Bryn Mawr Classical Review

Within Greek and Indo-European word-formation, nominal composition plays an important role, which is testified to by the variety of productive compositional types and the frequency of compound words in all registers. Ventris' decipherment of Linear B has enabled us to observe how compounding works in Mycenaean, but up until now the lack of a comprehensive study of Mycenaean compounds has blurred the full picture and hindered the comparison of Mycenaean forms with those of first-millennium Greek.¹ The special issues concerning Mycenaean -- e.g. the nature of the documentation (consisting solely of archive documents) and the frequently open interpretation of words written in a syllabic script -- add further complexity to any attempt at analysing its compounds. It is therefore a welcome improvement that we now possess a monograph dealing exclusively with the first stage of Greek compounding, as attested in the Mycenaean archive tablets.

Waander's book, the seventh issue in the Collana di filologia e antichità egee directed by Louis Godart and Anna Sacconi, aims to collect all Mycenaean compounds while offering a structural analysis (regarding which types are attested) and dealing with interpretative issues. The volume is thus divided: Introduction (pp. 1-2); Chapter I: Types of compound (pp. 3-6); Chapter II: Compound verbs (+ derivatives) (pp. 7-12); Chapter III: Compound adverbs (pp. 13-14); Chapter IV: Nominal compounds 1: 1V ('verbal government compounds') (pp. 15-20); Chapter V: Nominal compounds 2: Prepositional government (pp. 21-28); Chapter VI: Nominal compounds 3: 2V (pp. 29-42); Chapter VII: Nominal compounds 4: Type 2-to (/a-ktitos/) (pp. 43-48); Chapter VIII: Nominal compounds 5: Possessive compounds (pp. 49-64); Chapter IX: Nominal compounds 6: Endocentric compounds (pp. 65-68); Chapter X: Derivational compounds (pp. 69-72); Appendix: Compounds (?) 'incertae sedis' (pp. 73-74). Bibliography, alphabetic index, and a convenient reverse index end the volume.

The strength of this work is that it provides a comprehensive list of Mycenaean compounds (including forms that could potentially be compounds and whose interpretation is controversial) and offers a handy working tool for those interested in nominal composition. As the above summary of contents shows, forms are analysed according to compositional class and the author devotes a whole chapter to each class/subgroup. Within individual chapters compounds are arranged alphabetically; translation into English and a short commentary follow. References in these sections are far from extensive: it is probably assumed that the reader interested in exploring their interpretation further should refer to Aura Jorro's Diccionario Micénico and the references there contained.² This approach shows that Waanders's aim is more that of providing a workable corpus than of producing a full-fledged interpretation of the issues at stake in Mycenaean compounding and a comparison with later Greek.

The chapters dealing with the classification of compounds provide the most valuable contribution. For instance, Chapter II presents a welcome overview of compound verbs, as most treatments of Mycenaean compounding refrain from including preposition + verb forms and their derivatives (twenty-three forms according to Waanders's statistics); Chapter IV instead collects all verb-first compounds, a type of controversial origin which in Mycenaean seems to be specialised for personal names (forty-one forms). Chapter V deals with prepositional government compounds: Waanders lists some sixty compounds, including those of difficult interpretation. The prevalence of ἐπί as the first member, clear in later Greek, can already be noted in Mycenaean: out of sixty compounds, thirty are compounded with /epi/ and /opi/. In Chapter VI, Waanders discusses compounds with a verbal second member. In first-millennium Greek this is by far the most frequent class, further analysable into root compounds, those with a verbal noun in -ος in the second member, and those compounded with agent nouns in -της, participles, verbal adjectives in -τος, and verbal nouns.³. Waanders's list confirms the productivity of this type: there are about one hundred and thirty-one compounds in this class (though note that the list excludes those compounded with a verbal adjective, collected separately in Chapter VII: twenty-six in total). Because of its semantics (designation of agents and actions) this is also the class that most frequently features in the appellative lexicon, with numerous terms indicating occupations or functionaries, such as a-re-pa-zo-o 'unguent-boiler', di-pte-ra-po-ro 'leather-bearer'(?), ku-ru-so-wo-ko 'goldsmith', ku-wa-no-wo-ko-i 'enamel-workers', etc.

Possessive compounds are another frequent category. Judging from Waanders's list in Chapter VIIII more than a half of the circa one hundred and sixty possessive compounds are personal names, although the exact interpretation of many is far from certain. Chapter X deals with what Waanders calls 'endocentric compounds' and in my opinion would better be termed 'determinative compounds' (i.e. noun-noun or adjective-noun compounds, 'endocentric' being a term that applies to other categories as well). These compounds number seventeen in total: an interesting datum, corroborating the theory that this was the last type to develop in Greek.⁴ The last chapter discusses a few forms of difficult interpretation that might derive from syntagms.

