Talk:Weak suppletion

Weak suppletion is a contradiciton in terms. Suppletion implies that two verbs (nouns etc) supplement each other - the idea that a paradigm which is made from a single etymological root can be called suppletion seems nonsense. I do not believe that this is a real linguistic term. I will give those who wrote this two or three days to provide reliable sources to prove me wrong, and if they can't will call for rapid deletion. --Doric Loon (talk) 05:48, 6 June 2008 (UTC)

May I just add that I have done a Google search and found only a very small number of refereces, few of which appear to be authoritative sites. It looks as though one PhD thesis suggested this terminology and a few other lightweight internet sources followed him. This is not enough to have the terminology used in Wikipedia. Besides, even if we do accept this as alternative terminology, it does not deserve its own article since the concepts are fully discussed elsewhere. At the very most we need a brief note on such pages as "irregular verb" saying that this terminology is also possible; in that case this article would become a redirect. But I would resist even that unless a printed book by a recognised scholar can be found using the phrase. --Doric Loon (talk) 13:03, 6 June 2008 (UTC)

weak suppletion
You make a good point, suppletion either is or isn't. However, among the "is" suppletion there may well be a spectrum from "full suppletion" which encompasses totally (etymologically) unrelated words [go-went] to "weak suppletion" which would refer to words which are phonologically disparate but etymologically connected [child-children]. Some suppletion experts do refer to "weak suppletion", namely Andrew Hippesley, Marina Chumakina, Greville Corbette and Dunstan Brown (see particularly their article in Studies in Language (2004:vol 2, issue 2)).3.262ly (talk) 22:54, 30 June 2008 (UTC)3.262ly


 * I agree. "Weak suppletion" is not a completely unreasonable concept: dritte is not derived from drei by a transparent morphological process. In that sense it is somewhere in between a straightforwardly derived form and a fully suppletive one. That said, calling it suppletion does muddy the waters somewhat, and it's not really clear what's gained by categorising it as an odd case of suppletion rather than an odd case of derivational morphology. But that's a different question. As you indicate, some people in the field do use it, and that's enough to justify the article. However, I would recommend editing it to make clearer who originated the term, how far it's propagated, and if there are any explicit detractors. garik (talk) 10:10, 1 July 2008 (UTC)

The reason it is not a logical concept is because the word "suppletion" implies something being brought in from elsewhere - think of the meanings of words like "supply, supplement" - the metaphor is that a gap in the paradigm is filled by something "supplied" from outside it. That is true with go-went. It is in no sense true with three-third, because third always WAS the ordinal of three: they have been in the same paradigm ever since it was a regular paradigm; so no historical process of supplying-the-gap ever took place. The irregularity has a different, non-suppletive origin, which can be explained philologically. The whole point of the term "suppletion" is to make that distinction.

However that is not the point. Even if a term is absurd we can mention it if it has currency, and we can give it its own article if it covers a concept not covered elsewhere. The point is: Either way, this article ought to go.--Doric Loon (talk) 12:58, 1 July 2008 (UTC)
 * this certainly fails on the second of these criteria, because "weak suppletion" contains no other concept than "irregular paradigm" and we have other articles on that. Therefore there is no purpose in this article.
 * and it probably also fails the first criterion since linguists are not generally using it. For that reason I would be against using the term at all. But I may be wrong here, and if you can demonstrate currency, you could argue for briefly mentioning "weak suppletion" as a synonym for "irregularity" in some other article.
 * I would also like to see more sources than just http://wals.info/feature/description/79 using this term before being content that this is a well established term. In its current state, I'd suggest a redirect to Suppletion and a single sentence there saying something along the lines of "The term weak suppletion has been suggested in case where two morphologically related forms share phonological material but the alternation between them is unique to that pair of forms, e.g. English buy ~ bought, where the two forms share the initial consonant [b], but no other verb has an alternation between X[aI] in the present and X[O:t] in the past and past participle". (It occurs to me that at least one instance of synchronically weak suppletion is historically strong suppletion: the Irish word for "person, human" is duine [din'@] in the singular and daoine [di:n'@] in the plural. This would count as synchronic weak deletion since the two forms share a lot of phonological material, but historically they come from completely different roots: one of them is related to English "die" and originally meant "mortal", while the other is from the Indo-European "dhghem-" root and is related to Latin "homo". At the moment I can't remember which is which, though.) —Angr 14:53, 1 July 2008 (UTC)
 * Oh, person/people also share a [p], so they could also be considered synchronically weak suppletion but diachronically strong suppletion. Another weird one is choice/choose, with an idiosyncratic vowel alternation but a recurring pattern of voiceless fricative in a noun paired with a voiced fricative in a verb. What's particularly weird about this example is that choice is a loanword from French while choose is a native Anglo-Saxon word. (The French word itself is borrowed from Germanic, though.) —Angr 14:56, 1 July 2008 (UTC)

--The article on "suppletion" defines suppletive words in a very narrow and traditional way: words must not be cognate (=must not have a common origin) and one must be an inflected form of another, eg good-better. Igor Melchuk ["Suppletion: Toward a logical analysis of the concept”. Studies in Language, 1994, I 8:2, 339-410] suggests another way of understanding suppletion, namely those signs are suppletive that are semantically coincident but cannot be mapped onto one another using any synchronic phonological process. According to this definition, words that are etymologically related can be suppletive if through diachronic processes it is no longer possible to derive one from another using phonological rules. This definition entails moving the static function of "supplementing" to a more dynamic category. When, in the 16th C middle english thrid became third by metathesis, this set up an alternation of the stem meaning '3': thri-thir(d). As long as metathesis was productive in English we can relate the two signs. How do we relate the two signs after metathesis died out as a productive phonological process? Obviously they are related by meaning but they are no longer related by any morphological or phonological process, just like "good-better." The stem for 'three' supplements itself by itself (or at least a variant of itself).

I agree that the article on weak suppletion should be deleted but that a line in the article on suppletion contain some reference to it as a possible subtype of suppletion. I also think that the statement that suppletive forms must be cognate might better "...when the two words are not cognate, or when two cognates are phonologically disparate." or something like that.128.187.0.164 (talk) 19:29, 1 July 2008 (UTC)3.262ly


 * Just for clarity, does Melchuk explain in what way a suppletive paradigm in his sense of the word differs from an irregular paradigm in the traditional sense of the word? Because if they are the same, his redefinition gives us a loss of precision in diachronic analysis with no gain anywhere - we no longer have terminology to describe clearly what we used to call suppletion, and have two synonyms to describe irregularity. Why?  --Doric Loon (talk) 21:24, 1 July 2008 (UTC)
 * As I understand it, a suppletive paradigm is one where the alternation is unique to that paradigm (like buy~bought, where no other verb has that alternation), while a garden-variety irregular paradigm can have several members (sing~sang~sung, sink~sank~sunk, shrink~shrank~shrunk; write~wrote~written, ride~rode~ridden, drive~drove~driven, etc.) —Angr 17:21, 2 July 2008 (UTC)

OK, I've done the redirect as suggested by Angr. And I've written a short paragraph in Suppletion including the material offered by user128.187.0.164. This can probably be improved. --Doric Loon (talk) 09:36, 2 July 2008 (UTC)