Jump to content

Talk:Comparative method (linguistics)/Archive 2

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

the problem of PIE aspirates

I'm gonna revert this anonymous addition:

They state that it is extremely unlikely, or maybe even impossible, for a language to have a voiced aspirated (breathy voice) series without a corresponding voiceless aspirated series (though this latter position has been shown to be erroneous).

Shown how? Is there a living language with such a phonology? —Tamfang (talk) 05:25, 17 October 2009 (UTC)

Yes. One of them, (the Bario dialect of) Kelabit is even mentioned on Glottalic theory. Also, some Kwa languages are said to have such a system by Steward. Curiously, some Kwa languages have a four-way system with all possible combinations of voice and aspiration, but Mbato has lost the unvoiced aspirates of all series. Florian Blaschke (talk) 17:05, 27 December 2009 (UTC)

Gobbledeygook

", with alternative lexicostatistical methods such as glottochronology generally considered to be unreliable as there are cases when lexicostatistics does not work at all. Potential problems with the comparative method have also arisen as a result of a number of advances in linguistic thought, in large part due to some of the "basic assumptions" of the comparative method. However, as Campbell (2004:146-7) observes, "What textbooks call the 'basic assumptions' of the comparative method might better be viewed as the consequences of how we reconstruct and of our views of sound change."

I do apologize but I had to take this out. I doesn't mean anything, basically. Not enough information is given to properly interpret the abstracts. For example, what cases are you talking about? We don't know, I'm sure, and you aren't telling us. And now we have some weasel words, "generally considered", but what does unreliable mean? How so? And what do you mean, lexicostatistics does not work? The next sentences are totally disordered and that quote by Campbell gives us zero assistance. This is zero-grade explanation; nothing is presented, nothing is explained. You think you are reading something but you can't figure out what. You know, I don't know of any other language that allows you to do that, to say something without saying anything. I'm going to start calling English the null language. If I want to be diplomatic I speak French but if I really want to say nothing I switch to English. That way you can't be criticised for something you said because you haven't said it.Dave (talk) 03:32, 18 December 2009 (UTC)

"Although the following article deals only with the role of the comparative method in demonstrating genetic relationship, it is important to realize that this is only one application of the comparative method, which has rightly been described as the central tool of historical linguistics. For example, André Martinet uses the comparative method in his influential Economie des changements phonétiques (2005[1955]) to study the evolution of sound systems over time and, via this, to develop generalizations about the nature of sound systems as synchronic entities."

First phrase is not true. The article does go beyond the evaluation of genetic relationships, and therefore required some reorganization. Then we have this business about Martinet, which purports to be a different use but is described as being one and the same with the goals listed just before the quotation from Schleicher. This article is what used to be called a name-dropper, it throws a lot of famous names around without any explanation of why. We can get the names from any of the literature, what we need is the threads of meaning, the summary of what is going on, why these names are being used. The style is that of the "little professor", a child trying to talk over his head about undeveloped ideas in quasi-grandiloquent style with frequent lapses into informal English, and saying nothing really but doing a lot of name-dropping. You have to tell us how these people fit into the thread. You keep telling us the thread is one thing and then going on to tell us it is something else. Writing is hard work, my friend. You can't sell us linguistics door-to-door. Did you have a summer job selling vacuum cleaners?Dave (talk) 09:14, 18 December 2009 (UTC)

Plural used

The box says "Altaic languages", but only Turkish is mentioned. 1782 and 1786 are both mentioned for Jones. —Preceding unsigned comment added by 217.42.19.146 (talk) 09:22, 18 December 2009 (UTC) In the file, Mongol appears. It is not in the box in the article. —Preceding unsigned comment added by 217.42.19.146 (talk) 10:42, 18 December 2009 (UTC)

