Base_JSON/prefixJ/json/J76/J76-4007.json

{
    "paper_id": "J76-4007",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T02:51:21.937718Z"
    },
    "title": "P R O C E S S I N G C A S E Y O R I C K WILKS PRESENT A D D R E S S : DEPARTMENT O F LANGUAGE AND LINGUISTICS, UNIVERSITY OF ESSEX",
    "authors": [
        {
            "first": "E S S E X",
            "middle": [],
            "last": "Colchester",
            "suffix": "",
            "affiliation": {},
            "email": ""
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "",
    "pdf_parse": {
        "paper_id": "J76-4007",
        "_pdf_hash": "",
        "abstract": [],
        "body_text": [
            {
                "text": "Thls a n a l y s i s can be c o n t r a s t e d w i t h \" h i t \" , whose frame would be",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "( -------OBJECT (AGENT 1 INSTRUMENT) )",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "where t h e o v e r l a p p i n g b r a c k e t s mean that t h e two c a s e s , a g e n t and i n s t r u m e n t a l a r e semi-optional, i n t h a t a t l e a s t one of t h e two must appear.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "The window h i t",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "would b e ill-formed on t h a t view because, a l t h o u g h i t c o n t a i n s t h e o b l i g a t o r y o b j e c t , i t c o n t a i n s n e i t h e r t h e a g e n t nor the i n s t r u m e n t of t h e h i t t i n g , However, s e l e c t i o n r u l e \" , which s a y s t h a t , i f t h e r e i s an a g e n t , t h a t w i l l be t h e s u b j e c t of any a c t i v e sentence produded; i f t h e r e i s no agent b u t t h e r e i s a n i n s t r u m e n t , then t h a t w i l l be t h e s u b j e c t and s o on.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "I n t h i s paper, I Ishall c o n c e n t r a t e , a s i s normal i f n o t d e s i r a b l e i n A 1 and computational l i n g u i s t i c s , Qn q u e s t i o n s of a n a l y s~s r a t h e r than g e n e r a t i o n .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "The g e n e r a l p r o b l e n i n a n a l y s i s ( t h a t i s n o t emphasised i n g e n e r a t i o n ) i s t h a t of t h e s e l e c t i o n r e s t r i c t i o n s on t h e c a s e s , which can be i l l u s t r a t e d by l o o k i n g a t t h e s i m p l e s t of t h e systems u s i n g case a n a l y s i s (Simmons 1973) . Simmons wishes t o map a l l of t h e sentences ( 4 ) (7) onto t h e same semantic network because a11 f o u r , i n some s e n s e , r e f e r t o t h e szme e v e n t : Murcia (1972) . This paradigm, f o r a c t i v e f o r m s of a verb l l k e \"break\", has the form:",
                "cite_spans": [
                    {
                        "start": 319,
                        "end": 333,
                        "text": "(Simmons 1973)",
                        "ref_id": null
                    },
                    {
                        "start": 514,
                        "end": 520,
                        "text": "(1972)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "(8) AGENT * OBJECT INSTRUMENT AGENI * CBJECT INSTRUMEIJT * OBJECT OBJECT A",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "The l i n e s of (8) a r e p a t t e r n s t h a t must match i n p u t word s t r i n g s i n l e f t r i g h t o r d e r s o a s t o a s s i g n the c a s e s they c o n t a i n , The l i n e s of (8) match each of (4) -(7) i n t u r n , where marks t h e p o s l t i o n of t h e v e r b (break) i n each l i n e of t h e paradigm, The l i n e s of (8) a r e no more than t h e p o s s i b l e case combinations a l l o w e d by the case frame 11 f o r \"break\" t o g e t h e r with an a n a l y t i c v e r s i o n of t h e s u b j e c t s e l e c t i o n r u l e \" , which a l w a y s makes the Agent t h e f i r s t ( s u b j e c t ) i t e m I n any l i n e of t h e paradigm (8) i n which i t occurs, There i s no need f o r t h e l i n e s of (8) t o be ordered In t h e i r a p p l i c a t i o n t o i n p u t s e n t e n c e s , although t h e r e i s one additional. item of information required before they can be applied a t a l l : t h e s e l e c t i o n r e s t r i c t i o n s , These t e l l us what i t i s t o b e an agent of \"break\": i n S i m o n s ' s scheme a noun marked ANIMATE, The s e l e c t i o n r e s t r i c t i o n s attached t o t h e cases i n (8) a r e e s s e n t i a l t~ t h e a p p l i c a t i o n of t h e paradlgrn, f o r only thus could w e know t h a t \"~o h n \" i n (4) was matched by ACZENT i n the f l r s t l i n e of (8).",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "It should be noted t h a t t h e r e s t r i c t i o n of AGENT i n (8) Conversely, the ANIMATE r e s t r i c t i o n on AGEPU' Ts i n (8) IS n o t n e c e s s a r i l y on AGENTS a s such although i t might t u r n o u t t o be so, These p o i n t s w i l l be important when we come t o Chamiak's arguments l a t e r .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "Notice t o o , t h a t t h e r e a r e n o t two d i f f e r e n t ways I n which a sentence can be ill-formed with r e s p e c t to the paradigm: one with r e s p e c t t o s e l e c t i o n r e s t r i c t i o n s and one with r e s p e c t t o the case frame (as has been argued by Bruce 1975) . ",
                "cite_spans": [
                    {
                        "start": 285,
                        "end": 296,
                        "text": "Bruce 1975)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "John broke i s ill-formed with r e s p e c t to the case frame f o r \"break\" a s explained e a r l i e r , However, i f w e look a t t h a t f a c t i n procedural /terms, such as t h o s e p r o v i d e d by Simmons' paradigm, w e c a n n o t deem t h a t f a i l u r e a s one of matching a l i n e o f (8) a s d i s t i n c t from (Bruce's v i e w ) n o t meeting t h e s e l e c t~o n r e s t r~c t i o n s PHYSOB , say, on t h e c a s e OBJECT on t h e f o u r t h l i n e of (8). The o n l y way i n which a sentence c o u l d i n d e p e n d e n t l y f a l l t o match any l i n e i n t h e paradigm of (8) would b e t h e t r i v i a l one of h a v i n g some Semantic n e t s do n o t immediately suggest t h e i r a s s o c i a t e d processes, whereas t h e r e p r e s e n t a t i o n s h e r e a r e intended t o be d i r e c t e d towards t h e processes t h a t o p e r a t e on them.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Thus",
                "sec_num": null
            },
            {
                "text": "There i s a c l e a r d i s t i n c t i o n i n t h e p r e s e n t system of represent-a t i o n between the knowledge s t o r e d and t h e p a t t e r n s sought i n language, on t h e one hand, and t h e language t e x t a c t u a l l y represented found, on t h e o t h e r ; where the l a t t e r may not e x a c t l y match what was being looked c)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "The present system of r e p r e s e n t a t i o n i s intended t o be more \"habitable\" i n Watt's (1968) sense of providing a language of semantic r e p r e s e n t a t i o n t h a t i s a p p r o p r i a t e t o the way humans express themselves.",
                "cite_spans": [
                    {
                        "start": 91,
                        "end": 104,
                        "text": "Watt's (1968)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "That i s a highly s u b j e c t i v e n o t i o n , perhaps, b u t h e r e i t i s taken to r e q u i r e a t l e a s t a dynamrc, o r readable s t r u c t u r e , which n e t s do not haverThj srequirement r e s t s upon another assuniption: t h a t our r e p r e s e n t a t i o n must have t h e \"one t h i n g a f t e r another\" f e a t u r e t h a t t e x t , $ have, r a t h e r than being s t a t i c and timeless l i k e most semantic nets (thoughdNorman and Rumelhart (1975) The case p r i m i t i v e s f u n c t i o n within a semantic dependency grammar (Hays ? 9 6 4 ) , intended t o express t h e meaning of word senses and, by extension, of t e x t s .",
                "cite_spans": [
                    {
                        "start": 448,
                        "end": 483,
                        "text": "(thoughdNorman and Rumelhart (1975)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "Each of the casg p r i m i t i v e s above w i l l have a dependent, which i s a type of e n t i t y f o r a l l t h e case prirfii t i v e s except WAY and GOAL, which take an a s s e r t i o n a s dependent,",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "The case p r i m i t i v e and i t s dependent ( e n t i t y o r a s s e r t i o n ) f o m a c a s e group which i s i n t u r n dependent on a p r i m i t i v e a c t i o n (except f o r WITH and POSS which depended on an e n t i t y",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": ", and may t h e r e f o r e .be o n l y semi-cases). This is best s e e n by example of t h e first s t r u c t u r e i n the system, t h e formulawhich e x p r e s s e s word s e n s e i n t h e d i c t i o n a r y , The formula f o r t h e action sense of \"break\" i s as f o l l o w s :",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "The g e n e r a l structure of such formulas h a s been e x p l a i n e d In Wifks (1968, 1972, 1975a, 1975b ).",
                "cite_spans": [
                    {
                        "start": 83,
                        "end": 108,
                        "text": "(1968, 1972, 1975a, 1975b",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "b)",
                "sec_num": null
            },
            {
                "text": "f n t e r l i n g u a l meaning of t h e sense of t h e word, and t h e p r i m i t i v e s t h a t comprise them are i n t e h d e d t o be i n t e r l i n g u a l (as a r e F i l l m o r e ' s c a s e s ) even though they happen t o be mostly Anglo-Saxon monosyllables. ii) The dependency w i t h i n a group 1s i n t e r p r e t e d d i f f e r e n t l y accord- f o r now t h e formula i s t o b e taken as no more t h a n a formal expression of t h e meaning of t h e a c t i o n 'break' t h a t can be used i n subsequent It i s t h i s f e a t u r e of t h e system t h a t e x p l a i n s why t h e head, o r p r i n c i p a l , element of a formula i s e a s i l y a c c e s s i b l e a t one end QE i t --rather than b u r i e d i n t h e c e n t e r a s i t would b e i f t h e formula w2reL i n SVO form, r a t h e r t h a n SOV form, a s above, i t s l e f t -which of course, a s i n t h e c a s e of 'break' may t u r n out t o y i e l d t h e same e n t i t y a s t h e preferred agent of t h e whole formula, though t h i s need n o t be t h e case. Moreover, i n t h e case of a c t i o n s w i t h i n a formula (i.e. n o t c o n s t i t u t i n g t h e head) t h e agent need not be marked though t h e o b j e c t must be i f i t i s an e n t i t y type. as p e r f e c t l y well-formed, while n o t t a k i n g account o f t h e usgge i n t h e % e n t i v e p a r t of t h e case frame f o r ' f i n d ' . g e t h e r , a s i n t h e following f o r \"I broke the g l a s s w i t h a rock\".",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "I have followed Fillmore's (1975j device h e r e of making 13easier t o read by p u t t i n g ~t i n SVO rather than t h e usual VSO Thus, f o r example, Schank's p r i m i t i v e a c t TRANS expresses t h e u n d e r l y i n g c o n t e n t of such a c t i o n s as \"buy\", \" s e l l \" and \"take\", and .ke would b e g i n t h e r e p r e s e n t a t i o n of 14The man took a book a s (Schank 1973, p. 196) -to. man show, 8s with \"break\" (10) e a r l i e r , t h a t the r e a l dependency of the two groups was on MOVE, However, t h e requirement 1s observed t h a t the case expressed by t h e p a r a p l a t e , where by template-skeleton I mean an e n t i t y l i k e a t e p p l a t e except t h a t , instead o f a formula a t each of i t s t h r e e n o d e s , i t has a function r a n g k g over formulas. But l e t u s turn t o :",
                "cite_spans": [
                    {
                        "start": 386,
                        "end": 407,
                        "text": "(Schank 1973, p. 196)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "(32) He b e a t t h e g i r l w i t h a w i t Yet, i f w e were analyzing the standard sentence \"He broke the window with a hammer\", w e would expect t o use a stack of paraplates under \"with\", containing a substack whose left-hand a c t i o n was CAUSE, and i n it f i n d t h e a p p r o p r i a t e p a r a p l a t e f o r tying together the two templates f o r t h a t sentence with an INST t i e . This l a t t e r information might seem t o overlap heavlly with t h a t contained i n the formula 10and t h e question a r i s e s whether i t need be s t a t e d twice, The r e s u l t of matching t h i s with templates, applying p a r a p l a t e s a s described above, and then performing case e x t r a c t i o n s can b e w r i t t e n i n summary form as fallows: Their r o l e , a s described in (Wilks 1973c (Wilks , 1975a ) i s to be data f o r f u r t h e r inferences using common-sense inference r u l e s , whose function i s not e s s e n t i a l l y connected with case and w i l l not be recapitulated here. I n the implementation described i n those references, the extractions were purely \"problem driven\", i n t h a t they were only done when some problem of reference resolution i n the t e x t demanded t h a t the representation be deepened, However, t h a t was a s t r a t e g i c consideration i n no way a consequence of the nature of extractions: they could equally well be data driven, and be executed a f t e r every matching of a sentence with templates,",
                "cite_spans": [
                    {
                        "start": 807,
                        "end": 819,
                        "text": "(Wilks 1973c",
                        "ref_id": null
                    },
                    {
                        "start": 820,
                        "end": 834,
                        "text": "(Wilks , 1975a",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "C C.Tohn",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "The difference between paraplates and extractions should be c l e a r from (38) because i t contains both an INST paraplate-imposed t i e between main clause and preposition phrase templates,and an INST extraction from the main clause template yielding an extracted template equivalent t o \"John uses a thing(=a shotgun)\", I n t h i s way, we a r e a b l e Lo r e t a i n a s p a r t of the o v e r a l l semantlc r e p r e s e n t a t i o n of text,and i n a message form appropriate f o r subsequent inference, both t h e s u r f a c e s t r u c t u r e (paraplate INST t i e )and t h e underlying s t r u c t u r e (INST e x t r a c t i o n inference) -As we saw i n the last s e c t i o n , Schank hblds t h a t these a r e one and the same thing.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "It i s t r u e , a s we s h a l l see, t h a t i n order t o e x t r a c t t h a t John uses a shotgun, the e x t r a c t i o n mechanism must consult t h e template t i e d by t h e )INST p a r a p l a t e but, as I wguedr i n detail, i n t h e l a s t s e c t i o n , t h i s In no way e s t a b l i s h e s t h a t the u r f a c semantics and the underlying sei~ a n t i c Let us see how the e x t r a c t i o n s i n (38) a r e a c t u a l l y obtai'zed.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "This w i l l r e q u l r e t h a t w e give more of the content of t h e f i r s t source template i n ( 3 8 ) , and i n p a r t i c u l a r the formula f o r \"fire+atl', A N ) , an e x t r a c t i q n from the \"break\" formula (10) : it has not been $oafinned by t h e Cext, and,",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 171,
                        "end": 176,
