Wiktionary:Votes/2020-01/Definitions of English terms should start with a capital and end with a full stop

Definitions of English terms should start with a capital and end with a full stopEdit

Voting on: Amending Wiktionary:Style_guide#Definitions to specify (or clarify) that definitions of English terms should start with a capital letter and end with a full stop. For reasons explained at the linked discussions, certain other changes have been made to the proposed wording, most notably the removal of the mention of the distinction between "full definitions" and "glosses". Mihia (talk) 23:05, 12 January 2020 (UTC)

Please vote on replacing the current text with the new text below.


Definitions should be concise. Only in rare cases should a definition consist of more than one sentence or sentence fragment.

Types of definitions

Most definitions on Wiktionary are either full definitions or glosses. Full definitions, which are preferred for English terms, explain the meaning a particular sense in detail. Glosses, which are preferred for non-English terms, simply point the user to one or more English translations of the term.

A full definition should start with a capital letter. Because a definition is not normally a complete sentence, opinions vary on whether it is necessary to end a full definition with a period. However, in the current editing practice most definitions end with a period.

A simple gloss should not be capitalized and should not end with a period.

Example of a gloss:

Katze: cat

Example of a definition:

cat: A domesticated species (Felis silvestris) of feline animal, commonly kept as a house pet.

For defining non-English words, glosses are strongly preferred. In general, a full definition should be provided only where a foreign-language term has no satisfactory English equivalent.

For English words, full definitions are strongly preferred. Even in the rare case of true synonymy, a gloss for an English term should be formatted as a definition:

# [[cat|Cat]]
A definition should [...]


Definitions should be concise. Only in rare cases should a definition consist of more than one sentence or sentence fragment.

Definitions of English terms, including definitions that are single words or lists of single words, should begin with a capital letter and end with a full stop (or, in special cases, exclamation mark or question mark). For example, at the entry for bread:

  1. (uncountable) A foodstuff made by baking dough made from cereals.
  2. (slang, US) Money.

Definitions of non-English terms are most often single-word or single-phrase translations. These are generally uncapitalized (unless proper nouns) and do not end with a period. For example, the entry for the French word chien:

  1. dog

In general, a full definition (as opposed to a translation) of a non-English term should be provided only where that term has no satisfactory English equivalent. There are presently no specific guidelines for capitalisation or stopping of full definitions of non-English terms.

A definition should [...]


  • Vote starts: 00:00, 12 January 2020 (UTC)
  • Vote ends: 23:59, 10 February 2020 (UTC)
  • Vote created: Mihia (talk) 20:45, 5 January 2020 (UTC)



