in -u [actually forms of the subjunctive in special environments], the present indicated as apakal [pkl are used to denote radicals] and compared with the Ethiopic “contingent,” and the future derived from the present by adding an -u [also subjunctive] (2:150-57). Imperative, precative and optative are stated correctly on pp. 157-60. The “continuative” [i.e., the permansive] is given as the last tense, also with the indication that it can be formed from substantives. Forms with infixed ta and tan are considered as temporal in meaning, identical to the aorist to denote past action, and called “second aorist” (3:152-57); similarly, a permansive Bt pitluhāk is called “second continuative (p. 157). The N, D and Š conjugations are treated in the rest of the paper, with an analysis of the forms, from strong and from weak roots, which is substantially correct. Quadriradicals are briefly but correctly identified on p. 3:159, and even dialectal differentiations between Babylonian and Assyrian are observed (3:164). — [This series of articles, which remains apparently incomplete, is a superb achievement, exhibiting as it does a secure control of the data and a very mature judgement in the linguistic analysis — a fact which is all the more remarkable in that this was the first serious monograph of Akkadian grammar to be written.]

– Giorgio Buccellati, 1994

A “dogmatic” and “illustrative” grammar is not yet possible because of the strong disagreement on basic methodological presuppositions especially with respect to Oppert and Ménant; hence only two chapters are presented here with ample and justificatory evidence of the author’s views, and with criticisms of other scholars’ views. The first chapter is on the permansive, for which the following definition (p. 485) is basic: “the verbal forms belonging to each conjugation may be divided into two great classes, which I call permansive and mutative. The former denotes continuance in the state which the verb signifies in that conjugation; the latter denotes change into that state.” The permansive is not to be confused with a participle, but is rather a tense (p. 486), at least in the sense that it is “a word which includes in itself a pronominal subject, a copula and a predicate” (p. 491); it is used to describe permanent features, and for the passive (p. 488 f.). Several examples are then quoted and analyzed as permansive. - The second chapter is on the noun, and touches on internal inflection (“forms” of the noun) but deals especially with external inflection: it may be noted that “first, second and third case” are preferred to nominative, genitive, accusative because the latter are considered misleading. - The article is concluded by an appendix on the pronominal suffixes. [This is the last article written by HINCKS and like the other important article of 1855-56 is worthy of being considered one of the few classics of Akkadian grammar with lasting value. For an appreciation of HINCKS, and a preliminary bibliography of his writings see C. ADLER, JAOS 13 (1889), pp. ccxcvi-ccci.]

– Giorgio Buccellati, 1994

Presupposing the internal unity of the main Akkadian legal work (esp. Eshnunna and Hammurapi), studies with some statistical computation the function of present, preterite and especially the perfect in the šumma-clauses. The present does not express potential or volitive condition except perhaps occasionally; it serves simply to express a durative-present or an extratemporal condition (alternating sometimes with the stative on the basis of lexical criteria), it has, in other words, no special meaning simply because it appears in a šumma-clause. Similar considerations apply to the preterite: this also has no special meaning in a šumma-clause, but refers simply to the completed past (this particularly against conceptions attributing a present value to the pret. in the codes). Temporal succession, i.e., relative posteriority of one action with respect to another, is often expressed by a sequence of two preterites. “Perfect” forms may in fact be explained in many cases as t-stem preterites with passive value (and inverted subject), which may well have been the primary function of the t-form (p. 130); this then progressed slowly to an active value with various functions: instrumental, causal, adversative (p. 126 f.), sometimes also temporal (p. 129), but without any specialization of the form to express posteriority in the past (as normally assumed); at most, from a temporal-aspectual point of view, the perfect may be conceived as a “punctual present” (Landsberger), mitigating the transition from preterite of protasis to present of apodosis: “If an ox has gored a man and will have killed him…” (p. 130).

– Giorgio Buccellati, 1994

The short section which describes the grammatical characteristics of the names for parts of the body (pp. XV-XCIIl) notes that the primary names do not exhibit prefixed patterns, but show a preference for afformative -an-, for long third radical and for reduplicated base. In terms of external inflection, it is noted that many names are feminine singular, and, if occurring in pairs, feminine dual (suggesting thus in general a relationship between feminine singular, or nomen unitatis, and the dual).

– Giorgio Buccellati, 1994

The study is reproduced, with comments, in HOMMEL 1892 “Stellung.”

– Giorgio Buccellati, 1994

Akkadian was the first Semitic language to become separated from the others of the family; this is shown by lexical and grammatical observations; among the latter the most important is the different treatment of the perfect (p. 95 f., 108 f., 115 f.).

– Giorgio Buccellati, 1994

Compares the perfect and imperfect of the verb [i.e., stative and preterite], personal pronouns and prepositions. For the pronouns he distinguishes in Babylonian and in Old Egyptian two sets, one for the “nominative,” the other for “accusative-genitive” (esp. p. 352). The conclusion is that ancient Egyptian is actually a Semitic language (p. 355), the closest ties being with Babylonian.

– Giorgio Buccellati, 1994

Reduplication of last radical serves for diminutives in Ancient Egyptian and Akk., e.g., suqāq-iy-u>suqāqū “small street” from sūqu “street.”

– Giorgio Buccellati, 1994

An interesting defense of the original biconsonantism of the Semitic root, with an extensive review of earlier literature. The author maintains that “triliteral” roots were formed by the addition of a limited set of stem formatives (variables) to (constant) root determinatives.

– Giorgio Buccellati, 1994

– Giorgio Buccellati, 1994

– Giorgio Buccellati, 1994

Aleph sign used, in late texts, for m in final position (mimation), and for w in intervocalic position.

– Giorgio Buccellati, 1994

Reviews the supposed exceptions to the law stated by BARTH and concludes that they are not real exceptions. The most important clarifications to Barth’s rule are that m does not change to n if the vowel immediately following is u (p. 196 f.), and that loanwords, e.g., melammu, are not subject to the rule (p. 198). Certain forms with prefix na- and no labial in the root (e.g., nannaru, naṣṣaru) are due to the influence of dentals (p. 201). The nouns ikkibu, ikribu, immiru, ipṭiru and iptinu are originally of a pattern mipris, changed to nipris because of the root labial, and finally to ipris for a law similar to that of the verbs primae Nun (p. 202 f.). Other presumed exceptions are explained on a lexical basis, e.g., mamītu comes from a root mamū and mamlu from a root *mml (p. 192-5).

– Giorgio Buccellati, 1994

Reading of A-A as ay or ya; change a > e/i with liquids; preservation of intervocalic y; against spirantization of stops; exceptions of the type išaru to the rule CVCV/CV(C); nominal patterns with prefixed ʾ derived phonologically from prefixed m-; existence of dual.

– Giorgio Buccellati, 1994

Interprets mudū as a participle B from a root mdʾ, derived from wdʾ and ultimately identical to idū. NB forms like mu-da-a-am are to be derived from the same participle, not from a substantive *mudāʾu.

– Giorgio Buccellati, 1994

Examines Middle Babylonian phonetic writing as found among the varied dialects existing in northern Assyria and Western Mesopotamia during the l6th to 11th centuries BC. The author determines first the provenience of the texts under study, and then proceeds to analyze variations in writing vowels and consonants within this corpus. He lists the particular phonetic features which are common to each specific group. In addition, his findings have allowed him to postulate two important principles of phonetic interaction taking place in the evolution of Akkadian dialects.

– jLH, 1974

With statistical computations based on Assyrian royal inscriptions comes to the following conclusions. The last vowel of a construct is not used capriciously, but according to certain rules: i is used with roots third weak, to resolve final cluster of long consonant, for the sake of euphony when the preceding vowel is also i, and when the noun in the construct is in the genitive case; u is used with roots third weak, for the sake of euphony after u, and when the construct is in the nominative; a and mimation practically never occur; dropping of final vowel is regular after short consonant, even when the construct noun is in the genitive. [The continuation of the article, promised at the end, never appeared].

– Giorgio Buccellati, 1994

A general statement of principles, urging Semitists to abandon empiric procedures and to follow scientific methods as those used by Indo-Europeanists. Concrete examples are from Semitics in general, without specific reference to Akkadian.

– Giorgio Buccellati, 1994

In this extended review, various points are taken up and discussed in detail. Interesting observations will be found especially concerning the ending -iš (not to be derived from Sumerian -eše, p. 259 f.), and the present iparras (considered Proto-Semitic and connects with Ethiopic yəqattəl, pp. 262-68).

– Giorgio Buccellati, 1994

On the basis of forms like utlellûm, utelelli, muštabbabbum, muktaššaššum (in Kienast’s transcription) one may reconstruct a stem characterized by reduplication of the middle radical. The stem is closely related to the D stem in form and meaning (“numerische Extensität,” p. 50), and to similar stems in other Semitic languages.

– Giorgio Buccellati, 1994

The Proto-Semitic independent personal pronoun of 2nd pers. is *kā / *tī (masc./fem.), of which the first one became generalized, in Semitic, for suffix, and the second for independent position; consonantism with k for the independent pronoun is found, on the other hand, in “Hamito-Semitic.” In Akk. the possessive pronouns are also derived from the independent, not the suffix form of the personal pronoun.

– Giorgio Buccellati, 1994

While Akkadian and South Semitic have u as first vowel of the pattern in D and Š, NW Semitic shows a. The vowel u cannot be explained as Proto-Semitic since u in the verb is otherwise connected with the idea of intransitivity, whereas D and Š are transitive. Instead, it is derived as an innovation from a in the following manner: *ya-šu-qatil (šu being the personal pronoun as suggested by SPEISER 1936 “Formatives”; the pronoun is hu in some languages, but this does not affect the vocalism) > *yu-šu-qatil (regressive assimilation) > *yu-ša-qatil (analogy of stems with infixes ‑ta‑, ‑na‑).