Thanks, I'll look into it shortly and report backDave (talk) 10:53, 18 December 2009 (UTC)
  • What you say is true. Many of the illustrations in commons are not completely accurate or leave features to be desired. I suppose the artist means the languages cited are instances of those groups. Full lists would require many articles. He didn't say instances, but it is manfest he must mean that. I was looking for an instance of a comparison table. I saw this. I did try to compensate by changing the caption. If you can find a better, place it in there. My philosophy is to use illustrations even though they may not be perfect is every respect. Have you got a better one? I suspect English may not be your first language, but no English speaker would conclude Turkic is the only Altaic language. Anyway the issue is now up for discussion. Let us see what the readers think (if anything). By the way, if you are interested in contributing on a regular basis, perhaps you should get yourself a user page.Dave (talk) 11:30, 18 December 2009 (UTC)
  • The 1782 is wrong. I fixed it. I'm just beginning to work seriously on this section now.Dave (talk) 11:34, 18 December 2009 (UTC)
  • The failure of the Mongol to appear is the result of the kind of graphic it is. That is a matter for the artist. However I expanded the picture a little and now it is there. Ordinarily WP like to keep these side pics to 300 px but in this case a little larger seems warranted. I don't like unreadable maps and tables anyway. That seems to be about the best I can do without swapping the picture out for another, but first we have to have another!Dave (talk) 11:43, 18 December 2009 (UTC)

Lyle Campbell

Lyle is an established author mentioned in the article. However, the article does not reference any of his published works. Instead we are told of an unpublished debunking passage located in preview form on Lyle's site. That isn't there now; the link is dead. Moreover, the putative forthcoming work is not on Google (maybe I missed it). It seems clear, if there really was such a passage Lyle does not want us to use it. However, there is more. All these claims made for Jones are false, he never said any such things and as far as I can see no one ever thought he did. If Lyle said all those things they certainly were misplaced. I doubt it. However, we can't now verify just what Lyle said. So, I'm taking out the better part of that passage. I will say, the linguists of that time all drew conclusions that later turned out to be manifestly wrong. Anyone reading Jones can see that. No one takes Jones or the others like him seriously as linguists. If we're trying to "debunk" Jones we have created a straw man to belabor. He happened to be the first known of the time to mention a group to which he did not even give name. No one has claimed he invented the proto-language. See Tree model. It is more important to get information right than to get something on WP. More work.Dave (talk) 11:08, 18 December 2009 (UTC)

I took this out:

"An insight often attributed to Jones is conceiving of the idea of a proto-language, and consequently of the type of "family tree" model of language development (one proto-language splitting into various daughter languages, some of those then splitting again into further languages), upon which the comparative method is based. However, Jones' role in the development of these ideas has recently been called into question. According to the comparative linguist Lyle Campbell, the widely quoted passage from Jones has been removed from its proper context, and a reading of his work reveals his ideas of linguistic development as less clear."

I don't know of anyone who attributes those things to Jones (see under Tree model. What role is that? How can you say the role has been called into question and not say what the role is? Moreover, we have already explained the family tree concept, why do it again? As I said above, Lyle has chosen to remove himself from you presentation. Just what are Jones' ideas of linguistic development? All he said was, Sanskrit and some others appeared to have common ancestor now not extant. What's wrong with that?Dave (talk) 12:00, 18 December 2009 (UTC)

The writing of the article

Someone worked pretty hard on this article and there is much good material. It got a C, and now I'm making quite a few changes - not on the technical material, mind you, but on the writing. You should have had more English composition. But anyway I think I owe it to you to explain why I am making these writing changes, and this turns out to be some suggestions on your writing. I do see some consistency of style. As a linguist or student linguist you know all about social registers. An article such as this takes formal English, but the register keeps changing to informal or conversational English. Never do that, it attracts attention to yourself and inserts your presence into the material. The writing should be transparent; one does not see the writer or his style at all, only the material. Stick with one register, formal English, throughout. Second, a lot of this is patois. Patois is a device used by one ignoramus to impress another. Save it for your girlfriend. In patois one tries and fails to use the technical terms correctly and half the time it is mainly clear that the speaker has no idea what he talking about. It can sound impressive, however. Don't try to impress us, it means you have a low opinion of us, that we can accept slick talk in place of the real thing. Whenever I see the slick I stop reading. The third thing I noticed is that decisions on what to say next depend less on strict logical order than on the purple passages. You are trying to "write good." Don't. First have something good to say and then organize it well. You will find that you won't have much trouble saying it. Don't try to look smart or be entertaining. We don't read this material for entertainment, although I must say WP can be a pretty good laugh. And finally, you aren't using natural language. If you read the Egyptian Book of the Dead you will see a goodly number a repetitious formulae as though the author think if he does not say things in the correct formula someone will go to hell for it. Don't keep repeating the subject over and over in successive sentences; use pronouns. You don't have to summarize the whole article in every paragraph; you can assume the reader has read the previous. These are my comments, take them as you will. If you really want to know how to write linguistics start studying the writing of the introductory textbooks referenced so frequently in this article. Of course, they had the advantage of a team of editors. As far as I can see they are written really well. That's their job. No reply necessary. You can just start reworking some of the awful linguistics articles I have had to read, many of which you no doubt worked on.Dave (talk) 00:24, 20 December 2009 (UTC)

