Bryn Mawr Classical Review 2007.05.15
Benjamin W. Fortson IV, Indo-European Language and Culture: An Introduction. Blackwell Textbooks in Linguistics, 19. Malden, MA: Blackwell Publishing, 2004. Pp. xviii, 468. ISBN 1-4051-0316-7. $44.95.
Reviewed by Gary Holland, University of California, Berkeley (firstname.lastname@example.org)
Word count: 2529 words
It is refreshing and reassuring to see an introduction to Indo-European linguistics taking its proper place in a prestigious series alongside well-received texts on government and binding theory, lexical-functional syntax, morphological theory, and phonology (among others). The author, Benjamin W. Fortson IV (hereafter F.), has given us an up-to-date, detailed, and accurate introduction to Indo-European linguistics and culture. This book represents the majority point of view concerning basic issues of Indo-European reconstruction, cetainly in its acceptance of three laryngeals, of a vowel system consisting of five short and long vowels, and in its rejection of the glottalic theory. It surpasses its competitors in English through the liveliness of its presentation, its attention to cultural matters, and its avoidance of overly personal and idiosyncratic views. Furthermore, F. has constructed a set of exercises after each chapter, thus allowing student readers to determine the relative importance of the concepts presented and to strengthen their grasp of the necessarily dense and detailed subject matter.
The organization of the book is very clear, falling into two major parts: the first eight chapters (pp. 1-153) offer an introduction to the comparative method and to the Indo-European language family (ch. 1), a survey of Proto-Indo-European culture and archaeology (ch. 2), an overview of Proto-Indo-European phonology (ch. 3) and morphology (ch. 4), presentations of the verb (ch. 5) and the noun (ch. 6), of pronouns and other parts of speech (ch. 7), and of Proto-Indo-European syntax (ch. 8). The second and more extensive part of the book, chapters 9-20 (pp. 154-411), offers descriptions of, and text samples (with translations and running commentaries) from the major Indo-European branches--Anatolian (ch. 9), Indo-Iranian I: Indic (ch. 10), Indo-Iranian II: Iranian (ch. 11), Greek (ch. 12), Italic (ch. 13), Celtic (ch. 14), Germanic (ch. 15), Armenian (ch. 16), Tocharian (ch. 17), Balto-Slavic (ch. 18), Albanian (ch. 19). Chapter 20 is devoted to Fragmentary Languages (Phrygian, Thracian, Macedonian, Illyrian, Venetic, Messapic, and Lusitanian). In addition to the exercises mentioned above, each chapter (except 19 and 20) concludes with suggestions for further reading and lists of concepts for review. Further, beginning with chapter 9, F. provides twelve sets of PIE vocabulary items, arranged according to semantic field. The book ends with a glossary of technical terms, a well-selected bibliography, and indices of forms arranged by language.
In the following I will comment on some of the chapters, calling attention to a very few errors and to a few points where I think that views different from the author's are preferable. Chapter 1 offers a very clear exposition of the comparative method. F. discusses the anomalous nature of the IE family tree, with its ten traditional branches radiating out from the node representing Proto-Indo-European, observing that "[i]mplicit in the traditional diagrams is the notion of a more or less simultaneous "breakup" of the proto-language into ten or more dialect areas (the future branches)" (p. 9). This situation stands in sharp contrast to the trees reconstructed for other language families, which typically exhibit much more binary branching.
Chapter 2 gives an excellent overview of Indo-European culture. In an attempt to make a connection between Hittite and the well-known horse sacrifice in Indic, Roman, and Irish traditions (where it is involved in the consecration of kingship) and the concomitant ritual copulation with the sacrificed horse, F. states "interestingly, though, in Hittite law copulation with animals was a punishable offence except copulation with horses or mules" (p. 25). One should note, however, that this law also states that the person who has committed such an act cannot approach the king, nor can he become a priest, presumably because he is made ritually impure by the act. Consequently, it does not seem that this Hittite law is immediately relevant to the Indic and Irish traditions. In the same chapter, in a presentation of Georges Dumézil's tripartite ideology (the requisite criticisms are duly made), F. states "[i]n the second-millennium-BC Mitanni documents ..., which contain the first attested words in an Indic language, the names of the gods invoked at the signing of a treaty are Mitra and Varuna (sovereign and priestly first function), Indra (the warrior god, second function), and the divine twins the Nasatyas (the third function)" (p. 28; cf. also p. 184). Here, one might also add that these are not the only gods mentioned in this treaty; they occur after a listing of at least 100 other divinities. The author(s) of the treaty apparently list all the gods they can think of on both sides as witnesses. Mitra, Varuna, Indra, and the Nasatyas are not privileged here.
