Zompist Bboard Again

KathTheDragon wrote: ↑Mon Nov 16, 2020 5:04 pm I hereby reiterate that you will have a far better time if you stop engaging Octaviano.

You are of course right. We have been through this many times: he never learned and probably will never learn. He just keeps posting bogus etymologies and dodging our questions. It is futile. Yet, the temptation to say, "No, you are wrong because ..." is strong, but I must learn better to resist it.

That said, I shall start a new discussion which has nothing to do with Talskubilos's bogus etymologies (though it would not surprise me if T. has an opinion on this). In the middle of the 20th century, the German Indo-Europeanist Hans Krahe discovered (or at least, believed that he discovered) a network of recurring river names in western Europe, the Old European hydronymy. He assumed that these names were from an IE language ancestral to Italic, Celtic, Germanic and Baltic. Later in the century, Theo Vennemann posited that these names are "Vasconic", i.e. from a language family whose only surviving member was Basque. Others objected that these names resembled each other merely by chance, and likened the whole thing to ley lines: a meaningless pattern falling out of the sheer mass of data points. Yet, the hydronymy seems to be more or less coterminous with the Bell Beaker culture and the area of predominance of the Y-DNA haplogroup R1b, which seems as if there appears to have been some kind of ethnic group involved that left its traces in three different sciences (though one must concede that most of the boundaries of the three are sea shores, making the match trivial). Or is this all coincidence? What do you think?

WeepingElf wrote: ↑Tue Nov 17, 2020 9:40 amYou are of course right. We have been through this many times: he never learned and probably will never learn. He just keeps posting bogus etymologies and dodging our questions. It is futile. Yet, the temptation to say, "No, you are wrong because ..." is strong, but I must learn better to resist it.

WeepingElf wrote: ↑Tue Nov 17, 2020 9:40 amThat said, I shall start a new discussion which has nothing to do with Talskubilos's bogus etymologies (though it would not surprise me if T. has an opinion on this). In the middle of the 20th century, the German Indo-Europeanist Hans Krahe discovered (or at least, believed that he discovered) a network of recurring river names in western Europe, the Old European hydronymy. He assumed that these names were from an IE language ancestral to Italic, Celtic, Germanic and Baltic.

It's quite possible OEH doesn't represent a single but several paleo-IE languages, a term introduced by the Spanish Indo-Europeanist Francisco Villar.

Richard W wrote: ↑Tue Nov 17, 2020 3:42 amI'm not sure what the 'sampling' process for Tavi's examples is. I suspect it is more a matter of going through an entire documented vocabulary and looking for cognates and picking out what look like good matches - probably many years' work.

You guessed right.

Talskubilos wrote: ↑Tue Nov 17, 2020 9:56 am<image>

You see? How do you have a discussion with this? Why do you all hate yourselves so? Block him to remove that temptation. It's worked very well for me.

Richard W wrote: ↑Tue Nov 17, 2020 5:54 amAnd generally, the shorter the list, the better the quality of each match. Huge dictionaries are generally very bad for the average quality of the comparisons. Loanwords, of course, are especially difficult, as the semantics are likely to be significantly changed in the course of borrowing.

This isn't necessarily so, especially when they're Wanderwörter relative to domesticated animals or pottery/technology inventions. In the case of PIE, words such as 'plough', 'dog', 'piglet', 'horse' or 'wheel' would fall into this cathegory.

Richard W wrote: ↑Tue Nov 17, 2020 7:22 am

bradrn wrote: ↑Tue Nov 17, 2020 6:20 am I did already mention this:

bradrn wrote: ↑Tue Nov 17, 2020 1:55 am … I do agree with you that the false positive rate rises with sample size — it is a trivial conclusion that the number of chance resemblances increases with the number of words you compare.

It is not a trivial consequence that the proportion increases. As you lengthen the lists, the proportion of words that do correspond historically goes down. Conversely, if you sampled words randomly from a fixed vocabulary, the proportion of false matches would not increase as you increased sample sizes. However, if you enlarge the vocabulary included from the other language, the proportion of false matches will increase.

