The 4th millennium BC stands out as a period of increasing interaction between the Caucasus, Anatolia, the Levant and Greece, stimulated by movements of groups of people at land and sea, including the Black Sea coast (Bauer 2011), which had both genetic (Damgaard et al. 2018; Lazaridis et al. 2017; Wu et al. 2018), cultural and linguistic consequences, including Anatolian, which split off during the early to mid 4th millennium BC from early Maykop groups in the northern Caucasus. After the middle of the 4th millennium steppe Maykop expanded north, leading to the formation of the Yamnaya Culture and Proto-Indo-European, which by the beginning of the 3rd millennium saw the development of ancient Tocharian and the first migrations towards the east (Altai) and the west (Europe). Thus, for reasons given below, I argue for the “Indo-Hittite” hypothesis, using “Proto-Indo-Anatolian” for the source of both (Proto-)Anatolian and the rest of the Indo-European languages, reserving “Proto-Indo-European” for the source of the non-Anatolian languages.
Cladistic hypotheses are ideally based on arguments that use cumulative evidence from a wide range of shared innovations inherited from a more recent ancestor. The majority of historical linguists would agree that the best evidence for subgrouping would be shared phonological and morphological innovations, while evidence from proportions of shared lexical cognacy is less reliable for linguistic subgrouping. Recent high-profile studies have appeared, however, that have been based exclusively on comparative lexical material. The results of these methods have been sharply criticised, but in spite of the criticisms to cognacy-based approaches, there remains some potential that the lexical cognacy may provide some useful data to supplement cladistic hypotheses as part of an overall assessment of the complete bundle of available isoglosses. If lexical cognacy judgements can be treated as a potential source of data for cladistic hypotheses, how can they be implemented in a methodologically rigorous way? This chapter focuses on case studies from methodological issues that have arisen in encoding Indo-European lexical cognacy data on the Indo-European Cognate Relationships (IE-CoR) database project based at the Max Planck Institute for the Science of Human History. These issues are illustrated through case studies from problems that have arisen in assigning cognacy coding to lexical data. As such this chapter contributes a discussion towards improving the reliability of cognacy data for cladistic analyses as a supplement to more traditional analyses based on comparative phonological and morphological criteria.
Hittite šeppit(t)- ‘(a kind of) wheat’ can be identified as a borrowing from Akkadian or another Semitic language, identical to the Mediterranean culture-word that has ended up in English as semnel. There is thus no basis for a PIE reconstruction *sép-it meaning ‘wheat’. The very suffix *-it, which Watkins (1978) interpreted as a marker for ‘(basic) foodstuffs’, is likewise a phantom. Balkan Indo-European *álbʰ-it and Iranian *arpucya-, both ‘barley’, are borrowings from separate dialectal reflexes of Turkic *arpa ‘barley’; the former acquired its suffix in analogy with *mél-it ‘honey’ because the two terms were interpreted as *albʰ- ‘white’ and *mel- ‘dark’, respectively—an important distinction in ritual contexts.
The derivation of Hittite ḫandā(i)- from *h₂ento- ‘weaving, fabric’ vel sim. (Ziegler 2014) is highly illuminating for the Hittite verb, given evidence presented below that the original meaning was ‘to align, make straight’. The sense ‘to warp, begin to weave’, like that in Latin ōrdior which Ziegler properly compares, is derived from ‘to align the threads of the warp’. Both cases provide further examples taken from weaving for a widespread IE metaphor ‘(morally) right’ < *‘made straight’ (seen also in derivatives of *h₃reg̑- ‘to make straight’), which competed with another based on *‘fitting’ from verbs meaning to fit together, *h₁ar- ‘to fit’ (intr.), *tek̑- ‘to fit together, join’ (tr.) (Melchert forthcoming), and *(hx)reith₂- ‘to join, mix’ (tr.) (Weiss 2015). Which if any of these extended uses dates from the proto-language requires further study.
