Indo-European Perspectives: Studies in Honour of Anna Morpurgo Davies

The first thing a prospective reader of Indo-European Perspectives (hereafter IEP) notices is the striking black and green cover, featuring an inscription in Hieroglyphic Luwian wishing “long days and many years and good abundance” to the honorand of this big, rich festschrift. Anna Morpurgo Davies has retired from the chair of comparative philology at Oxford, and an all-star group of linguists has contributed articles in her honor. As with all such collections, readers will find some chapters more interesting than others, but the over-all quality is high. It’s customary in festschrift contributions to include a sentence or two congratulating the recipient or noting how the recipient’s classes changed the writer’s life. Although such notes do appear in IEP, many of the chapters make a more significant acknowledgment of Morpurgo Davies’ accomplishments and influence by responding to or building on her work. In addition, many contain suggestions for further work. As a result, IEP is a good introduction to current research in Indo-European studies and would be valuable reading for students beginning to do research.

The forty-two chapters are grouped into six sections: work on Indo-European in general; Greek, the largest group with 18 chapters; Anatolian; the western languages including Latin; Indo-Iranian and Tocharian; and two papers on the history of Indo-European linguistics. The editor’s preface and two notes from the Press give a very brief overview of Morpurgo Davies’ career, fleshed out by the bibliography of her major publications from 1958 to 2001. There is an index of words discussed, but no index of passages, which would have been convenient. Each chapter has its own references, rather than a single unified bibliography for the volume. And footnotes are conveniently located at the bottoms of pages.

The chapters are as follows.

Part I on IE (Indo-European) in general contains five papers. First, Paolo Di Giovine discusses “Il perfetto indoeuropeo tra endomorfismo ed esomorfismo,” showing how the forms of perfect tenses differ in Indo-Iranian, Greek, Latin and later Romance, Germanic, and Celtic.

George E. Dunkel’s “Particles and Personal Pronouns: Inclusive *me and Exclusive *we” argues that these two particles, which seem to appear as suffixes in certain pronominal forms, are actually independent particles that appear in other compounds as well. The analysis here is as elegant as the language is polemical.

In “Etymology and History: For a Study of ‘Medical Language’ in Indo-European,” D. R. Langslow outlines a proposal for future work, beginning from the shared vocabulary (including the root of the word ‘medicine’ itself) and the studies that have been done on medical terminology in the various IE languages. Langslow’s bibliography is particularly extensive (4 pages) and will be of interest to scholars in history of medicine as well as linguistics.

The late C. J. Ruijgh offers “The Stative value of the PIE verbal suffix *-eh ₁ –.” This suffix is to be distinguished from the Greek aorist passive suffix, which Ruijgh argues was originally ‐ης‐. If this is correct, arguments that *-eh ₁ – is fientive (denotes “becoming”) become harder to defend. The chapter is carefully argued.

“The Third Donkey: Origin Legends and Some Hidden Indo-European Themes,” by Calvert Watkins, connects the legend of the daughters of Danaus, as dramatized by Aeschylus in his Suppliants, with similar stories in the Indic and Hittite traditions, RV 10.86 and the Hittite Zalpa tale. Watkins further argues that this myth is “intimately linked” (p. 65) with the IE horse sacrifice ritual. All three stories feature multiple births and illicit sex, either incestuous or bestial. In the Indic and Hittite versions wild donkeys feature as symbols for male sexuality, and a donkey or horse is sacrificed. But there is no trace of the horse sacrifice in Greek; where, then, is the third donkey? Watkins observes that Pindar Pythian 10.31-36 describes the sacrifice of sexually aroused donkeys, and suggests that the two parts of the original story have become detached from each other in Greek.

Part II treats Greek, from Mycenean to Modern. A. C. Cassio reviews a Homeric crux in “Spoken Language and Written Text: The Case of ἀλλοειδέα (Hom. Od. 13.194).” This line does not scan in the form in which we now read it. Cassio argues that when this line was composed, the word must have been pronounced “alleydya,” in three syllables, as if it were spelled ἀλλειδέα. In that case, the spelling has preserved the etymological form of the compound.