I was only referring to that last point, ‘enlarging the vocabulary from the other language’.

bradrn wrote: ↑Tue Nov 17, 2020 6:20 am (Though it should be noted that this advantage of shorter lists disappears when you make your lists too short: the shorter the list is, the harder it is to apply proper methodologies, since you can no longer confirm sound correspondences nor account for semantic shifts.)

The Swadesh lists and the like have very little use when comparing unrelated languages and looking for loanwords.

Yes, loanwords are another reason it becomes more difficult to have good methodology with shorter lists.

Talskubilos wrote: ↑Tue Nov 17, 2020 9:56 am

WeepingElf wrote: ↑Tue Nov 17, 2020 9:40 amYou are of course right. We have been through this many times: he never learned and probably will never learn. He just keeps posting bogus etymologies and dodging our questions. It is futile. Yet, the temptation to say, "No, you are wrong because ..." is strong, but I must learn better to resist it.

https://3starlearningexperiences.files. ... 05/lf3.png

Ah, I see I’m not the only one who likes The Upturned Microscope! (I’m surprised anyone else has even heard of it.)

But more seriously: if we’re misrepresenting your arguments, could you please explain to us where we’re going wrong? I know I’m not doing it deliberately, and I don’t think anyone else is either.

I suspect the specific point he feels is a strawman is "He just keeps posting bogus etymologies and dodging our questions." There just isn't really anything else in what WeepingElf's post it could be. Which amounts to "your criticism is inherently invalid because I say so". I've been dealing with a crackpot on Academia.edu who uses exactly the same tactic, albeit a lot more rudely: anyone who disagrees is automatically wrong and no scientific argument is given because obviously none is needed; any rational person would immediately agree with the crackpot.

KathTheDragon wrote: ↑Tue Nov 17, 2020 7:00 pmI suspect the specific point he feels is a strawman is "He just keeps posting bogus etymologies and dodging our questions." There just isn't really anything else in what WeepingElf's post it could be. Which amounts to "your criticism is inherently invalid because I say so".

Not exactly. Your last statement is actually a strawman.

KathTheDragon wrote: ↑Tue Nov 17, 2020 7:00 pm I suspect the specific point he feels is a strawman is "He just keeps posting bogus etymologies and dodging our questions." There just isn't really anything else in what WeepingElf's post it could be. Which amounts to "your criticism is inherently invalid because I say so". I've been dealing with a crackpot on Academia.edu who uses exactly the same tactic, albeit a lot more rudely: anyone who disagrees is automatically wrong and no scientific argument is given because obviously none is needed; any rational person would immediately agree with the crackpot.

Ah yes, the good old technique of ‘anything that contradicts me is wrong and not worthy of reply’. It’s just lucky that Talskubilos doesn’t do that! I’d say he’s simply a bit bad at explaining his ideas: after some questioning, I think it’s reasonably clear what he thinks, and it isn’t even that crackpotty. (As I understand it: he thinks that PIE is heavily over-reconstructed, the unrealistic words actually belong to substrates, and PIE is related to NC. The only place he errs is in taking his correspondences too seriously.) Certainly his techniques compares favourably with those of such luminaries as Greenberg!

bradrn wrote: ↑Tue Nov 17, 2020 6:36 pmAh, I see I’m not the only one who likes The Upturned Microscope! (I’m surprised anyone else has even heard of it.)

Neither did I, but Google can lead you to unexpected places.

bradrn wrote: ↑Tue Nov 17, 2020 6:36 pmBut more seriously: if we’re misrepresenting your arguments, could you please explain to us where we’re going wrong? I know I’m not doing it deliberately, and I don’t think anyone else is either.