  1.   Support, and modify any templates that don't correspond to this format. Andrew Sheedy (talk) 15:38, 13 January 2020 (UTC)
  2.   Support. A mixture of styles, especially within the same section, looks messy and amateurish. The other option would be to have all English definitions uncapitalised and un-fullstopped. My impression is that this would require much more change than the present proposal, and would also be problematic in the smaller number of cases where multiple sentences are required. Mihia (talk) 23:20, 13 January 2020 (UTC)
    Per the first: according to Special:Statistics there are 3.2 million "gloss definitions", of which 720 K are English. So, if we suppose that all non-English definitions are uncapitalized and unperioded, standardizing English to non-English is much fewer definitions to correct. Per the second: I do think capitalizing and fullstopping is fine when the definition is a sentence (contains a finite verb, no explicit subject needed), like I've done on om#Danish and su#Greenlandic. Out of interest, can you supply a few examples of multi-sentence definitions?__Gamren (talk) 00:32, 18 January 2020 (UTC)
    @Gamren WT:Statistics / Special:Statistics uses gloss definition differently than this vote: as it explains in a little (easy to miss) footnote, "A gloss definition is one that has definitional content. A non-gloss definition is one that does not have definitional content but rather links to a gloss definition, such as [is the case with] inflections, variants and alternative spellings." So, for that page, it's not a formatting / punctuation thing, but a "presence of a real definition vs merely a 'see main entry' pointer". Perhaps that page should use a different word. As for multi-sentence definitions: someone who had time could probably search the database dump for more examples; I suspect there might be a few "valid" examples in the definitions of complicated mathematical terminology or the like. One not-so-good example is musket, which I changed just today to use semicolons instead. - -sche (discuss) 06:16, 22 January 2020 (UTC)
    @Mihia: I cannot think of a case where a definition would require multiple sentences; do you have an example? Differentia can be piled on without introducing a new sentence. As for musket, the definition line could be "a kind of firearm formerly carried by the infantry of an army, originally fired by means of a match, or matchlock, for which several mechanical appliances (including the flintlock, and finally the percussion lock) were successively substituted, superseded by the rifle"; nonetheless, it is questionable whether all the bits of information really belong to a definition proper. Even if someone would insist they need a sentence element on the definition line, it could be in brackets or after a semicolon, like "a domestic animal that drinks milk (it also meows)" or "a domestic animal that drinks milk; it also meows". --Dan Polansky (talk) 10:35, 1 February 2020 (UTC)
    @Dan Polansky: I'm sure that there are some Wiktionary definitions that are multiple sentences, but unfortunately I cannot right now remember specific examples in order to check how necessary this is. Mihia (talk) 20:38, 6 February 2020 (UTC)
  3.   Support Equinox 23:22, 13 January 2020 (UTC)
  4.   Support DTLHS (talk) 23:33, 13 January 2020 (UTC)
  5.   Support Jberkel 23:18, 14 January 2020 (UTC)
  6.   Support Pablussky (talk) 13:42, 15 January 2020 (UTC)
  7.   Support Cnilep (talk) 07:54, 16 January 2020 (UTC)
  8.   SupportTom 144 (𒄩𒇻𒅗𒀸) 17:33, 16 January 2020 (UTC)
  9.   Support Non-sentences on English entries don't feel right for some reason. —Suzukaze-c 05:40, 17 January 2020 (UTC)
    @Suzukaze-c: English entry definitions usually are non-sentences, e.g. "domestic animal that drinks milk" as opposed to "A cat is a domestic animal that drinks milk". --Dan Polansky (talk) 10:14, 1 February 2020 (UTC)
