Arguments of 'sing' (p. 34) [via PaperHive@docloop]

[docloop[bot]]

Annotation imported from PaperHive.

Regarding this part:

The benefactives seem to pattern with normal ditransitives here.

Ash Asudeh wrote:

This passage is slightly inaccurate. On page 81 we write that the rule in our (65)/your (36) is required to parse the V-DP-DP string but in fact the lexical entry we give for 'sang' in (72) does treat it as taking a patient but lets it act like optionally transitive verbs like 'eat' in having an optional UNDERSTOODOBJECT template (see Asudeh & Giorgolo 2012). We do this because we argue that 'sing' is not a true cognate object verb, since there are different syntactic patterns compared to, e.g., the truly intransitive cognate object verb 'laugh'. So in fact 'sing' is a transitive verb but one with very specific restrictions on its complement. In any case, though, it's a bit misleading to say that 'the children' is 'contributed by the c-structure rule that optionally licenses a benefactive'. In fact, the c-structure rule is necessary to parse any string of V-DP-DP sequences such that the verb may be a benefactive and the first DP is an OBJ and the second DP is an OBJ_\theta. The *reason* we propose this way of doing things for English is that it seems to be the configuration itself that gives the benefactive reading (but it is impossible to tell versus a lexical rule approach, of course, just based on this sort of consideration) and because, as you observe on the next page (bottom): "The benefactives seem to pattern with normal ditransitives here." Therefore, a straightforward way to treat benefactives as 'normal ditransitives' is to use the same rule for both. However, of course we could instead move the @BENEFACTIVE template call to the relevant lexical entries, in which case we would capture the data in something more akin to your lexical rule approach.

Link to original comment. About docloop.

[docloop[bot]]

Annotation imported from PaperHive.

Stefan Müller wrote:

I do not see what is inaccurate here. Ditransitive verb means that we have V-DP-DP. I guess for this it is not important how you specify "sing". > In any case, though, it's a bit misleading to say that 'the children' is 'contributed by the c-structure rule that optionally licenses a benefactive'. In fact, the c-structure rule is necessary to parse any string of V-DP-DP sequences such that the verb may be a benefactive and the first DP is an OBJ and the second DP is an OBJ_\theta. Would "licenced" be better? This is the Standard-CxG story: whithout the benefactive-template the V' would not be licenced and the additional element exists only because of this c-structure rule. > However, of course we could instead move the @BENEFACTIVE template call to the relevant lexical entries, in which case we would capture the data in something more akin to your lexical rule approach. Yes, you could. This is what the book is about: I would like you to do exactly this since attaching it to the phrasal level results in a lot of problems.

Link to original comment. About docloop.