Preserved Learning and Retention of Pattern-Analyzing Skill in Amnesia: Dissociation of Knowing How and Knowing that ® Neal J. Cohen; Larry R. Squire Science, New Series, Vol. 210, No. 4466 (Oct. 10, 1980), 207-210. Stable URL: http://links.jstor.org/sici?sici=0036-8075%2819801010%293%3A210%3A4466%3C207%3APLAROP%3E2.0.CO%3B2-K Science is currently published by American Association for the Advancement of Science. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/aaas.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org. http://www.j stor.org/ Mon Aug 23 12:03:28 2004 of the anesthetic block (6, 7). These currents were also not influenced by internal anesthetics (Fig. 2C). This is in contrast to the effects of external anesthetics, which cause a marked alteration of the steady-state current-voltage relation for ACh-induced current (6, 7). To determine whether the anesthetics reached the internal membrane surface, the voltage-activated sodium current was examined in two myoballs. At a holding potential of - 80 mV, a 30-mseC pulse to - 40 mV elicited an inward sodium current. In the absence of anesthetic compounds, the sodium current was not affected by stimulating at a frequency of 1 Hz. However, 20 minutes after the addition of 1 mM QX-314 to the internal solution, the sodium current exhibited a use-dependent block, decreasing by about one-third in response to 30 pulses at 1 Hz. Extracellular 1 mM QX-314 was without effect on the current through the sodium channel. A use-dependent inhibition of sodium current by internal QX-314 was reported previously (9) . It is possible that an asymmetry of surface potentials could explain the asymmetrical effects of QX-314 and QX-222 on the ACh channel. If the extracellular negative surface potential is greater than the intracellular surface potential, it could concentrate the cationic anesthetic at the outer mouth of the ACh channel, thus increasing its effectiveness when applied externally. If the extracellular surface charge density is about 0.002 electron charge per square angstrom (10) , then the surface potential can be calculated by use of the Grahame equation (10,11). The calculated extracellular surface potential, - 41 mV, would increase the concentration of the monovalent cation from 0.1 mM in the bath to 0.5 mM at the outer membrane surface. This concentration is still less than the intracellular concentration, 1 mM. In spite of the fact that the intracellular concentration is greater, these anesthetics have no effect When applied intra-cellularly. It seems to us, therefore, that asymmetry of surface potential cannot explain this result. Extracellular anesthetics appear to block channels by hopping over an extracellular barrier and binding to an intra-channel site within the transmembrane electrical field. Apparently the blocker is unable to hop over an even larger intracellular barrier to reach the inside of the cell. If the blocker could be driven by voltage over an intracellular barrier, the block would be relieved at sufficiently negative potentials by a "punch-through" mechanism (12), which has not been observed for the ACh channel (6, 7). The large intracellular barrier may prevent access of internally applied anesthetics to the binding site. It is conceivable that the intracellular barrier is the selectivity filter of the channel (2). Regardless of the mechanism of action of anesthetics, it is clear from these experiments that the channel is not a symmetrical structure. Our data are in agreement with recent morphological evidence showing that the ACh receptor-channel complex has an asymmetrical structure within the lipid bilayer (13). Richard Horn* Malcolm S. Brodwick Cholinergic Division, Department of Physiology and Biophysics, University of Texas Medical Branch, Galveston 77550 W. Daryl Dickey Department of Human Genetics and Biochemistry, University of Texas Medical Branch References and Notes 1. T. M. Linder and D. M. J. Quastel, J. Physiol. (London) 281, 535 (1978); P. r. Adams, Biophys. J. 25, 70 (abstr.) (1979); P. H. Barry, P. W. Gage, d. F. Van Helden,/. Membr. Biol. 45, 245 (1979). 2. r. Horn and m. s. Brodwick, J. Gen. Physiol. 75, 297 (1980). Amnesia, a neurologic syndrome characterized by a deficit in the formation of new memories, can exist independently of other cognitive impairment. The defi- 10 r 8 - 3. The internal solution consisted of 85 mM sodium aspartate, 155 mM sucrose, 5 mM sodium-Hepes .3] or by the controls [f(10) < 1, P > .3]. For performance on repeated SCIENCE, VOL. 210 80 60 40 20 Nonrepeated word triads • N.A. o-o Controls (N=6) i'ii'_....._I I I i i 80 i 60 40 o 20 S 0 60 40 20 •-# Korsakoff (N=4) 0-0 Controls (N=6) Mill_I I I I I_i'll i/,.....i -1 \ I ECT (N=3) i I i I I_i I I I I_i ' ' ' i -.....' Repeated word triads °°o °oooo °oSS I i i I I_I I I 1 I_I I I I l.yl I I I I Fig. 2. Acquisition of a mirror-reading skill during three daily sessions, and