You and Richard are doing fine, but I can't say the same of other board members.

bradrn wrote: ↑Tue Nov 17, 2020 7:12 pmAh yes, the good old technique of ‘anything that contradicts me is wrong and not worthy of reply’. It’s just lucky that Talskubilos doesn’t do that! I’d say he’s simply a bit bad at explaining his ideas: after some questioning, I think it’s reasonably clear what he thinks, and it isn’t even that crackpotty. (As I understand it: he thinks that PIE is heavily over-reconstructed, the unrealistic words actually belong to substrates, and PIE is related to NC.

Not exactly. As Sergei Starostin proposed in the article I quoted before, PIE has a bunch of Caucasian loanwords, and there're also substrate loanwords with Caucasian pedigree in IE languages, but this doesn't imply a genetic relationship between these families. To complicate matters further, Caucasian has also "Eurasiatic" lexicon relative to wild fauna and flora.

Talskubilos wrote: ↑Tue Nov 17, 2020 7:23 pm

bradrn wrote: ↑Tue Nov 17, 2020 7:12 pmAh yes, the good old technique of ‘anything that contradicts me is wrong and not worthy of reply’. It’s just lucky that Talskubilos doesn’t do that! I’d say he’s simply a bit bad at explaining his ideas: after some questioning, I think it’s reasonably clear what he thinks, and it isn’t even that crackpotty. (As I understand it: he thinks that PIE is heavily over-reconstructed, the unrealistic words actually belong to substrates, and PIE is related to NC.

Not exactly. As Sergei Starostin proposed in the article I quoted before, PIE has a bunch of Caucasian loanwords, and there're also substrate loanwords with Caucasian pedigree in IE languages, but this doesn't imply a genetic relationship between these families. To complicate matters further, Caucasian has also "Eurasiatic" lexicon relative to wild fauna and flora.

So you weren’t arguing for a PIE–NC connection in these?

Talskubilos wrote: ↑Thu Nov 12, 2020 11:23 am Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

Talskubilos wrote: ↑Fri Nov 13, 2020 9:09 am Apparently, there're no more examples of *ttsˀ ~ IE *d but I could find one with *s instead: Caucasian *ttsˀăqˀV 'strength, power' ~ IE *seģh- 'to hold'.

This reminds me of the d ~ s alternation in *pard-/*pars- 'leopard' ~ Caucasian *bħĕrtsˀĭ (~ -ĕ) 'wolf, jackal' and *sinģh-o- 'leopard, lion' ~ Caucasian *tsˀæ:nqqˀV 'lynx, panther' I quoted before, so I bet it has something to do with the prehistory of PIE.

bradrn wrote: ↑Tue Nov 17, 2020 7:12 pm

KathTheDragon wrote: ↑Tue Nov 17, 2020 7:00 pm I suspect the specific point he feels is a strawman is "He just keeps posting bogus etymologies and dodging our questions." There just isn't really anything else in what WeepingElf's post it could be. Which amounts to "your criticism is inherently invalid because I say so". I've been dealing with a crackpot on Academia.edu who uses exactly the same tactic, albeit a lot more rudely: anyone who disagrees is automatically wrong and no scientific argument is given because obviously none is needed; any rational person would immediately agree with the crackpot.

Ah yes, the good old technique of ‘anything that contradicts me is wrong and not worthy of reply’. It’s just lucky that Talskubilos doesn’t do that! I’d say he’s simply a bit bad at explaining his ideas: after some questioning, I think it’s reasonably clear what he thinks, and it isn’t even that crackpotty. (As I understand it: he thinks that PIE is heavily over-reconstructed, the unrealistic words actually belong to substrates, and PIE is related to NC. The only place he errs is in taking his correspondences too seriously.) Certainly his techniques compares favourably with those of such luminaries as Greenberg!

