# Assignment: Lab 4, Continued Matrix Grammar Development
# Grammar: Yiddish, Chopra and Letcher

# Phenomena

## Coordination

The sort of phrasal combination that is accomplished by using "and" in English can be achieved in Yiddish using the conjunctive "un". Conjuncts of "un" can be sentences or noun phrases.  

The other marking strategy Yiddish uses to indicate coordination is the use of double conjunctives to signal intensification of linking between conjuncts. Double conjunctives serve a correlative function in Yiddish, similar to the use of "both...and" in English.
	
The double conjuncts -- "i...i", "hen...hen", "say...say" -- differ in terms of generality. The first, "i....i", is the most general, occurring with S, NO, and VP. The other double conjuncts "hen...hen" and "say..say" are used less often with sentences -- though this is a grammatical application of both -- but otherwise are very similar in terms of usage to "i...i".

When using the single conjunctive "un" only the last conjunct must be marked in the coordinate structure. However when using the double conjunctives, both conjuncts must be marked.

We decided to model double conjunctives as omnisyndetic; it is not possible to constrain a coordination strategy as only being applicable to two coordinands in the matrix, thus our implementation overgenerates. Our model for the single conjunctive "un" as a polysyndeton coordination strategy corresponds with its usage in Yiddish and should not be the source of any overgeneration.

		#@desc: Single conjunctive "un" coordinates two conjuncts.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {coordination}
		Ikh ze dos book un dos esn.
		Ikh ze dos book un dos es-n.
		1SG.NOM see;1SG 3SG.ACC.N book and 3SG.ACC.N food.
		I see the book and the food.	

		#@desc: Double conjunctive "i...i" correlates two conjuncts.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {coordination}
		Ikh ze i dos book i dos esn.
		Ikh ze i dos book i dos es-n.
		1SG.NOM see;1SG both 3SG.ACC.N book and 3SG.ACC.N food.
		I see the book and the food.

		#@desc: Double conjunctive "say...say" correlates two conjuncts.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {coordination}
		Ikh ze say yenem man say yenc bukh.
		Ikh ze say yenem man say yenc bukh.
		1SG.NOM see;1SG both DEM.ACC.3P.M man and DEM.ACC.N book.
		I see both that man and that book.

		#@desc: Double conjunctive "hen...hen" correlates two conjuncts.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {coordination}
		Ikh ze hen dos bukh hen dos esn.
		Ikh ze hen dos bukh hen dos esn.
		1SG.NOM see;1SG both 3SG.ACC.N book and 3SG.ACC.N food.
		I see both the book and the food.

		#@desc: Incorrect; double conjunctives requires an markers before each cojunct, not after.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {coordination}
		Ikh ze dos bukh i dos esn i.
		Ikh ze dos bukh i dos esn i.
		1SG.NOM see;1SG 3SG.ACC.N book both 3SG.ACC.N food and.
		I see both the book and the food.

		#@desc: Incorrect; double conjunctives requires marking both cojuncts, not just the last.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {coordination}
		Ikh ze dos bukh i dos esn.
		Ikh ze dos bukh i dos esn.
		1SG.NOM see;1SG 3SG.ACC.N book and 3SG.ACC.N food.
		I see both the book and the food.

		#@desc: Incorrect; double conjunctives requires marking both cojuncts, not just the first.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {coordination}
		Ikh ze i dos bukh dos esn.
		Ikh ze i dos bukh dos esn.
		1SG.NOM see;1SG both 3SG.ACC.N book 3SG.ACC.N food.
		I see both the book and the food.

		#@desc: Incorrect, because constituents cannot be coordianted via juxtaposition in Yiddish.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {coordination}
		Ikh ze dos bukh dos esn.
		Ikh ze dos bukh dos esn.
		1SG.NOM see;1SG 3SG.ACC.N book 3SG.ACC.N food.
		I see the book and the food.
		
## Non-Verbal Predicates

In Yiddish NPs, adjectives and prepositional phrases can be used to express predicates. The primary copula in Yiddish is the copular verb "zayn"; it is used to connect a subject to its predicate in a manner similar to the English "is".

A copula is obligatorily required in order to connect a subject to a noun phrase, prepositional phrase or adjective and this constraint does not vary with tense. It is worth noting that adjectives when used predicatively remain uninflected.

