Wiktionary talk:Votes/pl-2009-12/Definition layout

Pluralia tantum and such

Plurale tantum, collective nouns, various irregular and defective words should probably be exempted from this. E.g. Serbo-Croatian kamenje, it would be weird to have in the definition line:

1. (collectively) stone, rock

In such cases literal translation seems quite appropriate to me. --Ivan Štambuk 13:03, 30 December 2009 (UTC)

Good point. More generally, when I wrote the text, I only had lemma forms in mind, but nothing in my text specified that; so, it implied that children should have some sort of definition like "A young person." :-P   I see that msh210 has improved the proposal somewhat in that regard, though his edits introduced some other problems (such as the use of the odd term "inflection-free", which I think might be part of what has tripped up EP below). —Ruakh_TALK 19:54, 30 December 2009 (UTC)

But also keep in mind that Latin nascor (used as an example) is itself a "defective / irregular" by virtue of being a deponent verb. Some verbs in other languages do not have a corresponding verb in English, but rather are translated as a condition or state of being. If you allow for exceptions, then you have to allow for exceptions, and not be restrictive or picky in which exceptions are allowed. ELE is a general document, and should not fret over all the little details.

Other classes of words that would trip up the current wording are comparatives and superlaitves (which are lemma forms in some languages like Latin), and certain classes of participles in some languages (Latin again is the example that strangely came to my mind). How would you propose Latin (deprecated template usage) amandum and (deprecated template usage) amandus be defined, since neither belongs to a part of speech we use in English? --EncycloPetey 20:16, 30 December 2009 (UTC)

Re: "If you allow for exceptions, then you have to allow for exceptions, and not be restrictive or picky in which exceptions are allowed. ELE is a general document, and should not fret over all the little details.": Yes, definitely. On this I agree 100%. The document should state and illustrate general principles, but should neither attempt nor pretend to cover all cases. —Ruakh_TALK 20:28, 30 December 2009 (UTC)

Then one way to phrase this proposal (but which itself may need tweaking) is to say that "In general, links within a definition should point to English lemma forms (where one exists), and not to form-of pages." I cannot agree with restricting the way a definition is worded, but I can agree with restricting the linkages. --EncycloPetey 20:39, 30 December 2009 (UTC)

Having read the three sections hereof, I have come to agree with EP’s lemma-linking proposal; it seems to be the solution that yields the greatest utility by minimising the number of clicks made until a person reaches useful content, maintaining definitional literal substitutability, and not confusing readers by making them think that (for example) sum and esse are precisely synonymous. † ﴾^(u):Raifʻhār ^(t):Doremítzwr﴿ 15:33, 1 February 2010 (UTC)

Various objections

The specific example given for Latin should be removed, as the general ELE page deals with English and generalities, leaving details in which specific languages differ to subsidiary pages. Latin examples should not be inserted into ELE, as Latin has its own separate style page.

The wording currently given is also self-contradictory, as it says that nascor should not be translated with inflection, but translating it as "to be born" is in fact translating it with inflection of the present infinitive. The results of implementing this proposal would have humorous consequences on the wording of English entries like (deprecated template usage) can and (deprecated template usage) may. This proposal also has problematic consequences for Brythonic languages which have singulatives; that is, where the lemma (dicitonary) form is plural, and the singular is formed from the plural by the addition of an inflectional suffix. Welsh plant (“children”) is the lemma, while plentyn (“child”) is a singulative form. The proposal currently given forces undue expectations on languages to meet English standards that they will not meet by virtue of being different languages.

English adjectives would suffer the most under the proposal as currently worded, especially words like (deprecated template usage) sickening, since causing is not an English lemma form. --EncycloPetey 18:28, 30 December 2009 (UTC)

Thanks for your comment. I disagree with your first paragraph, because while I see where you're coming from, I don't see a way around it. Various parts of your second paragraph raise interesting points; I'll have to put some thought into how to address them. And I don't understand your third paragraph, because I don't see how the current proposal precludes “causing sickness or disgust” as a definition. (Nothing in the current proposal requires the definition to be headed by "an English lemma form" so far as I can tell.) —Ruakh_TALK 19:54, 30 December 2009 (UTC)

If there isn't a way around the problem I've noted in my first paragraph, then the proposal is fundamentally flawed, as it places a specific stricture on a specific language, which is then rescinded in the allowance for individual languages to have their own styles. The ELE must stick to English, or it becomes a stumbling block to the work in non-English entries.

