15,#

ClJilfJter 0,

1.i",'lnli,i",f of rim,/I al/nlysls. In hndllB Ihe nfpha.ganllna Ilril-ed ac:lIvlllcs Ullouch. Ihe sCSlllcnlnl tOOI'~ nnd sume slIfmislllnnl clrculls oclill! on the Rensha'" Dlld III In­

hihilory IlIlcfllelllolls, Illy aim hn\ hcen 1I111SIrllilve only. The descrlptioll Is CO\llI'I~le

cllolISh Ilt show Ih:ll ill Ihe elld we nre rCllching II Iimll (or Ihe scnsible: use of wlrlllldin~ralll\ 1IIIIIIcl:rnli~'e I'hyslolog)'. This Is dlle 10 Ihe hicrarchlc nalUle or Ihe nervous

. sys1rlll. Oil whkh Iltlchli/lIlS Jnckson laid so /Ilu,h el11phasis. There h conlrolllpOncUlltrol a1\(1 each parllClllnr ",tcllnol\1n Is really well undcrslood ollly at lIS own levrl orllllalysis. Itclllclllhcring Ihot Shcrrlnl!IOn ddillcd Inlegratloll as Inlerne/ioll (or 8 pur­I)(ISC. hlll(;linns mllsl nhn be nllrlbulcd to c1rcuilry. Tills ullirnaldy lIleans IIl1denland·IlIg of wiling tlingrnll\s hI n beh3vlof:J1 COllle.l.

, "' he lIif1kuhies rOllhonllng cOlllplelc hehavlorullnlc,prcl3lions of Ihe bewildcr- .in,. ("llIllpl""ily or hllcrn':lifll1S III' hlellln:hic syslems arc virlnall)' IMUflllOlllllllblc:.Known wiring dillllrlllll5 ~cnerl\lIy have 10 be reg:trdcl185 comlrnints Of bOUIIllnfY COII­dilinns tldinllll: p0\5ihle nllef/lallvu. for Ihls reBson, 011I besl InlerprclaliollS of fnn.:­linll h:I\'1: comislell III filling wiring <Il:lgraOls inlo broall conceplllal gencrOililnliollS.slIch 015 rcdprt"..;il Jnllcrvnllon. nll'ha0l!nl1lllla linbee wilh liS Il1lplkalions for 11I010­

lIelllun mClllbrane ,'Mcnllilh, IncdHlllh,"s (ur stOlbililllllon or nCllronal discharges,Iccl1"ad operaliolls, nrousal,loati c:ol1ll>Cllsatiun. itlcas Oil poslure. cle. Orallil, 19U

Modes ofCentIra~'PrOCeSSHlJllgnn'HlUlm'an'learlllling and

Rememberongby K.B. Pribram

In: T.J. Teyler (Ed.)Brain and Learning.

~,Stamford. Conn:Grey1ock'Press. 1977

~~._.---~-----,--,"-------_._-_._...__ .__ ._ ..,

Inrroe/llclion

'fllt: :tilll of litis ch~plcr is 10 relate 10 111111"'" brain processes Ihe

reviews of Ihe rccenl surge of research on prbblems 'of learning ilndmelllory n' Ihe: molecular and wiring circllillcvclwhieh make up Ihe re­mnill(kr of this volullle. As nole<.lby Granil 111 Ihe above quolation frol11his allaly~is (If Ihe role of muscle spindles, Ihis endeavor Is /lol an easyonc. Still, Ihc silualinllis nol os llt:spcrnle al it was n quarler of a century:lgn when Lashley mode his famous slatement 10 indica Ie lhat what we1hen knew lIbOll1 hruin ftlnction precluded learning frolll occurrillg ot all(Lac;hley, 1950). The phrase was, of coursc, made longue ill check, ulIlsuhscqllcnl research (as well Wi somc cllrlicr formulations-c.g., WilliamJ'lIlle~, 19511; S. S. SleyenS, 1951) havc hortle onl nn intuilion hiddcn illl.a~"lcy·s slalclnell(: lhat the limilations 011 coping with complex ClI-

·j

1411 /BrOlin & LCOlrning

vironrncnts, and thereforc the potentialities for ovcrcoming these limita­tiOllS, arise not 50 much becnuse of restrictions all the usc of the finalcommon path-the efferent control over Inovelllelit-as Sherringtollonee suggestcd (19,n), but on limitations in central processing-i.e., inbrain function (Oroadbcnt, 1974; Pribram, 1974). Once the problem isclearly framed in this fashion, the prmpeets for relnting the molecubrand wiring levels of neurophysiological research to those involvinghuman learning and memory become comdderably less bleak.

In this presentntion I will rely heavily on the results outained with asingle technique to delineate the sought-nfter relationship. This tech­nique, experimental psychosurgery-the study of the effects of localizedresections of the brain tissue on behavioml performances-focuses on aneuml structure and asks what role it displays in cnusing or overcomingthe limitations on ccntrat processing mnnifest in learning and remember­ing. This is'a rather different sort of question than that usually asked inbiological npproaches to the problem. These ordinarily investigate con-

. solidation, habituation, conditioning or'discrimination by inquiring intothe molccular or circuit changes produced while the beh:1Vior under con­siderntion is in force. In contr:1st, the psychosurgical question, by its Cn1­

ph<lsis on processing limitntions and potentunlities. is akin to that whichmotivntcs research on human learning and remembering.

With Ihis pnrallel in mind let m look at somc of the principles ofbrain organization important to learning and remcmbering which havebeen reliably established by psychosurgical experiment.

The Dis lribu led 5lore

The bcst known principles that have resulted from psychosurgicalresenrch are the ones that Lashley wns nddressing in his pcssimistic state­ment. One of the most persistent results obtained when restricted resec­tions arc madc of brnin tissue is "nothing." Lashley formulnled thisrcsult into his laws of (a) mass aetion-Ihat to be effective a brain lesionmust exceed a criticnl size-and (b)· equipotentblity-Ihat spared brainlissue can come to function in lieu of lhat which h:u been resected.

