For my MesoEuropean speculation, I'm trying to hammer out a type of "jargon" that would include Uralic, IndoEuropean, Etruscan, and Basque. While I don't think they're directly related, there are parallels between these that can be observed with closer inspection; though the problem of testability arises when you compare any two of these individually.
I've tried conquering this problem with subgrouping Uralic with IndoEuropean as a genetic entity, and Etruscan and Basque as a possible one, though it helps, it certainly doesn't fix the problem. The problem being any "cognate" forms may originate from none of these language families specifically, but may be present in all of them. Further, these "cognate" (really bad word for it) need not genetically be related to one another, but could have arisen areally.
For instance, haber and have in Spanish and English, used almost identically, but not technically cognate. Their similarity may or may not have influenced or influence the developement of either, but that doesn't change the fact that those communicating between these languages will tend to use a form "have" to translate "haber", even if the corresponding Germanic form of "haber" were to be used as well.
This creates an obvious testability problem to add to the already existing ones. Now, not only can't we distinguish what is loaned and what isn't (a problem even extending to both "subdivisions"), we can't even be sure what forms are really convergent by this (as I'm pretty sure haber and have developed ...)
An example of these likely common borrowings/convergences being:
A pronominal system with a specific type of vowel gradation holding specified meaning:
mu(n)/mo(n), mi(n)/me(n), ma(n)/mə(n)/m∅(n) - 1st Person
tsu(n)/tso(n), tsi(n)/tse(n), tsa(n)/tsə(n)/ts∅(n) - 2nd Person
su(n)/so(n), si(n)/se(n), sa(n)/sə(n)/s∅(n) - 3rd Indefinite/Human
tu(n)/to(n), ti(n)/te(n), ta(n)/tə(n)/t∅(n) - 3rd Person Definite
*/ts/ used because I didn't want to give the impression that both could be reconstructed at any branch, as it really does matter which you choose. With the only exception being IndoEuropean, which shows /s/ on verbs and /t/ in pronouns for the 2nd person, and the 3rd person forms show seperate roots for /s/ and /t/ in differing branches such as Uralic (assuming any validity to the speculation). For the 2nd person, Uralic shows /t/, like with Luwian, also an /ri/ developing from that /ti/ form. Though the 2nd person and 3rd person forms in IndoEuropean and Uralic "seem related", I can get to why, possibly some "thou, this" distinction that can be later explored.
Both "thou" and "this" correspond to the Basque "hi" and "hau" respectively. In the above, that would be ts > z /s/, and t > h.
Etruscan would show the same /z/ and /h/ developements as Basque in this framework.
Correction made: I confused the Basque singular "thou" hi with the plural zu. This does get rid of the association between /u/ and /i/ and number.Also ni "I" and gu "we" below.
Correction made: I confused the Basque singular "thou" hi with the plural zu. This does get rid of the association between /u/ and /i/ and number.Also ni "I" and gu "we" below.
*-(n) should be more accurately rendered -n, and is obligatory in PUralic it would seem, but optional (?) in PIndoEuropean, or at least both forms with and without it can be seen in the possessive forms. This has lead to the moi- moin- distinctions, maybe? Though it may have been dropped too, and have been originally necessary, this is unclear. Though both exist in Uralic and IndoEuropean, without the -n element, Uralic has it in its enclitic forms (which are nearly identical to their IndoEuropean counterparts). Not sure what the Basque and Etruscan would yield. Though I know that Basque lacks the final nasal. It appears like it may just differentiate pronominal and attributive forms, if this nasal ending is even to be differentiated at all from the base root (hence the parenthesis). As a very simple sound change could eliminate it from suffixes.
The very nature of a root without the -n appears like it could be problematic for both Uralic and IndoEuropean, despite being reconstructible for both.
*3rd Person definite forms are historically related both back to the plural suffixes and the demonstrative pronouns with definite meaning under this model.
Some have linked the PIE plural -s with the PU -t, but my own thinking is that these are somewhat seperate. The reason for this, is simplicity. Both s- and t- exist as indefinite and definite forms of demonstrative roots, and assuming its an enclitic demonstrative root in origin (easily reconstructed and uncontroversially as -es in PIE and -t in PU), and their parallel t- (deictic demonstratives in both, and technically all 4; often the reflexive pronoun in Uralic) and s- (IE 3rd Reflexive and PU 3rd Person pronoun).
Also, evidence from IndoEuropean using a collective ending for Neuter nouns and pronouns seems to suggest that these originally didn't decline for number in the singular ~ dual ~ plural like Human/Animate nouns originally did (Masculine by the time of PIE), rather, by an Individual vs. collective type distinction.
This gives us a clear PIE neuter singular demonstrative -d as cognate with the PU -t. Though these would have necessarily developed seperately. I'm going to come back to the reason why in the section on ablaut.
You can see the mess this would create, trying to resolve this as one common root, instead of a merger of two seperate roots into a suppletive type pronoun in either branch independantly developing, IndoEuropean or Uralic.
The ∅-grade and a-grade forms might possibly be seperate or they may not be. Their meaning seems to not necessarily have been distinguished in personal pronouns, but for demonstratives they lend a sort of general/distal deixis (giving a generalized meaning "that").
The i/e-grade form has two seperate usages, it would seem.
In demonstrative pronouns, this grade gave rise to the "near deixis", and is constructed the same way in PIE as that plural suggestion, and indeed -i- for near deixis in PU. Further, this form may actually be related to the PIE root "this" also beginning with j-, and the root kj- "this".
In demonstrative pronouns, this grade gave rise to the "near deixis", and is constructed the same way in PIE as that plural suggestion, and indeed -i- for near deixis in PU. Further, this form may actually be related to the PIE root "this" also beginning with j-, and the root kj- "this".
In some branches of IndoEuropean, this remained the plural ending, alternating with the plural -es in other branches. In yet other branches, the j- root actually gave rise to the definite declension such as with Lithuanian and Latvian, (as t- had done in Balkan Romance, Luwian, as well as Bulgarian and Macedonian). Curiously, Germanic shows -n- for that same definite declension (so-called weak stems), seemingly related to the word "a/an/one".
While the definite meaning clearly developed seperately, there seems to have been some construction similar to what we see in Basque as the original state for IndoEuropean. Whether or not the noun was always declined for case in these constructions, the following determiner became declined for case, gender and number. This easily explains the independant developements of these as independant case endings for the Indefinite and Definite declensions in some branches, but there is a controversial element to it. It may well get rid of some case endings.....
.... I'll get to that later, this has gotten way too long as it is, we'll deal simply with phonological issues for now.
While it doesn't seem related (for the most part), Sumerian also seems to have a similar system of Ablaut for its pronouns. It may not be related. Though a comparison can be drawn between Sumerian and IndoEuropean enclitic personal pronouns:
o/u-grade IndoEuropean and Sumerian genitive/possessive enclitics
e/i-grade IndoEuropean Oblique and Sumerian Ergative enclitics
a-grade If the same as zero-grade, this would be the equivalent of the root-form in Basque, IndoEuropean, and Sumerian.