For whom is your invective intended? This article has had 538 edits so far: 138 by you, 83 by Sjcollier, no more than 21 by any other single account – so no wonder if there's some inconsistency in register. —Tamfang (talk) 03:06, 21 December 2009 (UTC)
That's an interesting theory, Tamfang - I don't know for whom - don't care - totally non-personal. I thought I saw a uniform bad style. It isn't invective, it is a tart critique. Some people don't take us too seriously and need to be jogged a little. The tartness comes from having to deal with the authors of bad articles who would rather conduct edit wars than allow one word to be changed. If the shoe fits wear it, if not, don't. But, maybe the perception of an overriding style is off. Maybe it is the accumulation factor. So, if it does not apply, forget it, but the suggestions are intended for anyone writing these C or D articles.Dave (talk) 13:29, 22 December 2009 (UTC)

References not used

  • Goddard, Ives (1994a). "A New Look for Algonquian." Paper presented at the Comparative Linguistics Workship, University of Pittsburgh, April 9.
  • Goddard, Ives (1994b). "The West-to-East Cline in Algonquian Dialectology." Actes du Vingt-Cinquième Congrès des Algonquibustes, ed. William Cowan: 187-211. Ottawa: Carleton University.
  • Holm, John (1989). Pidgins and Creoles. Cambridge: Cambridge University Press.
  • Krishnamurti, Bhadriraju (2003). The Dravidian Languages. Cambridge: Cambridge University Press.
  • Comrie, Bernard (ed.) (1990). The World's Major Languages. New York: Oxford University Press.
  • Dixon, R. M. W. (1997). The Rise and Fall of Languages. Cambridge: Cambridge University Press.
  • Pedersen, Holger (1962). The Discovery of Language. Bloomington: Indiana University Press.
  • Picard, Marc (1984). "On the Naturalness of Algonquian ɬ." International Journal of American Linguistics 50:424-37.

This article has a large number of references; however, not too large for the topic. There is no need on the other hand to swell the Bibliography beyond what is used. I put these unused items here. Many articles have "Additional reading" sections. I like those. Many of these refs are very specific rather than general. I'm not sure those belong there. If anyone wants to create an additional reading section with these I would not disagree with that. I did not do it myself because the refs need a lot of work to get them into WP formats, and some of them require tedious Internet lookups. If you want to do the work, go right ahead, I will not object. Note also if anyone puts more notes in and wants to use these items, they would go back in the references section, properly formatted.Dave (talk) 13:36, 22 December 2009 (UTC)

Edit completion

I just completed a major edit on this article, which I hope will remove some of the obstacles to improving its grade and make it more useful in the matrix of other articles. The article is middling long (not really long). Try as I might I found I could not reduce the size below 53 kB. My edits are mainly formatting and English-language. I followed the policy of not removing referenced information. I did insist on following the references so I removed some unreferenced paragraphs that were written badly or made no sense, as I have explained above. If anyone really wants to cut down, I would suggest some sections have too many examples. Maybe there are too many tree diagrams. You use YOUR judgement whether you want to add or subtract, merge or split. I got this to the point where I view it as satisfactory so I may not be back at least for a while. You can always reach me on my discussion page and I will discuss almost anything but I have to warn you I will not bicker. If you really have to change something, do it, and see what the reaction is. It's not my article, it's yours. Bye now.Dave (talk) 13:50, 22 December 2009 (UTC)

wave model

Speaking of bad style:

The Wave Model has been proposed as an alternative model of language change.[60] Each wave, an isogloss, is a circle in the Venn diagram, but the circles are not to be seen as simultaneous or extending over the same areas. The language must be found most certainly at the intersection of the greatest number of circles. It tapers off to intermediate times and locations. Some isoglosses may not even be found in languages of the same family. The tree model presumes that all the circles coincide in time and space.