		#@desc: The copular verb 'iz' connects a subject to an adjectival predicate.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {cop}
		Der man iz gut.
		Der man iz gut.
		DEF.SG.NOM.M man COP good.

		#@desc: The copular verb 'iz' connects a subject to an NP predicate.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {cop}
		Der man iz a held.
		Der man iz a held.
		DEF.SG.NOM.M man COP INDF.ART hero.
		The man is a hero.

		#@desc: The copular verb 'iz' connects a subject to an prepositional phrase predicate.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {cop}
		Der man iz unter dos hoyz.
		Der man iz unter dos hoyz.
		DEF.SG.NOM.M man COP under DEF.SG.ACC.N house.
		The man is under the house.

		#@desc: Incorrect; adjectives used predicatively remain uninflected.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {cop}
		Der man iz guter.
		Der man iz guter.
		DEF.SG.NOM.M man COP good.
		The man is good.

		#@desc: Incorrect; a copula is required to connect a subject to an adjectival predicate.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {cop}
		Der man gut.
		Der man gut.
		DEF.SG.NOM.M man COP good.
		The man is good.

		#@desc: Incorrect; a copula is required to connect a subject to a noun phrase predicate.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {cop}
		Der man a held.
		Der man a held.
		DEF.SG.NOM.M man COP INDF.ART hero.
		The man is INDF.ART hero.

		#@desc: Incorrect; a copula is required to connect a subject to a preopositional phrase predicate.
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {cop}
		Der man unter dos hoyz.
		Der man unter dos hoyz.
		DEF.SG.NOM.M man COP under DEF.SG.NOM.N house.
		The man is under the house.


## Information Structure

Our primary descriptive grammar for Yiddish (Jacobs 2005) covers leftward movements in Yiddish -- which it terms 'topicalization' though the constituent being moved to the left is not being made the clause topic -- but it does not distinguish between topics and foci (Jacobs mentions that this has been omitted to conserve space). We consulted analyses by Diesing and Prince in order to cover this phenomenon in the testsuite.

However, according to Diesing's analysis, Yiddish has V2 word order with unmarked word order being SVO: Any constituent can be topicalized by being moved to the preverbal position (or forfelt in Germanic V2 terminology). 

The leftward movement of an object marks a shift of focus. The literature is divided as to whether leftward movement can also mark topic; if this happens, it's in a very small number of cases in which psych-verbs are involved. Furthermore, non-counting constituents (such as negative markers) can precede the topicalized constituent. Following Sounghoun's suggestions, we tried to implement this in our matrix grammar selecting  clause-initial position for focus, sentence-initial position for topic, and preverbal position for contrastive focus.

Diesing's analysis of Yiddish word order also makes finer distinctions about how word order within the VP can mark contrastive focus. However this is only possible in complex verb phrases because they allow for OV order in the verb phrase (with the auxiliary being in the second position). We included what we learned from Deising's analysis in our testsuite but it was not suited to implementation in the matrix.

The contrastive dimension becomes relevant when the main verb is freed from being in second position because the auxiliary fills that spot, thus allowing more movement variations that carry semantic content relevant to information structure. Objects that are in contrastive focus appear to the immediate-left of verbs; this holds true for both indefinite and definite NPs.

With respect to information structure as it is indicated by word order in Yiddish, it appears to be obvious that different orderings simply express different information structural semantics, so we have not included any negative examples since they aren't distinct from ungrammatical examples of word order in the corresponding section of the testsuite.

Since the unmarked order in declarative main clauses is SVO, Yiddish provides a way to put the subject into focus by use of a dummy "es" which corresponds to the English "it"; by placing "es" in the initial subject position, one prevents other constituents from occupying that position and being perceived as the topic.


__[TODO: Add the "es" example.]__

		#@desc: Yiddish default unmarked focus/topic word order.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Der man leyent dos bukh.
		Der man leyen-t dos bukh.
		DEF.NOM.M man read-1SG DEF.ACC.N book.
		The man reads the book.

		#@desc: Intranstive verb OVS verb order.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Dos bukh leyent der man.
		Dos bukh leyen-t der man.
		DEF.ACC.N book read-1SG DEF.NOM.M man.
		It's the book that the man is reading.