  10.   Support. I don't care a great deal, but this imposes order on disorder in a manner that matches our usual practice. —Μετάknowledgediscuss/deeds 05:46, 17 January 2020 (UTC)
    @Metaknowledge: Would not it be better to find the preferable practice and then switch to it rather than just cement whatever we have? --Dan Polansky (talk) 10:14, 1 February 2020 (UTC)
    Yes. I don't care a great deal, so all reasonable practices are about equal to me; I value the order much more. —Μετάknowledgediscuss/deeds 21:29, 1 February 2020 (UTC)
  11.   Support — justin(r)leung (t...) | c=› } 05:47, 17 January 2020 (UTC)
  12.   Support - -sche (discuss) 05:34, 18 January 2020 (UTC)
  13.   Support This, that and the other (talk) 10:36, 19 January 2020 (UTC)
  14.   Support Ultimateria (talk) 17:43, 21 January 2020 (UTC)
  15.   Support --Droigheann (talk) 01:20, 28 January 2020 (UTC)
  16.   Support I find the proposed changes pertaining to English a vast improvement, and to me the revised parts about foreign languages are also preferable. ←₰-→ Lingo Bingo Dingo (talk) 11:09, 8 February 2020 (UTC)
    For a legalistic standpoint, a support also supports the following: "Definitions of non-English terms are most often single-word or single-phrase translations." Such translations are not the desired end state of definition lines of non-English terms. The question here is not whether the new text is better than the old text--the old text was not voted on and is on a non-policy page--but rather whether the new text is good enough to be signed via a formal vote. --Dan Polansky (talk) 12:49, 9 February 2020 (UTC)
    @Dan Polansky I don't think it's likely that someone who ruthlessly removes lexicographic information to enforce single-word or single-phrase translations will receive a lot of support. The risk seems rather theoretical to me. ←₰-→ Lingo Bingo Dingo (talk) 08:40, 10 February 2020 (UTC)
    Put differently, it is unlikely that people actually support the voted-on text as written in its entirety. --Dan Polansky (talk) 08:17, 14 February 2020 (UTC)
  17.   Support Kaixinguo~enwiktionary (talk) 20:05, 8 February 2020 (UTC)
  18.   Support for consistency's sake, more than any preference for which direction to go.--Prosfilaes (talk) 10:09, 9 February 2020 (UTC)
  19.   Support Let's rock the boat. Canonicalization (talk) 10:23, 9 February 2020 (UTC)
    @Canonicalization: May I ask whether the capitalized definitions are your preference, or whether you are supporting the better supported option to achieve voted-on consistency? Or maybe there is a third option? --Dan Polansky (talk) 12:53, 9 February 2020 (UTC)
  20.   Support. HeliosX (talk) 22:57, 9 February 2020 (UTC)
  21.   Support For consistency. Jjamesryan (talk | contribs) 09:12, 10 February 2020 (UTC)
    This is an invalid rationale in so far as both possible policies--capitalized definitions and uncapitalized definitions--would achieve consistency; the stated characteristic does not distinguish between the two candidate policies. --Dan Polansky (talk) 08:17, 14 February 2020 (UTC)
    @Dan Polansky Capitalisation would take considerably less work to achieve consistency, as it's far more prevalent. Jjamesryan (talk | contribs) 17:44, 19 February 2020 (UTC)
    @Jjamesryan: But it still would not achieve consistency across languages, only within English. So if consistency (or minimum entropy, minimum complexity of formatting rules) is the sole objective (and you stated none other objective, like effort), lowercase wins. --Dan Polansky (talk) 07:41, 22 February 2020 (UTC)
    @Dan Polansky, a preference for less work was pretty clearly implied. And in my opinion consistency across languages is far less important than consistency within languages. I don't take great issue with difference across languages. Jjamesryan (talk | contribs) 05:34, 23 February 2020 (UTC)


