Base_JSON/prefixJ/json/J76/J76-3004.json

{
    "paper_id": "J76-3004",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T02:51:25.364236Z"
    },
    "title": "",
    "authors": [],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "",
    "pdf_parse": {
        "paper_id": "J76-3004",
        "_pdf_hash": "",
        "abstract": [],
        "body_text": [
            {
                "text": "The present paper describes a computational solution to the problem of locating words that are zeroed under conjunction. In this solution, which is based on general properties 3f conjunctional constructions, a mechanism locates zeroed elements in the conjunction strings and cross-references them d t h respect to elements in the h e a d construction. Constraints can then be a p p l i e d to elided conjuncts a s though t h e y w e r e expanded, and transformational expansion, which reconstructs the complete sentences underlying conjunctional occurrences, can be carried out straightforwardly by following the pointers which have been set up.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "The main innovation is called \"stacking\". It is a nond.eterministic programming device which causes restrictions ( i . . , subprograms applying detailed constraints) to be reexecuted on conjoined segments whenever the restriction is inovked on an element which has a conjunct. One particularly intricate problem in the computer parsing of natural language texts is the complexity introduced into the parsing system by conjunctions. This complexity is due to the richness of conjunctional constructions and to the material implicit in sentences containing conjunctions, The computafional problem can be divided into three parts: (I) generating parse trees which cover the occurrences of conjunction-headed strings in sentences;",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "(11) locating words in the sentences which are repeated implicitly in a \"zeroed,\" i.e., elided, form in particular positions in conjunction strings;",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "and (III) reconstructing the complete sentences underlying conjunctional occurrences when the application requires such an expansion.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "A generalized algorithmic solution to problem (I) was provided over a decade ago by the New York University Linguistic String Project (LSP) in the framework of linguistic string analysis (Sager e.t al. 1966 , Sager 1967 , Raze 1 9 6 7 ) , and similar devices have been described more recently, e.g.,by Woods in the framework of augmented transition networks (Woods 1973) . The present paper describes a compukational solution to problem (II), locating the words that are zeroed under conjunction. In this solution, which is based on general properties of conjunctional constructions, a mechanism locates zeroed elements in the conjunction strings and cross-references them with respect to elements in the head construction. Constraints can then be applied to elided conjuncts as though they were expanded, and transformational expansion, which was problem (111) above, can be carried out straightforwardLy by following the pointer's which have been set up.",
                "cite_spans": [
                    {
                        "start": 187,
                        "end": 206,
                        "text": "(Sager e.t al. 1966",
                        "ref_id": null
                    },
                    {
                        "start": 207,
                        "end": 219,
                        "text": ", Sager 1967",
                        "ref_id": null
                    },
                    {
                        "start": 358,
                        "end": 370,
                        "text": "(Woods 1973)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "The main innovation is called \"stacking.\" It is a nondeterministic programming device which causes restrictions e . , subprograms applying d e t a i l e d c o n s t r a i n t s ) t o be re-executed on conjoined segments wheqever t h e r e s t r i c t i o n i s invoked on a n element which h a s a c o n j u n c t . S i n c e t h e cons t r a i n t u s u a l l y a p p l i e s t o a combination o f words i n t h e main s t r i n g and i n t h e c o n j u n c t , t h e d e v i c e must e n s u r e t h a t t h e p r o p e r grammatical r e l a t i o n h o l d s between t h e s e wards. That i s , it t r e a t s t h e s e s e p a r a t e d words as a s i n g l e l i n g u i s t i c e n t i t y , o b t a i n i n g t h e e f f e c t of a f u l l expansion o f t h e conj u n c t i o n a l o c c u r r e n c e w i t h o u t c a r r y i n g o u t t h e p h y s i c a l rearrangement and copying or t h e p a r s e t r e e . The s t r a t e g y of t r e a t i n g e l l i p s i s i n t w o steps ( l o c a t i n g d e l e t e d elements and l a t e r c a r r y i n g o u t t h e p h y s i c a l e x p a n s i o n ) is See Appendix f o r a d e f i n i t i o n of t h e term \" s t r i n g \" as used formally i n t h e LSP grammar. t h e elements NVAR and RN (rumors h a s t i l y p r i n t e d ) . The d i f f e r e n c e i n t h e s e two t r e e s shows t h e a m b i g u i t v . i n t h e g i v e n s e n t e n c e .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SUMMARY",
                "sec_num": null
            },
            {
                "text": "Another parse t r e e of noun p h r a s e :",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "Hearsay and rumors h a s t i l y p r i n t e d",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "LNR I AND Q-CONJ L-- --% 1",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "Hearsay and l a r g e s t p a r t o f t h e grammar, and w i t h o u t them, t e x t p a r s i n g 6 i s o u t of t h e q u e s t i o n . I n a d d i t i o n , w e have found t h a t roughly one t h i r d of a l l t e x t sent e n c e s c o n t a i n c o o r d i n a t e o r comparative c o n j u n c t i o n s , many t i m e s i n complicated i n t e r r e l a t i o n . I t is t h e r e f o r e e s s e n t i a l t h a t t h e r e be a means f o r e x e c u t i n g r e s t r i c t i o n s on s e n t e n c e s c o n t a i n i n g c o n j u n c t i o n s . c e s of code which are used a g a i n and a g a i n . Then t h e s t a c k i n g mechanism can be implemented e f f -i c i e n t l y by changing t h e r o u t i n e s r a t h e r t h a n t h e more numerous r e s t r i c t i o n s . I n t h e LSP cage, t h e r o u t i n e s which w e r e modified are ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "!WAR TI N ------7 I *N J ' V E N P A S S 4 rumors ILVSA LVENR * ----. SA PASSOBJ RV SA t . A _---_I_.-l a ------. . l --A -. . ] *D p r i n t e d h a s t i l y 2. R E S T R I C T I O N S UNDER C O N J U N Z T I O N S ; S T A C K I N G .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "SA TENSE $8 VERB S A OBJECT ANDSTG RV SA L ------ l -L -L -$ -----A - -. . - --7- -I--- _I They spread rubrs i Y l 8 LAND SA Q-CONJ s SUBJECT S A TENSE SA- VERB SA OBJECT L-.--.- --. -----------A they* p r i n t hearsay",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "For the sentence shown i n Fig. 5 , They may spread b u t n o t p r i n t the rumors the execution of WSELl i s d i f f e r e n t . The OBJECT ( t h e rumors) h a s no ' conjunct b u t t h e VERB does. I n t h i s c a s e t h e COELEMl$NT(VEZ.B) r o u t i n e goes t o t h e first VERB (spread) from OBJECT and saves t h e second VERB ( p i n t ) or. looking a t t h e second VERB i n ) . Therefore, t h e well-formedness of p i n t rumors is also checked. sequences: heard f a c t s , heard rumors, b u t follow t h e t e n s e p o s i t i o n ( d i d ) .",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 27,
                        "end": 33,
                        "text": "Fig. 5",
                        "ref_id": "FIGREF8"
                    }
                ],
                "eq_spans": [],
                "section": "FIGURE 2",
                "sec_num": null
            },
            {
                "text": "T h i s c o n j u n c t i o n a l occurrence i s covered i n t h e ESP grammar by a node c a l l e d NULLC.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "The p r e t e r w i l l be \" l o o k i n g a t \" E2.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "$STACK-CONJUNCTS = VERIFY ITERATE $STACK-X.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "$STACK-X = DO $POSTCONJ; STACK.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "$STACK-CONJUNCTS locates a l l t h e c o n j u n c t s o f t h e node by i t e r a t i n g $STACK-X.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "I t t h e n r e t u r n s t o t h e s t a r t i n g node. $STACK-X goes t o e a c h conj u n c t by f i r s t e x e c u t i n g $POSTCONJ and t h e n c a l l i n g STACK, t h e o p e r a t o r which p u t s the c o n j u n c t o n t h e r e -e x e c u t i o n s t a c k . I n F i g . 4 , s t a r t i n g a t t h e f i r s t OBJECT, $STACK-TEST w i l l c a l l STACK f o r t h e second OBJECT. i s assumed t o be a t X.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "I t i s a l s o assumed t h a t t h e r o u t i n e which c a l l e d $STACK-FOR-LEFT-TO-X s t a r t e d a t some node Y , saved Y i n r e g i s t e r X200 and we& from Y l e f t one o r more nodes t o a r r i v e B t X. For i n s t a n c e , t h i s To produce a s y n t a c t i c a n a l y s i s of n a t u r a l language sentences the computer programusestwocomponents: a word dictionary ( F i t z p a t r i c k and Sager 1974) and",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "an English grammar (Sager The p a r s e r a n a l y z e s a s e n t e n c e by b u i l d i n g a p a r s e t r e e f o r the s e n t e n c e . , ETTHER OR , BOTH AND , which permit the logical combina- ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "-- -- -- tion of",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "7",
                "sec_num": null
            },
            {
                "text": "= ONE OF $AT-ATOM, $DESCEND-TO-ATOM, $DESCEND-TO-STRING.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SCORE-PATH",
                "sec_num": null
            },
            {
                "text": "$AT-ATOM = TEST FOR ATOM.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SCORE-PATH",
                "sec_num": null
            },
            {
                "text": "= DESCEND TO ATOM NOT PASSING THROUGH ADJSET1.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "$DESCEND-TO-ATOM",
                "sec_num": null
            },
            {
                "text": "$DESCEND-TO-STRING = DESCEND TO STRING NOT PASSING THROUGH ADJSET1.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "$DESCEND-TO-ATOM",
                "sec_num": null
            },
            {