I'll have to give some more thought into my third paragraph. You may be right, but I think the issue would be worth keeping in mind in any future rewordings of this proposal. --EncycloPetey 20:08, 30 December 2009 (UTC)

The policy should IMHO simply insist on two points:

1. Foreign-language lemma forms must be translated as English-language lemma forms, if it's an established lexicographical practice for the given language (which it is in most of the cases this change aims to encompass)
2. Language-specific exceptions overriding this rule are defined on the appropriate policy/guideline pages (WT:About XXX), as agreed by the relevant contributors.

Discussing countless details for all of world's languages won't get as anywhere. I'd also like to know which current languages with (some kind of) considerable content would this affect. To my knowledge that would be Latin, Ancient Greek and Albanian. Any more? --Ivan Štambuk 01:10, 31 December 2009 (UTC)

The affected languages that come to mind are — in descending order by the number of entries in [[Category:langname verbs]] — Latin, Hebrew, Aramaic, Armenian, Ancient Greek, Arabic, Albanian, and Sanskrit. (Come on, did you really think we had more Albanian than Hebrew?) But I'm sure there are others that I don't know of or am not thinking of. —Ruakh_TALK 04:00, 31 December 2009 (UTC)

Oh, and thanks for your point about if it's an established lexicographical practice for the given language; that's huge. Because that's really the big point, at least for me; if it were the Latin-to-English tradition to give naciōr as "I am born", then that would be one thing, but that's not how anyone does it, so it makes us look clueless when we try to do it that way. (And that is also my response to EP's point above about Brythonic singulatives: I don't know what the tradition is for handling those languages, but whatever it is, we should follow it. Which means the current proposal on the vote page is missing the most important point.) —Ruakh_TALK 04:06, 31 December 2009 (UTC)

The "Foreign-language lemma forms must be translated as English-language lemma forms" is the unworkable part, as the exceptions are extremely numerous in many, many languages. This doesn't work. A quick look in L&S, for example, skimming quickly at the start of the letter "L": labrosus - with large lips; lacrimosus - full of tears; lactarius - containing milk; lacticinia - a dish prepared with milk and eggs; laetificans - rejoicing; lagophthalmos - a person afflicted with a disease of the eye; laniarius - pertaining to a butcher. I skimmed all of these in less than 5 minutes without hunting exhaustively even the few pages I examined, and have marked in bold the words that are not lemma forms in English. We cannot arbitrarily restrict the wording of translations with a criterion that will have this many exceptions. Exceptions being this numerous causes it to fail right away as a general principle.

In response to Ruakh's comment about blindly following "tradition", please read Fielding's Tom Jones, Book V, chapter 1. If you think we must follow whatever tradition is in dictionaries, then we must remove the pronunciations from all Latin entries. We must strip the inflection lines to give only the ending of the genitive for nouns, the gender endings for adjectives, and have only certain verb endings. We should eliminate all Descendants sections. We ought to dispense with categories and links. Citations should be reduced to one per sense, or else eliminated. Etymologies with Ancient Greek in them should not have transcriptions. That is how the other dictionaries do it, so why are we doing things differently? Wanting to be "like everybody else" isn't a good reason for anything:

Nevertheless the people refused to obey the voice of Samuel; and they said, Nay; but we will have a king over us;

That we also may be like all the nations; and that our king may judge us

It wasn't good for them, and it's not good for us. Please point to a single shred of evidence that supports your opinon that your bone of contention "makes us look clueless", as I haven't seen any comments about that from anyone except a very few non-Latin editors have a bone to pick without cause. The people who are editing Latin are happy, and the people who are actually using our Latin entries have been complimentary about our work in that language when the entries exist. The only complaints I've seen are in regard to missing entries. So, wherefore the bee in your bonnet about Latin verbs? --EncycloPetey 06:08, 31 December 2009 (UTC)

Re: your supposed L&S "exceptions": These are not exceptions to the intended proposal. If they violate the current wording (I don't think they do, but if they do), then the problem is with the wording, not with the proposal.

Re: blindly following tradition: Hey, don't look at me, it's you who was instrumental in deciding we should blindly follow the tradition of listing Latin verbs under first-person singular present active indicative forms. ;-)

Re: Latin-verb bonnet-bee: The bee in my bonnet is actually about Hebrew verbs; I'm not all that bothered by what you Latin-y folks want to do. The reason the Tea room discussion focused on Latin verbs is that it was started by a person who I believe — contrary to your statement — does "actually us[e] our Latin entries".

