Laura Spinney, a British science journalist, has come out with a book called Proto: How One Ancient Language Went Global, about my former area of specialization, Proto-Indo-European. Not having seen it, I can’t say how accurate it is, but Laura Miller’s Slate review makes it sound like Spinney has been keeping up with recent developments, anyway:
It’s astonishing how much we’ve discovered about these languages that have gone unspoken and unheard for millennia. In the past two decades, new DNA analysis technologies, combined with archaeological advances and linguistics, have solved many mysteries surrounding the spread of the Proto-Indo-European (or PIE). For example, Anatolian, a now-extinct group of languages, was once thought to be the earliest offshoot of PIE, the first instance in which a new language split off from the mother tongue. But in recent years, genealogical analysis of human remains from the period shows no genetic connection between the people who spoke the Anatolian languages and the Yamnaya, a people of the Pontic–Caspian steppe region north of the Black Sea—now considered the source of PIE. The presiding theory now is that Anatolian isn’t the daughter of PIE, but its sister, with both being the products of an even more ancient lingua obscura. […]
Fortunately, Spinney is a stylish and erudite writer; it’s the rare science book that quotes Keats, Seamus Heaney, and Ismail Kadare. She also has a keen sense of the romance of her subject. Her vivid scene-setting takes us from the vast, grassy steppes where the nomadic Yamnaya grazed the livestock whose meat and milk made them exceptionally tall and strong to the perplexing Tocharian culture on the western border of China—whose capital was regarded by the Chinese as filled with “heavy-drinking, decadent barbarians,” famed for its dancing girls and “the flock of a thousand peacocks upon which its nobles liked to feast.” This latter culture—and not Sanskrit, as was long thought—may even be the source for the English word “shaman.”
Spinney illuminates the way that languages reflect the material reality of the world in which they are spoken. “Hotspots of linguistic diversity,” she writes, “coincide with hotspots of biodiversity, because those regions can support a higher density of human groups speaking different languages.” These are the places where the speakers of different languages are most likely to borrow words from each other, leaving clues to their encounters for later generations of scholars. […]
Genetic evidence has also revealed that while the Yamnaya did not venture all that far from the steppes where they domesticated horses and ate tulip bulbs, their more aggressive successors, the Corded Ware Culture (named for their distinctive style of pottery), carried the PIE languages all the way to the Atlantic Ocean. In much of Europe, this advance resulted in, as Spinney writes, “an almost complete replacement of the gene pool,” in particular the male chromosome. The Corded Ware men “had bred with local women and prevented local men from passing on their genes,” she explains; “Rape, murder, even genocide could not be ruled out.” However, a group of Danish scientists now believe that the replacement was not necessarily intentional—that plagues swept through Europe in the Late Neolithic period, diseases to which the newcomers from the steppes were resistant. In a related mystery, the population of Ireland is one of the few in Europe that has been genetically consistent since the Bronze Age, yet somehow Ireland also adopted (and still strives to preserve) Gaelic, its own Indo-European language. Usually genetic and linguistic change go hand in hand, but in this case, not.
Multilingualism predominated in the ancient world, where you might need different tongues to chat with your neighbor, perform religious rituals, and trade with the metal workers upriver. Monolingualism is a modern phenomenon, one Spinney links to the concept of the nation-state. Though in the 21st century humans move greater distances even more easily, languages seem to intermingle and influence one another much less than in ancient times. Spinney theorizes that “the desire to belong is as strong as ever, and as it becomes harder to see the difference between ‘them’ and ‘us’, linguistic and cultural boundaries are being guarded more jealously.”
I’m sure that, like any journalist, she makes mistakes when dealing with specialist material, but my heart is warmed by her desire to spread knowledge of PIE, and I like this quote:
“Prehistoric people undoubtedly had identities as complex and multi-layered as ours,” Spinney writes, “but we can be sure that nowhere among the layers was the nation-state.”
Thanks, Bathrobe!
Spinney was here earlier this month, too.
(Not this one).
The reconstructed lexicon of Proto-Indo-European has only about 1,600 words, and at its dawn the language may have been spoken by as few as 100 people—people who didn’t need words for such exotica as, for example, bees
A confusion between what we have the data to reconstruct, and the actual real language these people spoke. I doubt if any human language has a lexicon of only 1600 words.
That’s a tiny number. For comparison, the “Vocabulary” section of my Kusaal grammar – far too brief to qualify as a dictionary, and only there as a sort of language index to the grammar, still has about 1200 lexemes. And it doesn’t list numbers, proper names or unassimilated loanwords. It does have “bee”, though …
Yeah, I left that part out because it was so silly. I expect it’s the Slate writer’s silliness, not Spinney’s, but who knows?
The illustration in Slate is messed up, with one arrow going from PIE to English, and another going from PIE to Old English to Proto-Germanic to German.
Also, “…a matter of comparing syllable sounds and consonant pronunciations…” Shudder.
…but… that’s the exact same thing. No change at all, except in which fork in the tree gets to be called “PIE”. Nomenclature, not science.
The actual history is more complex and perhaps more interesting, with “the Indo-Hittite hypothesis” being widespread in the early 20th century, then falling out of fashion, and then gradually but decisively coming back and now being mainstream but the nomenclature not immediately following suit.
Yes…
…no, why?
Up to a point. I see no reason to think that, for example, most inhabitants of Ancient Egypt weren’t monoglots for millennia.
Slate review:
Isn’t “not Sanskrit” is a bit garbled, since the Tocharian word itself is supposed to be a borrowing from Sanskrit? I’d revise it to something like “the perplexing Tocharian culture may even have been the way station through which a Sanskrit word for ‘ascetic’ was passed on to Siberian languages, and from there eventually into English as ‘shaman’.” (Unless it’s a native Tungusic word.)
There’s an excerpt of Spinney’s book at GBooks. It’s good, as contemporary non-fiction books go (“Thing: How an Obscure Bit Changed Civilization”). Generally accurate (though, “The Middle English of Shakespeare”? Ouch.) She describes the Yamnaya clearly and vividly. What impressed me most was her accompanying archaeologists in the field two years ago, during the war. That must have taken some great deal of will, not to mention the tough logistics.
No kidding!
“But in recent years, genealogical analysis of human remains from the period shows no genetic connection between the people who spoke the Anatolian languages and the Yamnaya, a people of the Pontic–Caspian steppe region north of the Black Sea—now considered the source of PIE. The presiding theory now is that Anatolian isn’t the daughter of PIE, but its sister, with both being the products of an even more ancient lingua obscura.”
I don’t understand what point the “no genetic connection” is supposed to be making. If they’re trying to make some sort of analogy between genetic closeness via DNA and “sister” languages vs “mother-daughter” languages, it doesn’t make any sense, since siblings share the same amount of DNA as parents and children do. What implications are we supposed to actually draw about language from the apparent genetic unrelatedness of the Anatolians and Yamnaya?
> In a related mystery, the population of Ireland is one of the few in Europe that has been genetically consistent since the Bronze Age, yet somehow Ireland also adopted (and still strives to preserve) Gaelic, its own Indo-European language.
Maybe I’m being stupid here, but this seems like no mystery at all?
Corded Ware Culture (wikipedia tells me) lasted “from the Late Neolithic, through the Copper Age, and ending in the early Bronze Age.” So sounds like CWC arrived, brought indo-european lang(s) to Ireland, and things have stayed the same since? (“Same” from a very high-level view where “still speaking IE languages” counts, I guess.)
…okay wait as i typed this I remembered discussion from…some LH thread. Is the problem that we think Proto-Celtic is much more recent, so the original CWC migration cant have been (proto-)Gaelic speaking… Well, thats still not much of a mystery since apparently parts of the island switched to Germanic languages while staying “genetically consistent”. Its still got nothing on, say, Basque
====
Separately, I think i might also be confused about the claim “high linguistic diversity coincides with high biodiversity”. I guess it is true in New Guinea, but is it true elsewhere? And just in general, to any commenters who feel they do understand/agree with that claim, i would appreciate elaboration.
Apropos, I recently read Kadare’s first novel, Qyteti pa reklama, which was written in Kadare’s student days in 1959 but left unpublished for decades. It never got an English translation, and I suspect one reason is that the plot hinges on some details of the history of the Albanian language that are little known to audiences abroad. It’s a pity, because I could see its appeal to the crowd here. In one comic passage, the philologist narrator is willing to do anything to get out of the backwater he has been assigned to as a schoolteacher, “even embark on an Albanian etymological dictionary”.
FWIW, Nugteren 2020 (“Linguistic aspects of the term shaman in northern Eurasia”) argues for an origin in Tungusic.
the plot hinges on some details of the history of the Albanian language that are little known to audiences abroad.
Makes me want to learn Albanian! (A few decades ago I might actually have done it…)
As I may have previously mentioned, the school district where I live has in recent years started making its generic communications-to-parents available in all of English, Spanish, and Albanian, demographic changes having led the authorities to conclude that Albanian was the second-most-likely language to be understood by LEP parents of students in the district. I regret to say that Albanian is not (yet) offered in the formal school curriculum as an alternative to learning Spanish or French etc., but the town library recently offered a low-key sounding “Albanian for Kids” class taught by a high school student.
Separately, I think i might also be confused about the claim “high linguistic diversity coincides with high biodiversity”. I guess it is true in New Guinea, but is it true elsewhere? And just in general, to any commenters who feel they do understand/agree with that claim, i would appreciate elaboration.
Speaking from very little knowledge, I’ve read that linguistic diversity is high in mountainous regions—at least the Caucasus as well as New Guinea—and mountains certainly correlate with biodiversity because of the different climates.
People say “biodiversity” like it’s a good thing when another way to characterize what they are (sometimes) talking about is fragility/vulnerability resulting from some random species of butterflies or lizards or what have you having a very constrained geographical range and thus being at risk of extinction if anything goes awry with the factors in that localized area that had previously made it hospitable. You can make metaphorical analogies to languages, of course, but humans as a species have a quite extended geographical range (like pumas or coyotes or what have you) because they’re pretty adaptable to different ecological niches. Whether they bring a given language with them when one localized ecological niche becomes untenable may be a different story, of course.
Nigeria would be a good test case for whether this is particularly a mountain thing. It has its mountains, but it’s not like the Caucasus or New Guinea. It’s certainly very linguistically diverse; I don’t actually know if it shows particularly high biodiversity (or what would be a sensible point of comparison.)
Many parts of West Africa are more linguistically diverse than the Caucasus, area for area (including savanna-zone areas which are probably far from front-runners in terms of biodiversity.) I suspect the trope about the Caucasus owes quite a bit to the striking contrast with Europe.
Talking of which, the lack of linguistic diversity in much of the world is actually a pretty recent development, to an extent that probably erases any correlation there may ever have been with biodiversity. Modern nation-states are probably good at erasing biodiversity too, though, which might restore a correlation.
Amazonia is (as yet) another locus of great linguistic diversity. That one does seem likely to correlate with biodiversity.
(Indigenous) Australia might or might not count as a counterexample, depending on how you measure the actually somewhat slippery notion of linguistic “diversity.”
Other things that correlate with biodiversity on land are wet climates and warm climates, I believe.
And yes, it would be good to know whether that coinciding is supposed to be before or after language suppression by nations, or both.
ETA: If I remember correctly, to compare the biodiversity of regions of different areas, you divide the number of species by the square root of the area.
This rather pretty map purporting to show world diversity among vertebrates
https://upload.wikimedia.org/wikipedia/commons/c/c0/Map_latitudinal_gradient_of_biodiversity_mannion_2014.png
doesn’t (to my eye) correlate particularly well with language diversity.
