Wiktionary talk:Votes/pl-2018-04/Unifying on Inflection heading

Rationale
Making Wiktionary unified in this regard will make it slightly easier for readers and editors. Both practices are acceptable: 1) use Inflection heading, 2) use Declension or Conjugation depending on part of speech. The first practice reduces the number of heading types with no loss of information: once you know the part of speech, you also know whether the inflection is called "declension" or "conjugation".

Some languages use Inflection, other use Declension and Conjugation headings. I saw Latin use Inflection, e.g. in. Perhaps someone will help collect statistics. That said, even if a majority of languages currently uses Declension and Conjugation, I find Inflection preferable. --Dan Polansky (talk) 05:54, 28 April 2018 (UTC) BTW: Maybe the section headers could be more informative or have a clarificatory into: "Support unifying on Declension and Conjugation as far as possible = Oppose unifying on Inflection as single header"; "Oppose unifying on Declension and Conjugation as far as possible = Support unifying on Inflection as single header"? -84.161.32.86 14:18, 30 April 2018 (UTC)
 * I can understand the desire for consistency, but I don't feel a compulsion to mandate consistency, especially since I personally tend to use all three: Declension for nouns and adjectives, Conjugation for verbs, and Inflection for prepositions in languages like Irish that inflect their prepositions. —Mahāgaja (formerly Angr) · talk 06:13, 28 April 2018 (UTC)
 * What are the benefits for the readers and editors of your current practice? How is a new editor going to discover the differentiation that you apply, and why should they care? --Dan Polansky (talk) 06:21, 28 April 2018 (UTC)
 * The benefits for readers and editors is that more specific terms are used when they're appropriate. A new editor can also use the more specific terms if he wants, or he can use the more general term everywhere if he wants. —Mahāgaja (formerly Angr) · talk 06:29, 28 April 2018 (UTC)
 * So should the use be inconsistent even within individual languages? Should some Latin verbs carry heading Inflection while other Latin verbs Conjugation? --Dan Polansky (talk) 06:41, 28 April 2018 (UTC)
 * I wouldn't say that should be the case, but it's not intolerable if it turns out to be the case, either. Another advantage to using Conjugation and Declension is that most of our inflection-table templates are named "Template:xy-conj-..." and "Template:xy-decl-...". If we get rid of Conjugation and Declension headers, the reason for those names will be obscured. —Mahāgaja (formerly Angr) · talk 14:35, 28 April 2018 (UTC)
 * The vote makes it possible to move away from tolerable inconsistent state to preferable consistent state, then. --Dan Polansky (talk) 07:19, 29 April 2018 (UTC)
 * When students learn a language they usually use a traditional set of grammar words to describe that language. We'd be throwing all of that out. DTLHS (talk) 06:31, 28 April 2018 (UTC)
 * They also know the word "inflection", don't they? Are the readers of current Latin pages really impoverished, then? --Dan Polansky (talk) 06:41, 28 April 2018 (UTC)
 * Generally no. "Conjugation" and "declension" are commonly used by language learners. "Inflection" is used more restricted to linguists. --WikiTiki89 17:51, 30 April 2018 (UTC)
 * @Wikitiki89: I don't know how to verify that statement. In any case, yields inflection as the winner. That does not prove anything, but is suggestive. --Dan Polansky (talk) 05:58, 1 May 2018 (UTC)
 * "Inflection" also has meanings outside of linguistics that are more common and known to the general public. I think it's a mistake in our entry for to list the linguistic definition first. --WikiTiki89 11:07, 1 May 2018 (UTC)
 * @Wikitiki89: A discussion on that with Stelio is below on the page, in another thread which contains more Ngram reports, including one on "inflected forms". --Dan Polansky (talk) 11:10, 1 May 2018 (UTC)
 * As for "once you know the part of speech, you also know whether the inflection is called "declension" or "conjugation".":  That's not correct. One can know it's a verb but still doesn't have to know how the 'changing' is called. One can even find usages like "decline the verb [verb]", so the person knows it's a verb but still doesn't correctly call it "conjugate". (I'd guess declining verbs is more common than conjugating nouns, but dunno.)
 * Regarding voting for two proposals: I believe it is allowable to, for example, support "Decl&Conj" and conditionally support "Inflection" if "Decl&Conj" fails. But note also the eligibility requirements to vote at WT:VP. -Stelio (talk) 16:04, 30 April 2018 (UTC)
 * You may happen not to know the word "conjugation", but the only thing you will miss is a name, and a distinction with almost no difference. As Richard Feynman's father used to say, you can learn the names of a bird in over 20 languages and you still have not learned a thing about the bird. If the broader name "inflection" does all we need, let's use it, and make Wiktionary user friendly. On the other hand, the collapsible headers under the heading show the declension and conjugation terminology anyway. --Dan Polansky (talk) 06:03, 1 May 2018 (UTC)


