Changes to the grammar:

1. Modification
The head-adj-int and adj-head-int rules were already in the rules.tdl. When parsing a sentence without a modifier, neither rule was fired. Upon checking the russian.tdl, we found that the line +nvcdmo :+ [ MOD <>] was already there.
a. Adjectives
We added type adjective-lex in russian.tdl:
adjective-lex := basic-adjective-lex & intersective-mod-lex &
	      norm-ltop-lex-item &
  [ SYNSEM [ LOCAL [ CAT [ HEAD.MOD < [ LOCAL.CAT [ HEAD noun, VAL.SPR cons ]]>,
			     VAL [ SPR < >,
				 SUBJ < >,
				 COMPS < >,
				 SPEC < > ],
			   POSTHEAD - ]]]].
In lexicon.tdl, we added the following instance of an adjective (uninflected and just the stem):
boljsh := adjective-lex &
  [ STEM < "boljsh" >,
    SYNSEM.LKEYS.KEYREL.PRED "_large_a_rel" ].
IGT Example:
Jeti                devochki           chitajut         boljshimi               knigi.   
Jet-i               devochk-i          chita-ju-t       boljsh-imi               knig-i 
These-FEM.3PL.NOM   girl-FEM.3PL.NOM   read-3PL-PRS     big-FEM.3PL.INST        book-FEM.3PL.ACC
These girls are reading large books.  
We then parsed a sentence Ja budu chitatj boljsh knigu (I will be reading a large book) and observed that the index for the _large_n_rel of the adjective was tagged the same as the _book_n_rel predication of the noun kniga as well as the ARG2 of the predication _read_v_rel. we did not create any lexical rules for boljsh and just used the stem as an example.
b. Adverbs
We added type adverb-lex in Russian.tdl:
adverb-lex := basic-adverb-lex & intersective-mod-lex &
  [ SYNSEM [ LOCAL [ CAT [ HEAD.MOD < [ LOCAL.CAT.HEAD verb ]>,
			   VAL [ SPR < >,
				 SUBJ < >,
				 COMPS < >,
				 SPEC < > ]]]]].
In lexicon.tdl we added the following instance of an adverb:
bystro := adverb-lex &
  [ STEM < "bystro" >,
    SYNSEM.LKEYS.KEYREL.PRED "_quickly_v_rel" ].
IGT example:
Samoljot bystro letit.
Samoljot bystro let-it
Plane.NOM quickly fly-3SG.PRS
The plane is flying quickly.
We then parsed three sentences (since in Russian adverbs can appear anywhere in the sentence):
Ivan      bystro   letit        samoljotom.  
Ivan      samoljotom        letit        bystro.  
Ivan      bystro       samoljotom        letit.
Ivan is flying quickly by the plane.
All three parsed correctly. Looking at the MRSs,  the ARG0 of the adverb is coindexed with the ARG0 of the flying predication of the main verb, which is the INDEX value for the VP. 
2. Adjective agreement
First, we created another adjective type inheriting from adjective-lexeme to cater for different declensions (one being worked on first is for a subset of Russian adjectives called hard and for a further subset for those hard adjectives ending in a specific list of consonants (we named it 1 rather than list all of the endings):
hard-adjective-1-lex := adjective-lex.
Next, we looked at where the agreement happens for Russian adjectives and nouns and that is in case, number, gender and animacy (in the case of masc sg and plural accusative.
We created lexical rules for three cases: nom, acc and ins for masc, fem and neut single and plural.
Example:
hard-adjective-1-masc-nom-lex-rule := infl-ltow-rule & add-only-no-ccont-rule &
  [ DTR hard-adjective-1-lex ] &
	[ SYNSEM.LOCAL [ CAT.HEAD.MOD.FIRST.LOCAL.CAT.HEAD.CASE nom,
                   CONT.HOOK.INDEX.PNG [ GEND masc,
                                         NUM sg,
      	 	            PER 3rd ] ] ].
Features GEND, NUM and PER were applied directly to the adjective, while the feature CASE (appropriate for nouns only) was applied to the first thing on the MOD list of the adjective.
Finally, we added instances or the rules with the appropriate suffixes in the irules.tdl:
hard-adjective-1-masc-nom :=
%suffix (* oj)
hard-adjective-1-masc-nom-lex-rule.