Some details: pp. 31-32 the bold-faced cadences in the scansions are not differentiated clearly enough from the regular font. -- p. 34 the Old English example of Behaghel's law of increasing members is not clear: the last member of the conjoined set indeed consists of two elements, proper noun and adjective, but it is only three syllables long, as is Heorogar, the first member of the set.
The phonology chapter is a model of conciseness and clarity, with an excellent presentation of the laryngeal theory. A few details: p. 50 ll. 3-4, for palatal stop read palatal fricative. -- p. 58 F. states that "[t]he oldest Indo-Iranian languages, Vedic Sanskrit and Old Avestan, also reflect consonantal laryngeals in certain positions as hiatus or glottal stops". This is indeed the case, but at least in Vedic, consonantal reflexes of the laryngeals also make position. -- p. 67 *h2rnbhi should be *h2nrbhi.
The introduction to morphology is also clear and concise. A few details: p. 73 sect. 4.12 in a presentation of ablaut, F. cites the English forms sing, sang, sung, song, and states that "[t]he particular vowels of these words, in fact, are descended directly from ablauting vowels of PIE." Here, both sang and song should be PIE o-grade (Germanic sang in both cases), and there has been a split (within the history of English) in the vocalism conditioned by word class.
The presentations of the verb, the noun, and pronouns and other parts of speech are again very clear. A few details: p. 104 l. 7 for hümant- read humant- with long u. -- p. 104 sect. 6.6 F. asserts that Hittite hupparas writes an animate nominative huppar-s. The writing hupparas is ambiguous, since it might also represent a thematicized common gender nominative. -- p. 105 sect. 6.12 implies that the Hittite directive ends in long -a. The directive of the word for 'sea' is cited with a long final -a (reflecting scriptio plena, again on p.165), which, to my knowledge, is not attested. -- p. 115 sect. 6.51 the circumflex accent on theous should be an acute. -- p. 127 sect. 7.3 all the Old Avestan forms ending in vowels should have their length marked. Latin nobis and nostrum are in the wrong positions in the paradigm. -- p. 133 F. states that Modern French ne descends from Latin nec. This contradicts the generally accepted derivation of this item from a proclitic Latin non. French ni (Old French ne) indirectly continues Latin nec. -- p. 134 Hitt. kuiski is derived from *kwis-kwe. Since -ki and -ka are both found as indefinitizing particles in Hittite, it is unlikely that they come from IE *kwe (see Puhvel, Hittite Etymological Dictionary, s.v. -ki). At a minimum, there should be a comment on the special treatment of the labiovelar, since otherwise they are preserved in Hittite. -- p. 135 7.28 ON *hánn should have an asterisk; it is not attested, but inferred from the dative hánom.
In the very welcome chapter on syntax, much attention is paid to questions of the underlying order of constituents and to the various movement rules that account for the surface orders. F. provides good discussions of the processes known as Wackernagel's Law, of WH movement, of fronting and topicalization, and of relativization.
The chapter on Anatolian contains short grammatical outlines of the principal Anatolian languages and gives valuable text samples, with translations and running commentaries.
This pattern is followed in all the remaining chapters. One detail: p. 169 sect. 9.54 the derivational process in Luwian that consists in forming nominatives and accusatives with an -i- stem (not long -i-) should not be called i-mutation (since this term is used for i-umlaut in other sources), but rather i-motion.
The Indic chapter contains an outline of the historical phonology and grammar of Vedic and a text sample. It further has a concise discussion of Middle Indic. Some details: in the discussion of the reflexes of syllabic resonants followed by laryngeals (p. 189 sect. 10.37), F. doesn't mention the commonly held view that -ur- (both long and short) originally occurred after labials, and -ir- (both long and short) after coronals. -- p. 190 sect. 10.38, F. should comment on the transliterations of ai and au with long a found in older works (such as Whitney!) as well as on the long e, long o transliterations. -- p. 194 sect. 10.46, in the Vedic citation pati should be pahi (both with long a).
In the Iranian chapter (p. 205 sect. 11.12), Young Avestan is dated to the 9th or 8th c. BCE, simply on the grounds that it seems more archaic than Old Persian, which we know is from the 5th and 4th cc. BCE. Here, the methodology seems wrong. For instance, if we were given comparable sets of Modern Icelandic and Modern German texts, and applied this methodology, we would conclude that the Icelandic was much earlier than the German. Short phonological and grammatical sketches and example texts from both Old and Young Avestan and from Old Persian are given. On p. 214 sect. 11.34a, I was surprised to read that the etymology of Vistaspa was disputed. This chapter also contains an introduction to Middle and Modern Iranian.
