Talk:Suppletion

Latest comment: 2 years ago by Hattes in topic Swedish

Swedish edit

I came to this page to understand why Swedish is the only language I know of that doesn't have an irregular inflection of "good". Instead it has both "god, godare, godast" and "bra, bättre, bäst". However, the article seems to say it isn't so. Not being native Swedish I don't wan to edit, but surely "god, bättre, bäst" is not correct/common Swedish?

It isn't, and neither is "ond, värre, värst". This table, as it pertains to Swedish, is archaic at best. If edited, I would want it to include "dålig, sämre, sämst" which are the most "natural" words for "bad, worse, worst", but those of course would have different etymologies. Hattes (talk) 21:23, 22 March 2022 (UTC)Reply

[Untitled] edit

I could use some information on the origins of the various forms of aller/ir/andare from a Latin scholar. -- user:Montrealais

I have supplied them as well as I am able, at least for the French forms -- IHCOYC 20:42 Mar 20, 2003 (UTC)
There is no Latin verb andare. Spanish andar < Latin ambulare. 4pq1injbok 23:32, 3 August 2005 (UTC)Reply

Latin verb ferre, to bring, carry has the principal parts fero, ferre, tuli, latum, meaning Is this irregular verb a result of suppletion in a precursor language? Any Latin scholars care to comment?

fero, ferre, tuli, latum. I just dropped by to note that that salient example from Latin was missing, but I see you have already pointed that out.
In 5 years of Latin study, we were never offered any explanation for fero, but, after a lapse of 35 years, it's suppletion!
I too would like to know what its roots are. 99.238.134.143 (talk) 19:19, 10 August 2012 (UTC)Reply

Domesticated Land Animals edit

Most domesticated land animals and their meat have a particularly rich set of non-cognate terms depending on age, sex, and castration. For instance: sheep, ram, ewe, wether, lamb, mutton. Do these numerous terms come under the category of suppletion? If not, where do they go? I am keen to see a table of such words, and create one if it doesn't exist, as it is a particularly striking feature of the English language. Is this phenomenon seen in other languages?

Suppletion relates to inflection, which is a grammatical feature. Animal names fall under Lexical gender rather than Grammatical gender. ("Gender" in the wide sense of "type", not just sex.) That said, English does have less lexical gender (as well as less grammatical gender) than most other Indo-European languages. Regarding the list, Wikipedia is not a dictionary but have fun with the Animals list in Wiktionary. jnestorius(talk) 14:17, 25 May 2006 (UTC)Reply

Suppletion in English edit

"He has been to France" is not a suppletion but a standard use of the verb "be" with an adjective phrase. Movement is implied by the dative quality of "to France" as in "He got to the final level in record-breaking time" or "I was already to Chapter 12 when he started reading." "Go" is not a required element.

Weak suppletion edit

I have removed the following section:

Morphemes themselves can also be suppletive. In English, the plural suffix -en which occurs in oxen is used in lieu of the more regular suffix -(e)s. In Seri, the infinitive prefix has two suppletive forms: ica- with intransitive verbs and iha- (plus change in the vowel under certain conditions) with transitive verbs. Also in Seri, the passive prefix has two suppletive forms: p- (plus change in the vowel under certain conditions) before vowel-initial roots, and ah- elsewhere. The imperative prefix in Seri has more than one suppletive allomorph; omitting some details, these suppletive allomorphs include c- in negative imperatives, c- when followed by a root that begins with a short low vowel, null plus a vowel change before an intransitive verb that begins with a non-low vowel, and h- elsewhere. The verb meaning 'arrive' has two suppletive stems: -afp when the subject is singular and -azcam when the subject is plural. The noun for 'thing' has two suppletive forms: ziix means 'thing' and xiica is the plural. Seri person is cmiique and the plural is comcaac.