IGT example:
Jeta              devochka           chitaet        boljshuju         knigu
Jet-a             devochk-a          chita-e-t      boljshu-ju        knig-u 
This-FEM.3SG.NOM  girl-FEM.3SG.NOM   read-3SG-PRS   big-FEM.3SG.ACC   book-FEM.3SG.ACC
This girl is reading a large book.
3. Demonstratives and markers of definiteness
a. Overt personal pronouns
There are overt pronouns in Russian, so we constrained their INDEX values to be [COG-ST activ-or-more, SPECI +] in russian.tdl for all pronouns, as in the example below:
pronoun-3rd-sg-masc-nom-noun-lex := no-spr-noun-lex &
  [ SYNSEM.LOCAL [ CONT.HOOK.INDEX [ PNG [ 
                                           NUM sg,
                                           PER 3rd,
                                           GEND masc ],
		        COG-ST activ-or-more,
	    	               SPECI +],
                                     CAT.HEAD.CASE nom ] ].
There are no definiteness determiners or nominal inflections associated with discourse status, so we didn't have to model that.
b. Demonstratives
Since we are treating the demonstrative relations as adjectival relations, no matter how they are introduced (via pronouns, determiners, or quantifiers), this is what we have (based on one of the two types of the three-way distinctions introduced in Lab 5): 
"Jetot" is a proximal pronoun, which points at something that is near (both to the speaker and hearer), 
"Tot" is a distal pronoun, pointing to something less near (both to the speaker and the hearer)
proximal+dem_a_rel := demonstrative_a_rel. ; close to speaker
distal+dem_a_rel := demonstrative_a_rel.   ; away from speaker
c. Demonstrative determiners
We added the determiner-lex-supertype to matrix.tdl according to lab instructions.
We then created two subtypes in russian.tdl:
non-demonstrative-determiner-lex := single-rel-lex-item & determiner-lex-supertype.
demonstrative-determiner-lex := determiner-lex-supertype &
  [ SYNSEM [ LOCAL [ CAT.VAL.SPEC.FIRST.LOCAL.CONT.HOOK [ INDEX #ind,
                                                                                                                 LTOP #ltop ],
                                    CONT.RELS.LIST.REST #altkeyrel ],
              LKEYS.ALTKEYREL #altkeyrel & [ ARG1 #ind, LBL #ltop ]]].
IGT example: 
Jeta              devochka           chitaet        boljshuju         knigu
Jet-a             devochk-a          chita-e-t      boljshu-ju        knig-u 
This-FEM.3SG.NOM  girl-FEM.3SG.NOM   read-3SG-PRS   big-FEM.3SG.ACC   book-FEM.3SG.ACC
This girl is reading a large book.
When parsing this sentence we can observer that the ARG0 of _girl_n_rel is identified with the ARG0 of the quant_q-rel and ARG1 of the proximal+dem_a_rel and ARG1 of the _read_v_rel. We can also observe that devochka has active+fam as the value of its COG-ST feature.
4. Argument optionality
For subject dropping, the customization already assumes that the dropped subjects have the OPT-CS value in-foc.
For object dropping we created a transitive verb subtype constraining the first item on its COMPS list to OPT-CS in-foc:
transitive-verb-lex-opt-dni := transitive-verb-lex &
	[ SYNSEM.LOCAL.CAT.VAL.COMPS.FIRST.OPT-CS in-foc ].
Next, we changed a relevant verb subtype created by the customization to inherit from this new subtype:
tr-perf-nom-acc-add-i-verb-lex := transitive-verb-lex-opt-dni & past-regular-rule-dtr & suffix-add-i-rule-dtr & infinitive-rule-dtr & imperfective-prefix-po-rule-dtr & present-1-conj-2-rule-dtr & nom-acc-transitive-verb-lex &
  [ INFLECTED -,
    SYNSEM.LOCAL.CONT.HOOK.INDEX.E.ASPECT perfective ].
When we parse a sentence:
Ja kupil
I bought
We can observe in the MRS that the ARG2 of the _buy_v_rel has in-foc as the value for its COG-ST.
5. Grammar coverage
Analyze | Coverage: 62.6%
Analyze | Overgeneration: 5.3%