A great deal of misunderstanding of the nature of brnin organiza­lion hns resulted from Ihcenunciation of the principle when formulntedin this fashion. Psychologists especially were prone' to accept the ideathat all parts of the brain functioned alike (and therefore thntthe studyof the detailed ~tnalomical arrnngements of br3in organization hadbecome supernuous). Neurophysiologists and Ilcuro:lI1atonists, on theother hand, intimately acquainted with tho:: exqubilcly precise wiring ofthe brain, looked <Isknnc(: nllhe tyre of bcl1nvioral annlysis which was sogross as :0 miss the obviuus distinctions between brain parts.

The resc;lrch results of the past lwcnty-fivc YC;lrs can, as we shallsee, put these misconceptions to rest. The brain has been shown 10 bemade lip of systems which m;1l1ifcst different functions. Howcver, withinony system and even 10 some extent belwcen systems, the I;1WS of Ill;1SSnction and eqllipotentiality hold. Electricnl recording of brain potentialsevoked by sensory events hns tlemol1stratedthnt the basis for the laws is

.,...~

,..<

=.1

"", " .. -' .,., ~.,.. ,~....-,.,,-~--_._----~-

. ,'~

Modes of Central Processing in Human I.cilrfling .1nd Relllemuerincf t·19'

the distribution of input over n wide exlent of tissue (John, 1907;Pribrnm, Spinelli &. Kamback, 1967; Spinelli, Starr & Barrelt, 1%8; •Bachy-Ritn, 1972; Morrell, 1972; PrilJrall1, Nuwer &. 13arol1, 1974). Theconclusion has been reached, therdore, thal the lack of crfect ofrestricted resections of brain tissue is due to the encoding of input whichhas become fairly widely distributed withinany particul:lr neural systern.

The mechanism of distribution and the nature of the encoding pro­cess are meas of current research activity which rclate the molecular andcircuit level of inqufry to the behavioral. I have elsewhere (Pribram,Nuwer·& Baron, 1974) detailed the various possibilities .and the prob­ability tlult distribution depends on the interactions among hyper- anddc-polarizing potential changes at synaptic and dendritic locations, in­ternctions which then result in conformational changes in local mem­brane proteins (Pribram, 1971). The result of such an encoding processwould be a distributed store with holographic-like properties that makepossible the construction or reconstruction of the configuration of the in-

.put from any restricted part of the store .A second property of a holographic-like distributed store relev:lnt 10

the circuit level of inquiry is its facility for organizing associativememory. Whenever two inputs occur together during storage, the subse­quent occurrence of either alone will evoke a "ghost imagc" of the othcr.This ilssociative property of the holographic memory mechanism is animportant nltermllive to the step by step forging of neuronal cOllnec:ion.~

as n function of repetition and prnctice. As we shall see there is a COIl­

siderable amount of evidence that both types of associative processesoccur-one is a fairly rapid "imprinting" of input, the other is more ex­tended in time and critically depends on repetition.

But in order to fully appreciate the evidence it is first necessary toreview Ihe dnta that highlight the diversity of brain systems involved in\c:lrning and memory. Only against this backgrO\lIld clo the distinctionsnnd commonalities among processes become fully evident.

Sensory Specificity in Central Processing

The diversity of cognitive processes Is manifest in the first InstanceIn their sensory specificity. When resecllons of primate "association"cortex arc made, lhc expectation 'that some general assocbtivc or lenrll­Ing capacity would become impnircd is not borne oul. Learning deficitsdo result, but these nrc limited to .one or another sensory mode-whichmode is affected depends on the IOClI~ of the lesion within the extent of"as5.oci:ltion" cortex (Pribram. 1960). This experimental result renectsclinical experience with m:l11 where "agaosias" due to brain injury nrc, asn rule, restricted to one or another sell50ry mode. Even in the illtact per­son it is difficult to demonstrate cognitive processes that nrc not sensorymode specific (Wa\l:lch & Averunch, 195~). Most thinking is pursucdeither in terms of incipient sounds (:l\l(!itory), im:lgcs (visunl), fcels(somalosensory) or tastes (gustatory and Olr:lClory).

The sensory specificity of cognitive processes docs not preclude theiroperating on morc wholistidlly org:lni7etllllcchanisms. The disiribllln\

,,I. I

s,

{\j, ,

, I

, I, :

l i

\:,U IBldll1 & I.C.HIIIIII;

storage is a goou candidate for providing the matrix of such operations.Thus a paradox exists-the association cortex operates within a wholisticmatrix. but the operations nrc sensory mode limited. The paradox isresolved by evidcnce that the sensory specificity is due to the discrete ol/f­pilI from localized POri ions of the association cortex to one or anothersensory projection system (Ulum, Chow & Pribram, 1950;Pribram, H. &Il:my, 1956; Pribram & Bagshaw, 1953; Dewson, Pribram & Lynch,1969; Pribram, 1%9). Cognitive processes arc thus found to be akin tomotor or command functions (sec, e.g., Mountcastle, Lynch,Georgopoulos, Sakata & Akuma, 1975 for a detailed analysis ofresponses to desired objects in mediate extrapersonal space). In fact thecritical output p:llhwny from the associatiofl cortcx is to the basalganglia, structures that have classically been considered to be motor infunction (see below).

MOlOr Functions as Centr(,1 Control ProcessesThese data fit with others that have revised our view of the opern­

tion of the motor systems of the brain. The classical view held that motorcontrol was cxerciscd directly on muscle to shorten it or to incrcase itstOllC. Over the pasl 25 years it has become cvident Ihat an even largcr I

share of the control issues to the muscle spindles, receptors Ihat arc con­nel:led in parallel with Illtlsch: fibers and thus monitor their contractions.Ct:ntral r~gulatory mechanisms depend on signals from the spindle 10 ad­just the system as necessary. Control is exercised by modulating themonitor through feedback loops, IIlll(:h as Ihe selling of a thermostat canbe changed by reselling lhe control dial (see, e.g., reviews in Miller,Gatant~r and Pribrall1, 1960; Pribram, 1960; Granil, 1970: Pribram,1lJ71; Granit, 1975). Thus the mechanism of motor control is akin to thatof homeostasis ralher than that of a piano keyboard: Motor systems inthe brain to a large extent send signals to receptors, not effectors.

