{
    "paper_id": "J78-3040",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T03:04:14.239817Z"
    },
    "title": "The 'Representation o f D a r t p b l e lnformation i n Memory: When What M i ght Have Been L e f t Unsaid I s Said",
    "authors": [
        {
            "first": "Rand",
            "middle": [
                "J"
            ],
            "last": "Spira",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Richard",
            "middle": [
                "J"
            ],
            "last": "Loseph Esposi",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "",
            "middle": [],
            "last": "Vondruska",
            "suffix": "",
            "affiliation": {},
            "email": ""
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "",
    "pdf_parse": {
        "paper_id": "J78-3040",
        "_pdf_hash": "",
        "abstract": [],
        "body_text": [
            {
                "text": ". However,a p o i n t o f t e n overlooked i s t h a t these same know7edge structures , w i t h the i r information about the world's order1 iness', may allow f o r more e f f i c i e n t processing and memorial representation o f e x p l i c i t information i n d i scourse, i n a d d i t i o n t o t h e i r r o l e i n d e r i v i n g i m p l i c i t information. This paper w i l l be concerned w i t h the psychological pro5ess i n g o f (imperfectly) p r e d i c t a b l e o r d e r i v a b l e informat i o n t h a t i s nevertheless e x p l i c i t i n discourse.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "Despi t e t h e f a c t t h a t most research on i n f e r e n t i a l Processeq i n comprehension has been concerned w i t h generation o f i m p l i c i t informat i o n , much i n f e r e n t i a l l y r e l a t e d information i s embod-ied e x p l i c i t l y i n cftscourse. W e are r e f e r r i n g here p r i m a r i l y t o pragmatic inferences, i.e., i m p l i c a t i o n s t h a t are u s u a l l y but n o t necessar i 1 y true. Language i s infrequent 1 y characterized by absolute redundancy; semdn t i c content i s r a r e l y \"repeated ,I1 except f o r speci at purposes such as emphasis. However, pragmatic inferences a r e o n l y imperfectly p r e d i c t a b l e . I f you read t h a t a karate champion h i t a block, uncertainty i s reduced by a l s o reading t h a t the block broke, d e s p i t e the f a c t t h a t t h a t outcome i s u s u a l l y t o be expected. S i m i l a r l y , i t would nat be considered unusual when r e l a t i n g the events a t a b i r t h d a y p a r t y t o mention t h a t there was a 'take wi t h candles blown out by the celebrant. Many things go i n stereotyped ways but r e q u i r e e x p l i c i t mention because thE stereotype does not describe a l l p o s s i b l e cases.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "P r e d i c t a b l e Information i n Discourse",
                "sec_num": null
            },
            {
                "text": "Throughout t h i s paper, \"predictable1' is used as a shorthand fol-\"imperfectly p r e d i c t a b l e , o r characterized by s i g n i f i c a n t l y l e s s than perf e c t uncertainty.\"",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "P r e d i c t a b l e Information i n Discourse",
                "sec_num": null
            },
            {
                "text": "processed? As was mentioned above, a t t e n t i o n has been p r i m a r i l y devoted t o the processing o f i m p l i c i t p r e d i c t a b l e information, leaving l j t t l e guidance on b e present issue. However, i n a v a r i e t y o f t h e o r e t i z a l o r i e n t a t i o n s , there i s a common i m p l i c a t i o n about howrpredictable information would be d e a l t w i t h : simply out, e x p l i c i t information, whether p r e d i c t a b l e o r not, receives s u f f i c i e n t processing t o be encoded i n long-term memory. For example, Kintsch (1974) assumes \" t h a t subjects process and s t o r e [an inference] whether o r not i t i s presented e x p l i c i t l y t 1 (p. 154). I t i s d i f f i c u l t to-imagine discourse representation t h e o r i s t s , who argue for'the e x p l i c i t representation i n memory o f imp1 i c i t inferences ( e . g . , Frederi Rsen, 1975 , Meyer, 1974 , arguing t h a t e x p l i c i t inferences aren o t represented. I n schema t h e o r i e s (e.g., Rumel h a r t f Ortony, 1977) , e x p l i c i t discourse i n f o r m a t i o n i s used t o bind schema variables, again suggesting t h a t p r e d i c t a b l e information would receive e x p l i c i t m i n t a l representation.",
                "cite_spans": [
                    {
                        "start": 568,
                        "end": 582,
                        "text": "Kintsch (1974)",
                        "ref_id": null
                    },
                    {