I don't think "suppletive" is being used in the same way here. (I may be misinterpreting the above paragraph: in which case, by all means reinstate it, but try to make it clearer: I don't see how it qualifies as "suppletion" in the sense used in the article.) "Suppletion" can be defined:

  • Bloch, Bernard (1942). Outline of Linguistic Analysis. Linguistic Society of America at the Waverly Press. p. 48. Suppletion may be regarded as an extreme kind of internal change, in which the entire base—not merely a part of it—is replaced by another {{cite book}}: Unknown parameter |coauthors= ignored (|author= suggested) (help)
  • Joseph E. Emonds (1985). A Unified Theory of Syntactic Categories. pp. pg.170 §4.5. ISBN 9067650927. {{cite book}}: |pages= has extra text (help)

In traditional terms, the following types of highly irregular items in the lexical categories are said to be "suppletive."

Verbs: go/went; be/are; Latin ferre 'to bring' / tuli 'brought' / latus 'brought'
Nouns: person/people
Adjectives: good/better/well

Ordinary irregular variants such as stand/stood; catch/caught; woman/women are not generally considered suppletive.

  • more broadly to mean anything that language students would call "irregular". Used in synchronic theoretical morphology. As in:
  • Veselinova, Ljuba N. (2005). "79: Suppletion According to Tense and Aspect". In Martin Haspelmath (ed.). The World Atlas of Language Structures. OUP. p. 322. ISBN 0199255911. Suppletion is defined as the phenomenon whereby regular semantic relations are encoded by unpredictable formal patterns. Cases where the paradigmatically related forms share some phonological material are examples of weak suppletion, as in English buy versus bought, while cases with no shared phonological material are instances of strong suppletion, as in English go versus went. {{cite book}}: More than one of |pages= and |page= specified (help)
  • "Piapoco and Natural Morphology Theory" (PDF). University of Wisconsin-Madison. pp. pg.44, footnote 19. Retrieved 2007-10-09. {{cite web}}: |first= missing |last= (help); |pages= has extra text (help); Missing pipe in: |first= (help)

The Natural Scale of Morphotactic Transparency illustrates the hierarchy in which morphological and phonological rules cause some morphemes to be more transparent than others. The scale is: Total Transparency > Resyllabification > Phonological Rules > Morphological Rules > Weak Suppletion > Strong Suppletion.

Someone with access to the following might fill this in better:

  • Dressler, Wolfgang U (1985). "On the predictiveness of Natural Morphology". Journal of linguistics. 21: pp.321-338. {{cite journal}}: |pages= has extra text (help)
  • Dressler, Wolgang U (1986). "Explanation in Natural Morphology, illustrated with comparative and agent-noun formation". Linguistics. 24 (4): pp.519-548. {{cite journal}}: |pages= has extra text (help)
  • Mel'čuk, Igor (2006). "Chapter 8: Suppletion". In David Beck (ed.). Aspects of the Theory of Morphology. de Gruyter. ISBN 3110177110.

Probably the article should distinguish between the broad/weak and narrow/strong senses, but I think it should concentrate on the latter. jnestorius(talk) 21:03, 9 October 2007 (UTC)Reply

Table of Comparatives/Superlatives edit

The column labeled "Comparative/Superlative" contains mixed information, so that it is not clear (to one, like me, who is unfamiliar with these languages) whether the words in the rows for Romance languages are comparatives or superlatives. I suspect someone expanded an existing table without amending the label or the structure. Perhaps this column should be broken into two, yes? fuper (talk) 15:54, 20 January 2008 (UTC)Reply

No, the same word is both comparative and superlative depending on the construction. The multiple words are for the multiple languages. I'm unsure how this might be conveyed more clearly. --jnestorius(talk) 20:04, 20 January 2008 (UTC)Reply

Persian forms should be here too: khub, behtar, behtarin. — Preceding unsigned comment added by 85.65.94.223 (talk) 17:31, 6 February 2012 (UTC)Reply