                "text": "The CQRE routine locates the sentence word corresponding to a higher level grammatical element E by descending to a terminal node (\"atom\") from E. This is done by $DESCEND-TO-ATOM. When C O W descends from E it does n o t look at structures which are adjuncts, i . e . , on list ADJSET1. Thus for One rumor hastily printed can ruin careers shown in Fig. A 2 above, the routine CORE, starting a t SUBJECT, w i l l n o t s e a r c h below t h e l e f t -a d j u n c t node LN ( a r r i v i n g m i s t a k e n l y a t o n e ) and w i l l . arrive a t N (the noun rumor). Sometimes t h e s t a r t i n g l o c a t i o n of CORE w i l l be a n a t o m i c node. T h i s is provided f o r by $AT-ATOM, which t e s t s whether t h e c u r r e n t node is an atomic node, i . e . , o n l i s t ATOM. Sometimes a s t r i n g occurs i n a p a r t i c u l a r s e n t e n c e i n place of a noun.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 349,
                        "end": 357,
                        "text": "Fig. A 2",
                        "ref_id": "FIGREF6"
                    }
                ],
                "eq_spans": [],
                "section": "$DESCEND-TO-ATOM",
                "sec_num": null
            },
            {
                "text": "I n H i s p r i n t i n g rumors r u i n e d careers, shown i n F i g . Thus CORE s t a r t i n g a t SUBJECT i n OR DO RIGHTR(X) .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "$DESCEND-TO-ATOM",
                "sec_num": null
            },
            {
                "text": "= ITERATE GO LEFT UNTIL TEST FOR X SUCCEEDS.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE LEFTR ( X )",
                "sec_num": null
            },
            {
                "text": "= ITERATE GO RIGHT UNTIL TEST FOR X SUCCEEDS. G i v e n t h a t X and Y are e l e m e n t s of some s t r i n g , COELEMENT starts a t Y and goes t o X. COELEMENT uses several o t h e r basic r o u t i n e s : ROUTINE LEFTR(X) goes.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE RIGHTR ( X )",
                "sec_num": null
            },
            {
                "text": "l e f t from Y u n t i l i t locates X; ROUTINE RIGHTR(X) s e a r c h e s t o t h e r i g h t o f Y to locate X; and combining t h e two, ROtjTINE COELl.(X) s e a r c h e s b o t h s i d e s of Y t o f i n d X. I n F i g .~4 COELEMENT(SUBJECT), s t a r t i n g a t VERB i n ASSERTION, locates SUBJECT by e x e c u t i n g $SAME-LEVEL. ROUTINE LEFTR(SUBJECT) successf u l l y locates SUBJECT, which is to the l e f t o f VERB; t h i s s a t i s f i e s COELl(SUBJECT), which s a t i s f i e s $SAME-LEEL. I f X is i n a s t r i n g segment,",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE RIGHTR ( X )",
                "sec_num": null
            },
            {
                "text": "COELEMENT w i l l locate it by e x e c u t i n g $X-IN-SEGMENT. I n Fig. A 4 , COELEMENT(OBJECT), s t a r t i n g a t TPOS, f i r s t locates VINGO by executing COELl(STGSEG). I t t h e n locates OBJECT by c a l l i n g r o u t i n e ELEMENT(0BJECT).",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 68,
                        "end": 78,
                        "text": "Fig. A 4 ,",
                        "ref_id": "FIGREF23"
                    }
                ],
                "eq_spans": [],
                "section": "ROUTINE RIGHTR ( X )",
                "sec_num": null
            },
            {
                "text": "A different s i t u a t i o n occurs when COEUMENT(TP0S) starts a t OBJECT. T h i s by e x e c u t i n g $ASCNT and t h e n goes t o RN by e x e c u t i n g t h e r o u t i n e RIGHTR(RADJSET). RIGHT-ADJUNCT goes t o VENPRSS by e x e c u t i n g t h e CORE r o u t i n e .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE RIGHTR ( X )",
                "sec_num": null
            },
            {
                "text": "Thus t h e p a s s i v e s t r i n g VENPASS ( h a s t i l y p r i n t RV may immediately f o l l o w t h e verb as i n H e r a n q u i c k l y , shown i n ~i g . ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE RIGHTR ( X )",
                "sec_num": null
            },
            {
                "text": "A S S E R T I O N parse t r e e of H e r a n q u i c k l y . ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "F I G U R E A 5",
                "sec_num": null
            },
            {
                "text": "*See Appendix f o r a d e t a i l e d d e s c r i p t i o n of t h e routines'.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "T h i s w a s done because it i s f a s t e r t o e x e c u t e a n a d d r e s s t h a n a r o u t i n e .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [
            {
                "text": "A l l t h e guests came who were expected.PA SUBJECT S A TENSE S A VERB S A OBJECT RV S A ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ASSERTION paxse t r e e for",
                "sec_num": null
            },
            {
                "text": "When a node name appears a l o n e ( w i t h o u t any r o u t i n e name) as t h e subject of a r e s t r i c t i o n s t a t e m e n t , e . g . , OBJECT i n the r e s t r i c t i o n s t a t e m e n t OBJECT I S ENPTY, t h e r o u t i n e STARTAT is invoked, with t h e node name as argument, e , g . , STARTAT(OBJECT).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "STARTAT ROUTINE",
                "sec_num": null
            },
            {
                "text": "I t is assumed that X i s one l e v e l below t h e c u r r e n t nodeinwhichcaseDOWNl(X) locates X , o r t h a t X i s t h e c u r r e n t node i n which case TEST FOR X i s , s u ccessful.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "ROUTINE STARTAT(x) = EITHER DO DOWN~(X) OR TEST FOR X -",
                "sec_num": null
            }
        ],
        "bib_entries": {},
        "ref_entries": {
            "FIGREF0": {
                "uris": null,
                "type_str": "figure",
                "text": ". . . . . . . . . . . 3 . Less c o m m o n d e l e t i o n forms 1 7 4 . Implementation . . . . . . . . . . . . . . . . . . . . 0 References . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . 1 Appendix -The LSP parsing s y s t e m . . . . . . . . . . 2 A COMPUTATIONAL TREATMENT OF COORDINATE CONJUNCTIONS Carol Raze",
                "num": null
            },
            "FIGREF1": {
                "uris": null,
                "type_str": "figure",
                "text": "i m p o r t a n t f o r a s o l u t i o n t o c o n j u n c t i o n s f a r s e v e r a l r e a s o n s . F i r s t , it i s c o s t l y and o f t e n f r u i t l e s s t o expand a c o n j u n c t i o n a l s t r i n g u n t i l one i s s u r e one h a s a good parse. Second, t o g e t a good p a r s e one h a s t o e x e c u t e x e s t r i ct i o n s o n t h e g e n e r a t e d p a r s e t r e e . i n c l u d i n g t h e c o n j u n c t i o n s u b t r e e s , and t h i s r e q u i r e s l o c a t i n g d e l e t e d eldments. And t h i r d , t h e d e c i s i o n as t o whether c o n j u n c t i o n a l c o n s t r u c t i o n s s h o u l d be expanded o r n o t i s a p p l i c a t i o n s p e c i f i c . T h i s a r t i c -l e is d i v i d e d i n t o f o u r s e c t i o n s : 1. The ~e f i n i t i o n of c o n j u n c t i o n S t r i n g s ; 2. R e s t r i c t i o n s under Conjunctions; S t a c k i n g ; 3 . Less Common ~e l e t i o n Forms; and 4. Implementation. An appendix d e s c r i b e s t h e LSP system. 1. THE DEFINITION OF CONJUNCTION STRINGS C o n s t r u c t i o n s c o n t a i n i n g c o o r d i n a t e c o n j u n c t i o n s CONJ have t h e o v e r a l l . form: A CONJ B, where A and B a r e s t r u c t u r e s o f t h e same t y p e i n t h e gramm a r . I n l i n g u i s t i c s t r i n g grammar, A and B are p a r t i c u l a r e l e m e n t s o r sequences of elements i n d e f i n e d s t r i n g s of t h e grammar. An element o r a sequence of elements i n a s t r i n g o c c u r r e n c e may be c o n j o i n e d by a c o n j u n c t i o n a l s t r i n g which c o n s i s t s of t h e c o n j u n c t i o n followed by a n o t h e r o c c u r r e n c e o f t h e same t y p e of s t r i n g element ( o r e l e m e n t s ) t h a t p r e c e d e s t h e c o n j u n c t i o n . I n Hearsay and rumors can r u i n c a r e e r s we have a n d . r m o r s (AND + SUBJECT) conj o i n e d Hearsay (SUBJECT) t h e ASSERTION s t r i n g , which c o n s i s t s o f t h e elements SUBJECT + TENSE + VERB + OBJECT. (See Appendix f o r s t r i n g d e f i n it i o n s . ) And i n Rumors can r u i n c a r e e r s and can cause much h a r d s h i p we have and can c a u s e much h a r d s h i e (AND + TENSE + VEFU3 + OBJECT) conjoined a f t e r Rumors can r u i n c a r e e r s (SUBJECT +'TENSE + V E N + O B J E C T ) . Computationally, i n c l u d e a l l t h e c o n j u n c t i o n a l combinations l a r g e grammar i n t h e grammar d e f i n i t i 6 n s would complicate t h e grammar and make it immense. I n s t e a d , an i n t e r r u p t mechanism c a n be used t o a c h i e v e t h e same r e s u l t . * An i n t e r r u p t o c c u r s when a c o n j u n c t i o n i s reached while p a r s i n g s e n t e n c e . When an i n t e r r u p t o c c u r s , a c o n j u n c t i o n a l node is a t t h c h e d t o t h e p a r t of t h e t r e e b e i n g b u i l t . T o i l l u s t r a t e , a s i m p l i f i e d t r e e of-t h e t y p e t h e LSP system would g e n e r a t e f o r t h e noun p h r a s e Hearsay and rumors h a s t i l y p r i n t e d i s shown i n F i g . 1. A f t e r Hearsay i s matched a s a noun, an i n t e r r u p t o c c u r s , and t h e s p e c i a l p r o c e s s node ANDSTG i s a t t a c h e d t o t h e r i g h t of Hearsay. Howe v e r , r e s t r i c t i o n t h e LSP grammar l i m i t s t h e i n s e r t i o n s p e c i a l c e s s node t o o c c u r r i n g only i n LXR t y p e sequences o r i n s t r i n g s , ' and t h e r e f o r e *The mechanism d e s c r i b e d h e r e f o r g e n e r a t i n g c o n j u n c t i o n s t r i n g s was f i r s t programmed by James Morris f o r t h e 1966 I P L v e r s i o n of t h e LSP system (Sager e t a l . 1 9 6 6 ) , I t was expanded i n t h e FAP v e r s i o n o f t h e LSP system, programmed by t h e a u t h o r (Raze 1 9 6 7 ) , and i s a l s o p a r t of t h e c u r r e n t FORTRAN implementat i o n by Ralph Grishman (Grishman 1973): A s n o t e d above, t h e LSP s y s t m i s n o t t h e o n l y system t o employ a dynamic d e v i c e f o r g e n e r a t i n g c o n j u n c t i o n p a r s e t r e e s , +An LXR d e f i n i t i o n c o n s i s t s of t h r e e elements: a p o s i t i o n f o r t h e l e f t a d j u n c t s o f X, a c o r e p o s i t i o n f o r t h e word c l a s s X I and a p o s i t i o n f o r t h e r i g h t a d j u n c t s of X.",
                "num": null
            },
            "FIGREF2": {
                "uris": null,
                "type_str": "figure",
