Indo-European Matters Even More

In this paper I am going to present an attempt at a solution to four of the most difficult individual problems in Slavic historical morphology:

1. CS o-stem masc. nom. sg. *-ъ
2. CS ā-stem gen. sg. *-y
3. CS ā-stem nom. pl. *-y
4. Old North Russian masc. nom. sg. -e

The solution I am proposing requires clarification of several other problems, including the following:

1. Old North Russian ā-stem gen. sg. -ě
2. Old North Russian ā-stem nom. pl. -ě
3. CS o-stem masc. nom. pl. *-i
4. CS e/o-vb. ipv. 2/3 sg. *-i

Most of what I say has been said before. What is new in my proposal is that I subsume individual solutions to several seemingly individual problems under one pre-Proto-Slavic phonetic change (R = resonant, S = fricative):

*ā̆ > Proto-Slavic *ə̄̆ /_ (R)S

The distribution of phenomena such as corruption, elision, and hiatus in Homer can provide valuable insights into their history within the epic tradition. For example, the correction of long diphthongs appears to be a secondary development, resulting in particular from the modification of preexisting formulae.

The distribution of Homeric diphthong elision is equally revealing. It shows some specific restrictions: only short αι and οι can be elided, and all the examples are accentually light, but nominative plurals in -αι and -οι are never elided, and while infinitives in -σθαι are often elided, infinitives in -σαι never are

From an examination of the entire corpus of elided diphthongs in Homer, it becomes clear that the key to this distribution is the restriction of diphthong elision to speech. Everything else, including the phonologically inexplicable restriction to certain morphological contexts, follows directly from this.

In this way, it is possible to prove that diphthong elision is not a metrical device, nor a reflection of any phonological differences, but rather a technique of literary characterisation.

This conclusion is reinforced by an examination of the distribution of diphthong elision in later authors; the different distribution patterns in these various authors can be directly linked to differences in genre and poetic structure. In this way, we can trace the evolution of diphthong elision in Greek verse from its origins within epic to its differing fates in later poetry.

This paper will deal with the origin and dispersal in the Slavic languages of compound agent nouns with a verbal first member. English ‘pickpocket’ sums up the basic characteristics of such a compound: it is made up of a verbal first member and a second member that typically refers to the object of that verb. The compound itself refers to the agent of the verbal act.

The Slavic representatives of such compounds will be discussed from two angles: a morphological one and a historical one.

In terms of morphology, these compounds have been topical for discussions of compositional morphology since the earliest times. Scholars have noticed a conspicuous number of cases in which their first members are formally identical to 2sg. imperatives; thus e.g. Rus. Vladiměr, literally ‘rule-world (a name)’ and vertigolova, literally ‘turn-head → heedless’. The Slavic forms have been interpreted as evidence of a general tendency to form compounds based on imperative clauses, and this has influenced the way scholars have analyzed the ‘pickpocket’ compounds of other branches (Italian batticarne, Greek φερέοικος etc.).

To the morphologist, it is somewhat unsatisfying to have to assume derivation from an inflected form, and a closer inspection of the Slavic material reveals that the imperative analysis is less than compelling.

Compounds of the ‘pickpocket’ type are found in most European languages, and often in similar semantic spheres, and they are generally thought to have been prolific in Proto-Indo-European. From a historical perspective, the Slavic representatives are highly revealing. On the one hand, such compounds are typical of the onomastic vocabulary in all branches of Slavic; thus e.g. Rus. Vladiměr. In this respect, Slavic can be grouped with Baltic, Germanic and Greek in the sense that they preserve an ancient tradition of double names. On the other hand, we find ‘pickpocket’ compounds in a somewhat baser use, namely as highly expressive agent-nouns typically denoting womanizers, thieves and misers; Rus. vertigolova ‘heedless’ is among the more neutral examples. That type is particularly typical of West Slavic languages; it shares its basic properties with similar formations in Romance and its immediate neighbours, more specifically Middle English and Middle High German.

In short, the Slavic ‘pickpocket’ compounds mark two isoglosses: an ancient one harking back to Proto-indo-European and a more recent one that was established in Medieval Europe.

I have come to find myself in agreement with Dirk Boutkan’s ideas about the development of vowels in Germanic final syllables in most regards, especially his and the Leiden school’s fundamental idea that the presence or absence of final consonants would have had more influence on the Germanic vocalism than tonal differences, as presented by Scherer, Hirt, and Streitberg, among others.

At one point, however, I disagree with Boutkan, namely on the development of the diphthongs PIE *ei̯, *eu̯, *ēi̯, and *ēu̯ in final syllables and, as a result of that, on the Proto-Indo-European shape and origin of some of the endings of the Germanic i- and u-stems. To avoid accepting o-grade endings like PIE *-ou̯s (gen. sg. of u-stems), I suggest that:

• PIE *ei̯ > Germ. *ai̯ in final syllables
• PIE *eu̯ > Germ. *au̯ in final syllables
• PIE *ēi̯ > Germ. *ei̯ > *ii̯ > *ī in final syllables
• PIE *ēu̯ > Germ. *eu̯ > *iu̯ in final syllables

Counterexamples do exist (e.g. Proto-Norse gen.sg. *-īz, which is more likely to be from PIE *-ei̯s than **-ēi̯s), but these may be explained in other ways, such as I intend to show in the presentation.

