Talk:Free variation

Doublets

How does free variation compare to doublets? Should it be mentioned in the article? 77.85.7.45 (talk) 12:58, 4 January 2016 (UTC)

Regional Accents

The article says: "...free variation is exceedingly common and along with differing intonation patterns is the most important single feature in the characterising of regional accents."

This may be true, but I have never heard it. At any rate, it needs a citation. (And since I have never heard it, I'm afraid I can't provide a citation. Mcswell 20:11, 24 August 2007 (UTC)

I don't have the kind of book handy which would give a citation, but I can assure you it is true. Let's suppose an Englishman, a Scot, an American and an Australian all say the same single-syllable word. That is hardly enough to have identifiable intonation patterns, but you can already tell the accents because the vowels are placed differently. That's free variation. --Doric Loon 21:47, 24 August 2007 (UTC)

Well... except it isn't really free variation. It's variation conditioned by dialect. Free variation, I thought, is when the two sounds in question can vary freely within a given dialect, or even idiolect. —Angr/talk 21:52, 24 August 2007 (UTC)

Interesting. As soon as I get to my office I can certainly give you a citation for "free variation" being defined as both variation within an idiolect and variation between speakers including between dialects; i.e. all allophonic variation which is not complimentary distribution. I believe the the author is Schulz, and the book is called something like Englische Phonetik und Phonologie, but I will find it. I must admit, it always bothered me that the same term was being used for what are obviously quite different things, so possibly my source is not a good one. (If I remember correctly, he only uses free variation when talking about allophones, though other people use it for phonemic variation too.) --Doric Loon 08:42, 25 August 2007 (UTC)

A random question mark?

"more stupid or stupider (?)"
What is that question mark doing there?--Adûnâi (talk) 18:08, 2 January 2018 (UTC)

"only a very small proportion of English words show such variations"

Aside from Free Variation being a bright red flag in itself in that variation is seldom truly free, the claim seems to be that British and American phonological differentiations are realized in "only a very small portion of English words." On the face of it, that's spectacularly counterfactual. I suspect that what was meant is buried in "such variations", and that the intention was that what would appear to be phonemically distinct forms of "the same word" are infrequent. Anyone care to repair this? Barefoot through the chollas (talk) 17:30, 28 May 2020 (UTC)

I'm not sure what the problem is. It seems to me to be pretty clear what is meant, and I think you have also understood it quite correctly. Whereas the free variation of allophones is extremely common, with every single English phoneme existing in multiple forms, the free variation of phonemes only occurs in a very small number of words, such as "tomato" or "either". I'm sure we can make a list of words where native English speakers differ in the pronunciation, not because they pronounce a particular phoneme differently but because they use a different phoneme, and if we stick to common English words, the list would be short enough to be manageable. Of course there would be far more if we include rare words, technical words, literary words, recent loan words, etc, where people are not sure how they should say them.

I don't understand "free" variation to mean that the speakers are not conditioned by their upbringings or groups - then there would be little or nothing in language that is "free". It simply means free of the phonetic environment, i.e. the allophones are not in a rigid complimentary distribution. --Doric Loon (talk) 17:47, 30 May 2020 (UTC)

"I think you have also understood it quite correctly" I guessed correctly as a result of having been a practicing academic linguist for decades. It may be that the root of my needing to guess in the first place and your "I'm not sure what the problem is" lies in a clash of subfields of linguistics, escalated by the often cryptic nature of the text in the article. Re the first, in the world of what might be described as "Language variation and change" (diachronic linguistics, dialectology, sociolinguistics, and numerous blends and offshoots in those galaxies of the linguistics universe) it's not a huge stretch to fear than an assertion such as "In the case of allophones, however, free variation is exceedingly common" tout court could set a dissertation committee's hair on fire and the poor candidate back to remedial coursework. But the candidate and this article can be saved by fleshing out and clarifying what's intended, which is the second point: clarification and exemplification for both "civilians" and linguists-in-the-making. This, for example, which might seem straightforward and clear at first glance:

free variation is the phenomenon of two (or more) sounds or forms appearing in the same environment without a change in meaning and without being considered incorrect by native speakers.

My native speech is relatively easily identifiable as U.S. South Midland (or Lower Midwestern, Mellencamp phonology, for familiar reference). If I were to come out with (/təˈmɑːtoʊ/ while amidst local native speakers in any context other than mimicking BrE, that would be incorrect by just about any criterion, other than that (or even though) I did manage to communicate the meaning 'pulpy edible fruit...' It's not all that difficult to tease out the distinction between variants producing no semantic disturbance and the same variants being identifiable as and/or functions of varieties of different lects.