retention 3 months later. The ability to mirror read unique (nonrepeated) words was acquired at a normal rate by amnesic patients. The ability of amnesic patients to mirror read repeated words was inferior to the control rate because amnesic patients, unlike control subjects, could not remember the specific words that had been read. i i ii i_i i i i i_i i i i i ..i i i i i 1 2 3 94 1 2 3 94 Day 208 word triads, which depends in addition on memory for the specific words that were repeated from block to block, only the amnesic patients showed marked forgetting [/(6) = 2.78, P < .02]. The striking deficit in memory for specific items exhibited by the amnesic patients was also illustrated by their poor performance on a recognition memory test administered after day 3 of testing (19). For the nonrepeated words, recognition d' was amnesics, 0.36 ± 0.1; controls, 1.46 ± 0.2 [/ (18) = 3.07,P < .01]; for the repeated words, recognition d' was amnesics, 1.97 ± 0.4; controls, 3.79 ± 0.1 [/ (18) = 4.24, P < .001]. Upon being questioned, none of the amnesic patients reported that words had been repeated during the task, even though by the end of session 4 the set of repeated words had been presented 20 times. All of the control subjects reported spontaneously that words were frequently repeated. The finding of intact acquisition and retention of mirror-reading skill in amnesic patients adds to the set of learning skills known to be preserved in amnesia, and suggests that the class of preserved learning skills may be considerably broader than was previously thought. The traditional view has held that the role of motor information is crucial in determining which tasks amnesics can learn. This view is based on the premise that motor information enjoys a special neurologic status and is thereby spared in amnesia. However, amnesic patients have recently been found to be impaired in a test of short-term kinesthetic-motor memory (20), which indicates that motor performance is not uniformly spared in amnesia. Moreover, amnesic patients can learn skills that are apparently not perceptual-motor in nature, including the present pattern-analyzing skill. Thus, there have been reports of numerical rule learning (21) and eyeblink conditioning (22) in amnesic patients, and anecdotal observations (23) that amnesic patients are often good at learning testing procedures even when they fail on the tests (24). If, as Kolers suggested (11, 12), acquiring reading skills involves the learning of encoding operations or procedures rather than the remembering of specific results of these operations, amnesic patients seem to provide a clear example of this distinction. Amnesic patients can apparently learn the encoding rules or procedures for acquiring skills, but can remember little or nothing of the information that results from applying them. Whether a task can or cannot be learned in amnesia seems to depend on the na- Nonrepeated Repeated Word triads Fig. 3. Forgetting between sessions as measured by the difference in reading time between the last block of each session and the first block of the subsequent one. The interaction of the subject group by type of word was significant 18) = 6.08, P < .03)]. ture of the information and not on the extent of motor involvement demanded by the task. We propose that perceptual-motor and pattern-analyzing skills belong to a class of operations governed by rules or procedures; these operations have information-processing and memory characteristics different from those operations that depend on specific, declarative, data-based material. Although the distinction we have drawn between these classes of information may not permit all tasks to be sharply dichotomized, it should prove useful in predicting what is affected or spared in amnesia. This distinction between procedural or rule-based information and declarative or data-based information, which is reminiscent of the classical distinction between "knowing how" and "knowing that," has been the subject of considerable discussion in the literature of cognition and artificial intelligence (25). The experimental findings described here provide evidence that such a distinction is honored by the nervous system. Neal J. Cohen Department of Neurosciences, University of California, San Diego, La Jolla 92093 Larry R. Squire Veterans Administration Medical Center, San Diego, California 92161, and Department of Psychiatry, University of California, San Diego References and Notes 1. B. Milner, Clin. Neurosurg. 19, 421 (1972); G. Talland, Deranged Memory (Academic Press, New York, 1965); L. R. Squire, in Electroconvulsive Therapies: Biological Foundations and Clinical Application, W. B. Essman and P. Abrams, Eds. (Spectrum, Jamaica, N.Y., 1980). 2. W. B. Scoville and B. Milner,/. Neurol. Neurosurg. Psychiatry 20, 11 (1957). 3. B. Milner, S. Corkin, H.-L. Teuber, Neuropsy-chologia 6, 215 (1968). 4. S. Corkin, ibid., p. 255. 