{ "paper_id": "J75-3009", "header": { "generated_with": "S2ORC 1.0.0", "date_generated": "2023-01-19T02:40:42.554871Z" }, "title": "I . SOME PSYCHOLOGICAL CONSTRAINTS ON SENTENCE COMPREHENSION MODELS", "authors": [], "year": "", "venue": null, "identifiers": {}, "abstract": "In t h i s paper I consider t h e question of how an a u t o m a t i c sentence recognizer would have to look in order to be compatible wLth present psycholinguistic knowledge about speech c o m p r e h m h The basic premise is that psycholinguistic considerations are of potential interest to computational theories (s e e , e. g. , ~chank(1972)) Let me begin by summarizing some characteristics of speeck processing which we know either from experiments, or which are i n t u i t i v e l y clear. First, there i s some evidence t h a t the clause i s a u n i t of processing. For instance, ~aplan(1972) showed that after a clause boundary is passed, the cons-bitaent words b f t h e completed clause are relatively inaccesai'ble, as measured by w m d recognition l a t e n c y. The e f f e c t was independent of the serial position of the word for which recognition time was tested. T h i s suggests that sentences are processed clause by c l a u s e , w i t h o n l y the semantic content regularly retained after the clause boundary is passed. The surface words (and g fortiori the syntactic s t r u c t u r e) of the clause would tend t o be erased after each clause boundary. 1 *Thie paper i s based on chapter VII o f my doctoral dissertation (~e i m o l d (forthcoming)). I wish to thank Thomas G. Bever , Jame-a Higginbotham, and D *Terence Langendoen f o r h e l p f u l suggestions. l~h e f o r t i o r i J refers to the fact that the syntactic s t r u c t u r e Another study supporting the clause as unit of proceasing is Abrams 6 ~e v e r (1 9 6 9). These a u t h o r s found t h a t r e a c t i o n time to s h o r t bursts of noise \" c l i c k s \") superimposed on sentences was longer for clause-final clicks than for clause-initial ones. This would point to the clause as unit of perception, under the assumption t h a t processing is more intensive towards the end of a p e r c e~t u a l unit, and that reaction time to e x t e r n a l stimuli is a valid ineicator of t h e intensity of internal procesging. or a review of other studies in support of the clausal processing theory, the reader i s referred t o Podor, Bever & ~arre%t(1974), where arguments a r e a l s o given f o r the clause as a decision point across which ambiguitiis are ,normally at l e a s t , not carried-.) Secondly, it seems t h a t as we l i s t e n to speech, we simultaneously have access to both the syntactic and semantic propertiee of what we hear. That is, there appears to be ~a r a l l e l orocessing of t h e syntax and the semantics of a clause. One finding explained by this assumption i s t h a t so-called \"irreversible\" passive sentences like (1) are perceptually no more complex t h n n their active counterparts (the air1 ~i c k e d the rlower,in this c a s e). By c o n t r a s t , 'reversiblew passives l i k e (2) take longer to verify visa -vis p i c t u r e s than t h e ~orre~pondin$ active sentences (~lobin(1.966)). presumably contain8 surface wards a s terminal nodes. Hence if the eyntax were regularly preserved the surface words should remain easily accessible, too.", "pdf_parse": { "paper_id": "J75-3009", "_pdf_hash": "", "abstract": [ { "text": "In t h i s paper I consider t h e question of how an a u t o m a t i c sentence recognizer would have to look in order to be compatible wLth present psycholinguistic knowledge about speech c o m p r e h m h The basic premise is that psycholinguistic considerations are of potential interest to computational theories (s e e , e. g. , ~chank(1972)) Let me begin by summarizing some characteristics of speeck processing which we know either from experiments, or which are i n t u i t i v e l y clear. First, there i s some evidence t h a t the clause i s a u n i t of processing. For instance, ~aplan(1972) showed that after a clause boundary is passed, the cons-bitaent words b f t h e completed clause are relatively inaccesai'ble, as measured by w m d recognition l a t e n c y. The e f f e c t was independent of the serial position of the word for which recognition time was tested. T h i s suggests that sentences are processed clause by c l a u s e , w i t h o n l y the semantic content regularly retained after the clause boundary is passed. The surface words (and g fortiori the syntactic s t r u c t u r e) of the clause would tend t o be erased after each clause boundary. 