 * Request for Clarification:
 * Is this vote intended to apply to all languages? Or just Latin?  I'm a bit confused about the intended scope of this after reading through the thread of the posts above.  ‑‑ Eiríkr Útlendi │Tala við mig 21:39, 28 April 2018 (UTC)
 * It is to apply to all languages. From the vote: "Thus, for instance, switch all Latin verb inflection headings to Conjugation"; italics on "for instance" mine. --Dan Polansky (talk) 07:08, 29 April 2018 (UTC)

Inflection vs. Conjugation and Declension

 * Original title: Terminology by Language

This looks like a sensible suggestion to me: giving a consistent presentation across individual entries, particularly for section headers, is exactly what ELE is there for. I'm in favour of using the more specific terms (declension and conjugation) as being appropriate and also in line with templates, but I wonder whether there is any variation by language? If a particular set of terms is used commonly for an individual language's grammar, then I support the use of that localised terminology being used. I don't know of any examples of this though; are there any, in anyone's experience? -Stelio (talk) 08:31, 30 April 2018 (UTC)
 * @Stelio: It is not clear to me why that is appropriate. A column in a database table or a spreadsheet should ideally have one hypernymic name even if it can contain differentiated content. I believe the terminological differentiation on the heading level does not add any true information to the human users and bots. The reader learns that the thing is called declension or conjugation anyway, in the collapsible header, to be seen e.g. in Latin, which has Inflection heading. --Dan Polansky (talk) 05:55, 1 May 2018 (UTC)
 * It looks like you're responding to my statement that, "I'm in favour of using the more specific terms (declension and conjugation)..." rather than my question (in this section) on whether terminology varies by language. Do correct me if I'm wrong. So on that point:
 * Declension and conjugation are correct terms, and an appropriate level of specificity. You say, "A column in a database table or a spreadsheet should ideally have one hypernymic name even if it can contain differentiated content." As a facetious example, we could replace all of the various "Noun", "Verb", etc. headers with "Part of Speech" and it would be correct, and we could then describe the part of speech in the content of a section; but in practice we use the more specific term as appropriate for a header. A less facetious example is that we have "Proper Noun" as a section header separately from "Noun", using a term correctly and with greater specificity even though "Noun" on its own would also be correct. Personally, I too believe a column in a database table or a spreadsheet should be given a name that describes its complete content; but I also believe that a column of data should contain content of the same type, and that mixed data types should be split into separate columns as appropriate. However, I also think that this analogy is not appropriate for what we are discussing here, which is entry layout of an article (a very different beast to columns of data).
 * Aside from declension and conjugation being both correct and more specific, I feel that inflection is less well known as a technical term. I think of declining a noun and conjugating a verb. I do not think of inflecting a noun or inflecting a verb. This is in line with WikiTiki89's comment above.
 * It seems clear to me now (having looked through various examples and read other sources) that it is sufficient to use "Conjugation" for verbs and morphemes that form verbs, and "Declension" for all other parts of speech. I haven't seen a specific example of a term for which "Inflection" is required because neither "Conjugation" nor "Declension" can be used on their own. Without that, I don't see a need for using "Inflection" as a header in any case. (Much as I wouldn't expect to see "Part of Speech" as a header for any term.)