		#@desc: Non-counting constituent precedes the topicalized constituent (subject).
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Neyn, ikh trink nit dem vayn.
		Neyn, ikh trink nit dem vayn.
		No, I didn't drink that wine.
		No, I didn't drink that wine.

		#@desc: Non-counting constituents precedes the topicalized constituent (direct object).
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Neyn, dem vayn trink ikh nit.
		Neyn, dem vayn trink ikh nit.
		PART.NEG DEF.DAT.M wine drink;1SG 1SG.NOM NEG.
		That wine isn't drunk by me.

		#@desc: definite article used to specify an object in a VP with head-final order.
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Der man hot dos bukh nekhtn gelyent.
		Der man hot dos bukh nekhtn ge-lyen-t.
		DEF.ART man have;3SG DEF.ART book yesterday PST.PTCP-read-PST.PTCP.
		The man has the book yesterday read.

		#@desc: head-final order providing contrastive emphasis (indefinite article specifying object).
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Der man hot nekhtn a bukh gelyent.
		Der man hot nekhtn a bukh ge-lyen-t.
		DEF.ART man have;3SG yesterday INDF.ART book PST.PTCP-read-PST.PTCP.
		The man read a book yesterday.

		#@desc: head-final order providing contrastive emphasis (definite article specifying object).
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {info}
		Der man hot nekhtn dos bukh geleyent.
		Der man hot nekhtn dos bukh ge-leyen-t.
		DEF.ART man yesterday the book read.
		The man read the book yesterday.

## Matrix Yes-No Questions

In Yiddish, matrix yes/no questions are formed a number of ways. Firstly, verb initial word order is used to indicate matrix yes-no questions. This is illustrated below:

		Source: a:229
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Farshteyt er shpanis?
		Farshtey-t er shpanis?
		understand-2SG 2SG.NOM Spanish?
		Does he understand Spanish?

		Source: a:228
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Zaynen zey meshug-e?
		be-3PL 3PL.NOM crazy-NOM.N?
		Are they crazy?

This could be analyzed as either as movement to VSO word order or subject verb inversion. In order to determine which, a complex verb phrase involving an auxiliary is required. Our primary reference grammar does not provide such an example, however we found one such example elsewhere, in which only the auxiliary and not the main verb moves to the initial position, telling us that this should probably be analyzed as subject-verb inversion:

		Source: a:229
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Farshteyt er shpanis?
		Farshtey-t er shpanis?
		understand-2SG 2SG.NOM Spanish?
		Does he understand Spanish?

		Source: a:228
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Zaynen zey meshug-e?
		be-3PL 3PL.NOM crazy-NOM.N?
		Are they crazy?

The optional marker "tsi" can also be used preceding the verb in yes-no interrogative clauses:

		Source: a:229
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Tsi farshteyt er shpanis?
		Tsi farshtey-t er shpanis?
		whether understand-2SG 2SG.NOM Spanish?
		Does he understand Spanish?

		# incorrect position of tsi
		Source: author based on a:229
		Vetted: f
		Judgment: u
		Phenomena: {Matrix yes-no questions, word order}
		Farshteyt tsi er shpanis?
		Farshtey-t tsi er shpanis?
		understand-2SG whether 2SG.NOM Spanish?
		Does he understand Spanish?

		# incorrect position of tsi
		Source: author based on a:229
		Vetted: f
		Judgment: u
		Phenomena: {Matrix yes-no questions, word order}
		Farshteyt er tsi shpanis?
		Farshtey-t er tsi shpanis?
		understand-2SG 2SG.NOM whether Spanish?
		Does he understand Spanish?

		# incorrect position of tsi
		Source: author based on a:229
		Vetted: f
		Judgment: u
		Phenomena: {Matrix yes-no questions, word order}
		Farshteyt er shpanis tsi?
		Farshtey-t er shpanis tsi?
		understand-2SG 2SG.NOM Spanish whether?
		Does he understand Spanish?

Topicalization from standard yes-no interrogative word order (VSO) is possible, in which case use of "tsi" is not possible, and the construction is accompanied by rising intonation:

		# topicalization of subject from yes-no interrogative word order
		Source: a:229
		Vetted: t
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Er farshteyt shpanis?
		Er farshtey-t shpanis?
		2SG.NOM understand-2SG Spanish?
		Does he understand Spanish?