                        "text": "A N )",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "replaced by a f u l l formula f r o m a source templake, a s were the o b j e c~ and agent of ( 4 0 ) , L e t us t u r n f i n a l l y t o t h e sentence matched i n i t i a l l y as: It i s , of cours'e, e s s e n t i a l here t h a t t h e extractions a r e applied i n a fixed order, so t h a t the lNST e x t r a c t i o n can make u s e of the SUBJ e x t r a c t i o n and know t h a t \"hanrmer\" i s no,t avaixable t o b e the r e a l agent of an e x t r a c t i o n from (42). It i s n o t t h e case t h a t every l i n e of t h e specialist f o r case X, say, i s of t h e form \" i f t h e head of t h e source formula i s Y do 2, else ,,. \" This p o i n t w i l l be important m t h e next s e c t i o n when we consider Charniak' s arguments concerning c as?. T h i s way of l~k i n g a t things assumes t h a t John must i n e f f e c t , already be marked as an agent, hammer a s an I n s t q e n t o r Object and so on, Simmons' a n a l y s i s paradigms f o r t h e same example a r e , a s we saw a t t h e beginning, an a n a l y t i c v e r s i o n of t h z t r u l e , The d i f f e rence between those two approaches and t h e p r e s e n t one i s twofold, Charniak demands t h a t t h e \"semantics-based systems\" i n A1 aJso do s u r f a c e case a n a l y s i s , of t h e type found i n illm more's papers. The meaning of case CASE w o u l d then be \"the s e t of inferences m e t a n W k e &bout X, knowing only t h a t X i s i n case",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "They are i n t e n d e d t o express t h e",
                "sec_num": null
            },
            {
                "text": "In r e p l y , I t h i n k Charniak misjudges where the h e a r t of h i s argument l i e s by g i v i n g undue prominence t o r a t h e r b i z a r r e arguments l i k e ( i ) and ( i i ) , whereas t h e s e r i o u s ones a r e those l i k e (iv) and ( v i ) t h a t draw a t t e n t i o n , a s t h i s paper has t r i e d t o do from a q u i t e d i f f e r -I r ent p o i n t of view, t o t h e need t o r e l a t e t h e use of case i n semantics- (iv) and (vi) These arguments are p a r t l y t r u e but t h e i r points a r e aet by presenting a system, l i k e t h e preference semantlics one, t h a t ilses case both a s a semantic r e p r e s e n t a t i o n and a parsing mechanism (the paraplates) ,",
                "cite_spans": [
                    {
                        "start": 464,
                        "end": 468,
                        "text": "(vi)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Secondly",
                "sec_num": null
            },
            {
                "text": "(v > It i s t r u e t h a t no system, Fillmore's fncluded, has given (i) , and the answer i s simply t h a t t h e mechanism goes t o t h e f i r s t p o s i t i o u because t h a t i s where t h e agent has been put; i n j u s t t h e way t h a t d o l l a r s a r e w r i t t e n before the decimal p o i n t and cents afterwards y e t t h a t f a c t does not allow u s t o dispense with t h e notions of d o l l a r s and c e n t s i n f i n a n c i a l calculations. However, t h e important point here i s t h a t t h e \"seeker of t h e agent\" would seek i t generally: i t would not matter whaE t h e a c t i o n ( s t r i k i n g ) i n t h e extracted template was. ",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 70,
                        "end": 73,
                        "text": "(i)",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Secondly",
                "sec_num": null
            }
        ],
        "back_matter": [
            {
                "text": "'Ted #ire,\" and the PRCS Bulk Data c.ip.lbi1it.y for trnnsmitti~g funds bcthccn 4CII: sites. IVhile t h e project 11~1s rcccit~d l i t t l e public cxposure or m -t o u r l c e -]dent o u t s i d e of file trade press, i t s i n i t j ' l l phnscs have bccn undcnvay since July, and it is sdlcdulcd for comi~pl.c-t im (incl udiog tr~nsmm ssion of 1 ivc pciyaicnt s) by Dcce~nber , 1976 F{+rI++,>aZ add p r~o a t e reaclion. The l~rovisaon of I:FT zerviccs by the l'cJcral Rcsurvc has bccn t h e subject of COIILSICII~, from both tk Federal and private s c c t o r s . ELirlicr t h i s year, PI ~~~u t i~n t i r i g g c~l c r a l l y upon the provj si on of EFT $crviccs by t h c QcJccr:ll Rcscrve, John Pgcr ( .in thc actual provj si on of @k1g scrvic-cs . At:cu,t(li ngly, i l l l ' a~~l t .lily c~;~p c l i i ng j u s t i f i c a t i o n fir guvc~~luacnt e n t r y i n t o tlli's m I rkot, t h e p~ i v a t e ~~c c t o r should be l e f t t o provide clcotronic iullds tr.lnsfcr scrvices i n ,in c~i v i~* o~! a c n t that is free from govcmment ~p e r~i t ' i o n i l involvc.nc~it .\" 'Illis vlc:tpoint llas been rcjxfbrc[.rl by co~luncnts from potential p r i v a t~ vcndors of 1:F' I' scrviccs :.hich 113yc I,VCII r i l e d llri tll tllc hTd1. i 411aal Cb;~o-~i ss j on @n I:F' T (?ICIJFIF) ;, s11q11 g~.(>~n).; Il,~ve cnllcd for prllrlic hc irin& on illc cx~h:~ilgc' pmj ect pri or t o i t s l'ull i~ mi) tc~ncntd t i on.",
                "cite_spans": [
                    {
                        "start": 369,
                        "end": 385,
                        "text": "Dcce~nber , 1976",
                        "ref_id": null
                    },
                    {
                        "start": 708,
                        "end": 709,
                        "text": "(",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "annex",
                "sec_num": null
            },
            {
                "text": "cv .ippc.? 1 i og j n FiJk'S-L t17jcY r y T'~,~vr.l, 01' 1' A s~i 5 i . l l l t Ci11cra1 Co1111sel TI~VHMS ? f c~i i~~j l l t \\'lid i.lldt tlie. 1) 1-oj c c t lu:s l L , Lcd i Re \"2114 1 cpt I ti 011s t l c~ P I O~Y , I P I I~ of an on-1 i i~o L ,~~I J~I I j t i , \" J I I~ CJI Led' i,t \"illove v! i t lrout ~-(>g'~-rd t o -the u1 shcs of tho #CI:FT, Congress, and the priv'ate scclor is t o how tke 11lt i m~t c con~plcxion o f tlfe nat i on1 s CFT systctlls should c~l )~~~d r . ?bs:-~l',gf a~t < v { t y , h'hi lc the red's proposed p r o j e c t lists as its obj e c t i r e~ a campi l a t i o n of recomrncndations and possible pi-~blerns involved i n EFT systems, a planning document f o r the project * c l e a r l y comtcmplates an expanded Federal government role; in p a r t i d u l a r , the document s t e t c s that the p i l o t \"should be conducted in t h e most r e a l i s t i c cn~itonm@nt possible. . . .[~]onsidcration should be given t o t h e d c s i r a b i l i t y o f selecting p a r t i c i p a n t s t h a t have a broad base of interregional a c t i v i t y , whizh wow2d facilitate e$pmion of the progrm t o other r.c~gllijot ACH's in thrrlfut*rerlt (Enphnsis supplied).While the XCEFT has not taken a definitive stand on the project, i t has bccn a t o p i c of discussion a t t'accat ~acctings of the Cortunission, 1!1c f ' l~t i~t c s 5 w i I 1 1 kc In i n i t i a l step tow I L .~ . I I *~ ( 5s t o E K C C I I~ ive l3r:,f1cll inf o l j ,~.~t ig911 ~j w~, t~i :~s (Scc ;,'te:;h~~:yl.~~ Z ? c~u~t , //76)* t11is 1.1w11th ~I I C I I CIIG ( ; c r~c r~~l ccou!it ing ol'ri qc i r~v~r i t o r y of ovcr 1,000 \\ysttLtl1s i n 53 d j t'fcrcr~t I:xw.,utivc a3o11~ics is n:dc avai lz~blc an-line t o slcrsbcrs, of Conkrcss through t h c SCOIIPIO data l):~sc ~n a intni ncd by thc I,i:)rary of Congrcss .1-n addition, SCORPIO i s s c h c d~l~d t o be augmented with a new d a i l y index to the ~w p e s s i o~1 Record. The file, which will be updated d a i l y will be accessib3;e %y-170 s u b j e c t terms, and will contain complete data back to t h e beginning of the second session of the194th Congress (from 1/19/76). Available data will incbude summaries of f l o o r actions, recorded votes, and the location of Ftcmber's s totcments made during debate. I n c o n t r a s t t o Cgcr, however, IIouser has taken a neutral stand on t11.e l e g i sl a t i o n now before Congress which would reassert.ATGTfs monopoly s t a t u s i n comri\\unicntions (~us'rling ton Report, 7/76) l!ouser s a i d that USP would colltil~uc t o be irltercstcd i n d a t a cununui~ications, privacy and e l e c t r o n i c funds t r a n s f e r ; he f u r t h e r announced t h a t t h e Ofrice would develop by t h e end of t h e year a document which would t d k e a long range look a t t e l c c o~~n u n i c a t i o n s , and would c o n t a i n statements o f p r i n c i p l e s by 0'1'P and private industry, respectively. While Houser indicated t h a t OTP was not presently taking p o s i t Jascph J . Ryan has been appointed as special a s s i s t a n t t o group created by t h e OSTP l e g j s l a t i a n . The Cdmmittcc i s t o survey t h e ovcrall Federal science, engineering and tcqhnolozy effort, submitq an intcrim rcport with rccomncndations one year a f t c r its. inception and a f i n a l rcport with. 'recdmnlunclations within tyo years. While Dr, Ram0 hap not )'ct outlincd h i s program, he is known t o place n high p r i o r i t y on computer technology ~n d partjcularly. o i its appliortif~h t o iiaprovo productivity. I 7/76). The Hoyse Subcommittee on Commtlrticati~na had sought testlmony on three primary topics : (1) the advantages aiid disadvantages of Federal Comrnunicatians Commisgion (FCC] policy allowlng ~o~npetition in the teletommunications industry, (2) the idcntifiaation of unresolved issues relating to competition, 2nd (3) identification of further information necessary prior to a subcommittee decision AT&T Teseimony. ATGT hairm man John D. deButts testified that the proposed legislation would not affect the profitabiility of his company, since ATGT would in any case petition the FCC for rates which would allow sufficient earnings to attract capital. DeButts re~terated A T U s prlmary argument, however that such competition would likely shCf* cost burdens from business to resiaential users of the system. D-e.Bvtdt;s mainta,inecl under questioning that competition permitted by the FCC had required ATET to adapt its design standards and operating methods to competitive equ;ipment, had complicated network planning, and had divided markGts which copld more economically be handled (by virtue of economies of scale) by a single system.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "In d n i n t c~v j",
                "sec_num": null
            },
            {
                "text": "Representatives of nunerous companies competitive with ATGT in both voice and data coiiununications testified during the second day of hearings. Ph~1i.p Wittaker of Satellit0 Business Systcms (SBS) argued t h a t ATgT was blrt one of many companies ifl carnmand of the technology and that f u~t h e r national economic development required the resources and expertise oi, these conlpanies in a competitive environment, in order to meet the needs of s p e c i a l i z e d users. i-Ie maintained ;that SBS :gill make available services not now offered nr proposed by other carriers and which will be significantly advantagous to users. Telenet Vice President Philip Walker submitted that if the AT$T legislation were effective in 1974, Telenet's applications to provide packet switching services [which are not provide-d by AT4T) would never have been approved. Walker proposed that ATET be perfiltted to provide competitive services only through subsidiary companies Peter McCloskey of the Computer and Business Equipment Manufacturers Association (CBEMA] testified in favor of competition in the manufacture of terminal equipment, arguing that the market itself provides the mechanism necessary to allow users a choice of equipment at the lowest possible prices FCC C h a z . W i Z e y . Richard Wiley opened by describing exrsting exceptions to monopolistic netwo'rk control, including independent telephone compahics and iddependent ly produced equipment used by t el cyhone companies . Wi 1 cy also stated that the proposed legislation would jeopardize significant new developments such as data comnunications and EFT systems, and would allow concentrated control over this \"increasin~ly vital and sensitive information system.\" The FCC had in May of this year submitted written comments oppos ing the proposed legislation, on the grounds that it would inhibit the development a f consumer alternatfues, and allow higher casts; the FCC also maintained that ATBT would be allowed to collect and dispense \"substantial subsidies without effective Federal oversight .Speakrng for the Justice Department, Assistant Attorney General Baker prafqed FCC regulation for allowing competition where possible in the common carriar field Baker stated that competition brought substantial benefits, and ursed the Congress to pursue such colnpetitive solutions. Baker maintained that existing data to subs.tantiat6 AT(Tfs claims of economy of scale was in~onclusive; he indicated that local message switching may involve economies of sgalc, but that the manufhcture of terminal equipment probably did not.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Response from compet;t<ve vendors.",
                "sec_num": null
            },
            {
                "text": "The Scnate in recent bearlngs considered three alternative plans to restructure its complex syqtom of 174 subco~fttees which relate to science and technology. The Temporary Selecr committee to Study the senate Conmittee System isfocusing on this ated by considering primarily three alternative plans which, if adopted. wolrld bd subject to furyeher nodif'cation. The first plan would retain the existing committee organization, but would significantly reduce. the number of committee members and reorganize committee jurisdic~ions . The Senate is expected to deal with these proposals when the 95th Congress convenes next year. The General Accounting 0fEic.e (GAO') has released a report-on \"Ways to Improve &mug-nt of Federally kcnded ~o @ u t e r i z e d ~o d e Z s I~ (#-enc luse $1.00) . The r e p o r t concludes that appropriate standards from* t h e Depaqtment of Commerce, and guidance from t h e General Services Administration should be provided f~r improving management of such models. GAO developed and proposed a phased approach fo'r planning, managing and controlling the development o f computerized models which consists of f i v e phases: problem d e f i n i t i b n , preliminary design, d e t a i i e d design, e v a l u a t i~n , and maintenance. According t o GAO,, t h i s approach will reduce costs overruns and r e s u l t i n models b e t t e r s u i t e d t o the needs for, which' they are designed.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "SENATE SCIENCE REID TECHNqLbGY COE1MITl'EE AE~P@WIZATIQN",
                "sec_num": null
            },
            {
                "text": "The White Hpuse Office of Management and Bud e t (0f1l33 recently released a d r a f t Statrdard Occupationat CZassif<~atiofi f SO^ lkmual wTiich categorizes t h e o c c u p a t i o n a~ s t r u c t u r e of t h e U. S. work force. I t is a n t i c i p a t e d t h a t the Manual waald be used by t h e Department of Labor arrd other Federal agencies f o r categorizing manpower data collected o r required by t h e Federal government. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "OBM PROPOSED STANDWD COMPUTER SCrENCE C JASSIFICATIONS -",
                "sec_num": null
            },
            {
                "text": "C I V l I, SIJIIYlCE .:Ob!llU' L%R SCl IJN'I'I S' T. OC(:IJIJA'l'lONAI, S'l'lZNl)ARD He had earlier undertal sn s i g n i f i c a n t research f o r AFTPS i n connection with t h e AFIPS Stt;dy on Professionu~ism, and the AFIPS roundtabl on t'Professionali.m i n t h e Camputer F i e l d t r held i n 1970. M c C a~t e r ' s journdli&tic insight i n t t a Wh-shina ton and r e l a t e d information processing issues is expected t o bs a s t r a n g W e o f f e r P e n d~r .and Linda our b e s t wishes i n t h e i r work with AFIPS, and welcome them t o the s t a f f :t o use material 'in t h e A~IPS' ~ashCngton Report t h a t where an a r t i c l e t i t l e appears with an l1(:)\" clcaranco must f i r s t be obtained from the AFIPS Washington O f f i c e Docu a r e a v a i l a b l e on rcqucst t o t h e AFIPZ noted, make checks payable t o \"ARIPS.ll II",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Richardsbn.",
                "sec_num": null
            }
        ],
        "bib_entries": {
            "BIBREF1": {
                "ref_id": "b1",
                "title": "UCLA Department of ~i n g u i s t i c s )",
                "authors": [
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Celce-Murcia",
                        "suffix": ""
                    }
                ],