                "text": "s e t r e e o f noun p h r a s e : Hearsay and rumors h a s t i l y p r i n t r e j e c t e d i n t h e l o w e s t l e v e l . T h i s r e s t r i c t i o n a v o i d s v a r i o u s r edundancies c r e a t e d by c o n j o i n i n g a t i n t e r m e d i a t e l e v e l s and r e g u l a r i z e s t h e p o i n t s o f c o n j u n c t i o n i n t h e p a r s e t r e e . When i n s e r t i o n f a i l s , t h e parser d e t a c h e s t h e s p e c i a l p r o c e s s node and c o n t i n u e s p a r s l n g as if no i n t e r r u p t had o c c u r r e d . Another i n t e r r u p t may o c c u r , however, a f t e r t h e n e x t element i n t h e tree has been s a t i s f i e d . Thus f o r t h e noun phrase i n F i g . 1 an i n t e rr u p t o c c u r s a f t e r NVAR i s s a t i s f i e d and ANDSTG i s a t t a c h e d t o t h e r i g h t o f NVAK. ANDSTG c o n s i s t s ofand followed by t h e g e n e r a l c o n j u n c t i o n a l s t r i n g Q-CONJ. Q-CONJ c o n t a i n s a r e s t r i c t i o n which g e n e r a t e s a d e f i n i t i o n f o r Q-GQNJ. I t s d e f i n i t L o n c o n s i s t s o f a s e t of a l t e r n a t e v a l u e s : t h e f i r s t v a l u e i s t h e element t o t h e l e f t of t h e i n s e r t e d node, t h e second c o n s i s t s of t h e two elements t o t h e l e f t of t h e i n s e r t e d node, e t c . Thus,in F i g . 1 Q-CONJ i s NVAR (rumors). The p a r s e r resumes by c o n s t r u c t i n g element RN o f LNR. Here t h e r i g h t a d j u n c t h q s t i l y p r i n t e d a d j o i n s t h e c o n j u n c t i o n o f h e a r s a y and rumors. Another a n a l y s i s f o r t h e noun p h r a s e h e a r s a y and rumprs h a s t i l y p r i n t e d i s shown i n Fig. 2. I n t h i s case h e a r s a x h a s no r i g h t a d j u n c t and Q-CONJ c o n s i s t s of",
                "num": null
            },
            "FIGREF3": {
                "uris": null,
                "type_str": "figure",
                "text": "s e n t e n c e w i t h a c o n j u n c t i o n p r e s e n t s a problem f o r t h e a p p l i c a t i o n of d e t a i l e d c o n s t r a i n t s t o t h e p a r s e t r e e ( r e s t r i c t i o n s ) , an e s s e n t i a l p a r t of t h e p a r s i n g p r o c e s s . The t r e e s t r u c t u r e w i l l be d i f f e r e n t from t h a t assumed by t h e r e s t r i c t i o n ; c o n j u n c t i o n a l s t r i n g s w i l l have been i n s e r t e d and khe c o n j u n c t i o n a l s t r i n g s themselves may be t r u n c a t e d v e r s i o n s o f d e f i n e d s t r i n g s . To a p p r e c i a t e w h a t t h i s problem means, one must keep i n mind t h a t a grammar l i k e t h a t of t h e LSP f o r p r o c e s s i n g E n g l i s h t e x t s e n t e n c e s i s very l a r g e by comparison w i t h grammars used i n most o t h e r n a t u r a l language p r o c e s s i n g systems, which a r e d i r e c t e d t o p a r t i c u l a r s u b s e t s of English. The LSP grammar c o n s i s t s of approximately 3500 l i n e s . The r e s t r i c t i o n s c o m~r i s e by f a r t h e",
                "num": null
            },
            "FIGREF4": {
                "uris": null,
                "type_str": "figure",
                "text": "One s o l u t i o n t o t h i s problem i s t o r e w r i t e a l l t h e r e s t r i c t i o n s so t h a t they t e s t f o r c o n j u n c t i o n s and accommodate t r u n c a t e d segments. This w a s done i n e a r l i e r v e r s i o n s of t h e LSP p a r s e r but t h a t involved a tremendous amount of d e t a i l and b o t h i n c r e a s e d and complicated t h e grammar enormously. A s a n a l t e rn a t i v e , i n some c a s e s t h e Sentence can b e expanded s o t h a t t h e r e s t r i c t i o n s o p e r a t e on corgplete s t r i n g s . But i n o t h e r c a s e s t h i s i s n o t p o s s i b l e because t h e expansion n e c e s s i t a t e s t h e i n t r o d u c t i o n o f c e r t a i n t r a n s f o r m a t i o n s which should be done l a t e r . I n t h e p r e s e n t system w e t h e r e f o r e u s e a general solut i o n whereby r e s t r i c t i o n s a r e re-executed a u t o m a t i c a l l y f o r c o n j u n c t i o n a l occ u r r e n c e s . Thus, i n t h e a n a l y s i s shown i n F i g . 1 of Hearsay and rumors h a s t i l yp r i n t e d t h e s e l e c t i o n a l r e s t r i c t i o n w i l l be executed a u t g m a t i c a l l y f o r t h e sequences hearsay p r i n t e d and rumors p r i n t e d . This i s e q u i v a l e n t t o expanding t h e s e n t e n c e i n t o two a s s e r t i o n s , namely, (Someone) p r i n t e d hearsay and (Someone) p r i n t e d rumors. The a c t u a l expansion i s performed i n t h e t r a n s f o r m a t i o n a l phase ( i f t h e a p p l i c a t i o n r e q u i r e s e x p a n s i o n ) , which t a k e s p l a c e a f t e r t h e s u r f a c e a n a l y s i s i s o b t a i n e d . However, f o r t h e c o r r e c t s u r f a c e a n a l y s i s it i s c r u c i a l t h a t t h e r e s t r i c t i o n s o p e r a t e on expanded o r complete s t r i n g s . Thus, t h e c o n j u n c t i o n computation performs some of t h e f u n c t i o n o f expansion p r i o r t o t h e t r a n s f o r m a t i o n a l phase. To a p p l y r e s t r i c t i o n s t o s e n t e n c e s c o n t a i n i n g c o n j u n c t i o n s , a nond e t e r m i n i s t i c programming mechanism, which we c a l l t h e s t a c k i n g mechanism, w a s i n c o r p o r a t e d i n t o t h e parser. T h i s mechanism s a v e s t h e c o n j o i n e d s t r u ct u r e s o t h a t t h e r e s t r i c t i o n can be re-executed f o r t h e c o n j o i n e d s t r u c t u r e . I n p r i n c i p l e , t h e same method can be Japplled t o any grammar c o n t a i n i n g r e s t r i ct i o n s as long as t h e s t r u c t u r e a s s i g n e d t o t h e c o n j u n c t i o n a l o c c u r r e n c e i s of t h e same t y p e as t h e one a s s i g n e d t o t h e segment that precedes the c o n j u r l c t i o n -An e f f i c i e n t r e s t r i c t i o n component of a grammar makes u s e of r o u t i n e s f o r p i e",
                "num": null
            },
            "FIGREF5": {
                "uris": null,
                "type_str": "figure",
                "text": "t h o s e which l o c a t e a b a s i c t y p e o f s t r u c t u r e i n t h e p a r s e t r e e , such a s t h e roubinasCQRE, HOST, E L W N T , LEPT-ADJUNCT, e t c . * I n a d d i t i~n t o l o c a t i n g s t r u ct u r e s t h e modified r o u t i n e s must a l s o t e s t whether o r n o t t h e s t r u c t u r e s a r e c o n j o i n a d . When a r e s t r i c t i b n c a l l s a r o u t i n e which l o c a t e s a conjoined s t r u ct u r e , t h a t r o u t i n e i n t u r n c a l l s a n o p e r a t o r which s a v e s t h e conjoined s t r u ct u r e (~) a l o n g w i t h t h e p l a c e o f t h e r o u t i n e w i t h i n t h e sequence of o p e r a t i o n s b e i n g executed t h e r e s t r i c t i o n . The o p e r a t o r p u t s t h i s i n f o r m a t i o n r e -e x e c u t i o n s t a c k . The r o u t i n e r e t u r n s t o t h e o r i g i n a l s t r u c t u r e l o c a t e d and t h e r e s t r i c t i o n i n t e r p r e t e r e x e c u t e s t h e r e s t o f t h e r e s t r i c t i o n . A t t h i s p o i n t t h e r e s t r i c t i o n i n t e r p r e t e r \"looking a t \" t h e o r i g i n a l s t~u c t u r e l o c a t e d by t h e r o u t i n e . When a r e s t r i c t i o n i s s u c c e s s f u l t h e r e s t r i c t i o n i n t e r p r e t e r u s e s t h e i n f o r m a t i o n on t h e r e -e x e c u t i o n s t a c k t o resume e x e c u t i o n of t h a t restrict i o n . The r e s t r i c t i o n i s resumed a t t h e p o i n t immediately a f t e r t h e c a l l t o t h e r o u t i n e s o t h a t t h e r o u t i n e i t s e l f i s n o t c a l l e d a g a i n . I n s t e a d however, t h e = s t r i c t i o n i n t e r p r e t e r w i l l \"look a t \" t h e conjoined s t r u c t u r e p r e v i o u s l y l o c a t e d and sayed by t h e r o u t i n e . For example, when t h e v e r b -o b j e c t s e l e ct i o n r e s t r i c t i o n WSELl i s executed on t h e s e n t e n c e They p r i n t e d h e a r s a y and rumors, it c a l l s t h e CORE r o u t i n e t o o b t a i n t h e c o r e o f t h e o b j e c t . The CORE r o u t i n e w i l l go t o h e a r s a y , l o c a t e and s t a c k rumors, and r e t u r n t o hearsay. WSELl w i l l be s u c c e s s f u l f o r p r i n t e d hearsay. WSELl w i l l be resumed t h e r e f o r e a t t h e p o i n t a f t e r t h e c a l l t o t h e CORE r o u t i n e and rumors, which was saved by t h e CORE r o u t i n e , w i l l be plugged i n as though it was j u s t o b t a i n e d bv t h e CORE r o u t i n e . WSELl w i l l be s u c c e s s f u l f o r p r i n t e d rumors. A t r a c e of t h e e x e c u t i o n of WSELl which shows t h e r e -e x e c u t i o n of t h e r e s t r i c t i o n on t h e conj u n c t a p p e a r s i n F i g . 3 . Various c a s e s of WSELl w i t h examgles a r e e x p l a i n e d i n d e t a i l below. An e f f e c t o f dynamic g e n e r a t i o n o f c o n j u n c t i o n a l parse trees is t h a t i n a d d i t i o n t o l o c a t i n g c o n j o i n e d v a l u e s , p r o v i s i o n must b e made f o r r e s t r i c t i o n s t o f u n c t i o n p r o p e r l y on t h e non-conjunctional grammar, i n new s i t u a t i o n s due t o c o n j u n c t i o n s ; a r o u t i n e may b e o p e r a t i n g i n a s t r u c t u r e i n t o which conjunct i o n a l s t r i n g s have been i n s e r t e d , o r i n a t r u n c a t e d v e r s i~n o f a d e f i n e d s t r i n g o r host-adjuruct sequence. For example, t h e LSP RIGHT-ADJUNCT r o u t i n e i s assumed t o s t a r t a t a node X i n a n LXR t y p e node. Without c o n j u n c t i o n s t h e RIGHT-ADJUNCT r o u t i n e g o e s one node t o t h e r i g h t from X t o a r r i v e a t t h e r i g h t a d j u n c t o f X. I n Fig. 2 f o r t h e noun p h r a s e rwnors h a s t i l y p r i n t e d , RIGHT-ADJUNCT, s t a r t i n g a t t h e c o r e p o s i t i o n NVAR subsuming rumors, goes one node t o t h e r i g h t t o a r r i v e a t i t s r i g h t a d j u n c t RN ( h a s t i l y p r i n t e d ) . However, w i t h con j u n c t i o n s t h e r o u t i n e must go r i g h t u n t i l l a n d s a t non-con j u n c t i o n a l node. Thus, i n Fig. 1, f o r t h e noun p h r a s e hearsay and rumors h a s t i l y p r i n t e d , RIGHT-ADJUNCT goes from NVAR ( h e a r s a y ) past t h e c o n j u n c t i o n s t r i n g (and rumors) t o RN. When RIGHT-ADJUNCT starts i n t h e c o n j u n c t i o n a l s t r i n g a t NVAR ( r u m o r s ) , RN i s n o t t o i t s r i g h t . To go t o RN, t h e r o u t i n e l o c a t e s t h e corresponding p r e c o n j u n c t i o n element NVAR ( h e a r s a y ) and goes t o RN from t h e r e . To i l l u s t r a t e how t h e s t a c k i n g mechanism works w e w i l l e x p l a i n i n d e t a i l how r e s t r i c t i o n WSELl i s executed f o r s e v e r a l s e n t e n c e s w i t h c o n j u n c t i o n s . WSELl i s housed i n t h e c e n t e r s t r i n g ASSERTION:* *WSEL~ i s a l s o housed i n o t h e r s t r i n g s c o n t a i n i n g o b j e c t and v e r b e l e m e n t s , b u t f o r o u r example w e w i l l o n l y c o n s i d e r t h e s t k i n g ASSERTION. I n t h e s t a t e m e n t W S E L~ = IN ASSERTION: IF ALL OF SOBJECT-NOUN, NOUN = THE C O E X 1 OF THE OEUECT X10 IS N OR PRO. $GOVERNING-VERB = AT X10, COELEMENT VERB X4 EXISTS. SFORBIDDEN-NOUN-LIST = THE CORE OF X4 HAS ATTRIBUTE N.OTNOBJ X 5 . The above r e s t r i c t i o n s t a t e m e n t s have t h e f o l l o w i n g f u n c t i o n s : $OBJECT-NOUN checks t h a t t h e c o r e of t h e o b j e c t p o s i t i o n i s a noun o r pronoun; $GOVERNING-VERB tests t h a t a v e r b eoelement e x i s t s ; and SFORBIDDEN-NOUN-LIST checks t h a t the g i v e n v e r b h a s an a t t r i b u t e NOTNOBJ.