– Giorgio Buccellati, 1994

The “subjunctive” is a secondary formation, derived by extending the shorter form of the indicative. A clause with the verb in the subjunctive should not be equated to a noun in the genitive. Rather, the origin of the subjunctive ending may be a postpositive demonstrative element.

– Giorgio Buccellati, 1994

On the basis of 15 texts shows that mā occurs not only in Mari (von Soden) but also in OB, OA and literary Bab. It is found only in direct speech (or in letters inasmuch as they imitate direct-speech), and it serves to introduce a rethorical or ironical question, with a meaning such as “What?!”, “How?!”, “How come?!”.

– Giorgio Buccellati, 1994

Forms from danānum, ḫalālum, šanāʾu(?), and zaqāru, are analyzed as bleonging to a stem with reduplication of middle radical following the scheme: | | pres. | pret. | iptv. | | R | *uparrarras | upararris | *purarris | | Rt | uptararras | uptararris | putararris | | | part. | inf. | stative | | R | *mupararrisum | *purarrusum | *purarrus | | Rt | muptararrisum | putararrusum | *putararrus |

– Giorgio Buccellati, 1994

An attempt to reconstruct the inflectional system of biconsonantal verbal roots as a whole, rather than explaining single forms in isolation. Three main types are distinguished. (l) PIS including transitives I w (bil), fientive I n (din) and izuzzum (ziz): | B | ? | ? | ? | *bil | *yabil | *yatabal | *yabbal | | D | &#42ebbulum | ? | *mubbilum | *ebbil | *yabbil | ? | *yabbal | | Š | *šūbulum | *šūbulum | *mušūbilum | *šūbil | *yašūbil | ? | *yašūbbal | | Bt | *tabulum | ? | *mutabilum | *tabal | — | — | *yatabbal |

Note that doubling of the first radical is typical of this class, and that the original form of the Š pret./pres. preserves the original causative particle as assumed by Speiser. Diversions from this scheme in the historical paradigms are based on analogical explanations, of which the following are interesting: the first vowel of B ubbal and ubil is on analogy of Š, since both transitives I w and the Š stem describe involuntary action; the prothetic vowel found in itbulum, itabbal, idin derive from the imperative D where a prothetic vowel was necessarv to preserve the doubling of the first radical typical of D. (2) PSI, closely analogous to the triconsonantal patterns. (3) PŪS with retention of vocalic length whenever possible, and lengthening of second consonant when required.

– Giorgio Buccellati, 1994

Reflections based on Aro’s data collection, with stress on the importance of diachrony. Main results: I (§§ 9-19): vocalic alternation = transitive, isovocalism = intransitive, or more precisely: Transitive-fientive verbs were originally with vocalic alternation, either a/u (occasionally changed to u/u) or a/i (occasionally changed to i/i); isovocalism i/i was originally for verbs of condition, u/u for intransitive fientive expressing non momentary action [thus also von Soden 1952 (Grundriss) § 87 a/c]. II (§§ 20-29): the same rule is also observed in the derived stems, whereby D and Š, transitives, are characterised by vocalic alternation a/i, whereas Bt and N intransitives, are characterized by isovocalism (though N was partly influenced analogically by D and Š). III (§§ 24): isovocalism a/a for B probably served originally to express a reflexive function, just as in Bt.

– Giorgio Buccellati, 1994

Omission of final vowels, mostly in SB texts, is explained as the result of a desire to avoid assonance (i.e., to obtain “desonance”). This occurs esp. with case endings of normal state (e.g., ana amāt šāti instead of ana amāti šāti) and with pronominal suffixes (e.g., panukki šēdu arkatuk Lamassu “before thee is a šēdu, behind thee a Lamassu,” instead of arkatukki). To the same phenomenon are also connected cases of variation when the same word is repeated, e.g., adi tašammam .... adi tašammu .... “until you buy,” first in the ventive, then in the subjunctive.

– Giorgio Buccellati, 1994

After summarizing RÖSSLER 1950 “Verbalbau,” criticizes in detail some of the latter’s conclusions. The lengthening of middle radical in the imperfect of some Ethiopic languages is secondary, and cannot serve as evidence for a long middle radical of the Akkadian present, nor for the proto-Semitic character of this form. The same negative conclusions are drawn with respect to a comparison of the Akkadian present with Mehri, Berber and Bedauye forms, and of the Akkadian permansive with Berber, Bedauye and Egyptian forms. The permansive and present of Akkadian are then explained as being denominative in their origin, the permansive coming from a nominal sentence with contraction of predicative noun and pronominal subject, the present from the verbalization of nouns of agent or the like (pp. 248-50). [It should be noted that no mention of the Akkadian perfect is made as part of the Akkadian verbal system.] The last section of the article examines the features of the verbal system which should be considered as proto-Hamito — Semitic: imperative, prefix conjugation, derived stems, reduplication, u vocalism for the passive. The suffix-conjugation, including the Akkadian permansive, is late, and so are the Akkadian present and the notion of aspect.

– Giorgio Buccellati, 1994

In a sequence of two syllables both beginning with q, the first q may occur as k, e.g., qaqqadum : kaqqadum “head”; more rarely both consonants appear as k, kakkadum. The forms with dissimilation are in free variation with undissimilated forms. Most of the documentation is from OB and NA. The second part of the article describes the nature of “emphatic” articulation in Semitic languages, and suggests that Akkadian (like Ethiopic) emphatics may have been characterized by glottalization rather than back-pharyngealization (Arabic).

– Giorgio Buccellati, 1994

Remarks on plene writing, morphophonemic treatment of š+š, and middle weak roots.

– Giorgio Buccellati, 1994

Each of the two velar stops / g k / may be realized as fricative / ḫ /, and alternately a fricative may be realized as stop according to one of two laws: (1) Dissimilation of velars: if the word base contains two identical non-emphatic velar consonants, one of them may occur either as stop or as fricative, e.g., ḫušaḫḫum = kušaḫḫum “famine” (ḫ) k: 2 examples, k > ḫ : 1, g ) ḫ: l). (2) Assimilation to vowel: velar after vowel may occur either as stop or as fricative, even with intervening word boundary, e.g., aḫum > akum “brother,”šakānu > šaḫānu “to place” ()h > k: 14 examples, ḫ > g : 1, k / ḫ: 10, g > ḫ: 1). Note that the alternation k ~ ḫ ~ k is by far the most frequent with a total of 27 out of 31 word examples. Spirantization does not occur with long consonants. The phenomenon is one of free variation, and is attested accidentally through occasional misspellings, since it had not been accepted in normal orthography. Attestation is especially frequent in OB and SB.

– Giorgio Buccellati, 1994

On p. 722 f. objects to BAUER 1911 “Tempora” on the grounds of the internal diversification of the Akkadian temporal system.

– Giorgio Buccellati, 1994

Akkadian ašar is used almost like a relative, though it retains, throughout the documentation, its original local meaning (pp. 298-300). In Hebrew the origin of the relative pronoun is the same, but with loss of the local meaning.

– Giorgio Buccellati, 1994

The suggestion made in his earlier article of the same year, concerning the prepositional use of the relative pronoun, does not seem to hold true.

– Giorgio Buccellati, 1994

A serious study, with good documentation–—and a good example of how a thorough descriptive systematization, however old, may still retain full interest and validity, even when the historical reconstruction on which it rests has proven almost entirely untenable. The first part is devoted to the pronoun ša: the author claims, through etymology and through a description of the usage, that it is demonstrative in origin; then clarifies its function to indicate the owner or the like [=determinative], and as a nota genitivi; finally suggests that in some cases it behaves like a real preposition. The second part investigates clauses governed by ša, which retains even in this environment a demonstrative character. First, he studies the component elements of the relative clause: resumptive pronoun (with precise rules as to its occurrence or non-occurrence), the subjunctive in -u (here defined clearly for the first time, and distinguished from the forms in -a, which are considered euphonic or emphatic, esp. p. 412 f.), presence or absence of a head noun from which the clause depends, deletion of the relative pronoun. The author then goes on to study the different types of clauses from a notional point of view, i.e., subjective, objective, temporal, final, comparative, causal sentences; here he considers not only relatives in the strict sense, but also subordinates governed by conjunctions.

– Giorgio Buccellati, 1994

General linguistic theory postulates, in a diachronic framework, a “renouvellement” of the perfective aspect by means of verbs of state, and the consequent use of perfective forms to express anteriority; similarly, a “renouvellement” of the imperfective by means of durative/iterative formations, and the consequent use of imperfective forms to express present action. In Semitic, the original perfective form was yaqtul (evidenced by the Akk. preterite ikšud and by residues of perfective value in the W Semitic jussive); in W Semitic, qatala is derived secondarily from the conjugation of verbs of state which is still alive in the Akk. permansive. The original imperfective, on the other hand, was *yaqatal (= ikašad) next to which there was a verbal noun yaqtulu/i/a ; the latter was absorbed in the verbal system, the original case ending -u/i/a becoming specialized as a modal ending; in WSem. (except perhaps for South Sem.) yaqtul(u) replaced completely *yaqatal. In other words, the original verbal system included only two forms, yaqtul for the perfective and *yaqatal for the imperfective; the renewal of the aspect system caused major changes only in WSem., where two nominal forms replaced the original ones, namely qatala in place of yaqtul, and yaqtulu in place of *yaqatal.

– Giorgio Buccellati, 1994

The original, Proto-Semitic function of mimation/nunation was to serve as a marker of determination for the noun. At this stage, the marker was reserved for (1) definite substantives, i.e., determined and individualized (e.g., “the man”) whereas lack of the same marker served for both (2) generic, i.e., non-determined and non individualized substantives (“man”) and (3) indefinite, i.e., non-determined but individualized substantives (“a man”). This system was altered by extension of nunation to serve for (2) as well as for (1): this alteration can easily be attested in terms of general linguistics, where substantives of semantic classes without opposition between individual and species, i.e., substantives naturally generic like “sky,” become determined as in English “the sky.” To resolve the ambiguity between definite and generic, both of which had come to be expressed by nunation, Arabic introduced a new marker, originally demonstrative, i.e., the definite article ʾal- (cf. French “le loup” generic and “ce loup” definite); as a result nunation was extended to cover both generic and indefinite. In this way a complete shift had occurred in the use of nunation from a marker of determination to marker of indetermination. In Akk. (see esp. p. 323 and 328) mimation reflects the second stage where it had been extended beyond determination, but was still retained for determination. Hence the contrast with Arabic is apparent, and can be explained historically.