I found Chapter III, which lists three possible compounded adverbs, the least convincing of all. Are the forms here listed adverbs formed in a different way from nominal compounds or are they rather instances of compounds used as adverbs? By way of clarification I give an example: could *pe-ru-si 'last year' (reconstructed from the adjective pe-ru-si-nu-wo) be interpreted as a prepositional compound? Perhaps these forms could have been dealt with in a different way: as iterative compounds in the case of a-mo-ra-ma (and we-te-i-we-te-i, cf. below) and as derivational compounds in the case of za-we-te and pe-ru-si- (see also my comments below). The criterion behind this chapter is misleading: while in the rest of the book compounds are classified according to structure and relation between the members (verb-first, verb-second, endocentric, etc.) those in Chapter III are classified according to function.

Waanders's work clearly shows that compounding is a central word-formation device in Mycenaean Greek. This said, and given the importance of Mycenaean compounding within the larger frame of Indo-European nominal composition, there are a number of issues that this book strikingly does not address. Not all of them may be relevant for all readers, but this reviewer feels that they ought to be pointed out nonetheless.

The main concern is with the theoretical framework of the book. While not openly stated, it is obvious that Waanders is inspired in his analysis of Mycenaean compounds by classic categorisations such as Wackernagel (1905) and Risch (1974).⁵ As the Mycenaean corpus is limited and rather standardised, it is hardly surprising that the book fails to provide an innovative classification of compounds and that the terminology it uses is the traditional one. However, when it comes to the definition of what a compound is, Waanders's treatment is unnecessarily superficial and potentially misleading for those unfamiliar with the subject. I will give two examples. The introduction, devoted to the definition of 'compound', maintains that "Theoretically, the formal (morphological/word-level) status of such constructions as Indonesian matahari..., the (indefinite) construct state in Arabic and Hebrew, the qualifying relationship in Turkish (as against the possessive relationship), and comparable combinations in, e.g., Chinese, is not unproblematic" (p. 3). This statement is not followed by an explanation of what 'construct state', 'qualifying relationship', and 'possessive relationship' mean: readers are left wondering about the arcane differences between these types of constructions, and in what ways nominal composition in the languages mentioned may differ.

After a few general remarks on the criteria used to distinguish compounds from non-compounds (comments similarly blighted by an alarming degree of subjectivity: e.g. the statement on pp. 2-3 that 'instances like Latin iusiurandum 'oath'...with both 'members' inflected, do not look like compounds to me'), Waanders provides a definition for Greek and Latin compounds: "units which contain two lexical morphemes, and which are experienced by the language users as one word under one accent, and, in the case of nominal and verbal stems, inflected only at the end; ideally, the constituent morphemes are bound morphemes, in which case there is no room for doubt" (p. 3). This definition ought to be complemented by two necessary observations: (1) our understanding of the Greek perception of 'word' is not as uncontroversial as it appears from Waanders's statement; hence, using 'one word under one accent' as a criterion for the compound status of Greek words is highly problematic; (2) there is plenty of inscriptional evidence showing that two or more words may be univerbated (probably under the same accent, e.g. o-u-di-do-si in Mycenaean or ὁ παῖς on Greek vase inscriptions), but this does not entail that they belong to nominal composition. Using stress as a criterion to decide on the compound status of Greek words is likewise problematic and further elaboration on this point would have been beneficial.

Experienced readers may not censure this book for failing to provide a fuller analysis of central concepts such as 'endocentric' and 'exocentric'. Nonetheless, in some cases the absence of more elaborate discussion results in obscurity. There are several examples of such shortcomings in this work. I will limit myself to two. Firstly, Waanders ascribes a frequent 'compounding suffix' to Greek compounds for which he provides neither an example nor bibliographical references.⁶ Secondly, the book contains no comprehensive overview of compositional types (determinative, bahuvrihi, etc.) either in Greek or cross-linguistically: after the general remarks in Chapter I (pp. 3-6: three pages!) the subsequent chapters deal with types such as prepositional government, verbal government, etc. which are never unequivocally defined.

Readers of this review may feel that much of its criticism is marginal and should only concern those who are interested in the minutiae of Mycenaean compounds, but it is my belief that a book that does not place any topic into its broader picture, from both a theoretical and a historical point of view, misses its chief goal. Except for the interpretation of individual words, students of nominal composition as a linguistic phenomenon will learn little from this book, as it attempts no kind of comparative approach. For instance, Chapter III deals with forms used as adverbs but does not mention that a-mo-ra-ma 'every day' may represent the continuation of an old Indo-European type, that of 'iterative' compounds,⁷ with all the consequences this has for the relation of Mycenaean with the parent language. Similarly, the volume does not deal with the incidence of personal names in the corpus of Mycenaean compounds (and in Mycenaean vocabulary in general). As a few compositional types are best represented by personal names, this is of some importance for our understanding of the phenomenon. For instance: is the fact that verb-first compounds (pp. 15-20) are mostly limited to personal names relevant for the origin of this controversial type in the parent language?⁸

Discussion of central issues in Waanders's book is often subjective and not substantiated by sufficient (if any) bibliographical references. Examples are: 'tmesis' (p. 5); the development of determinative compounds (Chapter VIII); and all forms whose interpretation is based on DMic without further reference to later literature. The bibliography itself is rather succinct (five pages), and oddly lists specialist articles while omitting most of the basic references on Greek nominal composition. ⁹ As a few of the listed works seem never to be quoted in the body of the book, it would be useful to know what criterion Waanders followed when deciding what to include and what to exclude from the bibliography.