* I f a l l t h e s e c o n d i t i o n s a r e s a t i sfied t h e s e l e c t i o n check $NOCOMMON i s made. $NOCOMMON = L I S T S X 1 AND XS HAVE NO COMMON ATTRIBUTE. If t h e noun d o e s have a s u b c a t e g o r y t h a t i s on t h e subcategory l i s t of NOTNOBJ, t h e n $NOCOMMON f a i l s and t h a t noun i s n o t a c c e p t e d as t h e o b j e c t of t h e g i v e n v e r b . Consider t h e parse t r e e f o r t h e s e n t e n c e shown i n F i g . 4 , They s p r e a d rumors and t h e y p r i n t h e a r s a y . In $OBJECT-NOUN, i n o r d e r t o go t o t h e c o r e of t h~ OBJECT, OBJECT must f i r s t be located. T h i s is accomplished by t h e routine (The f i r s t f o o t n o t e i s c o n t i n u e d from p. 8,) $GOVERNING-VERB we can t h e r e f o r e u s e t h e r o u t i n e COELEMENT w i t h argument VERB. I n t h e a c t u a l r e s t r i c t i o n a more g e n e r a l routine VERB-COELEMENT i s used. T h i s r e s t r i c t i o n was d e s c r i b e d in full, w i t h o u t r e f e r e n c e t o i t s o p e r a t i o n on conjunction s e n t e n c e s , i n Sager and Grishman 1975. *NOTNOW i s a s s i g n e d in t h e word d i c t i o n a r y t o t r a n s i t i v e v e r b s ; i t s a t t r i b u t e s f o r a g i v e n v e r b a r e t h o s e noun s u b c l a s s e s of t h e grammar which a r e n o t approp r i a t e noun o b j e c t s of t h e g i v e n v e r b ( i n s c i e n t i f i c w r i t i n g ) .",
                "num": null
            },
            "FIGREF6": {
                "uris": null,
                "type_str": "figure",
                "text": "Trace olf WSELL STARTAT(0BJECT). STARTAT both l o c a t e s OBJECT and c a l l s t h e s t a c k o p e r a t o r f o r each conjoined OBJECT. rhus f o r t h e sentence i n Fig. 4 , STARTAT w i l l go t o the f i r s t OBJECT (subsuming cumors) and w i l l save t h e second OBJECT (hearsax)t. Wheh STARTAT i s compLeted t h e CORE r o u t i n e i s c a l l e d t o l o c a t e t h e c o r e of OBJECT. I n t h i s exmpl'e it l o c a t e s the noun rumors.$GOVERNING-VERB goes t o VERB, which i s a coelement of OBJECT. I t does t h i s by f i r s t l o c a t i n g OBJECT (which was saved i n r e g i s t e r X10 by $OBJECT-NOUN) and by c a l l i n g the COELEbIENT r o u t i n e . Thus i n Fig. 4 , COELEXENT (WEW) goes from t h e f i r s t OBJECT t o i t s coelement VERB ( s p r e a d ) . I t l o c a t e s a l l conjuncts of this VERB and determines whether o r n o t t o call the s t a c k o p e r a t o r . VERB (spread) has a c o n j u n c t b u t i n t h i s case COELEIlENT(VERB) w i l l n o t s t a c k t h e conjoined VERB ( p r i n t ) because t h i s VERB has i t s own coelement OBJECT. When COELEMENT r e t u r n s , t h e r e s t r i c t i o n i n t e r p r e t e r is looking a t the f i r s t VERB. WSELl i s s u c c e s s f u l f o r spread rumors. Since t h e r e i s something on t h e r eexecution s t a c k , t h e execution of WSELl i s resumed, It i s resumed i n $OBJECT-NOUN a t t h e p o i n t immediately a f t e r t h e c a l l t o STARTAT(OEUECT), However, t h i s t i m e t h e r e s t r i c t i o n i n t e r p r e t e r i s l o c a t e d a t t h e second OBJECT. The c o r e (hearsay) of t h e second OEUECT is obtained and t h e r e s t of t h e r e s t r i c t i o n i s executed f o r t h e secona time. I n p a r t i c u l a r , COELEMENT(VERB) goes t c l t h e second VERB ( p r i n t ) from t h e second OBJECT and WSELl is s u c c e s s f u l f o r print hearsay. F I G U R E 4 P a r s e t r e e of They spread rumors and they p r i n",
                "num": null
            },
            "FIGREF7": {
                "uris": null,
                "type_str": "figure",
                "text": "t h e re-execution s t a c k . WSELl i s s u c c e s s f u l for spread rumors. Its execut i o n is t h e r e f o r e resumed i n $GOVERNING-VERB a t t h e p o i n t j u s t a f t e r t h e ~1 3 . t o COELEMENT(VERB), However, t h i s time t h e r e s t r i c t i o n i n t e r p r e t e r i s",
                "num": null
            },
            "FIGREF8": {
                "uris": null,
                "type_str": "figure",
                "text": "a r s e tree of rhey may spread b u t n o t p r i n t t h e rumors. e sentence shown i n Fiyure 6 , They heard and p r i n t e d f a c t s and. rumors, t h e execution of WSELl i s again different from t h e previous examples. OBJECT itself has no c o n j u n c t b u t t h e core of OBJECT does. Thus, when CORE i s called i n SOELECT-NOUN, the CORE r o u t i n e w i l l l o c a t e N ( f a c t s ) and w i l l place the second N (rumors) on t h e re-execution s t a c k . Likewise, t h e COELEMENT(VERI3) r o u t i n e w i l l not f i n d a c o n j u n c t for t h e VERB position i t s e l f However, t h e c o r e of the VERB has a conjunct. When CORE i s called in SFORBIDDEN-NOUN-LIST, it w i l l locate t v ( h e a r d ) and w i l l place t h e second t v ( p r i n t e d ) on t h e r e -e x e c u t i o n stack. T h i s w i l l r e s u l t i n WSELl being executed f o r a l l f o u r v e r b + o h j e c t noun f a c t s , p r i n k e d xumows.",
                "num": null
            },
            "FIGREF9": {
                "uris": null,
                "type_str": "figure",
                "text": "Parse t r e e of They heard and p r i n t e d f a c t s and rumors. n t e d OBJECT RV SA -r---. --. --. I c I -----.I--)L_ -2 l l o w i n g is an example of how t h e s t a c k i n g mechanism h e l p s t o r es o l v e s y n t a c t i c ambiguity. I n t h e s e n t e n c e H e p r i n t e d r u m o r s and his f r i e n dalso, there are two possible p a r s e s . I n one a n a l y s i s t h e o b j e c t o f p r i n t e s c o n s i s t s o f t h e c o n j o i n e d nouns rumors and f r i e n d . T h i s a n a l y s i s I C C U~~ be r e j e c t e d by WSELl (assuming p r i n t to have iTkIb-i\\Wi as 3 valiie of :~oT;;o~\\T) i n the fallowing manner. When the CORE r o u t i n e is callctl in SOBJECT-NOUN, t h e noun rumors is l o c a t e d and t h e conjoined noun f r i e n d i s placed on t h e re-execution s t a c k . P r i n t e d i s l o c a t e d by $GOVERNING-VERB, P r i n t e d rumors i s s u c c e s s f u l and WSELl i s re-executed f o r p r i n t e d f r i e n d . However, p r i n t e d has NHUMAN on i t s NOTNOBJ l i s t and f r i e n d has a noun subcategory NHUMAN. Thus $NOCOMMON f a i l s and t h i s a n a l y s i s i s r e j e c t e d . Another a n a l y s i s of t h i s sentence c o n t a i n s a second ( i m p l i c i t ) occurrence of p r i n t e d rumors: H e p r i n t e d rumors and h i s f r i e n d a l s o ( p r i n t e d rumors), The conjoined s t r i n g c o n s i s t s of SUBJECT ( h i s f r i e n d ) followed by VERB and OFUECT, which b o t h are assigned t h e v a l u e s NULLC (Section 3 ) . 3 . LESS COMMON DELETION FORMS Some c o n j u n c t i o n a l s t r i n g s d i f f e r from t h e usual form of c o n j u n c t i o n a l occurrences t h a t t h e d e l e t e d elements are n o t contiguous t o t h e con j unct i o n . For example, i n H e p r i n t e d t h e f a c t s b u t h i s f r i e n d d i d n o t , t h e v e r b and o b j e c t p r i n t e d t h e f a c t s have been d e l e t e d i n t h e c o n j u n c t i o n a l s t r i n g ; t h e p o s i t i o n s of t h e d e l e t e d verb and o b j e c t i n b u t h i s f r i e n d d i d n o t do n o t immediately f 01 low b u t ,",
                "num": null
            },
            "FIGREF10": {
                "uris": null,
                "type_str": "figure",
                "text": "NULLC i s a u t o m a t i c a l l y s a t i s f i e d without subsuming a s e n t e n c e word i f c e r t a i n c o n d i t i o n s a r e m e t . When it o c c u r s i n a p a r s e , it r e p r e s e n t s t h e f a c t t h a t a r e q u i r e d element of a s t r i n g -c o n j u n c t h a s been d e l e t e d i n such a p o s i t i o n t h a t t h e r e s u l t i n g p a r s e t r e e does n o t have s i m i l a r s t r u c t u r e s on e i t h e r s i d e of t h e conjunction. Thus, i n t h e p a r s e of t h e preceding s e n t e n c e f a shown i n Fig. 7 , t h e v e r b and o b j e c t of t h e a s s e r t i o n following b u t both have t h e v a l u e NULLC. Before a NULLC node i s accepted, s e v e r a l \"zeroing\" r e s t r i c t i o n s must be s a t i s f i e d . The g e n e r a l c o n d i t i o n s f o r acceptance of a NULLC node are t h a t",
                "num": null
            },
            "FIGREF11": {
                "uris": null,
                "type_str": "figure",
                "text": "a r s e t r e e of H e p r i n t e d t h e f a c t s b u t h i sfriend d i d n o t . ------_-_-----A\" --. . 4 -. -^---L Z . -. .& ---L -..--I L -3------. .be w i t h i n a c o n j u n c t i o n a l s t r i n g ; 2 ) t he p r e -c o n j u n c t i o n a l element which corresponds t o t h e d e l e t e d element sub9ume some s e n t e n c e w o r d ( s ) ; and 3 ) t h a t the z e r o i n g p a t t e r n be one o f a few t y p e s known t o be a c c e p t a b l e for c o n j u n c t i o n a l strings w i t h this type of d e l e t i o n (see examples below). Thus i n t h e example s e n t e n c e shown i n F i g . 7 , the verb of t h e a s s e r t i o n f o l l o w i n g but w i l l have t h e v a l u e NULLC, and c o n d i t i o n s 1-3 are m e t as f o 1 l o w s . i -1) NULLC i s i n a c o n j u n c t i o n a l s t r i n g --t h e string headed by b u t .",
                "num": null
            },
            "FIGREF12": {
                "uris": null,
                "type_str": "figure",