                "year": 1972,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Celce-Murcia, M, (1972) o n , (Los Angeles: UCLA Department of ~i n g u i s t i c s )",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "type_str": "figure",
                "text": "s paper h a s t h r e e purposes: f i r s t l y , t o d e s c r i b e how c a s e information IS d i s t r i b u t e d i n t h e p r e f e r e n c e semantics system o f language u n d e r s t a h d i n g , and t c show what p r a c t i c a l use i s made o f t h a t i n f o r m a t l o q . Second] y, t o argue t h a t t h a t way of doing t h l n g s h a s a d v ,~n t a g e s o v e r t a o a l t e r n a t i v e s : (a) p u t t i n g a l l c a s e i n f o r m a t i o n I n one p l a c e , and (b) n o t u s i n g any case i n f o r m a t i o n a t a l l , b u t o n l y t.he names of E n g l i s h p r e p o s i t i d n s . T h i r d l y , I wish t o -u s e t h e p o s i t i o n s e s t a b l i s h e d e a r l l e r c o u n t e r some r e c e n t arguments by and o t h e r s that t h e n o t i~n of c a s e i s n o t i n f a c t functioning i n any n a t u r a l language u n d e r s t a n d i n g systems t h a t f a l l w i t h i n what could be c a l l e d the A r t i f i c i a l Intelligence paradigm. A theme t h a t r e c u r s i n I 1 the p a p e r i s t h a t t e n d e n t l o u s d i s t l b 1 c t i o n s , such as s u r f a c e \" , \"deep\" and \"conceptual\" c a s e , must b e expounded i n p r o c e s s i n g terms i f t b e y a r e t o make sense. The p a p e r owes a g r e a t d e a l t o discussions w i t h Eugene Chanliak, Graham Ritcshie, M a r g~r e t Klng and F r e d e r i c k Parker-Rhodes. The mistakes, a s u s u a l , are a11 my own6",
                "num": null,
                "uris": null
            },
            "FIGREF1": {
                "type_str": "figure",
                "text": "e i n p a r a p l a t e s . . . . . . . . . . . . . . . . 2 . . . . . . . INTRODUCTION T h i s p a p e r i s Intended t o d e s c r i b e t h e case handling procedures of t h e p r e f e r e n c e semantics (Wilks 1 9 7 2 , 1975a)system o f n a t u r a l language understanding by g i v i n g a more complete account than In prevlous p a p e r s , and I n p a r t i c u l a r I s h a l l d i s t i n g u i s h t h e application o f case t o the p a r s i n g o f p r e p o s i t i o n s t r u c t u r e s I n E n g l~s h from subseqyent i n f e r e n c e s u s i n g c a s e , Case information i n t h l s system I S s t o r e d i n t w o d l f f e r c n t p l a c e s : i n what a r e c a l l e d formulas and p a r a p l a t e s r e s p e c t i v e l y . I s h a l l a r g u e a r e a s o n a b l e * p r o c e s s i n g account o f case r e q u i r e s t h i s . I s h a l l w n t r a s t t h i s p o s i t l o n b r i e w w i t h those of Schank and Riesbeck, who seem t o m e t o advocate a s i n g l e type of case information, and a no c a s e view r e s p e c t i v e l y . More i m p o r t a n t l y , I h s h a l l argue a g a i n s t a 9 r e c e n t p o s i t i o n of Charniak t h a t A r t i f i c i a l I n t e l l i g e n c e ( A I ) n a t u r a l language systems do n o t i n f a c t make any u s e d case, I s h a l l d l s c u s s h i s arguments and u r g e t h a t , although t h e r e a r e systems t o whlch h i s arguments do apply, t h e y do n o t apply t o t h e one described h e r e , a t l e a s t h o t i f c a s e i s t o have a n y t h~n g l i k e i t s normal meaning, The u s e of case i n f o r m a t~o n i n A 1 comes from t h e work of F~l l m o r e (1968), i n which t h e underlying s t r u c t u r , e of a s e n t e n c e i s displayed, i n essence, a s an a r r a y of argument values far a p r e d i c a t e , where t h e -redi,:ate i s t h e verb of tk sentenceo The corresponding values a r e the c a s e p a r t s of t h e s e n t e n c e , each of a d i f f e r e n t case type, and, f o r any given verb, t h e g e n e r a l p a t t e r n of c a s e s i t t a k e s is c a l l e d t h e c a s e frame of t h a t verb. Thus, i f t h e h o r i z o n t a l l i n e s denote t h e v e r b p r e d i c a t e , t h e c a s e frame f o r \"break\" could b e w r i t t e n as: ( -------OBJECT (AGENT) (INSTRLJIENT) ) which means t h a t t h i s verb must t a k e an o b j e c t , and cen b u t need n o t t a k e an agent and i n s t r u m e n t case. These l a t t e r two c a s e s a r e optiona l h e r e and t h a t o p t i o n a l l t y i s indicated by t h e p a r e n t h e s e s round t h e c a s e names. So, on t h i s view: (1) John b r o k e would be ill-formed b e c a u s e i t gives u s o n l y t h e ( o p t i o n a l ) agent of breaking, which i s John, b u t omits t h e o b l i g a t o r y o b j e c t t h a t i s broken.",
                "num": null,
                "uris": null
            },
            "FIGREF2": {
                "type_str": "figure",
                "text": "The window broke would b e p e r f e c t l y well-formed w i t h r e s p e c t t o t h e frame f o r 'break', T h i s s o r t of c a s e a n a l y s i s 1s normally z a l l e d deep c a s e * t o distinguish i t from t h e s u r f a c e c a s e m a n i f e s t a t i o n of word i n f l e c t i o n s i n Greek, Lati'n and German e t c . It should b e n o t e d , too, t h a t h e r e , a s througho u t t h e paper, no a t t e m p t i s made t o s t i c k t o t h e a c t u a l l i s t of c a s e names used by any a u t h o r , s i n c e these vary s d much from one t o the n e x t , and no p o i n t of principle hangs on arYy p a r t i c u l a r l i s t of cases, ~i l l r n o r e ' s l i n g u l s t i c theory i s , n a t u r a l l y enough, a g e n e r a t i v e one i n t h a t t h e procedures i t s u g g e s t s would be t h o s e f o r g e n e r a t i n g s e n t e n c e s from an u n d e r l y i n g s t r u c t u r e of a verb p l u s i t s c a s e argument values. What t h e s u r f a c e f o r m would be, glven any p a r t l c~l a r underl y i n g s t r u c t u r e , i s determined By what F i l l m o r e c a l l s t h e \"subject",
                "num": null,
                "uris": null
            },
            "FIGREF3": {
                "type_str": "figure",
                "text": "are p e r f e c t l y well-formed with r e s p e c t t o t h e f r a n c f o r \"break1' given e a r l i e r .Simmons p a r s e s s u c h sentences u s i n g an augmented transi t i o n network (Woodst1970) and a n o t i o n of case paradigm d u e t o Celce-",
                "num": null,
                "uris": null
            },
            "FIGREF4": {
                "type_str": "figure",
                "text": "t o nouns m%rked ANIMATE i s not n e c e s s a r i l y a r e s t r i c t i o n p e c u l i a r t o 'break', b u t r a t h e r t o the c l a s s of verbs f o r which (8) i s t h e ( a c t i v e ) paradigm.",
                "num": null,
                "uris": null
            },
            "FIGREF5": {
                "type_str": "figure",
                "text": "The case frame expressed by t h e paradigm, and the corresponding s e l e c t i o n r e s t r i c t i o n s are i n d i v i s i b l e . So, for example,",
                "num": null,
                "uris": null
            },
            "FIGREF6": {
                "type_str": "figure",
                "text": "For t h e s e l e c t i o n r e s t r i c t i o n a s s o c i a t e d w i t h a c a s e d e f i n e s what ~t i s t o match a c o r r e s p o n d i n g l i n e o f (8).",
                "num": null,
                "uris": null
            },
            "FIGREF7": {
                "type_str": "figure",
                "text": "number of arguments (say, f o u r o r z e r o ) n o t c o r r e s p o n d i n g t o any l i n e of (8), CASE I N PREFERENCE SEMANT?CS Case i n formulas T h i s system b u i l d s meaning s t r u c t u r e s and i n f e r e n c e r u l e s from eighty p r i m i t i v e semantic elements, These a r e of e i g h t types, one of which c o n s i s t s of t h e case elements a s follows; *DIRE t h e general DIRECTION ca'se element. Like a l l the p r i m i t i v e elements whose names a r e preceded by an a s t e r i s k , i t i s equivalent t o a c l a s s of o t h e r p r i m i t i v e s , i n t h i s case t h e following four: TO d i r e c t i o n towards FROM d i r e c t i o n away from something UP i n an upwards d i r e c t i o n THRU d i r e c t i o n through some o t h e r thing. INST the INSTRUMENT case, i n d i c a t i n g t h e instrument used i n some action FOR t h e RECIPIENT case, i n d i c a t i n g t h e normal r e c i p i e n t of an a c t ion I N t h e CONTAINMENT case, i n d i c a t i n g what contains some o t h e r t h i n g LOCA t h e SPATIAL LOCATION case, indicating t h e p l a c e of an a c t i v i t y o r t h i n g TLOCA 'the TIME LOCATION else, i n d i c a t i n g the time l o c a t i o n of an a c t i v i t y WAL t h e -, PURPOSE case, i n d i c a t i n g t h e purpose of an a c t i v l t y SOUR the SOURCE case, i n d i c a t i n g the substance from which some o b j e c t came WAY t h e MANNER case, ~n d i c a t l n g the manner o r method by which an a c t i v i t y was performed OBJE t h e OBJECTIVE case, indicatang t h e o b j e c t of an a c t i o n SUBJ t h e AGENT case,. i n d i c a t i n g t h e i n s t i g a t o r of an a c t i o n , ' s u b j e c t ' here b e i n g taken t o r e f e r t o a semantic, r a t h e r than a s u r f a c e , s u b j e c t WITH t h e ACCOMPANIMENT case, i n d i c a t i n g t h e acco~npanier of an e n t i t y POSS t h e POSSESSIVE c a s e , i n d i c a t i n g who owns some t h l n g These case p r i m i t i v e s a r e u l t i m a t e l y t h e names of r e l a t i o n s i n t h e system of semantic r e p r e s e n t s t i o n , j u s t a s i n the f a m i l i a r semantlc n e t r e p r e s e n t a t i o n s (Simmons 1973) t h a t i n d i c a t e t h e i n s t r u m e n t a l i t y of say, t h e a c t i o n of s t r i k i n g by a l a b e l l e d a r c such as: s t r i k e 'hammer The representations described h e r e a r e n o t of t h i s s u p e r f i c i a l form f o r t h r e e reasons: a)",
                "num": null,
                "uris": null
            },
            "FIGREF8": {
                "type_str": "figure",
                "text": "s t i n c t i o n 1 s not always easy t o work i n t o a semantic n e t s t r u c t u r e .",
                "num": null,
                "uris": null
            },
            "FIGREF9": {
                "type_str": "figure",
                "text": "hdve constructed n e t s containing ordered a s s e r t i o n s , though t h e s e a r e no longer semantic n e t s i n t h e c l a s s i c sense).",
                "num": null,
                "uris": null
            },
            "FIGREF10": {
                "type_str": "figure",
                "text": "E'ormulas are t r e e s of l e f t -r i g n t dependencies b u t t h e dependence i s i n t e r p r e t e d d i f f e r e n t l y acccrding t o t h e type of t h e subformula. W e have a l r e a d y mentioned t h e dependence of an e n t i t y ( o r a s s e r t i o n ) on a case p r i m r t i v e t o form a c a s e group, such a s (*'tIb% SUBJ), which means t h a t an agent i s ( p r e f e r a b l y ) human. These c a s e groups (except WITH and POSS) a11 depend r i g h t w a r d s on some a c t i o n , so t h a t (*HUM S U N ) i n (10) depends on t h e main p r i m i t i v e a c t i o h of t h e whole formula STRTK, The whole formula i s t o b e i n t e r p r e t e d a s \"breaking\" b e i n g a ~' k~~~i n g , done p r e f e r a b l y t o a *PHYSOSJect, and by a *HUMan SUBJect, u s i n g an INSTrument t h a t i s a THING and w i t h t h e GOAL o f CAUSing t h e *PINSOBJect t o BE NOTWHOLE. T h i s i n t e r p r e~a t i o n can b e c o n s t r u c t e d from t h e f o l l o w i n g _generaL r u l e s f o r t h e b u i l d i n g and i n t e r p r e t a t i o n of formulas: i) Each subgroup i n t h e f o r m u l a c o n s~s t s of a left mernber depending on a r i g h t member, and l e f t o r r i g h t may b e e i t h e r a s i n g l e p r i m~t i v e element o r a n o t h e r group, Thus, i n (*EIUM SUBJ) w e have a c a s e group, known t o b e such because t h e r i g h t m o s t m e m b e r of i t s p a i r i s t h e govern o r and SUBJ i s t h e p r i m i t i v e element naming t h e Agent caqe. One level h i g h e r (*HUM SUBJ) . depends on STRIK, t h e main p r i m i t i v e o f t h e whole formula, to form an a s s e r t i o n group. S i m i l a r l y , each of t h e other maln s u b p a r t s of t h e formula (whose heads a r e r e s p e c t i v e l y OBJE, INST an6 GOAL) depends on STRIK t o form i n each c a s e an a c t i o n group which always c o n s i s t s of an a c t i o n and any c a s e group that i s n o t an Agent group (nor a POSS o r WITH group),",
                "num": null,
                "uris": null
            },
            "FIGREF11": {
                "type_str": "figure",
                "text": "i n g t o t h e t y p e of t h e group.Within an a c t i~n graup ((*PKYSOB OBJE) STRIK) t h e dependepce i s t h a t of an a c t i o n ' s o b j e c t on t h e a c t i o n and OWE does no more than name t h a t r e l a t i o n .I n t h e cases group -(*PHYGOB OBE) t h e r e l a t i o n i n t e r n a l l y i s no s o r e than t h e p r e f e r r e d t y p e of c a s e f L l l e r Cphysical o b j e c t ) on t h e name of t h e case. With a s u b s t a n t i v e group l i k e (LINE THIYG) t h e dependence i s i n t e rp r e t e d a s s p e c i f i c a t i o n , i.e. l i n e a r o b j e c t . *PHYSOB i s a name ~f a c l a s s of primitive elements which i n c l u d e s THING, but a l s o o t h e r p r i m i t i v e s l i k e MAN. In c a s e subfarniulas, e&c@pt afid WAY, t h e l e f t m o s t item i s always t h e p r e f e r r e d e n t l t y type, t o f u y c t l o n i n t h e corresponding rightwards-nwed case. T h i s l e f trnost item 1 9 , i f you w i l l , the ' s e l e c t i o n restriction' f o r t h a t c a s e r o l e f o r whatever a c t i o n i s b e i n g coded: i. e. i n t h e formula above, f o r 'break'. The r e a d e r should n o t confuse t h i s with being a r e s t r i c t i o n f o r t h e a s s o c i a t e d p r i m i t i v e STRIK, T h i s p o i n t w i l l b e discussed l a t e r , b u t",
                "num": null,
                "uris": null
            },
            "FIGREF12": {
                "type_str": "figure",
                "text": "e r e n c e and p a r s i n g routines, However, s e l e c t i o n r e s t r i c t i o n \" h e r e i s t o b e read as I p r e f e r r i n g t h e agent of \"break\" t o be hurnaa1,say,I have d e s c r i b e d elsewhere (Wilks 197 c) how when t e x t r e p r e s e n t a t i o n s are assembled t h a t w i l l b e o n l y a p r e f e r e n c e on t h e agent of \"break\", and t h e system w i l l n o t baulk a t assembling a r e p r e s e n t a t i o n f o r \"The dog broke h i s bowl\" where t h e agent i s n o t human, though t h e system would p r e f e r a hwnan a g e n t i f i t could f i n d one. I write of \"assembling r e p r e s e n t a t i o n s u because t h e elements l i k e *Hull i n t h e formula above a r e n o t s l o t s t o be f i l l e d by, i n t h i s case, t h e agent of some breaking, The formulas a r e ' b l u e p r i n t s ' f o r how r e p r e s e n ta t i o n s a r e t o b. assembled elsewhere from whole formulas. When a r e p r e s e n t s t i o n f o r \"The man broke .the window\" i s assembled t h e whole formula above ( t o g e t h e r w i t h a PAST element) w i l l s t a n d a t some appt o p r i a t e node of a h i g h e r -l e v e l r e p r e s e n t a t i o n .",
                "num": null,
                "uris": null
            },
            "FIGREF13": {
                "type_str": "figure",
                "text": "s t f o r a l l p a r t s of speech s o , f o r ex@mple, a formula f o r an e n t i t y w i l l have n o t a p r i m i t i v e a c t i o n head l i k e STRIK, b u t an e n t i t y head l i k e T H I N G o r MAN o r STUFF ( f o r substance). Note t o o , t h a t t h e p r e f e r e n c e r e s t r i c t i o n s for case need n o t be simple a s above b u t can be a s complex a s r e q u i r e d , i n c l u d i n g f u r t h e r c a s e r e s t r i c t i o n s r e c u r s i v e l y . So, f o r example, if we had a formula i3r 'sew' i t might w e l l have a c a s e sub-fomrrula t o be i n t e r p r e t e d as: done with an INSTrument t h a t i s p r e f e r a b l y a LINEar THING WITH (accompaniment case) an a p e r t u r e (THRU PART). i i i ) I n o r d e r t o makc t h e formula \"habitablet' t h e agents and o b j e c t s a r e compressed, i n t h a t they can be agknts and o b j e c t s f o r more than one p r i m i t i v e a c t i o n , Agents and o b j e c t s of a c t i o n s i n a formula a r e normally sought t o t h e l e f t of t h e p r i m i t i v e a c t i o n element. I f t h e whole formula i s f o r an a c t i o n (as above f o r 'break') t h e two l e \u00a7 tmost subparts of t h e formula w i l l always be t h e p r e f e r r e d agent and o b j e c t of t h e head p r i m i t i v e , i n t h a t order. For any a c t i o n s w i t h i n t h e formula (such as CAUSE i n t h e formula f o r 'break') i t s preferred agent and o b j e c t are normally t h e next agent and o b j e c t t o",
