Wiktionary:Beer parlour/2015/April

discussion rooms: Tea room • Etym. scr. • Info desk • Beer parlour • Grease pit ← March 2015 · April 2015 · May 2015 → · (current)

---

International Journal of Lexicography

Hi, I'm not sure where to post my question. Is there someone who has access to the International Journal of Lexicography? I'm interested in an article on Dictionary illustrations because we have some recent discussions about it on Czech Wiktionary (they are strictly prohibited there and routinely deleted) and this article is supposed to provide good insight. However I don't want to spend $28 on it (I tried the DeepDyve free trial subscription but they have this journal only since 2004). Thanks to anyone for any help. --Auvajs (talk) 05:38, 2 April 2015 (UTC)

@Auvajs I have access. I've downloaded it and I'll be emailing you with the pdf shortly. —Μετάknowledge^{discuss/deeds} 06:12, 2 April 2015 (UTC)

Wow, that was so fast! Thanks very much! :) --Auvajs (talk) 06:19, 2 April 2015 (UTC)

Rádo se stalo. —Μετάknowledge^{discuss/deeds} 06:45, 2 April 2015 (UTC)

Images are really useful in dictionaries. I have an Italian language dictionary - sometimes I struggle to quite make out what the definition means, but then they have it in a page of pictures and I think "Oh! One of those!". SemperBlotto (talk) 07:06, 2 April 2015 (UTC)

@SemperBlotto: In case anyone didn't know: Wiktionary:Picture_dictionary. —Justin (koavf)❤T☮C☺M☯ 07:15, 2 April 2015 (UTC)

Suggested modification to sign language syntax

In British Sign Language (BSL) there is a hand shape that is frequently used which can't be described using the current notation (usually called '10' http://www.signbsl.com/sign/10). I suggest adding ...i@... (for [i]nside or t[i]p) to Entry_names, Detailed description of phonemes used in sign entry names to fix this. ...i@... would be underneath ...p@... in the Detailed description of phonemes used in sign entry names for 'The thumb tip contacts the inside of a finger of the same hand'. Nathanael Farley (talk) 07:41, 2 April 2015 (UTC)

On etymology referencing

I recently added a recommendation on WT:ETYM that references for the reconstruction of proto-words should preferrably be centralized on their appendix pages. User:Dan Polansky reverted this with the comment:

I see no discussion supporting such a policy, and it is easy to verify that Wiktionary by and large does not reference etymologies so the putative policy contradicts long-term common practice

This was re-reverted as minor by User:Angr. Fair enough I guess, the page is a draft proposal after all. But I support having some discussion regardless.

There is a vote Reconstructions need references in preparation, but rather than editing that right away I'd like to sound the community for a few propositions — since, in addition to my comment on reference formatting it seems that we lack agreement on reference requirements in the first place as well.

Still, note that in principle these are separate questions. Wiktionary undeniably allows references, and my original comment applies to how they should be organized when present. But regardless, for starters, my thoughts on reference requirements:

1. All etymological information should be verifiable. Which applies to several types of information, e.g.:
   • A given set of words being related to each other at all.
   • A given set of words being related in a specific fashion: usually, common descent or loaning in a specific direction.
   • Reconstructions (phonological, morphological and semantic) for proto-forms of words that are related by common descent.
2. Wiktionary is a work in progress, and edits to etymology sections do not need to immediately have every possible detail sourced.
   • It would probably be a good idea to bring in a {{citation needed}} template and other similar ones for tagging information whose validity someone contests, or which someone thinks should just be cited more clearly. (Also, these two things probably need different inline tags.)
3. Elementary synthesis of sources is allowed. E.g.:
   • Phonetic reconstruction: if we can source to Smorgle that *k in Proto-Foobar is reflected as Foo h, Bar k, and if we can similarly source to Zoop that the Foo word hu and the Bar word ku are related, at this point there is no problem in creating a Proto-Foobar entry *ku, even if we for the moment have no exact citation for this proto-form in the literature.
   • Phonetic reconstruction: if Smorgle gives a Proto-Foobar word *tata, Zoop gives a Steamy Foo reflex haha for this etymological set, and Shroobadoo has argued that Proto-Foobar *t should be reconstructed as *☕ whenever Steamy Foo has /h/, we can just fine reconstruct Proto-Foobar *☕a☕a rather than *tata, even if Shroobadoo never treated this particular word.
   • Etymological affiliation: if von Papperson states that word-initial *pr was forbidden in Proto-Foobar but evolved in early Bar; Böp states that in Swampy Foo there are loanwords such as prumf that have been acquired from Bar; and Zoop states that Swampy Foo pripi and Bar pripi are related, there is no problem in asserting that the former is probably a loan from the latter.