                "text": "The c o r r e s p o n d i n g p r e -c o n j u n c t i o n a l v e r b s u b s u m~s p r i n t e d .3) The NULLC verb i s n o t contiquous t o t h e c o n j u n c t i o n b u t , b u t f o l l o w s t h e t e h s e i n t h e c o n j u n c t i o n a l s t r i n g , t h i s c o n s t r u c t i o n b e i n g one of t h e allowed p a t t e r n s of z e r o i n g .I n a d d i t i o n t o e s t a b l i s h i n g c o n d i t i o n s f o r a c c e p t i n g NULLC nodes, the \"zeroing1' r e s t r i c t i o n s have a n o t h e r v e r y i m p o r t a n t f u n c t i o n . They locate the m a t e r i a l t h a t would have been r e p e a t e d i n t h e conjoined p h r a s e b u t which w a s \"zeroed,\" t h a t i s , d e l e t e d o r elided, and t h e y save t h a t i n f o r m a t i o n . When the a p p r o p r i a t e s t r u c t u r e i s l o c a t e d f o r a WLLC node, t h e l o c a t i o n is saved v i a a n o d e -a t t r i b u t e mechanism.The r e s t r i c t i o n assigns t o t h e NULLC node t h e node a t t r i b u t e LINKC, whose v a l u e p o i n t s t o t h e p r e -c o n j u n c t i o n a l s t r u c t u r e t h a t would have been repeated. Once t h e LINKC attribute i s assigned, any c o n s t r a i n t can l o c a t e t h i s s t r u c t u r e and Ehus i d e n t i f y t h e words t h a t have been d e l e t e d . I n the LSP system t h e actual physical rearrangement of the parse t r e e i s n o t dane u n t i l t h e t r a n s f o r m a t i o n a l phase. However, t h e a c t u a l \" f i l l i n g i n \" trans-f o r m a t i o n i s s i m p l e and s t r a i g h t f o r w a r d because t h e main b u l k of t h e work ( l o c a t i n g t h e material i d e n t i c a l t o what i s z e r o e d ) i s done by t h e r e s t r i ct i o n s i n the p a r s i n g phase. For c o n j u n c t i o n s t r i n g s with t h e NULLC t y p e of d e l e t i o n , r e s t r i c t i o n s c a n be executed o n l y a f t e r t h e f i l l i n g i n of d e l e t e d elements h a s been done. Some sxamples of t h e NULLC d e l e t i o n types which are covered by t h e LSP grammar a p p e a r i n p a r e n t h e s e s i n t h e f o l l o w i n g s e n t e n c e s : H e p r i n t e d f a c t s but h i s f r i e n d d i d n o t ( p r i n t f a c t s ) . H e washed t h e d i s h e s b u t s h e d i d n o t want t o (wash t h e d i s h e s ) . He s h o u l d have t r i e d to wash t h e d i s h e s b u t he d i d n o t ( t r y to wash t h e d i s h e s ) . \" H e ate supper and s h e (ate supper) t o o .H e p l a y e d t h e played t h e p i a n o and s h e (played) t h e drums.",
                "num": null
            },
            "FIGREF13": {
                "uris": null,
                "type_str": "figure",
                "text": "f t and (he left) f a s t .H e t r i e d 'to wash t h e d i s h e s b u t s h e was n o t w i l l i n g t o ( t r y t o wash t h e d i s h e s ) . * *This sentence c o n t a i n s a n ambiguity as t o how much material w a s d e l e t e d : t r y t o wash t h e d i s h e s vs. wash t h e d i s h e s . 4 . IMPLEMENTATION The b a s i c r o u t i n e s * of t h e LSP grammar were modified t o handle conjunct i o n s . They were modified t o l o c a t e t h e a p p r o p r i a t e conjoined s t r u c t u r e ( s ) and c a l l an o p e r a t o r which saves t h o s e s t r u c t u r e ( s ) . I n a d d i t i o n , t h e r o u t i n e s were modified t o f u n c t i o n p r o p e r l y i n t h e non-conjunctional g r m a r f o r t h e new s i t u a t i o n s which occur when c o n j u n c t i o n s a r e p r e s e n t . Although t h e s e changes involve some of t h e b a s i c r o u t i n e s only, t h e r e s t r i c t i o n s use t h e s e r o u t i n e s so o f t e n so that t h e changes have t o be a s e f f i c i e n t a s p o s s i b l e . Otherwise t h e e x e c u t i o n time of a r e s t r i c t i o n would be g r e a t l y i n c r e a s e d , To save much r e p e t i t i o u s moving around t h e t r e e , p o i n t e r s a r e a t t a c h e d t o t h e a p p r o p r i a t e nodes of t h e t r e e v i a a n o d e -a t t r i b u t e mechanism which i s d e s c r i b e d i n Sager and Grishman 1975. Each element E2 i n a c o n j u n c t i o n a l s t r i n g of t h e form El CONJ E is a s s i g n e d a node a t t r i b u t e c a l l e d PRECONJELEM 2 which p o i n t s t o t h e corresponding element E i n t h e s t r i n g -3rior t o conjunc-1 t i d n . Likewise each E i s a s s i g n e d a node a t t r i b u t e c a l l e d POSTCONJELEM 1 which p o i n t s t h e corresponding element E i n t h e 2 post-con j u n c t i o n a l s t r i n g . The node a t t r i b u t e assignments a r e done by a well-formedness r e s t r i c t i o n housed on the c o n j u n c t i o n a l s t r i n g . Once t h e node a t t r i b u t e s a r e a s s i g n e d , t h e r o u t i n e s can q u i c k l y o b t a i n ( o r check f o r ) conjoined v a l u e s of a node by u s i n g i t s node a t t r i b u t e POSTCONJELEM. And when a r o u t i n e i s c a l l e d from i n s i d e a t r u n c a t e d s t r i n g segment, it can q u i c k l y move t o t h e corresponding pre-conjunction element by o b t a i n i n g t h e node a t t r i b u t e PRECONJELEb1. From t h a t p o i n t t h e r o u t i n e can then l o c a t e t h e a p p r o p r i a t e elemeat. For example i n F i g , 1, t h e second NVAR (rumors) h a s been a s s i g n e d node a t t r i b u t e PFU3CONJELEM p o i n t i n g t o a NVAR ( h e a r s a y ) . Using t h e node a t t r i b u t e PRECONJELEM *See Appendix f o r a d e t a i l e d e x p l a n a t i o n of t h e b a s i c r o u t i n e s . of t h e second NVAR, RIGHT-ADJUNCT goes t o t h e f i r s t NVAR and then goes t w o nodes t o t h e r i g h t t o RN. Not a l l r e s t r i c t i o n s may be re-executed f o r sentences c o n t a i n i n g conj u h c t i o n a l occurrences. For example, t h o s e r e s t r i c t i o n s t e s t i n g number agreement have t o be changed t o e x p l i c i t l y t e s t f o r t h e occurrence of a conjunction. Therefore, t h o s e r e s t r i c t i o n s must use r o u t i n e s t h a t do n o t s t a c k . Each r o u t i n e that calls t h e s t a c k o p e r a t o r has a c o u n t e r p a r t which does n o t . The r o u t i n e s are w r i t t e n s o t h a t i n each p a i r , t h e one which s t a c k s c a l l s on t h e nonstacking v e r s i o n as a subpart. W e w i l l now go i n t o t h e d e t a i l s of t h o s e r o u t i n e s t h a t were modified f o r conjunctions. The e x p l a n a t i o n s here w i l l be concerned w i t h t h e modificat i o n s only. h s t of t h e b a s i c r o u t i n e s f a l l i n t o t h r e e c a t e g o r i e s : 1) t h o s e which begin at. X or go down t o X. STARTAT(X), ELEMENT(X), LAST-ELEMENT and NELEMRT* a r e i n t h i s category. These r o u t i n e s concern o n l y one element o f a s t r i n g o r sequence. 2) T h e r o u t i n e s which go r i g h t o r l e f t t o X. HOST, RIGHT-ADJUNCT, LEFT-ADJUNCT, NEXT-ELEMENT, PREVIOUS-ELSMENT and COELEMENT(X) a r e in t h i s category. These r o u t i n e s involve two elements of a s t r i n g o r Sequence. 3)'The r o u t i n e s which s t a r t at o r go up t o X. IMMEDIATE(X) , STARTAT(X), PRESENT-ELEMENT, I T , PRESENT-STRING, IMMEDIATE-NODE a r e in t h i s category. Because t h e conjunction m o d i f i c a t i o n can be g e n e r a l i z e d depending on which category a r o u t i n e i s i n , only s e v e r a l r o u t i n e s a r e needed t o handle v a r i o u s conjunction o p e r a t i o n s . The actual m o d i f i c a t i o n t o many of the r o u t i n e s t h e r e f o r e c o n s i s t s of addins a c a l l t o one of t h e f e w r o u t i n e s which handle c o n j u n c t i o n a l o p e r a t i o n s . *Core i s a l s o i n t h i s category; however, t h i s r o u t i n e needs t o perform some e x t r a o p e r a t i o n s t o g e t t o t h e conjoineu core word. The r o u t i n e t h a t h a n d l e s s t a c k i n g f o r r o u t i n e s of c a t e g o r y (1) above i s c a l l e d $STACK-TEST : * $STACK-TEST = I F $POSTCONJ THEN $STACK-CONJUNCTS . $POSTCONJ \" THE PRESENT-EWmNT HAS NODE ATTRIBUTE POSTCONJELEM. (GLQEAL) It i s assumed t h a t t h e r e s t r i c t i o n interpreter i s \" l o o k i n g a t \" X when $STACK-TEST i s c a l l e d . I f an element E h a s a c o r r e s p o n d i n g element E i n a con-1 2 j u n c t i o n a l s t r i n g E w i l l have t h e node a t t r i b u t e POSTCONJELEM. T h i s p r o v i d e s 1 a q u i c k test t o d e t e r m i n e whether o r not a node h a s a c o n j u n c t . I f El does n o t have t h e node a t t r i b u t e POSTCONJEEM, $POST60NJ f a i l s and $STACK-TEST i s f i n i s h e d ; i f E h a s t h e node a t t r i b u t e POSTCONJELEM, t h e a t t r i b u t e POSTCONJEEEM 1 h a s a v a l u e , namely E 2' When $POSTCONJ i s f i n i s h e d t h e r e s t r i c t i o n i n t e r -",
                "num": null
            },
            "FIGREF14": {
                "uris": null,
                "type_str": "figure",
                "text": "I t will r e t u r n t o t h e f i r s t OBJECT, b e f o r e e x i t i n g , a s a r e s u l t of t h e VERIFY con-mand.The r o u t i n e s t h a t h a n d l e s t a c k i n g f o r t h o s e r o u t i n e s i n c a t e g o r y 2 above are c a l l e d $STACK-FOR-LEFT-TO-X and $STACK-FOR-RGHT-TO-X r e s p e c t i v e l y . W e w i l l o n l y go i n t o t h e d e t a i l s of $STACK-FOR-&FT-TO-x s i n c e $STACK-FOR-RGHT-TO-X i s similar. When $STACK-FOR-LEFT-TO-X i s c a l l e d , t h e r e s t r i c t i o n i n t e r p r e t e r *In o u r system $STACK-TEST i s a c t u a l l y a g l o b a l a d d r e s s i n s t e a d of a r o u t i n e .",
                "num": null
            },
            "FIGREF15": {
                "uris": null,
                "type_str": "figure",
                "text": "o c c u r s when t h e r o u t i n e COELEMENT(VERB) i s c a l l e d from OEUECT i n Fig. 4. $STACK-FOR-LEFT-TO-x = IF SPOSTCONJ THEN VERIFY $STACK-IF-NO-Y-RGHT. (GLOBAL)If X has a corresponding element i n a c o n j u n c t i o n a l s t r i n g it w i l l have t h e node a t t r i b u t e POSTCONJELEM.If X does n o t have t h e node a t t r i b u t e POSTCONJEL!?M, $STACK-FOR-LEFT-TO-X i s f i n i s h e d . If it does have node a t t r i b u t e POSTCONJELEM, $STACK-IF-NO-Y-LEFT i s executed t o determine whether o r n o t to stack t h e c o n j u n c t (s) : $STACK-IF-RO-Y -RGHT = I F $POSTCONJ THl EN EITHER ALL OF $NO-Y-TO-RIGHT, Y-TO-RIGHT = NOT ITERATE GO RIGHT UNTIL TEST FOR X200 SUCCEEDS. $DO-STACK = STACK. Looking a t Fig. 4 assume COELEMENT(VERI3) i s c a l l e d when t h e r e s t r i c t i o n i n t e r p r e t e r i s a t OBJECT (rumors). The COELEMENT r o u t i n e l o c a t e s t h e f i r s t VERB ( s p r e a d ) . The premise o f $STACK-FOR-LEFT-TO-X i s s u c c e s s f u l because t h i s VERB has a c o n j u n c t and $STACK-IF-NO-Y-RGHT i s executed. The premise o f $STACK-IF-NO-Y-RGHT i s s u c c e s s f u l : a c o n j u n c t i o n i s found, i n t h i s c a s e t h e second VERB ( p r i n t ) . The r e s t r i c t i o n i n t e r p r e t e r remains a t t h e second VERB w h i l e t h e r e s t of t h e i m p l i c a t i o n i s executed. I n t h i s example, t h e r e i s OBJECT, ( h e a r s a y ) t o t h e r i g h t of t h e second V W ; t h e r e f o r e $NO-Y-TO-RGHT f a i l s . A s a r e s u l t VERB ( p r i n t ) i s n o t s t a c k e d . I n F i g . 5, i f COELEMENT (VERB) i s c a l l e d when t h e r e s t r i c t i o n i n t e r p r e t e r i s a t OBJECT ( t h e rumors) t h e VERB (spread) is located. It has a conjunct, which is stacked. In this case $NO-Y-TO-RGHT i s s u c c e s s f u l s i n c e there i s no OBJECT to the right of the second VERB. SSTACK-IF-NO-Y-RGHT is recursive so that all t h e c o h j o i n e d structures are located and t e s t e d . In the above examples if SPOSTCONJ is t r u e starting at the second VERB then $STACK-IF-NO-Y-RGHT goes to the next corresponding post-canjunctional VERB and determines whether or n o t to s t a c k it, In F i g . 