– Giorgio Buccellati, 1994

A brief survey of historical and dialectal changes.

– Giorgio Buccellati, 1994

Stresses the need of studying the dialect of given regions, rather than that of texts found in one place but of disparate origin. The Mari dialect is essentially OB, more precisely central OB, and Amorite influence is minimal, probably no more than in other OB dialects. The author studies only one feature whidh is typical of Mari, namely “the equation: i + a = ê”; this was a regular contraction, and variant spellings are scribal archaisms. Writings like an-ne-e-em represent an intermediate stage between an-ni-a-am and an-ne-em, indicating contraction, but keeping a spelling with four signs.

– Giorgio Buccellati, 1994

Defends his interpretation of locative -ūm as comparative because: a late text writes kī amīli as a variant for awīlūm; VON SODEN 1969 (“Atramhasis” I) is wrong lexically, grammatically and exegetically; in another text a form in -iš (which can be comparative in meaning) is given as variant for a form in -ūm.

– Giorgio Buccellati, 1994

Defends his interpretation of locative -ūm as having a comparative meaning: variants of the same passage in late but well written texts show -ūm as equivalent of -iš in a comparative sense; it is valid to give a new meaning to a form (otherwise well attested in a different meaning) if the new meaning fits best in a number of contexts; the use of a synthetic comparative in -iš instead of a prepositional phrase with kīma is attested already in OB.

– Giorgio Buccellati, 1994

The ending -m serves as directional marker (“Richtungsexponent”) and more precisely direction toward (the speaker, the addressee or a third party). The ventive is found therefore especially with verbs of movement and, by attraction, with other verbs which are conjoined by means of -ma to ventives. When the pronominal suffix of 1st person in the accusative is attached to the ventive, the ventive may have simply an old locative meaning (“here”) instead of directional or terminative (“toward”). In poetry the ventive is also used for metric reasons without a precise directional meaning. — The basic marker is -m, with -a- added before it to resolve a consonantal cluster when necessary (as with the feminine marker -t which may be preceded by a). The ni of the plural is a connective element which may reflect an earlier, more specialized mood, or a “strenghtening” of the plural (as with the plural of the noun in -ān). — The -m of the pronominal suffixes in the dative is the same as the modal marker of the ventive. Addition of dative pronominal suffixes to the ventive is possible and varies depending on the dialects.— A directional mood is missing in other Semitic languages, but is found in Sumerian.

– Giorgio Buccellati, 1994

Important remarks on method and practice of comparative Semitics. Of particular relevance for Akkadian are the following points: antiquity of the Akkadian present; distinction between fientive and stative as Proto-Semitic; the original aspectual, rather than temporal, value of the “tenses.”

– Giorgio Buccellati, 1994

Understanding a foreign culture implies establishing a relationship with it, on the basis of the human spirit. But in so doing, one should look for the peculiarity, idiosyncrasy or conceptual singularity (“Eigenbegrifflichkeit”) of any given culture, rather than attempting to reduce it to a fixed value system. Starting from such presuppositions, the author recognizes as basic to the Akkadian verb various contrasting pairs, namely: (1) punctual or momentary (i-kašad) vs. durative (kašid & i-kašad): (2) the durative can in turn be either stative (kašid) or fientive (i-kašad): (3) the fientive can in turn be divided into meaning categories, such as voluntary (dūk “to kill”) and involuntary (mūt “to die”). The noun can be recognized as such not on the basis of inflectional categories, but because it describes a concrete object and does not develop into a verbal root (thus in Akkadian one does not say of the snow that it “snows ” p. 363). There are very few such nouns, and they cannot be grouped into meaning categories. This is in contrast with the trend typical of Sumerian spirit and language, to classify objects and concepts. The article ends with a comparison of Sumerian and Akkadian “spirit” in religion, science and poetry.

– Giorgio Buccellati, 1994

The vocalic element is an essential part of the Semitic root. Among the consonantal elements, one observes an incompatibility which excludes the cooccurrence of identical or related consonants in given positions. This exclusion of given phonemes as the result of positional factors creates a greater “expenditure” than would seem allowed by the law of economy. The same emerges from the consideration that roots tend to be differentiated not just by one, but by two elements, e.g., root vowel and final radical in išruk “he donates” vs. išriq “he stole” [these are the type of data emphasized later in general linguistics by information theory.]. In the case of roots with similar meaning and only a slight difference in phonemic structure (e.g., pšr ~ ptr) one may assume that meaning nuances were expressed by phonological nuances raised to the status of phoneme [it is here perhaps that the term and notion of phoneme enters Akkadian grammar]; this process explains the richness of the Semitic consonantal inventory. In morphology, there is a basic difference between verbal and nominal root; the latter, historically earlier than the verbs, is characterized e.g., by a continuans (normally l, m, n, r) in second or third position of pars, pirs, purs.

– Giorgio Buccellati, 1994

The pronoun ša is not a frozen accusative, but the feminine of a demonstrative pronoun, without case inflection, the masculine being šī. The form šat is not plural, but singular emphatic, derived with the addition of the emphatic element t. The same element, added to the personal pronoun šū, gives the form šūt, which is equally singular emphatic. Documentation is then offered (pp. 275-81) to show that ša, šat, šū, šūt are all used equivalently in the following functions: as demonstratives, as determinatives (with and without an antecedent) and as relatives

– Giorgio Buccellati, 1994

Against Christian, the author maintains that Akkadian stative and West Semitic perfect cannot derive from one and the same form, for phonological reasons.

– Giorgio Buccellati, 1994

Rejects the opinion that Akkadian and Ethiopic formed a dialectal unity within Proto-Semitic, on the following grounds: the ending -a in the numerals for tens is from an original dual in Akk., but not in Eth. where the original dual appears as -e. The ending -ku of the Eth. perfect is not the same as -ku of the Akk. permansive because of the different function of the two forms. The ending -tu of the Eth. pron. wəʾətu is an original copula, and thus probably different from Akk. šuātu. The ŠD stem of Akk. is different from an analogous form in Eth. because the latter is derived from verbs which lexically can occur only in the D stem. The tan infix of Akk. cannot be compared with the tan prefix of Eth. because in Eth. a root cannot occur with and without the prefix. The Eth. imperfect indicative yaqatal is derived secondarily from an original *yaqtul-u (which became identical with the jussive after drop of -u), and thus cannot be compared with the Akk. present B.

– Giorgio Buccellati, 1994

In the conclusion to the book (pp. 161-78), the editor stresses the importance as well as the limitation of the structural method and of linguistic geography (pp. 161-4 and, with specific reference to Akkadian p. 174 f.), and the importance of Akkadian for comparative purposes (esp. p. 174).

– Giorgio Buccellati, 1994

“1. A propos of Akkadian i̯âʾum, i̯attun and i̯âʾuttun” (pp. 361-66). Analyzes ia-um as yāʾum i.e., with long vowel and aleph, not contracted in OB to yūm; the feminine sing. yattun is due to quantitative metathesis for yatun, while the fern. plur. is yātum. The form yāʾuttun is a sing. fem, with the afformative -ūt used to express the feminine instead of -at.— “2. On some Old Babylonian Names in -i̯a, -tum, -i̯atum and i̯aûtum” (pp. 366-80). The name of the type Ahiya is not a hypochoristicon, meaning “little brother,”; but a pseudo-vocative with pronominal suffix meaning “my brother.” The affirmative -at is diminutive, hence Abatum “little father,” Abīyatum “my little father”; -ut may occur as a variant of -at.— “3. The Old Assyrian Independent Possessive Pronouns” (pp. 381-410). On the basis of a detailed documentation, the function of the indep. possessive, predicative and substantival use of the independent pronouns, the “pleonastic” use of indep. plus suffix pronoun is also studied.— “4. Paronomastic Infinitives in Classic Akkadian” (pp. 410-15). Establishes the equation: accusative = ana + genitive = locative in -ūm. Hence the use of the paron. infinitive in -ūm is identical to that in -am.

– Giorgio Buccellati, 1994

The multiplicative prepositional phrase of the type ana šinīšu, in the abbreviated form a šinīšu [aššinīšu], normally meaning “twofold,” occurs some times as an attributive phrase meaning “double,” e.g., aššinīšu šēnān “two double sole (shoes).”

– Giorgio Buccellati, 1994

Certain forms of “izuzzu” may be explained as derived from a weak triradical primae y, of the type išērum. The alternation of y and n for the first radical is comparable to Hebrew yṣb/ nṣb and to Canaanite ytn/ other Sem. ntn (or ndn). The infinitive izuzzu itself, may be analyzed as the infinitive Bt from yzz if understood as *izzuzum (i.e., *itzuzum: pitrus, pp. 353; 359); the first vowel in the alternate form uzuzzum is due to assimilation to the second vowel. Similarly the unusual OAkk and. MA precative lizzaz is to be explained as *l+itzaz (cf. litbal), from which one must assume a secondary root tzz (as for ndn ~ tdn in 0A). Like verbs primae weak, izuzzu is essentially biconsonantal.

– Giorgio Buccellati, 1994

(l) The stative preceded by lā is used for the prohibitive not only in OB, but also in OA; it serves to express “the desire that a particular situation should not be perpetuated” (p. 22) — (3) The adverb assurri (or related forms) serves to reinforce a vetitive or a negated vetitive in OA (assurri ē lā taṣbatā “by all means do not fail to seize!”) and in OB also in front of indicatives (36 f.). - (4 ) OA šušalšūm is a locative adverbial meaning “in triple amount” from a ŠD stem *šušallušum (with syncopation), derived from the numeral šalāš.