The first two sentences are clear enough (though I don't see why "not ... simultaneous"). The last sentence is a bit obscure but I got the point. What the heck do the other three sentences even mean? —Tamfang (talk) 16:35, 22 December 2009 (UTC)

And why remove the wave model link?! —Tamfang (talk) 16:56, 22 December 2009 (UTC)

That is what I thought, Tamfang. Argue away in whatever monologue you care to select. I'm sorry if you took it personally. Maybe you should think harder about your writing style and your ability to express ideas. If it helps you to resent me, go right ahead, but it won't improve your style. For the link to Wave Model, I believe you are speaking of the "main" template. Wave model (linguistics) is not the main article. The section is not about the wave model. If you really want to see a link, it might be best to put in a "see also" template or some such thing. If you need help selecting a template, they are covered in a help section. Let me get it for you: Wikipedia:Template messages. You may have to do a little work to find the right one, but since you are showing such an interest, that should be no problem. For myself, I think we have plenty of links to it and do not need another. I did put one in the diagram. As for your not understanding a Venn Diagram, that is a serious deficit for a linguistics person. Did you read the linked article on Venn diagrams? Have you had any logic? Just what did YOU think diagrams such as this are? If you lack understanding of logic, you need to start with such expressions as A AND B, A OR B, A XOR B, which are more or less standard. In computer programming these are relational operators. Without that I am afraid you can't really understand the wave model diagram. Did you think they are pictures of waves? What happens when one wave intersects another? Anyway we can't explain all that in one article such as this one, which is not about the wave model anyway. There is a link to the Venn diagram. I would say, you are trying to pin your confusion concerning the subject matter on my style, and what is more, since your approach is clearly confrontational, I am going to let you get clean away with it. Take responsibility for your own education, you're supposed to be a grown-up. I'm through with this article so if you are going to make changes to it, it is all the others you will be facing. I've done my bit. Don't call me. I'll call you.Dave (talk) 00:42, 23 December 2009 (UTC)
Eh? The passage that I criticized here did not replace anything I wrote, so my ox is not gored. (Without checking the history, I'd have a hard time pointing to any sentence in the article that I did write.) I often lament that I have many years' practice in making myself misunderstood, but I don't think my fallibility makes my criticisms entirely worthless. ¶ As for "confrontational" – oh I am such a dark kettle. ¶ I'm not familiar with this usage of the word 'monologue'. ¶ Nor is it helpful to say that only my own maleducation is to blame if I can't make sense of gibberish that happens to use technical terms. The sentence containing "Venn diagram" is one of the two that I described as "clear enough", so that barb is wide of the mark.
When I wrote the above, the number of links to Wave model in that section had gone from one to zero, a deplorable development whether that link is in Template:Main or otherwise. I see that there is one now. (Thanks for the disambiguation, though Wave model redirects.)
I'll repeat the sentences that give me trouble.
  • The language must be found most certainly at the intersection of the greatest number of circles. It tapers off to intermediate times and locations.
One way to make sense of this is to take "The language" and "It" as meaning a core standard where the essence of the language is to be found, in contrast with the mere 'dialects' of the periphery; a politically incorrect concept. Another is to suppose that the Venngram in question shows only those changes relevant to the history of a (hypothetically) specified dialect; I imagine it's sometimes done that way, but I certainly wouldn't assume it, and if so why is 'Innovation A' shown? I eagerly await suggestions of other ways to understand the passage.
  • Some isoglosses may not even be found in languages of the same family.
On second thought, I understand this to mean that isoglosses may meaningfully be drawn between un"related" members of a Sprachbund (though without the Sprachbund concept one would think such isoglosses redundant!), and such isoglosses may not happen to cut through any family. Though a valid point, that's a somewhat confusing change of subject from what I take to be the usual application of Wellentheorie, viz describing changes within a dialect continuum, which is more relevant to this article.
—Tamfang (talk) 05:02, 25 December 2009 (UTC)

External links

The two external links are really to private sites. I don't think they are appropriate, but the previous editors had them, so I am going to leave them. I am insisting on cite web, however, according to WP policy. We aren't interested in editorial interpretation of the link. Cite web puts it in as it is, telling us who and what. Now the link to UT publishes material copyrighted I am sure by Kathleen Hubbard. I do not know at all if they have permission to publish it. It is a document on the classics department web site and that has been made public. What their use of it is and whether we can link to it I do not know. You legal persons now can decide. The other one is Professor Matthew Gordon's linguistic course aid. Is this an encyclopedic source? You decide. I presume he does not mind the whole world taking his course as he has made the site public. There is some interesting information in there, but, the thing is, its nature is didactic not encyclopedic. I believe both links are strictly transitory. In a few months, a year perhaps, someone will be having to remove the links as dead. Are they worth the trouble? You decide.Dave (talk) 00:57, 23 December 2009 (UTC)

