Talk:Diaphoneme/Archive 1

OED

I have the impression that the first Oxford English Dictionary used diaphonemes, whereas the second changed to IPA representation of Oxford dialect. Is there an authoritative statement? —Tamfang (talk) 17:59, 20 June 2010 (UTC)

The abstract to a review of the second edition of the OED says: "The worst innovation is argued to be the phonetic transcription, a literal transfer from the Murray system to the International Phonetic Alphabet (IPA), which results in a flawed, purportedly diaphonic transcription violating IPA principles & confusing 19th- & 20th-century pronunciations." — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 00:04, 21 June 2010 (UTC)

Trask's definition

FWIW, here's the definition given in Trask's The dictionary of historical and comparative linguistics (my italics):

A single abstract segment which occurs with the same distribution in several different varieties of a language but which is represented by a phonetically different phoneme in each.

Seems incompatible with the more abstract description given here. Perhaps Trask is not of the generativist school -- I don't know. Grover cleveland (talk) 06:16, 24 June 2010 (UTC)

I don't understand what he means by "a phonetically different phoneme". Sounds like he might mean that the dialects are phonemically equivalent, in which case wouldn't they be accents rather than dialects? — kwami (talk) 06:21, 24 June 2010 (UTC)

That is not the only definition Trask has provided. In A dictionary of phonetics and phonology, he provides two similar, though conflicting definitions of diaphone. His definition of diasystem implies that he is under the impression that diaphon(eme)ic representations are different from systems that set up more units than occur in any one variety.

Two varieties that differ in their phoneme count may still be considered different accents in the technical sense. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 07:02, 24 June 2010 (UTC)

I would guess that "a phonetically different phoneme" means "a phoneme which, in the same environment, would have a different phonetic realization". However, the problem I see with his (first) definition above is that it seems to rule out use of diaphon(em)es across accents with systemic differences. For example, we might want "car" to be something like |kɑr| across both rhotic and non-rhotic accents (as in WP:IPAEN): however according to my understanding of Trask the diaphon(em)e |r| cannot be present in this word because it is not "represented by a ... phoneme" in non-rhotic dialects (unless zero is considered a phoneme). Grover cleveland (talk) 23:58, 24 June 2010 (UTC)

Well, it is /kɑr/ in conservative RP, but maybe I get your point. So we couldn't have a Traskian system for dialects with and without the caught-cot merger, right? What of the pin-pen merger? — kwami (talk) 06:43, 25 June 2010 (UTC)

No it's not. It's /kɑː/ in all forms of RP. Grover cleveland (talk) 21:27, 25 June 2010 (UTC)

In the description of RP given at the British Library, it's /kɑr/, since they describe RP as having linking ar.[1] (I don't know how accurate that statement is these days, and they are not clear as to when people have linking ar as opposed to intrusive ar.) — kwami (talk) 05:46, 27 June 2010 (UTC)

Are you guys looking for published works that represent diaphonemes as we do at WP:IPA for English? I hear Rudolph Troike's (1969) Receptive competence, productive competence, and performance (Georgetown University Monograph Series on Languages and Linguistics 22.63-74) has an example of an abstract diaphonemic system using mergers. Going past English, Wheeler's chapter on Occitan in The Romance Languages does something like this for Occitan. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 17:19, 27 June 2010 (UTC)

• Kwami, the website which you link to (which is scarcely a scholarly work on phonology, BTW), represents RP as having both linking and intrusive R. Therefore it provides no support for a view that "car" is /kɑr/. Grover cleveland (talk) 22:52, 1 July 2010 (UTC)

Are you sure about that? It looks like it points out that "conservative RP" does not have linking R calling the absence of linking r "zero linking r" — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:29, 1 July 2010 (UTC)

Maybe I'm misunderstanding the website, but it seems to say that "conservative RP" has "zero linking R" -- i.e. no R sandhi at all, while non-conservative RP has both linking and intrusive R. Grover cleveland (talk) 23:58, 14 July 2010 (UTC)

Wells says that he himself has linking ar only after the vowels which most recently dropped following ars before consonants, and that younger people don't have linking ar anywhere. But AFAIK (which isn't far), people trying to sound conservative still use linking ar, at least when reading aloud. — kwami (talk) 01:54, 2 July 2010 (UTC)

No he doesn't. He says that he lacks R sandhi after his THOUGHT set. I don't think he says anything explicit about NORTH (there are very few words with the NORTH vowel word-final) but I would assume he lacks it there too. Grover cleveland (talk) 23:53, 14 July 2010 (UTC)

Alas that we can't ask Trask. (I had the good fortune to be on a couple of listservs with him, and consistently found his remarks entertaining.) —Tamfang (talk) 07:38, 24 June 2010 (UTC)

Great article

Thanks for all your work on this, Aeusoes1! 69.12.144.158 (talk) 18:25, 16 July 2010 (UTC)

No problem. There's more on the way, too. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 03:37, 17 July 2010 (UTC)

GA?

Could this be a good article candidate? I confess almost total ignorance of the GA/FA processes, but this certainly looks good to me. Lfh (talk) 18:46, 7 April 2011 (UTC)

We could give it a try. I believe the article would get some constructive criticism in the process, which would be helpful even (or especially) if it doesn't get FA/GA status. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:58, 7 April 2011 (UTC)

GA Review

This review is transcluded from Talk:Diaphoneme/GA1. The edit link for this section can be used to add comments to the review.

Reviewer: -- Cirt (talk) 06:11, 20 May 2011 (UTC) I will review this article. -- Cirt (talk) 06:11, 20 May 2011 (UTC)

Successful good article nomination

I am glad to report that this article nomination for good article status has been promoted. This is how the article, as of May 21, 2011, compares against the six good article criteria:

1. Well written?: Indeed, quite well written. However, I would note there are a few places with long sentences, and run-on sentences. Could use a bit of coypediting, from third party previously uninvolved users, with this in mind, going forward. I would suggest doing this, under the process of the Peer Review. Passes here.