1 *Thie paper i s based on chapter VII o f my doctoral dissertation (~e i m o l d (forthcoming)). I wish to thank Thomas G. Bever , Jame-a Higginbotham, and D *Terence Langendoen f o r h e l p f u l suggestions. l~h e f o r t i o r i J refers to the fact that the syntactic s t r u c t u r e Another study supporting the clause as unit of proceasing is Abrams 6 ~e v e r (1 9 6 9). These a u t h o r s found t h a t r e a c t i o n time to s h o r t bursts of noise \" c l i c k s \") superimposed on sentences was longer for clause-final clicks than for clause-initial ones. This would point to the clause as unit of perception, under the assumption t h a t processing is more intensive towards the end of a p e r c e~t u a l unit, and that reaction time to e x t e r n a l stimuli is a valid ineicator of t h e intensity of internal procesging. or a review of other studies in support of the clausal processing theory, the reader i s referred t o Podor, Bever & ~arre%t(1974), where arguments a r e a l s o given f o r the clause as a decision point across which ambiguitiis are ,normally at l e a s t , not carried-.) Secondly, it seems t h a t as we l i s t e n to speech, we simultaneously have access to both the syntactic and semantic propertiee of what we hear. That is, there appears to be ~a r a l l e l orocessing of t h e syntax and the semantics of a clause. One finding explained by this assumption i s t h a t so-called \"irreversible\" passive sentences like (1) are perceptually no more complex t h n n their active counterparts (the air1 ~i c k e d the rlower,in this c a s e). By c o n t r a s t , 'reversiblew passives l i k e (2) take longer to verify visa -vis p i c t u r e s than t h e ~orre~pondin$ active sentences (~lobin(1.966)). presumably contain8 surface wards a s terminal nodes. Hence if the eyntax were regularly preserved the surface words should remain easily accessible, too.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Abstract", "sec_num": null } ], "body_text": [ { "text": "( 4 )~q i r l { w i t h a green hat who wore a green hat g r e e t e d John.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "", "sec_num": null }, { "text": "J o h n ate t h e cake C afterwards .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "aster the guests left.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "Z v l d e n t l y , w i t h a green h a t in ( 4 ) is r e l a t e d to who wore a meen hat, and t h e adverb afterwards in ( 5 ) can be replaced by f i l l 1 adverbial clauses like after t h e m 1 e s . t~ l e f t . (1) it should be a clause-by-clause processo2, where my n o t i o n of \"-lause\" includes some things traditionally regarded as phrases; as soon as the i n t e r p r e t a t i o n of a clause is completed, its syntactic structure is erased;", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "( 2 ) there should be parallel syntactic and semantic processing of each clause1 conetLtuents. An example f o r the three stages is given in (13).", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "(13) The boy laughed. We can now translate t h e structures 111 (13) i n t o E n g l i s h .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "PSR r (THEX 1 BOYX) (By) (E t . : PAST t ) I: L A U G H Y~ 3 ISRI (IIIEX~BOYX &HUMAN% & ~ADULTX m 0 * } ( 3~) ( 3 t~ PAST t & -FUTUaZ t * . ) C LAUGHyt & HUMANY & ANIKATEY~ & BLNFyt", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": ")", "sec_num": "5" }, { "text": "\"%e x such t h a t x. is a boy is invowed in somE vent suoh t h a t there is some y and some time which is PAST, and . y is laughing at time t.