Stephen Colvin’s “Social Dialect in Attica” explores the linguistic implications of an ostracon which reads ὀστρακίδο instead of ὀστρακίζο. The ostracon dates from the early 5th century (so omega and omicron are not distinguished) and shows that someone in Attica at that time pronounced the zeta as a delta. This is a feature of the Boeotian dialect. Colvin argues that the zeta/delta isogloss does not precisely match the Attic/Boeotian border: some variety of Attic shared this feature with Boeotian. This variety would be lower-class and probably stigmatized. The chapter is a nice contribution to the sociolinguistics of ancient Greek, the first of several in IEP.

Emilio Crespo’s topic is also sociolinguistic, “The Attitude of the Athenian State towards the Attic Dialect in the Classical Era.” Crespo observes that in 5th-century inscriptions documenting oaths to be sworn by the Athenians and another polis, the words of the oaths are given in Attic dialect. This is true even in the few cases where we have a non-Attic copy of the inscription. On the other hand, it is not universally true for 4th-century inscriptions. Crespo suggests that the 5th-century treaties show Athens imposing its own dialect on other states and further suggests that this is “decisive for the creation and expansion of the Koine.” (p. 117)

In “Rules without Reasons? Words for Children in Papyrus Letters,” Eleanor Dickey collects all the addresses and references to children in letters from the 3rd century BC through the 3rd century AD. These are often phrases of the form “say hello to your children for me,” though some letters are addressed to the writer’s son or children and are headed with formulae like “X to his son Y, greetings.” Dickey observes that the vocative form is τέκνον, while the heading of a letter addressed to a child will be υἱῷ or θυγατρί as appropriate. The words to refer to a child in the third person are υἱός or θυγάτηρ in the singular, but παιδία and τέκνα in the plural, with no obvious distinction in meaning. In the classical period, however, τέκνον is poetic and υἱός is relatively rare. Clearly something has changed. Dickey draws no conclusions from her analysis but ends her chapter with the hope that these “peculiar patterns” (p. 129) can be explained somehow.

“Langage de femmes et d’hommes en grec ancien: l’exemple de Lysistrata,” by Yves Duhoux, shows that male and female characters in Lysistrata speak differently. They use different words, swear by different gods, select different particles, and choose different grammatical constructions. Duhoux suggests that a similar study could be done for the other plays of Aristophanes with prominent female characters, and perhaps also for Euripides. He also suggests, tentatively, that the difference Aristophanes makes in Lysistrata may reflect real differences in Athenian speech, though this is much harder to prove.

Ivo Hajnal wonders: “Die Tmesis bei Homer und auf den mykenischen Linear B-Tafeln: ein chronologisches Paradox?” He concludes that the main reason tmesis is not attested in Linear B is stylistic, and most instances of tmesis in Homer are innovations of the poetic language. Tmesis was available to writers of Mycenean Greek (that is, the relevant adverbs had not yet become permanently attached as verb prefixes) but they chose not to use it.

Henry Hoenigswald, who died before the book appeared, contributed Ἑλλήσποντος, in which he determines that the proper name must be one word, not two ( Ἕλλης πόντος), in Homer. The name always appears at line end, so if it were two words, it would violate Meister’s bridge, which forbids word end after contracted fifth biceps.

“Aspect and Verbs of Movement in the History of Greek: Why Pericles Could ‘Walk into Town’ but Karamanlis Could Not,” by Geoffrey Horrocks, gives us a brief look at Modern Greek. The resources available to Pericles that have since disappeared are the dative case, which gives Ancient Greek two kinds of prepositional phrases, one for place where and one for place to which; and the use of verb prefixes, to produce verbs that are nearly synonyms but have different intrinsic Aktionsart. As a result, Modern Greek cannot directly express the goal of motion. Horrocks acknowledges the contribution of Melita Stavrou to some of the work in this chapter, notably the discussion of aspect.

Joshua T. Katz, in “The ‘Swimming Duck’ in Greek and Hittite,” gives an etymology for Greek νῆττα, arguing that it is in fact cognate with Latin anas, but distorted by the influence of the verb νέω, ‘swim.’ Although this idea has been suggested before, Katz fleshes out the details and suggests that the Hittite word lahanza,¹ which most people have agreed must refer to some kind of bird, means ‘duck’ and comes from the same postulated pre-form as the Greek word. The argument is surprisingly complex but, I think, ultimately convincing.