In particular, the biggest splodge in Africa seems to coincide quite well with Narrow Bantu; though this comes back to how you propose to measure linguistic diversity. Sure, there are lots of Bantu languages, but the internal diversity of the whole group is only about the same as Oti-Volta: and much less than “Gur” or Chadic.
Amazonia and bits of SE Asia light up spectacularly; but not North America – nor New Guinea, particularly.
Again, the current spread of Bantu only happened over the past three thousand years (less, for much of it), and will have probably overrun a previously more linguistically diverse map. But that is all very speculative.
I suspect that a good bit of what correlation there is is actually just the result of historical accident.
Spinney may indeed “have been keeping up with recent developments”, but from what I have read she seems to be uncritical in following recent trends/fads. Accompanying archeologists on digs might seem to be admirable, but as Max Müller put it well over a century ago, skulls do not speak. Neither do archeological artifacts, and neither does DNA. This is a fact all too many linguists, archeologists and geneticists seem to wish would go away.
And as a journalist whose first and foremost priority is capturing her readers’ attention, she really does not need to examine/evaluate recent controversies in the field: the spread of Indo-European is one of those uncontroversial historical events which is so fascinating in and of itself that sticking to the accepted facts would in no way make her book less interesting.
“Spinney may indeed “have been keeping up with recent developments”, but from what I have read she seems to be uncritical in following recent trends/fads. Accompanying archeologists on digs might seem to be admirable, but as Max Müller put it well over a century ago, skulls do not speak. Neither do archeological artifacts, and neither does DNA. This is a fact all too many linguists, archeologists and geneticists seem to wish would go away.
And as a journalist whose first and foremost priority is capturing her readers’ attention, she really does not need to examine/evaluate recent controversies in the field: the spread of Indo-European is one of those uncontroversial historical events which is so fascinating in and of itself that sticking to the accepted facts would in no way make her book less interesting.”
Who gets to decide what’s an accepted fact versus a recent trend/fad?
Hippophlebotomist: Well, there are several accepted methods. If recent “cutting-edge” “discoveries” fail to account for objections made generations ago against similar such “discoveries” back in the day -like Max Müller’s, quoted above- it is reasonable to be suspicious. When we observe a mismatch today -okay, mismatches, and quite a lot of them, actually- between genetics and language classification, “scholarship” which takes it for granted that genetics and language families must have matched perfectly in the past would deserve to be taken CUM GRANO -okay, MONTE- SALIS.
Another red flag: when the *only* recent scholarship quoted involves applying new techniques to hone, refine and perfect the party line -err, I mean, the scholarly consensus. If recent scholarship which casts massive doubt (when it does not wholly refute) generations of earlier scholarship is unquoted, when its very existence seems unknown, then, again, skepticism seems appropriate.
(One of my favorite examples of the latter: A French Indologist has made an excellent case recently that the accepted archeological consensus regarding the expansion of Indo-Iranian into Central + South Asia and the Middle East is probably…utter nonsense. Why? Because, based on everything which we know -via the earliest texts-of the material culture of Proto-/Early Indo-Iranian speakers, we would expect early communities of speakers of these languages to have left NO tangible evidence of their presence detectable by archeologists today. The implications for archeological research seeking to identify Proto-Indo-European speakers/communities ought to be obvious).
How can a community leave no evidence at all of its presence?
Do you mean, no evidence that could be interpreted as specifically pertaining to that particular community? I can certainly imagine that.
David Eddyshaw: The latter: Early Indo-Iranian communities were nomadic, trading extensively with any (probably all) non-nomadic communities they came into contact with, and as a result we would not expect there to exist any archeological marker today which would unambiguously point to Indo-Iranian speakers. Since Indo-European, like Indo-Iranian, spread over territory which had already been inhabited by humans long before the Indo-Europeans’/Indo-Iranians’ arrival, I strongly suspect no specifically Indo-European archeological marker exists either.
The scholarly article in question is Gérard Fussman’s chapter in this book: https://www.amazon.fr/Aryas-Aryens-Iraniens-Asie-Centrale/dp/2868030726
“scholarship” which takes it for granted that genetics and language families must have matched perfectly in the past”
I don’t think that’s a fair characterization of the most recent archaeogenetics research that comments on linguistic connections. There’s plenty of work here that’s explicit about the fact that Uralic speakers, Basques, and the ancient inhabitants of Etruria all have considerable steppe ancestry but are non-Indo-European speaking, while the genetic impact of the steppe in areas such as Greece is much smaller. I don’t think this is something that gets swept under the rug in the actual studies that investigate these populations.
“Because, based on everything which we know -via the earliest texts-of the material culture of Proto-/Early Indo-Iranian speakers, we would expect early communities of speakers of these languages to have left NO tangible evidence of their presence detectable by archeologists today.”
This is a point where I think the perception of what archaeology is capable of finding lags behind the actual state of the field. The increasingly widespread adoption of techniques such as micromorphology, sedimentary DNA, and wider comparative isotopic datasets, on top of refinement in survey methodologies mean that things that were conventionally seen as invisible to the classical excavation approach have come into focus. Work by scholars like Shishlina has shown that a lot more about these otherwise elusive pastoralists is knowable than we used to think. None of these can be directly tied to whatever languages these peoples may have spoken, but it’s less of a fool’s errand than it’s sometimes made out to be.
Re IE and Anatolian, I think Spinney’s point, whether made clearly by either her or the reporter, is that Yamnaya appears to belong after, not before, the split. Of course this has no direct bearing on the degree of relationship between the languages.
The original 2015 Nature paper on the genetics of (to put it as carefully as possible) people buried in Corded Ware contexts already pointed out that the highest percentage of Yamnaya ancestry among people alive today is found in the Estonians.
Fair enough.
“Early Indo-Iranian communities were nomadic, trading extensively with any (probably all) non-nomadic communities they came into contact with, and as a result we would not expect there to exist any archeological marker today which would unambiguously point to Indo-Iranian speakers.”
I’ll have to track down a copy of the work in question to evaluate the argument, but since that book was published we’ve gotten a lot more data from firmly dated pastoralist sites like Ojakly with Corded Ware-derived ceramics and buried individuals whose ancestry is primarily from populations found buried in Corded Ware contexts in Eastern Europe (See Narasimhan et al 2019). They’re clearly interacting with local BMAC groups, but are still discernible as a distinct community, e.g.:
“Despite some overlaps in the domestic animal species utilized at Ojakly and at nearby farming-focused sites in the Murghab, there is a clear contrast in terms of the subsistence focus and practices, beyond what would be expected if these groups were specialized economic sub-sets of a single socio-cultural tradition.” Faunal remains from Ojakly, a Late Bronze Age mobile pastoralist campsite in the Murghab region, Turkmenistan – (Rouse, Woldekiros, & Cerasetti 2022)
In critiques of the steppe hypothesis old and new, there’s been a lot made of the fact that the Andronovo influence seems to not reach beyond Central Asia into Indo-Iranian regions of the Iranian Plateau and South Asia, but even this is called into question by recent finds:
”Another thing to bear in mind is the cycle of population movements and the interaction between Central Asia (Oxus civilisation and Andronovo culture) and Iran during the 2nd millennium BCE (Luneau 2019). […] The ceramic assemblage of Estark-Joshaqan consists of grey ware pottery and light reddish pottery which is hand made with punch or comb decoration similar to the Central Asian Andronovo group. Javad Hosseinzadeh’s preliminary studies indicate that traces of Andronovo ceramics are visible not only in Estark-Joshagan but also in Saram and Qoli Darvish. Therefore, it is meaningful to assume movements of population and interaction of populations in all parts of the central plateau with those in Central Asia, north-east Iran, north Iran and north-west Iran during the 2nd millennium BcE. […] We should also note that there is some new archaeological evidence, both from pottery decoration and style and from mortuary practices, that shows some similarity with those steppe cultures” – “Tappeh Sialk, the Glory of Ancient Kashan” (Nokandeh, Curtis, & Pic, eds, 2019)
It’s possible that the arrival of Andronovo style ceramics and burials, DOM2 horses, and genetic ancestry in Iran is completely disconnected from the fact that this is when Iranian languages show up in the textual record (at least per “Linguistic Groups in Iran” by Ran Zadok in The Oxford Handbook of Ancient Iran), but it seems at least worth looking into rather than declaring ignoramus et ignorabimus.
Even good journalists may exhibit the tics and quirks of journalistic style, and I think upon rereading that the odd phrase “the perplexing Tocharian culture” seems to be one of those. It sounds zippy at a superficial level, but makes no apparent actual sense. Does the journalist (this is Miller rather than Spinney’s phrasing) think “perplexing” is a synonym for “understudied” or “imperfectly understood due to limited data” as opposed to “puzzling” or “baffling”? Especially if we are talking about the culture rather than the language, what’s “perplexing” about it, in terms of having features that are peculiar or highly unexpected for the time and region? Admittedly, hat has previously linked to a blogpost entitled “Indo-European Etymological Dictionaries for the Perplexed: Tocharian Languages,” but I think that’s a different usage or different focus (plus a literary allusion) that sort of treats the perplexedness as a (potential) feature of the student rather than the language itself.
EDITED TO ADD: Is there actually anything that weird about Tocharian as a language group in a broad cross-linguistic context? Or is it just conventionally thought of as “unusual for an Indo-European language” because it deviates in various regards from many other better-studied IE languages, which is often parsimoniously explained by the ancestral version having supposedly branched off from the main IE lineage earlier than anything else other than Hittite?
I assumed what’s perplexing is not the culture but the location of the language – bad writing, but I hardly noticed.
No.
I assumed what’s perplexing is not the culture but the location of the language
I agree. “Why is Tocharian where it is? How did it get there??” used to be the meme people casually interested in the topic glommed on to.
I noticed it, but passed over it as being at least not as bad as “the mysterious Akkadian language”, which came up in some newspaper article we once discussed.
The location is not so far away from where Sogdian and other Iranic languages were long extant to seem particularly puzzling to me. As you migrate across the steppes, some turn right at Albuquerque and others don’t, innit?
Look, journalists are gonna journalist. There’s no point expecting them to stick to academically approved forms of explanation; they’re going to get stuff wrong, and they’re going to overstate, and they’re going to use silly analogies. But lots more people read journalism than academic literature, so we can only be pleased when their batting average is decent.
Genealogical tree of Indo-European languages
Enigmatic
/ \
Perplexing Mysterious
/ \ / \
Befuddling Baffling Confusing Puzzling
/ \ | / \
Flummoxing Dumbfounding Bewildering Arcane Impenetrable
🙂
But you left out “OURS”.
A true linguist knows that their own language is the most baffling of all.
“Hotspots of linguistic diversity,” she writes, “coincide with hotspots of biodiversity, because those regions can support a higher density of human groups speaking different languages.”
just jumping into the pile here, and echoing DE a little…
there are certain kinds of correlation between centers of biodiversity and linguistic diversity (new guinea, upland southeast asia, etc.), but they don’t have a damn thing to do with what a given “region can support” – that’s just warmed-over biologizing of human culture*. the correlation is a direct result of what james scott calls “the friction of terrain” making certain regions harder for states (and the industries they support) to control, which means that they have been subjected to less systematic clearing and agricultural use (whether for lumber, grain, vegetables, flowers, or industrial inputs) on the one hand, and on the other have been refugia for many iterations of groups of people moving out of state space. that’s also why both often peak separately in mountains, marshes, and (in some areas) steppe – current examples include the biodiversity hot spots in the wetlands and hollers of the largely linguistically flattened u.s. southeast and the linguistic diversity center in the not remarkably biodiverse caucasus.
.
* ever more common, thanks to e.o. wilson’s sociobio spawn, their evopsych cultist buddies, and of course steve sailer and his “Human Biodiversity Institute”, who are more explicit about the shared far-right politics that drive all the various wings of the project.