\" (14) (a) r t h e DDI a (THSV~ --) re-1", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The f i r s t , L e e , t h e preliminary S I i , sayst", "sec_num": null }, { "text": "(b) ~~O S T ~f l I ( E X ) E B O Y X I ( c ) Elaurrhed MVB PAST], (~' y ) (E t r PAST t) L L A U G~~~ 1", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The f i r s t , L e e , t h e preliminary S I i , sayst", "sec_num": null }, { "text": "Notice t h a t each of t h e deftnitions consists again of a prefix and a matrix. An example is given in ( 3 7 ) d~e r e as elswhere in t h i s paper, \"#\" atande for initial and \"$\" for final clause boundary.) ( 4 5 ) # J ohn btlieved $ ( LYcompll t h a t the 2ake was pcdecncd 4) If we assumed t h a t tpt2=t3 in ( 4 9 ) , then the t w o conjoined phrases pnd round and round* should be redundant in the same sense in which Fido ia a dog and is a doe: and is a dog is. However, (49) can q u i t e n a t u r a l l y be interpreted To formalize t h i s , we can make use of Pattern ittcbing.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The f i r s t , L e e , t h e preliminary S I i , sayst", "sec_num": null }, { "text": "There", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The f i r s t , L e e , t h e preliminary S I i , sayst", "sec_num": null }, { "text": "as", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "( T H E X~J O H N X ) L B E L I E V~~X an41 ( T H E~I C A K E~) [~~E D N E D~~", "sec_num": null }, { "text": "F o r instance, the encyclopedia would contain a p a t t e r n like (51), and t h e r e would furthermore be a meaning r u l e like (52).", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "( T H E X~J O H N X ) L B E L I E V~~X an41 ( T H E~I C A K E~) [~~E D N E D~~", "sec_num": null }, { "text": "(52) (AU.,tl,t2) LCAUSE l ) be p r e f i x h e a d v a r i a b l e s of a single type T (where V i precedes vj; f o r i< j(m) , and l e t ul,e.~,u, be type T argument places of MVB (where ui precedes u for i < j~n ) 8 then each V i n o t yet linked to any m a t r u -c o n s t i t u e n t is linked to the MVB-argument ui, unless MVB has the feature PASV in t h e syntactic structure.Far example, let ( 3 3 ) be t h e preliminary SR of (32):(32) J o h n gave t h e c a t some m i l k .", "type_str": "figure", "num": null }, "FIGREF15": { "uris": null, "text": "three prefixes a r e n o t yet linked to any matrix-constituent, and t h e i r head-yariablee &, y, p, are of the same type. Hence they a r e linked as in ( 3 4 ) , yielding, after simplification, t h e f i n d SR (35). Finally, 3 describes the svntax-sensitive rule. I t i s seneuive t o the fdature 3?Aw of the verb and l i n k s the constituents marked S7Mr OBJl, 0BJ2, and AGenT to the main verb, insofar as this has not been done by earlier strategies. f i r s t M V B -a r g u m e n t is linked to t h e p r e f i x head-~&t?ia5le pointlng to: {a) the SW-phrase if MVB is n o t marked PASV; (b) the AGT-phrase if t h e r e is such a phrase: (2) link theprefix head-variable pointing to 8 (a) an ORTI-phrase to the second MVB-argument1 (b) an OBJ2-phrase to t h e t h i r d WB-argument;( 3 ) if MVB is PASV Unk the prefix head-variable pointing to the SBJ-phrase to the remaining free (but not t h e f irst-) MVB-argument.", "type_str": "figure", "num": null }, "FIGREF16": { "uris": null, "text": "# The churchSBJ had been givenPASv t h e moneyow2 $ L#Pos~c~] by the bakerACT $ a PSRI f i r s t clalaset ( T~B~: C H U R C H~) ( T H E~~~O N E Y X ) ( E~~X~Z t l i I V~z~z~2~ 7 Links; x pdints to OBJ2. hence x = z g ; y pomt.8 to S 0 J of a PASV verb, hence y=z2 by o p t i o n ( 3 ) of rule ( 3 6 ) l FS3 f6r first clause8 ( T H E Y~C~~U R C X~) (THEX:MONEYX) (~2~) C G I V E~~Y~ 1 PSR f o r by-clauser (THEZIBAKERZ) Linkr z points to AGT-phrase, hence z=zl by option (lb)i V X I I a DIFFEi3ENT XOD3S OF PROCESSING L e t me interrupt here to consider t h e p r a c t i c a l problem of constructing an automatic sentence recognizer. Some aspects of t h e thapry I have j u s t sketched may n o t be optimal for a ~omputational model, even though they seem appropriate for a psychological m o d e l . For instance, to a pefson engaged in normal conversation accuracy of understanding is n o t very c r u c i a l most of t h e t i m e \" Often, t h e goal may only be t o '-get the essentials,\" and if some mistake & s made, it is simply corrected l a t e r on. This \"normal mode of processingw is what the p s y c h o l o g i c a l model sets out to d e s c r i b e . Now, in the c a s e of an artificial i n t e l l i~r t n c e system, one would probably demand h i g h e r accuracy, so as to minimize the need f o r correctionsm