                        "start": 897,
                        "end": 915,
                        "text": "Frederi Rsen, 1975",
                        "ref_id": null
                    },
                    {
                        "start": 916,
                        "end": 929,
                        "text": ", Meyer, 1974",
                        "ref_id": null
                    },
                    {
                        "start": 1047,
                        "end": 1060,
                        "text": "Ortony, 1977)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "How i s e x p l i c i t but predictable information",
                "sec_num": null
            },
            {
                "text": "I f anything, one would expect e x i s t i n g t h 2 o r i e s t o p r e d i c t t h a t e x p l i c i t -inferences would r e c e i v e a s t;onger memor i a 1 represen t a t i on than unp r e d i c t a b l e information, given t h e i r g r e a t e r contextual support. For example, i n t h e i r a s s o c i a t i v e network model, HAM, Anderson and Bower (1973) argued t h a t the g r e a t e r the number o f interconnections between information, the g r e a t e r the l i k d l i h o o d t h a t information w i t h i n the interconnected network would be r e c a l l ed, This view w i 1 1 be r e f e r r e d t o as the \"storage of expl i c i t inferences\" (SEl) hypothesis.",
                "cite_spans": [
                    {
                        "start": 342,
                        "end": 367,
                        "text": "Anderson and Bower (1973)",
                        "ref_id": "BIBREF0"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "How i s e x p l i c i t but predictable information",
                "sec_num": null
            },
            {
                "text": "An a l t e r n a t i v e hypothesis i s t h a t p r e d i c t a b l e information, however c e n t r a l t o a discourse, i s taken f o r granted, processed on!y superf i c i a l l y and receives an attenuated c o g n i t i v e represen t a t ion o r no-endur inq representat i o n a t a1 I . I f needed subsequent t y , i f can be der i v e d . This iview w i l l be r e f e r r e d t o as the \" s u p e r f i c i a l processing o f e x p l i c i t inferencest' (SPE I ) hypothesis. Processing expl i c i t inferences i n such a manner has the advantage o f a c o g n i t i v e economy o f represen t a t i o n (besides a f i ke 1 y reduct i o n i n processing time),",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "How i s e x p l i c i t but predictable information",
                "sec_num": null
            },
            {
                "text": "Most i n f o r m a t i o n t h a t i s a c q u i r e d w i l l never be used again. I t would then seem t o be more e f f i c i e n t t o devote e x t r a processing e f f o r t t o the occasions when the i n f o r m a t i o n i s needed ( i . e . , by d e r i v i n g i t when remembering) r a t h e r than e x e r t i n g e f f o r t toward s t a b l e encoding a t the time o f comprehension.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "How i s e x p l i c i t but predictable information",
                "sec_num": null
            },
            {
                "text": "There are considerable problems i n designing an e m p i r i c a l t e s t o f the hypothesis t h a t e x p l i c i t pragmatic inference5 i n discourse a r e n o t represented i n long-term memory.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experiments on the Representation o f E x p l i q i t Inferences",
                "sec_num": null
            },
            {
                "text": "I f one merely t e s t s memory f o r the inference, f e i l u r e t o remember couTd be a t t r i b u t e d t o n o t s t o r i n g the informat i o n o r t o s t o r i n g and then f o r g e t t i n g i t , i f the inference i s rememberea', i t could be because i t was s t o r e d and then r e t r i e v e d , o r j t may have been generated a t the time o f t e s t w i t h o u t having been stored. C was e i t h e r presented p r i o r t o A and B (C-Before) a f t e r A and B, (C-After), o r n o t a t a l l (NO-C). When C was not included i n the s t o r y , i f S P E l i s c o r r e c t , the B element should be taken for granted, processed o n l y superf i-ci-1 l y , and n o t s t a b l y represented. Two o b j e c t ions t o t h i s argument can be made. The i n f o r m a t i o n might be stored, but remembering C might lead t o a d e c i s i o n t h a t t h e memory f o r B must be mistaken ( a k i n d of output i n t e r f e r e n c e ) .",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experiments on the Representation o f E x p l i q i t Inferences",
                "sec_num": null
            },
            {