2. Factually accurate?: Duly cited throughout, copiously, to WP:RS sources. Passes here.

3. Broad in coverage?: Yes, quite thorough, covering multiple different aspects of the subject matter - in a clear and concise structure. Passes here.

4. Neutral point of view?: Neutral tone throughout, no issues here. Passes here.

5. Article stability? Upon inspection of article edit history and talk page, no outstanding issues with stability. Passes here.

6. Images?: No images used in article, thus no problems, nothing to evaluate. Passes here.

If you feel that this review is in error, feel free to take it to Good article reassessment. Thank you to all of the editors who worked hard to bring it to this status, and congratulations.— -- Cirt (talk) 02:56, 21 May 2011 (UTC)

Labovian linguistic variable

The article has been containing a claim that the Labovian "linguistic variable" is a similar concept to the diaphoneme. That was the opening of a paragraph with several citations. However, none of the citations was about comparing the diaphoneme to the linguistic variable, which makes the claim unsourced. In any case, the claim is false. The two concepts have quite different purposes. Therefore, i deleted this material. Dale Chock (talk) 09:50, 2 March 2012 (UTC)

They do have different purposes, which the article outlines, but they also have some similarities in what they attempt to cover and the theoretical thinking behind them. Hopefully that's clear enough with the article's coverage. I have put an additional citation to a source that says, briefly, that they are essentially the same. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 14:30, 2 March 2012 (UTC)

Diaphonemes are interdialectal, sociolinguistic variables the opposite. Also, one value of a sociolinguistic variable is often null, which is not consistent with the definitions of allophone and diaphoneme. I notice you did not identify any of the similarities you attribute to these linguists' constructs. For you to just say "read the article" doesn't cut it. Dale Chock (talk) 07:13, 3 March 2012 (UTC)

Diaphonemes, unlike sociolinguistic variables, do not have sociological and stylistic implications. You wanted this article to be about the diaphoneme. Dale Chock (talk) 07:18, 3 March 2012 (UTC)

Most of those aren't accurate differences. Both diaphonemes and linguistic variables cover multiple varieties. Both are potentially cognitively real. It is false that a null value is not a viable option with diaphonemes or allophones. There are differences, but there are also enough similarities that a mention of the linguistic variable is worthwhile.

Keep in mind also that linguistic variable links here. If someone creates that article, we can subsequently shorten its coverage here. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:28, 3 March 2012 (UTC)

Spurious citation for "a linguist's construct".

The article read:

Wells gives the example of straight, late and wait, which rhyme in most English varieties but, because some dialects make phonemic contrasts with the vowels of these words (specifically, in regions north of England), a panlectal transcription would have to encode this contrast despite it being absent for most speakers, making such a system "a linguist's construct"[footnote: Wolfram] and not part of the grammar present in any native speaker's mind (which is what adherents of such a system attempt to achieve).[footnote: Wells]

The phrase "a linguist's construct" was in quotes because it is a direct quote from Wells (i have verified that it is). But it was also footnoted with a statement, "According to Wolfram (1991:25), the linguistic variable is essentially a sociolinguistic construct". This footnote was irrelevant. The variations being discussed, like distinct pronunciations for 'straight', 'late', and 'wait', are not sociolinguistic variations. Thus Wolfram's statement, while largely true, has no bearing on the topic of the paragraph. Besides, most of linguistic explanations invoke linguists' constructs. There is another flaw in the citation of Wolfram: it's a slight distortion of what Wolfram said, which was much more qualified:

The linguistic variable in early variation studies was obviously motivated

by the desire to reveal the most clear-cut pattern of social and linguistic covariation. In this respect, then, it was essentially a sociolinguistic construct rather than a strict linguistic one, a kind of language-based analogue of the

social variable in sociology.

Note the use of the past tense, not the present, and the context of a certain short time in history, for the theory of the so called linguistic variable was rapidly evolving. The article in its current state is full of improper use of the simple present tense. The impropriety is not grammatical, but factual: the editor is discussing theorizing, some of it 70 years old, not unquestioned theory. Dale Chock (talk) 10:48, 2 March 2012 (UTC)

I've switched the placement of the two citations, which should clarify who said what. I'm not so sure I've addressed your other concern. Are you saying that, according to Wolfram, the linguistic variable only functioned as a sociolinguistic construct in the beginning of its use by scholars but that it became less of one later on? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 15:09, 2 March 2012 (UTC)

As to your second question first, you keep unjustly reattributing what some source said to me. Next, Wolfram's observation was irrelevant. "Phonemes" and "clauses" are linguist's constructs. In this context, Wells meant that this particular application of this construct would not correspond to what speakers "know". It makes little sense to assert that sociolinguistic variables are not real to speakers. Sociolinguistic variables are highly charged for speakers. The fact that the phrase "linguistic variable" is a truncation of "sociolinguistic variable" has nothing to do with linguists contriving an explanation that has no psychological reality. And it probably needs repeating that the dissimilarities under discussion are not sociolinguistic. The point with this particular citation is that it exemplifies how you generally overlook nuances in your citations, and tend to cite everything you come across as if it were permanently established teaching. Dale Chock (talk) 07:06, 3 March 2012 (UTC)

I asked for clarification on your interpretation of Wolfram. That has nothing to do with justice, so please calm down.

You are saying that "construct" means something differently to Wells than it does to Wolfram, correct? I can see that. However, considering that the "social variables" that Wolfram is comparing them to are themselves the construct of the researcher (and not of the target population) there is still a parallel to the beliefs. As Wolfram states in his conclusion:

I think that we will find that the linguistic variable will again turn out to be most adequately justified as a convenient, largely heuristic construct that points the way to the "best fit" between linguistic and social factors.

As we can see, Wolfram is indeed asserting that linguistic variables are not cognitively real. You may not like this position, but that doesn't mean it is not pertinent to the article (even if we eventually decide to move coverage of the linguistic variable somewhere else).