		# topicalization of object from yes-no interrogative word order
		Source: author based on a:229
		Vetted: f
		Judgment: g
		Phenomena: {Matrix yes-no questions, word order}
		Shpanis farshteyt er?
		Shpanis farshtey-t er?
		Spanish understand-2SG 2SG.NOM?
		Does he understand Spanish?

		# tsi is not permitted in topicalized interrogative
		Source: author based on a:229
		Vetted: f
		Judgment: u
		Phenomena: {Matrix yes-no questions, word order}
		Tsi er farshteyt shpanis?
		Tsi er farshtey-t shpanis?
		Whether 2SG.NOM understand-2SG Spanish?
		Does he understand Spanish?

		# tsi is not permitted in topicalized interrogative
		Source: author based on a:229
		Vetted: f
		Judgment: u
		Phenomena: {Matrix yes-no questions, word order}
		Tsi Shpanis farshteyt er?
		Tsi Shpanis farshtey-t er?
		Whether Spanish understand-2SG 2SG.NOM?
		Does he understand Spanish?

		TODO check for grammaticality of these next two:

		# yes-no questions with negation
		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {Matrix yes-no questions, negation}
		Zingt nit der man?
		Zing-t nit der man?
		sing-1SG NEG DEF.NOM.M man.
		Isn't the man singing?

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {Matrix yes-no questions, negation}
		Tsi zingt nit der man?
		Tsi zing-t nit der man?
		whether sing-1SG NEG DEF.NOM.M man.
		Isn't the man singing?

As argued above, it seems like subject-verb inversion might be a good way of analyzing the word order of yes-no interrogatives, however experimentation with the selection of this option in the customization system resulted in a grammar that showed unfortunate side effects. This was that the grammar produced an additional (and spurious) parse for both intransitive and transitive declarative (SVO) sentences, with the INV feature of the verb being positive. Interestingly, these spurious parses are not converted into interrogatives, are not being picked up as interrogatives with the int-cl rule. 

We decided this negative interaction between the implementations of subject-verb inversion and V2 word order was not worth the gain in coverage of interrogative sentences. Instead we will wait until we are hand editing the TDL, and explore whether it is possible to constrain the grammar produced with subject-verb inversion turned on so as the two phenomena get along together, or whether we will need to completely implement this ourselves.  Since the optional clause initial "tsi" marker can only occur in the presence of VSO word order, and the different interrogative strategies available in the customization system are independent of each other, this option was ruled out for being used also.


## Adverbs

Yiddish adverbs are not subject to inflection. The base form of adjectives in Yiddish can be used as manner adverbs. Manner adverbs can also be formed with derivational affixes out of bases including adjectives, past participles, and nouns, however we will not address these. Manner adverbs in Yiddish appear in the final position of unmarked clauses, in which case they attach to the VP. This is illustrated below:

		Source: a:193
		Vetted: t
		Judgment: g
		Phenomena: {adverb}
		Dos kind zingt shayn.
		Dos kind zing-t shayn.
		DEF.ACC.N child sing-1SG nicely.
		The child sings nicely.

Adverbs can also be topicalized by being moved to the first position, resulting in them attaching to the sentence. This is illustrated below:

		Source: author based on a:193
		Vetted: f
		Judgment: g
		Phenomena: {adverb}
		Shayn dos kind zingt.
		Shayn dos kind zing-t.
		nicely DEF.ACC.N child sing-1SG.
		Nicely, the child sings.

The remaining possible positions (at least in a simple declarative  sentence such as this) are ungrammatical: 

		# adverb in wrong location
		Source: author based on a:223-224
		Vetted: t
		Judgment: u
		Phenomena: {adverb, word order}
		Der man shayn leyent dos bukh.
		Der man shayn leyen-t dos bukh.
		DEF.NOM.M man nicely read-1SG DEF.ACC.N book.
		The man reads the book nicely.

		# adverb in wrong location
		Source: a:223-224
		Vetted: t
		Judgment: u
		Phenomena: {adverb, word order}
		Der man leyent shayn dos bukh.
		Der man leyen-t shayn dos bukh.
		DEF.NOM.M man read-1SG nicely DEF.ACC.N book.
		The man reads the book nicely.