                "num": null,
                "uris": null
            },
            "FIGREF14": {
                "type_str": "figure",
                "text": "Thisproviso does n o t apply I n t h e formula above s i n c e t h e agent i s t h e same f o r CAUSE and STRIK, and CAUSE takes an a s s e r t i o n as o b j e c t , but within a formula a group (MAN STRIK) would always be i n t e r p r e t e d a s an group, MAN being an unmarked agent of STRIK, and not a s a man being s t r u c k which would r e q u i r e a marked o b j e c t i n t h e a c t i o n group i , e , ( (MAN OBJE) STRIK) .A l l t h i s implies t h a t some of t h e subgroups i n t h e formula f o r break' a r e not t h e apparent ones i.e, t h e dependent of GOAL, a s mentioned e a r l i e r , must be an a s s e r t i o n , whereas i t i s brncketted t o only (((NOTWHOLE K1ND)BE)CAUSE) which can only (during i n f e r e n c e procedures,called ' e x t r a c t i o n ' t o be described l a t e r ) become ana s s e r t i o n group by t h e a d d i t i o n of an agent found t o t h e l e f t namely (*HUM SUBJ), CAUSE also r e q u i r e s a dependent o b j e c t t h a t i s an a s s e r t i a n (hence (*PH'ISOB OBJE) w i l l n o t do a s i t s a b j e c t 'taken alone) and can t a k e , a,s dependent of t h a t group, an e n t i t y t o i t s l e f t marked e i t h e r OBJE o r SUBJ kfnichever i s c l o s e s t , Hence t h e dependent of ((NOTWHOLE KIND)BE) i s *PHYSOB and t h e \"'real\" dependent of CAUSE (found by inference) i s (*PHYSOB ( (NOTWHOLE KIND) BE) ) and t h e r e a l dependent of GOAL i s ( (*HUM SUBJ) (*PHYSOB ( (NOTWHOLE KIND) BE) ) )CAUSE) . T h i s compresai& of expression can be argued t o be \"habitable\" f o r a formula maker, It a l s o avoids t o a l a r g e e x t e n t t h e d e f e c t of some f u l l e r conceptual r e p r e s e n t a t i o n s of t h i s general type, pointed out by Sandewall (1972), t h a t if t h e e n t i t i e s l i k e (*HbM SUBJ) a r e put i n t o t h e r e p r e s e n t a t i o n many times but a r e inteqded t o r e f e r t o THE SAME IlUMAN, then t h i s must be i n d i c a t e d a s i t f r e q u e n t l y was n o t , Where such i d e n t i t y must be s p e c i f i c i n formulas,but cannot be achieved by t h e above compressed expressions, i t i s obtained by means of t h e p r i m i t i v e s SAME and NOTSAME: t h e same (or n o t ) , t h a t i s t o say, a s t h e f i r s t encountered token of t h e a s s o c i a t e d p r i m i t i v e when working i n from t h e top l e v e l of t h e formula, One point t h a t should emerse from t h i s i s t h a t those.who want t o u s e c a s e names a s t h e names of r e l a t i o n s , as i n semantic n e t s , and a l s o d e a l with s u r f a c e language, must be prepared t d e x t r a c t a number of such r e l a t i o n s from a s i n g l e occurrence of c e r t a i n formula s u b p a r t s , Thus, t h e formula f o r 'break' a b~v e would c o n t a i n n o t only t h e n e t l i n k s : but a l s o t h e q u i t e o t h e r type of l i n k s e r t s t h a t t h e p r e f e r r e d agent of breaking w i l l be human. However t h e top l i n k must n o t be i n t e r p r e t e d a s saying t h z t t h e p r e f e r r e d agent of t h e p r i m i t i v e STRIK i s human, because t h a t i s not an a s s e r t i o n i n t h e system a t a l l . A l l t h e top l i n k can say i s t h a t the primitive a c t i o n STRIK sometines t a k e s human agents. In soae o t h e r formula, f o r another s u r f a c e a c t i o n whose underlying p r i m i t i v e was a l s o STRIK, t h e p r e f e r r e d agent might be \"ANI, a wider c l a s s . Hence, i n t h i s system t h e r e a r e n o t s p e c i f i c semantic r e s t r i c t i o n s on.",
                "num": null,
                "uris": null
            },
            "FIGREF15": {
                "type_str": "figure",
                "text": "t h e dependents of t h e p r i m i t i v e s , as i n , say Schank (1973), I n t h p p r e s e n t system, such a r e s t r i c t i o n could emerge only i n d u c t i v e l y from a survey of a c o n s i d e r a b l e body of formulas, It i s worth c l a r i f y i n g t h i s i s s u e here: what do underlying r e p r e s e n t a t i o n s c o n s t i t u t e case frames f o r ? The i s s u e i s r e l a t e d t o two o t h e r s :",
                "num": null,
                "uris": null
            },
            "FIGREF16": {
                "type_str": "figure",
                "text": "f i r s t , t h e d i f f e r e n t r o l e s of c a s e frames i n a n a l y s i s and gen*ratian, and s e c o t~d f y , t h e procedural o p p o s i t i o n between c a s e frame b l u e p r i n t s , l i k e formulas3 and the ' f u l l e r * ' r e p r e s e n t a t i o n s o f g e n e r a t i v e sunantic-> i s t s t r e e s and Schankian ~c o n c e p t u a~i z a t i o n s ' . First , l e t u s n o t e t h a t i t has neyer been as c l e a r as mrght b e wished what c a s e frames awe f o r i n F i l l m o r e ' s work, The normal i n t r oductory account given e a r l i e r s t a t e s t h a t they a r e f o r s u r f a c e v e r b s l i k e ' b r e a k ' , but i n F i l l m o r e (1975) he argues t h a t i t i s not so simple because h e would want t o admit s e n t e n c e s like: (12) Noon found Harry sleeping",
                "num": null,
                "uris": null
            },
            "FIGREF17": {
                "type_str": "figure",
                "text": "On t h e o t h e r hand, he wishes t o avoid t h e p o s i t i o n of s a y i n g t h a t case frames a r e f o r underl y i n g v e r b s l i k e semantic primitives i n formulas, o r t h e underlying v e r b s of Generative Semanti~srepresentations which avoid c a s e s a l t o -",
                "num": null,
                "uris": null
            },
            "FIGREF18": {
                "type_str": "figure",
                "text": "p r e d i c a t e f i r s t ) form, It w i l l be seen t h a t i t i s p r e t t y s i m i l a r to t h e above f n m u l a for \"break1' except t h a t , i n o r d e r to avoid case n o t a t i o n , they have had t o r e s o r t t o such p h i l o s o p h i c a l l y suspect devices as s e p a r a t i n g t h e a c t of using from t h e b a s i c ' a c t ' i n s i d e the tree, even though there was r e a l l y o n l y one a c t i o n i n t h e whole business An extreme v e r s i o n of t h e view t h a t case frames belong o n l y t o the underlying s t r u c t u r e i s Schank's (1973) view t h a t case frames awe f o r underlying p r i m i t i v e a c t s and t h a t all cases that a p r i m i t i v e a c t takes, i t t a k e s o b l i g a t o r i l y .",
                "num": null,
                "uris": null
            },
            "FIGREF19": {
                "type_str": "figure",
                "text": "e arrows l a b e l l e d R and 0 i n d i c a t e R e c i p i e n t and O b j e c t i v e . . -c a s e r e s p e c t i v e l y , and t h e -Agentive c a s e i s i n f a c t i n d i c a t e d by t h e I1 double arrow l i n k i n g man\", t h e a g e n t , t o t h e a c t TRANS. The d e t a i l s h e r e need n o t concern u s , t h e p o i n t b e i n g t h a t Schank i s s e t t i n g up case frames, n o t f o r surface verbs of E n g l i s h , l i k e F i l l m o r e (1968), b u t f o r t h e s e p r i m i t i v e a c t s , of which h e has about twelve. From the p o i n t of view on c a s e expressed i n t h e system d e s c r i b e d here, b o t h t h e s e s t r o n g p o s i t i o n s have drawbacks, indeed t h e y have complementary ones, In . t h e f i r s t p l a c e , illm more's (1968) system, w i t h t h e a i d of which he wants t o c o n t r a s t v e r b s by means of t h e i r frames, o n l y becomes s i g n i f i c a n t i f I n t e r p r e t e d w i t h t h e a i d of some n o n -s u r f a c e ' r e p r e s e n t a t i o n of a c t i o n s . So, f o r example, Charniak h a s pointed o u t ( p e r s o n a l communication) t h a t t h e e a r l i e r c o n t r a s t of t h eframes f o r \" h i t \" and \"break\" i s s i g n i f i c a n t o n l y i f t h e r e i s Pome common, u n d e r l y i n g , a c t i o n t h a t t h e two v e r b s s h a r e , and which can b e thought of as being s u b s t i t u t e d f o r t h e h o r i z o n t a l l i n e i n t h e frames.Eor, i f t h a t i s no& s o , theq t h e c o n t r a s t of t h e frames f o r \"kT11\" and murder\" i s of no more i n k e r e s t than t h e c o n t r a s t between \" k i l l \" andI I remember\", o r any o t h e r random verb, I n o t h e r words, i t i s only because t h e two v t r b s plready' have something i n common, over and above t h e i r case frames, t h a t t h e comparison has p o i n t . Thus, t h e c o n t r a s t of t h e c a s e frames of only s u r f a c e verbs i s , if unsupplemented, uns a t i s f a c t o r y , Conversely, t h e r e may be c e r t a i n problems i n h e r e n t I n Schank's attempt t o both (a) r e l a t e s u r f a c e verbs t o underlying p r i m i t i v e a c t i o n s , and then d i s c u s s o n l y t h e l a t t e r , and (b) a t t h e same time make a l l p a r t i c i p a n t s i n t h e c a s e frames f o r p r i m i t i v e s o b l i g a t o r y . 11 So, f o r example, Fillmore would express t h e case frame f o r see\" a s (OBJECT DATIVE) and f o r \"learn\" (OBJECT AGENT). While Schank (1973 pp. 220-1) expresses both v e r b s by an underlying p r i m i t i v e PITRANS t o g e t h e r with a c a s e frame, f o r t h e p r i m i t i v e , containing a t l e a s t A, 0 and R (Fillmore would c a l l R by D), The i n d i v i d u a l l e t t e r s f o r cases assigned by d i f f e r e n t a u t h o r s need n o t d e t a i n us, nor need t h e i r c o n t r a s t i n g i n t e r p r e t a t i o n of t h e c a s e names, f o r t h e p r e s e n t p o i n t i s the p e r \u00a3 e c t l y general one t h a t , whether o r n o t Fillmore i s r i g h t with t h i s p a r t i c u l a r v e r b p a i r , i t i s highly l i k e l y t h a t t h e r e are p a i r s L_. of s u r f a c e v e r b s l i k e t h i s one whose s u r f a c e c a s e frames a r e d i f f e r e n t and whose Schankian p r i m i t i v e a c t i s t h e same, It follows-from (b) above t h a t , f o r Schank, t h e i r \"deep\" c a s e frame must t h e r e f o r e be t h e same too. Since, f o r him, every s u r f a c e v e t b has a main a c t expressing i t , t h e r e i s c l e a r l y going t o be a problem with t h i s consequence of (b) u n l e s s he i s prepared t o say t h a t t h e r e i s no necessary r e l a t i o n a t a l l between a v e r b ' s c a s e frame and t h e case frame of i t s corresponding p r i m i t i v e a c t . It may well be p o s s i b l e t o defend such a p o s i t i o n w i t h i n h i s theory, but he w i l l s t i l l be l e f t w i t h t h e d i f f i c u l t y that: v e r b s with q u i t e d i f f e r e n t semantic behaviour ( f o r F i l l m o r i a n c a s e d i f f e r e n c e s a t e not super\u00a3 i c i 8 1 ) have i d e n t i c a l behaviour i n h i s system, There i s bound t o be a l a c k of d i s c r i m i n a t i o n consequent upon Schank's all-cases-are-obligatory view u n l e s s some c a r e f u l avoiding a c t i o n i s taken, t h a t h e h a s not y e t , t o my knowledge embarked upon, However, Schank would probably n o t wish t o take advantage of t h i s l a s t p o s s i b i l i t y because he does n o t himself hold t h e view t h a t 11 conceptual case i s e n t i r e l y independent of s u r f a c e s t r u c t u r e conside r a t i o n s \" , as was wrongly a t t r i b u t e d t o him by Bruce (ibid.p,338), and f o r t h e simple reason t h a t he i n t e n d s t h a t case s t r u c t u r e i n conceptualizations s h a l l r e s o l v e t h e c a s e ambiguities p r e s e n t i n English p r e p o s i t i o n constructions. Schank (1973) makes t h i s q u i t e c l e a r , and I s h a l l r e t u r n tc~ ilt when d i s c u s s i n g p r e p o s i t i o n constructi o n s i n t h e n e x t section. Thus, s i n c e Schankian c a s e frames f o r p r i m i t i v e s a r en o t independent o f a l l s u r f a c e s t r u c t u r e c o n s i d e r a t i o n s , he does have t h e problem above presented by the c o n t r a s t of \"see\" and 11 learn\", The burden of the l a s t arguments have been to show t h a t t h r e e d i f f e r e n t p o s i t i o a s on the question of \"what a r e case framesf o r ? \" a r e u n s a t i s f a c t o r y : ( i ) t h a t they a r e simply f o r surface verbs : Fillmore's 1968 p o s i t i o n , argued against along the l i n e s sketched above a s e a r l y a s Schank (1969), ( ii) t h a t underlying s t r u c t u r e s containing primitive a c t i o n s do n o t r e q u i r e c a s e frames a t a l l : the Generative Semaqtics p o s i t i o n , f o r example, P o s t a l (1971), ( i i i ) t h a t underlying s t r u c t u r e s containing primitive actions a r e case frames f o r those primitives and, moteover, a r e a l l obligatory case frames: Schank's position j u s t discussed. Let m e now r e s t a t e the p o s i t i o n of t h e present system, as i t concerns formulas, Formulas a r e meaning s t r u c t u r e s f o r surface word senses, Formulas f o r s u r f a c e verbs can be i n t e r p r e t e d as case frame$ f o r t h e verbs, i n t h a t they contain case subparts a t t h e top l e v e l ( i , e o depending d i r e c t l y on the head a c t i o n primitive) t h d t the formula maker has considered necessary t o express as p a r t o f the meaning of the verb. Thus, the formula f o r \"break\" given e a r l i e r (10) contains case subparts a t the top l e v e l (i.eo depending d i r e c t l y on the head primitive STRIK) INSTrument, GOAL, OBJect and Agent(=underlylng SUBJect), The formula maker i s concerned only with specifying, as b e s t as he can, t h e semantic preferences of t h e p a r t i c u l a r surface word i n question, If it were a verb he would semantically s p e c i f y t h e p r e f e r r e d agent, say, of t h a t a c t i o n , w i t b n o i m p l i c i t r e f e r e n c e to agenthood a s such. The same goes f o r t h e o t h e r cases: h e would attempt to put those cases i n t o a f o p u l a t h a t he t h~u g h t necessary t o specify t h e meaning of t h e a c t i o n , Consider (16) He l i v e s i n Lugano I and (17) H e drank some Barbera i n Lugano, It i s l o g i c a l l y t r u e t h a t one must, i f one d r i n k s , do i t somewhere, b u t no one wohld hold t h a t t h e n o t i o n of l o c a t i o n yas involved i n e x p l a i n i n g t h e meaning of drinking, However, one might well hpld t h a t i t was impossible t o e x p l a i n t h e n o t i o n of l i v i n g , i n t h e sense of i n h a b i t i n g , without making c l e a r t h a t i t was done i n some l o c a t i o n . 11 Thus o n l y i n t h e formula f o r l i v e \" would we expect a l o c a t i o n c a s e subpart (o.,..LOCA), This d e s c r i p t i o n of t h e i n s e r t i o n of case s p e c i f i c a t i o n i n t o t h e formula f o r some given s u r f a c e verb does n o t correspond i n any c l e a r way t o F i l l m o r e ' s o b l i g a t o r y o r o p t i o n a l d i s t i n c t i o n , though i t seems c l e a r t h a t any formula should contain a t l e a s t -F i l l m o r e ' s o b l i g a t o r y cases f o r t h a t s u r f a c e verb and, as w e saw , (10) above, f o r \"break\", does t h i s , s i n c e t h e qnly o b l i g a t o r y case f o r \"break\" i s o b j e c t , The main reason f o r t h i s d i f f e r e n c e remains t h e e s s e n t i a l l y g e n e r a t i v e q u a l i t y of a Fillmorean c a s e frame, Much s o p h i s t i c a t e d l i n g u i s t i c argument i n t h e p a s t decade has gone i n t o aptempting t o prove t h a t t h e term \"gen'erate\" is n e u t r a l Getween \"analysetl and 11 produce\". One of t h e b e s t e f f o r t s i s t o b e found i n~y o n s ' (1968, p.155), But, although i t i s easy t~ s e e t h e productive r o l e o f , say, illm more's s u b j e c t s e l e c t i o n r u l e (SSR), i t i s very hard t o s e e what a n a l y t i c s i g n r f i c a n c e it could have; t h e s u r f a c e s u b j e c t i s , a f t e r a l l , u s u a l l y revealed by simple methods n o t r e q u i r i n g t h e n o t i o n of case. The SSR t e l l s one how to c h o~s e t h e s u b j e c t 8ivent h e case s t r u c t u r e and i n t h a t s e n s e i s i n h e r e n t l y generative i n i t s non-neutral sense J meaning 'productive'. The system presented h e r e however i s i n h e r e n t l y a n a l y t i c ( i t s productive a s p e c t s have been described i n ( H e~s k o v i t s (1973)), and t h i s accounts f o r a g r e a t d e a l of t h e d i f f e r e n c e of approach t o t h e n o t i p n of meaning s t r u c t u r e 5 Those with p r a c t i c a l acquaintance w i t h sentence a n a l y s i s and production w i l l need l i~t l e persuasion t h a t t h e two processes a r e n o t i n any sense simple i n v e r s e s of each o t h e r (nor does Fillmore himself b e l i e v e they a r e , s e e illmore 1972 p, 23). Charniak has argued (1975) t h a t t h e method af formula coding r e s u l t s i n top l e v e l case subformulas t h a t cannot appear i n t h e s u r f a c e form of t h e verb, and he c i t e s ( i b i d , p,16) t h e formula . f o r \"drink\" : where t h e containment group (SELF I N ) i n d i c a t e s t h a t t h e l i q u i d o b j e c t (FLOW STUFF) i s moved i n t o the animate agent (SELF), and t h e d i r e c t l o n group ((MAN(Tmu FART))TO) i n d i c a t e s t h a t t h i s movement i s i n the d i r e c t i o n of a human aperture, These two groups appear a t the t o p l e v e l of the formula, and thus dependent on t h e head primitive a c t i o n C~~~ However, l a t e r i n f e r e n t i a l procedures of e x t r a c t i o n (see below) would",