Although flut as varied as the Sensory mechanisms, n considerablediversity of organization also characterizes the motor control operationsof Ihe brain. Two major types of operation ar\: identified in the clinic:one is mainly concerned with postural readiness, lhe other with the ex­ecution of skills.

When discases :>trike the basal gangiia, patients show posturaldisturbances and more or less continuous involuntary movements such astremors. The lype of disorder depends on which of the basal ganglia isaffected. Faulty fee<Jback duc to the discase is held lcsponsible for thedisorders (OllCy, 1949). The dis! urbances have recently been at least par­tially overcome by the adlllinistration of DOPA (Dioxy phcnyalaline)which suggests that tile trernor~ arc due to the (kpletion of this catecholumine which is ordinarily found in especially high concentrations in thebasal ganglia (for r\:vicw s\:c Ungerstedt, 1974). When overdoses ofDOPA nrc adrninisler\:u cognitive disturbances lIppear-these and thoselhat accompany the [)ostural disturbances could readily rcsult from thecontrol by the basal gangliu over sellsory fum:t!ons noted above. Morc ofthis in a moment.

",

',\

'.',::,r'..~

i·

Ii,i ;1I: .~ :

, j: i

i, I

i I; :1: f'

.]~

~i:;t _ .•-;"

I.1

I---,-.-.__ _._---_._~

Modes of Central ProcessinG in Humiln learning and Rcm~rnb~rinl:1 151

The second major type of motor control centers on the cerebellum.When the cerebellar hemispheres are injured, the patient suffers a loss ofcontrol of skill-his movements become as awkward as when h'e initiallyundertook the activity. The development of a skill entails the smoothcoordination of various muscle groups and the elimination of extraneouscontHlctions. Precise timing is of the utmost imporlance and thecerebellum is most likely a very powerful computer that can calculate "infast time"-Le., before a movement must occur-what its outcome willbe (Ruch, 1951; Ecc.:1cs, 'lto llnd Szentagotha, 1967; Pribram, 1971).

The difficulties produced by cerebellar lesions in the execution ofskl1led acts become especially noticeable when patients inlend, Le., will,their movements. Voluntary control becomes manifest when a signalregulates two or more mechanisms in parallel. Dy contrast to a feedbackloop which is "c)osed", fhe parallel processing that defines voluntarycontrol forms an open or helical loop and is called a feed forward(Teuber, 1960; MacKay, J966; Pribrnm, 1971; McFarland, 1971).. Thenrrangement of the cerebellar output provides just such a multipledisposition of signals-fo the periphery, to the cortex and to the nuclei ofthe upper brain stem which connect to the basal ganglia (Eccles, Ito &Szentagotha, J967).

Cerebellar control over muscular conlraction is accomplished; muchos other central controls, largely through the regulation of receplors. Thequestion has not as.yet been investigated as [0 whether cerebellar outputcan regulatc sensory as well as motor receptors. However the senses arewell represented in the cerebellar cortex through input fibers. Thisrepresentation may well be the immediate origin of !>ignals fhatsimultaneously move a sense organ (e.g., an eye movement) and changethe setting of the brain's receiving mechanism for that sense, sufficient tocompensate for the~novement.

These considerations added to the finding. that the basal ganglia areinvolved in regulating sensory functions make it necessary to view themolor mcchanisms of the brain not just as movers of muscles but ns cen­tral conlrol processes thaI operate on n varidY of other ncuralmechanisms and even on the senses. Thus what an organism senses is tosome considerable extent what he is set to sense, i.e., whai he is compe­tent to sense and what he attends. Perceptual competence and attentionare therefore akin to motor skills,

Perception as Central ProcessWe are thus forced to Ihe conclusion that what is usually called a

motor skill is a somatomotor skill involving the somatosensory systemnnd Ihat what we ordinarily refcr to as perception is a viSlloll1otor.auditory motor' or other special sensory motor skill. This view Is sup­portcd by evidence (Malis, Pribram & Kruger, J(53) of a relatively directinput to the precentral somatomolor cortex from somatic receptors (skinand muscle). Also, in the visual mechanism.at least, vis\lomotor systemsabutt the areas receiving retinal input (as is the case, in the somatosen·sorymotor cortex), and these systC:lI1s have rectntly been shown to b~ im·

~."

'oj.,." 1

Ii

:d

I, I ~ I ,

~ I

. !

;'1I;i,!;

porl:,"t to thc percept ion of constancy: (specifically sizc constancy; abla­lion of this eorlex prouuces a monkey who nltentls retinal image sizewillie ignoring di~l:ll1cC clles-Ungerlcidcr, G:\IlZ & Pribr;lnl, 1977).

A caution is in order herc. The Yicw thaI perception involvcs asensorY-lOotor skill c10cs not mcan that perceplion is a central motorrcsponse essenti;llly devoid of any scnsory component-the positionlakcn by Sperry (1952) anti [-estinger, Burnhalll, Ono, & D;llllbcr (1967).Rather, neurology nnd psychophysics (e.g., Gibson, 19G6) as well aseveryday experience indicales that perceptual skills whether somatic,visual, nuditory or olher (e.g., guslntory-olfnctory) are sensory motorperformances of a special sort: perceptual acts Ihat encompass reliablyrepealable, I.e., invariant, environment-organism relationships. Themolorcomponenl of these acls is not so mUch a "response" to input as itis n control (often n readiness) over the sensory input mechanisms.

In humans a further complexity arises. Specialization occurs in thecontribution made by each hemisphere of the brain. Auditory-verbalprocesses dominate the adult left hemisphere (in most right handed per­sons) and visual-spatial processes play the major role In the right (seereviews in Dimond & Denumont, 1974). The differences betweenauditoryand visuallearriing and remembering noted above thus becomedramatized as differences in hemispheric function-differences betweenlert brain and right brnin skills.