That reminds me of Johanna Nichols’ spread zone/residual zone thing, which obviously captures something real, however much one might object to some of her details. (I’ve got a copy of Linguistic Diversity in Space and Time somewhere. I remember it as an odd mixture of insightful stuff and implausible special pleading. But I should have such insights …)
@JWB, Tocharian perplexed everyone when they first learned about it.
(1) the neat centum and satem partition (sort of THE most important partition).
https://en.wikipedia.org/wiki/Centum_and_satem_languages
It spoiled it. SPECTACULARLY.
(2) another branch of IE, lots of texts.
One need to be exactly… – hm. Not sure what is English for пресыщенный. Let it be blasé – to feel nothing about it:)
“They might have seemed quite rudimentary to us, lacking adjectives, say, or a fixed word order.”
Hey, I’m going to develop an inferiority complex this way:( There ARE adjectives in Russian, which means I seem only HALF rudimentary. Could she please omit the part about adjectives, so I’d seem truly rudimentary?
@rozele, the context:
she say that
– diversity peaked in the Neolithic
– languages were distributed unevenly
[then this quote, one of the factors that make the distribution uneven – not about the Neolithic, but with the Neolithic in mind]
– then “formation of the first states” diminished the diversity.
“They might have seemed quite rudimentary to us, lacking adjectives, say, or a fixed word order.”
Where’s that from, drasvi? Can’t see it in the article.
Kusaal is even less rudimentary than English on this basis: a clearcut adjective word class, and even less flexibility of word order. Clearly the language of the future.
[The word-order thing is particularly odd. Nineteenth-century language-rankers tended to be quite keen on free-ish word order as a sign of an Advanced Language, what with it being prominent in Latin and Greek and all, the pinnacle of human language evolution.]
@David E.: the percentage of the Smug-Intelligentsia class in Western countries with knowledge of Latin and Greek has plummeted rather dramatically since the 19th century, so it is highly unsurprising that the criteria that class uses when doing language-ranking would have shifted.
Also Chomskyanism.
DE, the book. Spinney’s book. About when humans only were learning to speak.
But I don’t mean to say it is bad. And I even like that she (a) thinks something (b) wrote what she thinks – for texts stripped of the authors’ thought I have WP.
And I don’t expect her thought be same as thoughts of someone who enjoys reading grammars. Or feelings of English speakers which she is referring here to to be like mine:)
But I felt it is natural for a halfling to comment:)
___
My impression is mixed. She worked harder on this book than I ever worked on anything (I’m lazy).
But (1) there is the usual problem as with all popular texts: when they speak about something you don’t know, you want a reference. And some idiot proclaimed that books with references are “scientific” and too clever for normal readers:( Even though references are nothign more than links – not “as references in WP”, as internal links in WP.
(2) she packed facts very densely. I don’t have time to think about anything. You feel as someone scrolling through the feed (newsfeed or social-network) – but this someone intends to skip most of the feed as uninteresting.
“facts” – not “facts” – everything. Jumps from Yerevan to Tbilisi in two lines and the next paragraph she is in South Russia.
Speaking of this, Y and LH said something I disagree with… because it is not true.
Neither Russia nor Ukraine require a great ‘deal of will’ to travel in.
No greater than Israel or a half of the Arab world.
And you know that.
Well, sure, there were Russian archaeologists working in the field, the ones she joined, and they felt safe enough. I’m sure she was guided by them or by other insiders as to what is safe and what is not. Still, for an outsider to go to that part of Russia in 2023 was not a trivial decision. She was there during the Wagner rebellion, a particularly unstable time.
If Derek Bickerton had been correct (though I see no reason to think he was). yer first human languages would actually have been characterised by fixed word order …
https://en.wikipedia.org/wiki/Language_bioprogram_theory
“Adjectives” are basically stative verbs in the sort of creoles Bickerton was interested in. so he might have agreed about that. Kinda.
@Y, perhaps I shoudl not have said “and you know that”.
I think my analogy is good. Israel is more or less safe for tourists, despite constantly finding itself involved in some perfectly hot war – not proxy or something. The whole Arab world is also in some sense at war with Israel. And if Tunisia supports “Palestinians”, Algeria have very specifically also supported Hamas. Will you think the city of Algiers becomes less safe when the fighting intensifies in the ME?
The horror (and the Ukrainian and Gaza wars are same scale) is concentrated, focused.
Weirdly our war works like this.
In Russia when you are not at the frontline, you are in Algiers or Tunis. In Ukraine you’re in Israel.
Anyway, we are not talking about in what cities a lady should not walk alone in evenigns and where she can do it at 1 am, something locals know and foreigners are not sure about. To understand that “safety” in Algiers is not defined by what is happening in Gaza you don’t have to be a local.
The thing is when my ex-wife (who I love dearly) is heading to the Black Sea coast with her children I don’t feel an urge to call and say “go to Montenegro or Egypt instead, there is a war in Russia“. No need.*
But there is of course something which applies to a Western journalist but not to foreigners from some other countries. I don’t know if they can behave as they always did (or even take photos of military columns?) or conversely are not allowed to travel in whole large regions far from the frontline. This definitely means obtaining all necessary permits and approval of her route within Russia.
*Thinking about that – I can’t remember if she went there during the war. Perhaps she did not. And if she did, then I think not without learning first if it is safe, as I would have done myself (because children).
Not with how many people a region can support, but how much isolation it can support – how much it impedes communication.
Neither Russia nor Ukraine require a great ‘deal of will’ to travel in.
You’re seriously claiming traveling in a war zone is no more difficult than anywhere else? Doing archaeology in Ukraine is just like doing archeology in New Mexico? Come on.
Whether the entire territory of a country currently embroiled in a war is a “war zone” in the usual sense depends on the war (and may also depend on the country). Of course, the entire territory of a country involved in a war may be in a state-of-emergency mindset where foreigners doing odd things that they claim are benign (just digging up old pottery and/or genomes, nothing to worry about …) may be greeted with heightened suspicion.
Specifically, foreigners who are citizens of a country which materially supports the other side.
“They might have seemed quite rudimentary to us, lacking adjectives, say, or a fixed word order.”
I took that to mean that lacking adjectives and having a fixed word order would seem rudimentary.
Thanks to everybody who commented on biodiversity—I learned some things.
“say”
@LH, can’t you simply tell if my analogy with Gaza, Israel and Algiers is bad?
Even playing the silly game of imagining “primitive languages” supposedly spoken by early humans who hadn’t yet quite got the hang of this “oral communication” lark, the idea of absence of “adjectives” seems bizarre.
“Hot!” “Tasty!” “Sexy!” “Nasty!” strike me as every bit as likely as holophrases (if not more so) as “Sabre-toothed tiger!” or “Run away!”
I suspect the author has been unconsciously influenced by the doctrine, implicit in a lot of school grammar (and in more sophisticated treatments, too) that language is essentially all about stating propositions, and every other type of language use can be paraphrased into strings of propositions.
@drasvi: Have a look at the map here and look at longer periods (the default shows only the events of last week). While the most dangerous areas in Ukraine obviously are near the front line, there is no region where one is safe from drone strikes. So your analogy holds for Russia – it’s big, and there are areas far away enough from the front lines that they are not in danger, and even nearer to the front lines it seems from what I read that Ukraine doesn’t target civilian infrastructure as indiscriminately as Russia does (still, targets on the Russian shores of the Black Sea have been hit and it’s not an area I would choose for my family vacation right now). But your analogy doesn’t hold for Ukraine; going anywhere there is a calculated risk much higher than visiting Algiers or Amman.
(And I say that as someone who has lived in Lebanon during times of conflict. In general, it looked more dangerous from abroad than from in the country, but that doesn’t mean there was no risk – at one time, a bomb went off at a place about 15 minutes after I passed it, and another time a bomb went off at a point I passed every day, even if a couple of hours before I would have used that route.)
@Hans, I compared Ukraine to Israel, not Algiers.
Algiers is Russia:
they support Hamas, but this support affects Palestinians and Israeli people, not them.
Russian cities are not affected either.
____
I also do not mean that Israel is receiving “as many” bombs, rockets etc. per a citizen as Kiev did in 2023.
I’m willing not to compare Ukraine-2023 to Israel-2023 – together with 7th October. That’s simply not what I mean.
What I mean is that both for Israel, Kiev and most Ukrainian cities the amount (2023 Ukraine and I don’t know which Israeli year we’re picking) is insufficient for … hm. For that you could rationally base your decision on it as the most important factor when deciding whether you should visit this city. Something like that. And for speaking of “great deal of will”
Thanks everyone for the comments about linguistic and ecosystem diversity, you raised a lot of interesting points!There’s a lot I want to respond to but I can’t keep it in my head all at once, so I may re-read the thread and come back with a second comment.
@J.W. Brewer
> People say “biodiversity” like it’s a good thing when another way to characterize what they are (sometimes) talking about is fragility/vulnerability resulting from some random species of butterflies or lizards or what have you […]
Yes, thank you! That’s one of the things I had in mind when I asked. The book Why Big Fierce Animals Are Rare is probably very out-of-date as an ecology primer (and probably was even when I read it as a teenager around 2010) but it really opened my eyes to this sleight-of-hand, or at least alternate way of looking at the environment. I grew up hearing much praise for the diversity of, say, rain forests and coral reefs, and while I don’t think anyone said so explicitly I had come to equate species diversity with strength. WBFAAR has a big section describing tundra ecosystems’ ability to bounce back from drastic changes, and how this could be seen as a kind of “strength” despite its relative paucity of species.
(Of course, WBFAAR was published in 1978, and I don’t remember if it discussed climate change. Nowadays there’s a lot of concern that arctic ecosystems will not, in fact, bounce back from the current drastic change…)
@David Eddyshaw
> Amazonia is (as yet) another locus of great linguistic diversity. That one does seem likely to correlate with biodiversity.
…I was also thinking of the Amazon, but I have no idea if it is (or was) linguistically diverse compared to other ecoregions in South America. And of course as rozele and others point out, the last 500 years of history would have drastically muddied the picture…
The third type of ecodiversity that came to mind for me oceanic — coral reefs & ocean upwelling zones. But of course you don’t have human languages spoken there. I do wonder about adjacent land?? E.g. was California + PNW more linguistically diverse thanks to the biologically rich upwelling zones just off the coast? I think I’ve read that this ecosystem allowed for much greater population density than other hunter-gatherer societies, but I can’t remember where I read that… okay, this reddit thread say that the West Coast of N. America was a lot more diverse than the East Coast and says it’s due to older settlement — citing Johanna Nichols as also brought up in this thread. I should read Nichols, and James C. Scott as rozele brought up….
I also definitely take rozele’s point that this seems driven by state capacity rather than ecological allowance, I was vaguely thinking of that too (I haven’t actually read James C. Scott himself but I’ve read a lot of discussion of his books on the internet 😛 ). But if Spinney’s interpretation of *why* the correlation arises is wrong and it’s instead downstream of rozele’s argument, well, it’s still a correlation.
…if it *is* a correlation. Looking at the map on Wikimedia that David Eddyshaw linked, I’m struck by the drastic difference between New Guinea vs. Borneo or Sumatra. A quick google search tells me that all three of these islands are “very linguistically diverse” but I don’t get any sense of the comparative scale between the three. (I also take note of David Eddyshaw about the Caucasus being very linguistically diverse compared to Europe but not so extraordinary compared to West Africa.) What about the highlands of Peninsular Malaysia, which are even more biologically diverse on that map than all three islands??
Found a map of the languages of New Guinea. The language density along the north coast of the island (vs. the interior and south coast) seems to kind-of line up with the biological diversity map, but the east vs. west differences in the biological diversity map (especially the Papuan Peninsula) doesn’t seem to show up. However, this is just me quickly eye-balling things, and I really hesitate to take a map of languages at face value. I’m sure this is a simplification of complex linguistic situations on the ground, and the decision of where to draw boundary lines (languages vs. dialects!) is famously subjective.
