Let us consider two languages that are clearly related to each other, and one of them is older than the other. How would an academic linguist determine if the older language is an/the ancestor of the newer language or the older language is an older variety of the newer language in the past?

I suppose branching is the key. If there is no branching, it would be reasonable to conclude that it is a historical older variety (I further assume that if the time gap exceeds a dozen or so generations of speakers, a search for lost sibling languages is in order).

However, my question is in the scenario when the newer language has siblings, and all siblings have good amount of mutual intelligibility. Further, all siblings are considered bona fide languages and not dialects.

In such a scenario, any of the siblings (ts speakers rather) can claim that the older language is an older version of itself, thus appropriating all the prestige and legacy of the older language. What are the criteria to resolve this appropriately?

I suppose there will still be 'proper older version' of every sibling after it branched. But what if there isn't clear evidence of that, or a consensus about that? Would a linguist have to peg each sibling to an arbitrary period and call it 'Old X'?

Please note that I am aware of politics of nationalism that play a role here (I am aware of the quip about language and dialect, and army and navy). I am looking for purely academic response.

EDIT: I use the words old and new in the normal sense they are used, meaning, a language X as observed in 1500 AD is newer to one observed in 800 AD. I use the word ancestor to mean a language, that certainly existed and was not constructed by linguists, that transformed into X and its siblings. It could be taken to mean lowest common ancestor in the language tree. Further, once a language X has its name attached by its speakers, and then on is referred to as X by its speakers, X as it was two centuries back is old X and not ancestor of X of later times.

  • What's the supposed difference between "ancestor" and "older variety"? Old English is both the ancestor of English and an older variety of it, and the same arguably goes for Proto-Germanic or Proto-Indo-European.
    – TKR
    Jul 14, 2016 at 16:31
  • 1
    all siblings have good amount of mutual intelligibility. Further, all siblings are considered bona fide languages and not dialects. -- this is self-contradictory insofar as mutual intelligibility is the main criterion for calling two linguistic varieties "dialects" rather than "distinct languages".
    – TKR
    Jul 14, 2016 at 22:31
  • 1
    It looks to me like your edit (once a language X has its name attached, by anyone) basically defines an answer to your question. If you define "older variety of a language" to mean "older form sharing a name" and "ancestor" to mean "older form not sharing a name", well, that becomes your criterion for distinguishing "older variety" from "ancestor". It's not a very meaningful distinction, though.
    – TKR
    Jul 14, 2016 at 22:35
  • 2
    This is Linguistics SE so we use the term 'dialect' in its technical linguistic meaning, but it also has a political usage. In linguistics 'dialect' refers to varieties that are mutually intelligible, while those that are not are referred to as different 'languages'. From the linguistic pov 'Scandinavian' is one language with dialects such as Swedish, 'Norwegian' and Danish. Politically these are called 'languages' to emphasise that they are separate nations. The reverse is true of the Chinese 'dialects', which are (in linguistic usage) different languages. Jul 15, 2016 at 13:28
  • 2
    @vin A linguist would not say that Indo-European is a language of which French is a dialect. Proto-Indo-European and French did not coexist in time, so one can't be a dialect of the other. And even in an alternate world where they did coexist in time, the two languages are radically different and would have zero mutual intelligibility.
    – TKR
    Jul 16, 2016 at 3:13

1 Answer 1


Languages don't actually have ages, and they aren't discrete things like people are. So we'll have to first interpret what you mean by "older language" versus "younger language". It's even harder to understand what the difference would be between X being an ancestor of Y versus X being an older variety of Y.

Let's take "American English as currently spoken in the Pacific Northwest" ("Y") to be the younger language, and "English spoken in the British Isles in 1600" ("X") to be the older language. It would be reasonable to say that X is an ancestor of Y, in the same way that one can say that Knudt Knudtsen is an ancestor of me. We would not say that I am a newer variety of Knudt Knudtsen, because Knudt Knudtsen is a discrete entity who clearly died. X is not a discrete entity which clearly died at some point in the past. What an academic linguist would conclude is that there isn't a difference between the possibilities that you describe -- they refer to the same fact.