8, after e x e c u t i n g t h e remise of $STACK-IF-NO-Y-RGWT the restriction interpreter is a t t h e second VERB. $NO-Y-TO-RIGHT i s t r u e and $DO-STACK is executed. ~t w i l l stack t h e second VERB. $STACK-XF-NO-Y-RGHT i s called recursively and the restriction interpreter is at the third VERB. Howevex, OBJECT (them) is present; t h e r e f o r e $NO-Y-TO-RIGHT will fail and the third VERB (ignored) will not be stacked. Routines COELEMENT, YOST, LEFT-ADJUNCT and PREVIOUS-ELEMENT use $STACK-FOR-LEFT-TO-X . Routines COELEMENT, HOST, RIGHT-ADJUNCT and n d -d i d p r i n t t h e rumors b u t w e i g n o r e d them. u t i n e s i n c a t e g o r y 2 were a l s o modified t o o p e r a t e p r o p e r l y i f they s t a r t i n a t r u n c a t e d segment of a d e f i n e d s t r i n g o r h o s t a d j u n c t sequence. For example, i f cOELEMENT(OBJECT) i s c a l l e d from t h e second VERB i n F i g u r e 5, t h e CQELEMENT routine w i l l n o t be a b l e t o go l e f t o r r i g h t t o OBJECT. I t must f i r s t go t o t h e corresponding p r e -c o n j u n c t i o n a l element and t h e n t r y t o go l e f t o r right t o X from there. This i s accomplished by $TO-PRECONJUNCTION-Y. $TO-PRECONJUNCTION-Y = E I T H E R SPRECONJ OR $ASSIGN-PRECONJELEM (GLOBAL ) SPRECONJ = THE PRESENT-ELEMENT-HAS NODE ATTRIBUTE PRECONJELEM. If the s t a r t i n g node h a s node a t t r i b u t e PRECONJELEM, SPRECONJ w i l l go t o t h e corresponding p r e -c o n j u n c t i o n a l node; othefwise the node a t t r i b u t e s PRECONJELEM and POSTCONJELEM have t o be a s s i g n e d . T h i s i s accomplished by $ASSIGN-PRECONJELEM: SZOCATE-CONJNODE = ASCEND TO Q-CONJ; GO UP; STORE IN X100, $ASSIGN-PRE-AND-POST a s s i g n s the node a t t r i b u t e PRECONJEmM t o t h e curr e n t node. $ASSIGN-PRE-AND-POST i s d e f i n e d i n r o u t i n e ~~~POST-CONJELEM which w i l l be d e s c r i b e d l a t e r . 8 f t e r t h e node a t t r i b u t e PRECONJBLEM i s a s s i g n e d , $ASSIGN-PRECONJELEM goas t o t h e c o r r e s p o n d i h y p r c -c o n j u n c t i o n a l I I Q~G by QXCcuting SPRECONJ. R o r example i n F i g . 5 , if CQELEMENT(0BJECT) i s c a l l e d fuon t h e second VERB, COELtEMENT w i l l c a l l $TO-PRECONJUNCTIONbY t o go t o t h e f i r s t VERB; t h e n kt : w i l l try t o go l e f t or r i g h t to l o c a t e OBJECT. Another t y p e of a d j b s t m e n t i s needed f o r r e s t r i c t i o n r o u t i n e s in c a t egory 3 above. The problem o c c u r s when a r e s t r i c t i o n is e x e c u t e d s t a r t i n g a t the c o n j u n c t i o n a 1 s t r i n g Q-CONJ. When t h e a e f i n i t i o n f o r Q-CONJ i s g e n e r a t e d f r o m t h e e l e m e n t s o f a s t r i n g , t h e w e l l -f o r m e d n e s~ r e s t r i c t i o n s housed in t h e ebernents are t r a n s m i t t e d a l o n g w i t h t h e elements. The r e s t r i c t i o n s on those e l e m e n t s , t h e r e f o r e , were w r i t t e n w i t h t h e a s s u m p t i o n t h a t t h e s t a r t i n g p o i n t i s the s t r i n g t h a t t h e r e s t r i c t i o n s were o r i g i n a l l y housed in--i.e., two nodes up from (2-CONJ, For example i n F i g , 5 , a l l r e s t r i c t i o n s i n ASSERTION assume to s t a r t a t ASSERTION, Thus, t h e same r e s t r i c t i o n , s t a r t i n g at Q-CONJ would f a i l i f , f o r example, w e were t o t e s t whether t h e immediate-node of, the second VERB i s ASSERTION. T h e r e f o r e , t h e r o u t i n e s i n category 3 execute $UP-THROUGH-Q i n i t i a l l y : $UP-THROUGH-Q = ITERATET $GO-UP-TWICE UNTIL TEST FOR Q-CONJ FAILS. $GO-UP-TWICE = GO UP; GO UP. $UP-THROUGH-Q g o e s t o t h e node which i s two nodes up from t h e t o p o f a n e s t o f Q -C O N J t s . R o u t i n e PRE-POST-CONJELEM a s s i g n s the node a t t r i b u t e PRECOKJELEM t o the e l e m e n t s o f Q-CONJ, I t i s assumed t h a t t h e s t a r t i n g node i s t h e node above Q-CONJ. To each elemen+ of Q-CONJ that is not on the C-NODE list (ANDSTG, ORSTG, BUTSTG, e t c . ) , it assigns t h e node attribute PRECOVJELEM. L i k e w i s e the corresponding pre-conjunction elements will. be assigned the n~d e attribute POSTCONJELEM. ROWLNE PRE-POST-COMJELEM = STORE I N X100; DO EJXMENT-(Q-CONJ) ; DO LAST-ELEMENT-; ITERATEVERIFY $ASSIGN-TESTUNTILGOLEFTFAXLS, $ASSIGN-TEST = E I T H E R TEST FOR C-NODE OR E I T H E R SPECONJ [COELl-] OR $ASSIGN-PE-AND-POST. The routine PW-POST-CONJELEM saves the starting point in register X100. It then uses the nonstacking routines* to go to the last element of Q-CONJ. The node attribute assignments start from the rightmo~t node of Q-CONJ and proceed left. $ASSIGN-TEST determines whether or not a node attribute should he assigned to a part.icular node. An assignment i s n o t n e c e s s a r y if t h e node is on the C-NODE List or if the assignment w a s already made for the node. If an assignment daes not have to be made,$ASSIGN-TEST is finished; if one has t o be made $ASSIGN-PW-AND-POST is executed: $ASSIGN-PRE-AND-POST = STOm I N XSOO; For each routine t h a t stacks, there i s a nonstacking counterpart, The nonstacking routines are used for restrictions where s t a c k i n g i s not desired-such a s the number agreement restrictions. $GO-LEFT = ITERATET SUPCONJ U N T I L GO LEFT SUCCEEDS; STORE XN X100. SUPCONJ = GO U P ; TEST FOR Q -C W J ; GO UP. $ASSIGN-PRE-AND-POST s a v e s t h e node t o be assigned i n r e g i s t e r s X 5 Q O and XO.It then goes to t h e node saved i n X l O O (which i s i n i t i a l l y t h e s t a r t i n y C -N W E ) and locates t h e corresponding pre-conj~nctional~element by e x e c u t i n g $GO-LEFT u n t i l i t f i n d s a node which h a s the same name as that i n register S5OO.That node i s saved i nX l O O so t h a t t h e s e a r c h starts t h e r e f o r t h e next node to be assigned. $ASSIGN-POSTCBNJEmbl and $ASSIGN-PRECONJELEM a s s i y n the node attributes. W e w i l l c o n s i d e r t h e case where t h e second OBJECT i n F i g . 8 i s b e i n g a s s i g n e d node a t t r i b u t e PRECONJELEM. BUTSTG i s saved i n r e g i s t e r X100. $ASSIGN-PRE-AND-FOST s a v e s t h e second OBJECT i n r e g i s t e r s X O arid X500. I t then s e a r c h e s f o r t h e corresponding p r e -c o n j u n c t i o n a l element by going l e f t from B~S T G . The f i r s t OBJECT i s found. $POSTCONJ f a i l s a t t h e f i r s t OBJECT and $ASSIGN-PRE-AND-POST remains t h e r e . Node a t t r i b u t e POSTCONJELEM i s a s s i g n e d t o t h e f i r s t OBJECT by $ASSIGN-POSTCONJELEM: $ASSIGN-POSTCONJELEM = A S S I G N THE PRESENT ELEbIENT NODE ATTRIBUTE POSTCONJELEM. $ASSIGN-PRECoNJELI2M = AS$ZGN THE PRESENT ELEMENT NODE ATTRIBUTE PRECONJELEM . When node a t t r i b u t e POSTCONJELEM i s a s s i g n e d , i f t h e r e i s a node saved i n r e g i s t e r X 0 , t h a t n o d e w i l l a u t o m a t i c a l l y b e a s s i g n e d a s t h e v a l u e o f a t t r i b u t e POSTCONJELEMbythenode a t t r i b u t e assignment o p e r a t o r . I n t h i s c a s e t h e second OBJECT i s i n r e g i s t e r XO. Therefore the f i r s t OBJECT i s a s s i g n e d node a t t r ibute POSTCONJELEM with t h e second OBJECT a s i t s v a l u e . A f t e r a node attrib u t e assignment i s made, r e g i s t e r XO i s a u t o m a t i c a l l y c l e a r e d by t h e program. This p r e v e n t s a c c i d e n t a l v a l u e assignments from o c c u r r i n g i n case t h e grammar w r i t e r forgets t o clear the register. I n t h e above example, a f t e r node a t t r i b u t e POSTCONJELEM i s a s s i g n e d t o t h e f i r s t OBJECT, the f i r s t OBJECT i s saved i n r e g i s t e r XO by $ASSIGN-PIG-AND-POST a n d t h e sebond OBJECT s a v e d i n XSOO i s a s s i g n e d the n o d e a t t r i b u t e PIJECONJELIEM with the first OBJECT as its v a l u e . Sometimes a r o u t i n e s t a r t s i n a n e s t o f Q-CON3 n o d e s ; the c o r r c s l~o n d i n q p r e -c o~i~i n c t i o n a l e l e m e n t i s not n e c e s s a r i l y locatecj o n the next h i q l~e t r level. I n F i g . 9 when the second LN (a ---f e w ) is beinq a s s i c j n r d node a t t l r ib u t 6 PFECONJEI,EM, Pra-POST-CONJELEM h a s to (70 ul) ttm Q-CCjN,T I e v e~l : ; tt3 f i~l d the c o r r e s p o n d i n g LN. I n SGO-LEFT, i f t h e ca1-1-esp1ontIi nq n o d e i s n o t o n t h c l e v e l b e i n g s e a r c h e d , SUPCONJ i s e x e c u t e d t o l o c a t e t h e n e x t l e v e l . I n Chc\" above e x a m p l e , when $GO-LEFT c a n n o t g o l e f t f ronl t h e s e c o n d NVAR ( w o m e n ) , --$UP-CONJ goes up t o t h e n e x t h i g h e r Q-CONJ a n d t h e n goes u p t o t h e next C-NODE where the search f o r a c o r r e s p o n d i n g node resumes. Thus L N (a f e w ) i s a s s i g n e d node a t t r i b u t e PRECONJELEM w i t h LN ( a l l t h e ) a s i t s v a l u e . F I G U R E 9 P a r s e t r e e of A l l t h e men and women a n d a f e w children were h u r t . i g . 8, when t h e third v e r b ( i g n o r e d ) is being a s s i g n e d node attrib u t e s by $ASSIGN-PRE-AND-POST, t h e r e s t r i c t i o n i n t e r p r e t e r goes l e f t from BUTSTG and a r r i v e s a t the f i r s t VERB. The node a t t r i b u t e s are chained. Theref o r e , node a t t r i b u t e POSTCONJEmM of t h e f i r s t VERB sHould have t h e second VERB a s its v a l u e and node a t t r i b u t e POSTCONJELm of t h e second v e r b should have t h e third VERB as i t s v a l u e . $ASSIGN-PRE-AND-POST gets t h e l a s t node of t h e chain a f t e r it a r r i v e s a t t h e f i r s t VERB. T h i s i s done by the s e c t i o n of code . . . EITHER ITERATE $POSTCONJ [STARTAT] OR TRUE; . . . , I n this c a s e VERB ( i g n o r e d ) i s a s s i g n e d node a t t r i b u t e PRGCON3ELE.M with v a l u e VERB ( p r i n t ) . 1. F i t z p a t r i c k , E. and N. S a g e r , The L e x i c a l S u b c l a s s e s of t h e L i n g u i s t i c St-ring P a r s e r , American J o u r n a l of Computational L i n g u i s t i c s ; microf i c h e 2, 1974, 2. Grishman, R . , The Implementation of the S t x i n g Parser of English. In N a t u r a l Language P r o c e s s i n q , R. ust tin, e d . , Algorithmics Press, New York, 1973. 3 . Grishman, R* , N. Sager, C. Raze, and B. Bookchin, The L i n g u i s t i c S t r i n g Parser. Proceedings of the 1973 N a t i o n a l Computer Conference, 427-434, AFIPS P r e s s , 1973. 4. Harris, Z.S., String A n a l y s i s of S e n t e n c e S t r u c t u r e , Mouton & C o . , The Hague, 1962. 5. Raze, C., The FAP Program for S t r i n g Decomposition o f S c i e n t i f i c T e x t s . S t r i n g Program Reports ( S . P . R , ) N o . 2 , Linguistic S t r i n g P r o j e c t , New York U n i v e r s i t y , 1967. 6. Sager, N., S y n t a c t i c A n a l y s i s of Natural Language. Advances in Computers, v o l . 8, 153-158, Academic P r e s s , I n c , , New York, 1967.7. Sager, N., A Computer S t r i n g Grammar of E n g l i s h , S.P.R. N o . 4 , L i n g u i s t i c S t r i n g P r o j e c t , New York U n i v e r s i t y , 1968.",