Given all this, I think the best option is to reword the note. How would you suggest we change it to reflect Wolfram accurately? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 17:05, 3 March 2012 (UTC)

1. I am asked: "You are saying that "construct" means something differently to Wells than it does to Wolfram, correct?" No, that is obviously NOT what i am saying.
2. Let Wikipedians take note of a basic point about current linguistic theorizing. Linguists rarely insist on the cognitive reality of their constructs. In this article as written before February 2012, there was a naive unawareness of this fact. It is one thing to set the goal, as Chomsky did, of discovering the psychological reality of language processing and language use. And the Chomskyans in their first 15 years or so became notorious for overweening confidence in their progress toward the goal. Furthermore, we can make mention of the hope that in the ensuing half century, genuine, if minor, progress may have been achieved. But the critics of Chomskyan theories are not necessarily in the habit of claiming that they have the "cognitively real" (AEsos's choice of words) analysis, versus Chomsky. No, most critics of whatever Chomsky's latest theory is at a given time, and perhaps many Chomskyans, are humble about claiming "cognitive reality". It's one thing to reject a proposal on the grounds that it seems wildly psychologically unreal on its face. But any such arguments, even when compelling, are still impressionistic at this stage of the science. Linguists, as far as i am aware, do not have mature evidence-based criteria for measuring the psychological reality or cognitive reality of their hypotheses and constructs. Of course, one of main subject areas of psycholinguistics would be discovering the representation of grammar in the brain, and psycholinguistics should have achieved some modest progress in that regard. In conclusion, we have one Wikipedian who places unwarrranted focus on this issue.
3. Just to ensure that another flaw is not lost sight of. AEsos's version tended to insinuate that the linguistic variable was "just a linguist's construct", unlike the diaphoneme, citing JC Wells as an authority, whereas Wells rejects the diaphoneme construct. Dale Chock (talk) 05:30, 8 March 2012 (UTC)

Newton (1972)

The article said that the usefulness of the diaphoneme is demonstrated in Newton (1972) (The Generative Interpretation of Dialect: A Study of Modern Greek Phonology). But with Google Books one can determine that the word 'diaphoneme' does not occur in that book. Dale Chock (talk) 11:13, 2 March 2012 (UTC)

As Kazazis (1976), cited in the article, says:

N makes no hard-to-swallow claims concerning the psychological reality of his posited underlying overall-pattern (diasystemic) forms, or of the rules which derive the modern phonetic representations of those forms.

While it's true that diaphoneme doesn't show up (nor do diasystem or overall-pattern), it only takes a bit of common sense to look at what Newton has presented, combine it with Kazazis's characterization of it as diasystemic and our given definition(s) of diaphoneme to see that we are presented with something diaphonemic. I hardly see this as a WP:SYNTH violation, though my understanding of the subject may be too sophisticated to see what is and is not common sense to a lay reader. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 23:02, 2 March 2012 (UTC)

It's synthesis of mistaken beliefs. You want to think diaphoneme is a general, uncontroversial concept that any linguist would adopt. No; it's a theory specific concept not used by generativists. Newton had no control over a reviewer's choice of words. The terms "overall pattern" and "diasystem" were relics of the 1950s. Dale Chock (talk) 06:40, 3 March 2012 (UTC)

I'm sorry, but it is false that those terms are "relics" of the 1950s as usage persisted even into the 1970s. It is also false that generativists would not or have not used the concept of the diaphoneme. What about Sledd (1966)? Troike (1971)? Reed (1972)? The attribution to Newton (1972) having a diaphonemic approach has been given. I think this is sufficient. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 17:28, 3 March 2012 (UTC)

"1966, 1971, 1972." 1966 was only six years after 1959! And 1971, 1972? I'm not talking about 40 years ago, i'm talking about now! And as for generativists using "diaphoneme" in the first two years of the 1970s, those were relic usages then. More importantly, the generativist notions about a diasystem were a rejection of the 1950s notions of a diasystem. Generative phonology eventually overcame the inertia of using "diaphoneme", "diasystem". And 40 years later, it is not general linguists, but dialectologists and specialists in a single language, who resort to this relic term. AEsos does not comprehend the key finding that there is no diasystem possible which is linguistically reasonable and cognitively realistic for a wide span of a dialect continuum. At best a diasystem can be constructed for two or three dialects that differ trivially from each other. Dale Chock (talk) 04:00, 8 March 2012 (UTC)

Even more compelling to refute your point: JC Wells (1982:68ff), which you quoted from extensively. Three hours BEFORE you posted this "I'm sorry, but it is false . . ." comment, I had posted how you missed the point of those pages. As you wrote for the article, Wells mentioned the diaphoneme, and he gave some admonitions about how linguists could go wrong using certain analyses or certain presumptions. As I noted at the time which was three hours earlier (Point No. 1, "Call for comprehensive revision"), Wells (1982:68ff) was the work of a generativist who rejected the diaphoneme concept and term. You did not learn the lesson. Dale Chock (talk) 04:53, 8 March 2012 (UTC)

It is true that generativists (including Wells) don't currently use the diaphoneme as part of their conceptual framework; similarly, the closely related diasystem is rare. However, if we're talking about generativists of the 1970s, it is clear from the examples I've listed that generative linguists at Newton's time could indeed (and did) use both the diaphoneme and diasystem as part of their conceptual framework.

