|WikiProject Linguistics / Theoretical Linguistics||(Rated B-class)|
Origin of the c
is there any explanation for the c? From the meaning, you could expect co-command or something. But what is it with all these arbitrary names, "c", "alpha", "theta" and what have you? You would expect that linguists would somehow care about their terminology. 220.127.116.11 17:51, 1 November 2005 (UTC)
- I think the c stands for constituent. But it could stand for command (the way the D in D-day stands for day, and the H in H-hour stands for hour). Angr (talk • contribs) 14:55, 22 March 2006 (UTC)
- From my understanding, it does stand for "command" but not in the same geminative way that D-Day stands for "day". When the terminology was first introduced there was also a motivator or process of some sort called "kommand," which is the German spelling. So when they were spoken of in terms of talks and presentations, one had to say "c-command" and "k-command". K-command fell out of practice (as things in academics, and specifically, syntax, often do) but the name "c-command" stuck. I don't have the reference at hand now, but I believe Andrew Carnie brought it up in his Generative Introduction to Syntax. JesseRafe 03:12, 3 October 2006 (UTC)
- JesseRafe is right; this is located in Chapter 3, page 77 of the 1st edition of Syntax: A Generative Introduction. Except Carnie points out that this etymology is just a theory and he doesn't mention evidence one way or the other. I have found some other articles (also here, here, and here showing that "kommand" did exist; but it was referred to sometimes as k-command and sometimes as cyclic-command, and c-command was already referred to sometimes as "constituent-command" (as in the first link above) so it's not totally clear if the term c-command came from the need for a distinction between kommand and command (as Carnie claims) or if it is a shortening of the [possibly] pre-existing term constituent-command. --Politizer (talk) 22:42, 24 August 2008 (UTC)
No need for rumors or secondary sources. The term is introduced for the first time in Tanya Reinhart's 1976 MIT dissertation entitled "The Syntactic Domain of Anaphora", on pages 32-33. There she writes "I will now suggest that the syntactic domain which is relevant to the application of the coreference restriction is to be stated in terms of a relation which I shall call 'constituent command' (hereafter, 'c-command')" -- and is accompanied by a footnote thanking Nick Clements for suggesting the "felicitous name 'c(onstituent)-command'. Clangiphor3 (talk) 17:19, 17 February 2010 (UTC)
Rjanag added a  tag, with the comment "independent reference is still needed to verify that that was the first use". Do we actually need another source stating "this was the first use", when she makes it clear in the cited source (her dissertation) that she made the term up (albeit with help)? Is her say-so (plus the general attribution of the term to her) insufficient? Clangiphor3 (talk) 00:43, 18 February 2010 (UTC)
- Section 2.5 on page 11 of "A theory of command relations" (1989) by Chris Barker and Geoffrey K. Pullum says: "The name 'c-command' (for constituent command) was originally suggested by G. M. Clements to denote the relation that Reinhart (1974) calls 'superiority'. As Reinhart states (1974, 105, n. 3), her superiority relation is identical to the inverse of the 'in construction with' given by Kilma (1964, 297)." The citation of "Reinhart (1974)" refers to proceedings from a (student?) conference. I'm not sure why there isn't any mention of Reinhart's redefinition of c-command in her dissertation two years later, and the references of the paper do not elaborate on who this Clements person is. (For what it's worth, this same source also describes k-command and m-command, as well as a bunch of other types of commands.) —Gordon P. Hemsley→✉ 04:22, 16 March 2010 (UTC)
I put in a request that a page be started for Dominance, so that it doesn't have to be defined on the c-command page and instead can be wikilinked to. Barring that, we could at least edit the Parse tree page to have a section listing structural relations (c-command, m-command and the now-defunct government, dominance, precedence, etc.)--because it doesn't seem right to have a definition of dominance just randomly stuck into the page on c-command. --Politizer (talk) 23:13, 24 August 2008 (UTC)
Backward Anaphora Constraint
A recent discussion on Language Log  reveals that the notion of c-command was proposed even earlier, in 1969, by Ronald Langacker, but under the name of "Backward Anaphora Constraint". The reference is as follows:
Langacker, Ronald. 1969. On pronominalization and the chain of command. In Modern Studies in English, eds. Schane and Reibel, 160–186. Englewood Cliffs, NJ: Prentice Hall.