The Central Processing 01 SkillsThe mechanism for le:trning sensory-motor skills has. been studied

and is well described by the observntions of ethologists (see the review byBateson, 1964) and those engaged in experiments on perccptuallenrning(e.g., Gibson, 1953). As Dateson points out, the laws of imprinting andthose of pereeplul learning are not altogether different. And, on carefulexamination, the laws of simple instrumental or operant conditioningclearly converge on the same processes from the .vnnt:1ge of 3 responseoriented psychology (sec f9r instance Premack,.19GS). In each instancenn innalefy determined environment-organism relationship is modifiedby experience in which consequent occurrenccs slwpe and differentialethe innalely determined proce,,;s. Shnping as well as imprinting consists ofrelatively gross initial modi rication of the environmcnt- organism rela­tionship which is accomplished rather rapidly (e.g., the imprinting otfollowing a moving object mther than random i1ivesligatory sensing; thepressing of a lever rather than random investigatorY,movement). This in­itial change is followed by slowly progressive differenli:llion of the rela­tionship (differenlialing lhe imprinled object so that first any similar andthen or.ly il per se elicits the response; pressing thc lever only when nnS D is present). .

The vnri:lbles Important to shaping nppear to be "stimulus novelty"and "response density." ·The role of stimulus novcllY (nnd therefore 'Ofstimulus familiarity or repetitiveness) has becn clearly documented forimprinting by Hess (1959) tlnd Dateson (1964). The relnted concept ofresponse rate or repetition density has been invoked by Prcmack (1965)

.j.

-.i'

Modc~ of CClltrill I'roccssillt.: in IlUIlIil" Ll',lIIl1lll: .1Ilt! f{erncIlIbcrilll:l bJl

to explain under what conditions any .'ipeeirie innate behavior can serve areinforcing function to rt!'other. E. Roy John has shown (1967) that in~

itially during the shaping process a great number of brain sites (especiallyin thc core of the brain) show electricnl activity. As di'icriminution(perceptual) learning proceeds, the loci involved become markedlyrestrictcd. Just what goes on in thesc more rcstricted locations is current­ly uncler study by Jamcs aids (aids, Disterhoft. Segal, Kornblith. &Hirsh, 1972).. No clear picture of relationship among loci has as yetcmergcd; perhaps, as we shall see shortly, no simple lime dependentcause effect process is involved.

These studies and the more common behavioral analyses suggestthat the differcnccs between the processes of learning and memory in dif­ferent modalities are atLributable to differences in their differentiatingmechanisms and that commonalities are to be found in those processesresponsible for initial imprinting and shaping. This conclusion supports.the intuition of those who are concentrating their Investigations of neuralmechanisms on problems such as habituation to novelty. simple condi­tioning, and consolidation of the memory tmce.

Centra! Processing in Orienting and its HabituationIn analyzing the mechanisms involved in the simplest form of neural

modirication resulling from experience. the npproach of iI'ikillg abouttlte contributions of specific brnin structures to the limitations andpotentinlities in central processing hns proved rewnrding. Decrementingof neurnl responses occurs in interneurons of all sensory systems whenthey are subjected to repetitive stimulation (sec reviews in Horn andHinde, 1970). Any change in stimulus elicit'i more continuous responsesin another more medially placed neural system and these two systemsconverge to produce the typical behavior of habituntion nnd dishabilu:l­tion (Groves :lnd Thompson, t970). The medial cells which for a timetrack or monitor the change in stimulus nrc located in that part of thecentral nervous system in which viscera-autonomic neurons originate.Thompson's unit analyses were carried out in the spinal corel but similardishabituating "arousal and monitoring" behaviornl effects me obtainedwhen the core structures of the brnln such as the mesencephalic reticularformation and the hypothalnmic region arc stimulated electric"lly (secrevicw by Pribram and McGuinness. 1975). Furthermore, the viscero­autonomic responses that ordinarily occur during orienting and dish,,·bituation arc abolished by resections of certain forebrain structures: theamygcl:1la (Dagshaw. Kimble and Pribram, 1965) and front;ll cortex(Kimble, Bagshaw and Pribrnm. 1965; Luria. Pribrnm and Hornsbya.19(4). There thus appears to be an intimate relnlionship between themomal nspect of orienting and dishabituation and the visceroautollomicnervous'iystem. .

Another peculiarity was manifest by these rescelioru. Whilevisccroautonolllie reactivity 10 novelty .W:!'i aholished, behavioral orient­ing remaincd intact. Not so, however, with regard to behavioral habitua­tion which was abolished along with the visceroautonomic oricntinr,

I·

~ '.

responscs. Thus uchavioral habituation to novclty-the appreciation offallliliarity-seems to lh:pentl on the occurrence of visccroautonomicreaction to novelty.

This is not to say thaI learning and remembering cannot occur in the"bscnce of habilualion. The evidence is clear that discrimination learn­ing, the making of seleclive diITercntial responses to cues, is unimpairedby the brain rcscclions lhal interfere with habituation (sec, e.g., Douglasnnd Pribram, 1969). Such resections do, however, severely in\pair thelearning and retention of the ability to perform adequately in tusks suchas discrimiriation reversal, delayed alternation and, in some instances(i.e., resection of frontal cortex), delayed response (Pribram, 1973).

The analysis of the neural mechanisms involved in the orientingrcaction to novelty and its habituation thus leads to the same points asthat obtained from the analysis of the neural mechanisms involved in thedevclopment (learning and remembering) of scnsory-motor skills: twoseparate mechanisms can be identified. In the case of skill 11 rllpidimprinting-shaping process can be separated from one that Is more ex­tended over time and lrials alld leads to an invariant organism- environ­ll1~nt relationshii1. The imprinting-shaping proccss appears to be akin 10ori~lIling and ils habitual ion, ami the brain structures direCtly involvcd I

have lillie 10 do with learning and renlembcring invariances. lnstend,lkrccts in orienting and ils habituation and thc learning and remcmber­ing of regularly v~rying .performances such as n:vcrsals, alternations anddd:.t}' tasks arc produced by resections of the identical brain structures.