The third point might be the most contentious. Several further caveats may be required, e.g. all claims used as basis for synthesis of sources should probably represent mainstream scholarship. --Tropylium (talk) 07:44, 2 April 2015 (UTC)

I agree on all your points – a dictionary is more than definitions – these are methods that "keep an honest man honest" and separate fringe from scholarly interpretation of facts. Also, scholarly consensus changes and providing references or mentions places the etymology on a timeline for future contributors. BTW, User:Dan Polansky also doesn't like my documentation method of including content that is not an attestation of usage, and doesn't like me referring to Wiktionary:Citations and Wiktionary:References since they are not policies. You are not alone. —BoBoMisiu (talk) 13:32, 4 April 2015 (UTC)

• Yes, "All etymological information should be verifiable" but that does not mean that the English Wiktionary should demand that parts of etymology are referenced using <ref> technique. I oppose such demand, and so does the overwhelming previous practice. --Dan Polansky (talk) 13:41, 4 April 2015 (UTC)

@Dan Polansky: I noticed your subtle moving the goalposts from a discussion about references in an entry into a discussion about a particular format of references (wrapping references in <ref>s). Referencing helps future contributors, shows quality and credibility of the content, demonstrates veracity, and is valid rationale to defend against accusations of plagiarism and copyright infringement. Transparency is a good. —BoBoMisiu (talk) 15:31, 4 April 2015 (UTC)

Moving the goalposts would be making the standard stricter to exclude what passed the earlier one. As to the point at hand: you do tend toward excess in citing. You don't need to nail down every single detail in a definition with a reference. Etymologies need to be referenced, yes- but not definitions. A descriptive dictionary defines terms the way people use them, not the way reference works specify is correct. If people use water beetle to refer to a cockroach, so do we- even though a cockroach isn't a beetle. Technical terms such as taxonomic names are somewhat of a special case, since correctness as a technical term is relevant. Still, the taxonomic literature is full of misapplied taxonomic names, and of changes in meaning due to splitting and lumping. In my area, most of the scrub oaks referred to as are really - for a long time no one paid attention to the difference. While true Q.dumosa only grows along the coast, there are all kinds of references in the literature in numerous disciplines to Q.dumosa being found/used, etc. far inland. Wiktionary isn't a journal of record, and citing details in the definitions as if we were is a bit deceptive and adds to clutter. That's not to say we shouldn't have links to other, more comprehensive sources, but only in moderation. This is a dictionary, so we try to keep things structured and streamlined for ease of use. Chuck Entz (talk) 17:12, 4 April 2015 (UTC)

@BoBoMisiu, Dan Polansky is not 'subtly moving the goalposts' as what he refers to was part of the original reverted addition. I think the way the server interprets <ref></ref> syntax is ugly, and I try to avoid it. Also, people over use it, they use it for citations instead of just writing out the citation next to the sense which is to be cited, which is the best way to do it. Renard Migrant (talk) 17:23, 4 April 2015 (UTC)

Also verifiable is not the same as verified. Verifiable means if you try to verify it, you can. That's what we want. If something's not contentious, don't add a reference for it, because we end up with references coming out of our ears and the actual definitions become hard to find even for experienced users. Renard Migrant (talk) 17:26, 4 April 2015 (UTC)

@Chuck Entz I agree with you, "If people use water beetle to refer to a cockroach, so do we- even though a cockroach isn't a beetle." Yet, having an entry that doesn't clearly distinguish for the user between what an authoritative sense and what is another just-as-real usage lacks something. As for "making the standard stricter to exclude what passed the earlier one", the three attestations of usage are unaffected, but it structures entries into the haves and the have not, i.e. those entries that are denied a connection to what is considered factual – that is a systemic problem that will not go away. As to ease of use, a collapsed section has all the content and yet provides a visually simple interface, which could be added by a bot, so that is not the same issue as ignoring what every school child is taught (to provide credit to your source and avoid plagiarism and copyright infringement).