Frankly, I don't know what he is up to! He never explains, and uses terms for hypothetical linguistic entities, such as "Paleo-IE", "Europic" etc., in his own senses which do not seem to agree with what these terms were originally coined for. He likes to appeal to Francisco Villar, a Spanish linguist who maintains some quite bizarre ideas about European linguistic prehistory (which I can't easily explain because I haven't really got them well, my Spanish being too limited to make sense of difficult linguistic writing). Basically, it seems, as if Villar (and apparently also Talskubilos) entertains the notion that the languages of Europe evolved in situ and frantically exchanged loanwords all the times, which created the illusion of the language family known to mainstream scholars as Indo-European. But I don't know whether I did not misrepresent either author.

At any rate, there are indeed worse crackpots than Talskubilos. His ideas are not 100% nonsense, though I feel that his procedures yield far too many false positives to be of any value.

WeepingElf wrote: ↑Tue Nov 17, 2020 8:32 pm

bradrn wrote: ↑Tue Nov 17, 2020 7:12 pm

KathTheDragon wrote: ↑Tue Nov 17, 2020 7:00 pm I suspect the specific point he feels is a strawman is "He just keeps posting bogus etymologies and dodging our questions." There just isn't really anything else in what WeepingElf's post it could be. Which amounts to "your criticism is inherently invalid because I say so". I've been dealing with a crackpot on Academia.edu who uses exactly the same tactic, albeit a lot more rudely: anyone who disagrees is automatically wrong and no scientific argument is given because obviously none is needed; any rational person would immediately agree with the crackpot.

Ah yes, the good old technique of ‘anything that contradicts me is wrong and not worthy of reply’. It’s just lucky that Talskubilos doesn’t do that! I’d say he’s simply a bit bad at explaining his ideas: after some questioning, I think it’s reasonably clear what he thinks, and it isn’t even that crackpotty. (As I understand it: he thinks that PIE is heavily over-reconstructed, the unrealistic words actually belong to substrates, and PIE is related to NC. The only place he errs is in taking his correspondences too seriously.) Certainly his techniques compares favourably with those of such luminaries as Greenberg!

Frankly, I don't know what he is up to! He never explains, and uses terms for hypothetical linguistic entities, such as "Paleo-IE", "Europic" etc., in his own senses which do not seem to agree with what these terms were originally coined for. He likes to appeal to Francisco Villar, a Spanish linguist who maintains some quite bizarre ideas about European linguistic prehistory (which I can't easily explain because I haven't really got them well, my Spanish being too limited to make sense of difficult linguistic writing).

Huh? When did he do any of that? At least, I don’t remember seeing any of those in his recent posts.

Basically, it seems, as if Villar (and apparently also Talskubilos) entertains the notion that the languages of Europe evolved in situ and frantically exchanged loanwords all the times, which created the illusion of the language family known to mainstream scholars as Indo-European. But I don't know whether I did not misrepresent either author.

And, as far as I can tell, this isn’t what Talskubilos thinks: my understanding is that he accepts both IE and PIE as valid, he just thinks that we have wrongly ascribed many words to PIE which are really from substrates.

bradrn wrote: ↑Tue Nov 17, 2020 7:30 pm So you weren’t arguing for a PIE–NC connection in these?

Talskubilos wrote: ↑Thu Nov 12, 2020 11:23 am Getting back to my former example, the correspondence between IE *don- 'reed' ~ *ned-o- 'reed, rush' and East Caucasian *ttsˀwǝ̄nHē/*Hnǝ̄ttsˀwē 'reed, cane' (§3.15 in Starostin's article) would correspond to PIE, whereas *yoini- 'reed' would be a loanword, with ttsˀw ~ yo. Different sound correspondences => different lexicon layers.

Loanwords can have non-trivial phonetic correspondences - look at Romance into English or Indo-Iranian into Finnic. If the period of lending is short enough, these correspondences can also be regular - though inconsistent borrowing is quite possible, which is one reason I have trouble with English loanwords in Thai. The PIE-NC connection I see being proposed is borrowing, not common inheritance.

Talskubilos wrote: ↑Tue Nov 17, 2020 12:03 am

WeepingElf wrote: ↑Mon Nov 16, 2020 4:13 pmWe know; you have said so multiple times. But you haven't yet given a valid reason for your disagreement.