 * Using an inflection section header with tables marked as declension and conjugation would be a solution too, and one that I could support (albeit I would prefer declension and conjugation headers). (I should note that not all current inflection tables include the word "conjugation" or "declension", since they are currently used in sections where those words are used as the headers. I guess that most templates do not have "conjugation" or "declension" headers.) However, you say below that the vote should not be made more specific, so perhaps specifying the use of inflection as a header and that the corresponding table templates should use declension or conjunction as appropriate may be beyond what you would propose for a vote. A vote for just inflection headers without the concomitant requirement for appropriate table headers lacks sufficient rigour for me.
 * -Stelio (talk) 08:17, 1 May 2018 (UTC)
 * I would like to follow up to say, I like the vote and I agree that we should unify the headings into a consistent wording. I simply disagree that "Inflection" is the most appropriate choice, believing that "Declension" and "Conjugation" are better for the reasons above. What is your rationale,, for rejecting "Declension" and "Conjugation" in preference for "Inflection"? -Stelio (talk) 08:23, 1 May 2018 (UTC)
 * @Stelio: Thank you. My rationale is that terms should not be specific unless that specificity actually discloses something important. On my job, I am asked to differentiate things not worth differentiating when filling certain forms, and that differentiation burns my brain cycles. To me, "declension" is just a name for "inflection for nouns, adjectives and certain pronouns" and "conjugation" is just a name for "inflection for verbs". And as for abandoning headings Noun, Verb, etc. in favor of something else, that would be doable, and M-W does it; in M-W:paper, their heading is for the headword itself, with a numerical index; the part of speech is in smaller font below. While I do not currently propose to do something similar, it is a possible and sensible way of doing things. (They used to have a beautiful website and now they have the sad tabletty thing, but that's a separate matter.) --Dan Polansky (talk) 08:42, 1 May 2018 (UTC)
 * As for "I feel that inflection is less well known as a technical term": suggests otherwise, while not being conclusive. --Dan Polansky (talk) 08:46, 1 May 2018 (UTC)
 * In that graph, declension is in decline, pun intended :). --Dan Polansky (talk) 08:49, 1 May 2018 (UTC)
 * Let us amicably agree to disagree, then. :-) I suspect that this graph rather more charts either the decline in the volume of texts relating to grammar, or the rise in texts in other subjects. only has grammatical senses, whereas  and  have senses in other fields as well which will skew the graph in their favour. -Stelio (talk) 09:02, 1 May 2018 (UTC)
 * You are right; the graph is inconclusive. Let's have a look at to see what sort of senses we find there. The overwhelming majority seems to be from linguistics. But that is still inconclusive. I would submit that we do not know, and that there is no good reason to believe that the "inflection" terminology would be inconvenient for the readers or that they know it less well than "declension" and "conjugation". --Dan Polansky (talk) 09:09, 1 May 2018 (UTC)
 * Perhaps these are more useful? Still not perfect (restricting them to their verb forms ("inflect") rather than noun forms ("inflection")), but perhaps more indicative of use in practice:
 * decline vs inflect for nouns
 * conjugate vs inflect for verbs
 * -Stelio (talk) 09:20, 1 May 2018 (UTC)
 * These are interesting and seem to side with you. Let me note that the syntax used by Google is "_INF", not "_CONJ". Let me also note that these search for verbs rather than the nouns "inflection" and "conjugation". I seem to understand what "_INF" does: it seems to instruct Google to consider inflected forms of the base, but I do not know what _DET_ does; do you perhaps have a link to Google's explanation? --Dan Polansky (talk) 09:28, 1 May 2018 (UTC)
