Here are my ideas for roı. These structures only account for unquantified DPs (or quantified DPs if we want them to take wide scope, which is probably not right).
The basic structure of coordination shall be this:
Both names, "CoP" and "&P" are used in the literature. I like &P because it's easy to tell apart from everything else, while CoP is very similar to CopVP.
Note that because the first DP c-commands the second DP, we get the intended reading for tu pỏq roı pó hó kủne "everyone and their dog".
Here is roı-&P in the context of a sentence (same as any other argument):
Now, the idea is that roı, like D, can morph to signal a relationship with a higher Q head. If the Q head is [∀], roı becomes ru. If the Q head is [∃], roı becomes ra:
tu > rusa > ra
(note the parallel vowels)
Example with [∀]:
Example with [∃]:
This approach allows us to capture the behavior of ra/ru DP coordination without needing to rely on clausal coordination.
Things get more complicated with quantified DPs, however, and we need to first decide how we want things to scope. A mea in front of the &P would be one way to model &P's opaqueness. I think the opaque reading is generally the more useful one (consider the transparent reading of tu kảto roı sa kủne where both ∀ and ∃ have scope over &P for an example of the weird meanings wide scope produces). On the other hand, the logical expansion of the opaque reading is much more complex. It would be nice if simple cases of roı with unquantified DP could keep the simple structure above.
Note that mea is already involved in the formation of QP. Q is normally defined to select for predicates. To combine with &P/DP, Q needs to use an alternate rule, which in the case of [∃] would be:
λrₑ. λS₍ₑ ₜ₎. ∃x : meaga(x, r). S(x)
If we don't want this, then all &P must come with a mea by default, and then the whole mea CopVP moves to Comp,Q. The result is the same, but there is more syntactic structure.
Here are my ideas for roı. These structures only account for unquantified DPs (or quantified DPs if we want them to take wide scope, which is probably not right).
The basic structure of coordination shall be this:
Both names, "CoP" and "&P" are used in the literature. I like &P because it's easy to tell apart from everything else, while CoP is very similar to CopVP.
Note that because the first DP c-commands the second DP, we get the intended reading for tu pỏq roı pó hó kủne "everyone and their dog".
Here is roı-&P in the context of a sentence (same as any other argument):
Now, the idea is that roı, like D, can morph to signal a relationship with a higher Q head. If the Q head is [∀], roı becomes ru. If the Q head is [∃], roı becomes ra:
tu > ru sa > ra
(note the parallel vowels)
Example with [∀]:
Example with [∃]:
This approach allows us to capture the behavior of ra/ru DP coordination without needing to rely on clausal coordination.
Things get more complicated with quantified DPs, however, and we need to first decide how we want things to scope. A mea in front of the &P would be one way to model &P's opaqueness. I think the opaque reading is generally the more useful one (consider the transparent reading of tu kảto roı sa kủne where both ∀ and ∃ have scope over &P for an example of the weird meanings wide scope produces). On the other hand, the logical expansion of the opaque reading is much more complex. It would be nice if simple cases of roı with unquantified DP could keep the simple structure above.
Note that mea is already involved in the formation of QP. Q is normally defined to select for predicates. To combine with &P/DP, Q needs to use an alternate rule, which in the case of [∃] would be:
λrₑ. λS₍ₑ ₜ₎. ∃x : meaga(x, r). S(x)
If we don't want this, then all &P must come with a mea by default, and then the whole mea CopVP moves to Comp,Q. The result is the same, but there is more syntactic structure.
roı is not as simple as it seems.