The same problem afflicts appeal to linguistic "sibling", which would probably be taken to be "a closely related language descended from a single earlier language". Well, it turns out that "American English as currently spoken in the Pacific Northwest" ("Y") isn't entirely uniform – there are documented dialect differences (east of Cascades / west of Cascades being prominent; there is an Enumclaw dialect). No two people speak the exact same language (i.e. have the exact same grammar and lexicon, even setting aside cases where one person knows a word that the other person doesn't know). The significance of that is that there is no clearly-defined boundary between language and dialect. There seems to be a feeling that Scots is not a dialect of English, and certainly I can't understand it at least the bit of it that I've heard. So by the mutual intelligibility test, most people would say that Scots is a distinct language from English. I don't think there is much support for thinking that [Geordie][1] is held to be a separate language, but it is completely unintelligible to me. Indeed, a number UK dialects are unintelligible to me, and some US dialects are too.

A problem with using popular opinion as the adjudicator of the dialect / language distinction is that people mostly don't have opinions except about their own language, and then their opinions are shaped by completely unscientific and non-linguistic criteria. Whether or not the various Kurdish languages and Saami languages are considered to be separate languages or dialects depends on non-linguistic ideology and political matters. (My grasp of the situation is that Kurds downplay dialect differences in favor of a single Kurdish ethnicity, because there is an ongoing struggle for independence that relies on the unity of the Kurds as a nation; whereas the Saami enjoy basic rights and a fair degree of political autonomy so recognition of linguistic differences is not a threat to their future. More could be said, but won't be).


I don't think using "referred to as X by its speakers and outsiders" will get anywhere, since people's name for their own language is frequently different from what outsiders call them (the general scheme is that the language-internal name translates as something like "language of the people" and the outsider name translates as something like "language of the barbarians"). Eskimo, Sioux, Berber, Hottentot and Lappish are all etymologically insulting exonyms, likewise the Russian word for "German".

If you discard outsider names and stick with self-names like Suomi (Finnish), Deutsch (German), ελληνικά (Greek), or Italian(o), then there might be a legitimate question of cultural identity. Perhaps we could say that "English" first existed when those Germanic tribes called themselves "English" rather than "Saxon" or "Jute" (or historical equivalents -- "English" is a modern word). Note that Irish (Gaeilge) and Scots Gaelic (Gàidhlig) are different languages, but the names of the languages reflect a continuous use of the indigenous name for the language. A historical change in the name of a language probably reflects a major cultural change more than anything else, although actual change in language can be a part of cultural change.

Anyhow, the way "ancestor" language is used in linguistics, it refers to what preceded historically, without care for the language / dialect division, though we tend to reserve the word for predecessors that are at least some number of hundreds of years old.

  • i have edited the question to explain a bit more on what I meant. Please have a look.
    – vin
    Jul 14, 2016 at 18:25
  • 1
    @vin: even after your edits, there is really no linguistic question here. It's all a matter of politics and opinion. It's just as meaningless to exclusively identify either modern language with the common ancestor as it would be in biology to exclusively identify either humans or chimpanzees with their common ancestor. Furthermore, there are no dividing lines among languages in nature, just as there aren't among species, so it's really impossible to precisely identify when one natural language evolves into another. Jul 14, 2016 at 22:57
  • Yes, outsiders referring to a language isn't quite 'rigorous', so removed it. Your reply is quite informative but I am not able to find an answer to my question (it could be my fault for failing to find). Could you please make it explicit? You have removed the part about the question's premise being wrong. So does it mean it isn't any more?
    – vin
    Jul 15, 2016 at 6:39
  • @sumelic I understand the problem. I have even hinted about that in my question (not sure if it was clearly stated though). However, that wouldn't stop linguistics from attempting to unravel the mystery. It is that methodology that I asking about. To reiterate, I don't want to say language X started on such a day in such a year, rather I seek techniques to investigate a chaotic situation where many interrelated languages identify with an older prestigious language for obvious reasons.
    – vin
    Jul 15, 2016 at 6:44
  • 1
    @vin: unravelling what mystery? Within a certain range of relatedness, calling two varieties of spoken speech "languages" or "dialects" is just a matter of terminology. There is no objective linguistic way to determine the best terminology to use, especially not in a chaotic situation. The whole concept of "appropriating all the prestige and legacy of the older language" is not linguistically meaningful. Jul 15, 2016 at 9:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.