		Source: a:223-224
		Vetted: t
		Judgment: u
		Phenomena: {adverb, word order}
		Shayn der man leyent dos bukh.
		Shayn der man leyen-t dos bukh.
		nicely DEF.NOM.M man read-1SG DEF.ACC.N book.
		The man reads the book nicely.

Since manner adverbs are not yet handled by the Grammar Matrix, no implementation has of yet been done.


## Attributive Adjectives

Yiddish adjectives occur between the determiner and the noun and agree with the case, gender and number of the noun they modify. Agreement is marked via a suffix on the adjective. The following examples show correct and incorrect positioning of adjectives in the noun phrase.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives}
		Dos voyl kind zingt.
		Dos voyl-e kind zing-t.
		DEF.NOM.N nice-NOM.SG.N child sing-3SG.
		The nice child sings.

		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {adjectives}
		Dos kind voyle zingt.
		Dos kind voyl-e zing-t.
		DEF.NOM.N child nice-NOM.SG.N sing-3SG.
		The nice child sings.

		# The adjective is almost an adverb here, but shouldn't be inflected. 
		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {adjectives}
		Dos kind zingt voyle.
		Dos kind zing-t voyle-e.
		DEF.NOM.N nice good-NOM.SG.N sing-3SG.
		The child sings nice.

The following examples show adjectival agreement in the nominative case:

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Der guter man zing.
		Der gut-er man zing.
		DEF.SG.NOM.M good-NOM.SG.M man sing;1SG.
		The good man sings.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Di gute froy zingst.
		Di gut-e froy zing-st.
		DEF.SG.NOM.F good-NOM.SG.F woman sing-2P.
		The good woman sings.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Dos gute kind zingt.
		Dos gut-e kind zing-t.
		DEF.PL.NOM.N good-NOM.SG.N child sing-3SG.
		The good child sings.

The following examples show adjectival agreement in the accusative case:

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Ikh ze dos gutn man.
		Ikh ze dos gut-n man.
		1SG.NOM ze;1SG DEF.ACC.M good-ACC.SG.M man.
		I see the good man.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Ikh ze di gute froy.
		Ikh ze di gute-e froy.
		1SG.NOM see;1SG 3SG.ACC.F good-ACC.SG.F woman.
		I see the good woman.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Ikh ze dos gute kind.
		Ikh ze dos gut-e kind.
		1SG.NOM see;1SG 3SG.ACC.F good-ACC.SG.N kind.
		I see the good child.

The following examples show adjectival agreement in the dative case:

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {determiner, word order, agreement}
		Di mame git dem gutn man dos bukh.
		Di mame git dem gut-n man dos bukh.
		DEF.NOM.F mother give;1SG DEF.DAT.M good-DAT.SG.M man DEF.ACC.N book.
		The mother gives the good child the book.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {determiner, word order, agreement}
		Di mame git der guter froy dos bukh.
		Di mame git der gut-er kind dos bukh.
		DEF.NOM.F mother give;1SG DEF.DAT.F good-DAT.SG.F child DEF.ACC.N book.
		The mother gives the good child the book.

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {determiner, word order,agreement}
		Di mame git dem gutn kind dos bukh.
		Di mame git dem gut-n kind dos bukh.
		DEF.NOM.F mother give;1SG DEF.DAT.N good-DAT.SG.N child DEF.ACC.N book.
		The mother gives the good child the book.

The following examples show adjectival agreement with a plural noun. Plural Yiddish adjectives forms make no distinction between case and gender:

		Source: author
		Vetted: f
		Judgment: g
		Phenomena: {adjectives, agreement}
		Di gute kinder zing.
		Di gut-e kind-er zing.
		DEF.PL.NOM.M good-NOM.PL child-NOM.PL sing;1SG.
		The good children sing.

The following are a handful of ungrammatical permutations of the above:

		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {adjectives, agreement}
		Der gutn man zing.
		Der gut-n man zing.
		DEF.SG.NOM.M good-NOM/DAT.SG.M man sing;1SG.
		The good man sings.

		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {adjectives, agreement}
		Ikh ze di guter froy.
		Ikh ze di gute-er froy.
		1SG.NOM see;1SG 3SG.DAT.F good-ACC.SG.F woman.
		I see the good woman.

		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {determiner, word order,agreement}
		Di mame git dem gute kind dos bukh.
		Di mame git dem gut-e kind dos bukh.
		DEF.NOM.F mother give;1SG DEF.DAT.N good-NOM/ACC.SG.N child DEF.ACC.N book.
		The mother gives the good child the book.

