=== Clean up & Refinement === 

* I had my transitive être specifying its PRED value in lexicon.tdl.  Pulled it out: 

être-transitive-verb := transitive-verb-lex &
  [ SYNSEM.LKEYS.KEYREL.PRED "_be_v_id_rel" ].

être_1SG-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                PER 1st ] ] .
être_2SG-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                PER 2nd ] ].
être_3SG-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                PER 3rd ] ].
être_1PL-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                               PER 1st ] ].
être_2PL-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                                PER 2nd ] ].
être_3PL-verb-lex := être-transitive-verb &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                                PER 3rd ] ].

;; LER Takes a NP Comp
suis_trans := être_1SG-verb-lex &
  [ STEM < "suis" > ].

es_trans := être_2SG-verb-lex &
  [ STEM < "es" > ]. 

est_trans := être_3SG-verb-lex &
  [ STEM < "est" > ]. 

sommes_trans := être_1PL-verb-lex &
  [ STEM < "sommes" > ]. 

êtes_trans := être_2PL-verb-lex &
  [ STEM < "êtes" > ]. 

sont_trans := être_3PL-verb-lex &
  [ STEM < "sont" > ]. 

And checked that it was still in the MRS:

#316 - NP predicate
Source: author
Vetted: f
Judgment: g
Phenom: cop
Le garçon est un enfant
le           garçon       est    un    enfant
the.3SG-masc boy.3SG-masc is.3SG a.3SG child.3SG-masc 
'the boy is a child' 

* Fixed the semantics of my embedded clauses by mapping the VAL list to the ARG-ST: 

embedded-clause-verb-lex := verb-lex & clausal-second-arg-trans-lex-item & norm-sem-lex-item &  
  [ SYNSEM.LOCAL [ CAT.VAL [ SUBJ < #subj > ,
                             COMPS < #comps &  
                                      [ LOCAL [ CAT.HEAD comp,
                                                CONT.HOOK.INDEX.SF prop-or-ques ] ] > ]] ,
    ARG-ST < #subj, #comps > ].  

And now these sentences parse with the correct MRSs:

#304 
Source: author
Vetted: f
Judgment: g
Phenom: emb-q, tam 
La fille demande si vous avez mangé les biscuits 
le fille demand.e si vous avez mang.é les biscuits
the.3SG-fem girl.3SG-fem wonder.3SG if you.2PL have.2PL eat.PST-PTCP the.3PL cookies.3PL 
'the girl wodners if you have eaten the cookies' 

#305
Source: author
Vetted: f
Judgment: g
Phenom: emb-d, tam 
Je pensais que tu manges lentement
je    pens.ais           que  tu      mang.es lentement
I.1SG think.1SG-PST-IPFV that you.2SG eat.2SG slow.ADV
'I used to think that you eat slowly.' 

* My predicative adjectives had the wrong semantics, and weren't enforcing SUBJ - ADJ agreement.  Fixed this:

adjective-lex := basic-adjective-lex & intersective-mod-lex &
   [ INFLECTED -,
     SYNSEM [ LKEYS.KEYREL.ARG1 #xarg,
              LOCAL [ CONT.HOOK.XARG #xarg,
                      CAT [ HEAD.MOD < [ LOCAL [ CONT.HOOK.INDEX #xarg,
                                                 CAT [ HEAD noun,
                                                       VAL.SPR cons ] ] ] >,
                            VAL [ SPR < >,
                                  SUBJ < >,
                                  COMPS < >,
                                  SPEC < > ] ] ] ] ] .

intersective-mod-lex := norm-hook-lex-item & no-hcons-lex-item &
  [ SYNSEM [ LOCAL.CAT.HEAD.MOD < [ LOCAL intersective-mod ] > ] ].

And checked the MRSs: 

#318 - Adjectival Predicate
Source: author
Vetted: f
Judgment: g
Phenom: cop 
La fille est petite
la          fille        est    petit.e
the.3SG-fem girl.3SG-fem is.3SG small.3SG-fem 
'the girl is little' 

And that the agreement is being preserved: 

#289 Dischord between adj and noun 
Source: author
Vetted: f
Judgment: u
Phenom: adj, agr
*les grand enfants mangent 
les      grand. enfant.s    mang.ent 
the.3PL big.SG  children.PL eat.3PL 
'the big children eat' 


* I realized that none of my forms of etre had any tense!  