 * Anyway, here is, and while it shows "conjugated" taking a lead, it does not support the notion that "inflected" is hard to understand. --Dan Polansky (talk) 09:30, 1 May 2018 (UTC)
 * And here: : inflected again takes a lead. --Dan Polansky (talk) 09:32, 1 May 2018 (UTC)
 * The plural deals a more deadly blow: . --Dan Polansky (talk) 09:35, 1 May 2018 (UTC)
 * Good work on the "form/forms" graphs. So the terms as verbs favour conjugate/decline, and the terms as adjectives favour "inflected". (I'll still stick by the more specific terms, personally.) Ngram documentation is here (look in the advanced section). The "_INF" tag allows for inflections of a word. The "_DET_" tag allows for determiners. So "conjugate_INF _DET_ verb" searches for "conjugate the verb", "conjugating a verb", "conjugated every verb", and so on. -Stelio (talk) 09:41, 1 May 2018 (UTC)
 * Thank you for the link, and thank you for working together on the discovery of relavant empirical facts :). --Dan Polansky (talk) 09:51, 1 May 2018 (UTC)
 * You're still missing major biases of the data. The word form itself in this context is much more common in linguisitics than among the general public when learning languages. When I took French in middle school and high school, the common ways we used the word "conjugate" were like these:
 * You should learn the conjugation of the verb along with the verb.
 * You forgot to conjugate the verb.
 * How do you conjugate être?
 * Here is the conjugation of être.
 * Let's practice conjugation.
 * I don't remember ever hearing the collocation "conjugated form(s)". --WikiTiki89 11:24, 1 May 2018 (UTC)
 * I still think that the data we have suggests that readers will not have any difficulty understanding the Inflection heading, even if there is a bias in the "inflected forms" search toward professional linguistics; that search showed a huge lead for "inflected forms", per . --Dan Polansky (talk) 11:34, 1 May 2018 (UTC)
 * One more Ngram: . --Dan Polansky (talk) 12:15, 1 May 2018 (UTC)
 * Let's do a case study. I just searched for "grammar of french" in Google Books, and individually opened each of the first ten books and searched it for the following forms:
 * of "inflect": inflection|inflectional|inflect|inflected|inflecting|inflects
 * and of "conjugate": conjugation|conjugate|conjugated|conjugating|conjugates
 * Out of the ten books I looked at:
 * 7 contained numerous instances of the forms of "conjugate"
 * 0 contained any of the forms of "inflect"
 * 2 contained neither (at least according to the search)
 * 1 was not searchable
 * And this is exactly what I expected, because from my experience the forms of "inflect" are used in linguistic theory and the study of morphology in general, while only the forms of "conjugate" are used in the study of a particular language (outside of comparative linguistics). And even in linguistic theory, the general study of morphology, and comparative grammars, forms of "conjugate" are still common. --WikiTiki89 19:30, 1 May 2018 (UTC)
 * On top of that, if someone who hasn't studied linguistics sees the word "inflection", they're likely to think it's talking about intonation. —Granger (talk · contribs) 19:42, 1 May 2018 (UTC)
 * @Granger: I cannot verify that statement; it may be true or not. --Dan Polansky (talk) 20:17, 1 May 2018 (UTC)
 * @Wikitiki: I cannot easily reproduce that in, but even if it is true, it does not follow that our users will have difficulties with the "Inflection" terminology; the Ngrams cannot all be ignored. Another tidbit is that Britannica online has an article on inflection, whereas for conjugation, it only has a directory page. --Dan Polansky (talk) 20:17, 1 May 2018 (UTC)
 * You spelled "French" wrong, maybe that is what was wrong with your search? --WikiTiki89 20:45, 1 May 2018 (UTC)
 * I did spell it wrong, but Google showed me the corrected search as the default, 'Showing results for "grammar of French" ...'. --Dan Polansky (talk) 20:56, 1 May 2018 (UTC)
 * Then I'm not sure why you were not able to reproduce my experiment. --WikiTiki89 21:03, 1 May 2018 (UTC)