		Source: author
		Vetted: f
		Judgment: u
		Phenomena: {adjectives, agreement}
		Di gutn kinder zing.
		Di gut-n kind-er zing.
		DEF.PL.NOM.M good-NOM.SG.N child-PL sing;1SG.
		The good children sing.

Since adjectives are not yet handled by the Grammar Matrix, no implementation has of yet been done.

## EMBEDDED CLAUSES

Subordinate clauses in Yiddish display the same V2 word order as
declarative main clauses and are introduced by the complementizer
'az', which, in some circumstances can be omitted:

		Source: a:233
		Vetted: t
		Judgment: g
		Phenomena: {embedded declarative}
		Zi meynt, az ikh bin meshuge.
		Zi meyn-t, az ikh bin meshuge.
		3SG.NOM.F think-3SG, COMP 1SG.NOM be;1SG crazy. 
		She thinks that I have gone crazy.

		Source: a:233
		Vetted: t
		Judgment: g
		Phenomena: {embedded declarative}
		Zi meynt, ikh bin meshuge.
		Zi meyn-t, ikh bin meshuge.
		3SG.NOM.F think-3SG, COMP 1SG.NOM be;1SG crazy. 
		She thinks I have gone crazy.

Additionally, constituents within the embedded clauses can be
topicalized:

		Source: a:232
		Vetted: t
		Judgment: g
		Phenomena: {embedded declarative}
		Ikh veys, az moyshe kumt morgn.
		Ikh veys, az moyshe kum-t morgn.
		1SG.NOM know;3SG COMP Moyshe come;3G tomorrow.
		I know that Moyshe is coming tomorrow.
 
		Source: a:232
		Vetted: t
		Judgment: g
		Phenomena: {embedded declarative}
		Ikh veys, az morgn moyshe kumt.
		Ikh veys, az morgn moyshe kum-t.
		1SG.NOM know;3G COMP tomorrow Moyshe come;3G.
		I know that it's tomorrow Moyshe is coming.

		Source: a:232
		Vetted: t
		Judgment: g
		Phenomena: {embedded declarative}
		Ikh veys, az es moyshe kumt morgn.
		Ikh veys, az es moyshe kum-t morgn.
		1SG.NOM know;3G COMP 1SG.NOM Moyshe come;3G tomorrow.
		I know that it's Moyshe that's coming tomorrow.

Since subordinate clauses are not yet handled by the Grammar Matrix,
no implementation has of yet been done.

While our reference grammar does not say anything about subordinate
yes-no interrogatives, from looking through the data in the Corpus of
Modern Yiddish it would appear that the interrogative marker 'tsi',
which translates as 'whether' is used as a complementizer instead of
'az' similar, to English. As to the word order in embedded
interrogatives, it appears that both SVO and VSO are possible, with
VSO having the effect of contributing more interrogative force to
the utterance. 

		Source: author based on examples in c
		Vetted: f
		Judgment: g
		Phenomena: {embedded interrogative}
		Ikh freg, tsi iz es efsher
		Ikh freg, tsi iz es efsher
		1SG.NOM ask;3SG, whether;COMP be;3SG 3SG.NOM.N possible.
		I ask whether it is possible?

		Source: author based on examples in c
		Vetted: f
		Judgment: g
		Phenomena: {embedded interrogative}
		Ikh freg, tsi es iz efsher
		Ikh freg, tsi es iz efsher
		1SG.NOM ask;3SG, whether;COMP 3SG.NOM.N be;3SG possible.
		I ask, whether it is possible

		Source: author based on examples in c
		Vetted: f
		Judgment: u
		Phenomena: {embedded interrogative}
		Ikh freg, az iz es efsher
		Ikh freg, az iz es efsher
		1SG.NOM ask;3SG, that;COMP be;3SG 3SG.NOM.N possible.
		I ask whether it is possible.

Our reference grammar does talk about embedded wh-interrogatives
though. In these constructions, word order is mostly SVO in the
embedded clause but there are some cases where inversion happens
where the resultant sentential force is more interrogative than
declarative.