—Ruakh_TALK 17:46, 31 December 2009 (UTC)

Please do not blame me for "blindly following". Mine was hardly blind, as it involved actual lexical and grammatical reasoning, not "we have to do it because we must limit ourselves to be like everybody else". That's your reason, and has never been mine. If you can't find a wording for the proposal that exactly fits your ideas, and your issue is really over Hebrew, then why not just set up details at Wiktionary:About Hebrew? If we have to follow "established lexicographical practice", then we won't have separate entries for any adverbs in quite a number of our languages.

The TR discussion was not contrary to what I said above. Please re-read my comments above to spot your error in understanding. --EncycloPetey 17:59, 31 December 2009 (UTC)

Don't worry, I'm not actually "blindly following", either. But I think that with many existing traditions, we derive a great benefit from following the tradition, whether or not another approach might have made more sense in a total vacuum. There's a reason that even the most new-fangled and innovative Hebrew-English dictionaries still translate עשה as “do” or “to do”.
I've read all the comments, both here and in the TR discussion, and no one thinks what you seem to think we think. *shrug* I just hope that by the time this vote passes, if it does, you'll understand it well enough that you'll be able to follow it without writing defs that you and I both find stupid.
—Ruakh_TALK 19:50, 31 December 2009 (UTC)

There hasn't been a version of this proposal yet that wouldn't yield disastrous results. I hope that by the time this proposal fails, and I hope it will, that you'll appreciate the underlying principle is flawed. --EncycloPetey 20:11, 31 December 2009 (UTC)

I was thinking of languages that these vote would affect (all the other ones you listed other than those 3 translate lemma as lemma). Another exception that comes to mind are Lojban gismu with their predicate-logice type of definitions (which make sense since they can map to several English PoS, verbs, nouns..). Also a probable exception are the similar entries for lexical roots (===Root===) in various languages, with descriptive definition lines. --Ivan Štambuk 16:23, 2 January 2010 (UTC)

Substitutability principle

One of the desiderata for the gloss-type phrasal definitions that are the most common form of English definition is that they be "subsitutable". That is that the definition could be substituted for the word it defines and be correct grammatically, at least in canonical sentences. The proposed wording seems to eliminate this for English.

Moreover the wording has the effect to dictating the structure of English definitions while allowing complete flexibility for other languages.

The substance and the form of this imply to me further reduction of our value to someone who is primarily interested in English. I am very strongly opposed to any wording that would militate against the substitutabilty principle for English definitions.

I am beginning to wonder why I bother contributing here. It has been fun, but there are other ways I can waste my time. DCDuring TALK 12:20, 1 February 2010 (UTC)

Re: "The proposed wording seems to eliminate this for English": I don't see how. For English, this proposal seems to reflect our universal practice. Could you give a specific example of a definition that you like but that would be barred by this proposal? —Ruakh_TALK 12:44, 1 February 2010 (UTC)

It may be that the problem is poor drafting or the very idea of a vote on this, given the genesis of the "issue". It seems like using a weapon of mass destruction where small arms or hand-to-hand combat would be more appropriate. If someone objects to practice within a given language (as I often do to such items as gratuitously non-standard L3 headers; use of archaic, obsolete, and rare English words in glosses; etc), the first question should be about whether there is harm caused either within or without that language's entries. The next question is whether there are non-intrusive, not-too-contentious means to resolve the problem. In this case the simple use of pipes in {{term}} eliminates the effect of the idiosyncratic lemmas of Latin in etymologies. I have had witting direct experience of other problems caused by idiosyncratic practices in Latin (or other languages for that matter).

My difficulty is with possible construction of:

Therefore, the definition given in a lexeme's main entry is not worded to reflect to the specific form that is used as lemma (dictionary form); for example, the Latin verb nascōr should be translated in part as “be born” or “to be born”, even though that specific form means “I am born”.

The wording seems to forbid (is not worded) wording a definition to reflect the specific form. The examples seems to make clear that such is the intent, using the word "should" with a non-substitutable definition form.