                "text": "However, C i s present whether i t occurs b e f o r e o r a f t e r A and 8, so such an e x p l a n a t i o n would not accotrnt if a s t o r y takes place i n a restauraQt, and a1 1 the r e s t a u r a n t -r e l a t e d informattion i s t y p i c a l then t h a t aspect o f the s t o r y can be s t o r e d w i t h the a b s t r a c t schema node \" t y p i c a l r e s t a u r a n t a c t i v i t i e s . ' ' However, when the pro- I n general, i t seems t h a t p l a c i n g Lnformat i o n i n a s e n t e n c e -i n i t i a l subordinate clause: lowers the-superficial processing c r i t e r i o n . 3and 4 preceded by a l i n g u i s t i c s i g n a l . However, i f t h e information c o n t a i n s s a l i e n t i m p l a u s i b l e aspects o r something c l e a r l y i r r e l e v a n t , a s i g n a l 1 i n g phrase such as \"as you would expect\" might r e s u l t i n mare a t t e n t i o n being devoted processes has been provided by Spiro (19771, who found a pervasivk tendency for subjects t o produce pred l c table mean ing6chang ing d i-s t o r t ions and importations i n t e x t r e c a l l under c e r t a i n conditions. I n general, when subsequently encountered information contradicted cont lnuatlon expectations derived from a target story, the s t o r y f requentiy was reconstructed i n such a way as t o reconci 1 e o r cohere w i t h the continuation information, This process o f i n f e r r i n g the past based on the present was termed accommodative reconstruction. A f t e r a long r e t e n t i o n interval, subjects tended t o be more confident that t h e i r accommodative r e c a l l errors had a c t u a l l y beenincluded i n the story than they were confident about the accurate aspects o f t h e i r r e c a l l . Why should such gross errors occur and then be assigned such high confidence? Part o f the answer surely involves t h e i r function i n prbducing coherence. S t i l l , l t i s somewhat s u r p r i s i n g that subjects should be so sure they read information that bore not even a distant i n f e r e n t i a l r e l a t i o nship t o hat they actually d i d read, Spiro suggested that the basis f o r such an e f f e c t may be i n the way information i s treated a t the time of comprehension; namely, i t i s s u p e r f i c i a l l y processed and n o t stored i n longterm memory. Then, when remembering, individual s should know ( a t least t a c i t l y ) that considerable amounts o f predictable o r derivable informat ion they have encountered w i l l not be available i n memory, I n t h a t case, r e c a l l would t y p i c a l l y involve d e r i v i n g a l o t o f missing information. Accordingly, i t would not be surprising t h a t subjects faced w i t h njemories that lack coherence would assume t h a t missing reconciling information was presented but only superficial 1 y processed a t comprehension, The information could then be derived a t r e c a l f with high.confidence. Hence the capacity f o r restructuring the past based on the present.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experiments on the Representation o f E x p l i q i t Inferences",
                "sec_num": null
            },
            {
                "text": "individual Differences A f i n a l caveat should be o f f e r e d regarding the s u p e r f i c t a l processing e f f e c t , but also appl icable t o al.1 research on xhema-based processes i n comprehension and memory. The assumpt i o n i s usually made that there a r e no q u a l i t at i v e differences between i n d i ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Consider c o n t i nuat ions",
                "sec_num": null
            }
        ],
        "back_matter": [],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Human essociat i v e m e m 9",
                "authors": [
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Anderson",
                        "suffix": ""
                    },
                    {
                        "first": "R",
                        "middle": [],
                        "last": "Bower",
                        "suffix": ""
                    },
                    {
                        "first": "G",
                        "middle": [
                            "H"
                        ],
                        "last": "",
                        "suffix": ""
                    }
                ],
                "year": 1973,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Anderson, J , R., 6 Bower, G. H. Human essocia- t i v e m e m 9 . New Yorkb Wiley, 1973, -",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "Schemata as scaffolding for information i n t e x t",
                "authors": [
                    {
                        "first": "R",
                        "middle": [
                            "C"
                        ],
                        "last": "Anderson",