…maybe I should similarly hesitate to take a map of “biological diversity” at face value?? I know much less about biology than I do about linguistics (not that I know much about linguistics compared to most of the commenters here, either).
I should read Nichols
Linguistic Diversity in Space and Time surely deserves its reputation as groundbreaking, and I’d certainly say it’s worth reading, even though I have some issues with it myself.
I suspect it would be fairly heavy-going unless one had a bit of background linguistic knowledge, but I daresay you do.
My favourite book of hers is actually her Ingush grammar, which is freely downloadable:
https://archive.org/details/johanna-nichols-ingush-grammar-2011-university-of-california-pdf
So she’s well into credit with me, as not only being a proper descriptive linguist, but also making her work open-access.
I’m pretty sure that there is in fact a correlation, overall, between biological and linguistic diversity, but also that it’s a pretty imperfect one in detail, and that the causal relationships involved are (and always have been) complex and far from static. (Certainly the one form of diversity does not cause the other, of course – it’s all a question of underlying common causes.)
There are certainly both theoretical and practical problems with trying to assign simple numerical measures to linguistic diversity; I imagine that the same its true of the biology, but many Hatters know a great deal more than I do about that.
“I also definitely take rozele’s point that this seems driven by state capacity” – again, Spinney does name states as the factor reducing diversity. I won’t be surprised if she agrees with rozele, she was explaining why before the rise of states there was (hypothetically) more languages there than here. (but even then there were conquests, the explanation is incomplete)
Even before the rise of modern nation states, there were empires which led to the propagation of single languages over wide areas at the expense of other languages; the Roman and Chinese empires are obvious examples, along with Aramaic hanging on the coat-tails of various empires, and Arabic, but Hausa must have been expanding at the expense of its neighbours for centuries, and the relatively large areas occupied by Manding or even Mooré and Gulimancema must have been established with significant loss of prior diversity.
Even before that, you’ve got Nichols’ spread-zone/residual-zone thing. Though IIRC, her idea of an archetypal residual zone is a mountainous area like the Caucasus, and I’m not clear that such things necessarily correlate with high biological diversity. (They do with the inselbergs in the Sahel, but that’s rather a special case, I imagine.)
Refreshing my memory about Linguistic Diversity in Space and Time, one of the major problems I had with it is that it seriously underestimates the diversity within Africa. I think that’s partly because she’s swallowed the Greenberg line about there only being four language families in Africa, and partly a reflection of the actual criteria she selected to assess typological diversity, which are in fact criteria likely to overestimate diversity in the Americas and underestimate it in Africa. (Also a sheer lack of adequate African data, as so often.) Given that her real motivation is in trying to account for a supposed difference in diversity between the Old and New Worlds, I think she comes a bit close to circular argument at times.
Moreover, the relative lack of diversity of much of Eurasia even in premodern times is straightforwardly the result of known historical episodes of invasion and conquest by speakers of e.g. Iranian and Turkic languages – it’s the same kind of phenomenon as with the spread of Romance or Sinitic or Manding or Hausa. Sure, land empires are likely to spread most easily over areas their armies can easily traverse, but lumping this together with her geographical concept of “spread zones” in general seems to be confusing rather different phenomena. And mixing up timescales in a way which undermines any attempt to explain any differences between Eurasian and American linguistic diversity.
In a nutshell, I think a lot of what Nichols attributes to simple geography is really attributable to politics. Politics, in the relevant sense, goes back for millennia – it’s not some modern invention.
@drasvi —
Your comparison of Ukraine to Israel makes me more inclined to agree with languagehat. I have a number of family members and friends who used to travel frequently between the US and Israel but have mostly stopped doing so since October 7th.
(I am aware that there are stereotypes about tourists from different countries having different risk tolerances — particularly Russians and Germans flocking to places that Americans dare not venture. I don’t think that applies here — I’m talking about people whoe already decided one way on that risk calculus, but changed their minds in response to the war.)
(I also think the day-to-day safety in a given city is a different question than the risk you face as an international traveler to that city. The current risk in Haifa or even Sderot seems pretty similar to what it was in 2022, but to get to those places from another country, you have to pass through the international airport, and the risks there have change — or so my family and friends seem to feel, at least.)
@David Eddyshaw —
Thanks for the link! I don’t have much formal training, I only took a single “Intro to Linguistics” class in college, but I loved it and have lurked on here and other linguistics blogs ever since. I sometimes try to read the papers you guys link the comments — with mixed success, but it’s worth a try. I’m looking at the Ingush book now 🙂
(The main issue for me reading nonfiction books these days is honestly just that I tend to get distracted before I get very far into them. Blog comment threads are much better on this front 😛 )
I am aware that there are stereotypes about tourists from different countries having different risk tolerances
When I lived in Nigeria, I was surprised at the insouciance displayed by American colleagues to the question of personal safety, before realising that they were using a somewhat different yardstick from me. (The murder rate per capita in Nigeria is about twenty times that of the UK – or of Ghana – but only about four times that of the US.)
[Slightly unfair: many of the US missionaries I knew in northern Nigeria were, if perhaps a bit strange even by my standards, notably courageous. As in, not backing down in the face of perfectly serious death threats.]
Also a question of what you’re used to, of course. I think I’ve previously mentioned a trainee I knew from South Africa, a country whose violent crime rate makes Nigeria look like Japan, who told me he’d never been so scared there as on a Saturday night out in Glasgow. Fair …
sarah, I chose Israel because it is not particularly peaceful.
I chose Algeria because not many tourists visit it (unlike Tunisia), it does not really try to attract anyone, and I think many people (I among them) simply don’t know if it is safe or not. And while I want to see it (naturally: I’m interested in Maghreb, its people and languages), many I think interpret their “don’t know” as “potentially unsafe” and then “unsafe”. Especially given their strange politics, and (for those who are not Muslims) that it is a “Muslim country uninterested in tourism” and that it was known to be unsafe some 20 years ago.
(The paragraph about Algeria is longer because I don’t think Israel needs any explanation)
I did not intend to present either country as an ideal place – and there is no contradiction.
I only wanted to give some sort of an impressionistic but quantitative estimate. And I don’t think anyone is going to say that a trip to Israel or Algeria must take a “great deal of will”.
‘“don’t know” as “potentially unsafe” and then “unsafe”. ‘ – a common source of confusion, by the way. Everything you know nothing about is “potentially unsafe”. And yes, practically “potentially unsafe” means for you simply “unsafe”. “Safe” is “known to be safe”.
Then people react emotionally. They become afraid of it. Once they’re afraid of it, they treat it as objectively unsafe, converting “unsafe” in the sense “not known to be safe” into something entirely different.
How safe you are is highly dependent on who you are – in every country, but in some, strikingly more than others.
In Nigeria, I was actually very much less likely to be murdered than a Nigerian was. My American colleagues, for reasons that will be immediately obvious, were personally much less likely to be murdered back home in the US than the national average.
The application to some of the other countries that have been mentioned does not really need spelling out.
Sometimes it is illogical. If you’re a tourist in Israel during 7th October, you’re an enemy for a Hamas fighter. But when you (if you’re not Israeli) cross the border to the West Bank you are a tourist in the West Bank, not an enemy.
drasvi, again, when I wrote about “a great deal of will”, I wasn’t thinking that she was in danger of being hit by gunfire or anything that specific. It’s just that it’s a risky place for a foreigner to be. I could think of specific reasons — for example, Americans were arrested and jailed, to be traded later for Russian prisoners. But even aside from that, subjectively and generally, she would be right to worry.
@drasvi: What takes will is highly dependent on personality; for some people it takes a lot of will to leave the house.
In any case, whether one considers Israel as risky as Ukraine depends on the period of comparison. Based on the information that is available to me, the risk of being hit by missiles in Israel was mostly limited to border regions except for a short period between when Hizbollah entered the war and when it was eliminated, because Hamas’s missiles were mostly short-range and primitive. In Ukraine, everywhere is in the range of sophisticated Russian drones and missiles, and while there are areas that are targeted less frequently, nowhere is really safe. Using your analogy, Moscow is Tel Aviv and large parts of Ukraine are Gaza if the risk of being hit is concerned.
I currently wouldn’t visit Ukraine if I don’t have business there and I certainly wouldn’t let my family go there on vacation.
On the east coast, it seems all languages (between New Brunswick and Georgia or so at least) were either Algonquian or Siouan – but the situation still differed strikingly from Europe (all IE, Uralic or Basque for narrower definitions of “Europe”) in that these two families were scattered about at random, so that most languages had at least one neighbor that was drastically different from them.
Species vs. subspecies is exactly like language vs. dialect; the similarities are stunning, and fractal. So, for any map that “simply” counts species, you have to ask about the species concept the map assumed (…if it didn’t just use different sources for different places or different branches of the tree without caring that the different sources assumed different species concepts).
The hotspots by and large remain hotspots across all of that, but, for example, the regions of Mexico with the highest number of endemic bird species are near opposite ends of the country according to two of the more common species concepts, in each case by a pretty large margin.
There is the option of calculating phylogenetic diversity instead. But that’s a lot of work. I don’t think phylogenetic diversity has ever been mapped except maybe on tiny scales.
I think the Slate article is very good. Minor inaccuracies are inevitable when a complex topic is condensed for a lay audience, but there’s nothing that’s plain wrong, preventing understanding, or reflecting a pet idea or prejudice of the journalist.
That was my take as well.
“exactly” – DM, I have doubts about this. You can always make your criteria for classification similar – and thus obtain similar issues. But apart of our criteria, there is the real world, and ALL borders between langauges are blurred to some extent: slightly for an Indo-European and Khoisan speaker who have met a minute ago (even then there is some blurring: it may be easier or harder for a Bushman to explain what she means than it is for a Hani or Huli speaker) and much more so for two Slavs or two speakers of Arabic dialects (who both listen to Levantine songs, watch Egyptian and European films, have or do not have* experience of communicating with speakers of other dialects and on top of that there are conventionalised practices in such communication).
*Thinking of a Tunisian tourist in Morocco who couldn’t understand a word and found herself utterly lost. It would be enough for her to develop some minimal mapping of Moroccan prosody to Tunisian prosody to learn to recognise many words and guess meaning of many others from context.
Well, according to Chomsky we all speak the same language anyway …
However, this may not be an ideal starting-point fof studying diversity …
Just been looking through Linguistic Diversity in Space and Time again; I’d forgotten that it pays almost no attention to genetic diversity, but is almost all about typology. (How far this can be leveraged into conclusions about language prehistory depends on how highly conserved the various features she selects really are. It seems to me that Nichols greatly overestimates the staying power of her typological features, but that’s a whole other issue.)
So she’s not really talking about language diversity in the sense that Spinney means at all.
There’s quite a mismatch between these two kinds of linguistic diversity (in both directions), and the Nichols kind probably doesn’t correlate at all well with biological diversity. So the spread zone versus residual zone thing is likely not very relevant after all.
@Hans, I’m Russian and against the war.
What this does mean is that most or all people here are not even remotely as disgusted and shocked by this war as I am. When they bomb Kiev I feel the same way as I would if it were Tula or Kazan – or as you would have felt if one half of Germany bombed the other. That simple, and without any “but”.
Yes, you can think of many factors that could diminish this disgust and some of them may work for other Russians who are against the war. None works for me.
