Linguistic Miscellany Thread

[jal]

But we do-- we can see the Indo-European, Sino-Tibetan, and other families taking over huge territories over recorded history, and what we know of those areas from say 5000 years ago is that there were more families there. Where Europe is now almost entirely IE, and the Middle East is almost entirely Semitic, there were multiple language families or isolates.

As far as I know, Indo-European had taken over its current areas in Europe and India quite some time before recorded history (a pleonasm?). And my point was that it is all (educated) guessing how this happened, and especially what it replaced. And I really wonder how we actually know "there were more families here", as the only non-IE language we have in Europe is Basque. For all we know (exaggerating a little), IE replaced Basque everywhere. Yes, we probably have some substrate words, and some hydronomy, etc., but to confidently claim that IE and Semitic replaced multiple families, and going from there to say that there were more languages back then than we currently have IE languages - that seems to be a long stretch.

Again, I don't get this. With agriculture, you get single families extending over entire continents. How is that more diverse?

We don't know what these replaced. Could be one single family replacing another, and that other family could be of very low population density, and hence without much diversity.

It seems far more likely to me that Eurasia once resembled the precolumbian Americas: dozens of families. The Americas did have agriculture, of course, but it's precisely the areas with long-term agriculture (i.e. Peru, Mesoamerica) where diversity is less.

That may seem likely, but that doesn't equal proof. And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.[/quote]


JAL

[bradrn]

For all we know (exaggerating a little), IE replaced Basque everywhere. Yes, we probably have some substrate words, and some hydronomy, etc., but to confidently claim that IE and Semitic replaced multiple families, and going from there to say that there were more languages back then than we currently have IE languages - that seems to be a long stretch.

Well… this is trivially false, because at least Rhaetic existed too. Possibly there’s some relationship between Basque and Rhaetic, but I wouldn’t bet on it. It seems pretty certain that IE replaced multiple families, then.

(And that’s not even considering the IE languages spoken outside Europe!)

[zompist]

But we do-- we can see the Indo-European, Sino-Tibetan, and other families taking over huge territories over recorded history, and what we know of those areas from say 5000 years ago is that there were more families there. Where Europe is now almost entirely IE, and the Middle East is almost entirely Semitic, there were multiple language families or isolates.

As far as I know, Indo-European had taken over its current areas in Europe and India quite some time before recorded history (a pleonasm?).

No; as I said, it was during the period of recorded history. Around 3000 BCE IE had not yet taken over Spain, Italy, Greece, Anatolia, Iran, or India.

And I really wonder how we actually know "there were more families here", as the only non-IE language we have in Europe is Basque.

You're forgetting Etruscan, Iberian, and Tartessian, which are well attested. Pelasgian and Eteocretan are known from place names (and Greek testimony). There are still questions about Pictish and Rhaetic.

There was a huge drive in the 1980s/90s to reduce the number of language families, possibly even down to one. A lot of this work was sloppy and there's been a backlash— note the recent discussion here casting doubt on Afroasiatic. So the "everything non-IE in Europe is Basque" idea has not held up well.

And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.

It's the only mechanism we see in actual history for a language family taking over a large area. And there's a nice natural experiment: the Americas, where empires were limited in extent, and there's an enormous number of language families and isolates.

[Travis B.]

And don't forget the Uralic languages (whether Finnic, Sami, Ugric, or like) either.

[Raphael]

And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.

It's the only mechanism we see in actual history for a language family taking over a large area.

Huh? What about imperial conquests as a result of warfare between different agricultural societies?

[Richard W]

Around 3000 BCE IE had not yet taken over Spain, Italy, Greece, Anatolia, Iran, or India.

And it still hasn't totally taken over Spain, Anatolia, Iran or India - in fact it seems to have lost ground in the middle two.

And I really wonder how we actually know "there were more families here", as the only non-IE language we have in Europe is Basque.

And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.

It's the only mechanism we see in actual history for a language family taking over a large area. And there's a nice natural experiment: the Americas, where empires were limited in extent, and there's an enormous number of language families and isolates.

What's the connection between agriculture and IE before this millenium?

[WeepingElf]

But we do-- we can see the Indo-European, Sino-Tibetan, and other families taking over huge territories over recorded history, and what we know of those areas from say 5000 years ago is that there were more families there. Where Europe is now almost entirely IE, and the Middle East is almost entirely Semitic, there were multiple language families or isolates.

As far as I know, Indo-European had taken over its current areas in Europe and India quite some time before recorded history (a pleonasm?).

No; as I said, it was during the period of recorded history. Around 3000 BCE IE had not yet taken over Spain, Italy, Greece, Anatolia, Iran, or India.

I think jal meant "before recorded history began in the relevant regions". Sure, recorded history began about 3000 BC in Egypt and Mesopotamia - but not in Europe or India.

And I really wonder how we actually know "there were more families here", as the only non-IE language we have in Europe is Basque.

You're forgetting Etruscan, Iberian, and Tartessian, which are well attested. Pelasgian and Eteocretan are known from place names (and Greek testimony). There are still questions about Pictish and Rhaetic.

