Notes
-------------------------------------------------------

   * In my attempt to implement the past perfective tense, I'd listed
     the conjugations of the verb avoir (to have) as slots of a single
     auxiliary type.  Since the customization system did not allow me
     to specify features of these types, I revised this to make each
     PER-NUM combination its own aux slot.  This was an implementation
     detail and didn't really affect the test suite.

I think you mean "aux type", not "aux slot".


#256 With both est-ce que and S-V inversion  
Source: author
Vetted: f
Judgment: u
Phenom: q
est-ce-que mangez vous les biscuits?
est-ce que mang.ez vous    les     biscuits 
are        eat.3PL you.2PL the.3PL cookies.3PL-masc 
'are you eating the cookies?' 

  This parsed, I suppose because the customization system doesn't
  allow one to specify that if one method is being used, the other
  shouldn't be allowed.

Interesting.  I think you can rule this out by saying INV - on the
complement of est-ce-que.  (Tdl editing.)

  Where the subject was coordinated, the verb should then become
  plural.  Unfortunately, at this time, it does not.

More precisely, it's not required to, but this example does parse:

#262 Coordinating the subject 
Source: author
Vetted: f
Judgment: g
Phenom: crd
la fille et le garçon mangent les biscuits 
la          fille        et  le           garçon       mang.ent les     biscuits
the.3SG-fem girl.3SG-fem and the.3SG-masc boy.3SG-masc eat.3PL  the.3PL cookies.3PL-masc 
'the girl and the boy are eating the cookies' 

... in fact, it parses twice.  Your no-drop rule appears to be optional,
when it shouldn't be.

  Where 2 VPs were coordinated, it did not parse.  I suspect it is
  because the second verb doesn't know to combine with the subject of
  the entire sentence.

When I parse "Tu manges le biscuit et penses" with your grammar,
I get 12 parses.  Part of that is a problem with the interaction between
subejct verb inversion and object drop.  (The inversion lexical rule should
say [OPT -] on the inverted subject ... I need to fix that in the customization
system, but you can fix it in your grammar, too.)  Another part is the
same optionality of the pronoun-no-drop rule above.

  So, unless I have a means to categorize these based on where they
  can go, it looks like they can go either before or after the verb,
  with the risk of altering the semantics of the sentence.

Yes, the feature [POSTHEAD bool] lets you specify where the adjectives
show up.  If you have one that means different things in different positions,
then you really have homophones, with different PRED and different POSTHEAD
values.

You should also create some ungrammatical examples with the wrong
position for the wrong adjective (e.g., la fille grande).

  I'm not sure how I'd go about finding ungrammatical examples other
  than the usual lack of agreement ones.  The placement of the outer
  clause looks to be able to go at the beginning or at the end of the
  sentence (Tu as mangé tous les biscuits, je pense! Avez-vous mangé
  tous les biscuits, je te demande!) however, this may be more a
  marker of conversational French.

They *are* marked, by que/si.  You can make ungrammatical examples
by leaving these out, or putting them at the wrong end of the embedded
clause, or putting them both there, or giving the wrong one for the wrong
verb.

Ignoring the post-posed ones (Tu as mangé tous les biscuits, je pense) for
now is fine.

#292 without the copula.  Becomes an adjectival phrase.
Source: author
Vetted: f
Judgment: u
Phenom: cop
les garçons dans le parc 
les     garçon.s      dans    le           parc
the.3PL boys.3PL-masc in.PREP the.3SG-masc park.3SG-masc
'the boys are in the park' 

There's no adjectival phrase here.  The whole example does have
an analysis as an NP...

  * Information Structure - Topic and Focus are marked with the
    construction 'C'est' + NP which is usually at the beginning of the
    sentence.  Animate pronouns are then put into the accusative case
    (followed by qui, or who); inanimate pronouns are not (since there
    is no accusative case for them), and are followed by 'que'.

I'd be surprised if that were used for both topic and focus.  Knud Lambrecht
has looked deeply into the information structure of French, if you want
to get more details.

  * Embedded Clauses - In French, these are marked with verbs that
    require a prepositional complement.  As of yet, there is no means
    to enter these prepositional complements.  Sample sentences are:

That's a CP complement, not a PP complement.

  Of the 173 positive items, 26 failed to parse, because they had
  adjectives, adverbs, c'est, or the demonstrative suffixes that I
  haven't implemented, so, of 147 parsable items, I got 290 parses.

This is relatively high ambiguity, and you should look into what's causing
it (aside from the issues I noted above).

  Of the 123 Negative items, I have 16 parses.  Of those 16, 12 are
  past perfective examples.  I think these parsed because in the
  auxiliary verb section, where I specified each per-num of the verb
  avoir as its own aux slot, there was no place to specify features
  such as person and number; consequently, my examples that show
  dischord between the subject and the verb parse.

Indeed --- we meant to add this, and haven't yet.  You'll be able
to do it in the tdl, though.

  The other 2 examples (2 parses each) 1 was a conjuncted sentence
  where the subject was a conjunction and the verb didn't pluralize

"conjuncted" isn't a word.  I'd say you're talking about a sentence with
a conjoined subject.