Let us, for the sake of the argument, assume that the readers are not used to the term Inflection, and will not initially know what it means. Assuming that, once they land in Wiktionary, how long will it take them to find out? They navigate to Inflection heading, and what will they see? For one thing, they will see a collapsible element saying Declension or Conjugation. But let us assume that even the collapsible elements would be renamed to Inflection, which is not proposed in this vote; then, what will the readers see? They will see a table of inflected forms. How hard can it be to find out and get used to it? That is all under a pessimistic assumption that I find unlikely. --Dan Polansky (talk) 20:24, 1 May 2018 (UTC)
 * Even if it users can get used to it, it would still confuse one-time users. And we'd still be inferior terminology. Shouldn't we use the most appropriate terminology whenever possible? --WikiTiki89 20:45, 1 May 2018 (UTC)
 * As for one-time users, I don't know whether that is a thing. And I do not believe Inflection to be inferior; to the contrary: let's not differentiate what is not worth differentiating. I do not believe "declension" and "conjugation" to be "most appropriate"; I believe it to be a terminological distinction that, on the face of it, is there only so that there are more terms to learn. --Dan Polansky (talk) 20:56, 1 May 2018 (UTC)
 * By "inferior" and "most appropriate", I mean in terms of accepted usage. Even if it's not a meaningful distinction, we should use whatever is the accepted terminology. --WikiTiki89 21:03, 1 May 2018 (UTC)
 * No, we should not use the 'most appropriate' (Terminology contested and in need of proof.) terminology. We should use the terminology which produces the least work for editors while having no unacceptable detrimental effect for consumers. The sole argument brought forth against the proposal so far is hinging on the very contestable claim that an inflection table under a header with a term unknown to the user is not self-explanatory. Even if that were true, which I doubt, hampering copy/paste work by having to swap header names and hampering cross language work and uniformity (and thus ease of navigation) as well as being forced to basically dart toss names for languages with very loose concepts of noun and verb seems a very high price to pay for the minuscule benefit not making a group of users of unknown size have to look up an unknown word they encountered in a linguistic (lexicographic) context because they cannot make sense of the table in the first place, which realistically means they've most likely no idea what declension or conjugation is either. If they can't make sense of the table, calling it 'conjugation' won't suddenly make them. So I don't see what actual benefit is served with this pointlessly detailed naming practice. Korn &#91;kʰũːɘ̃n&#93; (talk) 14:56, 3 July 2018 (UTC)
 * No, my argument was that we should call things what they are generally called in similar works and not come up with our own terminology. I never suggested using "Terminology contested and in need of proof". --WikiTiki89 15:05, 3 July 2018 (UTC)
 * I didn't say that, I said that I and Dan are contesting the terminology you used, which was that of 'most appropriate' and 'inferior' terms. And 'we should do X' is an expression of preference, not an argument. The argument was 'people might not understand the term inflection which would confuse them as to the section'. Korn &#91;kʰũːɘ̃n&#93; (talk) 18:28, 3 July 2018 (UTC)
 * That's what I just said is not my argument. And I just told you what my argument is. --WikiTiki89 18:49, 3 July 2018 (UTC)