In “Names in -e and -e-u in Mycenean Greek,” John Killen argues that we do not have pairs of personal names with endings in ‐ης and ‐ευς. In some cases the ‐ευς names are nicknames or short versions of compound names. E. Risch has argued that the ‐ης names are, too, and that a single long personal name may generate two different nicknames. Killen disagrees. He points out that many of the ‐ης names are from Knossos and are probably not of Greek origin, and that some of the other words that Risch interpreted as names are now agreed to be common nouns. It is even possible that some of the doublets are the same name in two forms, one the original non-Greek form and the other with a more Greek-looking ending.

The title of Charles de Lamberterie’s paper is confusing at first. ” Sella, subsellium, meretrix : sonantes-voyelles et ‘effet Saussure’ en grec ancien” refers to the Greek words for sella and meretrix, not the Latin words themselves. The ‘chair’ words in question are θρόνος and θρῆνυς and the problem is their etymology. De Lamberterie, returning to an observation of Saussure, argues that both come from an original * d ^h rh ₂. θρόνος then develops by metathesis from the o-grade form * θόρνος, and θρῆνυς comes from a zero-grade form with suffix -nu. The derivation of πόρνη from the root * perh ₂ meaning ‘sell’ is similar.

Michael Meier-Brügger’s “Zu griechisch τυρός‘Käse'” is a short, clear demonstration that in Mycenean the word for ‘cheese’ must have been an i-stem, not the o-stem it becomes later. That is, the attested forms are not diminutives but i-stem forms.

Torsten Meissner addresses “Two Mycenean Problems,” namely re-ke-to-ro-te-ri-jo and po-ne-to. Each of these words attests spelling variation that, according to Meissner, cannot simply be chalked up to the difficulties of the writing system for Mycenean Greek. In the first case, the name of a festival ( λεχεστρωτήριον), we have re-ke-e-to-ro-te-ri-jo in addition to the shorter form cited above. Meissner argues that the shorter form is the regular one, showing the normal form of λέχος in composition and the form with the additional -e- seems to be remodelled based on the form of the noun within Mycenean. The verb po-ne-to appears to be a contracted thematic form, with long e from the contraction of two short e. Other verbs among the relatively few attested forms seem to vary between thematic and athematic forms. Meissner points out that in Aeolic and some other dialects, verbs with stems ending in vowels, including many denominatives, have athematic forms, while in Attic (and others), they have thematic forms. If the paradigm was “split” in Mycenean between thematic and athematic forms, this could account for the diverging development, as each dialect chose different forms to generalize from. A contract form like po-ne-to would be a step along the way.

“On Some Greek nt- Formations,” by Martin Peters, considers a group of nouns which he argues are related to participles with the suffix -nt-. In particular, the hapax εγκυαρ in an inscription from Miletus (Del. ³ 725.6) is certainly somehow related to ἔγκυος, ‘pregnant.’ Suppose there was a derived adjective with the -nt- suffix, perhaps * enkuwont-, and suppose its neuter nominative plural, with zero grade of the suffix, came to be used as a noun. This would be * enkuwata, which could then be re-interpreted as belonging to an r/n-stem noun, from which a new nominative singular could be derived, the attested εγκυαρ. Peters suggests that ὑφέαρ‘mistletoe’ was formed in just the same way from ἐπιφύομαι or, more precisely, its Arcadian variant * opi-p ^h uwat-.

In “Accentuation in Old Attic, Later Attic, and Attic,” Philomen Probert looks at the 2nd-century AD grammarian Herodian’s references to “earlier” and “later” forms or “Attic” and “non-Attic” forms. She concludes that for Herodian “old Attic,” “later Attic,” and generic, ordinary, or Koine Greek are three different varieties; in particular, what he refers to as “later Attic” is not the same as Koine, since there are classes of words which these two varieties accent differently. This is then one more way in which Koine is more than just a development of Attic Greek.