Newton does not pretend that his diasystemic/diaphonemic treatment is cognitively real. Because the article uses Newton (1972) to illustrate the point that the use of diaphonemes can still be linguistically reasonable despite this lack of cognitive reality, it should be clear that this is not intended to be an example of a cognitively real diaphonemic analysis. Is that not clear with the way it's worded? I admit that I may not always be able to see what's written clearly and what is vague or unclear. Or is your issue that you disagree with the point Newton is being used to illustrate? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 05:01, 8 March 2012 (UTC)

Insistence on consensus is invalid at Wikipedia

Recently, an editor has insisted on there being consensus before edits are made. I have seen this ploy once in a blue moon at other articles. It is a type of WP:OWN. It is incompatible with how Wikipedia works. Dale Chock (talk) 13:35, 5 March 2012 (UTC)

It is exactly how wikipedia works. You need to collaborate and if there is not consensus for including material it stays out.·ʍaunus·snunɐw· 18:20, 5 March 2012 (UTC)

Sorry Maunus, but in this situation it's the opposite. Dale is removing several paragraphs of information. In this sense, he needs to convince others that there is a better place for it. Simply removing it is disruptive. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 21:42, 5 March 2012 (UTC)

While it would be fun to point out the problems in that accusation, I would rather stick to the content of the article. If you really believe that I am acting improperly, I suggest you report me in the appropriate channels rather than attempt to use veiled threats as a rhetorical strategy in the talk page. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 17:36, 5 March 2012 (UTC)

In reply to all the above three replies. [I corrected indents above]

• Let readers be aware that prior to my very recent participation, editor AEsos had this article to himself. 99% of its content is from him. From 24 June 2010 to February 2012, the expected "crowd sourcing" did not occur, and the hoped for review by editors with expertise did not occur. If there were some editor who wanted to say that World War II ended in 1947, then to treat that editor as an equal collaborator would spoil the mission of Wikipedia — the above two commenters disagree (one of those commenters is like an editor who would say that about WW2). I am convinced that the mission of Wikipedia is to produce/be an encyclopedia, a high quality one; and convinced that what matters in editing an article is that statements would survive review by experts (amateur and/or professional), and that statements be well sourced.
• AEsos wrote that a linguistic work from 1951 was providing a generativist perspective! What he knows is the International Phonetic Alphabet. He has shown in two articles he edited (this one and Diasystem) and their Talk pages that he has no expertise with which to "debate the content of" a linguistics related article. His only means of replying to an opponent's reasoning is to pluck a word here and a word there; i.e., a mere "Smith said such-and-such". Almost all his edits are revertible because he doesn't know what he's talking about. I can point readers to paragraphs and paragraphs of my documentation of that fact.
• Understand that AEsos is calling for "consensus" BEFORE editing, not after editing. That means he's demanding that new editors get PERMISSION from old ones (again, he's been the sole old editor for the past year and a half). That is a sort of assertion of ownership. AEsos was the first to accuse an editor of ownership in regard to either of these two articles. It wasn't on this talk page, but in a Wikiquette discussion. So AEsos comes across as a hypocrite.
• Maunus seems to confuse "get consensus" with "collaborate".
• By the way, the Reliable Source requirement is a minimum requirement for inclusion, not a sufficient requirement. Dale Chock (talk) 04:00, 8 March 2012 (UTC)

I can see, because you have not reported me, that you don't actually believe that I am attempting to own the diaphoneme article. Moreover, your he-started-it comment that I was the "first to accuse an editor of ownership" tells me you either genuinely can't tell the difference between your behavior and mine or that you are just trying to throw muck at me to see what will stick.

I'll spell out what makes your accusation baseless. There's nothing in WP:OWN that says or implies that being the primary contributor constitutes article ownership. Similarly, reverting the blanking of parts of an article with GA status does not constitute ownership. While not technically vandalism, your edits, (shown here and repeated here) were disruptive to both the reader's experience and the process of coming to an agreement in the talk page.

Most importantly, my characterization of your behavior at diasystem as OWNy stems from both your article editing behavior and your talk page incivility, the latter of which you have not addressed except to point out that you can be nice if you want to be. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:05, 8 March 2012 (UTC)

Redlink, could be potential for possible article

1. Common core (linguistics)

-- Cirt (talk) 02:36, 21 May 2011 (UTC)

This would be worth at best a stub. It is a trivial concept, only a collection. Not that it's fallacious, but it's trivial. German doesn't have the 'th' sound of English, and English doesn't have the 'ch' sound of German. So these two consonants are outside the common core for German and English. It's just a term of convenience from ca. 1955, and it isn't even used. The "common core" of multiple phoneme inventories was not conceived as being a functional entity in and of itself, nor as being a distinctive substructure. Dale Chock (talk) 03:06, 11 March 2012 (UTC)

I agree. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 03:11, 11 March 2012 (UTC)

Call for comprehensive revision

temporary placeholder to prevent bot archiving 19:06, 5 May 2012 (UTC)

This article is in need of rewriting for to achieve accuracy and remove bloat. The work of essentially one person since summer 2010 (not counting "housekeeping" edits and copy edits), it is replete with misinformation about linguistic theory. (Its author, who also wrote Diasystem alone between summer 2010 and January 2011, does not even realize the difference between structuralism and generative grammar: in February 2012 at Talk:Diasystem, he explicitly described Trager and Smith (1951) An outline of English structure as being written from a generative grammar perspective—this AFTER he was advised that generative grammar didn't exist at that time).

This editor's username requires cutting and pasting if you want to transcribe it with minimum effort. I use the simplification, AEsos.

Here are some of the problems with this article.