After a great deal more thought and discussion about all the ramifications of this, I would be inclined to make mandatory substitutable definitions unless enclosed in {{non-gloss definition}} or similar template to so mark them for future contributors who may have other solutions. DCDuring TALK 13:26, 1 February 2010 (UTC)

I am very confused. Your initial comment seemed to complain about English definitions, and that's what I asked about, since as far as I'm aware, this proposal exactly reflects our current universal practice for English — including your practice. (Or did I misunderstand your initial comment?) Your new comment seems to bypass English entirely and discuss Latin instead. If you have a Latin complaint, that's fine, but can that go in a separate section, rather than as a putative reply to my comment? (Or am I misunderstanding your new comment?) —Ruakh_TALK 14:29, 1 February 2010 (UTC)

The genesis of this proposal was principally (deprecated template usage) nascor. The sole use of (deprecated template usage) nascor as an example seems to confirm the narrowness of the motivation, though I know other languages have been discussed. OTOH, the draft proposal's wording is not limited to Latin or to non-English languages. Given the high turnover of voting contributors and admins we cannot depend on collective memory of intent to provide "common sense" guidance on how to interpret policy votes. I am looking at the proposal in terms of its implications for English definitions and, especially, for the "substitutability" principle. I would not want the wording of any proposal about definitions to be inconsistent with that principle.

In the section excerpted above the words "is" and "should" suggest that there should be a departure from the substitutability principle, which principle, though stated nowhere in our pages AFAICT, lies behind the practice of many dictionaries and our own best practice in English (other languages not having phrasal definitions for the most part). If that is not what is intended, I would like the point made explicitly. Ie, that we prefer substitutable wording for all definitions excluding those worded as non-gloss definitions (speech acts, interjections, function words [many conjunctions, prepositions, pronouns, determiners], certain adverbs, etc). If we are choosing to compel translations from non-English languages to be into a standard form appropriate for an English PoS, as we seem to be, we should say so more unambiguously. If we can't contemplate all the consequences for all languages (Lojban PoSs come to mind.), we should limit the scope of our drafting to a specified list of languages or explicitly allow for exceptions.

I believe that we need to be more circumspect in how we make proposals that have a bearing on multiple languages. My own focus will, necessarily, always be on English and, by choice, on all the users interested in the English L2 section (especially those not participating in these discussions). But those with advanced practical and academic learning in other languages need to contemplate consequences for such proposals on their languages (and on unrepresented users whose need they may understand) as well. DCDuring TALK 17:50, 1 February 2010 (UTC)

I'm not aware of any English dictionaries that follow a "substitutability" principle; for example, see http://dictionary.reference.com/browse/stand. You can't replace "stand" with almost any definition on that entire page: the verb definitions almost all start with "to", the noun definitions almost all start with "a". "We to rise to an upright position on the feet"? I don't think so. What I'm proposing is that we take the same approach for other languages: define the lexeme at the lemma entry, and define it in the same sort of way that we define English lexemes, without regard for whatever specific inflected form happens to be the lemma. —Ruakh_TALK 19:00, 1 February 2010 (UTC)

I do not object to the inclusion of "a" (in countable nouns) or "to" (for verbs) or "the" for proper nouns or certain (rare ?) senses of common nouns. The traditional and mnemonic role of these seems beneficial. Both Landau and Atkins & Rundell commend substitutability as a desiderata for definitions, while mentioning other approaches (situational full sentences, etc). As other dictionaries practice substitutability, transitive verb definienda have substitutably transitive definitions, marked with "to" in the definition. The substitutability principle cannot be mandated, however desirable following it usually is, because there are so many places where it is unlikely to work.

I never doubted that there was good intent to the proposal. (At the very least, I always AGF.) I am concerned about the wording. After the intent is long forgotten, the wording remains. We seem to have had persistent difficulties with the received wording of CFI and ELE. Perhaps we can anticipate some of these problems in the nippable stage. I'd prefer to make a constructive suggestion on the wording but I don't understand the issues with respect to other languages. DCDuring TALK 20:12, 1 February 2010 (UTC)

I'm sorry, I just still don't get what you're objecting to. As far as I can tell, the current and proposed wordings are both completely consistent with current universal practice for English entries (ours as well as other dictionaries'), and are both completely consistent with substitutable definitions as you seem to mean that term. If you believe otherwise, then please be way more explicit, and please give specific examples if at all possible. —Ruakh_TALK 20:53, 1 February 2010 (UTC)