These dala lead to the inescapable conclusion that two rather dif­ferenl brain-behavior syslel11s, or rather sets of systems, operate duringlearning nnd remembering. One, as we have seen, leads 10 the processingof invariants. The olher is more concerned in processing recurrent varia­tion. III this set uf syslems, for learning (i.e., habituation) 10 occur,changes must be of sufficicnt regularily 10 be computable. When recur­rencc resulls in a residue of the computation, the residue provides thecontext in which further experience can be processcd, and any subse­quent recurrence is trealed as·" fnmiliar", thus precluding orienting.Visceroautonomic reactivity appears. 10 be integral to such computations(Pribram,197l).

Two Modes of Cenlral Processing in Man

A well dcsigned series of experiments has been performed thataimed at relnting diverse sets of data on the limitations of central pro­cessing. Usually these data arc considered in experiments on selcctive at­tcntion, sensory search, undmernory scanning in siudies of hutnnnicarn­ing and remembering. The results provide additional insight into the dif­ference betwecnthe two selS of systems we have been delineating nnd thepsychological processes which they control (Schneider and ShiffrinIIn7). The experimcnts did Ilol involvc manipulation of brain structhres;however, in cOll1ra51 Itl mmt sllIllies 011 cognitive behavior on nlanwhich lise verbal or pselldoverbal malerial, these experimental proce­dures rescmbled those lhal chnrncterizctllhe nOll-human primatc psycho­~lIq;ical rescurt:h thai gave rise 10 lhe distinction between the two Iypes of

Moues of Ccnlral Process inc in Hum,tn lcarninc and Remernbcrinl:! 155

learning and memory processes (sce below and review by Pribram, 1969).the human work centers on a display framing a variable number andtypes of patterns-e.g., numerals, lellers or geometric forms. A variablenumber of such frames can be prescnted to a subject aftcr which a singlepresentation is made containing olle or more of the pallerns (numbers,leuer or geometrical for IIIs) that have prcviously appeared-or alter­natively, patterns that have not appeared. The specific experiments con-

. sisted of embedding the to-be-remembered patterns In frames containingsimilar types ("same-distraclors") or different types of patterns ("dif­ferent-distractors"). Thus the numerals 7 and 2 might be enibedded in

.frames fUl and nn in Ihe same distractor condition, while theyl.12J L!~J fiXl r21l1

would appear in frames lMtl and l.!!9J in the different distractor con-dition. Frame size (Le., number of patterns per .frame) and number offrames in the to-be-remembcrcd (memory) set could be varied at will oneach episode (trial sequence).

Using this technique, dearctlt differences between the same and dif­ferent distractor were obtained in the effects of varying frame size nndsize of the memory sct on the number of correct responses and onrcsponse latency. To simplify a large number of results in a variety ofcondilions: in the different-distractor condition, rcsponsc appcared to beautomatic in thatlateney was short and varied lillie with episode (frameand memory set size). However, many errors were made originally wh~nframe amel memory set size were large and these droppped out onlygradually with practice. Dy contrast, when same-distractors were usedlatency of response varied (linearly) according to episode (frame andmemory set size) indicating lhat processing demanded active searchspecific to the episode. Practice had little effect-however, subjectscould eliminnle or uttenunte the influence of snme-distractors wh~n pro­perly instructed or when they themselves adopted an appropriale'response strategy (e.g., exhaustive vs self-terminal,ing, Le., speedy. scun­ning). In the same-distraclor, episode sensitive, condition, the poshionof the distractor proved to be an important variable: when self­terminating scans were adopted they oflen eliminated certain positionsfrom scan thus leading to error; when scanning was exhaustive, the posi­tions of the distractors had a sizeable influence on reaction time. No suchposition effects were obtained in the different-distractor conditions inwhich processing was automatic.

As noted, these experimenls are similar to many that have been per­formed on non-hultlan prirnilles. In the animal expl:rirnents, however,brain rescctions were carried out so that the two modes of processingwere related to specific brain systems. The resuILs show the posterior con­vexily of the brain to be involved in the aUlomatic type of processingwhich in earlier reviews (Pribram and Melgcs, 1969) WllS called "par­tkipatory", and Ihe fronlOlimbic forebrain in the episodic Iype(previously called prcparatory).Episodic and Automatic Central Processing

'II

\.\

""""ji'l '.~ 'i~:... !r--, •

\,I.I! •

i:

I.i·I

! I,.iiI

II.!

i;::

~Il

Terminology nlw;lys poscs difficult problems. 111. this ins lance"episodic" nnd "nutom:ttic" were chosen 10 describe the two centralprocesscs because Ihey best COl1l1ote the body of evidence that is covered.The term "episodic" is Tulving's (1972; Tulving ~nd Gold, 1968) and isb:lsed on his data that indicate thaI a memory process specific to 1n-

. ddents or episodes cnn be distinguished (rolll that which organb:es thelong term mcmory store. This spccificity to situntion is what Chomsky(1963), QulJlian (1967) and Pribram (1971) have called context specificityand context sensitivity to distinguish it from the context-free processesthat handle invarinnt relationships.

Shrirfin :lnd Schneider call the context dependent; episodic memoryprocesses "controlled" because memory search Is found to be self pacedIn their same-clistractor situations. Thclr term "311tomatic" refers to thecontext free proces~ing of invarinnts which proceeds according to theproperties of the ~ituation rather than t110se imposed by the organism. Itis adopted here because it emphasizes the automatic nature of the processfor which the term "senrch", which Shriffin and Schneider use, Is reallyinappropriate (sec Pribrnm, 1971, Chapter 17).

Kinshela (personnl communication) has also emphasizcd theautomnlic nature of the processing of invariants. He has developed andtested a mathematical model which shows that U1C selection of invariantsproceeds automatically from considerations of the amount of "noise"and thc structural redundnncy ill the slluation. Garner (1962) ha5specified the tmdeoff between external redundancy (how many featuresor dlmensions·of a situation spccify a diffcrence) :Iud the Internal reclun­cl;\I1cy (how much di ffcrenliation has the organism alrendy achieved).Kinshela's model docs not distinguish between external and internalredundancy but Wilson (1968) and Pribram and McGuinness (1975) h:1VCformulated a model of dwnnel competence based on an informationtheorctic approach to psychosurgical data which takes this relationshipinto account.