                        "suffix": ""
                    },
                    {
                        "first": "R",
                        "middle": [
                            "J"
                        ],
                        "last": "Spfro",
                        "suffix": ""
                    },
                    {
                        "first": "M",
                        "middle": [
                            "C"
                        ],
                        "last": "Anderson",
                        "suffix": ""
                    }
                ],
                "year": 1978,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Anderson, R. C., Spfro, R. J., & Anderson, M. C. Schemata as scaffolding for information i n t e x t . American Educational Research Journal, 1978, i h press,",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "Cogn i t ion and the symbol i c pr6cksses",
                "authors": [],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Cogn i t ion and the symbol i c pr6cksses.",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "laquage processins",
                "authors": [
                    {
                        "first": "",
                        "middle": [],
                        "last": "Erlbaum",
                        "suffix": ""
                    }
                ],
                "year": 1975,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Erlbaum, 1975. laquage processins. Cambridge, Mass. B o l t Beranek & Newman Inc., 1975.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Comprehens~on and the given-new contract",
                "authors": [
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Clark",
                        "suffix": ""
                    },
                    {
                        "first": "H",
                        "middle": [],
                        "last": "Haviland",
                        "suffix": ""
                    },
                    {
                        "first": "S",
                        "middle": [],
                        "last": "",
                        "suffix": ""
                    },
                    {
                        "first": "E",
                        "middle": [],
                        "last": "",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Clark, H, H., & Haviland, S, E , Comprehens~on and the given-new contract. I n R. 'reedle (~d .) , Discourse processing.",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "text": "I t i s now widely accepted t h a t natural language comprehension i s a c o n s t r u c t i v e process. Information in m c o u r s e i n t e r a c t s w i t h a v a r i e t y of impinging Eontextual f a c t o r s (inclwdlng, most promi nentl y, the comprehender ' s pre-exi s t i ng , knowledge) i n an ? c t lve, c r e a t i v e process t h a t r e s u l t s i n understandings not d e r l v a b l e by any s o l e l y l i n g u i s t i c o r l o g i c a l a n a l y s i s ( c . f . , Bransford & Mecarrel 1, 1975; Spi ro, 1977, i n preqs) . Acceptance o f the c o n s t r u c t fve v i~w o f comprehension e n t a i 1 s a concomi tan? del imi t a t ion o f the range o f possible theories o f mental representat ion. Knowledge s t r u c tor es must possess sbme c a p a b i l i t y f o r detecting t h e ptmgmatic, as we1 1 as l o g i c a l ,: ihpl fcat ions o f the incomplete data contained i n discourse ( c . f . ,",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF1": {
                "text": "I n o t h e r words, knowledge s t r u c t u r e s must contain considerable jnformaticin about the way the w o r l d u s u a l l y works. This c h a r a c t e r i s t i c o f representation i s useful and e f f i c f e n t because natural and s o c i a l contexts do produce s u f f i c i e n t c o n s t r a i n t s on w o r l d l y events and ideas as t o make them, t o a l i m i t e d extent, o t d e r l y a~d predictable.",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF2": {
                "text": "S p i r o and Esposi t o (1977) developed a paradigm n o t s u b j k c t t o the a m b i g u i t~e s o f i n t e r p r et a t i o n o f the more simple design discussed above The p r i m a r y manipulation g f i n t e r e s t involved subsequentlyrvi t i s t i n g the f o r c e o f an e a r l i e r e x p l ; c i t inference. i f t h e inference i s n o t stored, c e r t a i n p r e d i c t a b l e e r r o r s i n r e c a l l i n g i t should be made. I n the f i r s t experiment, s u b j e c t s were presented s t o r i e s 'which contained informat i o n A, B, and C such t h a t B was s t r o n g l y i m p l i e d by A except i n the presence o f C. For example, the A, B, and C elements i n one s t o r y (about a demons t r a t i o n by a k a r a t e champion) could be para-phrased as f o l l o w s~ A: The k a r a t e champion h i t t h e block. B: The b l o c k broke. C: He had had a f i g h t w i t h h i s w i f e e a r l i e r . I t was i m p a i r i n g h i s concentration.",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF3": {
