Indo-European and Romance are Very Different as to Diachronic Linguistics

No, "Language Divorce" is Not my Amateur Term for Divergent Evolution! · Proto-Languages - How Are they Reconstructed? · Sabellian and some more, but first Vulgar Latin · Indo-European and Romance are Very Different as to Diachronic Linguistics · More on Language : Latin to Romance · More on Latin to Romance and Middle English to English · More on language in general

By contrast with "Sabellian and some more, but first Vulgar Latin", this one deals with the part of Diachronic linguistics that Brugmann dealt with, and why it is not as well known, why it is more of a guess work.

Q

What evidence do we have of the existence of an Indo-European language?
https://www.quora.com/What-evidence-do-we-have-of-the-existence-of-an-Indo-European-language/answer/Hans-Georg-Lundahl-1

Hans-Georg Lundahl

amateur linguist

July 12th, 2023

The evidence are the “daughter languages” if the theory is true.

They could also be evidence of Sprachbund phenomena.

Q

What is an attested language in linguistics? What is an unattested language in linguistics? What is the difference between an attested language and an unattested language in linguistics?
https://www.quora.com/What-is-an-attested-language-in-linguistics-What-is-an-unattested-language-in-linguistics-What-is-the-difference-between-an-attested-language-and-an-unattested-language-in-linguistics/answer/Hans-Georg-Lundahl-2

Hans-Georg Lundahl

amateur linguist

28.VII.2023

If you have heard someone speak English, English is an attested language.

If you have read hieroglyphs, Old Egyptian is an attested language.

Any language that must have existed (or maybe need not have existed) BUT which is not attested in either of above ways, orally or in writing, these accessible to the culture of the linguists, is an unattested language. We know unattested languages must have existed, but it is harder to know if a specific unattested language existed. It depends on how good the reasoning is behind the reconstruction.

Q

When reconstructing Proto-Indo-European, do linguists give more weight to the sounds and forms of the earliest attested languages?
https://www.quora.com/When-reconstructing-Proto-Indo-European-do-linguists-give-more-weight-to-the-sounds-and-forms-of-the-earliest-attested-languages/answer/Hans-Georg-Lundahl-1

Answer requested by

John McCarthy

Hans-Georg Lundahl

amateur linguist

31.VII.2023

By now, this is no longer needed, if there was a PIE.

Since Brugmann, the sound changes have been so fixed that one can permit oneself the luxury to use later attested ones, that are considered to have gone through fewer changes in sound and form.

My Greek Professor told me (some time prior to spring term 1993) that if you have a Greek form and a Lithuanian form for the same word, you can confidently say what it sounded like in PIE. Now, Lithuanian is not attested prior to AD 1000, and not attested in fullblown texts prior to 1500 AD.

Wiki says:

Among Indo-European languages, Lithuanian is conservative in some aspects of its grammar and phonology, retaining archaic features otherwise found only in ancient languages such as Sanskrit[7] (particularly its early form, Vedic Sanskrit) or Ancient Greek. For this reason, it is an important source for the reconstruction of Proto-Indo-European despite its late attestation (with the earliest texts dating only to c. 1500).[4]

Lithuanian language - Wikipedia
https://en.wikipedia.org/wiki/Lithuanian_language

Footnote 4 is to "Lithuanian Language". Encyclopedia Britannica.

Footnote 7 is to Smalstieg, William (1982). "The Origin of the Lithuanian Language". Lituanus. 28 (1). Retrieved 7 August 2016 – via Lituanus.org.

Please note here, the timespan for Proto-Indo-European being supposedly spoken (according to the most but not only theory of there being one) is c. 4500 – c. 2500 BC.

Proto-Indo-European language - Wikipedia
https://en.wikipedia.org/wiki/Proto-Indo-European_language

So, 4500 BC (if that date were relevant for Indo-European) is 3700 years prior to Homer. Homer is 2800 years prior to us. Mycenaean Greek goes back another 800 years or so. But the texts are a corpus inferior to Homer and much more monotonous and therefore much less informative about the language. Hittite is comparable to Mycenaean Greek or a bit older, so is the language of the Mitanni, and Vedic Sanskrit would actually be younger than that, since closely related.

In other words, even “the earliest attested languages” (of the Indo-European “family”) will not get us all that far into the direction of actual Indo-European if ever there was one single such language. In fact, for each “subfamily” (we deal with clear actual language families), the earliest attested within that one is used. Gothic and Proto-Norse, depending on whether you prefer texts that can be read while sitting down or simple one to ten word inscriptions on artefacts, are the oldest for the Germanic family. Old Latin is the arliest (I think) for the Italic one. Homeric or Mycenaean Greek (same preference choice) is the oldest one for the Hellenic one. Old Irish or Gaulish / Lepontic (same preference choice) is the oldest one for the Celtic one.

All of these oldest attested languages per family are between themselves not at all as closely or as obviously related as the languages within one family. Swedish and German are far closer than Homeric and Latin. At least as close as Old Slavonic and Lithuanian, probably still closer than that (and note, one can argue whether Slavic and Baltic languages are two families or one).

So, the methodology of using earliest attested language for each “subfamily” is obviously a very good method when it comes to getting the reconstructions of PIE as close as possible to the actual language spoken by people who left their language, supposedly, to both Homer and Virgil, both Vyasa and the anonymous author of Táin Bó Cúailnge. But it is not in any way, shape or form a guarantee that these attested languages actually go back to a single non-attested one, which one is trying to reconstruct.

Speaking of reconstruction efforts. Since the sound laws in Brugmann are between the “Lautstand” (phonetic state) of one specific reconstruction and the daughter languages, when you change reconstruction you also change sound laws, and when you change sound laws you change reconstruction. There is still no single reconstruction for PIE that all Indo-Europeanists agree on as being correct, though the one used for Pokorny is used as a reference. Any one reconstruction will leave important gaps in the explanation of why we find this form or that sound or this word used in that sense in a given daughter language. Any reconstructed proto-word will have forms that do not obviously match with that word in some of the languages. Wasp is vespa in Latin. Fine, no doubt this was for some reason the same word. Dito for Lithuanian vapsva. But it is “osa” in Slavic languages, which according to the sound laws in Brugmann is where vapsva would be going, but it is in itself too unlike vapsva, vespa, wasp to prove it is one word with the other.