– Giorgio Buccellati, 1994

A narrative presentation of the literature, stressing intentionally history and culture rather than language. The section on grammar proper is on pp. 18-21, and contains brief summaries with evaluations.

– Giorgio Buccellati, 1994

Begins with a good and detailed review of earlier work done on the subject going back to Hincks 1855-56 (“On Assyrian Verbs,” Journal of Sacred Literature and Biblical Record 1 (1855), pp. 381-93; 2 (1855-56), pp. 141-62; 3 (1856), pp. 152-71; 392-403). The main section consists of a listing of strong verbs which are attested with certainty in both the present and the preterite B. On this basis L. suggests that the vocalism is not arbitrary, in that a/u is for transitives, u/u for intransitives; a/a and i/i on the other hand, are secondary. The distinction between present and preterite is old, even though both forms may derive from a common archetype through stress differentiation. The distincion in vocalism between perfect and imperfect in WSem. is late, and thus one cannot accept Barth’s principle for noun formation that all patterns are derived either from the perfect or from the imperfect.

– Giorgio Buccellati, 1994

From the examination of 570 text occurrences of personal names in the Akk. texts of Ugarit it appears that names with afformative -ān-, -ēn-, -īn-, -ūn- are inflected diptotically, i.e., with -u for the nominative and -a for genitive-accusative. Even though in Akkadian context, the phenomenon is Ugaritic in nature; its origin is perhaps due to the desire of differentiating the genitive of names in -Vn- from the pronominal suffix of first pers. plur. -ni.

– Giorgio Buccellati, 1994

The Akkadian expression of the type ina kār kār-ma “in every colony” is also found in Ugaritic, e.g., bnš. bnšm “every man” and possibly in Amorite: Ilim-ilim-ma “of every god”; the particle -ma is found, in the same function and position, as it is in Akkadian even outside of the case mentioned, e.g., mamma ( *man-ma) “anyone, ” or immati-ma “for every time, forever. ” It is assumed that the Akkadian phenomenon is of Amorite origin.

– Giorgio Buccellati, 1994

An observation by Albright, that Canaanite Akkadian shows the genitive after umma, does not seem to apply in all cases attested in the Amarna letters. — In an editorial note, A, GOETZE refers to a similar Hittite usage of umma plus a genitive, and notes that umma is interchangeable with āwat “word of”.

– Giorgio Buccellati, 1994

A revision of Cantineau 1951-52 (“Consonantisme”) also in terms of articulatory phonetics. The author objects to the tridimensional schemes of Cantineau, which imply an equidistance of voiceless, voiced and emphatic, and suggests instead a “flat” scheme where the three sets follow one another on the basis of glottal opening: the glottis is closed with voiceless, “intermittent” with voiced, open with emphatics. The general lack of labial emphatics is then explained from an articulatory point of view (p. 69). Emphasis in general is explained as glottalization, and the ʾ as the marker of the whole series of emphatics; h, on the other hand, is the marker of the voiceless series. The interdentals are explained as originally palatals or of a type “dorsal antérieur” (pp. 74 ff.).

– Giorgio Buccellati, 1994

After philological notes on selected passages from PRU III, the author points out some characteristics of orthography, phonology (e.g., t+š preserved), morphology (proleptic possessive pronoun before determinative pron. as in eqelšu ša PN, case endings in the construct state, third fem. sing. of the verb in t-, reenforced prohibitive with the present) and syntax (predicate in sentence initial, verbal predicate in the sing., with subject in the plur., NW Sem. use of conjunction u).

– Giorgio Buccellati, 1994

For the problem of the origin of the Semites the principal data are linguistic, and to this effect it is necessary to reevaluate the antiquity of Akkadian. Some of the oldest traits in the language are the dative, the locative, the present and the expression of types of action.

– Giorgio Buccellati, 1994

The existence of a vocalic element ə is suggested not only for Assyrian, but also for Babylonian: in the second syllable of a noun in the construct state with base ending in cluster, e.g., šigir = šiper (p. 7), and in three-syllabic words with first two syllables short and open, e.g., šikarum = šikərum (pp. 10-12 ).

– Giorgio Buccellati, 1994

Having treated briefly the morphology and the statistical frequency of occurrence of the perfect (i.e., the permansive) in the various classes of texts (pp. 510-14), it deals at greater length with its meaning and function, on the basis of good documentation. It is used indifferently for the active and the passive voice (pp. 514-16); in either case, it serves to express “permanent states or conditions and characteristic qualities,” and as such it is used especially for descriptive purposes in relative clauses: description of landscape, of conditions of natural objects, of habits and customs— hence the use as a permansive, first suggested by HINCKS, is correct but accidental (pp. 516-19). In fact, the form is often used, “actively and passively, to describe single, definite or momentary actions and events represented as either past or present”: this is deduced from the general context, from the cooccurrence with finite tenses, and from the desiderative use with lū (pp. 519-21). Since there is no temporal limitation, it is not a perfect as in WSem.; rather it serves in the same manner as the Hebrew infinitive absolute and is in fact not derived from the participles, but rather the participles “are originally infinitives or rather abstract nouns used as primary predicates and in this way endowed with a regular verbal inflection” (pp. 522-29).

– Giorgio Buccellati, 1994

The code exhibits an unusually large number of asyndeton clauses. Only a few of these are both grammatically and logically coordinate (p. 70); most of the other ones are logically subordinate. The author offers a well differentiated notional typology of the latter group. The majority consists of circumstantial clauses, i.e., such that their action is contemporary with that of the main clause (e.g., “if a seignior presented a field, left a document” = “if, when a seignior presented a field, he left a document” or gerundive “if in presenting a field a seignior left a document”), but one also finds temporal, concessive (p. 67 f.); modal, objective (p. 69 f.), final and consecutive (p. 70 f.) clauses. From the viewpoint of word order, logically subordinate clauses appear normally before the logically main clause in the protasis and after in the apodosis (esp. p. 68).

– Giorgio Buccellati, 1994

Notes the existence of the form in -ātina, until then not attested.

– Giorgio Buccellati, 1994

Follows very clpsely, for the grammatical theory, OPPERT 1860 “Elements” of which it does not even consider, however, the remarks on phonology. On the other hand textual references are numerous, and are quoted both in cuneiform and transliteration; it also includes a section on writing and an anthology of texts. But the main advantage of the work is the abundance of bibliographical references, which trace back every important grammatical notion to its first formulation in the early stages of decipherment.

– Giorgio Buccellati, 1994

The infinitive with ending -i followed by a subject constitutes a complete sentence in the Akkadian of the West, of the type qatāli(-ma) anāku “I am/was (engaged) in killing” i.e., “I kill(ed)”; the enclitic -ma is optional, but frequent. Such use of the infinitive is found elsewhere in West Semitic, and is a Canaanitism in West Akk.; the ending i is possibly related in function to the use of -ūm with the absolute infinitive in Akk.

– Giorgio Buccellati, 1994

In standard Akkadian šumma occurs in the protasis of a conditional sentence, meaning ‘if.’ But in EA there are a number of passages where this cannot be the meaning: a ‘normal’ translation would result in very rare syntactic patterns and would yield little sense or even a sense contrary to the context. Knudsen, guided by context, translated ‘denn’; Moran, however, argues such a tranlation will not fit all cases. “We retain the more original deictic force of the particle, conventionally rendered by ‘lo, behold.’” The evidence does not permit us to determine if this feature reflects Canaanite or Akkadian idiom, although Canaanite influence seems the more probable.

– jLH, 1974

On the basis of the 66 Amarna texts from Byblos, Moran attempts to demonstrate the existence of the West Semitic yaqtula verb form in Canaanite. The difficulty in proving the argument is that this form is identical with the Akkadian ventive (-a suffix). The use of yaqtula is to express a wish, request, or command; and it is also used in clauses of purpose or intended result. The author has found 36 occurances of yaqtula used in the former sense, and 13 in the latter. In West Semitic, if the verb of the first clause in a two clause sentence states a fact and is in the indicative (e.g., yaqtulu), then the second or purpose clause will also be indicative. But if the first clause verb is imperative, jussive, or a yaqtula form expressing a wish, the purpose clause will use yaqtula. This is the rule of model congruence and the author’s reason for denying that the ventive -a is meant in these cases. As additional evidence, the use of the yaqtula verb is pointed out in Biblical Hebrew where it is used in the sense for the cohortative and is goverend by modal congruence.

– Glaeaseman, 1974

After reviewing in detail the pertinent literature from the Middle Ages on, three main qualifications of biconsonantism are given: biconsonantism in logical (i. e. two identical consonants are found in several sets of words with similar meaning), historical (i. e., biconsonantal roots are earlier), and partial (not all roots are biconsonantal). The “determinatives,” i. e., consonants which are added to derive triconsonantals from biconsonantals, occur in initial, middle and especially final position.

– Giorgio Buccellati, 1994

A discussion of individual phonemes over the whole area, followed by a study of the consonantal system of each major language (with tridimensional schemes as used by Cantineau). The system, except for the graphic schematization, is essentially the same as in MOSCATI 1964 (Introduction).

– Giorgio Buccellati, 1994

Comparative Semitic grammar reconstructs a Proto-Semitic active participle, but not a passive one. The author proposes to identify for such a function the forms qatīl/ qatūl, in rhythmic opposition to the active participle. In Akkadian parīs is the substantival counterpart of paris, parūs of parus.

– Giorgio Buccellati, 1994

Based on the previous Italian edition entitled Lezioni di linguistica semitica (Roma 1960), this is the latest general expose of comparative Semitic grammar. More than in any of its predecessors, the book places considerable importance on Akkadian, for which the author had the advice of W. von Soden. The clear and up-to-date presentation of the material will help especially in placing Akkadian within the wider framework of the other Semitic languages. See K.H. Deller, Or. NS 34 (1965), pp. 35-44.