5. B. Milner, in Physiologie de I'Hippocampe (Centre National de la Recherche Scientifique, Paris, 1962). 6. D. N. Brooks and A. Baddeley, Neuropsycho-logia 14, 111 (1976). 7. N. Cohen and L. R. Squire, Soc. Neurosci. Abstr. 4, 256 (1978). 8. S. Corkin, Neuropsychologia 3, 339 (1965). 9. M. I. Posner, Science 152, 1712 (1966); J. Exp. Psychol. 75, 103 (1967);_and A. F. Kon- ick, Organ. Beh. Hum. Perform. 1, 71 (1966). 10. In a series of studies of the ability of normal subjects to read geometrically inverted text (11,12), Kolers found excellent acquisition and long-term retention of pattern-analyzing operations "that are directed at the surface lexical features of the text" and that can be distinguished from memory for the results of these operations, "the semantic or other grammatical content of text that is the subject of most contemporary studies" (11). 11. P. A. Kolers, Cognit. Psychol. 7, 289 (1975); in Levels of Processing in Human Memory, L. Cermak and F. I. M. Craik, Eds. (Erlbaum, Hillsdale, N.J., 1979). 12. _, Science 191, 1280 (1976). 13. H.-L. Teuber, B. Milner, H. G. Vaughan, Neuropsychologia 6, 267 (1968); L. R. Squire and P. C. Slater, ibid. 16, 313 (1978); L. R. Squire and R. Y. Moore, Ann. Neurol. 6, 503 (1979). 14. The patients had received a series of seven to ten bilateral treatments. B. Cronholm and C. Blomquist [Acta Psychiatr. Scand. 34, 18 (1959)] and B. Cronholm and J.-O. Ottosson [Br. J. Psychiatry 109, 251 (1963)] found anterograde amnesia 1 week after a series of two to seven and a series of three or four bilateral treatments, respectively. 15. Six normal control subjects were matched to N.A. for age (control mean, 38.8 years; N.A., 42 years), subtest scores on the Wechsler Adult Intelligence Scale (WAIS) (information; 21.7 versus 22; vocabulary: 60.0 versus 66), and educational background (14.8 years versus 13 years). Six alcoholic control subjects were matched to the Korsakoff patients for age (48.5 years versus 51.0 years), WAIS subtest scores (information: 17.2 versus 17.5; vocabulary: 50.7 versus 46.3), and educational background (10.7 years versus 12.2 years). 16. Their deficit was also observed in delayed recall of prose material. After a 15-minute delay, N.A. and the Korsakoff patients could remember none of the material learned 15 minutes earlier, and the ECT patients recalled an average of 2.1 segments. By contrast, the six control subjects recalled 6.8 segments and the alcoholic control subjects recalled 4.7 segments after the same delay. 17. N.A.'s reading time fell outside the 95 percent confidence interval for only 3 of the 20 blocks and never during the final 8 blocks. For Korsakoff patients and their alcoholic controls, there was no main effect of subject group [F(l, 18) = .02, P > .3] and no interaction of subject group with either block number [F(4, 32) < 1, P > .3] or test day [F(3, 24) < l,P> .3]. 18. For both the amnesic patients and the control subjects, we calculated for each block the ratio between reading time per repeated word triad and reading time per nonrepeated word triad. Compared with amnesic patients, control subjects benefited more rapidly from the repeated word triads [main effect of subject group: F (1, 18) = 18.17, P < .001; interaction of subject group by block number: F(4, 72) = 3.72, P < .01; interaction of subject group by test day: F (3 , 54) = 3.86, P < .03] and to a greater extent (lowest value of ratio attained: amnesics, .30; controls, .15; average value of ratio attained: amnesics, .48; controls, .31). 19. Each subject inspected a list of words which included the 15 repeated words, 45 nonrepeated words, and 60 distracter words and marked those words that could be remembered from the 3 days of testing. 20. L. Cermak and B. Uhly, Percept. Mot. Skills 40, 275 (1975). Cohen and Squire (6) studied delayed lever-movement reproduction in the absence of 10 OCTOBER 1980 209 visual cues. In a procedure that separated distance cues and end location cues, both amnesic patients and control subjects accurately reproduced movements in a no-delay condition, but the amnesic patients were impaired relative to controls after delays of 12 and 60 seconds. 21. F. Wood, thesis, Duke University (1974); B. T. Volpe, W. Hirst, M. S. Gazzaniga, Soc. Neurosa. Abst. 5, 1079 (1979). 22. L. Weiskrantz and E. K. Warrington, Neuropsychology 17, 187 (1979). 23. S. Corkin (4) referred to this phenomenon as "sparing of testing habits" after M. Glickstein and R. W. Sperry [J. Comp. Physiol. Psychol. S3, 322 (I960)]. 