Studies in linguistic phylogenetics

Methodology, terminology and the relationship between Indo-Iranian and Greek

Public defence of PhD thesis by Simon Poulsen.

Read thesis (pdf). 

Assessment committee

  • Professor Birgit Anette Rasmussen (Olsen), University of Copenhagen (chair)
  • Associate Professor George Hinge, Aarhus University
  • Professor James Peter Timothy Clackson, University of Cambridge

Moderator of the defence

  • Associate Professor Seán Vrieland

Copies of the thesis will be available for consultation at the following three places:

  • At the Information Desk at Copenhagen University Library South Campus - Humanities and Law, Karen Blixens Plads 7
  • In Reading Room East of the Royal Library (the Black Diamond), Søren Kierkegaards Plads 1, 1221 Copenhagen K
  • At the Department of Nordic Studies and Linguistics, Njalsgade 76, 4A.2.11.

 

This thesis examines the relationship between Indo-Iranian and Greek, two branches of Indo-European that are on the one hand frequently suspected of forming a clade or having been spoken close to each other in prehistory, but on the other hand preserve most of the linguistic material on which Proto-Indo-European is reconstructed. It consists of three studies in article form and an introductory section dealing with the methodology of higher-order subgrouping and its application to the hypothesised latest common ancestor of Indo-Iranian and Greek. First, it assesses the strengths and weaknesses of traditional phylogenetics, computational phylogenetics and prehistoric dialectology. This section reveals that traditional methods have a weakness when comparing very conservative and abundantly attested languages with much more scarcely attested languages, but that computational methods rely on the least salient evidence to avoid the data bias. Then it assesses the phonological isoglosses shared by Indo-Iranian and Greek and reconstructs the phonological system of the latest common ancestor. This reveals that it is impossible to reconstruct a younger ancestor than Proto-Indo-European, and that the proposed innovated phonemes are either inherited (*a, *b) or parallel (voiceless aspirates). It assesses the morphological isoglosses shared by Indo-Iranian and Greek which reveals that the two branches are – unsurprisingly – conservative representatives of non-Anatolian Indo-European. Their unique isoglosses are either unidentifiable in other branches or rather parallel and arose form system-internal pressure than contact. Finally, it applies an archaeolinguistic approach to narrow down when and where the alleged proto-language or ancestral dialect continuum was spoken: 2000 bce north of the Black Sea.

Article 1 presents a survey of the term “Indo-Greek” and of published tree topologies. It concludes that the “Indo-Greek hypothesis” does not exist in a narrow phylogenetic sense, and that no (available) family tree places the two branches as
sisters. Article 2 discusses the role of loanwords in linguistic phylogenetics under different methodological approaches and supplies two case studies. It concludes that *pelek̑u- ‘axe’ cannot be a borrowing from semitic, and *(H)a(i)̯g̑- ‘goat’ is unlikely to be Caucasian. Article 3 is a Maximum Parsimony analysis of East Iranian; a group of languages often considered a Sprachbund. The results are
inconclusive, but the paper offers new insights into the practical applications on computational analysis on grammatical data.