Looking back over the edit history of the article, it looks like the example goes through cycles of being correct/incorrect. Occasionally somebody fixes the example, but then someone else changes it to be incorrect or incomplete. Since I've just reverted the most recent change, to bring the example to a correct state, let me explicitly say why all the statements in the example section are now correct:
- M does not c-command any node because it dominates all other nodes.
- A c-commands B, C, D, E, F, and G. It certainly dominates none of these nodes, since A dominates nothing. None of these nodes dominates A because only M dominates A. Finally, the first branching node above A is M, and M dominates these six nodes.
- B c-commands A. Indeed, B doesn't dominate A, A doesn't dominate B, and the first branching node over B is M, which dominates A. B does not c-command M because M dominates B, and B does not c-command the other five nodes because B dominates each of those other five.
- C c-commands D, F, and G. Clearly C dominates none of these and none of these dominate C. The first branching node above C is B, and B dominates D, F, and G. C does not c-command B or M because B and M both dominate C. C does not c-command E because C dominates E. C does not c-command A because the first branching node over C is B, and B does not dominate A.
- D c-commands C and E. D does not dominate C and E, clearly, and C and E both do not dominate D. Moreover, the first branching node over D is B, which does dominate C and E. D does not c-command F and G, because it dominates them. D does not c-command B or M, because it is dominated by them. Finally, D does not c-command A, because B is the first branching node over D and B doesn't dominate A.
- E c-commands D, F, and G. Clearly E c-commands none of D, F, or G, since E in fact dominates nothing. Nor do D, F, or G dominate E since E is only dominated by C, B, and M. Also, the first branching node over E is B, and B dominates all three of D, F, and G. So E c-commands each of D, F, and G. On the other hand, E does not c-command any of C, B, or M, because each of these dominates E. Finally, E does not c-command A because the first branching node above E is B, and B does not dominate A. Yes, I think this is an error that needs correcting.
- F c-commands G. Neither one of F or G dominates the other. Also, the first branching node above F is D, and D does dominate G. So F c-commands G. On the other hand, F does not c-command D, B, or M, because each of these dominates F. Finally, F does not c-command A, C, or E, because none of these is dominated by D, the first branching node above F.
- By symmetry, G c-commands F and nothing else.
I have just reverted an edit which removed the fact that E c-commands D. In the past, there have also been editors who thought that E c-commands nothing (but it actually c-commands D, F, and G), or that E c-commands C (it doesn't, because C dominates E), or that F and G c-command nothing (but they c-command each other), or that B c-commands nothing (but it actually c-commands A). Various editors have also fixed the example, for example Noccca on 4 Jan 2011, Rjanag on 19 Jun 2011, me on 18 Aug 2012, and me today. So basically on three occasions the example has been fixed (I think), only to be made incorrect again by subsequence edits.
If anyone feels that the current statement of the example is incorrect, please state your reasons in this section when you make the relevant edit, just so the rest of us can understand what the issue is.
2/1/17 - Edited to replace "E c-commands D, F and G," with "E c-commands nothing." The reason for this is that the definition of c-command in use by Andrew Carnie (2013) directly contradicts the claim that "E c-commands D, F and G," which states "Node A c-commands node B if every node dominating A also dominates B and neither A nor B dominates the other." Since, in the example, the node immediately dominating E does not dominate D, F or G, then E does not c-command any of those nodes. 18.104.22.168 (talk) 09:42, 1 February 2017 (UTC)LingStudentUW
- If that is Carnie's definition, then it isn't the one used in the article. The example is supposed to be according to the definition used in the article. Perhaps there are two (or more) definitions in use? W. P. Uzer (talk) 11:37, 1 February 2017 (UTC)
So what does c-command mean then?
The article fails to explain what c-command means, other than to say it means X is a sibling or uncle of Y. Surely there is more to it then that. — Preceding unsigned comment added by 22.214.171.124 (talk) 00:13, 24 December 2013 (UTC)