                "text": "I t would be d e r i v a b l e i f needed. However, i f C i s presented.after A and 0 , memory f o r B should be-impaired s i n c e B was n o t s t o r e d and C w i l l b l o c k i t s d e r i v a t i o n from A a t the time o f t e s t . On the o t h e r hand, i f C occurs i n the t e x t p r i o r t o A and 0, then B i s not s t r o n g l y i m p l i e d by-A. 8 cannot be taken f o r granted w i t h the assumption t h a t i t can be generated l a t e r i f needed. Here B should be s t a b l y represented and memory f o r B should n o t be impaired. However, i f SEl i s c o r r e c t , memory f o r B should not be a f f e c t e d by whether C i s b e f o r e o r a f t e r A and B, s i n c e B i s ,tored whether i t 1s imp1 i e d by A (C-After) o r n o t implied by A (c-Before).",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF4": {
                "text": "f o r d i f f e r e n t i a l e f f e c t s of C-placement. The o t h e r p o s s i b i l i t y i s t h a t 8 i s represented i n C-After, b u t the r e p r e s e n t a t i o n i s a l t e r e d OJ-c o r r e c t e d when t h e C informat ion-i s encountered. T h i s p o s s i b i l i t y was i n v e s t i g a t e d i n t h e second experiment. I n the f i r s t experiment, the f o l l o w i n g ,pre'dictions of the SPEl hypothesis-were tested. More e r r o r s i n response tg questions about the presented p r e d i c t a b l e i n f o r m a t i o n (0) should be made i n the C-After than i n the C-Before c o n d i t i o n s . E r r o r s can be erroneous judgments t h a t n o t h i n g about t h e i m p l i e d i n f o r m a t i o n was presented, c a l l ed 8-Ment i o n e r r o r s (e.g. , the s t a r y d i d n o t mention whether the b l o c k was broken), o r , when the subject be1 ieves t h a t something about B was mentioned, remembering i n c o r r e c t l y what was s p e c i f f c a $ l y s a i d i n the d i r e~t i o n o f conforming w i t h the C information, c a l l e d B -t n e o r r e c t e r r o r s (e.g., i t s a i d i n t h e s t o r y t h a t the b l o c k d i d n o t break when he h i t i t ) . Confidence i n e r r o r s o r the l a t t e r k i n d were a l s o analyzed. I f s u b j e c t s are as c o n f i d e n t about these e r r o r s as they a r e about t h e i r accurate responses, I t would be even more d i f f i c u l t t o m a i n t a i n the hypotheqis t h a t the e x p l i c i t i nferences were represented. I n the No-C c o n d i t i o n , B-Mention e r r o r s may occur s i n c e B would n o t be represented p c c o r d i n g t o the S P t l hypothesis. The more important p r e d i c t i o n regarding the No-C c o n d i t i o n i s t h a t B -I n c o r r e c t e r r o r s should n o t occur more o f t e n than i n the C-Before c o n d i t i o n . Otherwise, the d i f f e r e n c e s between C-Before and C-After might be a t t r i b u t a b l e t o heightened accuracy due t o g r e a t e r s a l i e n c e o f t h e i m p l i e d informat i o n i n the former c o n d i t i o n r a t h e r tttdn g r e a t e r inaccuracy due t o a f?ai l u r e t o s t o r e the i m p l i e d i n f o r m a t i o n i n t h e l a t t e r c o n d i t i o n .9College s u b j e c t a read e i g h t t a r g e t v i g n e t t e s each c o n t a i n i n g A and B information, and C i nf o r m a t i o n included o r not and placed as a f u n c t i o n o f which o f t h e three c o n d i t i o n s subjects were randomly assigned to. C informat i o n was always on a separate page from ttte A and 1 i n f o r m a t i o n , and s u b j e c t s were i n s t r u c t e d t o n o t l o o k back a f t e r reading a page. A f t e r reading a l l the v i g n e t t e s , t h e subjects-were t e s t e d f o r t h e i r memory f o r the v i g n e t t e s . O f p a r t i c u l a r i n t e r e s t Mere the two types o f questions, mentioned above, concerning t h e B i n f o r m a t i o n (remember, 0 was always e x p l i c i t i n the s t o r l e s ) . -The r e s u l t s supported the hypothesis t h a t pragmatic inferences presented i n t e x t a r e s u p e r f i c i a l l y processed and do n o t r e c e i v e a s t a b l e andenduring representation i n memory. i n the C-After condition, subjects tended e i t h e r t o r e p b r t t h a t the inference was n o t presented I n the text Or t h a t the opposite o f t h e inference was presented. Furthermore, confidence i n these e r r o r s was as h i g h as confidence I n c o r r e c t memo r i e s . I t i s d i f f i c u l t t o r e t a i n the n o t i o n t h a t inferentes a r e deeply processed and s t a b l y encoded hen the C-Af t e r man i pu 1 a t i on can produce e r r o r s 1 i k e remembering the b