Pronoun structure

What inner structure do pronouns have? If pronouns are DPs, then where's the determiner? And what's up with this súna lıq = "you girls" construct?

High-level overview

As a speaker, all you need to know about this proposal is this:

• We can think of pronouns as determiners, and allow things like jí Toaqpoq = "a Toaqpoq such as me".
• This means the same thing as jí but presupposes that the speaker is a Toaqpoq.

Underlying syntax

Personal pronouns

Officially, we know that ⟦ jí ⟧^g,c = speaker(c).

We can imagine jí_DP as consisting of a determiner jí_D and a null nP.

Then by analogy with regular determiners, ⟦ jí_D ⟧^g,c is $\lambda P:P({\textrm {speaker}}(c)).{\textrm {speaker}}(c)$.

To draw this analogy further, it's tempting to say that the nP is just the same "covert raı" that can be seen (or unseen) after other determiners.

But then the idea of allowing jí Toaqpoq to mean "a Toaqpoq such as me" practically suggests itself! This denotes the speaker, and presupposes that they are a Toaqpoq.

There's a similar grammatical structure in English: in a sentence like "Will you jerks shut up?" the word "jerks" doesn't limit the domain of "you" but merely adds a presupposition.

Anaphoric pronouns

Officially, we know that ⟦ hó_i ⟧^g,c = g(i).

We can imagine hó_DP as consisting of a determiner hó_D and a null nP.

Then by analogy with regular determiners, ⟦ hó_i,D ⟧^g,c is $\lambda P:P(g(i)).g(i)$. This is just the official denotation of bound-the. But the difference (I think) is that the index is now "calculated" or "bound" rather than generated by the nP.

I don't think it's as tempting to allow constructions like hó Toaqpoq, but there's also no harm in being consistent.