1. It is written in accordance with the false belief that the 'diaphoneme' concept, which is largely a relic in linguistic theory, is universally used. He attributes its use to the generativists Newton (1972) and Wells (1982), for example. The very premise of the article is a violation of WP:NPOV. (Wells rejects the diaphonemic approach, although with exceeding tact. He introduces the concept only to immediately declare it problematic and covertly abandon it. AEsos, with his long quotations from that same section, seems not to have noticed.)
2. The opening sentence is: "In linguistics, a diaphoneme or diaphone is a phoneme viewed through its dialectal variants, called diaphonic variants or diaphonic allophones." Already at "is a phoneme" this sentence is fundamentally wrong.
   1. This editor does not sense the incompatibility of this definition with the source he cites: "Kaufman (2006:64) defines a diaphon(em)e as 'a phonological category that relates the statistically significant sound correspondences among phonemes across the dialects of a single language'." Aside from the inaccuracy, let us note the strange phrasing (not an isolated example), "viewed through".
   2. "Diaphonic allophones"! This theoretically invalid phrase is the editor's invention. (I Googled it and the only other hits are from sites that plagiarize Wikipedia.)
3. Usually, points are made after being stripped of historical and theoretical context.
4. He seems to confuse language standardization with a panlectal theoretical linguistic structure. This pertains to his "examples" of Arabic orthography and Kurath's dialect atlas. (I extend this expression "language standardization" to the ephemeral, nonofficial uses as in a dialect atlas or a linguistic research work)
5. Without justifying himself, AEsos equates Bickerton's polylectal framework with the concept of diaphoneme (see heading, "Cognitive reality"). AEsos falsely credits Trudgill (1974) with adopting the notion of "polylectal" ("The description of a cognitively real polylectal grammar came with Trudgill (1974)'s set of rules for the speech of Norwich"). Indeed, on p. 120, Trudgill rejects what he supposes would be Bickerton's multisystem solution to the problem at hand. (The word "polylectal" doesn't occur in Trudgill (1974).)
6. He belabors how people can understand conversation they hear in a different accent without acknowledging what authors (like Chambers and Trudgill) have acknowledged, the simple explanation of the redundancy of language. I.e., there may not be any phenomenon to research and explain.
7. In his eagerness to find diaphoneme belief under every bush, AEsos sometimes ascribes diaphonemicity or diasystemicity apparently without foundation.
   1. "James Sledd[33] put forth his own diaphonemic system that accommodated Southern American English". But when you check the PDF's of Sledd (1966), the words 'diaphoneme' and 'diasystem' (and their derivatives) do not appear.
   2. Across the two articles Diaphoneme and Diasystem, the editor simultaneously cites Weinreich as having deemed Trager and Smith (1951) as a half baked attempt at a diaphonemic/diasystemic analysis, and insists on affirming that it is diaphonemic. But it does not seem that Trager and Smith themselves used the "diaphonemic", because the term does not occur in the book review in Language 1955, 31.2: 312ff by James Sledd. (T&S (1951) appeared three years before Weinreich coined the term 'diasystem'.)
8. He claims that Labov's linguistic variable is "a similar concept" to the diaphoneme. (I deleted the associated paragraph, noting both inaccuracy and being off topic. See previous section above.)
9. His research did not uncover that Weinreich's concept of diasystem was debunked by research by structuralist linguists within a decade of its launch. (I discussed this development in detail in the new Diasystem.)

The article was rated a Good Article in May 2011, which compounds the travesty of it.

This editor has even adopted an improper citation practice. Say he wants to cite Chomsky X without reading work X. He has come across the mention of Chomsky X in Smith Y. Instead of writing "Chomsky said W<footnote: Smith Y>, he will write: "Chomsky said W<footnote:Smith Y, citing Chomsky>, then he will include both Smith Y and Chomsky X in the bibliography. The bibliography should only include works actually consulted.

As for how AEsos discusses criticism of his editing, consider this incident. He had inserted the same piece of information (which is accurate) into two articles: this one and Diasystem (although worded differently), about how one linguist criticized the work of another. At Talk:Diasystem, when I mentioned the presence of this information in the article, he adamantly protested it was not present in the article! (This was before I was aware he himself was the editor who inserted it.) Once I proved the point, he aggravated the weird behavior by not offering a, "Sorry, my bad", by not any acknowledgement at all. Dale Chock (talk) 14:42, 3 March 2012 (UTC)

Most of these are pretty easy to fix, I think. I have numbered both Dale's original points and continued the numbering scheme. Dale, I ask that you please continue this numbering scheme so that I and others are clear on the flow of each point's development. If you do not do so, I may be compelled to restructure the conversation so that each point is a separate section.

1. I think the very long criticism section goes to show the concept's reduced saliency. If the article doesn't make it clear enough in the criticism section that Wells rejected the approach, that seems like a fairly easy fix.
2. I would say the best fix regarding "diaphonic allophone" is removing it. Here are some definitions of diaphone(me) that I've found. We can take these and come to a reasonable definition:
   1. Robert Hall's (1977)review of several works: "an alternation of sounds between dialects in the same lexeme or group of lexemes, as in /æ~a~a/ in words like ask, grass, dance."
   2. Palmer (1931): "two or more phones [abstract speech-sounds] which serve jointly as units of meaning within the limits of a linguistic community."
   3. Collins & Mees (1999): "the range of sounds heard as realisations of a particular phoneme across language varieties, for example between different regional dialects."
   4. Trask (1996): "An abstract phonological unit posited for a language as a whole, represented in each of several distinct varieties of that language as a phoneme with the same distribution but with a significantly different phonetic realization: diaphones are sometimes represented by symbols enclosed in vertical bars."
   5. Crystal (2011): "an abstract phonological unit set up to identify an equivalence between the sound system of different dialects. e.g. the diaphone /ei/, as in English mate, is realized as [ei], [ai] etc."
   6. ELL (2006): "An abstract phonological unit that relates equivalent realizations of a phoneme in two or more dialects or varieties."
   7. Merriam-Webster online: a category or a member of a category consisting of the entire range of dialectal variants of an allophone
   8. Random House "[A] group of sounds comprising all the phonetically different dialectal variants of a given phoneme in a language"
   9. World English (see above link): "the set of all realizations of a given phoneme in a language"
3. Can you cite some examples of this and show how the theoretical context may be important?
4. The Arabic example is related, since orthographies can be diaphonemic by encoding for more than one variety. It's uncited anyway, though, so I don't have a problem with removing it for that reason. Kurath and McDavid's transcription system utilizes the IPA diaphonemically, which is important in establishing that the IPA can be used to represent diaphonemes. The transcription system of PEAS was also foundational in establishing Kurath's response to the system outlined in Trager & Smith (1951).
5. The article parallels usage I have found in the literature (such as Wells, 1982) where the terms polylectal grammar, diaphone(me), and diasystem are used interchangeably to refer to the same phenomena. Because Trudgill (1974) presents a diasystem that encodes multiple varieties, it fits the definition of polylectal grammar as presented in this article and in Chambers & Trudgill (1998). As for his rejection of Bickerton, Trudgill only accepts polylectal grammars that are part of speakers' linguistic competence. Thus, his rejection is not of polylectal grammars per se but ones that are not cognitively real. I should point out that Trudgill seems to have rejected his own Norwich diasystem in later works, particularly Trudgill (1983).
6. The article addresses this point, even citing Trudgill (1983). Did you miss this or is it not clear?
7. There actually is foundation. Allen (1977) explicitly characterizes Sledd (1966) as diaphonemic on page 226. In regards to T&S, McDavid's 1952 review of T&S says that they "describe all English dialects simultaneously in terms of a single over-all pattern" (p. 387). If that's not clear enough, he goes on to say that "when [a linguist] makes an analysis involving more than one dialect he should label this diaphonemic" (p. 389). That seems clear enough to me.
8. Let's keep the conversation regarding the linguistic variable in one place
9. While things are a bit contentious between us, I am glad you were able to help out in expanding the diasystem article. As I hope is clear, I only edited that article enough to fix the article's coverage. I did not continue adding to it after this correction in June-July of 2010, so I did not notice when, for example, Weinreich (1954) became available online a few months later. I'm not sure what relevance this has on this article though.