Fixed in my etre that took a NP COMP: 

être-transitive-verb := transitive-verb-lex &
  [ SYNSEM [ LKEYS.KEYREL.PRED "_be_v_id_rel",
             LOCAL.CONT.HOOK.INDEX.E.TENSE present ] ].

and checked it: 

#316 - NP predicate
Source: author
Vetted: f
Judgment: g
Phenom: cop
Le garçon est un enfant
le           garçon       est    un    enfant
the.3SG-masc boy.3SG-masc is.3SG a.3SG child.3SG-masc 
'the boy is a child' 

And my copular être:

adposition-taking-être := copula-verb-lex &
  [ SYNSEM.LOCAL.CONT.HOOK.INDEX.E.TENSE present ].

adp_être_1SG-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                    PER 1st ] ].

adp_être_2SG-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                PER 2nd ] ].

adp_être_3SG-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM singular,
                                                                PER 3rd ] ].

adp_être_1PL-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                                PER 1st ] ].

adp_être_2PL-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                                PER 2nd ] ].

adp_être_3PL-verb-lex := adposition-taking-être &
  [ SYNSEM.LOCAL.CAT.VAL.SUBJ.FIRST.LOCAL.CONT.HOOK.INDEX.PNG [ NUM plural,
                                                                PER 3rd ] ].

And checked them:

#318 - Adjectival Predicate
Source: author
Vetted: f
Judgment: g
Phenom: cop 
La fille est petite
la          fille        est    petit.e
the.3SG-fem girl.3SG-fem is.3SG small.3SG-fem 
'the girl is little' 

#320 - Prepositional Phrase Predicate
Source: author
Vetted: f
Judgment: g
Phenom: cop
Les garçons sont dans le parc 
les     garçons       sont   dans    le           parc
the.3PL boys.3PL-masc be.3PL in.PREP the.3SG-masc park.3SG-masc
'the boys are in the park' 

Added ASPECT to my past tenses:

past_imperfect := past_imperfect_endings-lex-rule &
  [ SYNSEM.LOCAL.CONT.HOOK.INDEX.E [ TENSE past,
                                     ASPECT imperfective ] ] .

#181
Source: author
Vetted: f
Judgment: g
Phenom: tam
Je mangeais le biscuit
Je    mange.ais        le      biscuit
I.1SG eat.1SG-PST-IPFV the.3SG cookie.3SG
'I was eating the cookie' 


Not entirely sure what, in the complex past tense should carry the tense/aspect marker.  I guess the auxilliary in the present + past participle is enough?  Looking at my conjugation tables, I see the following:
aux in present + past participle = passe compose
aux in imparfait + past participle = plus-que-parfait 
aux in the passe simple + past participle = passe anterieur
aux in the futur simple + past participle = futur anterieur 

Based on this, I put the tense/aspect markers on the auxiliary verb present tense conjugation:

avoir-aux-lex := inanimate_obj_pronoun-rule-dtr & negation_prefix-rule-dtr & arg-comp-aux-no-pred &
   [ SYNSEM.LOCAL.CAT.VAL [ COMPS.FIRST.LOCAL [ CONT.HOOK.INDEX.E [ TENSE past,
                                                                    ASPECT perfective ],
                                                CAT.HEAD.FORM past_participle ] ] ].


#192
Source: author
Vetted: f
Judgment: g
Phenom: tam
la fille a mangé les biscuits 
la          fille        a               mang.é       les    biscuits 
the.3SG-fem girl.3SG-fem has.3SG-PST-PFV eat.PST-PTCP the.3PL cookies.3PL-masc
'the girl has eaten the cookies' / 'the girl ate the cookies' 

* My strings of prefixes weren't appending properly, so negated sentences with a dropped object marker weren't parsing.  Fixed this by making a ne-rule-dtr:

ne-rule-dtr := word-or-lexrule. 

And making it the daughter of my negation prefix rule:

negation_prefix-lex-rule := infl-ltol-rule & same-ctxt-lex-rule & same-agr-lex-rule & same-head-lex-rule & same-hc-light-lex-rule & same-posthead-lex-rule & same-mc-lex-rule & local-change-only-lex-rule &
  [ DTR ne-rule-dtr ].

And a supertype of all the inflections that can negate:

er_verb_conjugation-lex-rule := inanimate_obj_pronoun-rule-dtr & negation_prefix-rule-dtr & infl-ltow-rule & add-only-no-ccont-rule & ne-rule-dtr & 
  [ SYNSEM.LOCAL.CAT.HEAD.FORM finite ] & 
  [ DTR er_verb_conjugation-rule-dtr ].

reflexive_verb_prefixes_conjugation-lex-rule := inanimate_obj_pronoun-rule-dtr & negation_prefix-rule-dtr & infl-ltow-rule & add-only-no-ccont-rule & ne-rule-dtr & 
  [ DTR er_verb_conjugation-rule-dtr ].

future_perfect_endings-lex-rule := inanimate_obj_pronoun-rule-dtr & negation_prefix-rule-dtr & infl-ltow-rule & add-only-no-ccont-rule & ne-rule-dtr & 
  [ DTR future_perfect_endings-rule-dtr ].

past_imperfect_endings-lex-rule := inanimate_obj_pronoun-rule-dtr & infl-ltow-rule & add-only-no-ccont-rule & ne-rule-dtr & 
   [ DTR past_imperfect_endings-rule-dtr ] .     
 
inanimate_dropped_object_marker := infl-ltol-rule & inanimate_obj_pronoun-lex-rule & ne-rule-dtr & 
  [ SYNSEM.LOCAL.CAT.VAL.COMPS < [ OPT +, OPT-CS in-foc ] > ]. 