It seems that part of the free variation problem stems from the term being used in two distinct ways, one of which abstracts out the real-life sociolinguistic (geolinguistic, etc.) conditioning which is always present in real language use (the traditional variante libre definition of French linguistics?), the other of which is fully informed by it. IMO that doesn't mean that a coherent article on free variation isn't possible, though; it means that once constructed, it's more informative and instructive. Barefoot through the chollas (talk) 20:44, 1 June 2020 (UTC)

That is really very interesting. I think this article began as an off-shoot from the article on allophones, which distinguished two kinds of allophones - those in complimentary distribution and those which are "free" of that kind of distribution rule. Free variation in that context means nothing more or less than that the variation cannot be pinned down to a phonological rule, but is rather a matter of speaker (or group) habits or preferences. (That group habits may be so entrenched that an individual speaker is not really "free" is beside the point; the word "free" means so many things that we can't use it sensibly unless we insist that it only means one thing at a time.) Now if it is true that the term "free variation" is used in linguistics in another way too, then the article does need an overhaul. But I would be out of my depth there, so I suggest, if you are sure of yourself and have sources, you should go right ahead and do it. --Doric Loon (talk) 17:45, 2 June 2020 (UTC)

You've struck at what may be the heart of the glitch with this: the word "free" means so many things that we can't use it sensibly unless we insist that it only means one thing at a time, the problem being that it doesn't mean just one thing in the usage free variation without further clarification. What's intended in the article is something like 'free of phonological conditioning'. That can run into some minor difficulties ([n] in a careful-speech rending of input vs. assimilatory [m]), but the issue is greater in tossing together the two very distinct categories of alternative phonemic structures (e.g. tomato, economics, either, route, data, Colorado) and variable allophony (realizations of the /p/ of stop) -- allophony is irrelevant to selecting one of the phonemic doublets, whereas position in the phonological utterance determines the possible (or at least likely) phonetic realizations of /p/ in natural speech (Stop! permitting some variants, Stop it! and Stop talking! each determining others). I'll give some thought to re-casting the article. Some things I don't know what to do with, since I'm reluctant to simply delete something that's there presumably for good reason, but needs further elaboration to makes sense to readers. I can't flesh out this puzzler, for example, without knowing what it's intended to mean:

English's deep orthography and the language's wide variety of accents often cause confusion, even for native speakers, on how written words should be pronounced. That allows for a significant degree of free variation to occur in English.

It may mean that the spelling of something like the English toponym Warwick can lead to a spelling pronunciation with two instances of [w] for those who have never heard it; if so, an example or two would have clarified the statement. Other matters are easier to deal with, such as free variation being actually infrequent: https://www.thoughtco.com/free-variation-phonetics-1690780. Non sequiturs can be re-stated and converted to illustrative examples, so that along with differing intonation patterns, variation in allophony is the most important single feature in the characterization of regional accents can lead into the various parochial realizations of the /t/ of water, etc. The bit about rhotics under English examples is probably incomprehensible to most readers, perhaps best deleted or replaced with more familiar rhotic vs. non-rhotic English and linking R. And all examples should present examples, i.e. illustrate. I've just noticed that there's a very good page labled Variation (linguistics), with no link from here to that or from that to here -- and no mention of free variation there. Barefoot through the chollas (talk) 18:45, 6 June 2020 (UTC)

A very frequent phenomenon in Greek

I don't know if you think we should add any more languages than English, but Greek includes this feature very often. Indeed, as the consonant clusters mp, nt and wk (μπ, ντ, γκ) slowly changed in Late Ancient Greek, they have now two alternative forms, either b (μπ), d (ντ) and g (γκ) or mb (μπ), nd (ντ), and ng (γκ). Although the former choice is preferred for foreign loan words. LightningLighting (talk) 00:13, 16 March 2021 (UTC)

Misnomer

The Kager quote added by User:Barefoot through the chollas is misleading to the reader. Kager is quoted as saying that the term "free variation" is justified because "no grammatical principles govern the distribution of variants"... but Kager seems to be misinformed here. Grammatical principles do govern the distribution of variants; they merely do so in a probabilistic way rather than a deterministic way. Using that quote alongside the sociolinguists' objection to the term leaves the reader with the false impression that the sociolinguists' objection doesn't include the fact that "free" variation is grammatically constrained. AJD (talk) 23:49, 8 September 2021 (UTC).

