dr jasper holmes, On 11/12/2009 14:38:
> I think you're right, Dick. There's no reason why relations can't be
> arguments or values of relations generally, but I think the arguments
> of syntactic relations (like eg complement) must be words. Not least
> because of their word order properties.
Would you be willing to countenance relations that are also words, then? Something that both isa Relation and isa Word?
--And.
> Jasp
>
> 2009/12/11 Richard Hudson <[log in to unmask]>:
>> Thanks for the rephrase, And. I suspected that was what you were getting at.
>> As always, it's a good idea worth exploring.
>>
>> Take your example (1).
>> (1) I want it back.
>> You're suggesting that "it" is the value of the relation 'subject/x' (i.e.
>> an instance of the general 'subject' relation) and that "back" is its
>> argument, and that 'subject/x' is the complement of "want".
>>
>> One difficulty I can see peeping over the horizon is that some dependents
>> are words, while others are relations. Similarly, some order rules will
>> apply to words, and others to relations. I can understand what it means for
>> a word to follow another word, but I don't understand what it means for a
>> relation to follow a word.
>>
>> Or am I missing something?
>>
>> Dick
>>
>> And Rosta wrote:
>>> Michael Turner, On 10/12/2009 15:15:
>>>>> Somewhere along the line Dick and I (or just Dick, or Dick and some
>>>>> other people) realised that the relationships might actually be blobs
>>>>> too (because they behave in so many ways just like Things), and that's
>>>>> what you get in LN. I'm not certain that all the ramifications of this
>>>>> have been worked out as yet, but I think I am convinced that it's
>>>>> correct.
>>>> We'll figure it out, I hope. But for now, can we at least stop calling
>>>> things (note the lower-case t) names like Blobs and Arrows? The map is
>>>> not the territory.
>>> Translating and reposing the key bit of my question:
>>>
>>> Can a Relation be an 'argument' or 'value'? (And if not, why not?)
>>>
>>> Jasper says the answer is nowadays Yes (which pleases him and me). As I
>>> noted in my previous msg, it has important implications.
>>>
>>> For example, in trad WG "want" in "I want it back" has two complements,
>>> "it" and "back", and "it" is subject of "back". However, there is now a
>>> further possibility, that "want" has a single complement, and this
>>> complement is (or isas) the Subject relation whose argument and value are
>>> "it" and "back". IMO this is the Right Analysis.
>>>
>>> --And.
>>>
>>>
>> --
>> Richard Hudson; www.phon.ucl.ac.uk/home/dick/home.htm tells more about
>> me, my work, my views on Israel and my family.
>>
>
>
>
|