Different types of evidence: lexicon vs. morphology

Not all elements of the language have the same probative force when comparison suggests a historical connection. The strongest lexical evidence for any language family is shared morphological paradigms, especially highly irregular or suppletive paradigms with bound morphology, as Fortson (2004) points out (for example, I'm sure other works on historical linguistics do the same). Pronouns are also generally considered likely if not unassailable indication of genetic relatedness (but if there is no evidence for past contact and the languages are not spoken anywhere near to each other, and there is no independent indication that they ever were, or were in contact, except possibly in the remotest past, borrowing can be ruled out with high probability, which makes the case of Kusunda interesting indeed). Independent lexemes, on the other hand, do not constitute compelling proof for relatedness at all, even when the sound correspondences are regular (for example, Latin loanwords in Welsh do not prove that Welsh is a Romance language, even though regular sound correspondences can be established), or when the languages are spoken in a very extensive geographical area such as Eurasia (as Wanderwörter can travel very far distances). The article should somehow mention this important fact that not all evidence is equally good, but I can't tell where it would fit best. Anyone? Florian Blaschke (talk) 00:29, 24 December 2009 (UTC)

I see that the widely accepted method of subgrouping by means of (preferrably non-trivial) common innovations (as long as relative chronology does not contradict - wait, how can an article about the comparative method omit a subject as important as relative chronology?) is not mentioned either, even though I'm sure that the handbooks treat it.
Also, where's proportional analogy, or paradigmatic levelling? The only form of analogy given as example is an instance of serial analogy, a rather marginal form typically affecting only single lexemes.
Analogy (at least proportional analogy, which is a well-known and easy to observe process) generally deserves to be treated as a complication, rather than a problem with the method itself, as accounting for analogical changes is part of the method.
It is worth pointing out that sound change is regular, but creates irregularity, while (proportional) analogy is irregular, but creates regularity, and that the two forces are in perpetual competition, cancelling out each other's effects. Florian Blaschke (talk) 17:59, 27 December 2009 (UTC)

Dubious Piraha Claims

The language Piraha has been subject to a number of dubious claims of uniqueness by Daniel Everett. "Piraha Exceptionality: a Reassessment", Language, 85.2, 355-404. The claim that its pronouns are borrowed from Nheengatu is not transparent. Thomason and Everett argue their case based upon a complex phonemic analysis.

Our claim is that the basic Piraha pronouns are nearly identical to those of Nheengatu and Tenharim. Superficially, however, the Pirah~a pronounsdon't look much like the Tup i-Guaran i pronouns; so this proposal will not be convincing without some additional information about the phonology of Piraha that shows how the phonetic realizations of the Tup i-Guaran i forms align with the Pirah~a phonemic system. "Pronoun borrowing" Sarah G. Thomason & Daniel L. Everett University of Michigan & University of Manchester

The claim should be attributed rather than presented as fact.μηδείς (talk) 16:40, 5 July 2010 (UTC)

The Uralic and Altaic pronoun table - false cognate

What is the point of having the pronoun table on the "comparative method" page? The resemblance of the Baltic-Finnic pronoun "sinä" and Turkic "sen" is obviously coincidental: the change of t -> s before "i" is a Baltic-Finnic innovation. All other Uralic languages have a "t" in that position (*tun, *tinä). Actually Finnish has the verbal 2nd person suffix "-t", and also the plural form "te" shows the original "t".