It would have been helpful for the reader if the frequent and complicated abbreviations used in the book were glossed in a list of abbreviations rather than in the introduction on p. 4. Besides, the list here provided is not even complete: for instance, there is never an expansion of PN and MN. In many other works PN stands for 'personal name'; here it is used for 'place name' (while personal names are abbreviated with MN = 'man's name' and WN = 'woman's name').

More detailed concerns regarding individual Mycenaean words and further minor quibbles follow. a-mo-ra-ma is not a 'compound adverb' (p. 13): most probably, an iterative compound which only happens to be used as an adverb; it would be interesting to know in what respect the author considers Bartonek's interpretation of za-we-te as /tsaweteros/ surprising (fn. 27 p. 13); it is not clear to me why Chapter III does not include we-te-i-we-te-i 'every year', another iterative compound with a clear adverbial usage (it is to be found nowhere in the whole book); if the origin of the verbal first member in ti/si compounds is disputed (p. 17), why is their 'internal syntax' unequivocally verb-complement? Some scholars have suggested that the first member was a noun; discussion of why forms ke-u-po-da, da-i-ja-ke-re-u, and da-i-wo-wo 'show a remarkable morphology' (pp. 19-20) is needed; the possibility of interpreting some compounds with a prepositional first member as possessive (p. 21) does not apply only to prepositions which have a 'reciprocal meaning' (whatever that means): cf. the type ἐπιάργυρος 'with silver on top'. ¹⁰

Finally, after reading this volume students of first-millennium Greek may be left with the impression that Mycenaean and Greek nominal compositions have little in common: there is no statistical information on the individual composition types vis-à-vis later Greek and no comparison between the two stages of the Greek language is ever attempted. Does Mycenaean nominal composition differ from later Greek nominal composition in any respect? Are some compositional types more or less attested? Do Mycenaean compounds show characteristics in common with other I-E languages but not Greek? These are examples of some of the basic questions that one would like a work on Mycenaean nominal composition to tackle, and which this work clearly does not.

In conclusion, Waanders's book does a good service to Mycenologists and to scholars working on nominal composition in that it provides a comprehensive corpus equipped with a basic structural analysis. However, it will be of limited use to those who are not working in the field. On a more general level, this book exemplifies a tendency that has dominated Greek and Latin linguistics in the past, namely the reluctance to use material from modern languages and to analyse Greek and Latin also in the light of results achieved by general linguistics and typology. As this is hardly a bulky volume (one hundred pages printed in 14 point type with 1.5 line spacing), the lack of further discussion of central issues is striking and constitutes its most serious flaw.

1.  Two useful articles that deal with Mycenaean compounds, though only partially, are M. Buzalkovska-Aleksova, 'Some Parallel Elements in Mycenaean Compounds -- Appellatives and Personal Names', in S. Deger-Jalkotzy, S. Hiller, O. Panagl (eds.), Floreant Studia Mycenaea. Akten des X. Internationalen Mykenologischen Colloquiums in Salzburg vom 1.-5. Mai 1995, vol. 1, Wien, pp. 177-184; and T. Meisser, O. Tribulato, 'Nominal Composition in Mycenaean Greek', Transactions of the Philological Society 100 (3), 2002, pp. 289-330.
2.  F. Aura Jorro, Diccionario Micénico, Madrid 1985-1993.
3.  See E. Risch, Wortbildung der homerischen Sprache, Berlin - New York 1974, pp. 194-212.
4.  See E. Risch, 'Griechische Determinativkomposita', IF 59 (1), 1944, pp. 1-61 and id., 'Griechische Determinativkomposita', IF 59 (3), 1949, pp. 245-294 (both reprinted in id., Kleine Schriften, Berlin 1981.
5.  J. Wackernagel, Altindische Grammatik, vol. II.1, Göttingen 1905; E. Risch, cit.
6.  See on the other hand A. Debrunner, Griechische Wortbildungslehre, Heidelberg 1917, p. 73.
7.  Cf. P. H. Salus, 'The Types of Nominal Compound of Indo-European', Orbis 14 (1), 1965, pp. 38-62.
8.  For the classic view that the type originated in the onomastic lexicon cf. C. Frei-Lüthy, Der Einfluss der griechischen Personennamen auf die Wortbildung, Heidelberg 1978.
9.  Examples are O. Landau, Mykenische-griechische Personennamen, Göteborg (out-dated, but still a useful reference work); A. Debrunner cit., the seminal articles by Ernst Risch cited in note 3; and various articles by J.-L- García Ramón dealing with individual Mycenaean personal names.
10.  See H. Forster, Zur Geschichte der griechischen Komposita vom Typus ἐπίχρυσος, Zurich 1950.