Terminology by Part of Speech
Mahāgaja mentions other parts of speech above, using Irish as an example. I suggest that an improvement to the vote's wording would be to make it more specific: "Use Declension headings for nouns and adjectives, Conjugation headings for verbs, and Inflection headings for other parts of speech." In practice, what other parts of speech do we have inflection sections for? (Mahāgaja has already pointed out prepositions.) -Stelio (talk) 08:31, 30 April 2018 (UTC)
 * Suffixes for sure, not sure what else. —Mahāgaja (formerly Angr) · talk 15:20, 30 April 2018 (UTC)

Following a little thought, some examples above of words with existing inflection tables for discussion. Bah. The original wording, although less precise, avoids the pitfall of onerous prescriptiveness. -Stelio (talk) 17:25, 30 April 2018 (UTC)
 * Perhaps the original vote wording ("Use Declension and Conjugation headings whenever such distinction is applicable.") is better, then, as it doesn't require a list of all relevant parts of speech?
 * Or, "Use Declension headings for nouns and noun-like parts of speech, Conjugation headings for verbs, and Inflection headings for other parts of speech"?
 * Or indeed, "Use Conjugation headings for verbs and morphemes that form verbs, and Declension headings for other parts of speech (such as nouns)"? (No "Inflection.")


 * (ec) Searching a database dump for pages that contain =Inflection= without containing =Noun=, =Proper noun=, =Adjective=, or =Verb=, I find Numerals (δύο, два), Participles (abditus), Particles (إن), Prepositions (ب, על), Postpositions (alitse, bikʼi), Circumpositions (naartoe, doorheen) and Circumfixes (in- -is), Pronouns (hij), Relatives (luhlaza, banzi), Suffixes... Adverbs (daarom), Conjunctions (أن), Determiners (deze, weinig), Interjections (ليت), Letters (Å), Phrases and Idioms and Prepositional phrases (with real tables like murtaa leipä and saada vihiä and għand bżonn, and with null content that should be a usage note like pieniin päin)... and even Prefixes, although they're iffy (keski-, sala-). Also: "Adjectival noun" (明らか) and "Verbal noun" (რეკვა). - -sche (discuss) 17:45, 30 April 2018 (UTC)
 * What an interesting collection. I've never heard of "relative" as a part of speech; it looks like a stative verb that actually means "to be [Adjective]", just like adjectives in Burmese and many other languages. The inflection table at and  probably shouldn't be there as the first one is actually just the conjugation of  and the second one doesn't tell us anything useful that the conjugation table at  doesn't already tell us. Same goes for the inflection table at ; it doesn't say anything the inflection table at  doesn't already say. —Mahāgaja (formerly Angr) · talk 19:11, 30 April 2018 (UTC)
 * If you're interested, the main discussion of Relatives was here, with briefer mention more recently here; it appears to be a recognized part of speech in the grammars of at least some Bantu languages. In Abenaki (and probably many other Native American languages), words of the sort are handled as verbs. - -sche (discuss) 20:45, 30 April 2018 (UTC)
 * I've been torn on what to do with them in Burmese, too. Existing dictionaries tend to call them adjectives, and I've often followed suit, but grammatically they aren't different from verbs, and it's really only for the convenience of English speakers that most dictionaries (including Wiktionary) call (for example) an adjective meaning "big" rather than a verb meaning "to be big". I know I haven't been consistent about this for Burmese, and I'm sure there are entries in CAT:Burmese verbs that could just as easily be in CAT:Burmese adjectives. The linguist in me wants to eliminate CAT:Burmese adjectives and call them all verbs, but I'm worried that doing so would confuse people who use Wiktionary in addition to other resources, when the other resources call everything an adjective that can be translated into English with an adjective. And while I understand the desire to use conventional labels that may be different from language to language, the linguist in me also finds it absurd to use different terminology for Zulu, Burmese, and Abenaki to describe what is pretty clearly the same part of speech. —Mahāgaja (formerly Angr) · talk 06:19, 1 May 2018 (UTC)
 * I don't know if this is true of Burmese or not, but in Zulu and Abenaki, there are a limited number of true adjectives (e.g. "great", which doesn't inflect for number the way statives like "white" do). Rua (and apparently at least some scholars of Zulu) cited that against Meta's suggestion of considering relatives to be adjectives in Zulu. And while I've only looked into Zulu briefly, from what I've seen it seems like anyone who doesn't treat Zulu relatives as a distinct class does consider them adjectives, not verbs, so it seems there might still be cross-language inconsistency if "Relatives" were retired... :/ Anyway, I suppose this is getting off on a tangent and should perhaps be moved to the BP if you want to do away with relatives. - -sche (discuss) 20:17, 1 May 2018 (UTC)


 * The Japanese term with the "Adjectival noun" POS was a holdover from a past formatting convention.  The Japanese label for it is, literally "adjectival verb".  Whatever it's called, words of this type function neither as nouns nor as verbs, but rather as adjectives when followed by the sentence-ending copula  or the attributive particle , and as adverbs when followed by the adverbial particle .  To minimize confusion among English-language readers (the target audience for the EN WT), I've been gradually turning all of these into just "Adjective", as that's what most Japanese-learner materials in English call this type of word.  ‑‑ Eiríkr Útlendi │Tala við mig 21:54, 30 April 2018 (UTC)


 * For completeness of the list, let me also mention that aside from parts of speech already listed above, the following parts of speech, while I did not find them with "Inflection" sections, do have "Declension" or "Conjugation" sections:
 * Affixes (-nnguaq), Articles (οι, των), Combining forms (-лагать), Contractions (ἁνήρ), Numbers (parisataa, which may need to be changed), Ordinal numbers (fünfte, miljardis, második), Relationals (itan, ijpak and many other Pipil words),
 * some initialisms/acronyms/etc (some of which I changed to actual parts of speech), several entries where POS headers had been forgotten (which I fixed), several valid instances of numbered POS sections (and one odd but probably valid arrangement in astú and adaigh), and aproveitar a maré.
 * - -sche (discuss) 20:57, 1 May 2018 (UTC)


 * The vote should not be made more specific, IMHO, since what it is about is whether to conflate all inflection under Inflection heading or whether to use more specific terminology whereever possible. The vote should not try to specify the rules when to apply Declension and Conjugation since that would introduce an opportunity for error. For many languages and parts of speech, these rules are a known given, and the vote should not try to restate them. --Dan Polansky (talk) 05:38, 1 May 2018 (UTC)