After some theoretical discussion of the question of lexical matches in related but non-contiguous languages illustrated by examples from Romance, this paper re-examines the well-known phenomenon of lexical matches shared between Italic, Celtic, and Indo-Iranian. In particular, I examine the case of *k̑red-s dʰeh₁- ‘put heart’ > ‘trust’, which is continued in Latin, Insular Celtic, and Indo-Iranian. I show that there is no reason to deny the connection between *k̑red-s and *k̑(e)rd- ‘heart’. Most probably the Schwebeablaut is phonotactically motivated. If the first element of the collocation was indeed an s-stem, it may have been modeled on the related idiom *mens dʰeh₁- ‘put mind’ > ‘think’. Given that the meaning ‘think’ is not attested for the root *men- in Anatolian or Tocharian, it is possible that *mens dʰeh₁-, and consequently *k̑red-s dʰeh₁-, are relatively recent creations. The paper then examines the semantic history of the idiom in Italic, Celtic and Indo-Iranian. Contrary to the usual view, the item is not mainly a religious term but refers to the social phenomenon which I call the “credit act”, i.e. putting oneself or one’s property in the hands or power of another with the expectation that the other individual would give good in return. Since the participants in the credit act were often of unequal power (patron ~ client), the credit act concept was transferred to the unequal relationship between gods and humans.
It is well-known that the present tense of the verb *es- ‘to be’ in the Italic languages shows a mixture of what look as if they were thematic forms (e.g. Old Latin 1 sg. es-om) beside athematic forms (e.g. Latin 3sg. *es-t). A similar state of affairs is attested in the Celtic languages. Within the broader perspective of Indo-European, the thematic forms are puzzling, and efforts have been undertaken to explain them away as secondary. I argue that those efforts have not been successful. By combining the rich but complicated evidence provided by the Celtic languages with the Italic data, it becomes necessary to reconstruct a thematic beside an athematic present of *es- for Italo-Celtic and to hypothesize that the thematic forms were originally used after a focused constituent.
The Hittite myth of Telipinu, the Greek myth of Demeter and Persephone, the Norse myth of Baldr and the Indic myths of Cyavana reflect an inherited Proto-Indo-European mythical theme about “Non-Functioning Fertility Deities”, as shown by the fact that they display several phraseological and thematic parallels in their employment of poetic devices describing the non-functioning, or incapacitated, state of the protagonists and the consequent non-functioning condition of the cosmos around them. The use of these poetic devices is a reflex of inherited Proto-Indo-European poetic culture, as they systematically match phraseological collocations and themes attested in several Indo-European traditions describing the existential conditions of any non-functioning character (e.g. dead characters) and of any non-functioning cosmos (e.g. the world at the End of Time), respectively.
The Greek, Norse and (to some extent) the Indic narratives also attest structurally comparable scenes involving horses, whereas the Hittite myth does not, thus reflecting an innovation which must have taken place after the split between Proto-Anatolian and Core Indo-European (from which Greek, Old Norse and Sanskrit later developed). The chariots employed in the Greek and Indic narratives must reflect an even younger innovation (after the 21st century BCE).
Although seemingly in agreement in all the other basic numerals, Proto-Anatolian ‘four’ *mei̯-(e)u- is incompatible with Core IE *kʷetu̯or-. At the same time, the overall likelihood of ultimate Semitic origins of at least PIE *septm̥ ‘seven’ and further connections also to Caucasian and Uralic languages suggest that numerals were in flux in the region at the time just prior to the dissolution of the Indo-European proto-language. Since relatively recent numeralization is hinted at throughout the decimal series, the development of an independent numeral in Anatolian can be corroborated by typological evidence from the northern branches of Afro-Asiatic, whose speakers most likely played an important role in the Neolithic expansion.
Some features attested in Anatolian presuppose the existence of inherited patterns, such as they may be reconstructed for Proto-Indo-European (i.e. earlier than the split of Proto-Anatolian) on the strength of the evidence from the Core Indo-European languages. The Anatolian forms can thus only be explained in terms of continuity of PIE patterns which have undergone transformations such as remodelling of inherited structures and lexicalization. This is the case with Hitt. mimma-ḫḫi ‘to reject, refuse’, with pippa-ḫḫi ‘to move violently’ and CLuw. ūppa- ‘to carry (off)’, and with Hitt. nakkī- ‘important, prominent’, ‘difficult’, which cannot be explained ex Anatolico ipso and absolutely require a PIE source.