l o c k was n o t broken when the karate champion h i t 1 t. The r e s u l t s cannot Be a t t r i b u t e d t o i n t e r f e r e n c e produced by the i n f e r e n c e -v i t i a t i n g C Information a t output, since the B B e f o r e subjects would a l s o be subject to such Interference. Neither can the r e s u l t s be r t t r i b u t e d t o d i f f e r e n t i a l a v a i l a b i l i t y o f C a t 3u tput, perhaps due t o p r imacy/recency e f f e c t s r e l a t e d t o t h e p o s i t i o n o f F' in $he t e x t , since the information was almost a l w a y~ reca1,led. Also, unimportance o f the B i n f o r m t i o n i s not a v i a b l e a l t e r n a t i v e since B tended t o be c e n t r a l t o the s t o r y (e.g., i n a s t o r y about a karate champion's performance, iflformation about h i s success i n the' demon st ratio^ i s c e~t a i n l y important). One a l t e r n a t i v e i n t e r p r e t a t i o n t h a t remains i s that suolecrs do deeply process and s t a b l y encode the presented in-ference, b u t \"correct\" t h e i r representation when the i n f e r e n c e -v i t i a t i n g information i s presented. I f subjects are s t o f i n g B and then changing o r c o r r e c t i n g i t a t the time C i s preserrted, e r r o r s oh k should occur I n t h e C-After c o n d i t i o n no matter how soon the t e s t i s admj n i s tered a f t e r read i ng . However, i f the SPE l hyptrthesfs i s correct, when deiay i n t e r v a l s are b r i e f enough some surface memory f o r the superf i c l a l i y processed 0 information may remain,. reducing the numb@ of B errors, Accordingly, i n the second experiment subjects were tested qi'ther im&d i ately. a f t e r reading each s t o r y ('Interspersed Quest ions condition)-, as i n the f i r s t experiment, a f t e r the C n t i r e set o f s t o r i e s had been read tQues t ions-Af t e r cond i t ion) , hga i n , the C-Before and C-After manipulations were emp 1 oyed . The r e s u l t s o f the second experiment r e p l icated those o f the f i r s t one i n t h e Questions-A6ter cbndi tion. Furthermore, the C-af t e r e f f e c t was l a r g e l y absent i n the lnterspersed Querkions condition, demonstrating t h a t the , e f f e c t i s n o t due go s t o r i n g and then changing the representatloq of the B information ( t h e e x p l i c i t inference). Re1 a ted l ssues The discussion of i m p l i c a t i o n s o f t h e superf i c i a l processing e f f e c t w i 11 a t times &e 1 i m i red t o reading r a t h e r than 1 i s t w i n g . Most o f the following i s o f a specuSlative nature. Representation and Under 1 y i ng Mechan i sms Assuming some compatible representation system, what character i zes the processes t h a t produce the s u p e r f i c i a l processing e f f e c t ? At t h i s time, o n l y speculations about a1 t e r n a t i v e p o s s i b i l i t i e s can be offered. There ar< three p o t e n t i a l l y b e n e f i c i a l aspects o f s u p e r f i c i a l processing o f e x p l i c i t p r e d i c t a b l e information: c o g n i t i v e e m m y (the information need not be s p e c i f i c a l 1 y stored i n # long-term memory) , speed o f processing (you can prooess and understand such Information r a p i d l y ) , and a u t o m a t i c i t y of processing ( l e s s consCioqs e f f o r t and d i n g memory .space a r e 'required). Two simple, preliminary accounts o f the f i r s t f a c t o r , c o g n i t i v e economy, can be offered. The superf i c i at process i ng phenomenon appears most compatible w i t h a schema-theoretic mode o f representation. Perhaps v a r i a b l e bindings t h a t are d e f a u l t (or a t l e a s t h i g h p r o b a b i l i t y ) values are n o t e x p l i c i t l y i n s t a n t i a t e d when they a r e e x p l i c i t i n discourse ( b u t see the discussion o f Determinants o f Performance V a r i a b i l i t y below). However, one should not be o v e r l y persuadtd by t h e s i m p l i c i t y 04 such an account. Other types o f representation systems could a l s o account f o r the phenomenon. For example, a spreading a c t i v a t i o n mode? (e.g., C o l l i n s & l o f t u s , 1975) might p r e d i c t t h a t e x p l i c i t information i s n o t tagged i n memory when i t has been r e c e n t l y a c t i v a t e d w i t h some greater than c r i t e r i o n strength. T h i s issue w i l l receive f u r t h e r discussion i n the next section. Regarding speed of processing, several possibi 1 i t i e s may be o f f p m d : t h e information i s a c t u a l l y predicted, perhaps followed by a s e l e c t i v e scanning f o r p a r t i a l clues o f conf irmation (e.g., the word \"broke\" i n the karate champion example; perhaps such checks couM be made i n the v i s u a l periphery and, when posit i v e , r e s u l t i n saccades t h a t s k i p the predicted information), o r the expectation may be formed a f t e r beginning t o read the p r e d i c t a b l e informat i o n followed by skipping ahead to the next l i n g u i s t i c u n i t (\"Oh. They're t a l k i n g about t H i s now. We1 1 there's no doubt how i t W I 11 t u r n out. I can pass t h i s by.