Duine/daoine edit

Well, despite what I said about duine/daoine over at Talk:Weak suppletion, I'm commenting it out here because I can't find sources that confirm that the version I heard (that "duine" comes from *ghdhm-yo- from *ghdhem- "earth", while "daoine" comes from *dhewenyo- from *dhew- "die") is in fact the standard view. I believe it is the standard view, but in the sources available to me at the moment, all I can find is (1) people deriving "duine" from *ghdhm-yo- with no mention of daoine, (2) people deriving "duine" from *ghdhm-yo- and daoine from *ghdhoim-o- (cf. Latin hūmānus < *(dh)ghoim-), and (3) people deriving duine from zero grade *du-n-yo- and daoine from e-grade *deu-enyo-. But I can't find a source that comes and and says directly what I learned in graduate school, name that for all their surface similarity, the two forms are etymologically unrelated. —Angr 18:11, 2 July 2008 (UTC)Reply

Irregularity and Suppletion edit

Isn't "irregular" simply a pedagogical term and not a linguistic one? We say French etre, aller are "irregular" because they do not follow normal conjugation patterns. But aren't etre-suis suppletive? According to the defintion given under this heading one could never predict suis from etre, or "am" from "be" for that matter--making these traditionally suppletive forms "irregular." I would recommend instead something along the lines of

"The term 'irregular' is used in language learning to suggest that a given form or paradigm does not follow a usual pattern or rule, such as the conjugation of "to be" in English or "aller" in French. The term 'suppletion' is a linguistic notion that refers to pairs of words that have the same basic meaning but are phonetically distinct--such as "be" ~ "am." Whether or not the historical origins of the words in such pairs is critical to the notion of suppletion is a subject of current inquiry in linguistics."3.262ly (talk) 18:29, 3 July 2008 (UTC)3.262lyReply