Peter Schrijver compares Greek and Celtic words, a relatively unusual pairing, in his “Indo-European * (s)mer- in Greek and Celtic.” This root stands behind Greek words like μόρος, μοῖρα, and μέρος, all referring (more or less) to fate. Although Latin mereor is related, up to now no one has found forms from this root in other branches. Schrijver observes that there are cognates in Irish and in the Brittonic languages. He interprets Old Irish mart or mairt as “prognosticated or impending death, death fate” (p. 294). Despite the gloss given in the Yellow Book of Lecan, this word cannot be related to Latin mors : if it were a loan-word from Latin it would be * moirt, and its meaning is closer to ‘fate’ than to ‘death.’ The cognates in Welsh, Cornish, and Breton also have meanings like “premature death,” “unnatural action with fateful consequences,” and “miracle.” Hence, he concludes, * smer- is common IE.

In ” χαῖρε καὶ πίει εὖ (AVI 2),” Rudolf Wachter analyzes the corpus of so-called “little-master cups” and their inscriptions, most of which are variations of the formula of his title, many with an explicit accusative pronoun. He observes that πίει must be second person singular, future indicative middle of πίομαι, “you will drink,” not any kind of imperative or jussive subjunctive. Then χαῖρε, which is of course imperative, has the idiomatic meaning “take with pleasure” or “take and enjoy,” referring to the cup itself. But many of the cups with this formula are quite small, smaller than one would expect for a cup passed around the table at a symposium. Wachter suggests that these smaller cups are love-gifts, perhaps the first wine-cup for a young ἐρώμενος. This chapter is a nice use of both archaeological and linguistic evidence.

The last chapter in the Greek section is Andreas Willi’s “Flowing Riches: Greek ἄφενος and Indo-European Streams.” Willi reconstructs a secondary r/n-stem noun * H ₂ bh-r-, based on the known root * H ₂ ebh- meaning ‘river, stream.’ The noun ἄφενος‘wealth’ would be derived from an s-stem noun formed from the “weak” case stem of this new noun. The old Greek adverb ἄφαρ could continue the “strong” stem of the noun. “Wealth” would then be the property that flows to one. The argument is neat.

Part III is made up of five papers on Anatolian topics. The late Gillian Hart discusses “Some Problems in Anatolian Phonology and Etymology,” specifically a group of words that have sibilants in Hittite but dental stops in Luwian. She suggests that this may have been a result of sound changes affecting initial * h ₃ – when u or w is in the environment.

J. D. Hawkins, in “The Stag-God of the Countryside and Related Problems,” fills in the details of the relationship between and interpretation of three Hieroglyphic Hittite signs, numbers 461-463 in Laroche’s catalog Les Hiéroglyphes hittites (Paris: 1960). In particular, L.461 and L.463(1-2) are earlier forms, and L.462 and L.463(3), respectively, evolved from them. There are thus really two signs here, not three. L.463(1-2) denotes the stag-god, and L.462 can be provisionally taken as the syllabogram ma _x.

“A Luwian Dedication,” by H. Craig Melchert, distinguishes the two Luwian verbs uppa- and upa- (the first has a long ‘u’ and the second has a short one). Although it has been clear for some time that these roots must be distinct in Cuneiform Luwian, meaning respectively ‘bring’ and something like ‘grant’ or ‘furnish,’ it is harder to tell them apart in Hieroglyphic Luwian, where unless there is an associated determiner, the two verbs will be spelled the same way. Melchert argues that they are distinct and derives the second one from IE * bhag-‘apportion, allocate.’

Norbert Oettinger’s “Das Wort für ‘Jahr’ und hieroglyphen-luwisch yari-‘sich ausdehnen'” derives this word from * h ₁ yeh ₁, ultimately the same source as English ‘year,’ because a year is a long extent of time.

In “Dal nome comune al nome divino, proprio e locale: il caso di tasku- in anatolico,” Massimo Poetto argues that this word, a theonym, must originally mean a bag or sack, related to Italian tasca or German Tasche. From there it develops to refer to the scrotum, then to virility in general, making it an appropriate name for a god.

Part IV is a grab-bag of western IE languages: Italic, Old English, and Celtic. James Clackson argues, in “The Word-Order Pattern magna cum laude in Latin and Sabellian,” that although we see “interposed order of adpositional placement” (p. 402) in both of these languages, its origin is not the same. In Latin it comes from fronting; in Sabellian, from loss of a postposition following the noun.