The issue conccrns the theme of this chuptcr: the wny In which thelimitations and potcntialities of central processing can be understood.Much of the work in humnn cognitive expcrimental psychology andespecially that addresscd to allention, deals with the problem in terms of

". limitations on channel capacity. The psychosurgic:l1 data reviewed in the·first section show, however, that there is actually a fantastic excess ofchannel capacity in the brain-lesions hardly affeel capacity even whenthey involve 80·90 1l/1l of the channel. .

Thus, the limits on processing must stem from some other propertyof the channel than ils capacity. This properly has been shown (Pribrnm& Tubbs, 1967) 10 be the s,tructure of the redundancy of the process. Ananalogy helps to point out the difference bet ween a capacity and a com­petency concept: limits on capaclty can be conceived to be similar to anexoskeleton, whereas processIng limitatIons due to an in:H}equate struc­ture of redunuancy are more nkin 10 an endoskeleton. Endoskcleta havethe advanlage lhal they can be ncxibly "restruclured" when the situationdemanus. Rcstructuring becomes a cognitive skill (Pribram. 1971).

.1

..:I,

,',:,('.

Moues of C{!lltrill J'roctlssilltt ill IllIlll.111 LI~.Hllil1l: .1IHII~t'Il11:/lI"cri"I;' 157

TlltlS, rather than depending on limitations of capndty, ccnlral pro­cessing appenrs to partake of n skill-a competency. A strict definitionof channel competence is given in information theoretic terms by thestatement that competence is the reciprocal of equivocation whereequivocation is the sum of noise and redundnncy (Pribrarn & J\'IcGuin­ness, 1975). This definition of competency is identical to the mathemati­cal derivation used by Kinsheill.

To summarize, human experimental approaches to cognitive pro­cessing and those that have come from the study of the effccts of selec­tive brain resections on cognitive behavior have demonstrated the ex­istence of two distinct central processing mechanisms. One deals withspecific episodes, is therefore context (i.e., situation) dependent andnecessitates a considerable amount of centrally controlled computing ofthe regularities (recurring variables) that describe the ~itualion. The otherprocesses invnriances in the relationships between the organism and thesituation and thus processing proceeds. relatively automatically withrepetition.

Particular Brai(l Mechanisms Involved in Central ProcessingWe have therefore once again, and by still another set of experimen­

tal data, arrived at the distinction between two clearly different type.~ ormodes of processing important to tearning and remembering. One pro­cess is demonstrated to be involved in shaping, in oricnting and itshabituation, and in active control over specific episodes in expcrimentson sensory search, attention and memory. This process is drastically in­terferred with by Icsions in the frontolimbic pnrt of the forebrain. The se-

. cond process is demonstrated to be involved in pmctice and in the atlain­ment of sensory-motor skills and is shown by experiments on sensorysearch, nllention and memory to be nutomatic. This process is drnstic:lllyimpaired by resections of the posterior convexity of the cerebral cortcx.

Much more can be statcd about the relalionship bctwecn brnin anelthese two behaviornl processes. As nOled enrlier, I he nutomatic process issensory specific and may be different in the way it is structurcd in dif·

.' fcrent sensory modes. Such differences have as yet not been systematical­ly investignted. However, automatic processing has been shown 10 be asimultaneous parallel process involving the long term memory store (secreviews by Neisscr, 1967; Pribrnm, 1971j Schncider, 1975). It \\'a5 alsonoted enrlier that the studies of Roy John (1967) and James Olds (Olds,Disterhort, Segal, Kornblith and Hirsh, 1972) have failed to yield onyelcur ellt sequential, I.e., time dependent cntlsc~crfcct order in lhe "r­penrance of critical c1ectricnl events during such automatic processing.Rather many such events occur simultaneously in n variety of structures:md behavioral response oppenrs to result from a correlation amongthese events (see also Pribrom, Dny nnd Johnston, 1977).

A good tleal is also known ~bol\t the rdation~hip hetween brnlnsystems nnd episodic processing. Different systcms of the fronlolill1hicforebrain have different functions in the overall c1eterminalion ofepisodic control. As noted, resections of frolltnl cortex nnd nmygdnla

.. _--~- ..----- ._-- .. -- -"-"'_._-'--'-'-"--'-~----

.\ 1SU /Urain & LeMnine

r~slllt in impaired viscaroauionomic responses during orienling anti a. wbsequent failure of behavioral habitllotion to occur. Orienting andI habitualion are nol per se controlled processes-in fact thc orienting

r-t:nction is oftcn referred 10 as the orienling rencx. However, orientingdoes serve as a signal thai interrupts ongoing brain"behavior activllies­I.e., it signals novelly, a distractor. If the distracting event is repeated.habiluation resulls unless it is overridden by some other mechanism. Ac­cording to the data prescntcd hcre two types of override appear to occur.One type sorts cvents into differcnlial categories and by praclice developsdifferential sensory motor skills to cope with the differences. The otherOV~I'COI11CS hahituation by "effort"-Le•• by coordinating the tendencyto relurn io prior aUlOlIlalic processing wilh the tendcncy to continue to .oricnt.

Again. considerably more can be slaled regarding the neural strue­lures involved in making such coordinations. Automatic processing hasbeen shown to involve the connections from the cortex of the posteriorconvcxilY 10 the basal ganglia of the forebrain. syslems on which scnsorymotor rcadincss depends (Pribram, 1977). As already reviewed. theIcndcncy 10 orienl is a function of a fronto-Iimbic forebrain (andhypothalamic) syslem. The functions of the rem.liness and orienlingsyslcms are coordinaled by a third. the hippocampal system. Evidencefor stich coordinalion and Ihe errort involved has been reviewed in delailelsewhere (Pribrall1, 1971; Pribram and McGuinness, 1975; Pribral11 andIsaacson, 1976; Pribralll. in press).