Finally, let me explain my unconventional citation practice: let's say I'm reading X, an article on a topic I am doing research for. X briefly summarizes Y, a work I do not have access to. Y makes a claim that I would like to include in the article. However, I think it to be poor practice to attribute claims to Y if I have not directly read it. I therefore include the claim, with the citation mentioning my indirect access as "Y, cited in X" or "X, citing Y." If, at a later date, I or someone else gains access to Y, they can corroborate that the statement is accurate and remove mention of X. I think it's fair practice to include both works though, because I did not directly consult Y, I can see it being fair to move it down to "further reading."

There is a parallel issue where, although I was not able to access Y, X is making a statement about Y in context that accessing Y cannot verify. For example the footnote that reads " Wells (1982:69) citing Jones (1962), credits Jones with pushing the concept." Accessing Jones (1962) in this instance wouldn't corroborate that Jones was instrumental in pushing the concept. At the same time, it's nice to point to the foundational works for interested readers.

I can see how this practice can be problematic, though I've tried to be transparent about it for readers and other editors. Perhaps there's something I can do differently; I'm open to suggestions. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:57, 3 March 2012 (UTC)

Addendum. I have since combed through the citations and made the distinction between a) citations of sources I was only able to access indirectly and which can be modified by editors able to corroborate that source Y did say what source X said they said, marked by "Y, cited in X" and b) citations involving sources I may or may not be able to access directly but are being talked about or pointed to, marked by "X, citing Y." This should clear things up a little as we come to an agreement of what's appropriate. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:44, 4 March 2012 (UTC)

In reply to the reply numbered "2". AEsos listed nine definitions for "diaphoneme". Sources which are lay works are not academically valid for our purposes. The dictionaries for linguists (including the glossary in Encycl. of Ling. and Lgs.) are not too informative because they do not set their answer in the context of the history of linguistic theory. These definitions, which in each case are the most theoretical possible ones, are merely what some linguists in some phase of the history of linguistic theory wanted "diaphoneme" to be. If we compare diasystem, the point that was unappreciated prior at Wikipedia prior to February 2012 is that Weinreich's pretension for "diasystem" was shot down through research, and it would seem the entire field agreed on that. Now, it's true generativists took up the quest, sometimes using "diasystem". 40-50 years later, there may still be generativists who think that generative phonology achieves diasystem structures (under another label), but others disagree. "Diasystem" and "diaphoneme" are now empty of their original pretensions, as far as I can document. A "diasystem" is now nothing more than a dialect continuum, while a "diaphoneme" is nothing more than a set of corresponding phonemes in several dialects. The point to learn is that there is no longer any interest in building a common structure to such interdialectal correspondences. To take the cited example of Hall (1977), linguists no longer posit an abstract entity, linguists no longer posit the existence of a "master phoneme" or "parent phoneme", a diaphoneme (which means "interphoneme", "cross phoneme") that generates all the items in the interdialectal correspondence, claiming for the abstract entity ("linguist's construct!") either cognitively reality or grammatical usefulness. A diathing has reduced to just a correspondence set. Those bare definitions being cited out of works from 1996 or 2006 are in effect footnotes to history. Dale Chock (talk) 06:02, 8 March 2012 (UTC)

Are you proposing the definition of diaphoneme that we put in the article to be "a set of corresponding phonemes in several dialects"? I'm not a fan of this definition. I think it suffers from some of the problems you've laid out.

Anyway, from what you've said, we can eliminate the lay dictionaries (7, 8, 9); 1 and 2 are trapped in historical context, so they're out, too; and 3 is giving a contextual definition (specifically, Jones's). The remaining three, (4, 5, 6) share a lot in common in their characterization, particularly that a diaphoneme as an "abstract phonological unit." I notice that there's variation between whether what's being compared are phonemes or phones and we might want to be neutral to that effect. The first two are more in line with a characterization of the diaphoneme as an analytical tool or as something proposed rather than definitively accepted by linguists, which I know is one of your concerns.