Jay H. Jasanoff’s chapter, ” Plus ça change : Lachmann’s Law in Latin,” considers various explanations for the apparent rule in Latin that verb roots originally ending in a voiced non-aspirate stop have a long vowel in the past passive participle. For example, ago has a long ‘a’ in the participle actus. The rule does not always hold, though: verbs like sedeo ( sessus) and stringo ( strictus) have short vowels in their participles. Jasanoff argues that the old Neogrammarian explanation is the right one: first the consonant in the participle is voiced by analogy with the perfect active, then the vowel is lengthened (and the consonant devoiced) by a regular sound law. The main point of the chapter, however, is a discussion of how several schools of linguistic thought have explained this quasi-rule over the last forty years. The essay is a good case study on the dangers of “extreme positions” (p. 415) in linguistic analysis.

Don Ringe’s “Old English mathelian, maethlan, maelan” is the first of two papers on Old English. Here Ringe argues that these three verbs are actually only two. The first is a denominative formation within Old English; the third is a derivative from the second, which is inherited from Proto-Germanic. The key piece of evidence is the restricted metrical environment of mathelian in verse: only its preterite third singular appears ( mathelode), and only as the end of the first half-line of a verse. Ringe concludes that this must have come about when poets attempted to preserve a formula originally using the older verb after sound changes had made the form unmetrical in the poets’ dialect.

“I nomi delle figure dei miti greci nelle lingue dell’ Italia arcaica. The First Traces of Achilles and Hercules in Latin,” by Helmut Rix, begins in Italian and moves to English. Rix argues that the names of Achilles and Heracles came directly into Latin from Greek, not via Etruscan, and are attested as early as the 5th century. Etruscan did influence the forms of the names, but only later. The earliest attestations are on engraved gems, which are older than the better-known inscribed mirrors and vases. The gem inscriptions are often simply taken as Etruscan, but as Rix points out, some of them are unambiguously in Latin.

Next comes the one Celtic chapter, Paul Russell’s “Old Welsh Dinacat, Cunedag, Tutagual : Fossilized Phonology in Brittonic Personal Names.” These proper names show archaic features, notably the vowel between the two parts of the compound, but Russell argues that they are revivals not survivals, taken from the equivalent Latinized names.

The second Old English contribution is Patrick V. Stiles’ “Consumer Issues: Beowulf 3115a and Germanic ‘Bison.'” This line in Beowulf includes a verb generally given as weaxan, ‘grow,’ but in context this makes no sense. What’s required is a verb meaning ‘consume,’ and Stiles revives an old conjecture of Ferdinand Holthausen, wesan or weosan. Finite forms of this verb are not attested elsewhere in Old English, but it has cognates in Gothic and Old High German, as Holthausen saw. In the 75 years or so since Holthausen proposed this, we have much more information about the root, which has descendants in Hittite, Old Iranian, Celtic, and Tocharian as well as Germanic. Stiles also connects to this root another word whose current etymology is unsatisfying: ‘bison,’ usually referred to * weis-, ‘stink.’ As Stiles points out, smell is not especially characteristic of bison, but eating may be. It therefore makes sense to derive the noun from the participle wesand, ‘the big eater.’

Jürgen Untermann considers “Die hispanische Heerschau des Silius Italicus.” He shows that in his catalog of the Spanish armies in book 3, Silius gives the names Greek forms, even though he could have had access to the correct forms of the names, for example from Livy. Accuracy seems not to be a concern of this poet.

Although Part V is called “Indo-Iranian and Tocharian,” only one of the papers is on Tocharian. First, however, is “On Vedic Suppletion: dash and vidh,” by José Luis García Ramón.² Here the question is how to recognize suppletion. The two verbs in question both mean ‘honor’ or ‘worship,’ usually with two arguments, the god and the offering or sacrifice. While most Vedic verbs of sacrifice, offering, or the like take the recipient in the dative and the offering in the accusative (“present offering to god”), or the recipient in the accusative and the offering in the instrumental (“honor god by means of offering”), these two take a dative and an instrumental (“give honor to god by means of offering”). One of these roots exists only in the present and perfect, while the other is primarily a thematic aorist. García Ramón concludes that these two roots form a suppletive paradigm. Their unusual construction, moreover, developed as the two verbs influenced each other.