In the chapter on Greek, F. characterizes ζ as "a voiced affricate dz" which is "also pronounced as a consonant cluster zd" (p. 226 sect. 12.11). Against the notion that ζ represents a voiced affricate, it should be noted that ζ behaves as a cluster: it always makes position in metrical texts, and no matter what its etymological source, final -ν disappears before it in compounds, as it does before στ-. -- p. 233 sect. 12.42, Doric éstan is derived from *e-sth2-nt, a root aorist with vocalized laryngeal. F. should comment on the syllabification here and on the sonority hierarchy that would yield such a result.
The Italic chapter contains an excellent overview of Latin-Faliscan and the Sabellic languages, with useful sample texts. A few details: p. 262 sect. 13.55, two rather than three dots are used to represent the interpuncts in the Faliscan inscription. Although F. cites this text by its number in Vetter, he does not follow Vetter's reading. Gabriella Giacomelli's La lingua falisca (1963) is listed in the bibliography, but her reading is not followed either. -- In the Oscan text (Ve 11) p. 268 sect. 13.77 tancinud should be corrected to tanginud (again on p. 269).
The Celtic chapter offers an excellent overview of the older Celtic languages, with phonological and morphological sketches, and sample texts in Gaulish, Celtiberian, Old Irish, Old Welsh, Breton, and Middle Cornish. A few comments: p. 276 F. states that -m > -n in Gaulish, but forms in final -m are preserved in the texts from Larzac and Naintré. -- p. 280 sect. 14.21 and 22: the IPA sign is wrong: the mid back unrounded vowel sign is used for gamma (also on p. 292). -- p. 291 sect. 14.56 Welsh <ll> is the voiceless lateral but the IPA sign for a velarized l is used instead. Similarly, Welsh <rh> is characterized as aspirated [hr] but this should be simply a voiceless [r].
The Germanic chapter follows the previously established pattern, again with outline historical grammars and sample texts with running commentaries. A few comments: p. 308 for ON rera 'he rowed' read reri. -- p. 309 sect. 15.33 F. states that "[f]rom the available evidence it appears that runes were originally restricted to use in magic or religious rituals." On the contrary, the earliest runic inscriptions seem to have nothing to do with magic, since they are possessors' names, makers' names, or object names, e.g., raunijaz on the Ovre Stabu lance head (2nd c.). In the same section, F. states that "[a]n early transitional version of the Runic alphabet was used to write the Germanic name harigasti ... " Although I agree that runes must have developed from a North Italic alphabet, there is no evidence that the alphabet that occurs on the Negau B helmet is anything other than just a North Italic alphabet. The use of the sign[khi] to write [g] has an early parallel in the Venetic alphabet. -- p. 310 sect. 15.36, F. should point out that the Gallehus horns are no longer extant. -- p. 325 sect. 15.76, F. gives the Second Merseberg Charm as a text sample. I am surprised that F. was able to resist citing the parallel passage from the Atharva Veda. -- p. 328 sect. 15.84, the seventh century is given as time of onset of Viking raids; the late eighth century is more accurate. -- p. 331 sect. 15.97, F. should point out that the negative suffix -at, -a is restricted to early verse.
A noteworthy feature of this book is the chapter devoted to fragmentary languages. This, again, is for the most part excellent, but in the presentation of Venetic there are a number of misstatements:
"Some of the similarities [with Italic, GH] include the development of certain voiced aspirated stops to f (written variously ph or vh), such as the dative pl. louzerophos 'for the children' (< *-bhos) and the s-aorist vhagsto 'he did' (cp. Lat. fac-ere 'to do', late PIE *dhh1-k-). However, other outcomes are seen as well that confuse the picture: alongside the dative plural ending in -phos is also one in -bos (e.g. andeticobos 'for the sons of Andetios' [?]), and *dh became z in the word louzerophos seen above (from *loudhero-)." (p. 406 sect. 20.17)
Here F. inconsistently confuses a simple transliteration of the Venetic script with a phonetic interpretation. F. further states that "[t]he Venetic alphabet does not contain separate letters for the voiced stops" (p. 407). The aspirated stop signs <phi> and <khi> are generally interpreted as representing [b] and [g] respectively (the latter implicitly accepted by F., cf. vhagsto), and the sign <zeta> is ordinarily interpreted as representing [d] (cf. G. B. Pellegrini and A. L. Prosdocimi, La Lingua Venetica, and Michel Lejeune, Manuel de la langue vénète). Thus, the problems noted by F. disappear, and the parallels with Italic (specifically with Latin) become even more striking (IE *bh and *dh > Ven. f- in initial position, but Ven. -b- and -d-, respectively, in medial position). Such misstatements are so untypical of this book that it is a bit of a shock to read them.
I would like to conclude by stressing that this is an excellent textbook. I have taught from it, and the students in my class not only learned a great deal from it, they also seemed to enjoy the book almost as much as I did.