For example, Indo-European family is divided into groups, such as Slavic, Romance, Germanic, etc. Some of these groups can also be divided, but let`s just assume, that there is no further division. And, I suppose, the question is - is presence of, for instance, Proto-Slavic language, necessary for Slavic languages to be assigned to the same group (of Indo-European family)?
The answer may be obvious at the first glance, but there are a lot of cases, where proto-language is not present at all. And we still can say that languages B and С are in a genetic relationship (=parts of the same group, I guess), based on regular phonetic matches between them. So, I guess, languages can form a group, lacking a proto-language which unites them. But do they have enough "similarities" with Proto-Indo-European language, to be placed directly into Indo-European family, without an "intermidiate" proto-language, such as Proto-Slavic, is the real question.
3 Answers
Usually we are applying a tree model of language evolution, i.e., we assume that languages change over time producing splits into more and more branches, some of them becoming extinct, others to thrive.
Within that tree model, proto-languages exist at each splitting point and they can be reconstructed when the modelled tree is good and enough data are available.
However, this is not the only possible model. Especially when dealing with dialect continua, there is the wave model that postulates language changes originating at some point and spreading from that. This wave model predicts among other features that the central dialects of the continuum accumulate the most changes while the peripheral dialect preserve some conservative features. It may be still possible to postulate and reconstruct a proto-language in the wave model, but it is more difficult here.
And there are cases of language merger. There is no proto-language for Mitchif. Or at least, you have difficulties in claiming there is one—say Cree—with very heavy borrowing.
From the facts of attested Romance languages, one can reconstruct shared properties of the languages and say that there may have been a proto-language ("proto-Romance") with those properties, but the result is often unsatisfactory in leaving out widely but not universally shared features, which leads to the postulation of dialects of the proto-language. A quasi-solution is to recognize that the reconstructable proto-language is not necessarily the earliest form of the language where it is dialectally uniform, so before proto-Romance, there was another more uniform language, Latin, which simply cannot be reconstructed based on Romance. It's perfectly reasonable to assume that Slavic had a different form preceding what it was like at the reconstructable stage, but where it also wasn't just Indo-European.
A proto-language (hypothesis) is what you get by applying the comparative method to extant data. A challenge for reconstruction is controlling what are considered to be "extraneous" social features. Bantu historical linguistics provides ample opportunities to study this (remotely). Different subgroups in Bantu re-intertwine over time, and interact with non-Bantu populations as they migrate. The Bantu languages of southern Africa are broadly divided into two historical subgroups, the Sotho group and "the rest", but they converge on phonological points because of contact with Bantu-speaking cousins and some number of unrelated languages termed "Khoisan" (which are not all demonstrably related). This renders difficult attempts to develop a coherent historical development based on shared features, since some features were borrowed by languages that happen to be less-closely related or totally unrelated.
The main question is: was Proto-Slavic in fact a language or a group of closely related dialects? In a tribal nomadic society which occupied huge territories and had no writing one can hardly speak about a common language. The reconstructed entity named Proto-Slavic is thought to have existed for 2000 years, it's as long as from the times of Julius Cesar till now, so look at what happened with Latin during that time and imagine what Proto-Slavic actually was and how it should be treated. Besides, our reconstruction of Proto-Slavic is based on the dialects which are known to us, that is, on those that managed to survive till our time or at least till the time when they were recorded in writing like Novgorodian.
In fact, among the language groups of the Indo-European family, it is only the Romance group that has a proven single proto-language which is Vulgar Latin. The reconstructed proto-languages of other groups are just extrapolations, sets of features common to the closely related dialects that comprise each of the group and which we find in each group's earliest stage.
-
1So is this an answer or not? If yes, what is the answer? Be aware that no-one thinks that the language did not evolve. The proposed evolution is well documented in various scholarly literature, of course with some uncertainty and variation but usually contains some time frame for various changes that happened between Proto-Balto-Slavic and Common Slavic. Jan 22, 2021 at 10:30
-
It is just customary to select one particular stage for most dictionaries. Usually this stage is taken quite late, after the second palatalization - that is known to have happened everywhere (except for novgorodian) but after liquid metathesis, that did not happen everywhere (polnoglasie in the east, nothing in some occasions). That does not mean that the evolution is not acknowledged and deeply studied. Jan 22, 2021 at 10:39
-
We know that the nasal used to be full ns at some stage, but it is not the stage that is selected for most dictionaries. We can normally reconstruct the form at other stages by appliyng straghtforward rules. We also know that the differences at the Common Slavic stage in the 9th and 10th century were pretty minor, as documented in the Moravisms or Bohemisms of the Kiev Missal or the Prague fragments or any other dialectal features of other surviving documents and from the forms of the languages that resulted from these dialects. Jan 22, 2021 at 10:41
-
2@VladimirF - What I mean is not the presence of language change, but the constant process of branching and the absence of a single Proto-Slavic language. And don't tell me the Kiev Missal or the Prague fragments were written in the vernacular language of that time, they were in Church Slavonic, a standardized language. If explicit dialectisms penetrated even the Church Slavonic text, we can well imagine how more numerous they were in the vernacular languages of those times. Jan 22, 2021 at 10:53
-
But it is well documented that the dialect continuum was very homogeneous at the Common Slavic stage. There are good reasons to believe it was even more homogeneous a few centuries before that. Jan 22, 2021 at 10:59