[matrix] Pronoun Relations

[Emily M. Bender]

Picking up this thread again after some delay...

It seems to me that we have an interesting case here where there
is some opportunity to achieve greater harmony in MRSs across languages
and yet reason to avoid doing so (i.e., the practical realities of
current anaphora resolution technology for the ERG).

It also seems to me clear that we will need a different anaphora
resolution strategy in pro-drop languages than the one currently
being developed for English, since the pronoun equivalents have
no overt position in the string anyway.  I think it follows that
we don't lose much by going to a no-pronoun-rel strategy in the
Matrix while keeping the pronoun_rels for the ERG: we're going to
have differences between languages no matter what we do.

On Sun, Feb 04, 2007 at 10:14:21PM +0000, Ann Copestake wrote:
> Currently, what we get out of having the relation rather than just an
> index is the cfrom/cto information which allows us to take cataphora
> into account.  As far as I can see, it would be tricky to reconstruct
> this if we just have an MRS with indices and no relations for the
> pronouns.  For instance, the MRSs for A and A' below are presumably
> identical (modulo cfrom/cto and any information structure) but we need
> to be able to distinguish them.
> 
> A: Because the cat purrs, Kim likes it.
> A': * Kim likes it, because the cat purrs.
> 
> I have no clue about what happens for Japanese, but we'll have a more
> poorly-performing algorithm for English if we don't take this into
> account.  Maybe one should do this off the cfrom/cto values for the
> relations for which the pronouns are arguments, but the Lappin and
> Leass treatment, for instance, just involves giving a negative weight
> to pronouns coming before their putative antecedants.  

It seems like taking the cfrom/cto from the selecting predicate
should work.  I tried for a bit to come up with some examples 
in English that would tease the two apart, but without success
(largely because it's hard to separate pronouns from their selecting
predicates by enough material to stick in an antecedent).

?*Kim walks, because the grocery store is close, to it.


> Obviously this is a bit of a hack, but I would argue that it's
> currently a useful surrogate for a proper approach to information
> structure.  There is also a practical point that concerns evaluation
> of anaphora resolution: we need to be able to locate the pronouns in
> the text and it's a pain to do this on the basis of indices alone.  In
> fact, I think it may be impossible in the limit - `Kim gave it it' vs
> `Kim gave it to it' - silly example, but the point is it would entail
> non-trivial code writing to do this properly.

Wouldn't it be possible to relate the pronouns to their indices by
looking at the actual derivations, and going from pronouns to MRS,
rather than the other way around?
 
> Incidentally, I think that there's an argument that we should be
> distinguishing between the different pronouns in terms of their
> relations, since right now, we're only capturing distinctions that
> make it onto indices.  This implies, I think, that we'd have to have a
> politeness value on indices to record the German du/ihr/Sie
> distinction and presumably some sort of dialect feature to indicate
> non-standard pronouns (e.g., the `us' for 1sg in some BrE dialects).
> I don't know what sort of variation is possible, and maybe it's OK to
> say that all variations in pronoun use are either ignored or put on
> indices, but I do worry that this could lead to index proliferation as
> new languages are dealt with if we go down the route of removing
> pronouns.

Is this kind of variation relevant to anaphora resolution though?
I think the best representation of politeness is through a separate
relation predicated of that individual (perhaps in CTXT or Background,
rather than in the CONT).  If we look at reference resolution more
generally (rather than just anaphora resolution) I think we'll have
other reasons for taking more relations into account (beyond just
the noun-relation introducing the index).

> I think there are existing mechanisms to allow a clean transfer
> approach.  I believe that there is the possibility of monolingual
> rules in the transfer machinery.  So you write a rule that just drops
> all pronoun relations from languages like English before transfer.
> You need another rule to introduce them for generation, but this is
> very similar to what you'd have to write anyway for the null semantics
> items in the generator.  I believe that this is the right mechanism to
> use for investigating systematic differences between languages: it's
> far less disruptive than changing the Matrix.  It may well turn out
> that once an inventory of rules has been developed, an overhaul of MRS
> generally is called for.  But given that many of us are using it for
> applications other than semantic transfer, we'll have to try to take
> different needs into account.

Would you count leaving out pronoun-rels in the Matrix-derived grammars
as an overhaul of MRS?  It seems like a relatively trivial change to me
(which doesn't effect overall well-formedness, etc).

Emily