O\'ercoming the Limits on Central Processing. A further suggl:slion has been tendered in the form of a model(Pribram and Isaacson. 1976). The functions of the hippocampal systemare conceived to be similar 10 Ihose of the cercbellum (histologicalparnlkls abound) in Ihat bOlh arc critical to the development of fecd­forward, open-loop (helical) brain and behavior processes (see Pribram.1971; McFarland, 11)7). In Ihe case of the cerebellum this is suggested 10be accomplished by compuling, in fast time (as opposed to real lime). acorrelation belween feedback controlled spinal and brain slem seri­sorymolor rcflexcs on the one hund and similarly feedback controlledbasal ganglia rcadiness. In the case of the hippocampal system, ns wehave scen, Ihe coordination occurs (also in fast time) between fccdbackconrrolled fronto-amygdala-hYPolhalamic orientiilg reflexes nnd thebasal ganglia readiness /llcchanism. FceMorward is conceived 10 resullwhel1 Iwo sequcnli:llly operating feedbacks can serve 10 bias Ihe olher(Prilmllll, 1971; PrilJram and Gill. 1976). Feedforward is proposed 10 ac­counl for the succcss of biofeedback proce~urcs lhat introduce a secondexll:fnal feedback which uecomes coordinated with the inlernal, IhllSproviding as il wcre a "proslhcsis" thaI cnhances the ordinary limila­liulls of lhc crforlnll:chullislll.

Thesc Jilllilaliolls have been dcalt with expcrimcnlally as lill1ilaliuJlSor shorl (errn mel110ry or allerualcly of allcnlion span. The purpose oflhe studics on hlll11al1S bridly reviewed earlicr (Schneidl:r. 1975) was to

,,

~""'-~'."•."f'.' "f'.-" ...~ .... ,.. -..

References

Modes of Central Processing in Human learnlnc and Remcmbcrinu/ 159

!

II

show that the results of experiments on sensory search, on attention, andon memory scan could be accounted for by a single theorc:tical formula­tion. The success of the endeavor suggests that we can conceptualize thelimitations on central 'processing in a unitary fashion and theneurological data noted above support the formulation (see alsoPribram. 1974). . '. A practical consequence emerges from this analysis. Central pro­cessing limitations exist ubiquol.lsly, whether because of inadequate en­coding of earlier experience, brain injury or inadeqllille heredity (whichnil of us sense to some extent in the highly complex society only a com­bination of brains could have construcled). Thus sensorymotor pros­theses based on the episodic-automatic distinction delineated for brainfunction can become useful engil,eering and educational instruments fortherapy and for growth.

In conclusion. experimental psychosurgical studies of the integrativephysiology of the brain. when coupled with neurophysiological data andresults of human cognitive work on learning and remembering, are pro­viding broad but specific generalizations applicable to man. This essayhas delineated some of these: the distributed store,sensory specificity incentral processing, motor functions as central controls over input. andthe distinction between episodic and automatic prpcessing. The yield isrich and shows practical as well as theoretical promise.

lIaleson. P. P. G. Chances in chicks' responses 10 novd moving objects Ovcr the sensitiveperiod for imprinling. A nI",al. llehuv .• 1964./1.479-469.

llachy-Rila, P. Brai" Mechanisms in Sensory Substitutioll. Ncw Yor\.:: Academic Pre~s.

1972.

nagshaw, M. H., Kimble, D. P. & Pribram, K. H. The GSR of mon\;eys durins orientingand habituation allli after ablation of Ihe amysllala, hippo.:alllpus and inrcrotcmporalcortex. Ne"ropsycholol;ia. 1965, J, 111-119~

ilium, J. S., Chow. K. L., &. Priuram, K. H. A behavioral analysis of lhe Or~Jllilalion oflhe parielo·tclllporoprcoedpital COriCK. J. CO/lIp. Nellrul., /950. 93. 53-100.

IIroa-lI":III, D. E. riivbilln or f\lll.:liOIl alld inlq;ralioll. NCllrOIL'iL'm·cs·Stlldy /'roSrOlll. Ill.New Yor!;; MIT I're~~. I!J74.

•

'·'1~';f;~':!' :\\1 :~ i: "I

;'\"j :c"

~i. ~j ill," . i,l;

~~,:""" ,I.

~'f· . ,;'Il~i!l.': ;:1 11r::: '

~I :1'. ~.

ilL ", II~, ",. I

tr .J. ~:

I i~ ",I",\'.~~::/:'':.b;~

"

", , ':1'",I

:" I:.....: .Ji,l:,'.. .,~r I

,'t']'I'nl)

-\·1," ' \111'

T'ji,, .

~'I 'j,

J

q.,l".'f;,.

"" '"",r:"I.:,!

;'

~ I; ~ ~I' I

t l:~i 'j~. Ii.

< I:) :'1.. ,

::~" I

:~ I~ !~~ !/; ;·i

:q'·i

I ,

,,

.;:"

"

1&0 IOr"in t. Lr;lrning

Buc}' P. C. Introduction nnd Chpt. XIV. Errects of c:'tlirpnlion In 111:t1l. In TIr~ PacmtralMotor Cortrx, edited by P. C. Ducy (2nd cd.). Univ. of Illinois Prcu, 1949.

Chom~ky, N. rorm:tl proprrtic! of gr:lmm:trs. In R. E. Luc~, R. R. nush :IIld E. H.Gnbntcr (cds,), /(olldlrook of Motlrc/l/otical Psychology. New York: John Wiley &.S(lIlS,/I, 323-418,1%3 .

Dour1:ts .. R. J., ."- rribr3rn, K. II. Lcarning lind limbic lesions. Nruropsyc!lol., 1966,4,197-220.

Dewson, J. H. III, Prihranl, K. II. &. Lynch, J. Effccts of :tblalions of tcmpor:!1 cortexupon spe~h sound ?iscrimin:tllon In the monkcy. Exp. Nrurol.•1969, 24, S79.S91.

Dimond. S. J., &. Deaumont, J. G. (Eds.). 1I,mis"lltre Function In the Uuman Drain. NewYork: John Wiley &. Son~, 1974.

Dougb~, R. J., &. Prlbram, K. H. Dlslracllon nnd hnbitu:ltion In monkeys with timbi~ Ie.sions. J. Co",,,. Physiol. Psych0/. , 1969,69,473-480.

Eccles, J. C., lto,M., &. Szentagotha, J. The cercbcllum as a nl'IIrollal machine. New York:Springer-Verlag, 1967.

rcslin~er, L., nmnh:lm. C. A., Ono. II .. &. n:tmber, D. Effcrence and the conscious ex­pcricncc of pcrl:cptioll. J. Ex". Psycho/., 1967, U, 1-36.