                "num": null
            },
            "FIGREF16": {
                "uris": null,
                "type_str": "figure",
                "text": "This work was supported in part by Research Grants GS2462 and GS27925 from the National Science Foundation, Division of Social Sciences, and in part by Research Grant GN39879 from the Office of Science Information Service of the National Science Foundation.APPENDIX -The LSP p a r s i n g SystemThe LSP system o b t a i n s a s u r f a c e s t r u c t u r e a n a l y s i s i n t h e form of a s t r i n g decomposition of a sentence. I n accordance with l i h g u i s t i c s t r i n g theory (Harris 1962), each sentence is composed of elementary word sequences of a few given t y p e s , s t a t a b l e a s word c l a s s sequences ( c a l l e d l i n g u i s t i c strings). Each sentence contains one c e n t e r s t r i n g (an elementary sentence) and z e r o o r more adjunct s t r i n g s , adjoined t o t h e l e f t o r r i g h t of elements of t h e c e n t e r s t r i n g o r of o t h e r a d j u n c t s . I n a d d i t i o n , s t r i n g occurrences may be r e s t r i c t e d w i t h regard t o t h e subclasses of words t h a t can co-occur i n t h e s m e s t r i n g o r i n adjoined s t r i n g s . A s t r i n g as a whole may a l s o have a d j u n c t s t r i n g s ( c a l l e d sentence a d j u n c t s ) which occur a t s t a t e d p o i n t s i n t h e s t r i n g .",
                "num": null
            },
            "FIGREF17": {
                "uris": null,
                "type_str": "figure",
                "text": "A1 i s an example of t h e computer o u t p u t of t h e s t r i n g decomposit i o n of One rumor h a s t i l y p r i n t e d can r u i n c a r e e r s . Line 2 i n F i g . A 1 shows t h a t rumors can r u i n c a r e e r s i s t h e c e n t e r s t r i n g which has t h e form of an a s s e r t i o n . Rumor has a l e f t a d j u n c t s t r i n g LN whose decomposition i s shown on l i n e 3 and a r i g h t a d j u n c t s t r i n g RN whose decomposition i s shown on l i n e 4. LN c o n s i s t s of t h e q u a n t i f i e rone. RN i s t h e p a s s i v e s t r i n g c a l l e d VENPASS which c o n s i s t s of t h e past p a r t i c i p l e p r i n t e d preceded by t h e a d j u n c t h a s t i l y . I n this example t h e o b j e c t p o s i t i o n (PASSOBJ) a f t e r t h e verb i s n u l l .",
                "num": null
            },
            "FIGREF18": {
                "uris": null,
                "type_str": "figure",
                "text": "1968), b o t h o f w h i c h a r e g e a r e d t o h a n d l e E n g l i s h s c i e n t i f i c t e x t s . Theword d i c t i o n a r y a s s i g n s t o e a c h w o r d i t s m a j o r s y n t a c t i c c a t e g o r i e s , e.g., noun,verb, a d j e c t i v e , e t c . , w h i c h m a y i n t u r n h a v e subcategories. he grammar c o n s i s t sFIGURE A 1 Computer Output of t h e S t r i n g Decomposition of One rumor h a s t i l y p r i n t e d can r u i n c a r e e r s . 1. SENTENCE = INTRODUCER CENTER ENDMARK 2. 2. ASSERTION = S A SUBJECT SA TENSE S A VERB S A OBJXCT RV S A a r t s : a c o n t e x t -f r e e component and a s e t of r e s t r i c t i o n s . The cont e x t -f r e e component d e f i n e s t h e s e t s of c e n t e r and a d j u n c t s t r i n g s of t h e grammar. The d e f i n i t i o n s are w r i t t e n i n Backus Normal Form, A n example of a s t r i n g d e f i n i t i o n i s : <ASSERTION> ::= <SA><SUBJECT><SA><TENSE><SA><VERB><SA><OBJECT><RV><SA>. Each of the elements of ASSERTION i s a l s o d e f i n e d i n t h e grammar. I n t h e above example, S A (sentence a d j u n c t ) and RV ( p o s t -o b j e c t r i g h t a d j u n c t of verb) are a d j u n c t sets; t h e r e f o r e t h e i r occurrence i n a sentence i s o p t i o n a l . SUBJECT, TENSE, VERB, and OBJECT are p o s i t i o n s corresponding t o r e q u i r e d elements of t h e s t r i n g . Each p o s i t i o n may have a l t e r n a t e v a l u e s i n d i f f e r e n t sentences.",
                "num": null
            },
            "FIGREF19": {
                "uris": null,
                "type_str": "figure",
                "text": "The t r e e r e p r e s e n t s t h e p a r t i c u l a r combination of s t r i n g s and a d j u n c t s whose t e r m i n a l nodes combine t o produce a well-formed s e n t e n c e , o r more e x a c t l y a well-formed sequence of word c a t e g o r i e s which match t h o s e of t h e seritence words. A p a r s e t r e e of ASSERTION f o r One rumor h a s t i l y p r i n t e d can r u i n careers i s shown i n F i g . A2. The computer o u t p u t i nF i g .~l i s a compressed v e r s i o n o f t h e p a r s e t r e e i n F i g . A2. I n t h i s p a r s e t r e e t h e elements of a s t r i n g a r e shown a s a sequence of connected s i b l i n g nodes. Thus t h e e l ements of ASSERTION a r e shown one l e v e l below t h e ASSERTION node, i n t h e o r d e r i n which they a p p e a r i n t h e d e f i n i t i o n of ASSERTION, The t e r m i n a l nodes o f t h e tree a r e e i t h e r word class symbols which correspond t o s e n t e n c e words (*N = rumors) or are n u l l nodes. The n u l l nodes a r e a u t o m a t i c a l l y s a t i s f i e d w i t h o u t sthsuming s e n t e n c e words. I n a d j u n c t s e t p o s i t i o n s they r e p r e s e n t t h e f a c t t h a t a d j u n c t o c c u r r e n c e s a r e o p t i o n a l . ( I n t h e par~se t r e e diagrams t h e n u l l nodes are o m i t t e d . ) A s t a n d a r d t y p e of s t r u c t u r e t h a t i s f r e q u e n t l y seen i n t h e t r e e i s c a l l e d t h e \"LXR\" node. A n example is t h e LNR node i n F i g . A2. An LXR d e f in i t i o n c o n s i s t s of t h r e e elements: a p o s i t i o n for t h e l e f t a d j u n c t s of X , a c o r e p o s i t i o n f o r t h e word c l a s s X , and a p o s i t i o n f o r t h e r i g h t a d j u n c t s of X. The c o r e p o s i t i o n as a r u l e subsumes a s e n t e n c e word of c l a s s X . For example, i n F i g . A2,in t h e SUBJECT of ASSERTION, NVAR i s t h e c o r e p o s i t i o n of LNR and h a s t h e v a l u e N corresponding t o rumor. NVAR (Noun -V a r i a n t s ) w i l l have one of s e v e r a l a l t e r n a t e v a l u e s , namely noun, pronoun, Ving, e t c . The \"LXR\" t y p e s t r u c t u r e i s i m p o r t a n t i n that b o t h t h e r e s t r i c t i o n s and t h e conj u n c t i o n mechanism depend on t h i s r e g u l a r i z e d r e p r e s e n t a t i o n o f an element and i t s a d j u n c t s . The r e s t r i c t i o n s are a s e t o f d e t a i l e d well-formedness r u l e s which must be s a t i s f i e d b e f o r e a n a n a l y s i s i s a c c e p t e d . The r e s t r i c t i o n s may be s t r i c t l y grammatical, such a s one governing t h e c a s e o f a pronoun. Such a r e s t r i c t i o n w i l l succeed f o r They r u i n e d c a r e e r s b u t n o t f o r Them r u i n e d C a r e e r s . O r t h e",
                "num": null
            },
            "FIGREF20": {
                "uris": null,
                "type_str": "figure",
                "text": "Parse t r e e * E~ One rumor h a s t i l y p r i n t e d chn r u i n c a r e e r s . The t e r m i n a l nodes which are n u l l a r e n o t shown i n t h e p a r s e t r e e diagrams.I n o r d e r t o keep t h e t r e e diagrams a s u n c l u t t e r e d as p o s s i b l e t h e d e t a i l s of c e r t a i n s u b s t r u c t u r e s have been o m i t t e d b u t t h e s e n t e n c e w o r d( s ) subsumed by t h o s e s u b s t r u c t u r e s a r e shown. Three v e r t i c a l d o t s below which a r e s e n t e n c e words s i g n i f i e s such an omission. r e s t r i c t i o n s may e x p r e s s s e l e c t i o n a l c o n s t r a i n t s ; t h e s e w i l l succeed f o r sequences t h a t are c o n s i d e r e d p o s s i b l e w i t h i n a g i v e n a r e a of d i s c o u r s e . E g g . , i n normal d i s c o u r s e one can say They p r i n t e d rumors but n o t They p r i n t e d c r i t i c s . Each r e s t r i c t i o n i s compiled i n t o a s e r i e s of b a s i c o p e r a t i o n s and t e s t s which a r e performed on the p a r s e t r e e w h i l e p a r s i n g a s e n t e n c e . The r e s t r i c t i o n s mainly t e s t c o n d i t i o n s between t w o elements of a s t r i n g o r between a s t r i n g element and i t s a d j u n c t .",
                "num": null
            },
            "FIGREF21": {
                "uris": null,
                "type_str": "figure",
                "text": "i g u r e~3 i l l u s t r a t e s two of t h e r e l a k i o n s used by r e s t r i c t i o n s . For t h e Verb-Object s e l e c t i o n a l r e s t r i c t i o n of t h e grammar t o o p e r a t e on They p r i n t e d rumors, t h e coelement r e l a t i o n i s used t o t e s t whether t h e co-occurrence of t h e o b j e c t (rumors) and t h e v e r b ( p r i n t e d ) i s well-formed. O n t h e o t h e r h a n d , i n Rumors hastily p r i n t e d can r u i n careers t h e h o s t -a d j u n c t r e l a t i o n i s used t o test t h e noun rumors and t h e v e r b p r i n t e d s i n c e t h e v e r b h e r e a p p e a r s i n t h e r i g h t a d j u n c t s t r i n g of rumors. Because t h e r e s t r i c t i o n s u s e t h e s e b a s i c s t r i n g r e l a t i o n s (coelement, l e f t -a d j u n c t , r i g h t -a d j u n c t ) s o o f t e n t o t e s t f o r w e l l formed sequences, t h e s e r e l a t i o n s a r e encoded i n t o b a s i c r o u t i n e s (COELEMENT, LEFT-ADJUNCT, RIGHT-ADJUNCT, e t c . ) , which i n t u r n use b a s i c t r e e o p e r a t i o n s (up, down, l e f t , r i g h t , t e s t f o r x , e t c . ) , For example, i n t h e LNR sequence subsuming one rumor h a s t i l y p r i n t e d , shown i n Fig.A2, t h e r o u t i n e RIGHT-ADJUNCT goes from t h e c o r e noun rumor t o i t s r i g h t -a d j u n c t s t r i n g h a s t i l y p r i n t e d . FIGURE A 3 R e s t r i c t i o n s U s e S t r i n g R e l a t i o n s t h e b a s i c r o u t i n e s by t h e r e s t r i c t i o n s g r e a t l y f a c i l i t a t e s t h e f o r m u l a t i o n o f the r e s t r i c t i o n s . One r o u t i n e c o n t a i n s many tree operat i o n s . The u s e of t h e s e r o u t i n e s a l s o f a c i l i t a t e s m o d i f i c a t i o n of the grammar. I f t h e r e i s a b a s i c change i n t h e grammar, t h e a f f e c t e d r o u t i n e s a r e changed b u t t h e r e s t r i c t i o n s themselves do n o t have t o be. The e x t e n s i v e use of r o u t i n e s by the r e s t r i c t i o n s p l a y s a s i g n i f i c a n t r o l e i n t h e t r e a t m e n t of c o n j u n c t i o n s , as w i l l b e d e s c r i b e d l a t e r . B o t h t h e r e s t r i c t i o n s and t h e r o u t i n e s arewritteninaprogramminglanguage d e v e l o p e d f o r t h e L S P (Sager and Grishman1975). The s y n t a x o f t h e r e s t r i c t i o n l a ng u a g e i n c l u d e s t h r e e m a i n s t a t e r n e n t t y p e s . O n e p a r t o f t h e language i s s i m i l a r t o a s u b s e t o f English i n t h a t t h e s t a t e m e n t s c o n s i s t of a s u b j e c t followed by a p r e d i c a t e , For example, THE CORE O F T H E SUBJECT I S PRONOUN, A n o t h e r part of t h e r e s t r i c t i o n language c o n s i s t s of l o g i c a l c o n n e c t i v e s , such a s IF THEN",