Why on earth is "irregular" not a linguistic category? I've taught in a university language department for 15 years and we use the term as a perfectly standard scholarly way of describing language. It's true that the concept of regularity is particularly relevant for language learning - but also for psycholinguistics: in our native language we process regular and irregular verbs differently in our brains - or at least, that is one current theory.
Suppletion is a subset of irregularity. I think the paragraph expresses that quite clearly. So être:suis is irregular, and is also suppletive, wherease vouloir:veux is irregular but non-suppletive.
My main complaint about your suggested re-writing is its focus on "phonetically distinct": the definition of suppletion has nothing to do with phonetics, though the minority use described at the bottom of the page contains aspects of phonology. But at the top of the page we are concerned with the traditional definition, which is a matter of philology. These are three different things. --Doric Loon (talk) 19:24, 3 July 2008 (UTC)Reply
The first thing we need is an article Irregularity (linguistics). Then we can replace the weak suppletion section with a hatnote and maybe a see-also to that. While it's pretty clear that the same word is being used in 2 distinct senses, we really need a citable source that states this explicitly. jnestorius(talk) 20:07, 3 July 2008 (UTC)Reply
Incidentally, être-suis isn't etymologically suppletive since both come from *h1es-. —Angr 20:37, 3 July 2008 (UTC)Reply
Ouch, you're right! --Doric Loon (talk) 12:03, 4 July 2008 (UTC)Reply
Do normal speakers of French know that etre and suis are related? At least from a synchronic point of view speakers probably just learn the two forms separately as they would a true suppletive pair such as bon - meilleur. I guess my main point is that the term "irregular" (not given a specific linguistic sense in the OED, by the way--though it does give a more general "grammatical" meaning) is more of an amorphous catch-all for any kind of non-standard linguistic behavior: a verb doesn't conjugate right, a whole paradigm isn't one of the normal ones, suppletion. I guess that's why I was arguing earlier for the term weak suppletion (etre-suis, vouloir-veux) since "irregular" seems to mean just about anything including suppletion.3.262ly (talk) 21:07, 3 July 2008 (UTC)3.262lyReply
I think Doric Loon's point is that "suppletion" is not normally a term used in synchronic linguistics, only in historical linguistics. What French speakers know is that être is an irregular verb; what historical linguists know is that it's not suppletive. —Angr 21:10, 3 July 2008 (UTC)Reply
Ok, I'm convinced that the article is a good statement about the traditional view of suppletion. And, since weak suppletion is not seen by a majority of linguists as valid, they prefer to use the term "irregular," then it might be a good idea to get rid of the section on weak suppletion completely. If it stays, then I would suggest at least deleting the two sentences "However since the focus is entirely synchronic, it is possible that a paradigm which is suppletive in the traditional sense may be classed as weak suppletion if there happen to be sounds common to the whole paradigm. This would be the case with person:people, since they have a /p/ in common." The whole point of weak suppletion is not that there happen to be sounds in common, but that one variant was derived from another--the coincidence of sounds is totally fortuitous. These two sentences represent no linguistic thought that I know of. 3.262ly (talk) 22:28, 3 July 2008 (UTC)3.s62lyReply
Well, that's the question, isn't it? I haven't read all the literature on weak suppletion, but what I have seen suggests that for some of the people using this term in the context of a purely synchronic discussion, the focus is on whether the speaker "feels" the unity of the two parts, irrespective of whether they really are related. If that is right, the point about person:people being "weak" because of the shared /p/ should be kept, and possibly expanded by the observation that être:suis would be "strong" despite being cognate. However we don't have any really reputable sources for this strong/weak distinction yet anyway, so I may have misunderstood it. When synchronic linguists take our terminology and redefine it, the confusion is often hairraising - I so wish they could just invent their own! --Doric Loon (talk) 12:16, 4 July 2008 (UTC)Reply
(After trouble parsing the word "hairraising" without a hyphen; I thought it was a misspelling of "harassing" first) Well, since historical linguistics is basically the history of linguistics, it's not too surprising. Back in the days of SPE, underlying representations often amounted to reconstructed historical forms, and historical sound changes were restated in the form of synchronic phonological rules. —Angr 12:23, 4 July 2008 (UTC)Reply
Yes, let's not go back to those days. What's more, Doric makes a good point about synchronists redefining diachronic terms. However, there are quite a few reputable sources that are now doing just that. See, for example, the decent annotated bibliography online at http://www.surrey.ac.uk/LIS/SMG/Suppletion_BIB/WebBibliography.htm . It annotates 37 recent articles and books on suppletion, 17 of which define suppletion as not having anything to do with etymology. Instead, they see suppletion as alternating forms with coincident meaning but which cannot be mapped onto each other by using any synchronic phonological rules. In short, nearly half the linguists in this bibliography accept some form of weak suppletion (they don't all call it that). That being the case, does it make sense to update the subsection on Weak Suppletion? Possibly something along the lines of, "Some linguists do not consider a shared etymology as disallowing a suppletive relationship. For them, words that share meaning but cannot be mapped onto each other by means of synchronic phonological rules are 'weakly' suppletive. For example, French etre-suis, though historically from the same root, are weakly suppletive because there are no phonological processes that relate the two. This is not the traditional view of suppletion. Instead forms such as etre-suis are seen simply as irregular, not suppletive."3.262ly (talk) 22:51, 7 July 2008 (UTC)3.262lyReply

By all means update the section. But I don't like the wording "Some linguists do not consider a shared etymology as disallowing a suppletive relationship." That sounds as though there was an alternative view of the same concept. The point is that these linguists are using the same term for a fundamentally different type of concept. Make that clear. What is still not clear to me, though, is how these synchronic people divide strong and weak. If weak means everything which diachronic linguists would exclude, but the new approach would include, then it won't be hard to find a modus vivendi without too much confusion. But if weak means everything which shares phonological elements (i.e. including person:people but excluding suis:être) then we have to be very careful to keep separate discussions of diachronic and synchronic studies. --Doric Loon (talk) 23:00, 7 July 2008 (UTC)Reply