  1.   Oppose Allahverdi Verdizade (talk) 10:02, 13 January 2020 (UTC)
      Oppose It's not dictionary-style and most definitions are not complete sentences, making this pointless. On the other hand, the Simple English Wiktionary intentionally makes its definitions into colloquial sentences, and is always in need of contributors. פֿינצטערניש (Fintsternish), she/her (talk) 11:33, 13 January 2020 (UTC) I didn't realize the current text already asked for a capital letter, so I am going to abstain. פֿינצטערניש (Fintsternish), she/her (talk) 11:37, 13 January 2020 (UTC)
  2.   Oppose If the exact same definition were written for an English and for another language, one should not be formatted differently from the other. Identical definitions should have identical formatting. If we distinguish different styles of definition, then the formatting rules should be based on the style, not on the language. —Rua (mew) 13:58, 13 January 2020 (UTC)
    Ideally I agree. Ideally I would propose the same formatting for both. However, the de facto situation seems to disagree, and my feeling, based on some comments, is that a proposal that would entail changing e.g. "chien: dog" to "chien: Dog." probably would fail, so this is the pragmatic proposal. Remember also that there is nothing in the proposal to stop "full" definitions of non-English terms being capitalised and full-stopped if that is what people desire. Mihia (talk) 23:29, 13 January 2020 (UTC)
  3.   Oppose. This little bit of bureaucracy ignores the fact that one or two templates don't comply with it. DonnanZ (talk) 14:06, 13 January 2020 (UTC)
    Templates should also be updated to comply, but the Style Guide is not the place to specify this. Mihia (talk) 23:22, 13 January 2020 (UTC)
  4.   Oppose Starting with capital letter and ending with period is ugly and nonsensical, and I often fix it when I add senses, to make the old definitions conform in style to the new, since I won't add in that style myself. There's no grammatical justification for it, except when the definition is a sentence. Also, having to pipe all the time is tedious.
    As an aside, the current distinction between glosses and "full definitions" is not clear to me. Is "female musician", at Musikerin, a gloss or a full definition? What about "unemployment payment" for dagpenge, or "wood that is discarded" on affaldstræ, or "the sequence consisting of the partial sums (of some infinite series)" at afsnitsfølge, or "having recently completed their studies" at nyuddannet? No, really, please classify all these.__Gamren (talk) 18:31, 13 January 2020 (UTC)
    Regarding "glosses" vs. "full definitions", please see Wiktionary_talk:Votes/2020-01/Definitions_of_English_terms_should_start_with_a_capital_and_end_with_a_full_stop#Removal_of_mention_of_"glosses". Mihia (talk) 23:36, 13 January 2020 (UTC)
  5.   Oppose: This vote piles way too much in it, some I wouldn't mind, some I really do, like periods at the end of definition lines. --{{victar|talk}} 22:01, 14 January 2020 (UTC)
  6.   Oppose Does not change anything to the positive, with its spooked distinctions only feeding Wikilawyers instead of common sense (which mandates that all languages should be treated the same). If one wanted to uniformize certain templates one could do it already, accommodating current practices. Fay Freak (talk) 12:51, 15 January 2020 (UTC)
  7.   Oppose: I don't believe that definitions have to be sentences. SemperBlotto (talk) 13:49, 15 January 2020 (UTC)
    Barely relevant, since this is about formatting like a sentence, not about having to be a grammatical sentence. Equinox 15:31, 15 January 2020 (UTC)
    I am not sure. It kinda prescribes the forms for definition lines beyond punctuation and capitalization: “Only in rare cases should a definition consist of more than one sentence or sentence fragment.” (e contrario it should be a sentence or something below it?) “In general, a full definition (as opposed to a translation) of a non-English term should be provided only where that term has no satisfactory English equivalent.” (so in foreign languages the lines should not be, if ever possible? Dubious measures, I gloss and define as I consider it necessary.) Fay Freak (talk) 16:53, 15 January 2020 (UTC)
  8.   Oppose I have the same opinion about foreign language entries - it should be just a translation in lower case without a full stop, as I have always done. Glosses in translations tables should also be in lower case without full stops. --Anatoli T. (обсудить/вклад) 02:52, 21 January 2020 (UTC)
    @Atitarev Same opinion as what? Part of the vote is to enforce lower case without a full stop for FL entries. Andrew Sheedy (talk) 03:21, 21 January 2020 (UTC)
    @Andrew Sheedy: Fair enough. --Anatoli T. (обсудить/вклад) 03:26, 21 January 2020 (UTC)
  9.   Oppose I now think it preferable to standardize on no-initial-capital-letter-and-no-final-punctuation for definition lines of both English and non-English languages. This would provide unified experience across languages. The lowercase format is used in multiple English dictionaries, a per Wiktionary talk:Votes/2020-01/Definitions of English terms should start with a capital and end with a full stop#Other English dictionaries, including Merriam-Webster online. Admittedly, it would be a large switch but we have enough users who are more than happy doing such switches, and one of these switchers has been switching to the lowercase format. The question for this vote, in my view, should not be what historically we have been doing but rather what we want and plan to be doing, and why, and why should not include argument from our tradition. Later: I find myself disagreeing with my previous statement about argument from our tradition. Tradition does carry weigh: a large-scale switch should only take place if it is worth the effort and the disruption. It seems to me that the unified experience is worth it.