Compare also the following inflections (nominative sg - essive sg - essive pl):

vesi - vetenä - vesinä (water) käsi - kätenä - käsinä (hand) kuusi - kuutena - kuusina (six)

--Muhaha (talk) 13:39, 27 August 2010 (UTC)

So your point is that Uralic is actually closer to Indo-European?μηδείς (talk) 17:02, 27 August 2010 (UTC)

The chart adequately attributes the opinions held to the linguists who hold them, which is the only relevant matter - not anyone's OR POV, neither mine nor thine. It should occur to you that a pre-Altaic form in *tin- could easily account for the *sen- of Turkish, especially in light of the the *chi- of Mongolian. If anything, the chart could be expanded to include the mine/thine pronouns of PIE and the nyi/chi pronouns of Nivkh.μηδείς (talk) 18:49, 5 September 2010 (UTC)

The table should also include Proto-Uralic and Proto-Altaic (and PIE) reconstructions, and regarding the gender thing (he/se), PIE didn't have "she", but only animate and inanimate. --88.114.40.114 (talk) 19:35, 5 September 2010 (UTC)

Yes, it would have been better with the proto forms, and you are right about the pronouns as well. The problem is that this table is done with an image file, not editable wikipedia markup. It still has value since it does illustrate the beliefs of the long range comparativists like Poppe who are mentioned in the subcaption. μηδείς (talk) 19:40, 5 September 2010 (UTC)

Origin and development of the method

The history section of the Indo-European studies article gives a different impression of the origin of historical linguistics than this article does, mostly pushing the method much further back in time. Even in that article, there is no mention of Yehuda Ibn Quraysh, who sometime in the 9th or 10th century thourhg comparison of the phonology and morphology of Hebrew, Aramaic, and Arabic recognized that Semitic languages had a common origin and "are subject to the same linguistic laws". This reference states that the Indo-European studies in the 1640s by Van Boxhorn and Elichmann were "significantly more accurate" (and more comprehensive) than those 140 years later by William Jones, "who erroneously believed that Egyptian, Japanese and Chinese were part of Indo-European while Hindi was not, which suggests that his method was seriously flawed." The reference also mentions that 67 years before Jones, William Wotton had already attempted to reconstruct an Indo-European proto-language.

Perhaps I'm misunderstanding how narrow the concept of the "comparative method" is, in which case the difference of which with earlier methods could perhaps be made more explicit in the text. At the very least, the overcredited Jones' overexposed quote, which as far as I can tell told nothing new, can be removed from the origin section. Afasmit (talk) 00:22, 30 September 2010 (UTC)

Reading the source material for Ibn Quraysh it's obvious he is only restating the stereotypical religious based "all languages evolved from Hebrew" doctrine:

"The reason for this similarity and the cause of this intermixture was their close neighboring in the land and their genealogical closeness, since Terah the father of Abraham was Syrian, and Laban was Syrian. Ishmael and Kedar were Arabized from the Time of Division, the time of the confounding [of tongues] at Babel, and Abraham and Isaac and Jacob (peace be upon them) retained the Holy Tongue from the original Adam."

In other words, Abraham and Isaac and Jacob retained God's tongue, not that Hebrew and Arabic evolved from some third tongue that was neither. This whole section needs to be reverted to the status quo ante. μηδείς (talk) 02:48, 16 April 2011 (UTC)

Mistake

It is stated that latin word for "tongue" is "dingua", but as far as I can tell, it should be "lingua". Even so, linguistic is not absolutely my field of expertise, so I wanted to signal the thing instead of changing it. Anyway, yes, your are linguistic not dinguistic, so I'll be quite surprise if that is right.

87.0.49.19 (talk) 22:22, 28 March 2011 (UTC)

blench vs jones

Jones' work is highly cited and should be available, but Blench does not provide any reference for his claim that Jones believed that Chinese, Japanese and Egyptian "were" Indo-European. Given Jones' remarks about the detailed parallels between Sanskrit, Greek and Latin, and the fact that Jones didn't use the term Indo-European, it seems likely something else is going on here - perhaps a surmise on his part that the three tongues were related at some remove? In any case, we need a source in Jones for this, not an unsupported vague criticism in Blench. μηδείς (talk) 02:35, 16 April 2011 (UTC)

I have restored the status quo ante. What is unique about Jones is his positing of some no-longer existent independent source, Blench's unsourced cavils notwithstanding. Ibn Quraysh does not posit the origin of Hebrew and Arabic in some common and no longer spoken proto-tongue, but attributes Hebrew to the Patriarchs with other's being Arabicized away from God's original Hebrew. The same naive pre-scientific mythological view holds for the Romans. They believed their tongue was a debased form of Greek, not that it and Greek both evolved from some distinct third (proto-) language no longer spoken. μηδείς (talk) 02:59, 16 April 2011 (UTC)

What? Blench's criticism is not at all "unsupported" or "vague". By all means, take a look at Jones' remarks in the Third Address to the Asiatick Society and you will see that the criticism is mostly accurate. For example, Jones explicitly says that Hindustani belongs to a different stock than Sanskrit (unlike Latin, Greek and Persian), and is only influenced by it.