I’m not a young man who was born when our countries were already independent, everyone in my circle is connected to Ukraine this or that way. So Ukrainians are not abstract. Even if we ingore feelings of Ukrainians, I don’t think that the split of USSR is anyhow “bad” (and generally am not fond of empires and don’t care about states) and not tempted to think of the reality better than it is (as some sort of psychological mechanism). When it comes to human societies I’m a pessimist, I’m accustomed to the thought that we are horrible. On top of that I don’t adapt. I feel today what I felt in 2022, and I will keep feeling it as long as they keep fighting – it is not one of those things I’m going to get used to. I guard my idea of “normality”.
Keep this in mind. I don’t want to downplay the burden of this war for Ukrainians.
The reason why I chose Israel is that it is (1) “known” land for Westerners. It is not Algiers – many Westerners won’t go there even when Algeria is not involved in any war. Algeria is alien. Israel is believed to be intelligible.
Also there is a well developed tourist industry (it is developed in Tunisia as well, but Tunisia is though to be alien).
(2) a land from where you often hear about missiles, wars and things like that.
This example allows me to disentangle feelings caused by missiles and wars and feelings caused by “alienness”
So if Ukraine ranks below Israel, then it ranks below Israel. The list of countries from where you hear about “missiles and wars and things like that” is long – it is somewhere in this list. But other countries in the list are “alien”. People won’t go there anyway.
(and I think you’re – unintentionally – downplaying the Gaza war)
None of us here needs to have either war explained to us. You are attributing thoughts to other people that are invented by you. Ukraine is dangerous, Gaza is dangerous, what is the point in trying to compare the exact degree of danger each presents at any given point in space/time? This all started because you objected to the idea that working in Ukraine as an archeologist was any more difficult or dangerous than anywhere else, which is a very strange position to take and insist on. And let me point out that you’re the one who brought up Israel/Gaza.
Keep this in mind. I don’t want to downplay the burden of this war for Ukrainians.
And I hope I didn’t make it look like you did.
(and I think you’re – unintentionally – downplaying the Gaza war)
I don’t think it’s a contest. And I have no doubt that the human suffering in Gaza is bigger, because there’s hunger on top and the people are trapped. But my point still is that in your analogy, Ukraine is closer to Gaza than to Israel.
It may not sway your opinion, but the company group I work for uses professional risk assesors; we are allowed to travel on business to Israel, but not to Ukraine. So there are people analyzing the situation closely who came to the conclusion that Ukraine is more risky than Israel.
DE, stability of some arealisms (typological or not) must be reinforced by the size of the area, I think…
when the change is random. Becuase people reborrow them.
I don’t know if that works for less random changes.
@Hans, yes, not a contest. And to be clear, I can’t imagine a reason why you’d want to downplay it (I mean, of course there were multiple complaints from Arabs that the Ukrainian war diverts attention from Palestine and then similar complaints about the Gaza war. But it would not occur to me to think that you’d act within this system of coordinates)
David Marjanović: “On the east coast, it seems all languages (between New Brunswick and Georgia or so at least) were either Algonquian or Siouan”
Are you conflating Siouan with Iroquoian?
@drasvi:
I think that what matters may be not so much the size of the area in question, as how common it is within that area for people to be multilingual.
In West Africa, where that is indeed common, there actually is some noticeable convergence between genetically unrelated languages. That may indeed have a sort of stabilising effect overall, in preventing one language from drifting too far out of step with local norms, and indeed in conforming incoming languages to areal patterns (that’s certainly happened with both English and Shuwa Arabic.)
Mind you, the convergence is more in things like idioms, semantic ranges of lexemes and the range of categories that the grammar distinguishes from one another than in (most of} the kind of things Nichols uses in her typology.
I think large-scale very-high-level convergences may be more significant than the things like constituent order, alignment and head-marking versus dependent-marking that Nichols focuses on. (There’s some overlap, though.)
I recall reading something of Edward Sapir’s where he says that American languages don’t really do proverbs, for example, which seem, by contrast, to be a very striking feature of virtually every African language. Dunno if he was right, but hey, Sapir …
Are you conflating Siouan with Iroquoian?
Surprisingly, there were Siouan languages on the east coast too – not on the coast as such, but not far inland.
That said, the pre-Columbian languages of Ohio and West Virginia are so poorly documented it wouldn’t be hard to imagine an isolate or two escaping notice.
Miller, probably summarizing Spinney:
Incidentally, it’s just hours since Dmitry linked to a preprint of the Danish paper in question:
Martin Sikora et al: The spatiotemporal distribution of human pathogens in ancient Eurasia and the emergence of zoonotic diseases. biorXiv preprint (2023):
More specifically, they find that both the plague and louse-born relapsing fever were present at significant rates from a sudden start a little before 5500 BP.
It’s worth noting that these two diseases are believed to be low-hanging fruit — easy to detect in ancient DNA.
But even if the European Farmers and the remaining hunter-gatherers were decimated by epidemics, that would just make the land more open to incoming herders. It wouldn’t explain the gender imbalance in genetic legacy (and the near-complete replacement of the male ancestry).
* The occasion for Dmitry’s post was that the “recent preprint” has been published, but unfortunately not in open access. Pooja Swali et al: Ancient Borrelia genomes document the evolutionary history of louse-borne relapsing fever, Science, Vol 388, Issue 6749 (2025):
The Caucasus overall is biologically very diverse, and with a very high rate of endemism (which is arguably a better metric than species density when comparing to language density).
Most of the linguistic diversity of the Caucasus (in terms of language density) is in Dagestan. Chechnya is rather language-poor, for its size. I don’t know how biologically diverse Chechnya is.
If you look at “how many stably distinct language communities exist in an area”, England is as diverse as Dagestan, even looking only at English varieties. If you ask “how far apart typologically/phylogenetically are the linguistic varieties spoken at one area”, that is a different metric. Phylogenetic diversity can’t be exactly estimated, because you can’t tell if two languages are 20,000 or 100,000 years apart in time-depth.
Vanuatu and the Solomon Islands are very diverse linguistically; I’m just guessing, but I suspect that, as island countries, they are probably not very diverse biologically.
Islands are very prone to endemism. By that measure, the Solomons and Vanuatu are, I imagine, very diverse.
(And like everywhere in the Pacific, were a lot more so before people came.)
“I think large-scale very-high-level convergences”
DE, what is “large-small” and what is “high-low” here?
I mean, over wide geographical areas, and high-level, as in features like discourse organisation and semantic fields as opposed to the details of how such things are actually realised in individual languages.
It’s in this sense, for example, that Hausa and Kusaal, which are completely unrelated genetically, and whose homelands are hundreds of miles apart, nevertheless have enough in common that I found it helpful to refer to Hausa syntax several times in my Kusaal grammar.
Just to take a couple of concrete examples: Hausa aikata “work” and aika “send” are derived from the same root; compare Kusaal tʋm “work, send.”
Hausa and Kusaal both use main clauses in continuing a narrative which have formally subordinate characteristics. (The actual form this takes in the two languages is quite different.)
I could produce quite a few more examples. (Incidentally, both these features appear much more widely in West Africa. The “work”/”send” polysemy even turns up in Bantu, and many other African languages also use subordinate-like clauses in narrative when carrying on the uninterrupted sequence of events, including Wolof and Fulfulde, even.)
@DE, thanks.
I guessed more or less correctly what you meant by “high-level” (though it is difficult for me to think of a scale where your examples are intuively “high-level” and your examples from Nichols are intuitively “low-level”. If those were phonemes…) but its combination with “large-scale”…
@DE, about the size of an area and multilingualism – I meant a langauge area with arealisms… Well, yes, if you (artificially) make it so that all languages in a large area have a value N for a variable X (less artificially this happens as result of conquests – or, for lexical varables, many other processes, e.g. proselytising), there is (as you say) no reason to think that the size of the area will make N any more stable if speakers simply do not communicate to each other. If this is what you meant.
But if what you observe is “N in many (very different) languages in a large area” then N is here since the contact between them.
Well, there are low-level Sprachbund features too in West Africa (and beyond.) The labial-velar consonants are an often-cited example. (Though plenty of West African languages don’t have them, and they are not actually confined to this macro-region, they are still pretty characteristic of the zone.)
A classic paper on all this as it concerns West Africa is
http://www.ddl.ish-lyon.cnrs.fr/Fulltext/philippson/Guldemann_areal_typology_Africa.pdf
(Nichols’ book is referenced, as you’d expect.)
I should perhaps say that a lot of what Güldemann is up to here is undermining Greenbergian lumpery, and there has been some pushback from what I optimistically (and perhaps teleologically) like to think of as the Old Guard.
…Yes, I am, thank you. Interesting that I did this.
@DE, the Negro group!
Not the 19th-century version of it*, but Meinhof’s and Westermann’s (renamed as Sudan group).
____
* The reasoning in books from 19th cenutry that I read is often based on typology (which is not to say it was confused** with reconstructions). But that’s Semito-Hamitic. The Negro group for Müller is simply a list of languages spoken by a race, for Cust an unclassified (and largely unknown) pile, for Lepsius something very funny: an outcome of contact between truly African Bantu and invading Hamitic.
** But Meinhof*** already believes in its areal unity and then Westermann with Sudan and Meinhof with Hamitic do something a bit lazy: try to go beyond areal or typological claims without really updating the method.
***Though… We STILL don’t have a reconstruction for AA.
Iroquois languages are themselves a subset of “Macro-Siouan” for those who believe Macro-Siouan to have been a thing. I don’t have an informed view myself, other than a suspicion that negative reaction to Greenberg’s ur-lumper “Amerind” proposal may have kept many other specialists perhaps excessively far over to the splitter end of the spectrum.
In any event, I don’t know why geographical discontinuities in the range of Siouan languages should be any more surprising than e.g. comparable discontinuities in the range of Uralic languages. There’s no obvious reason to think that pre-contact North America didn’t have population movements and other historical developments broadly comparable to those we know (because the details are better documented) account for where Uralic-speakers did and didn’t end up. And of course after initial contact along the coasts there were often dramatic shifts in who did or didn’t live where (and what languages were or won’t spoken where) that occurred quite a substantial distance inland from where the European incomers had actually gotten to but that resulted from various chains of causation that the incomers were part of.
The only GOOD development associated with this war is that Russia imports Iranian vegetables.
(A sticker on a unusual pepper says it is from Isfahan)
Based on my experience with certain Iranians I have very warm feelings for this country. (this and that “country” is never “state” for me. Also as one can guess the Iranians I’m likely to befriend aren’t going to be fond of such a state. Most young educated people in Tehran are not going to be fond of it).
J.W. Brewer:
As some comments above alluded to, Tocharian couldn’t have come along with the Iranian languages, because it’s centum*, not satem. That’s what’s perplexing. If Tocharian split off before satemization, then where was it for thousands of years with no known traces, until it appeared in the mid-first millenium AD in those cities full of peacocks and Buddhist texts? Linguists deduce that the language shows a heavy phonological influence from Samoyedic, and after that it ran into its long-lost Iranian cousins and picked up a couple of layers of loanwords from them, so now they’re hoping for material/genetic evidence to corroborate the story.
At least, that’s what I’m getting from a few pages of Spinney’s book in the preview; I couldn’t follow the technical discussions here, where everybody assumed that background was already known, so I’m happy to get the beginner’s version from Spinney.
*it’s more complicated than that, but anyway, not satem.
I have always supposed thst satemization began somewhere in its present range, but before the IE settlement of India Magna, and spread both east and west.
A Conversation with Laura Spinney.
Linguistic diversity and human biodiversity tend, to some extent, to correlate because the more difficult travel is, so is, obviously, language and genetic interchange.
Linguistic diversity and human biodiversity tend, to some extent, to correlate
You just popped into a thread featuring multiple geneticists and linguists just to repeat a message you’ve been pushing for 30 years, without adding a single useful observation to the discussion already present? Muchas gracias.
(Rather looking forward to rozele’s next comment…)
Thanks, Lameen.