                "num": null,
                "uris": null
            },
            "FIGREF20": {
                "type_str": "figure",
                "text": "obligatory cases (only Agent presumably) f o r \" d r~n k \" appear a t t h e t o p l e v e l , and t h e o t h e r case groupscontainment and d i r e c t l o nmost c e r t a i n l y could appear a t t h e s u r f a c e with 'drink' as in:",
                "num": null,
                "uris": null
            },
            "FIGREF21": {
                "type_str": "figure",
                "text": "John drank t h e beer up t h r w g h h i s nose w i t h a straw and i n t o h i s b r a i n , where both cases appear a t the surface. Charniak i s confusing what can appear a t t h e surface with what i t would n o r m a l l x b e redundant t o say. The diifelrence i s made c l e a r i n (19) i f we choose t o say somet h i n g remarkable, and f a l s e . I n t h i s s e c t i o n on formulas I have t r i e d t o j u s t i f y t h e i n t e rr o l e of case description formulas : by claiming BOTH t h a t they give a meaning s t r u c t u r e t h a t , under inferences processes, can provide a r e a l underlying s t r u c t u r e f o r t e x t AND t h a t they must include enough of t h e s u r f a c e case behaviour (of the verb meaning being expressed) t o give procedural power i n analyzing the input surface sentences. It seems to me e s s e n t i a l t o p r e s e r v e h o t h these r o l e s , and t o avoid the path taken, i n t h e i r d i f f e r e n t manners, by Schank, Fillmore and the Generative Semanticists, qf displa$ing a f u l l underlying s t r u c t u r e d i r e c t l y without t h e processes t h a t reach i t . I argued e a r l i e r t h a t each of those t h r e e gave only a f i l l e d -i n or-f i n a 1 , s t r u c t u r g which i n i t s e l f gives no h i n t s as t o how you get there. The generative l i n g u i s t argues, of course, t h a t he i s n o t t r y i n g t o reach such a s t f u c t u r e a t a l l , b u t to generate surface s t r u c t b r e s C_ from i t , and the \" a r t i f i c i a l i n t e l l i g e n c e c r i t i q u e \" of t h i s aspect of generative l i n g u i s t i c s i s f a m i l i a r by now (my own version i s i n ilks \" 1975~). The wqy in which Schank emphasises a f i l l e d -i n s t r u c t u r e i s a q u i t e d i f f e r e n t and more i n t e r e s t i n g matter, His conceptualizations are f i l l e d -i n s t r u c t u r e s , with no procedural capacity. However, at t h e back of h i s e a r l y papers (i.e. the Appendix t o Bchank e t a: . 1970) a r e d i c t i o n a r y e n t r i e s o r s k e l e t o n s f o r c o n c e p t u a l i z a t i o n s , which c o n t a i n s e l e c t i o n r e s t r i c t i o n s on t h e s l o t s i n t h e conccp t u a l i z a t i o n s (and i n (Schank 1973, p. 229)). These o b j e c t s a r e much more fundamental t o h i s approach than i s g e n e r a l l y r e a l i z e d , f o r they do give i t more procedural power t h a n one could envisage from looking only a t t h e f i n a l c o n c e p t u a l i z a t i o n s , T h e i r r o l e i s somewhat l i k e t h a t of t h e formulas f o r a c t i o n s : they a r e b l u e -p r i n t s showing what t h e system would l i k e t o be t h e usage and c o n t e x t of a given a c t i o n . But t h e d i f f e r e n c e from formulas i s this: the Schank d i c t i o n a r y e n t r y h a s s l o t s , marked, say, HUMAN which, when they a r e a l l f i l l e d by s u r f a c e vords y i e l d the f i l l e d -i n conceptu a l i z a t i o n , The formula too has c a s e s p e c i f i c a t i o n s , l i k e (*AN1 S U B J ) , b u t t h e s e are n o t s l o t s , t o be f i l l e d i n , b u t d i r e c t i o n s a s t o how t o f i l l i n an agent s l o t i n a higher o l d e r e n t i t y c a l l e d a template whlch c o n s i s t s of whole formulas: e s s e n t i a l l y , a network based on an agent formula, an a c t i o n formula and an o b j e c t fdrmula (although any of t h e s e may be d u m i e s ) . Schank has no e q u i v a l e n t t o formulas f o r nouns o r a d j e c t i v e s , o r any p a r t of speech o t h e r than verbs. Thus, i f the formula (10) f o r \"break\" i s placed a t t h e a c t i o n node of a Lemplate, i t s agent p r e f e r e n c e txies t o ensure t h a t a formula f o ran animate e n t i t y w i l l go a t t h e corresponding agent node. There i s a m e t r i c f o r t h i s i n i t i a l p a r s i n g and matching described i n(Wilks   1972(Wilks    , 1975a.The r o l e of the formulas a s d a t a f o r p a r s i n g I h e r e i s a s follows: templates, matched onto sentences and clauses expressed a s s t r i n g s of formulas, t r y t o pick up s u r f a c e s u b j e c t s and u s e t h e formulas as blue p r i n t s f o r electing t h e b e s t template, by trying t o ensure t h a t t h e s u r f a c e s u b j e c t i s a l s o t h e agent. I n t h e i n i t i a l matching of t h e he harmner broke t h e window'' t h i s w i l l not be 11 p o s s i b l e because t h e formula f o r hammer\", t h e s u r f a c e s u b j e c t , w i l l be placed a t t h e f i r s t ( o r agent) node of t h e template, However, s i n c e t h e formula f o r \"break1' w i l l be a t t h e corresponding a c t i o n notle t h e system w i l l know a t m y point i n l a t e r processing t h a t t h e s u r f a c e s u b j e c t i s n o t t h e underlying (animate) agent as p r e f e r r e d by t h e \"break\" formula a t t h e a c t i o n node. of khe same template, This c l a s h w i l l , a s we shall see, lead t o l a t e r ext'raction inferencing t h a t d e a l s 'with t h e ei!\"gatiw paradigm in a u n i k r m manner. But t h i s c l a s his only observable i n a system which b u i l d s s t r u c t u r e s t h a t r e t a i n t h e l r preferences b u i l t i n : i.e. n o t i n one l i k e Schanks' t h a t j u s t f i l l s ia 11 s l o t s i n d i c t i o n a r y e n t r i e s . Fillmore's anomalous Naon found Harry t l sleeping\" causes no t r o u b l e here: t h e formula f o r find\" expresses a preference f o r an animate agent, In t h e template f o r t h i s sentence,",
                "num": null,
                "uris": null
            },
            "FIGREF22": {
                "type_str": "figure",
                "text": "t i s not s a t i s f i e d by t h e s u r f a c e s u b j e c t (a formula f o r noon\", with a head ( W N POINT)) b u t no case frame i s v i o l a t e d , Formulas a r e not used a s parsing mechanisms t o deal with p r e p o s i t i o n s t r u c t u r e s i n English, This i s done by o t h e r s t r u c t u r e s c a l l e d p a r a p l a t e s , t o which I w i l l now turn, s o as t o givc a f u l l e r account of them than i n t h e p a s t , i n (Wilks 1975a), I n what follows templates w i l l be w r i t t e n i n s h o r t form: not a s t r i p l e s of complex formula t r e e s , but a s square brackets round the s u r f a c e phrase o r c l a u s e t h a t t h e template i s f o r , The words a r e c l u s t e r e d within t h e brackets so a s t o r e p r e s e n t t h e t h r e e nodes of the template, Case i n p a r a p l a t e s P a r a p l a t e s a r e s t r u c t u r e s employed t o a s s e r t a c o n n e c t i v i t y between two templates, t y p i c a l l y between a template r e p r e s e n t i n g a main c l a u s e and one r e p r e s e n t i n g a p r e p o s i t i o n a l phrase, Thus, i f we were analysing (20) John l e f t h i s c l o t h e s a t t h e c l e a n e r s which would be represented i n i t i a l l y i n t h e system by two templates, t h e c o r r e c t p a r a p l a t e , when applied, would a s s e r t a s p a t i a l l o c a t i o n case ( LOCA) t i e between t h e two templates a s follows: [John l e f t h i s + clothes] LOCA G[( n = Dumv Agmt) a t the + cleaners]Each p a r a p l a t e corresponds t o one of t h e cases i n the inventory, which i s the same as t h e inventory used t o c o n s t r u c t formulas.Many parap l a t e s , however, may correspond t o a s i n g l e case, A p a r a p l a t e has t h e form of two template-skeletons connected by a l a b e l i n d i c a t i n g t h e",
                "num": null,
                "uris": null
            },
            "FIGREF23": {
                "type_str": "figure",
                "text": "t ma?ches e i t h e r p a r t of t h e paraplate m s t have formulas t h a t s a t i s f y the functiops i n the corresponding p a r t of the p a r a p l a t e , I f the functions i n both p a r t s of a paraplate are s a t i s f i e d by a p a i r of templates (and the template f o r the prepositional phrase i s nlormally considered t o be the right-hand p a r t , though t h i s need not correspond t o a c t u a l o r d e r of occurrence i n text') then t h e case l a b e l of t h e parap l a t e i s a s s e r t e d i n the representation as holding between the two templates. In e a r l i e r d e s c r i p t i o n s I have distinguished paraplates from inference r u l e s , but i n f a c t they can p e r f e c t l y w e l l be seen a s a form of inference r u l e s a s Schanis has argued. However, the e s s e n t i a l r o l e of p a r a p l a t e s i s as parsing s t r u c t u r e s f o r p r e p o s i t i o n a l phrases. So, i f we were representing \"~o h n picked up the s t a t u e made oul of wood on t h e t a b l e a f t e r lunch\" we would expect paraplates f o r the various case dependencies to c r e a t e t i e s a s follows: i n d i c a t e s source case, and TLOCA, time location. The p r e f e r e n c e semantics system a p p l i e s p a r a p l a t e s immediately a f t e r matching templates and chaosing t h e ''most p r e f e r r e d \" ones (Wilks 1975@). I n o p e r a t i o n , the system i n p u t s small English paragrdphs onl i n e , produces a semantic s t r u c t u r e f o r them, and from t h a t generates a French t r a n s l a t i o n (Wilks 197313, 1975a; ~e l j S k o v i t s 1973), Thus, a s regards Engliqh, t h e system i s an a n a l y t i c one, and i s faced with t h e standard problem t h a t a s i n g l e English p r e p o s i t i o n can introduce many cases (and can, of course, o f t e n be t r a n s l a t e d by a number of French prepositioris depending on t h e case). L e t u s consider \"by\", f u d c t i o n i n g i n t h e following sentences, a l l of which may be considered t o s t a r t , l i k e ( 2 3 ) , with \"He l e f t Lugano by ..,\", where I have i n d i c a t e d t h e apparent (though d i s p u t a b l e ) c a s e of t h e l a s t c l a u s e a t the r i g h t of each l i n e : (23) He l e f t Lugano by c o u r t e s y of t h e p o l i c e",
                "num": null,
                "uris": null
            },
            "FIGREF24": {
                "type_str": "figure",
                "text": "a p l a t e s a r e six-place e n t i t i e s , n o t a l l of whose p l a c e s need be f i l l e d , corresponding t o Agerit-of-first-template, Action-of-firsttemplate, Object-of-first-template, and so on f o r t h e second template.Here are f o u r p a r a p l a t e s t h a t should match onto t h e templates f o r t h e s e n t e n c e s above w i t h corresponding numbers, L i k e t h e sentences, t h e p a r a p l a t e s w i l l a l l have t h e same left-hand s i d e , which i s w r i t t e n o n l y ) (WHERE S I G N )The p a r e n t h e s e s c o n t a i n i n g t h e formula p a r t s a r e a l l t o b e i n t e r -p r e t e d as matching onto correspondibg p a r t of a template i f and only if t h e l a t t e r has the mentioned s u b p a r t s as i t s head of formula. Thus p a r a p l a t e (24)' matches (24) because t h e formula f o r f o r t h e o t h e r correspondences of sentences and p a r a p l a t e s , which w i l l then a s s e r t t h e c a s e l a b e l t i e w r i t t e n a t t h e r i g h t hand end of the sentence i n each case (and on t h e corresponding arrow i n t h e p a r a p l a t e ) a s hcildipg between t h e correspon'ding template p a i r , $ i s a dummy p l a c e h o l d e r , *DO covers a wide class of a c t i o n i , as does *REAL of e n t i t i e s , The p a r a p l a t e s above (24) ' t o (29) ' a r e a s u b l i s t of t h o s e stacked under t h e name of t h e p r e p o s i t i o n \"by\" and under t h e p r i m i t i v e a c t i o n -MOVE t h a t occurs i n a c t i o n p o s i t i o n of t h e l e f t s i d e of each p a r a p l a t e , There w i l l a l s o be o t h e r s u b l i s t s of p a r a p l a t e s stored f o r \"by\" under o t h e r p r i m i t i v e a c t i o n s , For example, (30) This p a i n t i n g i s by T i t i a n would, d u r i n g i t s a n a l y s i s , a l s o access t h e p a r a p l a t e s under \"by\" b u t i n t h i s c a s e t h e s u b l i s t under BE, The s u 6 l i s t s of p a r a p l a t e s , f o r a given p r e p o s i t i o n , t h a t are names by d i f f e r e n t p r i m i t i v e a c t i o n s a r e not ordered w i t h r e s p e c t t o -each o t h e r , However, w i t h i n any s u b l i s t , such a s t h e s t a c k (24) ' t o (29)' above, t h e p a r a p l a t e sa r e ordered, i n t h a t t h e p a r a p l a t e a t t h e t o p of t h e s t a c k i s a p p l i e d f i r s t , t h e next only i f t h a t f a i l s and so on. L e t us r e t u r n t o t h e example t o s e e why, F i r s t , INST case i s l a r g e l y a d e f a u l t case f o r MOVE a s i t i s cued i n by \"by\", i n t h a t almost any e n t i t y can b e an instrument h e r e i f we have no reason t o beli@ve i t is anything e l s e . Thus t h e more s p e c i f i c (24)' must be applied b e f o r e (25)' i n o r d e r t o match d i r e c t i o n case f o r (24) ' s i n c e , i f t h e o r d e r were reversed, (25) ' might match with what \"ought\" t o match w i t h (24)'. W e could imagine something very s p e c i f i c i n (25)' t o match t h e formula f o r c a r (such a s a formula expressing \"thing f o r moving humans\"), b u t t h a t would r i s k missing rt by c a t t l e truck\" which i s n o t normally used f o r t r a n s p o r t i n g people. However, i t should be n o t i c e d h e r e t h a t t h e p a r a p l a t e s a s d e s c r i b e d do n o t have t h e power t o d e t e c t an anomaly such a s \"John walked home by car\". The templat&matching t h a t s e n t e n c e would be q u i t e p r o p e r l y t i e d ( a s INST) by (25)'. Any anomaly i n i t would have t o b e revealed by l a t e r i n f e r e n c e , This i s an example of what Lyons (op.clt.) means by t h e c r i t e r i a of a h a l y t i c systems bein'g ( i n i t i a l l y ) weaker than those of p r o d u c t i v e systems. A s t r o n g e r b u t s i m l l a r argument a p p l i e s t o p u t t i n g (28) ' a5ove (29)' i n a p r e f e r e n c e s t a c k , given t h e v e r y weak c r i t e r i o n embodied i n (29)', t h a t almost any a c t i o n upon any p h y s i c a l o b j e c t would s a t i s f y i n d e f a u l t ; and one could extend t h i s t o p u t t i n g a (23)' above a l e s s s p e c i f i c (26) ' , The main p o i n t h e r e i s a s follows: i t i s c l e a r t h a t snch parap l a t e s must b e a p p l i e d i n some o r d e r , b u t it i s n o t c l e a r t h a t they should b e slmply ordered. For example t h e r e i s no reason why (28) ' should be o r d e r e d with respect t o (24)'.",