As of right now, these don't append onto my tense-marking auxiliaries, so sentences such as "je le-ai mangé" and "je ne-la-ai pas mangé pas" don't parse.  I tried interactive unification but nothing jumped out at it.  I'll tackle it with a fresh brain next week.  

#247 Future Perfect
Source: author
Vetted: f
Judgment: g
Phenom: tam, neg, pro-d
Vous ne-les-mangerez pas.
Vous    ne-les-mang.erez       pas.
you.2PL no-it-eat.3PL-FUT-PFV not
'you will not eat it'

#248 Past Imperfect
Source: author
Vetted: f
Judgment: g
Phenom: tam, neg, pro-d
Je ne-la-mangeais pas
Je    ne-la-mang.eais         pas
I.1SG not-it-eat.1SG-PST-IPVF not
'I was not eating it' 

=== Information Structure ===

== Topic ==

Topic is not overtly marked in French.  The subject of the sentence is generally the topic.  If the topic is not in focus, an overt NP is used.  If it is considered in focus, then a pronoun referent it used.  This is reflected in COG-ST in the existing grammar.  

== Focus == 

Focus is marked with a cleft construction, C'est + acc + qui (if the focus is animate) and C'est + acc + que (if the focus is inanimate).  I did search for other means of focus marking, and found several articles in _Romance Languages and Linguistic Theory 2000_ that dealt with the topic of prosody and focus in French, and even those dealt with the prosody of these c'est constructions.  I've asked my sources if prosody alone can be used to mark focus, as it can be in English.  Without thinking too hard, they don't think it can.  

=== VPM === 

I added the file semi.vpm to my directory and added the following lines:

PNG.PN : PERS NUM
1sg < > 1 sg
1pl < > 1 pl
2sg < > 2 sg
2pl < > 2 pl
3sg < > 3 sg
3pl < > 3 pl 
1per << 1 *
1per <> 1 !
2per << 2 *
2per <> 2 !
3per << 3 *
3per <> 3 ! 
*    >> ! !
!    << * * 

E.ASPECT : ASPECT
  aspect <> aspect
  imperfective <> imp 
  pefective <> perf
  aspect >> aspect
  no-aspect << aspect 



=== Test Suite Assessment ===

== Baseline ==

# of positive items: 229
# of items parsing: 214  93.4% 
    Reasons for non parses:
         5 instances of -ci and là noun suffixes (not implemented)
         6 instances of strings of prefixes 
         2 - C'est moi, qui.. 
         2 - prendre 
Total # of parses: 355
     Reasons for spurious ambiguity:
         negated sentences with overt object parse twice 
         conjoined sentences, or sentences with conjoined verbs parse twice
         sentences with a copular être + adposition parse 6 times.  These sentences in an embedded clause parse 30 times. 

# negative items: 152
# of items parsing: 4   2.63# 
     Reasons for parses: 
        2 instances of ne without pas, or pas without ne
        1 instance  conjoined subjects not raising the number of the verb 
        1 instance of the adverb being after the past participle, instead of the verb

== Lab 8 Test Suite == 

# of positive items: 229 
# of items parsing: 218  95.2% 
         5 instances of -ci and là noun suffixes (not implemented)
         2 instances of ne-la+aux  
         2 - C'est moi, qui.. 
         2 - prendre 
Total # of parses: 393 
    Reasons for spurious ambiguity: 
        Conjoined verbs/sentences # 276, 278, 279 in test suite 
        Je mange et mange et mange - 68 (!!) 
        Embedded declarative clauses with an overt object are parsing twice (Je pense que elle mange les biscuits) 
        Je pensais que to manges lentement - 4 times - 2 thanks to the je and the bare-np rule, and 2 because of manges and the basic-head-opt-comp 
        Il a pensé que elle mange une pomme - 5 parses - il a is being considered an S that takes an S comp in 4 of those cases
        Les garçons sont dans le parc - 6 parses 
        Embedded être copula sentences are parsing 30 times (je pense qu / je demande si les garçons sont dans le parc)
         Modal sentences without an object are parsing twice (il peut manger) 
        
=== Generation === 

The inv-lr is still keeping me from generating.  If it's commented not out (which it currently isn't), I am getting edge errors or lack of heap errors.  I've checked my heap size, etc, and this doesn't seem to be an issue.  I wonder if CJ does housecalls.. 