I'm no expert on any of this, but my main concern is that we find a reliable external source that explicitly states the variance in definition/interpretation of the term "suppletion"; rather than inferring it ourselves, which violates WP:SYN. Perhaps a source for such would be the intro to a synchronist's text, giving a history of the concept. jnestorius(talk) 08:42, 8 July 2008 (UTC)Reply
"Suppletion is a relation between signs X and Y such that the semantic difference between them is maximally regular while the phonological difference is maximally irregular." Igor Mel'chuk. 1994. “Suppletion: toward a logical analysis of the concept”. Studies in Language 18:2. 339–410. Hippeseley et al. define semantically regular alternants as having the same meaning and phonologically irregular alternants as those that cannot be related by ay possible phonological rule. They add, "We take the synchronic view of suppletion: stems which meet the definition are treated as suppletive irrespective of their etymology." Andrew Hippisley, Marina Chumakina, Greville G. Corbett and Dunstan Brown "Frequency, categories and distribution of stems" in Studies in Language, 2004:28,2 pp. 387-409. And David Fegert (Suppletion, natural morphology, and diagrammaticity" goes as far as to say, "Most contemporary morphologists have come to use the term suppletion to refer generally to any morphological relationship that is "maximally irregular"...This definition is purely synchronic and does not include the criterion that suppletive stems be etymologically unrelated." In Linguistics: an interdisciplinary journal of the language sciences 36.6, 1998. p1065. 3.262ly (talk) 20:09, 8 July 2008 (UTC)3.262lyReply

Those are very useful quotes which should certainly go into the article. They still don't answer my question, however, and it is crucial here, since it decides whether the diachronic and synchronic uses are complimentary perspectives on the same issue, or mutually exclusive uses of terminology which cannot both appear in the same discussion without great confusion. The shape of our article will depend on that. I suspect the latter, and if so, I imagine the way we will have to go is to divide the article into two main sections, the first on philology, the second on synchronic morphology. --Doric Loon (talk) 20:58, 8 July 2008 (UTC)Reply

I think you are exactly right. Within the context of philology/historical linguistics suppletion is strictly go-went. In synchronic morphology it includes both go-went and three-third. I like your suggestion of two main sections for the article. 3.262ly (talk) 22:14, 8 July 2008 (UTC)3.262lyReply
What about keeping the first part as it is, and replacing the section "Weak Suppletion" with the following:
The term “suppletion” is also used in contemporary synchronic morphology in regard to sets of stems (or affixes) whose alternations cannot be accounted for by current phonological rules. For example, the stem-final consonant in the pair oblige - obligate exhibits the alternation [ʒ]~[g]. These stems are related by meaning but are not related by any synchronic phonological process. This makes them similar to stems that are traditionally called suppletive (see above). The chief difference is that in true or full suppletion, the stems are not etymologically related. In morphological suppletion, sometimes called “weak” suppletion, however, the stems do have a common etymological source.
“Most contemporary morphologists have come to use the term suppletion to refer generally to any morphological relationship that is “maximally irregular...This definition is purely synchronic and does not include the criterion that suppletive stems be etymologically unrelated.” David Fegert, “Suppletion, natural morphology, and diagrammaticity” in Lingustics: an interdisciplinary journal of the language sciences 36.6, 1998 p. 1065.
Current usage of the term “suppletion, suppletive” in synchronic morphology, however, is not fixed.
Since I'm new to wikipedia talk, I'm hoping for some response before trying to change something that's already pretty good. 3.262ly (talk) 20:01, 10 July 2008 (UTC)3.262lyReply

Good start, reads well. I like the last sentence - that's part of the point - we can't pin this down, though we have more actual sources for synchronic than for dichronic use, but the diachronic use is so fixed that there is no controversy. I think you have understood "weak" suppletion differently from me, but you might well be right. Better try to source that precisely. What I would then like to see, as a short third section, is something on how the two interrelate, whether they are just different perspectives or whether they really clash. But if no-one has published on that, it might be OR, in which case we just have to live with the uncertainty. --Doric Loon (talk) 21:27, 10 July 2008 (UTC)Reply