From this, it seems that a good synthesis of the three definitions is "an abstract phonological unit that identifies a correspondence between related sounds of two or more varieties." What do you think? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:54, 8 March 2012 (UTC)

AEsos is putting words in someone's mouth. I didn't "lay out problems with" any definition of 'diaphoneme'. AEsos, by referring to an argument i didn't make, is making believe he has an argument: "i second what Dale said". But Dale didn't say it. Dale Chock (talk) 20:11, 9 March 2012 (UTC)

Since you didn't lay out problems with definitions of the diaphoneme, did you consider that I was referring to something else when I said "some of the problems you've laid out"? I think you're smart enough to figure out what I meant. Please address the substance of my post. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 23:11, 9 March 2012 (UTC)

Regarding no. "5". It is the usual combination of disregard for what the theoreticians considered important (see a heading added a few hours ago: "Linguists, their terminologies, ..."); WP:OR; incomprehension of the sources; misquoting; and refusal to identify sources. Trudgill (1974) explicitly stated he is dealing with a SINGLE variety, and i inserted that in the related Diasystem, so not only is AEsos reading the source carelessly, he's inattentive to the information being added to the other article. Also regarding Trudgill (1974), also at Diasystem, i pointed out that Trudgill discloses he is using "diasystem" in a way which one who has paid attention to the preceding part of the article can see is inconsistent with Weinreich (1954). After all, Weinreich (1954) had been debunked by the early 1960s (as detailed in that article). Wells (1982) of course rejects the use of "diaphoneme"--i've explained that here. As has been proven on this page, all AEsos's assessments of linguistic literature are uninformed and misinformed. For the rest, he merely asserts he's right without offering sources--as he consistently does. Dale Chock (talk) 13:42, 8 March 2012 (UTC)

In the chapter in question, Trudgill sets out to account for variation within Norwich English. He shows quite explicitly that Norwich English exhibits variation manifested as both phonetic differences between corresponding phonemes and differences in the phonological oppositions that speakers make. So I'm not sure what exactly Trudgill means by "variety" on page 135, but if he doesn't see varieties within this "variety" it is because he has chosen not to mark discrete (and arbitrary) boundaries between a larger phenomenon of variation. In other words, the Norwich Diasystem is an attempt to create a more abstract grammar, specific to that particular community, that accounts for variation. Pointing to a place where Trudgill called it a "single variety" does not undermine this overall strategy that makes the article's current characterization of his study appropriate.

If your problem with referring to Trudgill's Norwich Diasystem as a "polylectal grammar" is also that he did not call it that, I think this is an unnecessary standard. As I have said, the distinctions between diasystem, polylectal grammar, and overall system are both minute and variable in actual usage so that they are referentially equivalent. The article sets up what "polylectal grammar" means within itself, which I think is a fair approach.

You keep repeating that Wells rejected the diaphoneme, as though I am denying it. I already know this and have included this information in the article (as we are discussing above), so I'm not sure what your point is in repeating it. You aren't suggesting that any mention of Wells in the context of the diaphoneme must always make it explicitly clear that he rejected it, are you? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:54, 8 March 2012 (UTC)

Linguists, their terminologies, and major Wikipedia editing policies

temporary placeholder to prevent bot archiving 19:06, 5 May 2012 (UTC)

I see the need to make a general point, in view of several remarks above that point in a similar direction, and draw attention to deviations from Wikipedia policies.

Scholars invent terminology in order to distinguish their proposals from competing proposals. Oftentimes this is valid because there are really different entities of study, different strategies, and different predictions, etc. It is our duty to make clear the divergences in opinion between individual linguists and schools of theory. We are constrained by Wikipedia policy to report how linguists saw things, and not to declare insights we think they missed.

An example of the wrong attitude is this statement that editor AEsos wrote above:

As Kazazis (1976), cited in the article, says [regarding Newton (1972)]:

• N makes no hard-to-swallow claims concerning the psychological reality of his posited underlying overall-pattern (diasystemic) forms, or of the rules which derive the modern phonetic representations of those forms.

While it's true that diaphoneme doesn't show up (nor do diasystem or overall-pattern), it only takes a bit of common sense to look at what Newton has presented, combine it with Kazazis's characterization of it as diasystemic and our given definition(s) of diaphoneme to see that we are presented with something diaphonemic.

This original research depends on--here i go for the 11th time--the writer's incomprehension of linguistic teaching (here, a misunderstanding of the definitions of diaphoneme). We also go against our duties if we pluck stray quotes that support our point of view (WP:NPOV). Instead, we must judge how representative a quote is of professional opinion.

The citation of Kazazis to defend the idea that generativists believed in the diaphoneme is a case of being insensitive to the need to acknowledge theoretical differences. What differences am i referring to? The phrase "over-all pattern" was the special terminology of Trager and Smith in at least two publications (both from the 1950s). Possibly they used it in the 1940s too. "Diasystem" was a neologism by Weinreich in 1954. Editor AEsos himself consulted Stockwell (1959) for the related article, Diasystem. Stockwell elaborated on the differences between these two approaches to a shared hope of described multiple dialects with a single grammar. Certainly, the generative school of phonology (1960 onwards) shared either the identical hope or a hope that was nearly identical. However, what any Wikipedian needs to acknowledge is that beyond that shared goal, Trager, Weinreich, and the generativists parted ways.

As for the school of generative grammar, the diaphoneme concept pre-1960--at least as most general linguists, like Weinreich, understood it--is incompatible with generative grammar. Diaphoneme is an example of what generativists call taxonomic linguistics. In generative theory of that era, although there still is the belief you can "explain multiple dialects at once"--which sounds like "diaphoneme" to some people--generative theory REJECTS diaphonemes, diamorphemes, etc. While generativists still pursue an all embracing total structure, a "dialect over dialects", as it were, in principle they reject a "phoneme of phonemes" (= diaphoneme). They accept only "one taxon": phonemes.

For this article, AEsos relied on Wells vol. 1 (1982), gladly finding use of the term "diaphoneme". He didn't notice that Wells uses the term only two or three times in the entire volume, within the space of two pages. He didn't notice the more frequent use of taxonomic. He didn't notice that Wells finds fault with "taxonomic linguistics" and with the "diaphoneme", and discards them for generative theory.

Now you can see that when the reviewer, Kazazis, said "overall-pattern (diasystem)", he was banging together terminology from linguists who didn't see eye to eye (amply attested in Stockwell (1959) in portions which AEsos studiously refrained from citing in Diasystem). This passage from Kazazis is typical of the stray, indeterminate quotes that AEsos favors. One needs to be alert and sensitive to interpretations other than the plainest one. Like, maybe Kazazis was being sarcastic, knowing how Weinreich loathed the Trager-Smith analysis. (Note: by 1976, Weinreich was already long dead.) Maybe Kazazis, like AEsos, didn't know better and thought the perspectives were equivalent.