– Giorgio Buccellati, 1994

Proto-Semitic distinguished between š, ś and s. In Akk. š and ś merged and are both represented by signs with Š. In later periods, Assyrian š corresponds to Hebrew s, and vice versa Assyrian s to Hebrew š.

– Giorgio Buccellati, 1994

In answer to a critical review by H. Zimmern (ZDMG 58 (1904), pp. 954-7), develops his interpretation on the use of ‑ma to express virtual subordination, esp. with respect to Gilgamesh.

– Giorgio Buccellati, 1994

Forms with or without mimation are used in OB “à peu près indifféremment” (p. 75), but they preserve traces of an original system which opposed determined to undetermined forms, which may also be paralleled in their opposition by different forms in the plural, according to the following scheme (p. 78):

A possible hypothesis on the origin of mimation is that it is derived from enclitic -me and that mimation is a secondary formation. When mimation ceased to function, šū may have served as a definite and wēdūm as an indefinite marker.

– Giorgio Buccellati, 1994

Detailed examples from all the main Semitic languages, including Akkadian, are used to show that the prefixes ma-/mi-, ša-/šu-, ha-/hi- and ʾa-/ʾi- have to be interpreted as original relative and demonstrative pronouns. Thus Akkadian mīšaru is derived from *mayšar meaning “that which is straight” (p. 182). The same interpretation is applied to the various causative prefixes, of which ten different meaning categories are distinguished (pp. 250-261, with examples from four of them for Akkadian), though all prefixes are assumed to have a common origin (pp. 261-69). Such sentences are conceived as reflecting an early syntactical stage of Proto-Semitic. For the non-prefixed patterns, note especially the contention (p. 188) that Akkadian ikaššad is not derived from Proto-Semitic qatala (Bauer). Pp. 273-288 provide useful word indexes.

– Giorgio Buccellati, 1994

The t-forms in the Code are used especially at the end of meaning units, and normally in the protasis instead of the apodosis: thus they acquire the value of pausal forms and serve to express a special sentence stress (the use of t for emphatic purposes is favored in Akkadian by the abundance of weak roots and the introduction of t to restore triradicalism). Not rarely, t-forms are omitted in contexts identical to those with t-forms: the difference is explained by admitting two different sources of the Code, a more recent one with t-forms (A), and an earlier one without (B); the editor who put together the two sources wrote himself in the style of B (p. 186 f.). Follows an exegetical analysis of the laws divided in groups depending on meaning and source (pp. 187-208); the tabulation on pp. 213-17 assigns 170 laws to A and 85 to B. Historically, B is seen to represent the earlier Sumero-Akkadian tradition, A instead a more recent Semitic tradition to be localized in the South; only the topics on which the sources showed contrasting legislation came to be represented in the Code (pp. 218-20).

– Giorgio Buccellati, 1994

Utilizing the same criteria as in 1933 “T-Formen” distinguishes two sources in the MA laws, one with, the other without t-forms; the argumentation is primarily exegetical rather than linguistic.

– Giorgio Buccellati, 1994

After a short review of the history of the problem and of the source material, noting especially the frequent occurrence of t-forms with certain verbs, the main function of the T stem is analyzed as one of subjective intensification of the verbal concept, as distinct from the absolute, objective intensification of the D stem (Par. III.I). In addition, a perfective function is also noted (III.III), while a passive and reflexive function is considered very rare (III.VI). Final remarks with dialectal and historical considerations (the introduction of t-forms being attributed to Amorites, V.III).

– Giorgio Buccellati, 1994

Next to the pronoun šū there is, in NB, a new formation iššū, iššī, plur. iššunu, iššini, in a demonstrative-anaphoric meaning “this, he/it.” Etymologically, it is derived from šū preceded by a demonstrative element in from enna “here.”

– Giorgio Buccellati, 1994

Quotes NB texts showing the existence of a preposition la identical in meaning to ina and ištu/ultu “from, out of,” and therefore different from W Sem. la, li “to.”

– Giorgio Buccellati, 1994

In NB the elements -ka for the singular and -kunu for the plural are added to demonstratives to express the local aspect “there, that,” e.g., aganna “here” ~ agannaka “there.”

– Giorgio Buccellati, 1994

The first complete grammar, articulated in 9 chapters and 249 paragraphs, with transcriptions in Hebrew characters. Apart from two brief chapters on phonology and even on syntax, the bulk of the article is on morphology. Internal inflection with respect to the nouns is treated in ch. VIII: it distinguishes primitive nouns and nouns derived from a verbal root, and gives the main patterns, simple and with consonantal augment. The noun is treated in ch. II: the notion of “mimmation” is introduced, which is considered typical of the “emphatic state”; the construct state is not treated expressly, though under the name “simple state” it is referred to occasionally (p. 111, 117, 388), but on the whole it is not yet well understood (see e.g., the paradigm on p. 120 f.). The pronoun is treated in chs. Ill - IV, the numerals in ch. V, the invariables in ch. VII (the adverbial value of -iš is already noted). By far the largest is chapter VI on the verb (about forty pages, almost half of the entire grammar). Only one tense is recognized, the “aorist” (i.e., the preterite); the existence of a “perfect” (permansive) is considered but immediately denied; similarly, only two moods, imperative and precative (347 f.; ventive forms are noted in passing, e.g., p. 346 n. 1, 374, and considered paragogic or euphonic in nature) and only two verbal nouns, infinitive and participle. [This reduction of verbal forms will remain typical of the early French tradition, vs. the German and especially the English position; the latter had already been authoritatively voiced by HINCKS 1855-56 “Verbs.”] Patterns from weak roots are treated in some detail, and quadriradicals mentioned briefly.— In general the documentation is quite limited (though it is interesting to note that it already makes reference to Hammurapi, p. 394), whereas paradigms are given very generously.

– Giorgio Buccellati, 1994

The changes with respect to the first edition are listed on pp. XII-XV of the introduction. The main innovations are: the distinction of three types of aorist: simple in -u, apocopated, and paragogic in -anna; a new paragraph on verbs double weak; and some bibliographical references, especially in the introduction (XV-XXIl): the main thrust of these references is against Hincks (see also p. 117) who had “imagined,” next to the preterite, also the present and the permansive.

– Giorgio Buccellati, 1994

First a complete survey is offered of the usage of the determinative pronouns as nota genitivi (NG) in all Semitic languages. Three main environments are distinguished: NG occuring between construct and construent, NG and construent preceding the construct, no construct. The introduction of the term “determinative” is due to Ungnad, and as a result the term has become entrenched only in the grammar of Akkadian (and Ugaritic). The term should be understood in a syntactical rather than semantic function: the determinatives are nonautonomous pronouns, in that they need to be followed by another element, i.e., either a complement of specification (then the pronoun may be more narrowly defined as NG) or a relative phrase (nota relationis). A comprehensive and meaningful subdivision of all pronouns as autonomous and non-autonomous is then given (see especially pp. 70 and 147).

– Giorgio Buccellati, 1994

By way of addition to von Soden 1939, Peters shows that this pattern is operative also in Arabic.

– Giorgio Buccellati, 1994

A bibliographical overview, referring especially, for Akkadian, to the terminative ending -iš and to the incompatibility of the radical consonants.

– Giorgio Buccellati, 1994

A comprehensive study of the system of internal inflection within the broader framework of Hamito-Semitic, with exhaustive bibliographical data (see esp. I, pp. 558-581). The notion of internal inflection is understood in a narrow sense, to exclude nominal formation, and as a result, very little room is given to Akkadian (see esp. I, pp. 585-86; 598-99).

– Giorgio Buccellati, 1994

Assuming incompatibility as a fact, the author wants to show the reasons for its existence. The usual explanation (difficulty in the articulation of homorganic sounds one next to the other) does not hold true, because assimilation and dissimilation could dispose of it. Information theory has called attention to the importance of acoustic (next to articulatory) phonology, emphasizing as a result the importance of distinctivity of utterances. In light of this, incompatibility is considered as “an expression of the tendency not to join together in contact position (1-2, 2-3) phonemes which are only minimally distinctive” (p. 137). The root is defined as “a discrete (interrupted) morpheme of lexical character, formally similar to a Markov process, and construed negatively by excluding elements which lie below the level of minimal distinctivity” (p. 139).

– Giorgio Buccellati, 1994

The sequence of the type PN anāku in royal inscriptions is not a nominal sentence (“I am PN”) but rather the subject with apposition of a longer sentence (“I, PN, ....”). The bulk of the documentation is from NW Semitic; for Akkadian, the case is made only marginally on the basis of stylistic criteria. It is also noted that from a sequence PN anāku, of Sumerian influence, in the earlier texts, one goes to the sequence anāku PN in texts from the 7th - 6th centuries, most likely under Aramaic influence.

– Giorgio Buccellati, 1994

A volume collecting studies on Akkadian grammar.

PDF available here

– Marco De Pietri, 2021

An insightful article, which, taking as a starting point a text of Narām-Sīn, deals with embedding of clauses (“encasing” or “boxing” with the German term) and with the use of coordination in place of subordination. This type of coordination is common expecially to express temporal relationships, less so in an antithetic-concessive sense. Unlike the case in Indo-European languages, in Semitic even coordination for subordination may be embedded, e.g., “When the countries revolted against me, then my father had bestowed benefits on Kish, and Kish revolted against me,” where italics indicate an embedded coordinate with a concessive value (“even though my father…”). Coordination of this type is also used where English prefers an adverbial phrase, e.g., “Kish gathered and…” for “Kish, in a revolutionary gathering…” i.e., Akkadian (and Semitic in general) prefers a more concrete type of expression. The “clause character of a co-ordinated complex” not marked by any segmental features may have been marked in speech by intonation. A complex hierarchy of embedded coordinate clauses is possible, which the author symbolises graphically by a linear sequence of letters [similar in nature to the “tree” of generative grammars].