In “Tocharian B päst and its Vocalism,” J. H. W. Penney shows that the form päst must be derived from pest, not the other way around as has been thought. The word is an adverb meaning ‘away.’ It is usually assumed, based on what we know of the phonology of Tocharian B, that päst must be an unstressed form, but Penney questions this based on word order. It is still unclear why this form, with a vowel that elsewhere appears only in unstressed syllables, would have evolved as it did.

Given my own interest in meter, one of the first papers I turned to was Rüdiger Schmitt’s “Promising Perspective or Dead End? The Issue of Metrical Passages in the Old Persian Inscriptions.” The answer, alas, is “dead end.” Schmitt finds no evidence whatsoever for metrical passages in the Achaemenid royal inscriptions — not quantitative, not stress-based, not syllable-counting. Although for many years people have been trying to scan these texts, Schmitt shows that most previous attempts have involved fairly arbitrary contractions, expansions, and other distortions. There is some ornamental language in the inscriptions but that is not enough to turn them into metrical verse. Disappointing as the result is, the chapter is well argued, with some nice observations.

Nicholas Sims-Williams analyzes “The Parthian Abstract Suffix -yft.” This suffix in Manichaean Parthian is equivalent to, and presumably cognate with, Middle Persian -îh and New Persian -î, though the details of its derivation have been unclear. Sims-Williams gives an etymology, showing how the original Old Persian suffix developed in the various dialects of Middle Persian.

Elizabeth Tucker’s chapter “Denominative Verbs in Avestan: Derivatives from Thematic Stems” compares the evolution of these verbs in Avestan with their counterparts in Vedic. In both cases the verbs are built from nominal stems with the suffix * -yá- (in Indo-Iranian; * -yé-, -yó- originally in IE). There are two classes of denominatives in Avestan, those that lose the vowel before this suffix and those that retain it. In Vedic, on the other hand, this vowel always remains, giving the -ayá- verbs. Tucker suggests that the Avestan verbs that have lost this thematic vowel are the original type, and the -aiia- denominatives developed later from a re-interpretation of the (different) inherited suffix * -áya-. This re-interpretation could only have happened after the inherited accent was changed. As Tucker observes, the development of * -éye- verbs and * -yé- denominatives in Greek is similar, and similarly related to changes in the accent system.

Finally, Part VI comprises two papers on the history of IE linguistics. “The Celtic Studies of Lorenzo Hervás in the Context of the Linguistics of his Time,” by Javier de Hoz, presents this late 18th-c. scholar as a curious mix of modern-sounding ideas and unexamined theories. Hervás, a Jesuit priest, believed in the story of the Tower of Babel, which led to various problems in his conception of linguistic history. On the other hand, he was one of the first to list and classify the Celtic languages, almost entirely correctly ascertaining the relationships within the group. He resisted the then-common idea that Basque and Celtic were related, though more for nationalist reasons than linguistic ones. Much more work can be done here; de Hoz suggests in particular that the relationship of Hervás to Leibniz, and Hervás’s history of linguistics, would both repay further study (p. 567).

The forty-second and last paper in the volume is “Johannes Schmidt’s Academic Career and his Letters to August Schleicher,” by Klaus Strunk. These newly discovered letters fill in details of Schmidt’s various rejections on the way to ultimate success. The letters appear to be a study in academic politics, with linguistic observations on the side. Strunk summarizes the main ideas but does not quote extensively. He asks “whether an edition of the collection of letters … should be undertaken” (p. 582); this chapter is a good advertisement for such a project.

In sum, virtually anyone with an interest in IE linguistics will find something useful or exciting here. Some papers solve problems, others pose new ones; this makes the book particularly valuable for students who want to find out how the field works. But even the most erudite specialist — for example, Anna Morpurgo Davies herself — will find the book worth reading.

Notes

1. The standard diacritical mark for ‘h’ in transcribed Hittite words is not available in this medium.

2. Once again an issue with diacriticals; the first of these words has long A and the palatal sibilant normally romanized as s + acute accent.