24. It is not yet clear how the performance of amnesic patients on numerical rule learning or classical conditioning compares with the performance of control subjects. Only when amnesic patients perform as well as controls is it possible to argue conclusively for preserved learning or memory in amnesia. Good, but not normal, performance by amnesic patients could occur because some methods of testing yield good performance in all subjects [L. R. Squire, Neuro-psychologia 18, 369 (1980)] and not because the Diurnal variations in pain sensitivity have been demonstrated repeatedly in rats (/, 2). These variations are attenuated or abolished by the opiate antago1 nist naloxone, suggesting that the phenomenon is mediated by an endogenous opioid system (3, 4). Strong support for this suggestion is provided by the finding that fluctuations in pain sensitivity are closely paralleled by alterations in levels of endogenous opioids in the rat brain (5). It has been assumed that this diurnal cyclicity in pain sensitivity reflects the circadian rhythm common to other pitui- aspect of memory under study is preserved in amnesia. Thus, reports of good performance by amnesic patients on incomplete figures [E. Warrington and L. Weiskrantz, Nature (London) 217, 972 (1968)] or partial information [ibid. 228, 628 (1970); Neuropsychology^ 12, 419 (1974)] do not necessarily mean that these tasks demonstrate preserved function. Amnesic patients often do rather well in recognition memory tasks compared with free recall tasks, but the advantage of recognition memory over free recall applies to control subjects as well [G. Talland, Deranged Memory (Academic Press, New York, 1965), p. 231], 25. T. Winograd, Understanding Natural Language (Academic Press, New York, 1972); in Representation and Understanding, D. Bobrow and A. Collins, Eds. (Academic Press, New York, 1975). 26. Supported by the Medical Research Service of the Veterans Administration, by National Institute of Mental Health grant MH24600, and by NIMH Mental Health Clinical Research Center grant 1P50 MH 30914. 3 March 1980; revised 11 June 1980 tary hormone concentrations (/). We now present evidence suggesting that diurnal fluctuations in pain sensitivity reflect patterns of food intake rather than a circadian rhythm. Researchers reporting diurnal rhythms in pain sensitivity and endogenous opioid concentrations used standard light-dark cycles in conjunction with unrestricted feeding (1, 2). However, since rats given this liberty feed predominately at night (6), the possibility exists that pain sensitivity cycles are entrained by food intake patterns rather than light- dark cycles. This suggestion is supported by the finding that food deprivation can induce analgesia, which is markedly attenuated by naloxone (4). Related to this point are several studies implicating endogenous opioid systems in the modulation of food intake. Central injections of opioid peptides increase food intake in the rat (7), while opiate antagonists decrease food intake (Ó*). Also, the cyclicity in adrenal responsiveness to adrenocorticotropic hormone (ACTH) is primarily a function of food intake patterns (9). In view of these findings, we explored the hypothesis that diurnal fluctuations in pain sensitivity are entrained by circadian patterns of food intake. The subjects were 24 adult male hooded rats individually housed and maintained on a 12-hour light-dark cycle. They were randomly assigned to one of two experimental groups. In the first group (group A), free access to food was given only during the light phase of the cycle; in the second group (group B), only during the dark phase (10). Water was always freely available. After an 8-day period of adaptation to the restricted feeding schedule, the animals were tested for pain sensitivity by the tail-flick method. The latency of tail flicking in response to a thermal stimulus of fixed intensity was determined for each rat during the first 2 to 4 hours of both the light and dark phases (//). The food availability schedules for the two groups were then reversed, and the animals were tested again 8 days later. The results of this experiment clearly indicate that food availability can significantly alter pain sensitivity and that this effect is largely independent of the light-dark cycle. Pain thresholds were uniformly elevated after food deprivation, regardless of the phase of the cycle (Fig. 1A). A within-subjects, repeated-mea- Mediation of Diurnal Fluctuations in Pain Sensitivity in the Rat by Food Intake Patterns: Reversal by Naloxone Abstract. Rats maintained on a 12-hour light-dark cycle were tested for pain sensitivity after being deprived of food during either the dark or the light phase of the cycle. Diurnal fluctuations in pain sensitivity were observed. The fluctuations followed food intake patterns rather than a natural circadian rhythm, with food deprivation producing a decrease in pain sensitivity. The analgesic response produced by this mild food deprivation was strongly attenuated by naloxone or feeding, suggesting that endogenous opioid systems may be related to patterns of food intake. 6.0 i- Group B (N=24) ^5 čp 0) CO > to to S 1 O) c CO te a > 0) to a. to 01 ■C Q a Group A (N=24) 4 - B Satiated group (N=12) It 'T -C a. a. a> -q- u CD 2 S Deprived group