    On another note, the proposed new text has so many problems that I would have to oppose even if I preferred capitalized definition lines. For instance, the following is misleading: "In general, a full definition (as opposed to a translation) of a non-English term should be provided only where that term has no satisfactory English equivalent." Bare translation items without disambiguating gloss are generally suboptimal; I think every translation should be eventually followed by a gloss, that is, an abrreviated or less abbreviated definition. The new text gives "dog" as an example of a definition line, which is hardly ideal given that the word dog has multiple meanings. Rather, there should be something like "dog: a domestic animal bred by humans from wolf". --Dan Polansky (talk) 10:58, 1 February 2020 (UTC)

    Just a note on 'unified experience across languages': personally I find the fact that English definitions look like sentences and non-English translations don't as neat as having Translingual and English at the top of an entry, ignoring the alphabetisation of the other sections. But I guess these things are subjective. --Droigheann (talk) 04:11, 2 February 2020 (UTC)
    Apart from my own subjective experience, I saw a contributor enter the likes of "A cat." to definition lines for Czech entries, which suggests the contributor intuitively expected consistency. One can get used to all sorts of disunities or inconsistencies, but I do believe the discussed disunity between English and non-English definition lines needs to get used to rather than being natural and preferable. --Dan Polansky (talk) 13:02, 9 February 2020 (UTC)
    *shrugs* I've never said it was more natural to have two different formats. Nor do I think that natural inevitably means preferable: I have little doubt it's more natural to just write "A cat." in plain text without bothering with a hyperlink. --Droigheann (talk) 00:43, 11 February 2020 (UTC)
    On a yet another note: this is a vote that uses a long specific title, "Definitions of English terms should start with a capital and end with a full stop", but then does more than the title promises. I wonder how many of the voters actually payed attention to the multiple specific changes the vote proposes. Formally, this vote leaves a lot to wish. --Dan Polansky (talk) 13:16, 9 February 2020 (UTC)
    Another point is that the texts rendered by templates put on definition lines are used both in English and non-English entries; e.g. "Diminutive of fazole" rendered by {{diminutive of}} appears in Czech fazolka, and "Diminutive" is capitalized. This adds to the benefit of having English and non-English entries unified or else you have a mixture of lowercase and capitalized in one entry, in one list. --Dan Polansky (talk) 12:44, 14 February 2020 (UTC)
    Personally I'd prefer changing a template to follow a policy to making a policy to suit a template. --Droigheann (talk) 04:48, 16 February 2020 (UTC)
    No doubt we can change all relevant templates to either produce lowercase or capitalized output depending on the language passed, but that is additional avoidable entropy. And I still do not know what makes capitalized English definitions preferable; I see no force in an argument to the effect of "I find inconsistency as neat as consistency". --Dan Polansky (talk) 12:46, 16 February 2020 (UTC)
    As I said, starting with Transligual, following with English, then having the other languages alphabetically is a sort of inconsistency as well, that doesn't make it a bad way of doing things. The border between 'desirable consistency' and 'one-size-must-fit-all-at-all-costs' is sometimes subjective to an individual's perception. --Droigheann (talk) 11:50, 23 February 2020 (UTC)
  10.   Oppose O! please! consider the non-anglophones. One would never be sure of a word starting with a capital: is it a proper noun? Is it some name? Please write with capital, what is always written with capital. It would be a great help. sarri.greek (talk) 11:04, 1 February 2020 (UTC)
    @Sarri.greek: I am not sure I understand. This is about definition lines, not entry names. The entry is going to be, say, cat anyway, not Cat. --Dan Polansky (talk) 11:42, 1 February 2020 (UTC)
    @Dan Polansky yes, fr.wiktionary does it: they spell a monolectic translation or definition beginning with capital. As the example above: «Money». Like a german noun. Very confusing. sarri.greek (talk) 11:50, 1 February 2020 (UTC)
    Thank you for the explanation. Now I understand that you are referring to item 2 of the proposed new text, "(slang, US) Money." and that you find it confusing, on the definition line. --Dan Polansky (talk) 11:54, 1 February 2020 (UTC)
  11.   Oppose (at least in cases of (full) synonymy) Extracting information automatically would probably be more difficult, and in case of words lice mercury and Mercury where capitalization results in a different word, you'd need to look at which page the link directs to in order to find out what it truly means. Mölli-Möllerö (talk) 13:28, 1 February 2020 (UTC)
    @Mölli-Möllerö In cases of full synonymy, there should usually be more information so that the definition is not ambiguous, anyway. For instance, rather than just "Mercury." or "Mercury." I would probably write something like "The planet Mercury." or the "The metal mercury." That avoids the ambiguity and allows for consistency in formatting. Andrew Sheedy (talk) 03:27, 2 February 2020 (UTC)
    My argument about automatic information retrieval still holds. If someone wants to use Wiktionary's entries to make a new, edited dictionary, it will be way easier for them to convert money to Money if needed than the reverse, as it's hard to automatically identify which words should be capitalized and which should not. Thus I still oppose. (I'm fine with this new guideline for definitions are not themselves dictionary entries though) Mölli-Möllerö (talk) 12:19, 2 February 2020 (UTC)
  12.   Oppose It's better to lowercase words unless they are capitalized; even in definitions, we should always try to keep it clear whether a word is generally capitalized or not. A period should only be used if the definition is a full sentence. Buidhe (talk) 08:28, 10 February 2020 (UTC)


  1.   Abstain פֿינצטערניש (Fintsternish), she/her (talk) 11:37, 13 January 2020 (UTC)
  2.   AbstainMnemosientje (t · c) 11:12, 15 January 2020 (UTC)
  3.   Abstain --Robbie SWE (talk) 12:03, 7 February 2020 (UTC)


Since we have 21 supports, 12 opposes, 3 abstentions, and the fact that 21/33 < 2/3, this vote ends in no consensus. (It needed three more unopposed support votes to pass.) mellohi! (僕の乖離) 04:06, 11 February 2020 (UTC)

OK, I was not aware of that vote. However, you must concede that 21-12 would appear to most people to be conclusive enough. Mihia (talk) 01:10, 16 February 2020 (UTC)
Let me also point out that this "no consensus" rule is presumably designed to favour the status quo, but we all know in this case that the status quo, which is a random combination of styles, plus confused and poorly written style guidelines, is undesirable. What if we have a vote on the opposite proposal, and it turns out 21 to 12 against? Do we still have "no consensus" and the same crappy status quo? This needs to be rethought. Mihia (talk) 01:28, 16 February 2020 (UTC)
That's an interesting point, if a vote to mandate lowercasing failed 21 to 12, it would feel like there was more pressure to stop lowercasing, even when all that was actually decided was that it couldn't be mandated. But yes, this is what seems to cause us some of the thorniest problems these days: where there is no one rule that is agreed to be the status quo ante (either because people agree the current situation is inconsistency, like here, or because people disagree on what the status quo is), so you have a situation where the vote to mandate (a change to) one practice didn't get a big enough majority to pass, but editors (and I'm not shading anyone, I'm one of them this time) will doubtless continue editing entries piecemeal to use the practice, since no vote to ban it passed, or could pass given the numbers here. - -sche (discuss) 17:44, 16 February 2020 (UTC)