AEsos has insisted, in multiple remarks on this page, that Troike (1970), Reed (1971) used "diaphoneme". Obviously, this could have been due to inertia. But AEsos is missing a larger point: what did they mean by it? If they were in line with generative theory, then it didn't mean to them what it had meant to Weinreich and Stockwell in the 1950s. Then our task would be to explain multiple--incompatible--uses of "diaphoneme". Dale Chock (talk) 09:22, 8 March 2012 (UTC)

That's an interesting question. I have access to Troike but not Reed, so I can begin this investigation.

To be clear, Troike and Reed did not just use "diaphoneme"; they used the diaphoneme. That is, it's more than an issue of terminology; they used the concept as a fundamental basis of their analysis. So when you say that the diaphoneme is incompatible with generative grammar, these two examples contradict you.

The crux of your argument here is that, since structuralists came up with the concept and generativists abandoned it several decades later, the latter must have done so because there was too much structuralism inherent in the concept. That is, rather than the abandonment being because of the issue of cognitive reality (as the article presents), you are arguing ex novo, with no logical justification, and in spite of sourcing given to you, that it is because the concept is incompatible with generative linguistics. I hope you can understand my resistance to this interpretation. Do you have sources to back it up? — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 17:52, 8 March 2012 (UTC)

Castilian Spanish

It's strange that someone reverts an edition while stating that it is no doubt true. We have an issue here of slavish copy of a source which must obviously be wrong (I can't access the text), as the author apparently thinks that in all of the territory of Castile (historical region) there is no s-aspiration, something that is well-known to be false. We could put something like the northern Iberian dialects that the author calls "Castilian Spanish" at most. Or else, delete the whole thing, as the source proves to be unreliable. Jotamar (talk) 13:28, 16 May 2012 (UTC)

Our article Castilian Spanish says there are multiple meanings for the term. It's my understanding that "Castilian Spanish" can be used to mean Standard European Spanish, which is likely what the author in question meant. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 19:21, 16 May 2012 (UTC)

Well, in fact Standard European Spanish is almost as ambiguous as Castilian Spanish. I see it's of no use discussing this with you. Jotamar (talk) 15:34, 17 May 2012 (UTC)

My edit summary was with the understanding that you were attempting to expand the referents from just the standard dialect to those of northern Spain. I see now that you were attempting to restrict the referent, as you see "Castilian Spanish" being the group of Spanish varieties spoken in Castile.

Since the example is used to illustrate how two phonetically divergent varieties can be analyzed as having the same underlying structure, I don't feel like the point is undermined if the example is a bit oversimplistic. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 23:27, 17 May 2012 (UTC)

Theoretical and stylistic problems with the current state of the article

The article blends in Bickerton and Bailey's research without out discussion to justify doing so, assuming that it's relevant to the topic. That is, that their interest in dialectal variation somehow shares concepts with the work of Jones and Weinreich.

The article is not written like an encyclopedia article, but like a survey of research appropriate to a professional journal.

The article departs from WP:NPOV. Editor AEsos is obsessed with writing about diaphonemes and diasystems. He will not accept that they are relic terms in today's linguistics.

Aesos persists in an academically inappropriate citation practice pretending to cite original sources through other researchers which cite them. He apparently either thinks (1) the original researchers deserve the mention (2) the original researcher lends more authority. The practice is crappy. "Wells, citing Weinreich". Although it OK to stop there (to acknowledge that Wells got an idea from Weinreich), it is unsavory to actually provide a footnote and bibliographic entry for Weinreich, unless you consulted Weinreich. But in that case, it would be no use citing Wells. I have nicely advised AEsos three or five times that this is not have scholarship is done. Dale Chock (talk) 01:02, 4 May 2012 (UTC)

Welcome back, Dale. Since the above post is simply-restating what you've already stated, I've restored the relevant discussion from the archive. Feel free to continue the discussions from two months ago by responding to the three posts above that warrant response (dates [23:11, 9 March 2012; 16:54, 8 March 2012; and 17:52, 8 March 2012] marked in green for your convenience). — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 11:53, 4 May 2012 (UTC)

It's been two weeks and Dale has not furthered the discussion. His activity elsewhere prompts me to believe that he is not interested in elaborating on how this article violates WP:NPOV. So far, Dale has expressed the view that the article violates NPOV because it goes into great detail over an outdated concept and, he says, it misrepresents the context of scholarship used. I quote:

• "The very premise of the article is a violation of WP:NPOV. (Wells rejects the diaphonemic approach, although with exceeding tact. He introduces the concept only to immediately declare it problematic and covertly abandon it. AEsos, with his long quotations from that same section, seems not to have noticed.)"
• "Usually, points are made after being stripped of historical and theoretical context."

Since the article already covers criticism of the concept at length, Dale's critique of misrepresentation made little sense to me. So I asked for clarification/elaboration:

• "If the article doesn't make it clear enough in the criticism section that Wells rejected the approach, that seems like a fairly easy fix."
• "Can you cite some examples of [misrepresenting context of scholarship] and show how the theoretical context may be important?"

While Dale continued to comment, the issue of neutrality has not gone further as Dale has more-or-less repeated these points without actually responding to my rebuttals, even saying strange things like, "Editor AEsos is obsessed with writing about diaphonemes and diasystems. He will not accept that they are relic terms in today's linguistics."

This last quote, in fact, is the driving point that Dale most recently wished to express above in regards to article neutrality. I believe that it is a gross misunderstanding of Wikipedia policy to believe that WP:NPOV is violated when a topic is presented as interesting and relevant or when other editors believe it is so. To butcher a quote: Diaphonemes are interesting, and if you don't agree you can go somewhere else. If no one disagrees and Dale remains silent, I'll be removing the NPOV tag soon. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 02:23, 20 May 2012 (UTC)