I quoted an example where both sound correspondences and semantics are good, so Zju's point was a strawman.

WeepingElf wrote: ↑Mon Nov 16, 2020 4:13 pmProtolanguages aren't languages?!? You are making a fool of yourself. Of course, protolanguages are less well known than living languages (where we can go and ask speakers) or languages known from large text corpora like Greek or Latin. But they are languages, though incompletely known ones. At least, they are models of languages of the past. Not perfectly accurate, but reasonably close if the evidence is good and the methods are sound. Indeed, PIE is known better than some scarcely attested historical languages such as Hattic or Etruscan!

I disagree. The reconstructed PIE, with more +2000 lexical items, could never be anything like a real language!

One can hardly talk about sound correspondence when there's a single pair of words present and no sounds match up (Caucasian *ttsˀăqˀV 'strength, power' ~ IE *seģh- 'to hold'). Before you say 'But ablaut!', falling back to ablaut or wildcard characters* makes things worse, as it increases the potential for chance resemblances. The difference in semantics is not trivial, either. Both sound correspondences and semantics are bad.

* or reconstructed segments with unknown quality, such as PIE ģh. There are like what, half a dozen proposal as to how it was pronounced? I'm not sure.

In general, establishing relationships on the grounds of reconstructed forms is a much less rigorous methodology. Mind the asterisk:

Then we compare the “proto-branch” languages to reconstruct the most recent common ancestor of the whole family, don’t we?

No, we don’t. Proto-Indo-European was not reconstructed by comparing Proto-Indo-Iranian, Proto-Slavic, Proto-Italic, Proto-Celtic, Proto-Germanic, Proto-Anatolian, etc.,
...
Linguistic reconstruction is not conducted consistently in a bottom-up fashion, by piecing together smaller units before handling larger ones.
...
Thus, protolanguage reconstructions are not “data”. They are forever provisional and hypothetical. Using them as data is a category error.

I don't recall anyone claiming that the reconstructed PIE is the real thing, the same as the language that was spoken. It's a model of it, and it inevitably conflates multiple dialects spoken at different times.
If your argument rather is that some lexical items are loans and not native words, well, that's that. The word 'canoe' is both an English word and a loan from Taíno *kanowa. Similarly, PIE *h₂ébl̥ 'apple' is both a PIE word and (most likely) a borrowing from another language. If you argue that individual words were borrowed at a later stage than PIE, that's another point altogether and doesn't dismiss the validity of PIE and the rest of its lexicon.

Talskubilos wrote: ↑Tue Nov 17, 2020 12:03 am

WeepingElf wrote: ↑Mon Nov 16, 2020 4:13 pmOf course, you will get more positives that way - but they are false positives, making the whole endeavour worthless.

As I mentioned before, quality doesn't depend on quantity, and the likelihood of false positives (a better term than "chance resemblances" in this context) actually raises when more data is available.

Quantity is still prerequisite. One cannot establish much of anything with a single example, or even a couple. More are called for. Anything less than that and we may as well be looking at chance resemblances.

bradrn wrote: ↑Tue Nov 17, 2020 8:41 pm Huh? When did he do any of that? At least, I don’t remember seeing any of those in his recent posts.

Oh, we had quite a bit of argument earlier, over some purpoted etymology; the problem I had with these is not only that they're unproven, but that they're unprovable. I don't think that my objections were satisfactorily answered.
Neither was my recent objections that some proposals would lead to a extraordinarily high number of borrowings in PIE - or even, in several IE languages.
(This does happen in Australia, yes... But on what grounds should we posit the same for IE?)

And honestly, yes, it's worse than Greenberg. A restricted number of families would make sense for the Americas; the prevalence of N-M pronouns is hard to explain. There's really nothing suggesting that IE can be explained in terms of more or less random borrowings with random sound change between random families; as Pauli said, it's not even wrong.