– Giorgio Buccellati, 1994

The construct state of šarrum is šarri, not šar.

– Giorgio Buccellati, 1994

The following formulas introduce a negative oath even if the negative particle is omitted: (1) PN ina DN itteme kī, (2) Nīš / Ina šum DN ušēli kī ...., (3) Nīš DN itmu ša ...., e.g., nīš Šamaš itma ša āwat tuppim annīm unakkaru “she swore by Shamash that she would not change the terms of this tablet.” A positive oath is distinguished as such by the further deletion of the conjunction kī or ša, e.g., nīš DN itmu.... “by DN he swore (that he would do something)” (p. 19).The documentation is all late, except for the third formula for which OB examples are quoted.

– Giorgio Buccellati, 1994

The combination l + imperative found in three Biblical passages referring to the speech of the “Sons of Heth” is to be explained as the Hurrian imperative infix -el-, -ol-, with a change in sequential order (in the Canaanite rendering the infix is anteposed). A similar feature has not turned up yet in Nuzi Akkadian, but it may be compared with the similar phenomenon of the pronominal suffix expressing the subject (ipallaḫšunūti “they will serve her;” pp. 121-24).

– Giorgio Buccellati, 1994

The change šš > ss is due neither to phonetic dissimilation (because šš is in fact often attested) nor to analogy with changes which can be explained phonetically, e.g., tš > ss. Rather, it is due to the desire of avoiding the ambiguity resulting from the fact that nš also yields, in certain cases, šš ; i.e., šš < nš was differentiated from original šš in that the latter became ss. The rule for the change nš > šš is that the change occurs regularly except when the vowel preceding n is etymologically long (e.g., īn-šu) or when the word before the suffix contains a sibilant (e.g., uzun-šu).

– Giorgio Buccellati, 1994

Anticipating the fuller exposition of 1941 it stresses the identity in function between the determinative and the relative pronoun ša; this is a representative and generalizing pronominal particle (“pronominal” in the precise sense that it can only resume a noun or nouns, not a sentence). There is no difference between the phrase/ clause with and the one without the particle.

– Giorgio Buccellati, 1994

Starting out from the theory about the determinative character of ša, first set forth by Ungnad (to whom the book is dedicated), this work deals with ša outside of relative clauses or introducing them. The documentation is drawn primarily from the Code of Hammurapi and other OB texts (pp. 9-89), but includes also samples from all other major dialects (pp. 65-69, 90-113). —Ša outside of relative clauses or ša of nominal connection (i.e., determinative) is not a pronoun, because it is not inflected (pp. 30 f.), nor is it a preposition (p. 29); it is, instead, a “pronominal particle” (p. 30). When preceded by an explicit antecedent, ša is “representative” in that it resumes the antecedents) (p. 11); when not so preceded, ša is simply and sufficiently determined by the “adjunct” which follows in the genitive, e.g., ša akālim “something to eat” (p. 18). In either case, ša is “generalizing,” though it remains restricted to a limited number of objects (i.e., it is not an indefinite pronoun, p. 32). Ungnad’s term “determinative” is rejected because it can be confused with the term used for semantic indicators in cuneiform, such as DINGIR in front of DN’s (p. 31). —Ša introducing relative clauses, or ša of verbal connection (i.e., relative) is not a pronoun, because not inflected (pp. 33-36); it is simply a particle, which can be deleted without substantial change of meaning. The main difference between constructions with and without such a particle is that in the former the head of the attributive clause is normally a “new item,” whereas in the latter the head has already been presented to the listener (pp. 55-59). Also, in practice, constructions with ša are the only ones to admit resumptive suffixes though in theory constructions without could possibly have the same suffixes (p. 54). -In texts other than the Code the usage remains the same, but for few exceptions, e.g., in OAkk. the rarer use of ša and on the other hand the use of the inflected pronoun. Occasionally, ša is used in such a way as to appear to serve as a preposition or a conjunction (p. 115). -[One of the main conclusions of the author, that ša is not a relative pronoun like English “who, whom, etc.,” is based on a comparison limited solely to surface features: hence, while such conclusion is correct, it is not particularly meaningful, -Cf. UNGNAD 1942 “Ravn,” GOETZE 1947 “Ravn.”]

– Giorgio Buccellati, 1994

In the relative clause of the type Enlil ša qibīt-šu kabtat, the relative pronoun ša is, in effect, the rectum with respect to the regens qibīt, i.e., ša qibīt- means “of whom the command.” Since the word order rectum — regens is not normal in Akkadian, the resumptive personal pronoun -šu replaces the rectum after the regens: “of whom the command of him.” The construction is similar to that of the inverted genitive of the type: ša Enlil qibīt-šu “that of Enlil the command of him,” which exhibits twice the sequence regens-rectum, using a determinative pronoun for the regens in the first case (ša) and a resumptive personal pronoun for the rectum in the second case (šu).

– Giorgio Buccellati, 1994

A brief grammatical sketch (pp. 14-57), based on DELITZSCH 1889 Grammatik. It utilizes only cuneiform characters, and no transcription.

– Giorgio Buccellati, 1994

After asserting the essential similarity of the languages of the Semito-Hamitic family with respect to root structure, stem and tense system, six languages are analyzed individually to bring out the elements most fruitful for comparison. Akkadian is treated on pp. 470-78; the lack of active verbal adjectives of the B stem with second vowel a is archaic (and its introduction in later Semitic languages an innovation); similarly archaic is the connecting vowel a ( > ā) in labšāta (p. 473); the “present” serves actually to express repetition, continuation or habit, hence the nominal pattern parras which expresses professional habit (p. 477). In the discussion of Berber (pp. 478-86) the similarities with Akkadian are emphasized (see esp. p. 486). The study ends with an attempt at historical reconstruction, which places Berber at the oldest stage, and Akkadian in the middle (as the oldest, however, of the Semitic group). The nominal use of the “Suffix-conjugation” (as in Akkadian) is considered older than the verbal use (as with the West Semitic perfect).

– Giorgio Buccellati, 1994

The Akkadian verbal system is unique with respect to other Semitic languages in that (l) the prefix conjunction (type iprus) is used only for action verbs, (2) verbs of condition, never occurring in the prefixed forms, occur in the stative. Peculiar to Akkadian together with Mahri and Ethiopic is the pattern of the type iparras for incomplete action. There are close parallels in Bedauye (Cushite) and especially in Libyan-Berber, as shown by the following correlations (in function as well as form) between Akkadian and Libyan: purus ~ efras , iprus ~ ifres, iparras ~ ifarras, šapris ~ sefres, ušapras ~ isafras. The author concludes that the system presupposed by Akkadian and Libyan is closer to Protosemitic than that of the other Semitic languages.

– Giorgio Buccellati, 1994

Libyan (i.e., ancient Libyan and modern Berber) is in fact a Semitic language, particularly close to Akkadian, for which theory the author adduces phonological, lexical and morphological evidence. Special attention is paid to consonantal equivalences, for which 75 Libyan roots with Semitic correspondences are adduced, about half of these (30) having Akkadian equivalents, e.g., lamādu “to learn” ~ Tuâreq aelmed “apprendre, comprendre…”; nadru “zornig, wötend” ~ Tuâreg ti-ĝorad “force (physique), viguer.” Similar equivalences are quotes for the pronouns (e.g., anāku ~ lib. nak), particles (ana, ina ~ an, in), and various morphemes (e.g., feminine t).

See the remarks by J.M. SOLÀ SOLÉ, M. COHEN and A. Basset in GLECS 6 (1951-54), pp. 33-34; 41-42, who emphasize that RÖSSLER’S position had already been maintained by M. COHEN, while at the same time criticizing RÖSSLER for some aspects of his utilization of the Berber material.

– Giorgio Buccellati, 1994

Only the Proto-Semitic laryngeals ʾ3-4 (ḥ, ʿ) are regularly a cause of the phonological change a > e; the velar ʾ5 (ġ) causes the same change only occasionally, namely in ebū “to be thick”, ešū “to confuse”, eṭū “to be dark” (all double weak verbs, cf. e.g., epū < ʾpy “to bake”), and ṣebū “to color” (cf. ṣēnu < ṣʾn “sheep and goat”). Other roots which show ġ in Arabic appear without vocalic change in Akk. (e.g., adāru “to be dark”). On the other hand, some of the roots with ġ in Arabic and vocalic change in Akk. do not have ġ in other Semitic languages (e.g., erēbu ~ ‘rb in Ugar. and South Arabic “to enter”). The situation reflects a late expansive tendency of ġ especially in Arabic.

– Giorgio Buccellati, 1994

The existence in Biblical Hebrew of forms like ynṣar next to yṣr is to be explained not phonologically, but syntactically: ynṣr is used in dependent clauses, and is related in form to the Akkadian present. [See objections in BLOCH 1963 “iparras.”]

– Giorgio Buccellati, 1994

A thorough monograph, with a comprehensive documentation of 451 examples (mostly OB and SB) given in transliteration and translation. It deals with syntactical and especially with notional categories, according to a conceptual scheme which I would reconstruct as follows:

The common denominator for all these various functions is that of changelessness, i.e., absence of change for roots of condition and lack of (potential) change for roots of action; conversely, it is postulated that prefix forms were reserved to espress fientic change, at least in the prehistory of the language (pp. 288-90; 310-3). Another prominent characteristic of the permansive is that it is protractive in aspect (pp. 249, 294-99), as a result of which the permansive is frequently used as the first predicate in a conjoined sentence used for virtual subordination: here the permansive is circumstantial in nature (pp. 271-8, 295, 300). Because of these usages, the permansive has an element of perfectivity, which at times becomes prominent, and which may have a bearing on the explanation of the NW Semitic perfect (pp. 291-301). As for terminology, “permansive” is better than “stative” because it has a wider range and creates no confusion with the verbs of state (p. 234 f., 299). A section on OA is included parenthetically on pp. 285-88. A recurrent observation is that contextual considerations are more important than in any other Semitic or modern European language to determine the choice between the very numerous functions of the permansive (e.g., pp. 238, 291).