Russian examples edit

These pronunciation guides appear to be a confusing combination of IPA and intuitive English transliteration... /'luchshij/nai'luchshij/ /'hudshij/nai'hudshij/

I've edited them a bit so that they conform better to IPA but I'm not really an expert in this domain. Kasnie (talk) 03:56, 13 July 2008 (UTC)Reply

Latin sources for Romance "go" edit

This section of Examples has been bodged, and will need to be fixed. Five sources are numbered within the table, but only four are listed underneath. In the table French "allai" is given as "3 or 4", and Spanish "fui" as "5", but in the list #4 is Latin "fui", which clearly applies to Spanish "fui" not "allai", and there is no list #5. It looks like there have been two different models of sources, and one has been changed to the other but not completely - probably from five to four, the older #4 being removed and #5 moved up.

If "allai" does not come from "ambulare" it may come from the same source as Italian "andare" according to Wiktionary - namely a combination of "vadere" and "aditare". Altho where the N comes from is still a mystery. But I would have thought the French forms with their L's are much more certain derivations from "ambulare" than Spanish "andar" or Italian "andare/andrò/andai etc.". There is no accounting for the ND there.

But it might help to find some linguistics sources making it all clear? 86.0.169.202 (talk) 12:19, 19 October 2008 (UTC)Reply

The Online Etymology Dictionary says that Italian andare and Spanish andar are from Vulgar Latin ambitare, which I suppose is a frequentative of ambire. Phonologically, it's plausible as ambitare > ambidare > ambdare > andare. I don't know where Wiktionary gets its etymology of andare. I've just changed Wiktionary's etymologies of both andare and andar to reflect the Online Etym. Dict.'s etymology. It may not be right, but at least it's sourced. —Angr 16:55, 19 October 2008 (UTC)Reply

Note 4 of the table confuses the Latin "perfect" with English "perfective" constructions, which also exist in the languages referred to in this note. For instance "Have you been to France?" would be "Ha estat a França?" in Catalan and ¿Ha estado en Francia?" which are periphrastic constructions based on reflexes of a conjugated form of the Latin verb "habere" and the perfect participle of "stare" and therefore would bear no relation to "esse" at all. It also seems to state that English has no simple present, which is simply false, and this would also seem to imply that the preterite is a present tense form, which it is not. An example from Spanish: "Yo fui a Francia" translates into English as "I went to France" (which also carries the "perfect" implication, i.e., on vacation and now I'm back). The table also makes no mention of the "imperfect past" in these languages which is sometimes based on a suppletive stem from a different Latin verb (for instance Catalan "anava, anaves . . .," also see discussion of the Italian "andare" and Spanish "andar" above), and sometimes not (for instance Spanish "iba, ibas . ." which are direct reflexes of the Latin imperfect of "ire"). Spanish "Yo iba a Francia," could be translated as "I used to go to France." (i.e., habitually, for business reasons). I suggest removing the comparison to English in Note 4 completely, including the imperfect past in the table, and the inclusion of a note pointing out the semantic differences between the preterite and imperfect. Skip Purdy (talk) 19:41, 6 November 2015 (UTC)Reply

Person—People edit

I don't believe the English convention of using people as the plural of person is an example of suppletion in the sense used in this article, since each of those words has the full, regular set of inflectional forms: person, person’s, persons, persons’, people, people’s, peoples, peoples’. Rather, this is an instance of a usage convention and unrelated to inflectional morphology except as it may provide an example of how suppletion occurs. Grosbach (talk) 23:52, 18 September 2012 (UTC)Reply

That doesn’t matter. Many people say one person, two people, three people, which clearly shows suppletion. Two persons is literary. Peoples is a plural meaning “ethnic groups”. — TAKASUGI Shinji (talk) 13:46, 9 June 2019 (UTC)Reply