Sticky wicket, that's for sure. These are Kager's examples of "no grammatical principles govern the distribution of variants":

a. sentim[en]tality ~ sentim[n̩]tality

phonology (vowel reduction)

b. I know that John will leave ~ I know John will leave

syntax (complementizer)

If you have good examples in widely-known languages for exemplification of grammatical (i.e. system-internal) determination of (thus pseudo-)free variation, presenting them and the principles governing their use would be excellent for the body of the article, which is in pretty sad shape. A sort of ranking from genuine free variation (flapping or aspirating the /t/ in AmE see ya tomorrow may well be an example) to variants more constrained in canonical ways easy for readers to grasp would provide excellent illustration. (As for your last statement, I'm not clear on your meaning -- and I don't see where that interpretation is implied -- but I think it would take a lot of searching to find a sociolinguist who would claim that variation can't be grammatically constrained.) Barefoot through the chollas (talk) 01:11, 9 September 2021 (UTC)

I wouldn't say it's misleading as much as I don't see what the relevance is. The quote is pretty much a repetition of what the sociolinguistic objection is to and doesn't provide any rebuttal.

As for "misnomer", I find it could be better worded because "free variation" is already an established term that refers to a very specific situation. Whether that thing is truly what the parts that make up the term imply or whether it was coined based on an accurate assumption doesn't particularly amount to an argument, as much as one may indeed find it misleading and advocate for avoidance or an alternative. It just sounds like a prescriptive objection to an idiom on the grounds that it's not logical. Nardog (talk) 02:06, 9 September 2021 (UTC)

I'm reminded of those practice in-house miniconferences that good mentors hold for their Ph.D. students about to present a paper at a real conference for the first time. Now and then a student who hasn't done his background reading or didn't take the variation discussion seriously in class will describe x and y as being "in free variation". During the following discussion, one of the attending faculty will almost certainly say something to the effect of "You'll want to be careful with the notion of free variation. Make sure that the variation isn't actually conditioned by something you haven't noticed." Not infrequently the student will be led to realize that the variation of x and y is, indeed, not free, but describable coherently in linguistic or sociolinguistic terms. Whether the upcoming conference is an NWAV or an ICHL or a hard-core GLOW, the student has been done a big favor. -- Free variation does exist. But it's not common. The real-world fact is that the label free variation has been a red flag for decades, it's (usually) a misnomer and very often quite misleading because it misrepresents what it's labeling, and thereby the nature of human language. Most linguists who deal with variation regularly are very aware of that, and proceed accordingly. The result is not trivial, and expecting accuracy in use of the term is no more prescriptivist than expecting the label motorcycle to be used for motorcycles. (end of screed) -- I'm not enamored of the Kager quotation, but it's succinct, and makes the distinction fairly clear (he's explaining, not rebutting). If someone has a better one of similar purpose, I'd be happy to see it. Barefoot through the chollas (talk) 03:47, 9 September 2021 (UTC)

Ah, I now see the word was used in rather disparate ways in Ajd's version and yours. Ajd meant "there's no such thing as free variation, period", whereas you meant "it does exit, but most of the phenomena that get the label aren't actually what the term purports it to be", hence "very often" (i.e. not invariably). Thus calling it "a misnomer" in the former is an objection to terminology, while in the latter it's an objection to certain applications of the term. Either way I don't think the word is helpful (it's just too vague), but the solution would depend on the resolution of the other dispute (the one between you two), so I'd like to see it settled first. Nardog (talk) 06:25, 9 September 2021 (UTC)

Regarding one of the Kager examples, Tagliamonte & Smith (2005) found that complementizer deletion is favored by (1) first-person subject of matrix clause, (2) lack of elements such as negation or modal auxiliaries in matrix clause, (3) pronominal subject of embedded clause, (4) absence of adverbial elements between matrix verb and embedded clause, and (5) present tense in matrix clause. These are all grammatical factors that probabilistically condition complementizer deletion. Kager's claim that "no grammatical principles govern the distribution of variants" is simply false; this is a perfect example of why sociolinguists consider the term "free variation" to be misleading. AJD (talk) 06:45, 9 September 2021 (UTC)

Exactly. And not just sociolinguists. Krager goes on to sort of recognize the difficulties of FV, but his goal is to deal with apparent FV within OT, so he accepts FV for the argument. But Krager's not the point for the article here; his or another similar quote is merely an attempt to salvage the concept of FV enough to warrant the discussion. In fact, skipping over a number of basic details that need to be ironed out in the lead and further on, the somewhat rare status of genuine free variation calls into question the existence of this article. As it stands, it's mostly Amateur Hour. It's not clear why, but while people with limited knowledge of Neurotoxins or Condensed Matter Physics tend not to mess with those, it's not uncommon for articles on language to be subject to ongoing rookie editing. This article differs from most language articles I've seen in that very little of that has been cleaned up. I don't have time to do it myself, but I'm willing to help out. The lead needs minor work, but the rest needs to be re-done, starting with organization, such as category Phonology subsuming what is now Phonology and Pronunciation (strange at best that those are presented as separate), with examples from English and a few other languages. Then perhaps on to morphology... sorting out along the way questions such as are dreamed/dreamt loaned/lent morphological or lexical, and so on. Barefoot through the chollas (talk) 16:12, 9 September 2021 (UTC)