Zju wrote: ↑Wed Nov 18, 2020 12:58 amOne can hardly talk about sound correspondence when there's a single pair of words present and no sounds match up (Caucasian *ttsˀăqˀV 'strength, power' ~ IE *seģh- 'to hold'). Before you say 'But ablaut!', falling back to ablaut or wildcard characters* makes things worse, as it increases the potential for chance resemblances. The difference in semantics is not trivial, either. Both sound correspondences and semantics are bad.

Admittedly, this particular correspondence is weaker than others I presented before, but it doesn't invalidate them.

Zju wrote: ↑Wed Nov 18, 2020 12:58 amIf your argument rather is that some lexical items are loans and not native words, well, that's that. The word 'canoe' is both an English word and a loan from Taíno *kanowa. Similarly, PIE *h₂ébl̥ 'apple' is both a PIE word and (most likely) a borrowing from another language. If you argue that individual words were borrowed at a later stage than PIE, that's another point altogether and doesn't dismiss the validity of PIE and the rest of its lexicon.

Actually, the 'apple' word isn't even PIE-reconstructable, and it would be a Paleo-European substrate loanword cognate to Hittite šam(a)lu- and other non-IE words. The "anomalous" /b/ comes from denasalization of /m/.

Most interestingly, there're instances of two words A and B (and more rarely, even a third word C) which would be linked through a long-range correspondence within the reconstructed PIE. Take for example (A) *prt-u- 'passage, way' (Latin portus, English ford) vs. (B) *bhrodh-o- 'ford' (Russian brodъ 'ford'). These can be derived respectively from the verbs *per- 'to pass through' (Latin portāre, English fare) and *bhredh- 'to wade, jump over' (Russian bredu, bresti 'to wade', Albanian breth 'to jump'). Given that e ~ o is the "canonical" Ablaut pattern and external comparanda (e.g. Berber *a-barid 'road') have got voiced stops, my bet is (B) belongs to the "real" PIE and (A) is a loanword from another language.

Ares Land wrote: ↑Wed Nov 18, 2020 3:12 am

bradrn wrote: ↑Tue Nov 17, 2020 8:41 pm Huh? When did he do any of that? At least, I don’t remember seeing any of those in his recent posts.

Oh, we had quite a bit of argument earlier, over some purpoted etymology; the problem I had with these is not only that they're unproven, but that they're unprovable.

Aren’t most etymologies unprovable? Or was this one somehow more unprovable than most?

Neither was my recent objections that some proposals would lead to a extraordinarily high number of borrowings in PIE - or even, in several IE languages.

Yes, I would agree: I’d say this is the single strongest objection to Talskubilos’s ideas.

And honestly, yes, it's worse than Greenberg. A restricted number of families would make sense for the Americas; the prevalence of N-M pronouns is hard to explain. There's really nothing suggesting that IE can be explained in terms of more or less random borrowings with random sound change between random families; as Pauli said, it's not even wrong.

But Talskubilos, as far as I can tell, is not denying that IE is a family. He’s proposing that most reconstructed PIE words are actually from substrates. If he were arguing against the existence of IE, I agree that this would be not even wrong.

Talskubilos wrote: ↑Wed Nov 18, 2020 7:27 am Actually, the 'apple' word is't even PIE-reconstructable, but it's instead a Paleo-European substrate loanword cognate to Hittite šam(a)lu- and other non-IE words. The "anomalous" /b/ comes from denasalization of /m/.

Um, Hittite is an IE language, so how can you have a non-IE word which is cognate to a Hittite word?

bradrn wrote: ↑Wed Nov 18, 2020 7:49 amUm, Hittite is an IE language, so how can you have a non-IE word which is cognate to a Hittite word?

The same way an IE word can be cognate to a non-IE word, although in the case of 'apple' we're likely dealing with an ancient Wanderwort: Uralic *omena/*omVrV 'apple', Basque udari, udare 'pear' < *u-malV, Caucasian *mhalV-/*mhanV- 'warm'.