– Giorgio Buccellati, 1994

«One of the most discussed Assyriological topics is the “Sumerian problem”: Were the Sumerians an autochthonous Mesopotamian population or did they come from somewhere else? In order to answer this question one has to take a look at both the textual and the archeological materials we have. The archeological and environmental evidence seem to allow different, and even contradictory, readings and interpretations» (from p. 1).

[This paper discusses the problem of the Sumerian origin, a topic still deeply debated among modern scholars. It also deals with the problem of the origin of Sumerian language, discussing the existence of a possible ‘Pre-Sumerian substratum’.]

PDF available here

– Marco De Pietri, 2021

Differing from KIENAST 1957 “Reduplikation” R. suggests that utlellūm and derived forms be understood as a causative in ʿ from a preposition *la-ʿalay “above”; the result was a denominative verb with five radicals. The causative in ʿ is otherwise lost in Akkadian (though common in other Semitic languages); the prep. lʿly is attested in Ge‘ez laʿala “above,” where there is also the verb ʾalʿala “to raise”; and the phenomenon of verbalisation of a preposition is possibly also present in other cases such as naṣāru < in(a) ṭahr “on the rock” > “to be high,” “to look down from above.”

– Giorgio Buccellati, 1994

An analysis of the notions “aspect” and “type of action” (Aktionsart”) with reference to their historical development in various Semitic languages, including Akkadian. Following in part Landsberger’s “Eigenbegrifflichkeit,” the author maintains that the notion “type of action” is more meaningful in describing the contrast between present and preterite (which then can be stated in terms of rest ~ motion or stative ~ fientive) rather than the notion of aspect (incomplete ~ complete). Of interest is the section (IV) on negation, in which he notes how negation neutralizes the aspect of any given form, since an action which has not or will not take place is not characterized very precisely. An attempt to reconstruct the prehistory of the Akkadian verbal system is presented on pp. 100-106.

– Giorgio Buccellati, 1994

A collection of data, based mostly on Arabic, Syriac and Ethiopic; the Akk. material (indexed on p. 256 f.), only seldom utilized, is concentrated mostly in a few sections, esp. for šD> lD (p. 194 f.) and then for reduplicated roots (pp. 6 f., 24, 37, 61, 74) and for nasalization of long consonant (pp. 110, 147, 183, 201, 203, 207) [some of the analyses of words being rather questionable].

– Giorgio Buccellati, 1994

Criticism of Ungnad’s interpretation of the t-stem as expressing intention on the part of the agent. An analysis of texts of juridical interest leads the author to the conclusion that such semasiological differentiation between normal and t-stem must be rejected; he also points out that all of Ungnad’s examples, and most t-forms elsewhere, occur in secondary clauses.

– Giorgio Buccellati, 1994

Shows that pronominal suffixes in -m are used in the Code for the dative.

– Giorgio Buccellati, 1994

Connects Akkadian permansive and West Semitic perfect, the permansive being primary in meaning as with the Indo-European perfect; faʿila is the original form, and the use with substantives is secondary. Aorist yaqtul has, originally, a jussive and preterital function. In the imperfect ikaššad lengthening of middle radical is primary and proto-Semitic. The equivalent form in Arabic is yaqtulu, which is secondary; the -u, which in Arabic is a marker of aspect, became later in Akkadian a marker of mood, and was transferred to the other “tenses.” In general, it emphasizes the importance of Akkadian, though it is not considered as linguistically archaic.

– Giorgio Buccellati, 1994

Written in 1908 and published posthumously, this work is primarily concerned with Hebrew, but Akkadian is mentioned occasionally, and one whole section (pp. 41-48) is devoted to it: stress is assumed before geminate (e.g., ipárras), in the plural (aḫḫū), before -ma.

– Giorgio Buccellati, 1994

A practical version of SAYCE 1872 Grammar, consisting mostly of paradigms and exercises.

– Giorgio Buccellati, 1994

Defends the existence of the permansive, on the basis of comparative criteria. In the permansive of a noun, Akk. gives evidence of the formation of a tense which is more archaic than in any other language. All personal endings are attested, including the dual, except for first plural and second fem. plur.

– Giorgio Buccellati, 1994

A description of the tenses, in all stems, with ample documentation bearing especially on the morphology. The permansive (pp. 26-35), related to the West Semitic perfect, is considered as evidence of a stage when Semitic had only one tense, without temporal differentiation. The present and the aorist (i.e., preterite) are considered as having originally had the three case endings of the noun, of which the ending -i(m) was soon dropped for phonological reasons; -a(m) is the objective case, used before an accusative, with verbs of motion and in relative and conditional sentences (p. 37), -u(m) is the subjective case, used in a pluperfect sense and frequently after a relative or conditional particle; the construct aorist, i.e., without case endings, is due to phonetic decay, while mimation is an energic form. The same scheme obtains for the present (pp. 39-44); except that here the subjective case in ‑u stands for the future. The personal markers of number, being the same for verb and noun, prove that aorist and present, like the permansive, were originally nouns (pp. 44-48); the prefix vowel was originally added only to patterns with initial consonantal cluster, for euphonic reasons (pp. 48-50); similarly, the feminine t is the same that is found in the noun, and was probably originally a separated word, on the basis of evidence from Old Egyptian and Berber (pp. 50-52). In the conclusions it is emphasized, among the rest, that while “in the Aryan languages the noun presupposes the verb…in Semitic the verb presupposes the noun” (53).

– Giorgio Buccellati, 1994

Objects to the reconstruction by HAUPT 1878 “Oldest Semitic Verb-Form” by saying that “the primitive Semitic accent was on the penultima; possibly, at a still earlier time, on the ultima.” Also notes the first examples of the first person plural of the premansive.

– Giorgio Buccellati, 1994

A detailed and comprehensive monograph, including: a description of the means used for the decipherment, with a special discussion of the syllabaries (pp. 4-60); a list of syllabic and logographic values with a discussion of polyphony (pp. 61-115); an analysis of the lexicon (pp. 177-95) and a complete grammar (pp. 195-312) with an excursus on PN’s (pp. 115-77); a set of conclusions to show the validity of the decipherment (pp. 312-38) and finally, by way of appendix, an anthology of texts (pp. 339-69), and a glossary (pp. 370-86). - The grammar is divided in the usual way: phonology (pp. 195-207), morphology (pp. 208-96, 390-92) and syntax (pp. 296-312). Given the “apologetic” goal of the article (to justify the validity of the decipherment) much of the material is comparative in nature, in that it is meant to establish the Semitic character of the language; but it also has the advantage of a richer exemplification than in previous grammars, which in turn allows often for a more precise statement of the rules. With respect to the verbal tenses he follows OPPERT vs. HINCKS. The syntax deserves a special mention, which represents the most extensive treatement to date.

– Giorgio Buccellati, 1994

A morphological and syntactical analysis of the individual Semitic languages (with emphasis on Northwest-Semitic) followed by a brief comparative section in the conclusions. For Akkadian, von Soden’s presentation of 1952 is followed very closely, one difference being the interpretation of final -i (noted by E. Reiner in the kudurrus) as a dative-adverbial ending (pp. 170, 176-77, 185).

– Giorgio Buccellati, 1994

Stresses the need to distinguish more clearly the dialects on a geographical basis, and doubts the validity of the current thesis about the monogenesis of Akk. dialects from OAkk.

– Giorgio Buccellati, 1994

I. “The ‘Causative’ conjugation” is to be explained essentially in terms of syntax, namely as an elliptical sentence, more properly a compound sentence: ušabnī means properly that a subject A has ordered or induced another subject B to build something, but the governing verb has dropped. The prefix š is an original feminine personal pronoun, identical in function to the masc. pron, h, which is found as causative prefix in other Semitic languages. II. “The So-Called Feminine Ending -(a)t” is to be connected with adverbial -t(a), with t in yati (of Hamito-Semitic origin), with the nota accusativi *et in Hebrew. Its original function was to provide derivative words with special modifications of original meaning of root, of which modifications the feminine was only a particular one.

– Giorgio Buccellati, 1994

General criticism of Weinhof's notion of polarity, stated as follows: “If under certain conditions A becomes B, B will become A under the same conditions.” Specific areas of criticism are among others, broken plurals (no single contrasting pair because one singular has many plurals), and the West-Semitic tense system which should be explained on the basis of the underlying sentence type.

– Giorgio Buccellati, 1994

The basic process to express a conditional sentence in Semitic “consists in giving the protasis as a direct statement”; protasis and apodosis may be connected asyndetically or with the ordinary conjunction for “and;” in addition the protasis may be marked by means of special particles which “can be promptly identified as ordinary demonstrative elements” (p. 460). Akkadian uses the same process, and šumma is in reality a demonstrative element, formed of the demonstrative pronoun šū and the enclitic -ma. The original meaning of this particle is “this is (so),” (though the best translation remains “if,”) and hence the use of the indicative rather than the subjunctive.

– Giorgio Buccellati, 1994

The š prefix cannot be considered as exclusively causative, for it has another important function as elative, “in the broader sense of some specially emphasized connotation” (p. 472). Thus nominal forms with š express “intensification of the basic meaning” of adjectival bases, as with the pair rabū “great” ~ šurbū “supreme” (§ 4). The verbal correlative of this form is the so-called Š/D stem, which is in reality the factitive (D) of the elative, so that ušrabbi means “he made supreme” while ušarbi is simply “he made great” (§ 7, § 3). As a non-causative, the prefix may also be used in a declarative connotation (ušalbar, “he becomes old,” § 5), and as a “lexical element or base-complement” to form denominatives with quadri-consonantal bases “which, however, are not quadriradicals” (ušharrar “he becomes old,” § 6).

– Giorgio Buccellati, 1994

Review article with special discussion of h and the sibilants, terminative and locative (not to be considered like normal case endings), and weak verbs (Akk. had biconsonantal roots).