Likewise.
1-John Cowan, ktschwartz: Centumization/satemization is probably not a useful dichotomy here (incidentally, a similar dichotomy in Celtic linguistics, between P-Celtic and Q-Celtic, is largely rejected by specialists today: I can give some scholarly reference(s) to any interested hatter who requests I do so downthread).
It has been pointed out (by T. Burrow in THE SANSKRIT LANGUAGE, if memory serves) that Greek could be defined as a satem language in every way except the relevant palatalization. What makes Tocharian intriguing is not so much the fact that is a “centum” language, it is the fact that it (unlike Greek, nota bene!) seems to share next to no linguistic innovation whatsoever with Indo-Iranian, or indeed with any group of satem languages.
And it is this last fact (as opposed to the plain ole’ satem/centum difference) which makes it unlikely in the extreme that Indo-Iranian and Tocharian expanded eastwards together. So does the fact that there do not seem to exist any loanwords in Tocharian which must go back to Proto-Indo-Iranian. This is quite unlike what we find in Uralic, where a number of old, specifically Proto-Indo-Iranian loanwords can be found.
2-John Cowan: on the historical range of satemization, there is a relevant difference between Indo-Iranian satemization on the one hand and Baltic + Slavic satemization on the other: In Indo-Iranian satemization is a sound change that seems to have affected all inherited Indo-European words, whereas in Slavic and Baltic there are a number of exceptions to satemization (which otherwise appear to be fully native Indo-European words, in terms of diachronic phonology).
I suspect that in many of the extinct (and mostly unknown) Indo-European languages of Central Europe and the Balkans even more mixed satem/centum languages must have existed (Speculating wildly here: Imagine an Indo-European language where satemization takes place in monosyllables and in stressed syllables, and centumization in other positions. Or vice-versa, if you prefer).
Meaning that: the “border” between satem and centum must have fluctuated quite a bit historically, and thus we should be cautious to the tenth power: the historically attested range of centum versus satem languages is likely to be QUITE unlike its range at earlier points in history (A similar point was made when it comes to P- versus Q-Celtic, by the way).
3-On correlation between linguistics and human biodiversity: There is a possible instance I am aware of of a REVERSE correlation, where a radical decrease in linguistic diversity was accompanied by a (modest) increase of biodiversity: the Roman Empire.
On the one hand a huge number of languages were lost, annihilated by the spread of Latin and a number of other languages. But on the other, a little-known fact of the Roman Empire is that a large number of new plant varieties were created, increasing agricultural diversity (and thus reducing the risk of famine, which factor definitely explains why the Roman Empire lasted as long as it did). I forget who it was who said that in France the two most lasting legacies of the Roman Empire were the French language and French wines (at the time of the Roman conquest there was no wine production in Gaul, as no breed of grape at the time could survive a winter there).
This spread of new crops did not, as far as I know, trigger any large-scale extinction of plants or animals anywhere in the Empire. So: your typical province of the Roman Empire underwent a sharp reduction of linguistic diversity and a moderate increase in biological diversity.
Thoughts?
your typical province of the Roman Empire underwent a sharp reduction of linguistic diversity and a moderate increase in biological diversity.
I remember reading somewhere that a number of plants otherwise endemic to Africa or Asia still grow around the Forum in Rome, brought by the dung of animals taken there to entertain the masses.
But we know of at least one Roman-induced extinction: silphium. Probably there were others, too economically uninteresting to be recorded in written sources.
Probably two things happened:
1) Weise’s law – depalatalization before something like /r l n/, leading to grammatical alternation that was leveled out differently in different branches;
2) borrowings from Germanic, Italic and perhaps Celtic in Baltic and Slavic.
Weise’s law seems necessary to explain why Lithuanian today has both akmuo “stone” and ašmuo “knife”: of *h₂aḱ-mon-, the nominative would not have been depalatalized (*h₂aḱmō), while the genitive would have been (*h₂akne/os with regular reduction of *-Cmn- to *-Cn-); Lithuanian redistributed the variants by meaning, Slavic generalized the depalatalized variant, Indo-Iranian the palatalized one. The whole word seems missing from West IE (hammer has been proposed, but is more likely a loan from… the Finnic for the back of an ax).
On the subject of Proto-Indo-European, Richard Manning has started a website called Secrets of the Words You Know whose aim is “to widen among the general public with an interest in words knowledge of the basics of Indo-European language links, and doing so by taking advantage of how English – the magpie language – has borrowed so much of its vocabulary from other related languages.” On the home page, it says “A set of roughly 35 blogs (one sent every Monday to subscribers) will unlock the ‘tacit knowledge’ of related languages that English speakers share.” Check it out!
35 blogs? …Oh, 35 blog posts on a single blog. Sorry, pet peeve.
Yeah, I know. We grizzled veterans have to get used to these things.
I suddenly remember one crackpot who really did have a blog that consisted of a single post plus comments that were mostly by him. Eventually, he declared the blog full and made a new one, again with a single post…
I stand corrected, and better informed – though noting that much language change is due to the ignorant winning out over the purists!
I hope that my solecism won’t put people off the main idea that I am hoping to promote – that it’s well worth finding ways of widening knowledge of the basics of sound changes in Indo-European languages beyond the experts, and that English, thanks to its many borrowings, can be one good point of entry for this.
Oh, trust me, nobody will even notice except us grizzled veterans. Your usage is standard nowadays.
Kudos to Richard Manning. This is the first time I see IE historical phonology explained clearly at high-school level.
Yes, it ought to be very useful for that.
I’m finally reading it now, and I’m quite impressed! I learned a few things, too!
Nits to pick:
Blog 4: “There are early records of ‘Gothic’, usually seen as a separate ‘East Germanic’ language, but this has died out, being last recorded in Crimea in the 18th century.” The Wikipedia article “Crimean Gothic” has “Extinct – 18th century” in the infobox, but the text makes clear that that just means it was extinct by the 18th century, because, following the last mentions from the 16th century, a “report by Prussian polymath Peter Simon Pallas from 1794 states that he was unable to find any remnants of the language in Crimea, probably meaning that the language had become extinct by then.” Admittedly, he may have missed it quite narrowly, or maybe he should even have looked in the freshly founded Mariupol – the preceding sentence is: “The Eviction of Christians from the Crimea [link to article] in 1778 might have played a major role in the extinction.”
Blog 8: “And why the ‘g’? The Danish for the city is ‘København’, with the ‘v’ one would expect for a ‘haven’. The ‘g’ in English is due to translation of the Danish sound via a North German pronunciation, where the word for ‘harbour’ is ‘hagen’.” Not quite. There is a term Haag ~ Hagen that shows up in placenames from the Netherlands (the Hague = den Haag = ‘s Gravenhage, “the count’s estate”) all the way to Austria, but it referred to an enclosure (compare German Gehege and, as I just found, Swedish hage) and then perhaps an estate or maybe even a walled town; but it’s not related to a safe haven. Rather, the thing to keep in mind about Danish is that, as someone put it here a few years ago, “all vowels become [ə], and all consonants become vowels”. Today, København is pretty much [ˈkʰømhaʊ̯n]; I don’t know how far that had progressed at the time in question, but the v was probably a wide-open approximant [ʋ], while a g in this position would presumably have been a Spanish-style [ɰ], close enough for confusion; and g in such positions in Low German is generally speaking [ɣ]. In short, -havn was misinterpreted – folk-etymologized – as -hagen.
Blog 9: “indeed English and Dutch are among the very few Indo-European languages to have conserved what was probably the original sound of *w.” Most of Dutch has moved on to [ʋ]; AFAIK, only West Flemish is keeping [w].
Still Blog 9: “to help/hilfen” – helfen. The noun is Hilfe, the imperative is hilf (due to a vanished -i), and so is, in Upper German dialects, the 1sg (due to a vanished -u), but that’s all.
Also: “Pfeife (a musical pipe or whistle)” – also a pipe for smoking. Never a mere tube as in “pipeline”, though.
And: “‘passen’ (to pass)” – that’s the “cognate”, but the meaning, except in football, is “fit”. For “pass by”, passieren is sometimes used, but that word is much more common in the meaning “happen”.
Spelling: ß in beißen (but you can ignore that and point to Switzerland), uppercase for Volk.
More importantly: fünf. This is an *i rounded by the surrounding labial consonants.
The link to Blog 11 leads to https://wordsecrets.org/blogs/previous-blogs-list/previous-blogs-11/ , which doesn’t exist. The expected URL https://wordsecrets.org/blogs/previous-blogs-list/blog11/ works.
Blog 11: “Gasteropod (slugs and snails with a ‘belly-foot’)” – that’s what it should have been, but only the French have managed to correct it to gastéropode. Everyone else says gastropod.
Also:
This is probably all true, but it is a level of precision unmatched by the entire rest of the site – for the good reason that making the whole site like this would require a lot of lengthy explanations that would greatly slow everything down and eventually throw people off. I’d say the most important difference between an English /kw/ and an Italian or PIE /kʷ/ is that the former is aspirated, which creates a pause between the release of the /k/ and the /w/.
“The case for this conclusion has been strengthened by the discovery of three corresponding sets of guttural sounds in Luwian, an extinct Anatolian language.” But the “palatals” – I’d bet actual money they were palatalized velars, as found today (due to later developments) in Russian, Irish and most of Greek for example – had already become affricates in Luwian (written with the z series of cuneiform signs).
Blog 12: “a PIE root *meg-” – actually with *ǵ, as the Indic forms show (let alone the Iranian ones).
Worse: “and from Greek like scientific words beginning with ‘brachio-’, similarly meaning ‘short’. (‘Brachiocephalic’ people are ‘round’-headed as opposed to long-headed). All these words go back to a PIE root *mregh-u-, also meaning ‘short’.” But, consistent with the *-u-, the Greek for short is brachy–, and round-headed people are brachycephalic. Brachiosaurus had upper-arm bones that routinely reached the length of an NBA player; rather than “short”, it is an “arm lizard”.
“*dekm̥” – again ḱ; you get it right in the next paragraph.
I hope that helps!
København: Danish still has have (now = ‘garden’, not just ‘enclosure’ like Sw hage can be) and havn (now = ‘port’, not just ‘refuge’. Sw hamn). Copenhagen is a port city, though, and I don’t think any native speaker is confused as to which of the words it contains. Germans, now…
[k͡p g͡b] are quite distinct from [kʷ gʷ], and the former are very rarely found outside Güldemann’s “Macro-Sudan Belt” (where they’re very common.) It doesn’t seem likely to me that PIE had them. They’re pretty much as areal as clicks.
Nawdm has all four sounds as distinct phonemes, which is, admittedly, unusual for Oti-Volta (and I’m not sure how it came about.)
Kusaal, on the other hand, uses /k͡p g͡b/ for Hausa /kʷ gʷ/ in loanwords, e.g. bakpae “week” from Hausa bakwai “seven.”
The English cognate* of those words for “enclosure” is hedge.
* There are actually several others, but they are archaic or dialectical.
the former is aspirated, which creates a pause between the release of the /k/ and the /w/
Not if you have a voiceless /w/.
hedge: figures. And the other word is haven, I assume. So the misunderstood version of København should be Cheapmanhedge or some eroded version of that.
Pronounced Chunch.
@Lars: in Hrolf Kraki’s Saga by Poul Anderson, the Danes begin to call a certain mercantile area “Cheaping Haven”.
I think one of the dialectal words Brett mentioned is “Hay”, though the only region where I’ve heard of it is Buckland (the easternmost part of the Shire, at least in the 50th Anniversary edition).
Lars Mathiesen has explained why the element hafnium isn’t called copenhafnium: “… the Latin name was just what the clergymen of the 11th and 12th came up with. (The Købmannæ part was added after Absalom’s time, it seems, so didn’t make it into Latin).”