There was a huge drive in the 1980s/90s to reduce the number of language families, possibly even down to one. A lot of this work was sloppy and there's been a backlash— note the recent discussion here casting doubt on Afroasiatic. So the "everything non-IE in Europe is Basque" idea has not held up well.

Indeed not! Neolithicization of Europe, was, as Haak et al. have shown in 2010, by immigration of farmers from the Near East, which would have involved the spread of a small number of language families (maybe just one, but the Mediterranean and Danubian expansion routes may easily have involved different families), but some hunter-gatherer languages may have survived, perhaps in the mountain ranges, for quite a while.

It is controversial whether Iberian was related to Basque or not. There are some tantalizing similarities, but so far Basque hasn't really been helpful in understanding Iberian. If both descend from the language of the Neolithic farmers who came to Spain about 5000 BC or so, Roman-Era Basque and Iberian would be related about as closely as two modern IE languages of different branches (e.g. German and French): some resemblances would be noticeable, but one language wouldn't help much in understanding the other - which is exactly what we can observe. Southwest Hispanic ("Tartessian") has been claimed by J. T. Koch to be a Celtic language, but most relevant scholars doubt that. It may be another Neolithic farmers' language related to Basque and Iberian, or a surviving hunter-gatherer language.

Of Etruscan, we don't know whether it is native to Italy or came later. There is a hypothesis that it came from northwestern Anatolia about 1200 BC, which has been supported by Beekes and Kloekhorst, and the myth that Rome was founded by descendants of refugees from Troy may reflect this.

Greek is indeed full of loanwords from unknown languages (may have been more than one), and AFAIK nobody has convincingly shown them to be cognates of words in some other known language.

And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.

It's the only mechanism we see in actual history for a language family taking over a large area. And there's a nice natural experiment: the Americas, where empires were limited in extent, and there's an enormous number of language families and isolates.

Languages of course can spread by elite takeover - Latin/Romance is a case in point, and Turkic appears to be yet another. Perhaps also Celtic and Italic, which IMHO date from the Urnfield expansion. (While the Bell Beaker people have been shown to be of steppe origin and therefore probably spoke an IE language, that language, which may underlie the Old European Hydronymy, doesn't seem to be the ancestor of Italic and Celtic because it apparently merged PIE *o with *a, which happened neither in Italic nor in Celtic.)

[zompist]

And I have still doubts about whether there's a causal connection between agriculture and language diversity, especially a universal one.

It's the only mechanism we see in actual history for a language family taking over a large area.

Huh? What about imperial conquests as a result of warfare between different agricultural societies?

Yes, of course, I thought this was obvious from the discussion. Agricultural empires are very good at spreading languages.

[Raphael]

Huh? What about imperial conquests as a result of warfare between different agricultural societies?

Yes, of course, I thought this was obvious from the discussion. Agricultural empires are very good at spreading languages.

Sorry; I had the wrong impression that you were saying that only the replacement of hunter-gatherers by agriculturalists works out that way.

[Tropylium]

It's that time of the year again... I have heard multiple people pronounce pączki as [ˈpʰũʔtʃci(ː)] as of late. How in the world do you get from [ɒ̃], which would be close to the original Polish, to [ũ] of all things is beyond me...

Spelling pronunciation from "punchkey", i.e. the "P[u]njab" mechanism?

And don't forget the Uralic languages (whether Finnic, Sami, Ugric, or like) either.

The European range of Uralic postdates all the main IE spreads and has arrived at least to some extent over early Indo-European, in the south at least. Some of the higher latitudes, e.g. inner Finland, northern Komi or most of the Nenets Autonomous Okrug, do not even seem to have ever been primarily IE-speaking; in Finnmark, Sami has probably only spread over former Paleoeuropean languages, not bronze-age Germanic. (Russian, Norwegian, Swedish, Latvian, German etc. are all yet newer arrivals, of course, and Hungarian is the only case that happens recent enough to be during local recorded history.)

[chris_notts]

https://chrisintheweeds.com/2024/01/24/ ... rsive-nps/

"In any case, one interesting thing about Yimas is that it has two kinds of noun phrase structure, a tightly integrated but non-recursive structure, and a much looser, more appositional structure."

I just noticed that my Coastal Marind grammar describes an almost identical constraint, where a single prenominal modifier is permitted and multiple modifiers require a looser appositional structure. Coastal Marind is odder though, because the pre-nominal modifier slot is not open to adjectives, which instead must compound with the head noun to form a single phonological word.

Coastal Marind is Papuan but not closely related to Yimas as far as I know, so it may be an areal feature? Marind also has right edge stress, so my proposal that the Yimas version might be prosody related doesn't really work for Marind.

Edit: Marind also lacks integrated relative clauses, using non-embedded generic subordinate clauses for that purpose, so it has a very impoverished, mostly non-recursive NP structure. Possibly the only integrated recursive NP modification is possession (it's not clear to me, but possession looks integrated and not loose/appositional).

[bradrn]

Coastal Marind is Papuan but not closely related to Yimas as far as I know, so it may be an areal feature?