@Renard Migrant I agree with you, "verifiable is not the same as verified", that is not the same thing as not giving credit to your source. The notion that excluding things so it looks pretty is just style over substance that can be solved by collapsed sections. —BoBoMisiu (talk) 20:31, 4 April 2015 (UTC)

Etymological dictionaries usually do not give credit to sources on a per-entry level. The idea that you need to credit your source of information on a per-entry level or else you have a copyright infringment is wrong. Copyright protects expression, not information; you can take someone else's information but you cannot take their expression - a particular sequence of words or sentences; if you take someone else's expression, you are liable to have copyright infringement and referencing does not make it much better. Nonetheless, I when I was entering etymologies from a public domain source, I nearly always stated my source in the edit summary; I saw many editors of etymologies not do even that much. --Dan Polansky (talk) 18:03, 4 April 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘@Dan Polansky "Etymological dictionaries usually do not give credit to sources on a per-entry level" because they are paper. Deciding what is or is not someones creative work or potentially discovery is beyond most contributors; transparency is not beyond most contributors. —BoBoMisiu (talk) 18:22, 4 April 2015 (UTC)

That's not just because they're paper, it's also because they have onymous authors who have reputations to consider and who can generally be trusted to do their research. We don't have that luxury. Because we're a wiki that anyone can edit, we have to show that our etymologies haven't been pulled out of our collective ass. —Aɴɢʀ (talk) 18:41, 4 April 2015 (UTC)

Bravo! —BoBoMisiu (talk) 18:47, 4 April 2015 (UTC)

I refuted the copyright infringement assertion, and none of the things added above counterargued that refutation. As for the other subject whether we want to use <ref> in order to show which sources were used, I object to making that a demand or the recommended practice. An editor was adding <ref> to etymologies of Czech entries, and while I did not like it, I did not revert it. But it is one thing to tolerate the use of <ref>, and an entirely different thing to pretend that the English Wiktionary has a policy that requires contributors of etymology to use <ref>; I oppose such a policy, and claim that the use of edit summaries to state sources should be enough; in fact, we have not been chastising editors who provided zero edit summaries and zero references using <ref>. --Dan Polansky (talk) 18:57, 4 April 2015 (UTC)

Nobody can copyright that the English word lemma comes from the Ancient Greek word λῆμμα (lêmma). Also you've confused excluding things with not including things. You're basically arguing that literally everything should be included in every entry, relevant or otherwise. That's nothing to do with style over substance, that's insanity over substance. Renard Migrant (talk) 19:06, 4 April 2015 (UTC)

@Dan Polansky where did you refute it?

@Renard Migrant I agree with you, but chains of assertions that are copied from sources are someone's work. —BoBoMisiu (talk) 19:17, 4 April 2015 (UTC)

Nothing at WT:ETYM requires contributors to use <ref> or to provide sources for their etymologies at all. Doing so is recommended whenever possible; it is not required. Etymologies without references are more likely to be removed as possible bullshit than ones with references, but it is neither the case that all unsourced etymologies are bullshit nor that all sourced etymologies aren't. —Aɴɢʀ (talk) 19:25, 4 April 2015 (UTC)

@Angr I was reminded WT:ETYM is only a draft, so I started to ignore it, maybe I should follow it. I think references add transparency. Like you wrote, that bullshit still needs to be removed whether it is referenced or not. I disagree with Dan Polansky that an edit summary is equivalent to a fully cited reference. I think Tropylium's "Wiktionary is a work in progress" is valid and I suggest forking w:Template:Citation needed span from wikipedia. That template wraps the particular questionable content in a visually distinct box, making it easy to see where contribution is requested and what is suspect. A "synthesis of sources" is a good idea as long as it references where each piece of information comes from. That way a synthesis is transparent. I also think forking w:Template:Citation/core and changing the R: templates into wrappers with consistent parameters is a good idea and a step away from willy-nilly flat references toward structured machine readable data, but that is a future discussion. —BoBoMisiu (talk) 20:31, 4 April 2015 (UTC)

I think you'll find that forking templates from Wikipedia is not exactly popular around here. The people who propose doing so usually don't understand the difference between Wikipedia, with its large editor population, extensive bureaucratic infrastructure, and loosely-formatted, single-entry structure vs. Wiktionary, with its small editor base, simple rules/procedures and rigidly-formatted multi-entry structure, which requires that everything be concise and to the point. As for hiding the References section: that would still leave lots of superscripts, most of which link to statements of the obvious. If you really want to be thorough, how about word-frequency statistics? Binary checksums? Diagramming of the sentence structures? Those can all be hidden away in boxes, but will all lead to reader resentment for wasting their time if they follow the references. It reminds me of pulling over in the middle of nowhere because of a light on the dashboard: it was time for an oil change based on the mileage. If you have too much pointless information, hidden or not, the important information gets lost in the clutter. Chuck Entz (talk) 22:49, 4 April 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘So a discussion about objective etymology referencing comes down to the subjective – a delicate aesthetic preference of looking at a pretty entry without seeing those brutish superscript numbers? That reminds me of the quote from Amadeus: "It's quality work. And there are simply too many notes, that's all."