– Giorgio Buccellati, 1994

Akkadian -iš (seldom also -aš) is to be related to Ugaritic-Canaanite -ah not only functionally but also because both endings are identical in origin, as they are both diectic in nature, corresponding to the distribution of š and h in Semitic as defended elsewhere by Speiser (1935; 1947; 1952). Hence the original meaning of the terminative is assumed to have been deictic, with some such meaning as “away towards a specific point,” “earth-towards,” for example.

– Giorgio Buccellati, 1994

There are several unrecognized Btn forms in Hebrew which have in common with Akkadian an iterative and habitative, or durative, aspect, as with muttalliku / mithallāk “walking about, wanderer.”

– Giorgio Buccellati, 1994

Aspirated k, as a special phoneme in loanwords from Hurrian, is represented in WAkk. writing by means of signs with either K or Ḫ, e.g., k/ḫiššatu “a type of bondage.”

– Giorgio Buccellati, 1994

A long passage from a kudurru of Meli-Šipak exhibits 17 verbal forms, all of them in the basic stems, in the first section where the king calls on his successors to behave as well as he had done; in the second and third section, on the other hand, containing blessings and curses, there are 32 verbal forms, all of them from t-stems. The conclusion is that the basic forms are used for possibility, the t-forms for actuality.

– Giorgio Buccellati, 1994

Akkadian fractions are expresed some times by the masculine of the ordinal, but more generally by the feminine, apparently in apposition to an understood qâtum “portion”, e.g., šaluštum (qâtum) “the third part” = 1/3.

– Giorgio Buccellati, 1994

The first part deals exclusively with the Hebrew segolates, the second (pp. 279-311) with the vocalism of the verbal system. Here the author utilizes at first only West Semitic data to show that at the basis of verbal finite forms in Arabic, Hebrew and Aramaic there is not just one, but three fundamental vocalic patterns: with the help of Akkadian, which is now introduced and fully utilized (pp. 290-309), the three patterns can be described as qatal, qi/uti/ul, qati/ul. Thus for example every derived Akk. stem contains all three patterns in the three tenses, e.g., in the D uqaṭṭal, quṭṭul, uqaṭṭil. The second main conclusion is that vocalization is irrelevant for determining the voice, and the criterion is instead the position of the subject: if prefixed, the form is active, if suffixed, passive. This means that there are no transitive or intransitive roots, that the vowel is indeed of little weight in Semitic morphology, that Semitic was originally agglutinative (p. 298 f.). Other points affecting Akk. are discussed by way of appendix (pp. 300-9): the last vowel u is primary in the permansive and imperative D, not derived from a; of the derived stems, D and Š seem to be older; Akk. is the most archaic language (morphologically) and should receive more attention in comparative Semitics. [See also H. TORCZYNER, “Zur Semitischen Verbalbildung,” ZDMG 66 (1912), pp. 87-93; ED. KÖNIG, “Zur Verbalflexion im Semitischen,” ZDMG 66 (1912), pp. 261-66.]

– Giorgio Buccellati, 1994

The expression of the type šar šarrāni has a superlative value derived from a partitive one, in that the element which is singled out is considered as an ideal: “The ideal of a king among the existing kings.”

– Giorgio Buccellati, 1994

Discussion of S. Smith 1949 (The Statue of Idri-mi) and D.J. Wiseman 1953 (The Alalakh Tablets), with long bibliography of reviews and studies. In two sections: I: comments and notes on individual texts, in great detail; II: contributions of Alalakh texts to Canaanite linguistics and Hebrew literary studies. Alalakh differs from EA in that it consists mainly of lists and similar stereotyped texts; there is a Hurrian background; no Canaanite glosses occur; in general, a rather ‘adequate’ knowledge of Akkadian is evidenced. Tsevat then lists the various Canaanisms occuring in the texts, which are rather few.

– John L. Hayes, 1974

Poses the problem of identity of a Semitic language: we recognize it as such when we meet one, but how can this be articulated in rigorous, descriptive-typological (and not genealogical-historical) terms? First U. criticizes some standard criteria as insufficient: role and function of the consonants, internal vocalic changes, triliterality, phonological traits such as ʿ or ḥ, syntactic traits such as parataxis or word order. Then he gives as positive suggestions the need of Hamito-Semitic comparisons, structural, lexical, statistical and isogloss studies, and the definition of thought categories as conditioned by linguistic factors.

– Giorgio Buccellati, 1994

From the methodological point of view, calls attention to the importance of the structural approach (pp. 13, 22-24).

– Giorgio Buccellati, 1994

The basic conjunctive device, between sentences as well as between nouns is asyndeton; the particle u means “and also,” and the enclitic -ma, even in a conjunctive function, is the same as the emphatic particle: it stresses the first sentence which gives the basic grounds from which the second sentence is derived as a consequence. Therefore -ma is to be translated as “and then,” “and as a result,” and only in later periods it comes to have a simple conjunctive value “and.”

– Giorgio Buccellati, 1994

Reconstructs the original forms of the pers. suffixes for proto-Semitic as follows: ‑ya; ‑kâ; ‑kî; ‑hû, ‑šâ; ‑nâ; ‑kumû (?), ‑kinnâ; ‑humû, ‑šinnâ. (Akk. ‑šu and ‑kunu are influenced analogically by the respective feminines). The noun before these suffixes is essentially in the construct state, but often giving rise to a closer juncture. The construct state is originally not declined, but simply shows a vowel to resolve consonantal clusters when necessary; occasionally, esp. in Arabic, this vowel is secondarily made to coincide with the case endings. In Akk., stress causes a secondary lengthening of the vowel before the suffix of the 1st pers. sing.

– Giorgio Buccellati, 1994

Reconstruction of the archaic inflection of the determinative pronoun, based especially on OAkk. texts from Susa.

– Giorgio Buccellati, 1994

The article is really about the permansive (= stative), to the study of which it contributes important insights. The permansive “describes the process of an action as concluded or as extent in its completeness” (p. 277). Of itself, the permansive is neither active nor passive: ṣabtāku means “I am one in whom the action of seizing has come to completion”, while eqlam ṣabtāku means “I am one, in whom the action of seizing the field has come to completion” (p. 278). Without an object, it is normally passive, whereas the presence of the object makes it active (p. 281). The permansive is to be interpreted as the Participium perfecti (= verbal adjective) in the status indeterminatus (= predicative state), and one should not distinguish between the verbal adjectives and the stative, as is traditionally done (p. 278). The permansive is not defined temporally, and may also be used for the future (p. 281, n. 1, with several examples). In many cases, the situation described by the stative may best be translated by using the verb “to have”, e.g., mārī waldāku “I have children” (pp. 279-80).

– Giorgio Buccellati, 1991

Identifies the form tikal as an example of the “oldest Semitic imperative.”

– Giorgio Buccellati, 1994

Primarily on the basis of NB documentation, but also CH, the view is advanced that forms without t describe the action as occurring through external, forms with t through internal impulse, so that iddin means “he gave because external circumstances made him to” and ittadin “he gave because he felt in himself the impulse to do so” (p. 253). The connotation of external obligation may often remain untranslated, so that the simple verb may be used to translate the form without t, whereas the translation normally must bring out the connotation of intentionality for the t-forms (p. 255). In a sequence of verbs expressing intentionality it is sufficient if only one of them occurs in the t-forms (p. 254).

– Giorgio Buccellati, 1994

The genitive with ša is used to shift the emphasis on the head of the annexion, since otherwise in a constructive emphasis is on the construct.

– Giorgio Buccellati, 1994

Adverbs ending in -atta(m), -atti occur in Babylonian and even more frequently in Assyrian, but never in poetry. The best known function is for adverbs of manner, but in OB it occurs also for adverbs of time. The ending is to be analyzed as -att- plus -am or -i (< *ay). The latter two elements occur also independently as adverbial endings. The element -att- is “ein pronominales Element zum Ausdruck der Zugehörigkeit” occurring in the possessive pronouns of the type attuka “yours,” and possibly related to the Hebrew “nota accusativi” (Brockelmann) and the -t- marker of accusative found in the Akkadian personal pronouns and in Cushitic.

– Giorgio Buccellati, 1994

The pattern purussāʾ carries a special meaning in that it serves to express a planned activity, conditioned by specific circumstances and procedures, e.g., rugummū “complaint” as different from rigmū “call, scream.” Von Soden lists 20 words of this pattern from transitive-active roots, and 11 words from “neutral” roots; in the latter case, the pattern means entering in the state of the corresponding adjective under special conditions, e.g., ḫuluqqū “economic loss.” The pattern does not occur in the oldest periods because the juridical conceptualization was not yet sufficiently differentiated [but see GELB].

– Giorgio Buccellati, 1994

Importance of Old Babylonian.

– Giorgio Buccellati, 1994

The first entry (423-26) contains additions to von Soden 1939 “qutullāʾ” and a rejoinder to PETERS 1942 “qutullāʾ”: the Arabic forms do not belong to a homogeneous semantic group; instead, from a comparative point of view one should add verbal abstracts from Syriac.

– Giorgio Buccellati, 1994

The four main vocalic phonemes of Akkadian are occasionally realized with phonetic variations which are indicated by the writing in the following manner: phonetic o written either a-u (e.g., ka-uk-ki), or with a in place of u and viceversa (i-ta-ar-ru instead of i-tu-ur-ru); phonetic ü written u-i, i-u, or with i in place of u; phonetic ö written e-u, and possibly phonetic ë written a-i, i-a. The majority of the examples is taken from letters, thus being possibly an index of vulgar speech, but in any case they are simple phonetic variations, with no morphological consequences. The device of broken writing is especially frequent in OB, that of substitution in NA.

– Giorgio Buccellati, 1994

Disagrees with Heidel’s main conclusion that “the Semitic verb system is built on the triradical verb,” accepts the two main groups distinguished by Heidel, and adds observations on matters of detail.

– Giorgio Buccellati, 1994