I doubt it — Marind is in the south, Yimas is in the northeast. There’s a big mountain range between the two.

[chris_notts]

Coastal Marind is Papuan but not closely related to Yimas as far as I know, so it may be an areal feature?

I doubt it — Marind is in the south, Yimas is in the northeast. There’s a big mountain range between the two.

Oops! Still odd that the two cases I know of are both Papuan though.

[abahot]

How many languages have ever been spoken? There are seven-thousand-odd spoken today, roughly, but I'm wondering if we have an estimate for how many have been spoken at any point in time. I know this question is not very well-defined but even still there might be some way to put a rough number to it. Maybe a better question is to ask how many languages have ever died.

[jal]

Maybe a better question is to ask how many languages have ever died.

We can't possibly know, can we? We don't even know when humans started speaking, and when the first fully fledged, grammatically complex languages evolved. Some estimates are at least 100,000 years, which is 10 times the window into which we can currently look.


JAL

[Linguoboy]

How many languages have ever been spoken? There are seven-thousand-odd spoken today, roughly, but I'm wondering if we have an estimate for how many have been spoken at any point in time. I know this question is not very well-defined but even still there might be some way to put a rough number to it. Maybe a better question is to ask how many languages have ever died.

By some odd coincidence, I happened to be reading the chapter of David Crystal's Walking English last night where he tosses out an estimate of 150,000. I don't have the faintest idea how this estimate was arrived at though.

[Moose-tache]

Here's my *extremely* bar napkin-y estimate.

We start with Australia, using reasonable estimates of the number of pre-contact languages. We establish roughly how many languages are spoken in different parts of the continent, and end up with a figure of one language every 15,000 square kilometers in high productivity areas and one languages every 200,000 square kilometers in low productivity areas. If we then assign A and C climates the former status, and B and D climates the latter, we end up with a global total of 4032, lower than the number of languages attested today.

Like I said, this is not likely to be very accurate (it could easily be off by an order of magnitude). But it's not so crazy when you think about it. The highest density of languages by area is usually in agricultural areas like West-Central Africa or the New Guinea highlands. If there is a minimum population for a language community to maintain itself as a separate language, it's easier for farmers to reach that minimum on a smaller amount of land. Maybe the peak of language density is somewhere between the neolithic revolution and the bronze age? Just a possibility. Mobility, disease, and marriage practice would all have an effect as well.

Plus, we haven't really defined our terms very well. Density of actual languages is not the same thing as linguistic diversity between languages. And sweeping replacements can be observed at all technological levels, including Carib and Pama-Nyungan expansions among hunter gatherers. Australian languages are famous for violating some previously-held "universals," but if for example Dyirbal swept the continent replacing every other Australian language, it's not like each language lost would take with it a unique set of universal-violations, because some Australian languages are unusual in the same ways. What exactly are we measuring when we talk about loss of diversity? Is it quantity of mutually-unintelligible speech varieties? Is it the amplitude of linguistic uniqueness of each language? Is it something else entirely?

[abahot]

Maybe the peak of language density is somewhere between the neolithic revolution and the bronze age? Just a possibility.

This is actually a really interesting question about language density that I hadn't thought of! That sounds roughly reasonable. Although perhaps it occurred in places on the brink of an agricultural revolution, when hunter-gatherers started to become sedentary, increasing their population before large expanding groups started to displace neighboring peoples. I do not know much about this area, so maybe this is completely misguided. But it does seem reasonable that we should look for high population density without sweeping demographic changes.

[WeepingElf]

I am no expert on this, but I did spend some thought on the linguistic diversity of prehistoric Europe. In precolonial North America, which is perhaps best comparable to Mesolithic Europe, linguistic diversity (which is perhaps better measured in number of stocks, by which I mean language families with a time depth of up to 5,000 years, than individual languages) is highest in California and lowest in the Arctic. North America has, by conservative classifications, about 50 stocks and isolates, so Mesolithic Europe, which is about half the size, may have had up to about 20 or so. Their number would have been largest in the Mediterranean, where each of the big peninsulas may have housed 2-4 stocks; add to this a few isolates on the islands. North of the Alps, it would perhaps have been 3-6 stocks in the entire area from France to southern Russia. The area north of ca. 50°N would have only rather recently been settled by humans because it had been glaciered before, so there would be only 1 or 2 stocks there, which would not even add to the total number of stocks in Europe, as they were just extensions of stocks further south.

The Neolithic revolution, which we now know to have been by immigration of farmers, would have brought in 1 or 2 new stocks, while most Mesolithic languages would have disappeared, leading to the extinction of some stocks and the reduction of the remaining ones to a few language pockets in residual zones such as mountain ranges. So by the year 4000 BC, linguistic diversity in Europe would have been greatly reduced. Then, of course, came the Indo-Europeans, and we know, more or less, what happened.

[jal]

But do we measure diversity by the "stock" or by the language? Is an area less diverse when it's languages have diverged, say, 4000 years ago but there's 100 languages now, than the same area that has 10 languages that diverged over 6000 years ago?


JAL