\"); o r temporary binding o f a schema v a r i a b l e ( e s s e n t i a l l y a v e r i f i c a t i o n o f f i t ) may be more r a p i d than more durable i n s t a n t i a t i o n , o r l e s s metacognitive a c t i v i t y (pondering, studying, rehearsing, etc.) may be devoted t o p r e d i c t a b l e informat ion, given i t s d e r i v a b i l i t y ( t h i s a l s o r e l a t e s t o automat i c i t y , obviously). Regarding autom a t i c i g , i t seems l i k e l y t h a t the amount o f conscious processing required would be nega-t~v e l y c o r r e l a t e d w i t h the goodness o f f i t t o p r i o r knowledge. Thus conscious attempts t o make sense o f p r e d i c t a b l e i n f o r m a t i o n would be expected less often. Also, r e l a t e d t o the suggestions above regarding expectations and r a p l d i t y o f processing, the o p e r a t i o n o f some p r e a t t e n t i v e process ( I n the sense o f Neisser, 1967) is, a p o s s i b i l i t y . N a t u r a l l y , i t may be the case t h a t a11 o f these f a c t o r s a r e cont r i b u t ing. However, some o f ttfe f a c t o r s may be mutually exclusive. For example, i f d e f a u l t values are processed automatically, an expectat i o n and c o n f i r m a t i o n process may be redlrndant. Determinants o f Performance V a r i a b i l i t y Occurrence o f s u p e r f i c i a l processing and f a i l u r e t o s t o r e information probably depends on more than p r e d i c t a b i l i t y o r d e r i v a b i l i t y considered I n i s o l a t i o n . F o r o n e t h i n g , the YE: d e s i v a b i l i t y o f o t h e r information i n the d i scourse w i l l have an e f f e c f . The g r e a t e r t h e p r o p o r t i o n o f f i t t o one's schemata f o r the d i scourseas a whole, the more l i k e l y i , t i s t h a t conforming information w i l l be l e f t . t o be der i v e d .",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF5": {
                "text": "p o r t i o n o f f i t i s poor, i,e., some a t y p i c a l events occur, even t y p i c a l , p r e d i c t a b l e events may have t o be stored. Occurrence o f superf i c i a 1 processing i s a l s o l i k e l y t o be a f f e c t e d by the e x t e n t t o which the system i s taxed. When the system i s overloaded, as when t h e r e i s a l a r g e amount ofinformation t o be acquired o r the time t o acquire the h f o r m a t i o n i s l i m i t e d , more superf i c i a l processing and l e a v i n g o f i n f o r m a t i o n t o be derived probably goes on. Perhaps the system has f l e x i b l e c r i t e r i a f o r d e r i v a b i l i t y , reducing c r i t e r i a under overload c o n d i t i o n s and increasing them when processing load i s l i g h t (and when demands f o r r e c a l l accuracy a r e h i g h or' when subsequent a v a i l a b i l i t y of the informat i o n i s l i m i t e d ) . B r i e f l y digressing, Shere may be a temptation t o confuse s u p e r f i c i a l proc e s s i n g -o f d e r i v a b l e information w t t h skimming. However, skimming i s a s e l e c t i v e seeking and then deep process i_ng o f s i t u a t ional ly important i n f o r m a t i o n (see FRUMP, i n Schank t Abel son, 1975) whereas s u p e r f i c i a l processing involves s e l e c t i v e l ynot processing deeply i n f o r m a t i o n perceived as derivable, however important i t might be. I n o t h e r words, the same i n f o r m a t i o n t h a t might receive more a t t e n t i o n w h i l e skimming may receive less a t t e n t i o n i n normal s i t u a t i o n s i f the i n f o r m a t i o n i s derivable. T h i s w i l l happen t o the e x t e n t t h a t skimming r e s u l t s i n shallow processing o f e a r l i e r i n f o r m a t i o n t h a t i s the b a s i s f o r t h e d e r i v a b i L l t y o f t h e l a t e r information. Besides context-based v a r i a b i l i t y i n d e r i v a b i l i t y c r i t e r i a , research i n t h e psychology o f p r e d i c t i o n i n d i c a t e s the p o t e n t i a l o p e r a t i o n o f a genera1 b i a s i n determining the c r i t e r i o n f o r d e r i v a b i l i t y and s u p e r f i c i a l processing. For example, F i s c h o f f (1 975, 1977) ha3 found t h a t when people a r e t o l d t h a t some event has occurred, they increase t h e i r s u b j e c t i v e probabi 1 i t y e s t irnate o f the 1 i k e l ihood t h a t the event was going t o occur, S i m i l a r l y , estimat i o n af how muEh was known b e f o r e being-given a c o r r e c t answer increases when the answer i s provided. I n the case o f s u p e r f i c i a l processing o