I do not wish to claim that the present solution is the only possible and realistic way to explain the inflexion forms in question, but I believe that it has at least two advantages:

• avoidance of IE *-oi- and *-ou- forms in the i- and u-stems
• no difference between ‘acute’ and ‘circumflex’ diphthongs in Proto-Germanic Auslaut (admittedly, this goes for Boutkan’s and many others’ solutions, as well)

The old PIE s-stem *nébʰ-os- ‘sky’, which appears as Hitt. nepiš, nepišaš, Gr. νέφος, -ους, Skt. nábhaḥ, nábhasaḥ, OCS nebo, -ese, etc., is generally reconstructed with a genitive *nebʰ-es-(o)s. However, the linguistic material offers evidence for a zero-grade root *m̥bʰ-: Arm. amp, amb ‘cloud’, for instance, can, given the general transfer of the s-stems to o-stems in Armenian, very well represent *m̥bʰ-os-, the zero-grade form of *nébʰ-os-. It is therefore likely that the original Indo-European paradigm must be reconstructed as *nébʰ-os, *m̥bʰ-és-(o)s.

The reconstruction of *nébʰ-os and *m̥bʰ-és-os neatly accounts for some other formal difficulties in the IE daughter languages. I would like to argue that Lat. nimbus ‘cloud’ arose as a contamination form of the nominative with the genitive root, i.e. *nembʰ-os- or *nm̥bʰ-os-. It is further possible that Lat. imber ‘rain, storm, storm cloud’, which is usually taken to be derived from *imbri-, directly continues the oblique stem *m̥bʰ-és-.

As an open-ended question, I will furthermore investigate the possibility of reconstructing an even more archaic paradigm *dʰnébʰ-os, *dʰm̥bʰ-és-os given the problematic initial dentals of Lith. debesìs, Luv. tappaš ‘sky’. This paradigm also seems to find support in some Germanic forms, cf. OSw. dimba ‘fog’.

It is universally assumed that, with one exception (Breton kaezh, pl. keizh ‘miserable’), adjectives did not inflect in number in Breton and Cornish. In this talk, I will point to a hitherto overlooked instance of plural inflexion, occurring in the adjective bihan, byhan ‘little’. This adjective provides us with a full range of examples from different syntactic contexts, showing that Middle Breton and Cornish adjectival morphology closely matched that of Old Breton and Middle Welsh. The result of the investigation will also shed new light on British Celtic–Latin language contact, as well as provide a precise typological parallel to adjectival inflexion in the Scandinavian languages.

While Hitt. armae-zi ‘become pregnant’ is generally explained as a derivative of arma- ‘moon; moon-god; month’, the origin of arma- itself (and its Anatolian cognates) has remained enigmatic. Zeilfelder (1998) has justifiably rejected any connection with Hitt. erman- ‘sick’, as did Hilmarsson (1989) with Toch. B yarm A yärm ‘measure’.

In this paper, I will argue the superficial similarity with Lat. arma pl. ‘weaponry’ and Gk. ἅρμα ‘wagon’ (possibly with secondary aspiration, cf. Myc. a-mo ‘wheel’, not †a₂-mo, and Delph. ἄρμα ‘coitus’) is no coincidence, and that they all go back to PIE *ar-mo- ‘fulfilment (of a phase or a rite); outfit’. Several IE languages possess words with meanings that cover both ‘pregnancy’, ‘wagon equipment’ and ‘weaponry’ because they refer to rites of passage for young soldiers-to-be and maturing girls alike.

The various meanings of Lith. šarvaĩ, šárvas, šar̃vas ‘(1a) discharge after birth, placenta; (1b) menstruation; (1c) discharge from the mouth of the dead’, ‘(2a) armament; (2b) soldier’s outfit, weapons and ammunition; (2c) carapace’, and ‘(3a) dowry; (3b) burial object; (3c) kind of wagon, bottom board of wagon’ have been satisfactorily explained by Gliwa (2005) as from an earlier meaning *‘physical sign and ritual equipment in a rite of passage'.

One further meaning (4) of šárvas, however, is not mentioned by Gliwa: In Brodowski’s Lexicon Lituanicum (before 1744), the word is glossed as ‘December’. Formally, it corresponds to the second member of Cretan Διοσ-κούρος ‘sixth month of the year’, cf. the Cretan Κουρῆτες ‘young men in their capacity as warriors, guardians of the infant Zeus’, and κορή ‘young girl’. Remarkably, the Roman name of the other solstice month, mēnsis Jūnius, is derived from the name of Jūnō < *‘the new moon’ (Fortson 2005), originally a moon-goddess and the protecting goddess of lying-in women, derived from the same base as Lat. iuvenes, Umbr. iouies ‘young men of military age; warriors’, SCr. jùnâk ‘soldier’, Lith. jaunóji ‘bride’, jaunáitis ‘new moon’.

Uighur aram, Hunno-Bulg. alem ‘first month of the year’ have no counterparts in other Turkic languages and maybe IE borrowings, while a possible Uralic cognate is PFU *ärV-mV ‘year; autumn’ (Udmurt arm, Mansi oarem) < ‘ripening, fulfilment; (new) season’.

The ritual meaning is also reflected in other derivatives such as Lyc. ara ‘rite’, Ru. radu-nitsa ‘first Sunday after the first full moon on or following the vernal equinox’ and even Lat. rītus ‘rite’. The semantics of *ar- must have been about the same as in Komi ještõ- ‘be accomplished, be fulfilled; be in time (at arrival); be mature/ripe’ (< PIE *Hei̯k̂- ‘to have in one’s power’).