                "num": null
            },
            "FIGREF22": {
                "uris": null,
                "type_str": "figure",
                "text": "4 , t h e string NSVINGO s a t i s f i e s t h e SUBJECT O F ASSERTION. T h i s s i t u a t i o n i s p r o v i d e d f o r by $DESCEND-TO-STRING.",
                "num": null
            },
            "FIGREF23": {
                "uris": null,
                "type_str": "figure",
                "text": "w i l l l o c a t e t h e s t r i n g NSVINGO. ELEMENT ROUTINE ROUTINE ELEWNT ( XI = EITHER DO DOWN1 (X) OR $STRING-SEGMENT. $STRING-SEGMENT = DO DOWN^ (STGSEG) ; ROUTINE DOWN1 (X) DO DOWN1 (X) . = GO DOWN; ITERATET GO RIGHT U N T I L TEST FOR X SUCCEEDS.",
                "num": null
            },
            "FIGREF24": {
                "uris": null,
                "type_str": "figure",
                "text": "t i s assumed t h a t ELEEENT s t a r t s a t node Y and t h a t X i s an element of t h e s t r i n g~c o r r e s p o n d i n g t o Y .Thus ELEMENT l o c a t e s X bv s e a r c h i n g t h e l e v e l below Y .T h i s i s done by r o u t i n e D O W N l ( X ) which f i r s t goes t o t h e l e v e l below Y by e x e c u t i n g t h e command GO DOWN and t h e n s e a r c h e s t h e nodes on t h a t l e v e l u n t i l it f i n d s X.The l a t t e r s t e p i s accomplished by an i t e r a t e command:ITERATET GO RIGHT UNTIL TEST FOR X SUCCEEDS.* I n F i g .A 2 t h e e x e c u t i o n of EL=NT(SUBJECT), s t a r t i n g a t t h e s t r i n g node ASSERTION, w i l l l o c a t e t h e node SUBJECT. Sometimes i t i s more c o n v e n i e n t t o d e f i n e a s t r i n g Y by u s i n g t h e name of a n o t h e r s t r i n g X i n t h e d e f i n i t i o n o f Y i n s t e a d of naming a l l t h e elements o f y. T h i s i s t h e c a s e i n F i g . A 4 w h e r e V I N G O , a d e f i n e d s t r i n g o f t h e grammar appears as a string segment of NSVINGO. I n t h i s s i t u a t i o n n o t a l lt h e e l e m e n t s of NSVINGQ are on one l e v e l below NSVINGO but some a r e one l e v e l \"ITERATET i s an ITERATE command i n which t h e e x i t t e s t ( t h e p a r t w r i t t e n a f t e r U N T I L ) i s t h e f i r s t o p e r a t i o n performed.",
                "num": null
            },
            "FIGREF25": {
                "uris": null,
                "type_str": "figure",
                "text": "Parse t r e e of His p r i n t i n g rumors r u i n e d careers. o w t h e s t r i n g s g g m e n t VINGO of NSVINGO. $STRING-SEGMEPJT, t h e r e f o r e , s e a r c h e s one l e v e l below Y for a node on t h e s t r i n g segment l i s t STGSEG and i f it f i n d s one, it searches f o r X one l e v e l below the s t r i n g segment. I n F i g . A 4 , ELEMENT(OBJECT), s t a r t i n g a t NSVINGO, f i r s t locates VINGO by exec u t i n g DOWNl(STGSEG) and then locates OBJECT by executing DOWNl(0BJECT).COELEMENT ROUTINE ROUTINE COELI (X)ROUTINE COELEMENT(X) = ONE OF $SAME-LEVEL, $x-IN-SEGMENT, $Y-IN-SEGMENT.",
                "num": null
            },
            "FIGREF26": {
                "uris": null,
                "type_str": "figure",
                "text": "s i t u a t i o n i s handled by $Y-IN-SEGMENT, F i r s t , t h e s t r i n g segment VINGO i s l o c a t e d by t h e sequence GO UP; TEST FOR STGSEG. Then TPOS is l o c a t e d by r o u t i n e COELl(TPOS), which s e a r c h e s the same l e v e l as VINGO t o f i n d TPOS. RIGHT-ADJUNCT ROUTINE ROUTINE RIGHT-ADJUNCT = DO RIGHT-ADJUNCT-PO$; DO CORE. ROUTINE RIGHT-ADJUNCT-POS = EITHER $ASCNT OR TRUE; Do RIGHTR (RADJSET) . SASCNT = ASCEND TO AVAR OR WAR OR QVAR OR WAR. I t i s assumed that RIGHT-ADJUNCT s t a r t s a t the c o r e of an LXR t y p e node. I t goes t o t h e c o r e o f t h e r i g h t -a d j u n c t p o s i t i o n i n t h e LXR sequence, For example, i n t h e LXR sequence LN NVAR F W (one rumor h a s t i l y p r i n t e d ) , shown i n F i g . A 4 the r s u t i n e RIGHT-ADJUNCT-POS, s t a r t i n g a t N (rumors) ascends t o NVAR",
                "num": null
            },
            "FIGREF27": {
                "uris": null,
                "type_str": "figure",
                "text": "e d ) i s l o c a t e d t h e r i g h t a d j u n c t of the noun rumor, -LEFT-ADJUNCT ROUTINE ROUTINE LEFT-ADJUNCT = DO LEFT-ADJUNCT-POS; EITHER TEST FOR LN OR DO CORE. ROUTINE LEFT-ADJUNCT-PQS = EITHER SASCNT OR TRUE; DO LEFTR (LADJSET) . LEFT-ADJUNCT i s similar t o RIGHT-ADJUNCT except it goes to t h e core of t h e l e f t -a d j h n c t positi\"bn of an LXR sequence. Xf t h e l e f t -a d j u n c t position is LN, however, t h e routine stops there since it is assumed that further operations w i l l be specified t o locate a particular left adjunct of the noun, e . g . a quantifier ox an adjective. HOST ROUTINE ROUTINE HOST = CORE OF THE HOST-ELEMF,NT EXISTS. ROUTINE HOST-ELEMENT = ONE OF $AT-LADJ, SAT-RADJ, $AT-EWSUBJ I S TRUE. $AT-LADJ = EITHER TEST FOR LADJSET OR ASCEND TO LADJSET; GO RIGHT. $AT-RADJ = EITHER TEST FOR RADJSET O R ASCEND TO RADJSET; EITHER $RV-IN-STRING OR GO LEFT. $RV-IN-STRING = TEST FOR RV; STORE. IN X100; GO UP; TEST FOR TYPE STRING: EITHER $RV-IN-OBJECT OR $RV-IN-CENTER. $RV-IN-OBJECT = TEST FOR NTOVO OR NTHATS OR NSNWH OR PNTHATS OR PNTHATSVO OR PNSNWH; ASCEND TO OBJECT; DO VERB-COELEMENT . $RV-IN-CENTER = GO TO X100; DO VERB-COELEMENT. $AT-RNSUBJ = EITHER TEST FOR RNSUBJ OR ASCEND TO RNSUBJ; ASCEND TO SA; DO COELEME' NT (SUBJECT) .I t i s assumed t h a t the r o u t i n e HOST-ELEMENT s t a r t s a t node Y , which i s i n o r a t a n a d j u n c t p o s i t i o n i n a n LXR s t r u c t u r e . I t goes from the a d j u n c t posit i o n t o t h e c o r e p o s i t i o n of t h e LXR s t r u c t u r e . HOST t h e n goes t o the CORE of t h e node l o c a t e d by HOST-ELEEiENT. For exan~ple, c o n s i d e r the o p e r a t i o n of the HOST r o u t i n e on t h e p a r s e t r e e shown i n ~i g . A 2 where HOST s t a r t s a t t h e r i g h tadjunct s t r i n g VENPASS = h a s t i l y p r i n t e d , $AT-LADJ f a i l s b u t $AT-RADJ ascends t o RN by e x e c u t i n g 8SCEND TO RADJSET and goes left t o W A R . The CORE r o u t i n e t h e n l o c a t e s N ( r u m o r s ) . A s i m i l a r s i t u a t i o n o c c u r s when HOST s t a r b s a t LN. TEST FOR LADJSET s u c c e e d s and HOST-ELEMENT goes one node t o t h e r i g h t to NVAR. HOST c a l l s tne CORE r o u t i n e which l o c a t e s N ( r u m o r s ) . I f the HOST r o u t i n e s t a r t s i n o r a t RV ( r i g h t a d j u n c t s of v e r b ) , e x t r a maneuvering is n e c e s s a r y t o l o c a t e the verb. T h e r e are t h r e e p o s s i b i l i t i e s .",
                "num": null
            },
            "FIGREF28": {
                "uris": null,
                "type_str": "figure",
                "text": "s case, $RV-IN-STRING f a i l s because t h e node above RV ( i . e . , '~'ERB) i s n o t on t h e s t r i n g l i s t . HOST-ELEMENT t h e r e f o r e goes l e f t t o W A R , whose c o r e i s r a n . I n some c a s e s , RV f o l l o w s t h e o b j e c t of t h e v e r b as i n H e r a n t o s c h o o l quickly, shown i n F i g . A 6 . I n t h i s c a s e , $RV-IN-STRING s u c c e e d s . T h e node &ve RV i s t h e string ASSERTION. $RV-IN-CENTER locates the v e r b a l element VERB (ran) o f ASSERTION by c a l l i n g r o u t i n e VERB-C~ELEMENT (a g e n e r a l i z e d rout i n e for finding a verbal c o e l e m e n t ) . I n o t h e r s e n t e n c e s , RV i s situated i n the m i d d l e of a n o b j e c t s t r i n g . I n She t o l d him q u i c k l y t h a t he had t o l e a v e , shown i n F i g . A 7 ,RV i s s i t u a t e d a f t e r NSTGO (him) i n the object NTHATS of t h e v e r b t o l d . I n t h i s s i t u a t i o n $RV-IN-OBJECT l o c a t e s t h e v e r b a l element of ASSERTION by a s c e n d i n g t o OBJECT and c a l l i n g r o u t i n e VERB-COELEMENT.",
                "num": null
            },
            "FIGREF29": {
                "uris": null,
                "type_str": "figure",
                "text": "E R T I O N parse tree of She t o l d him quickly that he had t o l e a v e . arises in sentences where the right a d~u n c t df t h e subject noun does not immediately follow the noun. In the grammar t h e s e occurrences are covered by R N S U B J in t h e post-OBJECT s e n t e n c e adjunct posltion",
                "num": null
            },
            "TABREF0": {
                "content": "<table><tr><td>Applications. Revised from -SPR 4 (L968).</td></tr><tr><td>9. Sager, N., The String Parser for Scientific Literature, In Natural</td></tr><tr><td>Language Processing, R. Rustin, ed., Algorithmics Press, New York, 1973.</td></tr><tr><td>10. Sager, N. and Ralph Grishman, The Restriction Language for Computer</td></tr><tr><td>Grammars of NatuYal Language. Communi,c~tions of the ACM, 18, 390-4Q0,</td></tr><tr><td>1975.</td></tr><tr><td>11. Sager, N., Salkoff, M., Morris, J., and Raze, C., Report on the String</td></tr><tr><td>Analysis Programs, Introductory Volume. String Program Reports No. 1,</td></tr><tr><td>L i n g u i s t i c S t r i n g Project, New York University and University of</td></tr><tr><td>Pennsylvania, March 1966.</td></tr></table>",
                "type_str": "table",
                "html": null,
                "text": "8. Sager, N., A Computer Grammar of English and Its Applications, to be published by Gordon & Breach i n the series Mathematics and Its",
                "num": null
            },
            "TABREF1": {
                "content": "<table><tr><td>CORE ROUTINE</td></tr><tr><td>ROUTINE: CORE</td></tr></table>",
                "type_str": "table",
                "html": null,
                "text": "restriction statements. For example, IF THE CORE OF THE SUBJECT IS PRONOUN X1THEN Xl IS NOT ACCUSATIVE.Another type of restriction language statement consists of a series of commands which are used mainly for writing the routines. A command may consist of a basic tree operation such as GO UP, or a call to execute a restriction routbg, such as DO ELEMENT(X), or a call to execute another restriction statement, such as DO $1. There are provisions for saving a node in a register and for restoring a node from a register. For example, STORE IN X1, GO TO X1. The commands may also be logically combined.BASIC ROUTINES OF THE GRAMMARThere are about thirty basic routines in the grammar. Described here in detail are the ones which represent the major grammatical relations among words in a sentence and are important in the treatment of conjunctions.",
                "num": null
            }
        }
    }
}