Oh, that guy. He’s talking through his hat. (Though the story seems legit).
Romance in general seems to have restored the -e-; Spanish uses gasterópodo much more frequently than gastrópodo
Not quite; the German cognate of that one, with the same meaning even, is Hecke. We’re looking at a Proto-West-Germanic *ggj ( < Proto-Germanic *ɣj) here; it’s some sort of derivative from the common root.
Of course not, but the “blog” only says “perhaps closer to”, which seems like just too much emphasis on “this is a single phoneme, not a sequence of two” to me.
Oh! Nice!
The epenthetic e in str > ster seems not to be a frequent outcome (maybe it was earlier?) in Romance languages. But in English, str > ster (and tr > ter) are “all over the place”, at least in the spelling (but also in the pronunciation, I say “enter” and “center”, not “entr”or “centr”). Maybe *gasteropod was replaced by gastropod because people thought it was an epenthetic e .
The immediate cognate of Eng. ‘hedge’ is Ger. Hecke “hedge” < *hagjō- f. while Da. hæg, Norw. hegg, Sw. hägg “bird cherry (Prunus padus) < *hagja- m. This looks much like a feminine with collective meaning and its corresponding singular form. But it’s more complicated than that. There’s also Norw. hage “garden” and cognates < *hagō- m., which could be a more basic derivation, and *hag-n- (Ger. Hain) with what looks like an inchoative suffix.
Possible root cognates in other IE branches include Welsh caer. They are few and all have in common that they are derived by other suffixes than in Germanic. This suggests to me that the Germanic forms are separate (and recent?) formations from a verbal root. So which Germanic verb? There’s *hagōn- of ON haga “order, groom; fit, suit” and Ger./Du. behagen “please”, and also the related adj. ON hagr “skilled, competent”. If so, the basic sense of hage is not “enclosure” but “ordered area”, and the hedge is the defining element.
[Edit: This comment looks like a super-quick response to David M., but I’ve been thumbing on it for hours.]
I should add that B&L treat hage v. (Nyn.) and hag adj. under høyre “right (direction)”, seeing the root as an ablaut form to Gmc. *hōg- (-u- or -i- stem) of ON hœgr “simple, good, pleasant” and e.g. hógligr “easy”, which have no Gmc. cognates outside Norse and only tenuous comparisons elsewhere in IE. I believe it would be better to assume that hag- is the primary form and hōg- a lengthened form that underwent the regular change *ā > *ō.
Nonononono, this is about the η in γαστήρ… which I thought wouldn’t disappear in declension. Turns out it does, and Gastropoda is actually fine. Oh well. It’s all Greek to me.
Not “please”, but hard to explain; de.wiktionary, unusually, with an etymology section.
Ah, if that’s a loan from Very Low German Indeed, that explains why it’s so obscure…
David M.: Not “please”, but hard to explain;
Yes. I pondered “comfort” or “make comfortable” but decided it wouldn’t take me much closer.
Glancing through the Herkunft section, I recognize many of the forms from B&L.
@David Marjanović: I’m not sure what you are saying about hedge. You seem to state both that it’s a cognate and not. Is the point that it comes from a word that diverged farther back, in proto-Germanic perhaps, rather than proto-West-Germanic?
Low German? All the etymological/historical dictionaries list it as a late medieval contraction of MHG hagen, common as a part of ostmitteldeutsch toponyms. And obscure? What’s obscure about it? The word (already often used by Luther) was popularized by Klopstock, anybody who has read some German Classical or Romantic literature should have encountered it.
Oops, it was late enough at night that I had managed to forget that [ɣ] > [j] happens from Cologne to Berlin, it doesn’t have to be Frisian!
“Obscure” wasn’t the right word either; “archaic-literary” perhaps, plus terminology such as Hainbuche “European hornbeam“. …and now I see another German name is Hagebuche.
How did Friedrichshain in Berlin get its name? Was it unusual to use a local dialect word for a planned public park?
It’s perfectly standard, just literary to archaic nowadays. If you name things after the king, poetic vocabulary is expected.
Add to this that people don’t always know the words they use are actually regional.
Borrowed into English as hagberry; OED (2016): “the tree is perhaps so called on account of its use in hedges”. Some sources (including Wiktionary) explain hagberry as “the witch’s tree”, but that’s folk etymology.
Being used in hedges also explains hawthorn (as Trond knows).
Hexen have something to do with hedges, but what that is is a mystery.
…What I really meant by “obscure” is that the exact meaning has never been clear to me. Apparently it ranges from “small forest” to “5 planted trees”…?
Paul has “als poetisches Wort durch Kl[opstock] sehr üblich geworden für einen anmutigen Wald oder einen heiligen, den Göttern geweihten”. In other words, the German equivalent of Latin lucus (Forcellini: locus arboribus consitus et religione sacer).
kts: Being used in hedges also explains hawthorn (as Trond knows).
As Trond forgot — at least the Celtic part.
What I’ve been trying to get my head around in this thread is somewhat orthogonal: How are the different Germanic hag- words actually related.
David M.: Hexen have something to do with hedges, but what that is is a mystery.
Before opening your link, I wonder if they could be root cognates. A hag or Hexe could be someone practicing the art of *hagan “forming things after one’s will” — if we also can tie *hagan to ON *hugr “mind” (as e.g. Huginn “mythological raven”). If so, the homonymic connection with hedges would have been there from the start — not even a pun but a figura etymologica.
That makes a lot more sense. There is an OHG “think” word with hug-.
PWGmc *hagatussjā “witch”
I can’t see that working as a compound directly from the verb, but it could be from the noun *hagōn- in our postulated abstract meaning “planning” or some such.
The main motivation for pressing hugr into this is that it’s lonely. B&L make a good case that it’s derived from the same root as Eng. ‘high’, so semantically “height” -> “top” -> “head” -> “mind”, but I’d rather pretend not, since Huginn ok Muninn work better as parallel verbal formations.
The other motivation is that there’s already ablaut present in the haga set, so it would be nice to discover a complete strong verb.
The known word family is here; under “related terms” there are four verbs, one of them strong but not like that.
hagberry
Immediately made me wonder if this is also the etymology of hackberry; apparently the answer is yes.
From memory, Zenaga and Tamasheq both share a tree name äδäräš /adăras which similarly means “hedging”; it was a Sahelian species whose name escapes me.
David M.: The known word family is here
en.wiktionary:
Oh, *ḱek- (“to be able, capable”)! That’s what I tried to invent from the *haga- set.
Bjorvand & Lindeman’s etymology makes it identical to the homonym ON *hugr “hilltop”, which is unattested but reconstructed from placenames. A related formation would be Ger. Hügel. All this is supposedly from the same root as Eng. high, so first we’ll have to sort that out.
First, there’s Da. høj, Sw. hög (> Mod. Norw. høg/høy) “high”, which B&L take back to a *haugá- with suffix accent. The formally close Norw. haug “hillock, pile” should belong here as well.
ON hár/hór “high” and the other Germanic adjectives are from *háuha- with root stress.
So where do we place hugr? I think it must be from a zero grade *KHwg-, which means that it must have been formed before PGmc., though it may of course have been one form in an ablauting paradigm.
Anyway, when it comes to the identity of the two words hugr, this is neither here nor there. When they are identical in form, they can obviously also be reconstructed in the same way. The crux is semantical. There’s no trace of the intermediate meaning “head” in any language, and the meaning of hugr and derivatives such as hugsa “remember” are tilted toward mental abilities. It’s much easier to see that derive from be able.
But to connect hugr “mind” to hagr “able” and haga “make order” (and Wiktionary’s root), we need a hypothesis for the formal derivation. I said I want a strong verb but, unfortunately, I can’t get any of the conjugation classes to work out of the box.
ON *hugr “hilltop”
Inherited from proto-Scandi-Congo: cf Yom kʊ̄ɣə̀r “mountain.”
Obviously. But on the intermediate level:
Me: I can’t get any of the conjugation classes to work out of the box.
To relieve you from your state of breathless suspense: What I want to do is posit a preterito-presentic verb parallel to the one represented by Mod. Norw. måtte, Ger. mögen, OE magan.
And with that I leave you again.
…huh.
That’s certainly a challenge…
I had a clear enough idea in my head that I thought I could do research between the chores to add a few pieces a day and have a complete puzzle solved by the end of the holiday. Not so. I’ll be back when I have time.
*ḱk-í-s: That’s one fancy initial cluster. Are we 100 that the zero ablaut stage didn’t have just a teeny tiny schwa or something? And if it gives *hugiz in Proto-Germanic, that reminds me of the resolution of syllabic resonants. If I had been a PG speaker I’d probably have simplified the thing to a single stop — unless there really was something there to make a syllable out of.
We*’re actually pretty sure there was exactly such a thing – it just wasn’t phonemic.
* Andrew Byrd’s 2010 PhD thesis on PIE syllabification used to be on academia.edu. Now he has disappeared with all his works:
Even his coauthored works no longer list him in the metadata (example here: top: just Ryan Sandell; actual PDF: “Ryan Sandell and Andrew Miles Byrd”).
WTAF. He’s not on ResearchGate either.
Anyway. The idea is that some of the worst zero grades weren’t phonetically zero: *sés-ti, *səs-énti; *kʷétwōr, *kʷətwóres or something like that.
Byrd’s dissertation is at U Kentucky, here. Maybe he decided to break with a.e and RG?
(His CV says he is working on writing novellas in PIE.)
That reminds me of those much-touted 8-consonant clusters of Georgian. Do they allow of such sub-phonemic resolution as well, or are there other things at play? (I see lots of letters that look like they could be syllabic resonants or the like, but Hat would know). Also, put angstskrig in your pipe and smoke it. (= ‘scream of terror’).
There are definitely syllabic resonants, but there are also consonants piled up against each other with chthonic abandon, and the unwary learner may emit an angstskrig.
WP’s article on Nuxalk now has a recorded sample of one of those fearsome long vowelless words. I think I can imagine some linking schwas or such. Like all the rest of the word, they would be voiceless, which I am realizing now is true for much of the language in general. Other than a whole lot of voiceless stops, fricatives, and affricates, it has three vowels, two glides, and two nasals, none of which seems to be used very often.
Re angstskrig, English has this, e.g., rest stop, realised I think with the first t hardened to d or omitted altogether. Is the ng in parking space pronounced differently from the ng in waiting area? I think I can pronounce it in two ways in the latter case but only one way in the former case.
Here in the U.S., I don’t think I’ve heard or uttered a lot of [t]s in the first syllable of “rest stop”, “first step”,etc., though I’m capable of producing one there. I don’t think it’s a [d]
But the ability to have and be flexible about consonant clusters is one of English’s strengths. At this point I usually say that the hardest-to-pronounce English sentence that people have actually used (not just mentioned) is “Ten-twelfths equals five-sixths.” When I’m teaching, I pronounce all the indicated consonants in that sentence, but people usually make it easier somehow.
I have a [g] in “longer” but not in “singer” or “waiting”.
Awesome, thanks!
Probably he never was on RG; as a general tendency with numerous exceptions, the humanities are on a.e and the natural sciences on RG. Of course it’s possible that a.e annoyed him once too much.
…huh.
He’s probably going to discover problems, and quite possibly solutions, that nobody had even thought of before.
I’ve seen them denied on the grounds that 1) the stress goes on the first syllable and 2) words like the infamous gvbrdgvni have final stress, so they must consist of a single syllable preceded by 8 non-syllabic consonants. Well, I barely recognize Georgian when I hear it, but… what if the stress simply goes on the first vowel instead of the first syllable? Syllabic consonants not being stressable is after all widespread: English, German… Syllabic [v] is certainly unusual, but still less unusual than the syllabic voiceless obstruents of the Pacific Northwest.