T h i s is, in some ways* similar to t h e s i t u a t i o n where you p u t subjects in a psychulinguistic experiment. They u s u a l l y abandon the \"normal mode of processingw very soon and instead employ the strategies that guarantee best performance for the s p e c i f i c experimental task they are faced w i t h . To give an example, consider t h e common type of experiment where a subject h a s to v e r i f y sentences l i k e those in 138). (38a) 5 precedes 13. (38b) 5 is preceded by 13. ( 3 8 4 13 is preceded by 5 . In o r d e r to i n t e r p r e t such sentences c o r r e c t l y , h i s knowledge t h a t 5 in f a c t precedes 13 i s of no help whatsoever to t h e listener, because true sentences occuf t o g e t h e r w i t h false ones in this game. Therefore, he will soon drop all semantic short c u t s --which he normally e a p l o y s -and interpret the sentences purely on t h e basis of their syntax. In a way, therefore, such experiments do n o t r e a l l y t e l l us anything about normal speech processing. Still, such conditions o'f heightened aecuracy may be just the onee we want to apply to the artificial i n t e l l i g e n c e system. Let us therefore consider how such a non-normal mode of processing could be s i n u l a t e d in our t h e o r y . Now, looking back at the Linking S t r a t e g i e s in (22) above (~lnking by Variable Type, Pattern latching, C o n t r a d i c t i n n -Xlimination, Canonical Order, ana Syntax-sensitive M V B -~u l e ) , there is indeed ar. obvious way of simulating the \"high accuracy procedure,\" namely by dropping type 2, 3 , and 4 strategies. 'Phese c o f l s t i t u t e the \" s h o r e cuts\" which work 90% of the t i m e , but sometimes l e a d to misinterpretations. Notice that the result is still not a syntax-governed model, because most of the linking would s t i l l be handled by the syntax-free ne'chod of Linking4by Variable Type, and syntactic btruc t a r e ~o u l d s t i l l be erased in clause-intervals, As a matter of fact, in this last respect I t h i n k it is possible to g o even further than I have done here, and erase syntactic s t r u c t u r e a f t e r each major clause constituent (i.e., after each NP, adverb, or main verb), retaining only its functional feature, which is then simply integrated into the semantic representatbn. I have already done t h i s here for t h e constituent ~a i n~e r~, s o in a way t h i s woula o n l y be a l o g i c a l extension of my proposal Looking at t h e syntax-sensitive MVB-Rule ( 3 6 ) , it is evident that it rerers exactly to those functional features, namely SBJ, O B J 1 , O B J 2 , MVB, and AGT. It would seem, then, that even for purposes of a r t i f i c ' i a l i n % l l i g e n c e , i t may be preferable to o p e r a t e w i t h a p a r a l l e l processing m o d e l , thereby minimizing t h e size of t h e syntactic gltructl~re and t h e amount o f s y n t a c t i c o p e r a t i o n s . Apart from this, I would like to argue t h a t t h e essential ingredients. of Pattern Matching and Contradiction-Elimination are still required for any adequaze t J l e o~. W s -b r i~g back to t h e sixth type of Semantic Linking Strategies, namely Alternative Linking Strategies. I X s W A L T E R N A T N E LINKING STRATEGIES\" Alternative Linking Strategies apply if t h e normal strategies (types 1-51 have failed t o produce a semantically acceptable reading. These strategies r e l y heavily on the Semantic Knowledge Rules. The most important a n d m o s t general (and the only one to be discussed in the present paper) is the Obvious Connection Strategy in ( 3 9 ) . It says, roughly, t h a t if a variable g cannot be linked to the R n B then if t h e encyclopedia contains a rule conne~ting q to same o t h e r entity 11' then try and link t h i s n e w entity y o to the MVB. (393 O b v i o u s Connection S t r a t e a r If the head-variable u c d a prefix ( Q T F~U I A U ) cannot be l h k e d to i t s appropriate IWB-argument V , and both u and Instance, consider again sentence (9b)l (9b) They published Wodehouae immediatela he came over. s indicated in (40), the o b j e c t of PUBLISH must be some r i t t e n work, and Wodehoyse of course does not q u a l i f y as uch. Therefore, the l5nR between x odeho house) and t h e o g i c a l o b j e c t y e of P U B L~~H is r e j e c t e d . N o t i c e that t h i s s j u s t the kind of \"semantic anomaly test\" which was central o the C o n t r a d i c t i o n -E l i m i f i a t i o n Strategy Its i n t u i t i v e asis is obviousr the listener normally assumes t h a t t h e peaker is t r y i n g to make sknse, and therefore, h e w i l l e j e c t a l l non-sensical interpretations. 