                "num": null,
                "uris": null
            },
            "FIGREF25": {
                "type_str": "figure",
                "text": "There i s no formal t r o u b l e expressing a p a r t i a l o r d e r i n g of t h i s s o r t procedurally. I argued i n (Wilks 1975a) t h a t i n many c a s e s we might expect more than one template a t t a c h e d t o a fragment a t t h i s s t a g e and t h a t w e would \" r e s o l v e t h e ambiguity\" by p r e f e r r i n g whichever template matched h i g h e r up such a p r e f e r e n c e stack of p a r a p l a t e s . Order, then, i s important i n such paraplates s t a c k s , even i f i t t u r n s o u t t o be o n l y p a r t i a l o r d e r . If such a s t a c k i s o n l y p a r t i a l l y o r d e r e d then we may, i n t h e c a s e of a p r e p o~i t i o n a l p h r a s e w i t h two o r more c u r r e n t t e m p l a t e s competing t o r e p r e s e n t i t , expect adraw a t some stage: t h a t i s t o s a y , two p o s s i b l e t e m p l a t e s f o r t h e same p h r a s e may match w i t h p a r a p l a t e s t h a t are n o t o r d e r e d w i t h r e s p e c t t o each o t h e r . Such a s i t u a t i o n might w e l l correspond t o an example s u c h a s : (31) John jumped on t h e box where t h e l a s t p h r a s e i s g e n u i n e l y ambiguous between TO and LOCA c a s e ( i n t h e s e n s e i n which t h e s u b s t i t u t i o n of \"onto\" f o r \"on\" could o n l y be i n t e r p r e t e d a s TO). Thus under \"on\" we would p r o p e r l y e x p e c t MOVE ( f o r \"jump\") p a r a p l a t e s f o r TO and LOCA t o be m u t u a l l y unordered.",
                "num": null,
                "uris": null
            },
            "FIGREF26": {
                "type_str": "figure",
                "text": "h e r e d arm where t h e c a s e dependence o f t h e p r e p o s i t i o n a l p h r a s e c o u l d b e e i t h e r i n s t r u m e n t a l o r a s p e c i f i c a t i o n of which g i r l i t was.I n such c a s e s one might imagine sorie e x p e c t a t i o n a l f o r c e t o b e drawn from t h e c a s e c o n t e n t of t h e formulas.Thus, i f \"beatp' h a s an i n s t r u m e n t group i n its formula, we can conceive of u s i n g t h i s f a c t t o d e c i d e t h e draw on 11 the grounds t h a t \"beat\" r e a l l y expects\" an instrument, so why n o tg i v e i t o n e h e r e , and s e t t l e t h e q u e s t i o n . And t h e r e might indeed be somc p s y c h o l o g i c a l grounds f o r doing t h a t , But i n general, i t i s c l e a r t h a t t h e p a r a p l a t e s do not f u n c t i o n e x p e c t a t i o n a l l y , except i n t h e sense t h a t by looking f i r s t a t t h e p a r a b l a t e s most p r e f e r r e d i n t h e s t a c k ( f o r a given preposition and b a s i c a c t i o n ) one could be s a i d t o be expecting i t , And t h a t sense of \"expectation\" i s p e r f e c t l y c o n s i s t e n t with t r y i n g more than one candidate template, f o r a given phrase o r clause, a t t h e same time a g a i n s t t h e p a r a p l a t e stack, a s we might have t o in. a sentence l i k e (33) He beat the dog with a branch where we would n o t only be resolving the cake of t h e p r e p o s i t~o n a l phrase, b u t a l s o , and a t the sme time, the word sense of \"branch1', where t h e senses of \"branch\" might give r i s e t o a corresponding number of templates and we would always p r e f e r the one t h a t matched f u r t h e r up t h e p a r a p l a t e s t a c k , a s i n (Wilks 1975a),The reader may have been s t r u c k by an overlap of case content between t h e p a r a p l a t e s and formulas, F o r example, i n (10) t h e formula f o r \"break\", t h e head a c t i o n CAUSE has a dependent (THING INST).",
                "num": null,
                "uris": null
            },
            "FIGREF27": {
                "type_str": "figure",
                "text": "However, although t h e t w~ f o m s of information overlap t h e r e i s no reason t o b e l i e v e they a r e t h e same: t h e p a r a p l a t e s r e f l e c t t h e c a s e ambiguity of English p r e p o s i t i o n s w h i l e the formulas r e f l e c t only t h e s o r t s of c a s e i n f e r e n c e s one might subsequently want t o make ( i n e x t r a c t i o n , s e e below), There i s no reason t o suppose t h a t one form f o r d a t a can s u f f i c e f o r both t h e s e a c t i v i t i e s , However, Schank does make t h i s assumption, and t h e p o i n t h e r e may become c l e a r e r a f t e r some b r i e f r e c a l l of h i s p o s i t i o n on t h e p o i n t , I n (Schank 1973) an account i s given of case i n f e r e n c e s from a p r i m i t i v e a c t i o n . I n (34) John shot t h e g i r l ~i t h a r i f l e t h e underlying p r i m i t i v e i s PROPEL which 5 o m s t h e c e n t e r of any c o n c e p t u a l i z a t i o n f o r \"shoot\". An obligat'ory INST i n f e r e n c e i s always made f o r PROPEL o r , as ~t , was p u t e a r l i e r , EGT i s p a r t of t h e ( o b l i g a t o r y ) c a s e frame f o r PROPEL, and a l s o (aS i t h a w e n s ) an optiona l p a r t of t h e Fillmorean s u r f a c e frame f o r \"shoot\". I n ~c h a n k ' s scheme I n s t r u m e n t a l case i z~v o l v e s t h e i n s e r t i o n of another p r i m i t i v e a c t i o n i n t o t h e c o n c e p t u a l i z a t i o n o g u t t h a t i s n o t e s s e n t i a l t o rhep r e s e n t p o i n t which i s t h a t , f o r Schank, PROPEL \"expects\" an i n s t r ument, and c r i t e r i a can be p o s t u l a t e d such t h a t a r i f l e w i l l f i t those c r i t e r i a , i n a way t h a t h a i r w i l l n o t , a s i n (35) john-shot t h e 6irl w i t h long h a i r .Thus, f o r (34) an INST i n t e r p r c k t i o n i s achieved f o r \"with\", w h i l e i n (35), by d e f a u l t , t h e h a i r i s t a k e n a s a s p e c i f i c a t i o n of t h e g i r l .These s o l u t i o n s a r e , of c o u r s e , c o r r e c t b u t t h e d i f f e r e n t a s p e c t s of t h e phenomena do n o t seem t o connect i n p r o c e d u r a l t e r n s . So, f o r example, Schank i s p e r f e c t l y w e l l aware of t h e c a s e ambiguity of t h e p r e p o s i t i o n \"with\", and even l i s t s f o u r forms of i t ( i b i d , p. 231) corresponding to d i f f~r e n t cases, aloqg w i t h f o u r 11 c~n c e p t t u a l r e a l i z a t i o n s f o r t h e s y n t a c t i c item 'with (noun)' \", and specifies t h a t t h e y should be checked i n o r d e r \" f o r conceptual v a l i d i t y u , j u s t as p a r a p l a t e s a r e , The correspondence t o t h e n o t i o n s d e s c r i b e d I n t h i s paper (and i n Wilks 1973, i n , t h e same volume) i s reasohably clear: t h e i n f e r e n c e of a n instrument from PROPEL corresponds t o an instrument s p e c i f i c a t i o n i n a formula f o r \"shoot\" (and f o r Schank 1 i t a l s o comes from shoot': r a t h e r than more g e n e r a l l y from PROPEL, since o t h e r w i s e i t could n o t b e so s p e c i f i c about t h e instrument being a gun, as PROPEL does n o t d e a l g e n e r a l l y i n guns); whereas t h e ordered l i s t of case p o s s i b i l i t i e s f o r \"with\" i s n o t developed, b u t corresponds roughly t o a p a r a p l a t e s t a c k f o r \"with\", But h e r e i s t h e problem: i n Schank (1973) t h e two forms of i n f o r m a t i o n do n o t a c t u a l l y meet i n any g e n e r a l way, Schank m i t e s as i f t h e l i s t of p o s s i b l e c a s e f u n c t i o n s of \"with\" i s g e n e r a l (i.e. a c t i o n independent), b u t w e have shown t h a t i t may w e l l be s p e c i f i c t o d i f f e r e n t p r i m i t i v e a c t i o n s , i n that t h e r e may w e l l b e a s e p a r a t e p a r a p l a t e s u b s t a c k f o r e~r h p r i m i t i v e a c t i o n , and may be no s h o r t gel.yral l i s t of f u n c t i o n s of \"with\", $ Schank s u g g e s t s t h a t t h e -example (34) i s r e s o l v e d b e c a u s e PROPEL alld/or \"shoot\" \"expecrs a n instrument\"$ b u t that p i e c e of i n f o r m a t i o n does n o t , and cannot, t a k e accoupt o f t h e p r e -e x i s t e n t ambiguity of \"with\"* . i t j u s t happens To f i t t h i s example, b e c a u s e PROPEL \"expects\" an i n s t r u m e n t and i n s t r um e n t a l c a s e i s p u t a t t h e t o p of t h e ambiguity l i s t f o r \"with\" ( i b i d , But w i l l t h i s c o i n c i d e n c e hold i n g e n e r a l ? I f w e had been d e a l i n g n o t w i t h PROPEL b u t w i t h some e t h e r p r i m i t i v e , i t i s n o t c l e a r t h a t t h e same o r d e r f o r \"with\" would h e l p , I degcribed e a r l i e r a I1 M O m p r i m i t i v e u n d e r l y i n g leave\" ( t h e p r i m i t i v e and i t s name a r e n o t i m p o r t a n t , on-ly t h a t i t i s n o t PROPEL), and considered s e n t e n c e s l i k e ( 3 6 ) 1 l e f t Lugano w i t h Ywo pounds i n my pocket 11 which shows t h a t i n s t r u m e n t a l i t y i s u n l i k e l y t o b e t h e e x p e c t a t i o n 1 ' for \"with\" f r o m whatever p r i m i t i v e u n d e r l i e s \"leave\". Hence Schank' s l i s t f o r . \"with\" i s p e r h a p s s p e c i f i c t o i t s u s e w i t h PROPEL, and h e must develop sqme s t r u c t u r e analogous t o p a r a p l a t e s and not, assume $ i.e. a \" d i c t i o n a r y e n t r y \" f o r a p r e p o s i t i o n ( l i k e a p a r a p l a t e subtrack) w i l l be a f u n c t i o n of a p r i m i t i v e actioxl, t h a t t h e i n s t r u m e n t a l i t y of PROPEL s u f f i c e s : My p o i n t i s simply t h a t any attempt t o i n t r o d u c e g e n e r a l i t y h e r e w i l l show t h a t another s o r t of information s t r u c t u r e i s r e q u i r e d t o d e a l w i t h one s p e c i f i c t o p a r t i c u l a r p r e p o s i t i o n s , and t h a t g e n e r a l c a s e i n f e r e n c e s ( l i k e i n s t r u m e n t a l i ' t~f from PROPEL) could o n l y h e l p i n s p e c i a l cases. Any develppment of \"wlth\" l l s t s i n t o p a r a p l a t e substack s t r u c t u r e s would, I b e l i e v e , show t h a t t h e expecta t i o n a l i n f e r e n c e s from t h e p r i m i t i v e s p l a y a f a r smaller r o l e I n analysis,because t h e c r i t e~i a they contaxn w i l l have been p r a c e d u r a l l y expressed somewhere e l s e , It i s t r u e , however, t h a t t h e r e i s a s t r o n g e x p e c t a t i o n of a guninstrument from \"shoot\", and t h i s phenomenon does n o t f i t e a s l l y i n t o t h e p a r a p l a t e scheme. It might b e necessary t o ensure t h a t i f a s p e c i f i c instrument were mentioned i n a verb formula then t h a t could o v e r r i d e subsequent p a r a p l a t e matching when it; a r o s e , However, such s p e c i f i c instruments a r e n o t t h e norm f o r PROPEL, f o r this p r i m i t i v e presumably a l s o u n d e r l i e s \"break\" \u00a3-or ~chanl;, and t h e r e i s no such presumption of s p e c i f i c instrument with t h a t verb, Indeed, such an e x p e c t a t i o n would be bound t o l e a d t o e r r o r i f any o b j e c t found a f t e r \"with\" were taken a s an instrument, r a t h e r t h a n using a b r e a d t h -f i r s t approach l i k e p a r a p l a t e s t h a t considers t h e ranked p o s s i b i l i t i e s f o r t h a t preposition and t h a t a c t i o n . This i s e s p e c i a l l y t r u e f o r a system l i k e Schanks t h a t has no back-up and no possibility of recovery i f i t goes wrong, Formula and p a l a p l a t e information do sometimes overlap, and I suspect SchankTs ~y $ t e m , t h a t assumes t h a t t h e same mechanism can serve both f u n c t i o n s , draws much of i t s f o r c e by choosing an example from an a r e a l i k e ( 3 4 ) where they do overlap. But how, i n t h e examples of (23)-(29), could we work i n genexal with t h e ambiguity of \"by\" by assuming t h a t MOVE had sdme s t r o n g s i n g l e c a s e e x p e c t a t i o n , over and above t h e f a c t t h a tsome case p a r a p l a t e ( l i k e (24)' f o r T O ) would have n e c e s s a r i l y t o Appear a t t h e t o p of any substack. How could Schank express t h e action-dependent o r d e r i n g of t h e r e s t of t h e substack? Another p e c u l i a r f e a t u r e of SchankTs treatment of t h e s e phenomena i s h i s suggestion ( i b i d . p. 232) t h a t any o r d e r i n g of p r e p o s i t i o n f u n c t i o n s must go \"from the g e n e r a l t o t h e s p e c i f i c \" . Such l i m i t e d d a t a as a r e a f f o r d e d by substacks l i k e ( 2 4 ) ' -(29)' i n d i c a t e t h a t i t may t u r n o u t t o be t h e o t h e r way round, because t h e most general (right-hand) c r i t e r i a tend t o appear i n t h e p a r a p l a t e a t .the bottom of t h e substack, Much though not a l l , of what has been s a l d about Schank's approach a p p l i e s t o Riesbeck's (1974) implementation of i t . It i s a g a i n a s t r o n g l y e x p e c t a t i o n a l system, i n p r a c t i c e i n ~i e s b e c k ' s case, and t h a t l e a d s t o t h e mentioned d e f e c t s of a d e p t h -f i r s t a p p r~a c h ( i f implemented without back-up, a s i t i s ) , s i n c e t h e whole notion of \"preference\", i n case a s elsewhere (Wilks 1973a), depends on following a number of possibilities b r e a d t h -f i r s t f o r a while before comparing them and committing the system t o only one. Only i n that way could one t r e a t , i n a general way, examples l i k e (33) t h a t r e q u i r e comparison o f how h i g h up a p a r a p l a t e s t a c k a given i n t e rp r e t a t i o n comes r e l a t i v e t o o t h e r s . ~l e s b e c k ' s system generates expectations ( c a l l e d \"requests\"f from main sentence verbs and uses t h e s e t o a n t i c i p a t e t h e case p a r t s of t h e seritence. The main d i f f e r e n c e from Schank's approach, and which makes i t a curious implementation of i t , is t h a t i t d e a l s with p r e p o s i t i o n s not, as one might expect, by primitive a c t i o n s s e e k i n g c e r t a i n cases, but by verbs seeking a c t u a l p r e p o s i t i o n names. e x m p l e , prevent\" he seeks \"from\" and \"by\", jus-t as t h e computational l i n g u i s t i c systems of t h e f i f t i e s and s i x t i e s d i d , It a l s o has less power than Schank's theory (and p a r a p l a t e s ) i n t h a t Riesbeck s p e c i f i c a l l y says ( i b i d . p . 95) that r e q u e s t s (including p r e p o s i t i o n r e q u e s t s presumably) a r e not o r d e r e d , A l l t h i s seems t o r e q u i r e that a l l p o s s i b l e p r e p o s i t i o n sequents be s t o r e d for every s u r f a c e verb i n the system which leaves l i t t l e scope f o r semantic g e n e r a l i z a t i o n ( t h e aim of t h e whole e x e r c i s e , presumably). The strong point of d i f f e r e n c e , between Riesbeck's system and t h e one described here, concerns t h e role of u n i n h i b i t e d \"expectations\", $ The examples presented h e r e suggest t h a t one cannot base a system simply on t h e o b s e r v a t i o n t h a t if someone u s e s \"by\" a f t e r \"leave\" then i t w i l l b e expected t o i n t r o d u c e an instrument, For \"~e l e f t Lugano by day'' does not f i t t h a t and y e t cannot be sa4d t o be violating any axp,ectation, It i s simply t h a t t h e s t r u c t u r a l ambiguity of p r e p o s i t i o n s muse be accounted Eor i n a s t r u c t u r e d and general manner l i k e t h a t o f f e r e d by t h e pawaplates. The u s e f u l n e s s of expectationg i n any f i e l d , depends on (a) t h e high chances of Tt being r i g h t , and (b) knowing what t o do when i t i s f r u s t r a t e d , and ~i e s b e c k ' s system seems t o l a c k both t h e s e d e s i d e r a t a ,",
