Talk:body part

body part
RFV-sense "The portion of a work of writing which takes up the greatest part of the page." Added in [//en.wiktionary.org/w/index.php?title=body_part&diff=14272854&oldid=13718026 diff]. There is one quotation below this sense already, but it doesn't use this spelling and it's not clear to me that it supports this sense. The other two senses could be sent to RFD because they read like they're quite SOP indeed. - -sche (discuss) 19:58, 9 November 2013 (UTC)
 * Strictly speaking, the first sense seems SoP. But it is in accord with what some other dictionaries have (eg, Collins, Wordnet, Vocabulary.com). See . It seems somewhat conventionalized. The same cannot be said for sense 2, let alone sense 3. I'm not sure what test would validate/objectify that perception/intuition. DCDuring TALK 22:11, 9 November 2013 (UTC)
 * At COCA body and parts have a mutual information score of 7.07, reasonably high but a bit lower than "body adornment(s)", "body makeover(s)", "body lotion(s)", "body orifice(s)". "body image" and "body language" have lower MI scores. In contrast "body politic" has a score of 11.45. DCDuring TALK 22:37, 9 November 2013 (UTC)
 * Hm, what do those scores mean? - -sche (discuss) 03:26, 10 November 2013 (UTC)
 * It is a measure of affinity of one word for another. A high score indicates that two terms tend to collocate; 0 indicates complete independence; a negative number indicates the words co-occur less than completely independent terms. Mutual information explains the mathematics. This gives the specifics of COCA's calculation. I use COCA's MI measure only because it requires no calculation and is suggestive. But we are often interested in something more specific than what COCA's number measures: strict collocation (ie, the lemmas (not words) are directly adjacent in a specified order). That is, "part of the body" counts as a collocation of "part" and "body" in COCA's calculation, but has little bearing on whether body part should be an entry. I'm not sure exactly how to compute an MI-type index that would be better for our purposes. DCDuring TALK 04:08, 10 November 2013 (UTC)

As for the RFV'ed sense, the only dictionary in Onelook that recognizes it, is Wordnik and it uses Wiktionary as its reference! The quote is from a text discussing the classifier system in American Sign Language, thus it's hardly about a "portion of a work of writing". The definition written by Luciferwildcat would better fit the word "body", sense 2.4: "content of a letter, message, or other document, as distinct from signatures, salutations, headers, and so on". I would say speedy. This interpretation has already contaminated dozens of net dictionaries. --Hekaheka (talk) 08:37, 10 November 2013 (UTC)
 * Unfortunately we can't unring a bell. Some of the other sites that have copied us will keep the content they found for quite some time. OneLook itself last indexed Wiktionary in September 2010. As bad as the definition is we may as well follow our rules. DCDuring TALK 16:43, 10 November 2013 (UTC)
 * Is there a rule against correcting an obvious error at once? --Hekaheka (talk) 19:15, 10 November 2013 (UTC)
 * No, but we don't delete items that are in RfV before the process has run its course. It's quite possible the item should have been RfDed. It is always possible that there is a usage community that uses or used the term in a way that could be deemed idiomatic. DCDuring TALK 19:45, 10 November 2013 (UTC)


 * RFV-failed. - -sche (discuss) 22:34, 5 February 2014 (UTC)
 * You forgot to strike. --Hekaheka (talk) 18:48, 14 February 2014 (UTC)