40) They published Wodehouse. changes (THEX~WODMOUSEX) in the PSR to: {BY; (THEX r W O D M O U S E x ) (WRITTNWRK~ & CREATEX~)) Link added by Rule (391 I y'y', yielding the FSR: staying w i t h our sentenae ( 4 0 ) hey p u b l i s h e d h ode house), if you know t h a t Wodehouse was a w r i t e r you a l s o know t h a t he c r e a t e d w r i t t e n works3 and ths new entity i n t r o d u c e d by t h i s encyclopedic r u l e , namely \"written works created by W o d e h~u s e ,~ is the one which i s interpreted as . l o g i c a l o b j e c t o f t h e predicate PUBLISH. This then is the kind ofWsernantic detourw descrttJeb by t R e ObvTous ConnectLon strategy, and this strategy is applicable to t h e other sentences in ( 9 ) too 10 Now, I think t h e sentences in ( 9 ) exemplify something that happens all t h e t i m e in speech: namely o m i s s i o n o f t h e obvious. I a l s o think it w o u l d be extremely inconvenient if we had to ask people to use o n l y t h e i r best Sunday Grammar when conversing with an English-speaking r o b o t . Me m i g h t as well ask them n o t to use pronouns, or to speak at a c o n s t a n t p i t c h of 4.00 H e r t z . To be competitive, t h e r o b o t should undecstand Monday Gramar as well, and t h a t means, sentences l i k e those in ( 9 ) Xr CLAUSE-TO-CLAUSE LINKIKG For the remainder, 3: would like to d i s c u s s clause-to-clause 1Mcing. The d i e t i~l g u i s h i n g feature o f my p r o p o s a l , a s will 10 See ~chank(1972) f o r a d i f f e r e n t approach to t h e problem of recovering implicit information. He o u t l i n e s , f o r instance, a method by which \"I like books\" would be expanded i n t o t h e conceptual equivalent of \"I like to read books.\" SchankDs theory is baqed on wconceptual casesw and l e x i c a l decomposition r a t h e r than meaning postulates. For instance, he claims that \"John would be pleased by Il:ary8s going\" is a proper part of t h e meanhg of John w a n t s Marv to come home, w h i l e in my terms \"John would be pleased by !daryw s going\" is merely an allowable inference which may or may n o t be drawn. Failure to Eraw an a l l o w a b l e inference explains t h e possibility of holding contradictory b e l i e f s . For instance, somebody m i g h t judge John's uncle left to be t r u e , w h i l e at the same t i m e believing t h a t t h e brother o f John's mother or father left was false, because he failed to apply t h e mean%% rule relating uncle and brother of mother or f a t h e r . It is difficult to s e e how a theory b a s e d o n l e x i c a l decomposition would explain such f&c ts C e r t a i n l y , it w o u l d be unreasonable to claim that a person has not understood the sentence John's uncleJaft unless he a l s o is aware of the synonymy r e l a t i o n between t h i s sentence and the b r o t h e r of John's mother or f a t h e r Left be r e c a l l e d , is t h a t s y n t a c t i c structure is erased clause-byclause. The p r o b l e m , t h e r e f o r e , is to show t h a t sentences containing m o r e than one c l a u s e can in f a c t be interpreted correctly w i t h o u t r e f e r r i n g to t h e full syntactic structure of any (completed) earlier c l a u s e . Consider first wPost-clauses\" e , clause-final adverbs, temporal nouns, and preposition-phrases Tliey are already covered by the strategy of Linkillg by Variable Type discussed earlier. For instance, consider again sentence (25a). (yesterday t h e f a t h e r of t n e boy sang horribly in the b a t h . ) It c o n t a i n s t h e t w o P08-b-clauses h o r r i b l y and in the bath; and it was shown earlier how t h e y a r e linked to the main clalase by V a r i a b l e Type The o n l y smtac t i c information r e q u i r e d f o r t h i s operation was t h e marking of t h e MVB in t h e semantic representation* (41) indicates roughly how relative clauses are