Re: [tied] Latin versus *Proto-Romance

----- Original Message -----
From: <x99lynx@...>
To: <cybalist@yahoogroups.com>
Sent: Monday, February 03, 2003 7:19 PM
Subject: [tied] Latin versus *Proto-Romance


> Piotr, do you mean Latin is not the 'immediate' ancestor of Romance? In all the descriptions I've seen of the development of Romance, it seems as if the analysis seems to be saying Classical Latin > Vulgar Latin > Romance.

To claim that would be like claiming that all the accents of Modern English derive from "vulgarised" RP (not that there aren't people who believe in something of the sort). Classical Latin was a literary language, used by the educated, artificially codified by prescriptive grammarians and canonised in the high prose and poetry of the Golden Age. The colloquial language of Rome's polite society (known as "sermo cotidianus" and reflected in the more informal Latin texts) was of course strongly influenced by the official norm but it also had features not reflected in "serious" literature.

But standard Latin was only a tiny fragment of a much larger whole that existed independently of it. Latin was like any other language in that it had the normally expected range of regional and social variation. There was a continuum of sociolects between upper-crust Latin and "sermo plebeius" a.k.a. Vulgar Latin -- the latter being a cover term for a whole family of provincially differentiated "Latins" used by the uneducated or poorly educated classes (that is, by the overwhelming majority of Latin-speakers). It was the popular dialects, with their nonstandard and absorbed "barbarian" features, that eventually gave rise to the Romance languages. If you have no access to a handbook of Romance linguistics, here's a small discussion of it at:

http://www.orbilat.com/Proto-Romance/Proto-Romance.html

> How would the nominative > accusative shift, for example, make any sense if Classical Latin were not being treated as ancestral?

Non-classical Latin also inherited the cases, but the weakening and loss of final /-m/ (not quite alien to standard Latin, BTW) was earlier and more consistent there, since its pronunciation was free from the influence of the conservative standard spelling. There was also a tendency to drop final /-s/ (most consistently in the eastern dialects), as a result of which the nominative and the accusative were confused phonologically not only in the singular of "a"-feminines but in nearly all nouns, and the nom./acc. distinction collapsed. The levelling-out of older alternations, if any, was usually in favour of the accusative, since the accusative was more "regular", i.e. consistent with the rest of the paradigm (pe:s, pede(m), pedis, etc. > Romance *pEde [not *pe or *pes], getting rid of the oddball nominative form).


> PIOTR ALSO WROTE:
> <<Instead, we learn more about Latin and its internal differentiation. PIE counterpart of Classical Latin may have never existed, for all we know.>>
>
> Well, this seems important. If, just for the moment, we posit that the PIE counterpart of Classical Latin never existed, what would that say about what we are reconstructing or about what additional explantions we might need for the existence of the IE languages?

What the reconstruction shows us is a flat shadow of what must have been an ordinary language with at least some degree of internal (dialectal, stylistic, perhaps even social) differentiation and its normal share of grammatical idiosyncracies and typological untidiness. I don't want to imply that the attested branches originated from different dialects of PIE; more likely some of the oldest dialects vanished forever without leaving any offspring, and the boundaries of others were affected by waves of diffusion leading to areal convergence at a time when IE (or most of it) was still a continuous network of dialects. Some of the original variation perhaps survives in the daughter languages, but it's mixed with secondary variation of more recent origin. Still, no living language is homogeneous, and it would be naive to imagine that we'll ever arrive at some kind of "standard PIE" through comparative reconstruction.