@Chuck Entz the solution to your concern "that would still leave lots of superscripts" is to allow the user to choose through a user preference toggle.
References are not just for statements of the obvious. Your propositions about "word-frequency statistics? Binary checksums? Diagramming of the sentence structures?", are an unintentional red herring about potential uses of collapsed sections, they are about different propositions than the one presented by Tropylium which is about references. The revision as of 19:26, 1 April 2015 is:

Etymologies should be referenced if possible, ideally by footnotes within the “Etymology” section, secondarily just by listing references in the “References” section. The Reference templates are useful in this regard.
If a word descends from a common root with several words in related languages, and an appendix page exists for the reconstructed proto-form, references on details of the reconstruction are best placed on that page, rather than duplicated across the cognate entries.

For six weeks, I misunderstood: Etymologies should be referenced if possible, ideally by footnotes within the 'Etymology' section, secondarily just by listing references in the 'References' section.
I, for six weeks, created sub-sections within entry etymology sections. I would like it changed to:
Etymologies should be referencedinclude Wiktionary:References if possible(for policy, see Wiktionary:Entry layout explained#References), ideallypreferably by footnotes within the 'Etymology' section(see Help:Footnotes), secondarily just by listing referencesin theor simply adding the source to the entry 'References' section (for policy, see Wiktionary:Entry layout explained#References).

While Wiktionary:Entry layout explained is a policy, I would like to see a clear explanation about the actual status of Wiktionary:References is. Is Wiktionary:References a policy?

I think that the other paragraph would be more understandable as, If a wordterm descends from a common root with several wordsother terms in related languages, and an appendixa page exists for thethat reconstructed proto-formterm (for policy, see Wiktionary:Reconstructed terms), references on details ofabout the reconstruction are best placedpreferably located on that reconstructed term page, rather thanand not duplicated acrossin the cognate entries. —BoBoMisiu (talk) 15:39, 5 April 2015 (UTC)

Finding old deletion discussions, and related points

An attempt to create the entry "get behind" ([1]) leads to a page that says "You are recreating a page that was previously deleted" and then provides a log entry that reads "deleted page get behind (Failed RFD, RFDO; do not re-enter)".

1. How do I find the deletion reason/discussion? The link to "RFD" in the log entry just throws me into the current version of that page, which is useless long-term. Furthermore, the supposed RFD archive page ([2]) says that the archive is no longer active, but that "The current procedure is to archive the RFD discussion to the corresponding article's talk page". Er ... can anyone else spot the flaw in that procedure??

2. The explanation of "Failed RFD; do not re-enter" at [3] says "This term (in a particular language) failed WT:RFD. Do not re-enter it. You may re-enter a different term, especially in a different language." Suerly "Do not re-enter it" is too final? What if something previously rejected gains a valid meaning in the future? (I'm not suggesting this is the case for "get behind", but obviously it could happen.) Also "You may re-enter a different term, especially in a different language" is odd and pointless. 31.51.1.15 17:27, 2 April 2015 (UTC)

• Subsequent to writing the above, I have discovered that the talk page http://en.wiktionary.org/wiki/Talk:get_behind exists even though the page "get behind" does not. I did not realise this was possible. Is this how it is generally supposed to work for deleted entries? This is unintuitive, and I propose to updated the documentation to explain it. However, what is the expected procedure for navigating to such a page? I found http://en.wiktionary.org/wiki/Talk:get_behind by first navigating to a known existing talk page and then changing the last part of the URL. Is this what users are expected to do, or is there a more friendly way? 31.51.1.15

You're right that we should have some message that the deletion debate is on the talk page, or should be eventually on the talk page. Attempts to create talk page archived by bot are underway. I completely agree that sometimes it's really hard to find deletion debates because of change in archive methods over the years, and sometimes debates are not archived anywhere. Renard Migrant (talk) 23:16, 2 April 2015 (UTC)