integrated. (~c t u a l l~, t h e r e are some complications here, b u t t h e y a r e irrelevant to t h e present discussion; the main p o i n t here is t h a t again no reference need be made t o t h e syntactic strracture of the f i r s t c l a u s e . For d e t a i l s , s e e ~e i m o l d ( f o r t hc o m i n a where c o o r d i n a t e c l a u q e s , comp&rison-clauses, and various s l~b j e c t l e o s complement-clauses a r e treated as well.) (41) # The g i r l (if who was tired 3 ) giggled q. (TMEX:GIRLX)[--~ and (~y ) [ T 1 i ( E~y l link, y-xj integrated structure a ( T~I E X I GIRLX & TIi-iEDx) C --7 Next consider circumstantial clairses l i k e because t h e kan~aroo i u m~e d in ( 4 3 ) . The s t r a t e g y for these c l a u s~s , stat-ed in (42), does again not refer to t h e syntactic structure of t h e f i r s t clause* (42) Circumstantial -Clause Ruler Substitute the matria of t h e main clause f o r t h e free s-argument of the J U N C t o r of t h e circumstantial clause, and join the prefix of t h e main clause before t h e p r e f i x of the circumstantial c l a t~s e . (43) # The boy wap nappy $ ( # l b e c a u s e J U N C J t h e kangaroo jumped$) (THEIS : BOYX) [HAPPYX 1 and ( T H X~ 1 K A N G A~) CBEGAUSL ( Q ,JUMPY) true f o r complement-clauses like t h a t t h e cake w a s ~o i s o n e d in (45). The corresponding strategy is given in ( 4 4 ) . (44) C o m~l e m e n t Clause Ru1e.r Substitute the SR o f t h e complement-clause f o r t h e free s-argument of the MVB of t h e main c l a u s e .", "type_str": "figure", "num": null }, "FIGREF17": { "uris": null, "text": "FSRr (TH-I JOHNX )CBKLIFNE$ x, ( T H E~~C A K E~) ( P O I S O N E D~) 3 7 In sum, t h e p r i n c i p l e of clause-by-clause e r a s u r e of syntactic structure seems ihdeed compatible w i t h t h e requirements of clause-to-clause lirling. The exception i s t h e feature MVB, and it was suggested earlier that this syntactic feature i s integrated into the semantic representation Notice a l s o that certain a s p e c t s of syntactic structure are reboverable from our semantic representations. For instance, the order of t h e prefixes in the SR r e f l e c t s the surface order of the NP'a of a clause. Hence, if certain constructions require access to such syntactic aspects, this i8 s t i l l n o t incompatible with erasure o f syntactic structure in clause intervals. or instance, coordinate clauses and certain subjectless complement clauses do often require identification of the surface s u b j e c t . For detailed discussion, see Reimold (forthcoming) . )X -I r THE \"TEMPORAL SEQUENCE STEIATEGY \"The last strategy discussed here concerns t h e tense of consecutive clauses. When we interpret sentences S1, 52 in a t e x t or sentence conjunct, where S3, and S2 have the same tense, we o f t e n assign a relative chrondlogy to the events described by these sentences. I will r e f e r to t h e p r i n c i p l e assigning such a chronology as t h e Temporal Sequence Strategy.", "type_str": "figure", "num": null }, "FIGREF18": { "uris": null, "text": "Then h i s head jerkedtl back a little and he l a y s t i l l . t2 Assigning PAST(^^) t o the f i r s t clause and PAST(^^) t o the conjoined clause in (46) does not account for actual comprehension: the listener knows that t2 is l a t e r than tl, even though there is no overt sequewe marker ( e s g * , b e f o r e , after, then) A s a first approximation, the strategy might be s t a t e d as f 011 ows t (47 ) Tem~oral Sequence stratem (preliminary) I Given t w o main or conjoined clauses C1, C2 such t h a t C1 precedes C 2 and t h e t i m e tl of C 1 has t h e same \"tense predicatew ( e . g . , PAST, FUTURE) as t h e t i m e t2 of C 2 , then assume t h a t DIRFRBC t p = (47) w i l l require several m o d i f i c a t i o n s . F i r s t , t h e r u l e h o l d s f o r c e r t a i n nan-tensed clauses as well, as illustrated by (48) (~h e s r cameto Baghdad p 044) 8 (48) Then he was o u t , across t h e Khan, back into the Suq. . The f u l l interpretation of (48) must specify that t h e t i m e of o u t precedes t h a t of acrcws, which in t u r n precedes that of bacq. Since across and back had no overt tense predicate in 7 i48), the strategy must somehow be Liberalized to include S U C~ C~S~S . In