Also, he clearly fails to differentiate between the history of "races" (i.e., ethnic groups) and languages (and even between the history of spoken language and scripts), and often conflates his bizarre speculations about the origins of "races" (i.e. that modern Indians are related to Egyptians, Chinese and even "Peruvians") with his linguistic speculation.

if you remove that single fortuitious quote on IE, most of Jones' ramblings would be considered nonsense today, and certainly inferior to the contributions of many of his contemporaries or predecessors. The fact that he is so often glorified and quoted has more to do with the fact that he held a prominent position and was a pioneer, than the quality of his work as a scholar. Again, by all means, go and read him, instead of relying on secondary, tertiary, etc. sources.

Finally, his claim of a no longer existent independent source was neither that decisive not definitively accepted, since several decades later Sanskrit itself was still claimed to be the IE proto-language, as Schleicher's quote in this very article makes clear. What we really have with Jones is an easily quotable paragraph from a famous scholar of a prestigious institution. Certainly, the fact that he was somehow the "father" of comparative linguistics is an attractive narrative, but hardly accurate.KelilanK (talk) 16:12, 16 April 2011 (UTC)

Thanks. Perhaps you can provide a link to the incriminating remarks in the Third Speech to the Asiatick Society? Blench, whom I have read at length, largely admire, and whose website I have long had bookmarked, didn't provide a link or even identify the text. I understand your personal opinion of Jones differs from that of Scholars such as Winfred Lehmann. Have you written any works we can cite as reliable sources? The fact remains, whatever Jones' wider errors, which would be relevant in an article on him, and not on the development of the Comparative Method, that his supposition of some common source which no longer exists was an essential observation cited by later scholars. μηδείς (talk) 16:29, 16 April 2011 (UTC)

http://books.google.com.br/books?id=ZiYNAAAAYAAJ ----- pages 33 and 34 ("Third Discourse")contain references to Hindi ultimately having a "Tartarian" or "Chaldean" origin and being influenced by Sanskrit, as well as the famous "philologer" quote.

In pages 45-46, he seems to add "Chinese, Japanese", "Phoenicians" and "Peruvians" to the same "race" of speakers of "IE" languages.

In the Sixth discourse he concludes that Farsi is derived from "Chaldean" (and thus related to what we'd call Semitic languages) and unrelated to Avestan.

And see page 186 ("Ninth Discourse"): "that the first race of Persians and Indians, to whom we may add the Romans and Greeks, the Goths, and the old Egyptians or Ethiops, originally spoke the same language and professed the same poupular faith is capable, in my humble opinion, of incontestable proof". In this and adjacent pages, Jones again draws a distinction between the first and second Indians and Persians: Sanskrit and Avestan are what we now call IE, Farsi is Semitic, Hindi is either "Tartarian" or "Chaldean". Egyptian is also IE. And he again says, though with less confidence, that Chinese and Japanese might be related to the "Hindus" (apparently the "First" Hindus, i.e. Sanskrit), etc.KelilanK (talk) 20:48, 16 April 2011 (UTC)

Thanks, I read the material you linked to at Google. It was very, shall we say, interesting! Not the most sober, skeptical or empirical of minds, was he? Of course Copernicus and Newton had their own screwball ideas as well, and the comments on race are, strictly, divisible from those on grammar. I think adding something like, "While, as Blench comments (Blench ref), Jones held all sorts of ideas that would be dismissed as crackpot today (Link to Jones), his observation that (blockquote) is widely quoted by modern students of the comparative method. μηδείς (talk) 21:09, 16 April 2011 (UTC)
We could also use a link to the works of the Dutch Scythianists. μηδείς (talk) 21:09, 16 April 2011 (UTC)

Origin

The transition from "Origin" to "Application" in the text is too abrupt. How did the linguists of old really apply the method, by and large? And how has its application changed? The history part gives some clues, but very little.Xemoi (talk) 05:24, 12 December 2013 (UTC)

Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Comparative_method_(linguistics)/Archive_2&oldid=844845793"