Anyway, /tʰbilisi/ is real, the /tʰ/ is not voiced and the /b/ is not devoiced, and the closest thing to a vowel in that cluster is the voiceless aspiration.
…While I’m at it, given its large consonant inventory, it’s also interesting which consonants Georgian does not have. The complete lack of [j] is striking. There’s nothing that could be interpreted as any kind of /j/; any ai sequences, even in fresh loans*, contain a syllable break.
* at least the one word I was able to identify when I heard it
That’s mainstream. There’s a large area in England where there’s a [g] in all three, but it goes nowhere near London.
the stress goes on the first syllable
Not really. Wikipedia: “Stress is very weak, and linguists disagree as to where stress occurs in words.” Stress is not a useful concept in Georgian.
The plot thickens!
Me: I have a [g] in “longer” but not in “singer” or “waiting”.
DM: That’s mainstream. There’s a large area in England where there’s a [g] in all three, but it goes nowhere near London.
In the U.S., that [g] in “singer” is one of the things people noticed about my mother’s accent when she moved from Pittsburgh to Cleveland, she tells me.
I’ve seen in more than one place a mention of a distinction between singer with [ŋ] and finger with [ŋg], as if it were an English universal.
How rare is finger with just a [ŋ]?
@Y: English pronunciation of finger with just a [ŋ] is decidedly nonstandard. I’m not sure I’ve ever heard it from a native speaker. I remember, in my first year of German classes, our teacher pointed out the difference in pronunciation between the two languages, since in German, finger never has [ŋg]. In contrast, singer can go either way in English.
i associate generalized [ŋg] in that kind of context with both some some yiddishished englishes and (overlappingly) some suburban new york topolects (especially the ones that get referred to by outsiders through citing their [ŋg] in “long island” – usually as something like “a lawn guyland accent”).
@Y: “Only obstruents,” it claims, but that depends on analyzing the thing that’s written kw as [kʷʰ], but to me it sounds more like [kʰu̥] in the recording, and [-skʰu̥t͡s] seems easier to produce than [-skʷʰt͡s]. Also the /ɬ/s, ‘obstruent’ is a bit strongly worded for those.
But then there’s something like tts = ‘little boy’ cited lower; I think I can pronounce [tʰt͡sʰ], but only because the aspiration affords a tiny break between the first release and the second obstruction. In that aspect it’s not much different from Ancient Greek χθών or very careful old RP chthonic that didn’t simplify the cluster. Doing a longer sequence like that is just a skill issue.
That’s what both CEPD and LPD say, both for General British and General American pronunciation.
Wells gives the following rules of thumb:
Exceptions to the rules are longer and younger, which have /ŋg/ like finger despite the morpheme boundary.
In German (universally or nearly so), to find /ŋg/ you need a word boundary or at least a following unreduced vowel (Ungarn “Hungary”; contrast von Ungern-Sternberg without the /g/) or, in various southern realms, an immediately following /l/ or /r/. The only special case seems to be the conjugation of hängen, which comes from the conflation of a verb (obsolete hangen) with its former causative (obsolete henken) and maybe something else, too.
Danish is the same; before reduced syllables, i.e., -e and -ig, you get /ŋ/, otherwise /ŋg/. I know of no exceptions. Inge (girl), Helge (boy), afhængig vs. Inga (girl), Helga (girl), Ingo (boy). I can’t put my finger on any non-proper nouns with the latter pattern right now, but there must be some loanword somewhere.
I think the only surviving causative of strong hænge/hang/hængt is weak hænge/hængte/hængt, but at least in my idiom billedet hang på væggen fordi jeg hængte det der. I don’t have any YPOTD handy to check if this has also descended into chaos. (Wikt.en says they are often mixed up, which rings true. Using the weak preterite for the strong is the way it most often goes, that’s widespread in child language, but jeg hang billedet på væggen falls in a pattern with jocular imitations of substandard usages like jeg lå mig på sengen).
(Hænge ud is a straightforward calque of E hang out, but I’d be torn whether to use the weak or the strong preterite. Neither feels WRONG, though logic would indicate the strong because intransitive).
(‘Execute’ is henrette, ‘executioner’ is bøddel. I don’t remember encountering anything like Ge henken or Henker in Danish).
@Lars
We had Ge. Beil on another thread, but bøddel looks bizarre, kind of like executioner crossed with church functionary (Ge. Pedell, earlier with variant Bedell, both variants also in late Latin)
Compare German hinrichten.
Not as obsolete, by far, as hangen. And Henker, the word, is definitely not obsolete, even though the office is. There is also the synonym Scharfrichter.
Exceptions to the rules are longer and younger
I think that the forms with /ŋ/ (as opposed to /ŋg/) before a suffix happen only with the ultra-productive gerund/participle suffix -ing and the similarly ultra-productive “agent noun” suffix -er, which also share the property that they induce no formal irregularities whatsoever (so even, e.g. “going”, “goer”, “being”); they are almost clitic-like, rather than suffix-like: certainly so, from a purely phonological point of view.
Zwicky and Pullum’s criteria for suffixhood versus clitichood imply a neat dichotomy, but really there’s a cline.
https://web.stanford.edu/~zwicky/ZPCliticsInfl.pdf
As with definitions of “word”, phonological and morphosyntactic criteria for cliticness don’t neatly align with one another in reality. (This is a big issue in Kusaal, too.)
It’s perfectly possible for a historical flexional suffix to develop into a clitic: English possessive “s”k is a fine example. (Kusaal main-clause perfective -ya is another.)
Nobody* seems to have any doubt that “wronger” and “winningest” would be pronounced with /ŋ/ before the suffix, not /ŋg/, even if they’re doubtful about whether they’d actually use these words. Pullum thinks “wronger” is mysterious, but I don’t see why; aren’t we just generalizing from those other ultra-productive suffixes?
* That is, nobody who speaks General British or General American; I guess speakers of the non-ng-coalescing dialects weren’t surveyed.
Oops, yes: I was taught the entire comparative is an exception – which makes the majority intuition about wronger (which is, interestingly, also mine) the exception from the exception. Given the rarity of wronger, I want to blame analogy from a grammatically different word – but what? Surely not the agent noun longer, which is at least as rare?
Büttel “some kind of subordinate office in medieval towns”.
True.
Superlatives in -est are interestingly different from comparatives in -er, notably in that, at least in what is sometimes rather charmingly called “mannered prose”, they are quite a bit more freely formed: “beautifullest”, albeit twee, is more acceptable than “beautifuller.” I think that one could argue that “-est” is farther along toward the “clitic” end of the affix-clitic scale than comparative “-er” is (though not nearly as far as “agent”-noun “-er.”)
Come to think of it, I pronounce “bringest” (as in “thou bringest”) as /bɹɪŋɛst/, not /bɹɪŋgɛst/, but that probably has little or no connexion with how such forms were actually pronounced historically. I think the simplification of the word-final /ŋg/ cluster actually postdates the loss of the old 2nd person singular forms in standard English anyway.
@dm
Of course, Bedell = beadle, Büttel = butler? But I still don’t see how you get from butler to executioner (via “the butler did it”?).
I wonder if French bourreau might be related? The derivation from bourrer given in Wiktionary looks shaky to me.
Büttel = butler? But I still don’t see how you get from butler to executioner (via “the butler did it”?)
No. Butler is related to bottle, while Büttel is related to “bid” – a person who does someone’s bidding. As DM says, it’s a name historically used for low-level officials, doing, among others, bailiff or police work in apprehending culprits or defendants or executing court orders, from where it’s not a big step to hangman.
Obligatory reminder of the book on the complexity of clitics and adjacent phenomena (right before the thread descends into Udi endoclitics and most delightful Hebrew neo-roots).
“Nobody* seems to have any doubt” — that was careless of me, there *was* some dissension in the comments to the WRONGER post, as well as in the sci.lang thread from 1998 that Pullum linked (where Keith Ivey voted for “wronger” with a /g/).
Come to think of it, I pronounce “bringest” (as in “thou bringest”) as /bɹɪŋɛst/, not /bɹɪŋgɛst/
So do I, and /bɹɪŋgɛst/ sounds wrong and weird to me.
Wrongo! definitely has a /g/, but that’s because of bingo!
@Lars: I am not sure if you can really draw a clear phonetic line between [kʷʰ] and [kʰu̥], in all their variability.
When I heard that recording on WP my first thought was, “Why is he whispering?”. My second was “Oh.” And then I looked at the sound inventory and realized that the language in general could mostly sound like that. How do you sing or shout in Nuxalk?
Bøddel is from MLG bod(d)el, cf G Büttel and E beadle it says, something like ‘bidden by the court (of law)’. Related to the verb byde. Danish also has skarpretter, calqued from German. TIL I learned that a bøddel hangs people, while the skarpretter uses sword or axe; they have been conflated for centuries, though.
Pedel is the same word after a detour through LL.
Also also, Henker was in actual fact attested in Danish. In 1740.
@David Marjanović: Saying wrongo! with [ŋg] definitely sounds autological to me.
Yeah, I would not say wrongo! with [ŋg] (though I don’t remember ever saying it).
“Wronggo” from John O’Hara’s doubletalk expert.
I remember encountering “wrongo” as “Wrongo bongo!” which I’m sure has a [g], so I think “wrongo” does too, for me. (On the other hand, “wrong” and hence “wrongo” are CLOTH and “bongo” is LOT.)
But my intuition is for no [g] in comparative “wronger”, despite the analogy with “younger, “longer”, and “stronger”.
Y:
Well, that’s my point, really. The whole “Wow, Nuxalk has words with only obstruents” thing is a bit shakier if kʷʰ has a phonetic range that includes a vowel (albeit unvoiced). And you have to class fricatives like /ɬ/ and /s/ as obstruents…
The tts word is less contrived as an example, though. You really have to work that aspiration to get anywhere.
And yeah, how do you get angry in Nuxalk? Maybe you can pick words with lots of /ɬ/ and /s/ and sound like an annoyed snake, but from the sample words on Wikipedia it looks like a full-throated “drill instructor” bellow will be hard to carry off. Maybe they cherry-picked them just a little.
@Lars, Y
For me this is a bit like asking how West Country or Welsh people (sorry, D.E) express anger. Register, tone, pitch, volume. It is true that there are individuals who seem to have “permanent non-angry settings” for all these things; in case one meets such an individual, the rule is the same as with certain dogs: be alert and keep a respectful distance. Judging from overhearing some Spaniards in a casual friendly conversation which sounded to me as though they were about to engage in a fight to the death, I believe the problem of gauging a stranger’s anger level might apply equally to West Country or Welsh (sorry again, David) people.
Well, DE has his Glasgow roots to fall back on.
You lookin at me, Jimmy?
(This classic Glaswegianism is probably even more scary when spoken softly than well bellowed. We are a subtle people.)
I think a good analogy would be the way that languages manage to mark emphasis and focus perfectly well without the elaborate English clause-level stress system, which unsophisticated English speakers often attempt to carry over into other languages under the misapprehension that it’s a human universal.
Well, face to face there are many other cues than tone of voice. I don’t imagine that Mack the Knife had to raise his much; we didn’t evolve to judge emotion from sound clips on WP. And maybe those belligerent-sounding hispanohablantes were indicating that it was all a big joke in ways that bodily hampered Anglos and Danes just won’t notice.
“Obstruent” is the cover term for fricatives, affricates and plosives.
I’m not surprised. Un jardín de naranjas sounds like a threat all by itself.
@DM, I should have known that. But then I mean, “they say that as if it’s strange.” A word with only (unaspirated) plosives and affricates would be strange/hard to pronounce, in my opinion. Fricatives lowers the difficulty a little. But it still breaks the sonority peak model of the syllable.