                "num": null,
                "uris": null
            },
            "FIGREF28": {
                "type_str": "figure",
                "text": "For f u r t h e r d i s c u s s i o n of the i s s u e , see (Wilks 1975c, pp. 40-45) Case e x t r a c t i o n inferences Extraction inferences produce new template-like forms from t h e c a s e content of formulas embedded i n source templates: those t h a t have been matched with t e x t sentence f ragrnents, They a r e s t y l e d \"template-like\" h e r e only to i n d i c a t e t h a t they have n o t been matched w i t h input t e x t , And hence the i n f e r r e d p r o p o s i t i o n s they r e p r e s e n t have not n e c e s s a r i l y been s t a t e d e x p l i c i t l y i n the input t e x t , Let us f i r s t see t h e e f f e c t of doing t h i s , and then the mechanism t h a t does i t , I n what follows, we extend t h e \"short form\" of templates (obtained by w r i t i n g square brackets round English words, c l u s t e r e d a t t h r e e nodes t o show t h e d i s t r i b u t i o n of formulas i n t h e f u l l template) by w r i t i n g e x t r a c t i o n s a s English words i n s i d e double square brackets, L e t us consider (37) John f i r e d a t a l i n e of s t a g s with a shotgun",
                "num": null,
                "uris": null
            },
            "FIGREF29": {
                "type_str": "figure",
                "text": "u l l e t ) a+ sha tgun 1 The extracted templates are t i e d by dotted l i n e s t o the source template from which they have been extracted, and t h e case name on the dotted l i n e shows the case type of the extraction, The inferences cover both those t h a t must be t r u e ( l i k e the OBSE extraction, slnce to f i r e a t a l i n e of s t a g s 1s n e c e s s a r i l y to f i r e a t stags) and those, l i k e the GOAL extraction, t h a t a r e only l i k e l y . These extracted templates a r e not printed out (in the way i n which, MARGIE (Rieger 1974) daes) f o r they do not i n themselves c o n s t i t u t e t e s t a b l e output.",
                "num": null,
                "uris": null
            },
            "FIGREF30": {
                "type_str": "figure",
                "text": "s t r u c t u r e s a r e one and the same, r e q u i r i n g o n l y a s i n g l e representatlon, The e t t r a c t i o n mechanism c o n s i s t s of a 'specialist'l(to use Winograd's t e n d f o r each case (and f o r CAUSE, which i s t r e a t e d a s a semigcase during e x t r a c t i o n ) . It i s c a l l e d a f t e r an i n i t i a l semantic representation f o r a t e x t has been b u i l t up a s templates t i e d together by p a r a ? l a t e s and anaphora t i e s (See Wilks H73c, 1-975b) An e x t r a c t i o n , r e s u l t i n g i n a new doubJe-square-bracketted template, a s i n (38) above, i s made f o r each case (or CAUSE) sub-formula a t the top l e v e l of t h e formulas of each source template.",
                "num": null,
                "uris": null
            },
            "FIGREF31": {
                "type_str": "figure",
                "text": "may be considered a semi-full-form of [ 30 hn f i r e d + a t l i n e ( o f + s t a g s ) l i n t h a t t h e c e n t r e node has been expanded t o i t s formula b u t the o t h e r 4 t w o n o b s are l e f t i n \" s h o r t f o m \" , Each top-level case (or CAUSE) subfornula i s e x t r a c t e d by t h e a p p r o p r i a t e \"case s p e c i a l i s t 1 ' which has access t o the a d d i t i o n a linformation grovided by t h e whole template i n which t h a t formula i s embedded (and, a s we s h a l l see, t o o t h e r templates a s w e l l ) . By way of i l l u s t r a t i o n , we w i l l look a t t h e OBJE and GOAL e x t r a c t i o n s from t h e a c t i o n formula i n (39). dependent of t h e OBJE case i n (39) shows t h a t \" f i r e j a t \"l i n e ( + o f + s tags) I p r e f e r s an ANImate o b j e c t , but t h e formula i s i n a template whose o b j e c t i snot animate ( i t i s \" l i n e \" ) , and s o we have a f a i l e d preference, However, an animate o b j e c t (stags) i s a v a i l a b l e as a dependent of t h e surface o b j e c t i n t h e template, The e x t r a c t i o n process takes t h e form of f i l l i n ga new copy of t h e source template, and imposing t h e a v a i l a b l e p r e f e r r e d animate o b j e c t , t o y i e l d : C C John f i r e d + a t s tags] 1 ~x t r a c t i n g the GOAL, case i s more complex. A s we saw e a r l i e r , i n t h e formula (10) f o r \"break\", t h e real dependent of a GOAL case i s an a s s e r t i o n group, The appropriate f u l l dependent t o the left of GOAL i n (39) i s (*HUM SUBJ) (*AN1 OBJE) (STRIK $) where $ i s i n d i c a t e s t h e p o s i t i o n of \"GOAL\" i n (39). T h i s i s an a s s e r t i o n group as it appears i n a formul J. The e x t r a c t i o n takes t h e form of r e w r i t i n g t h i s in tmplate form (from SOV t o SVO form) and attempti% t o f i l l i n i t s nodes wit11 f u l l formulas matching t h e preferences: so, on r e w r i t i n g t h e above we get: an agent and o b j e c t can be i n s e r t e d fromithe whole template (39) so a s t o match the preferences expressed ( i f we incorporate the above 0BcJ;EFj e x t r a c t i o n as w e l l ) . Thus, we g e t an extracted \"short form\" template (converting STRIK to \" s t r i k e s \" f o r uniformity of expression, though it r e a l l y remains a s t h e p r i m i t i v e ) : The e x t r a c t i o n s f o r INST and CAUSE r e q u i r e t h a t we a r e a b l e to look o u t s i d e t h e source template f o r con\u00a3 irmation to o t h e r templates, .-Here, only INST receives any such c o n f i r m a t i m because t h e r e i s no mention of any \"bullet\" o r o t h e r m i s s i l e t h a t would confirm ( (THLNG MOVE) CAUSE) ) . Since the' p r i m i t i v e dependent on INST i s THING. the \"INST s p e c i a l i s t \" produces a copy template whose primitive a c t i o n i s USB (if the dependent had been another type of e n t i t y , the r u l e could have been d i f f e r e n t ) . The INST-tied template matched t o \"with a shotgun\" i n (39) i s accessed by t h e INST s p e c i a l i s t t o provide t h e f u l l o b j e c t formula f o r USE yielding the e x t r a c t b r i e f l y a t how e x t r a c t i o n s cope i n a novel way with t h e standard ergatxve paradigm rehearsed a t the beginniug of the I t paper, and i n p a r t i c u l a r w i t h the i n t e r e s t i n g farms The window broket' and h he h m -e r broke the window\", I n i t i a l l y t h e s e w i l l r e c e i v e a template match so a s t o y i d d a r e p r e s e n t a t i o n even more s u p e r f i c i a l than t h a t given i n Simmons' system. Under e x t r a c t i o n however, appropriate r e p r e s e n t a t i o n s a r e obtained and cued by t h e f a c t that t h e s u p e r f i c i a l s u b j e c t formula i n t h e templates does n o t , i n e i t h e r c a s e , s a t i s f y t h e animate agency p r e f e r e n c e of t h e formula (10) f o r \"break\" which i s t h e action node of both tkmplates, Thus h he window broke\" y i e l d s m i i t i a l l y : (41) C the+window broke Now, on e x t r a c t i o n , ~~~' s u E J (agent) s p e c i a l i s t sees not o n l y t h a t (a) t h e s u r f q c e s u b j e c t (window) does n o t specify satisfy the (*AN1 S U B J ) p r e f e r e n c e of (lo), b u t (b) t h e same surface subject does s a t l s f y the (*RQJYSOB OBJE) preference of (lo), which i s f i l l e d by o n l y a dummy i n the source template ( 4 1 ) , Thus t h e \"SUBJ s p e c i a l i s t \" produces a copy template w i t h t h e agency p r e f e r e n c e s a t i s f i e d : s p e c i a l i s t \" correspondingly produces : [:[: R broke window11 and t h e s e are immediately conflated, On t h e g e n e r a l p r e f e r e n c e m i l k s 1973a) p r i n c i p l e o f producing t h e f u l l e s t r e p z e s e n t a t i o n p o s s i b l e , as the e x t r a c t i o n : f [ some+animate broke windowl 3 where t h e agent formula (now, of course, a t l u e agent, n o t a surface s u b j e c t ) i s merely (THIS *",
                "num": null,
                "uris": null
            },
            "FIGREF32": {
                "type_str": "figure",
                "text": "[ h a m e r broke window1 and again consider the e x t r a c t i o n s f rom4 (10) f o r \"break\" when a t t h e a c t i o n node of (421, W e can produce t h e same SUBJ e x t r a c t i o n as f o r (41) above, and we can again, using t h e \"INST s p e c i a l i s t \" , produce an e x t r a c t i o n with USE as i t s main action prrmitive, as from (39), b u t i n t h i s example che d u m l e s i n the extracted template can h a t be f i l l e d i n from an INST-tied template, as i n ( 4 1 ) , and as would be the case with \"John broke t h e window with a hammer\", but from the surface s u b j e c t of the source template , i t s e i f . Thus we obtain an e x t r a c t i o n scheme f o r (42):",
                "num": null,
                "uris": null
            },
            "FIGREF33": {
                "type_str": "figure",
                "text": "Not a l l e r g a t i v e v e r b s would r e c e i v e i d e n t i b a l treatment, Thus, the relat'ion between:(44) John moved t h e stone and (45) The stone moved would not be the same a s between the corresponding examples above f o r \"break1', The a c t i o n formula f o r \"move\" i n (44) would have CAUSE a s head, but t h e a c t i o n (again \"move\") i n (45) would be a d i f f e r e n t sense and so a d i f f e z e n t formula, whose head would be MOVE, and which would not contain an instrument s p e c i f i c a t i o n . So we would not expect any extraction: , though w e might well expect an extractiorl equivalent t o (45) from the template f o r (44) There can be l e g i t i m a t e d i s p u t e about the l i m i t s t o which case = t r a c t i o n s should go: so, f o r example, some w r i t e r s (Parker-Rhodes 1975) would consider t h a t \"John wrote a l e t t e r \" should be an inference from 1 received a l e t t e r from John , and he describes a l a t t i c e scheme in which an AGENT case is generally and l e g i t i m a t e l y i n f e r r e d from a SOURCE case, That could only be achieved with the \"SOURCE s p e c i a l i s t 1 ' i n t h e present system from a very f u l l , s p e c i~i c , formula f o r ' l e t t e r T (and e x t r a c t i o n s can come from nouns and not only from verbs) i n which a petson) source was specified a t t h e top level of the formula, Two p o i n t s should be n o t i c e d , i n conclusion, about t h i s method of producing a qemantic s t r u c t u r e a p p r o p r i a t e t o the e r g a t i v e paradigm, F i r s t , t h e \"case s p e c i a l i s t s \" a r e not i n general dependent upon t h e p a r t i c u l a r f o m u l a i n which they o p e r a t e ( e t h e word t h e f o m u l a is f o r ) , n o r upon t h e p r i m i t i v e t h a t i s t h e fonmlla head, Each s p e c i a l i s t does, of course contain ranked s i d e c o n d i t i o n s but-th-y a r e n o t , except i n s p e c i a l cases, dependent on t h e head of t h e formula,",
                "num": null,
                "uris": null
            },
            "FIGREF34": {
                "type_str": "figure",
                "text": ", t h i s way of d e s c r i b i n g the e r g a t w e paradigm i s n a t u r a l l y d i f f e r e n t from F i l l m o r e ' s because, as has been noted, he i s p r i m a r i l y concerned w i t h r e s t r i c t i o n s of t h e generation (=productron !) of s u r f a c e sentences by means of h i s ordered Subject Selcztiozl Rule: \"I\u00a3 t h e r e i s an agent i t i s t h e s u b j e c t , e l s e i f t h e r e i s an experi e n c e r i t i s t h e sub,iect, e l s e i f t h e r e i s an instrument .... and so on\".",
                "num": null,
                "uris": null
            },
            "FIGREF35": {
                "type_str": "figure",
                "text": "i r s t , t h a t t h e p r e s e n t system i s p r e f e r e n t i a l i n that John does n o t have t o be.myrked as an agent: no confpsion would b e caused here, f o r example, i f he was used as an instrument, Secondly, and more importantly, t h e c o n s t r u c t i o n here of t h e e r g a t i v e r e p r e s e n t a t i o n f o l l o w s n o t1 )o n l y from the a p p l i c a t i o n of t h e c a s e specialists\", b u t from t h e g e n e r a l r u l e of p r e f e r e n c e(Wilks 1973b)  t h a t as f u l l ( o r maximally redundant) a r e p r e s e n t a t i o n should b e produced a s p o s s i b l e . TRis i s r e q u i r e d , i n a d d i t i o n t o t h e case s p e c i a l i s t s , t o produce t h e e x t r a c tion CCsome+animate broke window31 from, t h e less r&undnmt ergative, (411, On t h i s groudY I wuu1d argue t h a t t h e phenomenon is d e a l t with h e r e u s i n g a more g e n e r a l rule, and moreover, a g e p e r a l rule t h a t d r i v e s t h i s whole p r e f e r e n c e semantics system of a n a l y s i s . ILL d r e c e n t paper, Charniak (t975) has argued t h a t A1 systems t h a t use case n o t a t i o n i n t h e i r a n a l y s i s of n a t u r a l language a r e n d t r e a l l y making u s e of case, and t h e present system i s one of those he c r % t i c i s e s i n t h i s connection, Argunients along the same general l i n e s a r e t o b e founa i n Welin (1975). I t h i n k Ch-rniak's arguments a r e important and s t i m u l a t i n g , though i n sode ways misguided. They have t h e paradoxical sidk-effect of showing ;hat not even Fillmore 1 s r e a l l y making use of case. I s h a l l b r i e f l y sur,marise seven aspects of ~h a r n i a k ' s complex argument, an$ make some reply, (i) The p a $ i b i~n Argument, Charniak d i s t i n g u i s h e s between case n o t a t i o n and p o s i t i o n n o t a t i o n ( i b i d . pp. '3~4). I f we have an event of Jack opening t h e door with a key then, f o r Chamiak, a case form of that would be: (PREDICATE EV OPEN) (AGENT EV JACK) (OBJECT EV DOOR) (INSTRUMENT EV KEY) whereas OPEN(JACK, DOOR, KEY)' would be pure p o s i t i o n n o t a t i o n . He argues t h a t many A 1 systems claim t o be using case n o t a t i o n but i n f a c t a r e using only p o s i t i o n notbtion: dqing p l a c e s a r b i t r a r i l y t o pass p a r t i c u l a r case arguments t o a predicate, (ii) The evacuation argument. Charniak argues t h a t t h e standard b e n e f i t of case --t h e prdduction o f , say, t h e e r g a t i v e paradigm f o r 11 open\" from a case r e p r e s e n t a t i o n a s i n (i) above, together with a s u b j e c t s e l e c t i o n r u l e --ean be obtained equally well with t h e p o s i t i o n n o t a t i o n , (ibid. p. 17), Iq o t h e r words, the whole business can b e evacuated of case a l t o g e t h e r , with the case names remaim-11 i n g perhaps a s mnemonic v a r i a b l e names\" (ibid. p . 27). ( i i i ) T h e l i n e a r i t y argument. Charniak argues t h a t the e;xpression of meaning i n terms of case i s non-linear. Schematically, i f i t was l i n e a r i t would look l i k e : while t h e meaning of \"see\" would have EXPERIEHCER where the above has AGENT, However, i n A1 systems i t tends t o be don-linear, l i k e : MEANING (LOOK) =MEANING (OBSERVE ,AGENT) o r , i n o t h e r words, t h e meaning of MTANING(AGENT) cannot be computed independently of knowing t h a t t h e agent i s the agent of OBSERVE. (iv) T h e p r o d u c t i o n argument, Charniak argues t h a t t h e n o t i o n of case f o r Fillmore i s es s e n t i a l l y connected with i t s productive g e n e r a t i v e r o l e i n c o n t r o l l i n g s u r f a c e grammaticality, and t h a t since A I systems are preoccupied with a n a l y s i s , they a r e not making use of case, (v) The case-content argument, Charniak argues t h a t we a r e never t o l d what i t i s t o b e an agent as such, T h i s i s a subargument of ( i i i ) above t h a t one cannot compute MEANING(AGENT) independently, (vi) The s u r f a c e a n a l y s i s argument.",