f information i n discourse, i t i s p o s s i b l e t h a t the der i v a b i 1 i t y o f -information i s overestimated a f t e r i t i s e x p l i c i t l y encountered. I t seems t o be a fairly c o m m~n ' e x~e r i e n c e , f o r example, t o n o t w r i t e down an idea t h a t you a r e sure w i l l be d e r i v a b l e . agaia l a t e r , o n l y t o f i n d subsequent d e r i v a t i o n impossible. What i s being suggested here i s a source o f f o r g e t t i h g n o t u s u a l l y discussed i n memory theories: superf i c i a l brocessing o f i n f o r m a t i o n whose d e r i va b i l i t y has &en overestimated. The Form ~f Exprespion of Derivable l nformat ion, Semantic content, p r i o r knowledge, a8d task contexts a r e not t h e o n l y determinants o f perceived d e r i v s b i l i t y . The I i n g Q f s t i c fcmn i n which information i s expressed w i l l sometimes p r o v i d e s i g n a l s of what information i s already known 'br can be taken f o r granted, as when i n f o r m a t i o n i s expressed near t h e beginning o f a sentence (c.f., Ctark 6 Haviland, 1977, on t h e g i ven-new s t Pa tegy) , Tak 1 ng an example from Morgan and Green ( I n press) , compare sentences (1) and (2). (1) The government has n o t y e t acknowledged t h a t d i s t i l l e d water causes cancer. (2) That dist.i 1 l e d water causes cancer has n o t y e t been acknowledged by the government . I n (2) there i s a stronger imp1 ied presumption o f the t r u t h o f t h e p r o p o s i t i o n regarding d i st i l l e d water and cancer than there 7s i n (1).",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF6": {
                "text": "o f \"The k a r a t e champion h i t t h e block.\" (3) The b l o c k broke, and then he bowed. (4) R f t e r the b l o c k broke, he bowed. The b l o c k ' s breaking would appear t o be more taken f o r granted i n (4) than i n ( 3 ) . L i n g u i s t i c s i g n a l s o f p r e d i c t a b i l i f y o r d e r l G a b i l i t y need n o t be i m p l i c i t . Consider c o n t i nuat ions (5) , (6) , and (7) o f t h e same sentence as above. (5) Obviously, the b l o c k broke. (6) As you would expect,the b l o c k broke. (7) Natural l y , t h e block broke. Words 1 i k e \" c l e a r 1 y\" and phrases 1 i ke \"of course11 a r e e x p l i c i t 1 i n g u i s t i c s i g n a l s t h a t i n f o r m a t i o n t o f o l l o w r i s p r e d i m M e and can be s u p e r f i c i a l t y processed. However, one would expect t h a t such s i g n a l s could have t h e i r e f f e c t o n l y f o r i n f o r m a t i o n w i t h i n an acceptdble range o f p l a u s i b i l i t y . That i s , a p l a u s i b l e b u t n o t p r e d i c t a b l e c o n t i n u a t i o n may be more l i k e l y t o be taken (erroneously) as p r e d i c t a b l e when",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF8": {
                "text": "t o t h e c o n t i n u a t~o n ~n f o r n l b t i o n .I m p l i c a t i o n s f o r t h e Nature o f Discourse MemoryTo the e x t e n t t h a t discourse i s superf i c i a l l y prqcessed, mgmory must be r e c o n s t r u c t i v e r a t h e r than reproduct i ve. Rather than ret r i e v i n g traces o r i n s t a n t i a ti o n s o f p a s t experienge, the p a s t must be i n f e r r e d o r derived. Just as a p a l e o n t o l o g i s t r e c o n s t r u c t s a dinosaur From bone fragments, the p a s t must be recons t r u c t e d from t h e incomplete data e x p l i c i t l y stored. Evidence f o r such r e c o n s t r u c t i v e",
                "num": null,
                "uris": null,
                "type_str": "figure"
            },
            "FIGREF9": {
                "text": "v i d u a l s i n the manner i n which discourse i s processed. Howevet, Spi r o and h i s cot leagues have recent l y found that r e l i a b l e s t y l e differences can be predicted i n c h i l d r e n (~p i r o & Smith, 1978) and i n c o l lege students (~p i r a 6 T i r r e , i n preearation). Some individuals appear t o be more d i $course bound, tending toward over-re1 iance-on bottom-up processes. Others are more p r i o r know1 edge bound, tending toward over-re1 ance on top-down processes. For the a d u l t bottomup readers, p r i o r knowledge obviously must be used t o a c e r t a i n extent i n comprehension. However, where use o f p r i o r knowledge i s more optional, e.g., i n providing a scaffolding f o r remembexing information (Anderson, S p i~o , & Anderson, 1978), the bottom-up readers capi ta 1 i ze less. Whether the l a t t e r type o f individual w i l l evince less knowledge-based s u p e r f i c i a l processing (again an optional use o f p r i o r knowledge) I s a question c u r r e n t l y under investigat ion.",
                "num": null,
                "uris": null,
                "type_str": "figure"
            }
        }
    }
}