Page 1 of 1

How to make a (pseudo-)Paleo-Balkanic conlang?

Posted: 09 May 2018 17:13
by Zythros Jubi
Recently I've read A Concise Historical Grammar of Albanian by Vladimir Orel and The Language of Thracians (in its English translation, see: by Ivan Duridanov, as well as several papers by other Bulgarian scholars such as Bilyana Mihaylova. They came up with a hypothesis that the Thracian and Pelasgian (pre-Greek substrate) languages are closely related, and are characterized by a Grimm law-like sound shift (i.e. IE *t>*th *d>*t *dh>*d), intact PIE *s, plus Satemization (actual realizations are unknown, Duridanov supposed *k' *g' *g'h > *θ *ð *ð); in addition, syllabic consonants turned into *uC/iC. In contrast, in Proto-Albanian the first set of sound changes is absent, instead *Dh > *D ; only the other three are present, with *k' *g' *g'h > *ts *dz > Alb. th dh/d, and PIE *s > early Proto-Albanian *s (complex changes afterwards).
I wonder the relationship between Thracian-Pelasgian (if these two form a coherent group), Illyrian and Dacian: according to Duridanov's reconstruction (see Wikipedia entry for Thracian language & Dacian language), such a consonant chain shift is absent in Illyrian and Dacian. Perhaps the three groups were particularly close, and the chain shift occurred relatively late.
Anyway, has anyone here attempted to create a Paleo-Balkanic conlang, or Paleo-Balkanic-based one? I want to make one spoken in Rhodopes (Smolyan Province, Dospat and Sarnitsa, Chech region in Greece etc., less than 500,000 native speakers; perhaps as a descendant of Bessian; high proportion of Muslims) with typical Balkan Sprachbund characteristics: suffix for definiteness, merger of dative and genitive, substitution of infinitive by subjunctive (at least in some dialects, cf. Serbian and Albanian), synthetic future, 11 = "one on ten", and so on. The crucial question for me is the choice of PIE root: I'm unsure how related Thracian and Pelasgian are lexically, and according to Duridanov et al., Thracian is close to Balto-Slavic especially Baltic in terms of lexicon; Orel also pointed out that Albanian has numerous cognates with Germano-Balto-Slavic, then with Balto-Slavic, particularly Baltic, relatively rare with Germanic and Slavic alone; notably, somewhat more cognates with Greek-Armenian and Indo-Iranian. I also suppose the conlang to have more loanwords from Greek and Southern Slavic than Latin.
Any suggestions for choice of PIE roots and derivational affixes, and the proportion of roots of uncertain etymology/substrate (cannot be traced to PIE)? How about the reflex of *k', *s or *ts or *θ? I want to have voiceless aspirate changed to fricatives because of Greek influence.

Re: How to make a (pseudo-)Paleo-Balkanic conlang?

Posted: 10 May 2018 03:31
by Zythros Jubi
As for Proto-Albanian homeland, Orel said it was marshy land, not mountainous nor coastal, with few valleys; I guess Pannonia or Wallachia may fit. In terms of indigenous vocabulary, words for cattle and sheep herding is abundant; in PA there are much more Latin loanwords than Greek ones.

Then it comes the question, what will be my con-protolang called: I intended to name it Proto-Thraco-Pelasgian, but it may lead to confusion, like Mecislau's Proto-Ob-Yeniseic.

Re: How to make a (pseudo-)Paleo-Balkanic conlang?

Posted: 17 Jul 2018 00:00
by Ælfwine
Look into Romanian. A lot of early sound changes in it could be attributed to Dacian, which is thought to be close to Illyrian.

a > ə for example, in certain environments, seems common.