From: blog UROPI : Proto-Indeuropan (PIE), Uropi id moderni lingas - et les langues modernes -and modern languages
Yesterday or today can count as PI day (3.14 and 3.15 or 314:100 and 315:100 being limit values surrounding in two decimal precision the non-showable precise value of pi, however, 3.14 is the closer of the two).
Some linguist made a pun and made it PIE day ... as not in apple pie or mince meat pie but Proto-Indo-European ...
Now, there is one group which has never been seen as documented in writing or oral performance. Proto-Indo-European.
There are three of four or five groups which contributed to the theory which had very early contact : Italic (with Latin as well known example), Celtic (also marked close to Italic on diagram), Germanic (other branching, itself closer to Baltic and Slavic which are closer to each other than to it).
The fourth is Greek, which had later (but still early) contact with Italic.
The fifth is Indic (with Sanscrit).
So, four out of five main contributors to the theory of a single proto-language are more similar than they should be purely out of origin, due to early contact.
If Linear A turns out to be related to Indic or Iranian or both, that would be the fifth group of the original theory which had "very early contact".
And in the top branching of the diagram, closest to PIE, non attested, is Anatolian (attested with Nesili/Hittite, with Luwian and with a few more).
That is in the right position to have also had very early contact with ... the Italic and Greek early but not very early contact.
So, what if IE group is a Sprachbund?
Anatolian of some kind (Nesili or Luwian or Lydian etc) acting as a catalyst for Indo-Europeanisation, at least in some main features and the other in the "early contact" or "very early contact" contributing less important common or half common features?
This is a theory I have been brooding on, and so far I have had no refutation, except mostly that this is not the theory of the accepted expert linguists.
It was, I suppose, the theory of one, namely Trubetskoy, and even if this had not been the case, it would still be a valid theory if explaining things.
It has been met with two main objections:
- existence of regular sound correspondences, as would be expected with a common background changing in different ways (Latin to Romance model of language development)
- reluctance of languages to borrow base vocabulary and grammatical features.
And I have answers to both.
"existence of regular sound correspondences"
- regularity is not absolute;
- tentatives to make it so end up with a more and more contrived proto-language, which begins to sound like Klingon;
- regularity may also result from "backformations" in the case of words taking the other route than the typical one : if I know that Swedish kona - dialectal for kvinna, related to Queen and to γυνη as well as to Celtic bean, fenyw/benyw - in Danish becomes "kone", and hear Danish "pige", girl, I might take the word and in Swedish change it to "piga" or even "pika" (Norwegian has pike). As it happens, Swedish has the word piga, and in a slightly different meaning, maidservant. Changing Danish -e to Swedish -a is a backformation, compared to the Danish change from -a to -e.
- Regularity may also result from having a common text with dialectally different variations on the sound of letters, even possible with alphabetic script, and certainly possible with cuneiform and hieroglyphic syllabaries, since there the set values of a given glyph may change, especially if they don't look like devanagari, with consonant given and vowels other than short a added, but like early Aegean and Anatolian ones.
"reluctance of languages to borrow base vocabulary and grammatical features"
- ignorant of Balkan linguistics, where at least favouring of certain and disappearance of other forms in abstract grammar grids are made by bilingual speakers' relations to other languages;
- ignores that IE shares grammatical features with two neighbouring language families, Semitic and Fenno-Ugrian (Ablaut is a Semitic trait, endings for first and second persons are very clearly related in IE and in Finnish, the typologically rare double past, either imperfect or perfect, in IE, or in most IE, could be from Semitic tenses being more about perfective and imperfective, while Fenno-Ugrian is more about present versus past, Germanic shares the latter type with Fenno-Ugrian, and shares Grimm or part of it with Hungarian and Etruscan), the 8 case system is of complexity intermediate between Semitic 3 case and Fenno-Ugrian 15 case systems;
- and of course also ignores the Tsiganic or Gypsy languages, the various languages known as Romani : caló is spoken with Spanish-Portuguese type grammar, tattar-Romani with Scandinavian grammar, "Shelta shares its main syntactic features with Hiberno-English and the majority of its morphological features such as -s plurals and past tense markers." (from article). The latter remark is true for Scandinavian Romani too, like noun plurals in -(V)r, like verb presents, all persons, -(V)r, like verb pasts in -(a/e)de and I think the same applies to Caló ... yes, and here is another similarity between Caló and Tattarspråk : "Many Caló terms have been borrowed in Spanish (especially as slangisms and colloquialisms), often through Flamenco lyrics and criminal jargon (germanía)." (Also from its article).
I have even played with the thought that the main catalyst for indo-europeanisation of included groups may have been a language used like Romani. If the common word for grain is not specific between wheat and barley and rye, while IE commonalities are specific about horses, it might be because wheat was not one usual article such a far range trading people usually trasnferred, but horses were their own thing.
Other languages affacted by theirs would also have become IE language groups, but retain their own words for relevant cereals.
Hans Georg Lundahl
II Lord's Day in Lent
PS, I forgot the very obvious lacunae of IE commonalities in very common base vocabulary, hands, heads, colours, wheat and rye and so on are not commonalities, and neither is iron (iron, Eisen, jern is a Germanic commonality with Celtic Houarn, Haearn, Iarann, but not with Ferrum in Latin and Romance, unless Celtic were borrowed from a form of it changing f to h or zero, not with Geležis and Żelazo, not with Σίδηρος, and on Fenno-Ugrian side, Vas and Raud are unlike, the latter however borrowed perhaps from proto-Germanic or early Germanic, like proto-Nordic for red).
PPS, in Danish, I suppose "pige" is regular. If Norwegian "pike" is regular, the g in Swedish "piga" must be from a dialect of Danish or a dialect close to Danish. If the g in Swedish "piga" is regular, the k in Norwegian "pike" must be a backformation. Nw/Sw lök/løk Da løg. Da pige, Nw pike. The sound change has given rise to a sound correspondence and the sound correspondence contributes to words changing sound the other way, what linguists now call a backformation. Once it was simply called a sound correspondence.