		Source: a:233
		Vetted: t
		Judgment: g
		Phenomena: {embedded interrogative}
		Ikh veys nit vemen du host gezen.
		Ikh veys nit vemen du host ge-ze-n.
		1SG.NOM know;3G NEG WHO;DAT 2SG.NOM have;2SG PP-see-PP. 
		I know not whom you have seen.

		Source: a:233
		Vetted: t
		Judgment: g
		Phenomena: {embedded interrogative}
		Ikh veys nit vos du host gezen.
		Ikh veys nit vos du host ge-ze-n.
		1SG.NOM know;3G NEG what 2SG.NOM have;2SG PP-see-PP. 
		I know not what you have seen.

		Source: author based on a:233
		Vetted: f
		Judgment: u
		Phenomena: {embedded interrogative}
		Ikh veys nit az vemen du host gezen.
		Ikh veys nit az vemen du host ge-ze-n.
		1SG.NOM know;3G NEG that WHO;DAT 2SG.NOM have;2SG PP-see-PP. 
		I know not whom you have seen.

		Source: author based on a:233
		Vetted: f
		Judgment: u
		Phenomena: {embedded interrogative}
		Ikh veys nit az vos du host gezen.
		Ikh veys nit az vos du host ge-ze-n.
		1SG.NOM know;3G NEG that what 2SG.NOM have;2SG PP-see-PP. 
		I know not what you have seen.


# Coverage

We had a drop in coverage because of adding items to the test suite that included additions of adverbs and adjectives as well as interrogative questions. See the examples for these sections for a sample of the sentences that failed. 

Total items:            215
Positive items:         140
Negative items:          75
Parsing positive items:  53
Parsing negative items:  10
Coverage:                39%
Overgeneration:          13.3%


# Improvements and bug fixes to our grammar

We forgot to add tense=present to verbs inflecting in the present
tense. This didn't result in any changes to parse results however.

We forgot the -t form of the suffix lex rule involved in the past
participle circumfix. Adding this caused a couple of new items
containing such past participle forms. We also forgot to add
constraints such that this affix requires the presence of the ge-
prefix. This means that ungramatical items such as 'Ikh red-n', where
incorrect present tense verbal agreement was being parsed as the past
participle form. We fixed this by adding a require constraint on the
lex rule of the suffix to require the prefix.

Source: author
Vetted: f
Judgment: u
Phenomena: {agreement}
Ikh redn.
Ikh red-n.
1SG.NOM talk-1PL.
I talk.

We also discovered some spurious ambiguity that was introduced by the
introduction of argument optionality in lab 3. In particular, we were
getting two parses for transitive sentences with optional complements
verbs that were listed in the lexicon as having an optional argument
(ie transitive but OPT left underspecified on the argument). One of
the parses was correct, with the VP forming using the
basic-head-opt-comp-rule, but the other being derived with the
subj-head-rule and the basic-head-opt-comp-rule being applied in the
opposite order, with the basic-head-opt-comp-rule forming the spanning
parse. This ambiguity was seen in a simple sentence such as this:

Es zet.
Es ze-t.
3SG.NOM.N see-3SG.
It sees.

This ambiguity between the ordering of the rules seemed to be due to
an interaction between the V2 and argument optionality modules, where
the head-daughter of the head-subj or subj-head were not able to be
constrained to be COMPS < >, as this would prevent various
permutations of V2 word order that are required.

As suggested, we added the following TDL to yiddish.tdl so as to
remove the optionality of complements after being processed by a 
head-subj-rule:

  nolist := list.

  nocons := nolist & cons &
  [ FIRST unexpressed & [ OPT - ],
  REST  nolist ].

  nonull := nolist & null.

  basic-head-subj-phrase :+
  [ HEAD-DTR.SYNSEM.LOCAL.CAT.VAL.COMPS nolist ].

This was *almost* universally a success. All spurious ambiguity
associated with this problem was eliminated. However the following 
OVS sentence then (newly) fails to parse:

Source: a:224
Vetted: t
Judgment: g
Phenomena: {word order}
Dos bukh leyent der man.
Dos bukh leyen-t der man.
DEF.ACC.N book read-1SG DEF.NOM.M man.
It's the book that the man is reading. 

This should get a spanning parse with the head-subj-rule, however
interactive unification shows a unification failure at
SYNSEM.LOCAL.CAT.MC, getting the values na and bool.