t h i s connection, consider a l s o (49)(~hey came to Baghdad. pa711 Never. I thought, would t h e plane land. It went round and round*:! and roilndt,. t 1", "type_str": "figure", "num": null }, "FIGREF19": { "uris": null, "text": "\" t h e plane went-round and then round and then roundN b e . . as 45. -DIRPREC t2 & _t3 DIRPREC t3). Next, t h e Temporal Sequence Strategy seems blocked if some general semantic principles \\in ~a r t i c u l a r , Pattern Matching) suggest a chronology caR#licting with that imposed by the Temporal Sequence Strategy. For instance, , h e broketl h i s a r m . H e f e l l t Z off his bike. Here the second sentence in (sob) is interpreted as preceding the f i r s t sentence in (50b) temporally, counter t o what the Temporal Sequence Strategy would prediict. The reason is obviousr there is a perceived causal connection between the sentences, such that the second sentence describes the cause of the first. Since a cause must precede its e f f e c t , t2 must precede tl in (50b).", "type_str": "figure", "num": null }, "FIGREF20": { "uris": null, "text": "have to add t h e f o l l o w i n g principle to t h e Temporal Sequence Strategyr (53) Causal Connection Constraint: If C1, C 2 a r e not conjoined by p d , c a l l up a l l Encyclopedic Rules in the intersection of t h e M V B V s of C1 and C2. If 4 DIRPREC tl is heuristically derivable from a pattern, then add t h i s link t u C2 and do not apply t h e Temporal Sequence Strategy. 11", "type_str": "figure", "num": null }, "FIGREF21": { "uris": null, "text": "a sentence like pe broke h i s arm and fell o f f his b i k e . cannot be interpreted as \"he broke h i s arm because he fell off h i s b i k e . tB the \"Short Events Principle\" discussed below) was l i g h t i n g t a cigarette a n d leavingt2 the raom. y , I come to \u20ache m o s t general and important r e s t r i c t i o n on the Temporal Sequence S t r a t e g y , Phis r e s t r i c t i o n is of a subtle semantic nature: it states that t h e s t r a t e g y is applicable to clauses C1, C2 o n 1~ if the events described by C1 and C2 are unlikely to h e simultaneous.Consider, fqr instance, t h e following sentence: (56) She j u s t stood t h e r e and looked at h i m . It is perfectly p o s s i b l e f o r someone to stand somewhere and at t h e same t i m e to l o o k at somebody. Hence no temporal sequence is imoosed on stood and looked. By contrast, in (57) (~h e u came to Baghdad, p.76) wen3 and stood must be interpreted as seqyenced, since one .cannot at t h e s a m e t i m e a somewhere and stand somewhere elser (57) She wenttl out from the bar onto the terrace outside and s t o o d % by the railing... Note that it will not do to d e f i n e . t h e condition of \"possible simultaneity' d i r e c t l y f o r verbs. We cannot say, e . g . , t h a t clauses containing stand and a must be\"sequencad.\" For instance, in (58) and (5% s t o o d and went w o u l d be interpreted as simultaneous, the reason being, of c o u r s e , t h a t different agents are involved, which makes simultanei-ty c onc e ivabl e . ( 5 8 ) Jack s t o o d by t h e window. Jane went t o the door. (-59) Jane went t o the door. Jack stood by t h e window. It is clear, then, t h a t a d e t a i l e d semantic analysis is need6d to determine \"possible simultaneity of t w o events. The p r i n c i p l e can be s t a t e d as f o l l o w s : ( 6 0 ) possible Simultaneitv Constraint: C a l l u~ allEncyclopedic Rules in t h e intersection of the MVBms of. C1 and C2. Unless tl t2 is derivable ( s t r i c r t l y or heuristically) the Temporal Sequence S t r a t e g y is inapplicable There a r e some cases, however, where two events are nocmally ihterpreted as sequenced, even though they could, stric-tly speaking, be sirnultaneous~ For instance, it is theoretically possible to l i g h t a cigarette while g&tting up, and one certainly would n o t want to add a r u l e to t h e encyclopedia stating t h a t two such events are unlikely to be' sinnaltaneo~s~ Nonetheless, (61) is normally interpreted a8 sequences I (61) John got up and lit a cigarette.", "type_str": "figure", "num": null }, "TABREF0": { "text": "", "html": null, "type_str": "table", "content": "
Preliminary SR f -$~~t e r r n e d i a t e w F i n a l
", "num": null } } } }