                "num": null,
                "uris": null
            },
            "FIGREF36": {
                "type_str": "figure",
                "text": "He argues t h a t some such analyses, like t h e formula gi-ven e a r l i e r f o r \"drink\", q c t u a l l y exclude t h e s u r f a c e forms, but I showed e a r l i e r t h a t t h i s i s not so.( v i i ) The case-inf erence argument. Charniak argues t h a t 11 a system u s i n g case f o r semantic r e p r e s e n t a t i o n should provide case infe.zencesll.",
                "num": null,
                "uris": null
            },
            "FIGREF37": {
                "type_str": "figure",
                "text": "based understanding systems\" t o i t s o r i g i n a l Fillmorean r o l e as a c l a s s i f i e r of superficial f o r m s l l i k e p r e p o s i t i o n and o t h e r case cons t r u c t i o n s i n English. I would paraphrase Charniak's W r n e n t on this 1 t p o i n t , perhdps u n f a i r l y , a s claiming t h a t case\" takes i t s meaning from s u r f a c e phenomena and t h e r e f o r e any attempt t o l o c a t e i t only i n a deep t! semantics, wholly divorced from such phenomena, i s t o make case\" meani n g l e s s . My answer t 6 t h a t , and i t i s a s e r i o u s charge, i s t h e p r e s e n t a t i o n , h e r e and elsewhere, of a system t h a tboth p a r s e s s u r f a c e English by means of caseand expresses i t s underlying semantics i n t h e same n o t a t i o n , The t h i r d a s p e c t of CharniakTs arguments, l i k e (v) and ( v i i ) , i s h i s concern with t h e p r e c i s e content of t h e cases, This b o t h e r s everyone, including Fillmore, who had beer. unable t o provide i t , and so i f t h e production of such d e f i n i t i o n s i s made a necessary condition of u s i n g c a s e then F i l l a o r e , l i k e everyone e l s e , f a i l s the t e s t . Let Be, i p conclusion, reply t o t h e arguments above i n a l i t t l e more detail. ( i > i s , I t h i n k , simply confused, The two (case and p o s i t i o n ) a r e simply equivalent forms of information, provided one knows t h a t t h e \" f i r s t argument i s t h e agent\" etc. e t c . Charniak himself seems t o s e e t h i s when he admits ( i b i d . p, 4) t h a t one could have case n o t a t i o n without case, and case without case n o t a t i o n , ( i i ) shows only t h e t r i v i a l i t y t h a t any cornputatidla1 system c o u l d have a l l its top level notions replaced by gensyms, l i k e WQ5.567, and f u n c t i o n i n t h e same way, This would not however show t h a t t h e n o t i o n s being programmed were n o t e s s e n t i a l t o t h e meaningful f u n c t i o n and i n t e r p r e t a t i o n o f t h e system. I n any system doing p r o p o s i t i o n a l l o g i c , t h e n o t i o n AND can properly be replaced by a t r u t h -t a b l e v e c t o r 1000, but t h a t does not show that t h e notion AND represents i s vacuous, o r a mere mnemonic convention. ( i i i ) does not e s t a b l i s h t h a t systems must b e wholly l i n e a r , i n Charniak's sense. However, i t would c l e a r l y be a defect i n a system i f , every t i m e an inference about AGENT was made, t h e system had to look and s e e wl4at verb i t was the agent o f , o r what the p r i m i t i v e a c t i o n coding t h e verb was, Some examples of Schank's t h a t Charniak discusses do show t h a t t h i s access t o the head primitive i s somerlmes made, and Charnzak then argues t h a t the case i s ipso f a c t o dispensible, and t h e inferehces a r a b a t t e r seen a s inferences from that7-head Grimi t i v e aqtion, But, a s T pointed out. when d i s c u s s i n g e x t r a c t i o n , the I I case s p e c i a l i s t s i ' i n , t h e present system, a t l e a s t , do n o t generally inake such reference) back co t h e acr;fun head, So, f o r example, t h e \"GOAL-specialist\" did not have t o find o u t , b e f o r e operating, what the a c t i o n was t h a t a given GOAL was a goal o f .",
                "num": null,
                "uris": null
            },
            "FIGREF38": {
                "type_str": "figure",
                "text": "p r e c i s e d e f i n i t i o n of t h e cases beyond t h e most general i n d i c a t i o n s , such a s that agency i s concerned w i t h i n i t i a t i n g f o r c e etc. T h i s c r i t i c i s m i s t r u e f a r beyond case: i t a p p l i e s e q u a l l y t o notions o f thinghood, causation and movement etc, It i s j u s t a f a c t t h a t b a s i c semantic notions arelvague,and perhaps n e c e s s a r i l y so.It i s c e r t a i n l y no ground f o r lament, and does n o t prevent us speaking p r e c i s e l y i n everyday l i f e . Provided, t h e r e f o r e , t h a t a programed system can perfarm p r e c i s e l y using these vague notions, we can demand no more.The f a l l a c y i s t o imagine t h a t underneath t h e vague notions used i n thought, speech and w r i t i n g , t h e r e must l l e p r e c i s e c r i t e r i a f o r t h e i r use. This i s a p e r s i s t e n t f a l l a c y t h a t has received much a t t e n t i o n i n Twentieth Century philosophy.In t h e p r e s e n t system, a s i n many others, s p e c i f i c a t i o n s ( s e l e c t -i o n a l o r p t e f e r e f i t i a l ) a r e given of the agents of p a r t i c u l a r a c t i q n s [human f o r \"think? a$imate f o r \"see\", and: so on), One could argue t h a t \"agents a s such\" a r e , i f anything, the union df a l l such agents, but t h a t w i l l n o t , and cannot, y i e l d an i n t e n s i o n a l d e f i n i t i o n of t h e notion. E x t r a c t i a n operating on such specif i c a t i d n s ( t h e \"SUBJ-11 s p e c i a l i s t \" ) y i e l d s Pnferences such as t h a t some animate does so and s?\"' b u t these do n o t meet Charniakt s utll~and i n (vii) above. I t e i s worth notifig-at t h i s p o i n t t h a t Charniak puts h i s demand (or t e s t f o r \"using zase\") i n such a way ( i n p o i n t s (v) and ( v i i ) above) t h a t i t i s hard t o s e e what could s a t i s f y i t : what r e s u l t of computation could we conceivably expect as t h e v a l u e of MEANING(AGENT)? I would maintain t h a t e x t r a c t i o n s , a s described above, are most c e r t a i n l y \"case in\u00a3 erences\", b u t a r e not \"the s e t of inferences one can make about X , knowing only t h a t X i s i n case C A S~I y l b i d . p. 19) because t h a t s e t i s n o t t h e s o r t of t h i n g one wants t o know i n a language understanding s y s tern. The \"case s p e c i a l i s t s \" i n ogreration produced i n f e r e n c e s t a k i n g information from t h e a p p r o p r i a t e p o i n t s i n t h e r e p r e s e n t a t i o n ; but from knowing merely t h a t John i s an agent nothing i n t e r e s t i n g follows o r could follow! (And from t h a t f a c t nothing follows t o t h e e f f e c t t h a t agency i s vacuous, a s Charniak seems t o t h i n k ) , The most t h a t could follow would be t h a t John does somethfng o r i n s t i g a t e s sortne a r t . W e -could h a t d l y i n f e r a t t h a t point t h a t John w a s animate, because t h a t i s presumably information g e n e r a l l y attached t o \"John\", and, i n any case, some verbs ( l i k e \"calculate1') may have a s e l e c t i o n or preference r e s t r i c t i o n on t h e i r agent t h a t is more general than animate, so t h a t being animate could n o t follow from being an agent. To i n f e r anything u s e f u l , a s i n t h e e x t r a c t i o n s e a r l i e r i n the peaper, we have t~ look n o t j u s t a t what John i s t h e agent of (which would make t h e whole t h i n g c i r c u l a r f o r Charniak, by argument kl), but p o t e n t i a l l y a t a l l t h e information a v a i l a b l e i n t h e representation-. It i s , perhaps, s i g n i f i c a n t t h a t agent i s t h e case t h a t makes Charniak's p o i n t best. The o t h e r e x t r a c t i o n s answer i t b e t t e r , f o r example i n t h e way t h e I' INST-specialist\" c r e a t e s an e x t r a c t e d Pemplate w i t h a c t i o n USE; o r t h e way t h e \"GOAL-specialist\" e x t r a c t e d an i n \u00a3 erence \"John s t r i k e s stags\" i n a manner independent of t h e f a c t t h a t t h e source a c t i o n was \" f i r e at\", f o r t h e \"GOAL-specialist\" did not consult t h e source verb name o r i t s head p r i m i t i v e a c t i o n CAUSE. O r , again, t h e way i n which common-sense inference r u l e s (Wllks 1973c, 1975a, 197513, but n o t discussed i n t h i s paper) would operate onany such extracted GOAL template and i n f e r t h a t the corresponding agent wanted t h e GOAL (whatever i t was). And an inference r u l e l i k e t h i s l a t t e r can operate only i f -i t can l o c a t e the corresponding agent, Sob when we extracted t h e GOAL equivalqrit to \"~o h n s t r i k e s stags\", we tsould have ( i n order t o operare t h e new inference r u l e ) t o be a b l e t o f i n d t h a t John was the agent of t h e extracted template i f we whnted t~ in\u00a3 e r f u r t h e r t h a t John wanted t o s t r i k e t h e s t a g s . The agent (John) w i I l of course be t h e formula i n t h e f i r s t p o s i t i o n of t h e extracted template, and Charniak might then say t h a t t h a t does nok r e q u i r e agency b u t only access t o a f i r s t template position. This would b e a r e v e r s i o n t o t h e bad p o s i t i o n argument",
                "num": null,
                "uris": null
            },
            "FIGREF39": {
                "type_str": "figure",
                "text": "behind Charniakls position\" argument (i) is that no A I systems have such general r o u t i n e s , and hence an agent f o r one verb c o u l d be s t o r e d quite d i f f e r e n t l y f r o m that f o r a n o t h e r because there are no Such routines that would ever find this out! And that I believe is not t h e case, . Charniak, E m (1975) A b r i e f on c a s e . Working p a p e r s No. 22 (Castagnol* I n s t . f o r Semantic and C o g n i t i v e S t u d i e s ) .Fillmore, C. (1968) The dase f a r q s e , i n {eds) Bath & Harms, U n i v e r s a l s i n L i n g u i s t i c Theory.(New York: Hol t , R i n e h a r t F i l l m o r e , C. (1971) Some problems f o r c a s e grammar, i n (ed) ~'~r i e n , L i n g u i s t i c i c s : developments of the s i x t i e s (Washington DC : Georgetown UP), 35-56, F i l l m o r e , C. (1972) S u b j e c t s , s p e a k e r s and r o l e s , i n (eds) Davidson & Haman, Semantics of N a t u r a l Language (Dordrecht: ~e i d e 1 ) , 1 -2 4 . F i l l m o r e , C. (1975) The c a s e f o r c a s e reopened, L e c t u r e a t ~n i v , of B i e l e f e l d , Hays, D, (1964) Dependency t h e o r y : a formalism and some o b s e r v a t i o n s Language, 40, 511-525. -H e r s k w i t s , A. (1973) The g e n e r a t i o n of French from 2 s e m a n t i c r e p r e s e n t a t i o n . Memo No. AM 212 ( S t a n f o r d , C a l i f . Computer S c i e n c e Dept,) m o AoCeMos -18, 264-2740 Wilks, Y o (1975b) N a t u r a l language systems within t h e A 1 paradigm. Memo No, AIM-337 ( S t a n f o r d , C a l i f . : Computer Science ~e~t , ) , and Amer, J n l . Comp. Ling., -40 (1976), Wilks, Y. ( 1 9 7 5~) Seven theses d n a r t i f i c i a l intelligence and p a t u r a l lahguage, Mufio NO. 16 (Castagrtola: I n s t , for Semantic and C o g n i t i v e Studies), Woods, W. (1970) T r a n s i t i o n network grammars f o r n a t t~r a l language American Journal of Computational Linguis tiesTHIRD I N T E R N A T I O N A L J O I N T CONFERENCE ON P A T T E R N R E C O G N I T I b O N 8-11 NOVEMBER 1976 PROCEEDINGS: $18,75 FOR IEEE MEMBERS, $25 FOR OTHERS Pattern Recognition, P 0 , Box 639 S l l v e r Spring, Maryland 20901 R A N S I T I O N NETWORK GRAMMARS S M C~O U & K 5 FU a S U P E R V I S E D L E A R N I N G TECHNIQUE TO. I D E N T I F YSHORT NATURAL LANGUAGE SENTENCES D Coulo~I and D Kayser GRAMMATICAL INFERENCE BASED ON PATTERN R E C O G N I T I O N G B Porter A P A R S I N G ALGORITHM F O R IMPERFECT PATTERNS AND I T S A P P L I C A T I O N R Mohr and J P .HatonSYNTACTIC FEATURE E X T R A C T I O N FOR SHAPE R E C O G N I T I O N T Pavlidls INFERENCE OF REGULAR EXPRESSIONS L Miclet A MODEL FOR DEDUCING SEMAtdTI'C RULES OF INFERENCE USING A LAYERED O R G A N I Z A T I O N FOR REPRESENTING E X P E R I E N T I A S Kadota, S Hayashi, M Yam~rnoto, 'S Yajima, and M Yasuda THE R E C O G N I T I O N O F CHINESE CHAR.ACTERS (KANJI) USIdplG T I M E V A R I A T i O N OF P E R I P H E R A L BELT P A T T E R N T F u j i t a , M Nakanlshi, and K Miyata AN O P T I C A L CHINESE CHARACTER READER K Sakal, S H i r a i , TKawada, and S. ArnanoO P T I M A L I N F O R M A T I O N ECONOMY I N THE M U L T I F O N T DCR r N P U T SYSTEM 0Albertsen, E Muenster, and P PonsmngEXTRACTION OF THE MOST PROBABLE ENTRY FROM A DIRECTORY M NadlerA COMF9RISON OF C L A S S I F I C A . T I O N R E S U L T 6 BY MAN AND BY MACHINEH Niem,nnIMPRINTED CHARACTER R E C O G N I T I O l j U S I N G T E M P L A T E MATCHING ON S P A T I A L COMPLEX FREQUENCY P L A N E --E X P E C I A L L Y F O R KATAKANA LETTERS AND NUMERALSTanaka and K , OzawaLINGUISTIC.APPLICATIONS --AND NATURAL LANGUAGE PROCESSINGChalrmen T Kasvand, Canada, and N G Zagoruiko, USSRAPPLICATION O F A WEB GRAMMAR MODEL T U AN EARTH RESOURCES S A T E L L I T E P I C T U R E J Brayer and K FU NATURAL LANGUAGE UNDERSTANDING AS A PATTERN R E C O G N I T I O N PROBLEM J Maroy LEARNING STRUCTURED PATTERNS FROM ~X A M P L E S F Hayes-Roth and J McDermott RECOGNITION OF YAND-DRAWN F L O W C H A R T S C Suen and T Radhakrlshnan FOOT P R I N T S : A REPRESENTATION FOR R E S T R I C T E D M O T I B N I N OU'TDOOR SCENES B Bullock and S , Dudan1 CHARACTEAISING S Y L L A B L E S A S SEQUENCES O F MACHINE-GENERATED L A B E L E D SEGMENTS O F CONNECTED SPEECH: A STUDY I N SYMBOLIC PATTERN L E A R N I N G U S I N G A CONJUNCTIVE FEATURE L E A R N I N G AND C L A s s I F I c n T r o N S Y S T E M F'Hayes-Roth and J. BurgeSTRUCTURAL A N A k Y S I S OF H I E R A R C H I C A L SYSTEMS E R Caianiell0and R C a p o c e~l l iSPEECH RECOGNITIONChairmen K S Fu USA, a n d J I Nagumo, JapanAUTOMATIC EXTRACTION O F D I S T I N C T I VE FEATURES FROM CONTINUOUS SPEECH R De Mori and P Laface APPROXIMATION TECHNIQUES FOR THE LEARNING OF SEQUENTIAL SYMBOLIC PATTERNS M S FOX a n d F Hayes-Roth $PEECH UNDERSTANDING SYSTEM LITHAN AND SOME A P P L I C A T I O N S T S a k a i and S Nakagawa WORD RECOGNITION IN A MULTI-TALKER ENVIRONMENT USING S Y N T A C T I c METHODS R L Kashyap and M C M l t t a l PHONETIC S T R I N G R E T R I E V A L USING IMPERFECTLY LABELED PROTOTYPES S K Chang American Journal of Computational Linguistics Microfiche 56 : 7 4 S O C I A L S C I E N C E : A F C E T WORKSHOP Program o f a workshop held at Gif-sur-Yvette, 4 November 1976 AUTOMATIC ,DOCUMENTATION AND URBANISM. ADEQUACY AND V A L I D I T Y O F A DOCUMENTATION LANGUAGE F O R THE S A T I N 1 SYSTEM J . C . Daumas and E Raby, Centre de Documentation sur ltUrbanisme, Paris I N F O R M A T I C S I N S O C I A L S C I E N C E . TOWARD A CHANGE IN MODES O F THOUGHT A. Degenne, Laboratoire dtEconomle et de Sociologie du Travail, Aix-en-Provence ON SOME L I M I T A T I O N S O F SYSTEMS F O P EL,ABORATE C O N S U L T A T I O N O F DATA BANKS Guenoche and J V l r b e l , Laboratoire d'Informatlque pour l e s Sciences d e llHomrne, Marseille These papers are published In the Cahlers of an AFCET worklng group For infomation E Chouraqui & J : V l r b e l L a b o r a t o i r e d ' Informatlque pour l e s Sciences de L'Homme 31, Chemin Joseph Alguier Marseille 9 ' American Journal of Computational Linguistics M~crofl che 56 75 INTERNATIONAL MEETING PROGRAM LOUISA ( L L I N G~l~~l~~L L Y O R I E N T E D U N D E R S T A N D I N G AND I N D E X I N G SYSTEM F O R ABSTRACTS) -A PROGRESS REPORT J Nogl (Belgim)",
                "num": null,
                "uris": null
            },
            "TABREF0": {
                "html": null,
                "content": "<table/>",
                "text": "I n t r o d u c t i o n . . . . . . . . . . . . . . . . . . . . . . 4 . . . . . . . . . . . . . .",
                "num": null,
                "type_str": "table"
            }
        }
    }
}