Garner, W. R. Unctrfointji &Structure os Psye/lologlcol Concrpts. New York: John Wiley,I.5J·161·1962.

Gibson, E. J. Improvemcnt in perceptunl judgments as a runclion of controllcd practice ortr:tininll· PJ)'clrol. Dull., 1953, .50, 401-431.

Gibson. J. J. Tlrr SrnJi'S Considrrl!d OJ PCfCl!f1tlla/ Systrms. noslon: lIoughlon Mirmn Co.1966, p. 355.

Gmnit, n. The na.liJ of Motor COlltrol. Nl'w York: Ac:tdemlc Prcss. 1970.

Granlt, R. The function:tl role of the muscle spindles-f:tcls :1I1d hypothcscs. D;ain. 1975.9,Y, P:lTtl V, .5) J,SS6.

Groves, r.M. &. Thompson, n. F. lI:lbltu:tllon: A dual-procl'ss theory. Psych. Rev.,1970-77,419·4.50.

Grucninger, W., & Prlbl'llm. K. II. The crfects of SP31i:11 :IIld n0l1sJ1:ltl:t1 distrnctors on per.formancc l:ttency or monkeys wilh rronl:tllc~ions. J. Co"'p. Plfyso/. Psycho/., 1969,68,20J-209.

Hm, E. II. Imprinting. Se/rnce, 1959,/JO, 1]).141.

lIorn, G., &. Hinde, R. A. (Eds.) SI,or(. Tr"" Changes In Nellral Artivlty and Drhavlor.C:tmhridgc: C:Jmbridcc Univ. rre~s, 1970.

Jame~, W. Prlnclple.r of 1'.lydloloKY, Vol. I, II, 1950,688. New York: Dover Publlc:llions,Inc.

"

,l;.t

. :

-,-- ... -.) •.

Modes 01 Cenlral Processi,,!: In Human Learni"!: ilnd Remembcrincl 1&1'

John, E. R. MtchanlSII/S 0/MClllary. New York: Academic Press, 1967.468.

Kimble, D. P•• nag,~haw, M. H., & Prlbram K. II. The GSR of monkeys during orknlingnnd habituation aher seleclive partial ablations of thc cinculatc and frontal COrlcx.

Ntllropsyr:11ol., 1965, J. 13 1·128.

J<inshe1:t. personal communication

Lashley. K. S. In search of the engr:tm. In: Society for experimental Diology (Great Dri·lain) Physiological Mtchan/sms In Anlmol nt!hovior, 454·482. New Yorlc: AcademicPress, 1950.

luria, A. R., Pribram, K. H., & Homsknya, E. D. An experimental analy~is of thebehavioral disturbance produced by a left frontal arachnoidal endothelioma (mcn.Ingioma) Nturopsychol., 1964,2,257·280.

MacKay, D. M. Cercbr:l\ organization :Ind the conscious contror of aclion. In J. C. Eccles(Ed.) Droin and Cansc/oliS Exptritnct!, 1966,422-445. Ncw York: Springer. Verlag.

Malis, L. I., Pribram, K. H., & Kruger, l. Aclion potentials In "molor" cortex evoketl byperipheral nervcstlmulation. J. Ntllrophys/o/.,19B./6, 161-167.

MCP:Hland, D. J. FI!f!dback'MtC'1lll/1isll/s ;n An;mal Dl!lmv;or. London: Academic rre~s,

1971,279.

Miller, G. A., Gnlanler, E. H.. & Pribram, K. H. Plern.r and tlrt SlrllCl/lrto/ Bt!/rov;or. NewYork: Henry Hold & Co., 1960.

Morrell, F. Visual sysh:m's view of acoustic space. Nalllrr (London) 1972. 138. 44·46.

Mounlcnslle, V. D., Lynch, J., Georgopollla~. A., Snbln. 11., .'::. Aeunn, C. Posteriorparielal association eorlex or the monltey: Cornmaml fllnetion5 for operations wilhincxtrnpersonalsp:lec. J. Nf'lIropl,)'siol. 1975, Vol. VIIJ. No.4. 871·90R.

Neis~cr. U. Congn/flvt Psychology. Ncw Vor!;: Appleton·Century Crofts. 1967.

Olds, J., Disterhoft, J., Scgal, M., Kornblith. C., & \-lir~h R:"Lenrning eenlers or ral brainmapped hy mensurlng latcncies of conditioned unit responses. J. Nf'urollysinl. Vol.'

' . .....\X. No.2, 1972,202·219..

rrcmaelt. D. Reinforcement Theory. In DavId Levinc (Ed.) Nf'llrosk(l .'i)'lI/p(lJilll/l 01/

Mol;vnfiOlr. Lincoln: Univ. or Nehr.ulta Pren. 123·1 SR, 1%5.

Prillrtlm, 11., & lInrry, J. further behavioral analy~is of lhe par1clo.lemporClprcnccipilalcorlex, J. Nellrophy'siol:, 1956, /9,99·\06.

Pribralll, K. 11. A review of lheory In physlologicnl psychology. In Annrcol R(!yiI!", ofPsychology, 1960, 1-40 .• Ptll0 Alto, Calif. Annunl Revicws, Inc.

Pribram, K. II. The amnesllc syndrollles: dislnrhanecs In c(lclin~7 In O. A. TallalHl antINnncy C. Wnugh (Eds.), I'oll,%/:y of Mrlllary. New York: ACOItlcmic I'ress. 1%9,

127·157.

_.....~._ .....-..~ ---" _... -_ ..._.--- -.-'-----_._.

II

I'[fiI'

! 'I:I,; ,\:

Ir1

:1t 9

1(,2 /Ilr.lin & lCJrlli,,~

!'rihraJll.' K. II: l.allCl/llges of the /lrail/: EXfll!rillll!l/tul Parado);l!! ancl Principles InN'·/Iropsyc!lology. c/lcli:wooLl CIlUs, N. J.: Prenlice-Il;tll, Inc., 1971.

I'rilJraJll, h. II. Th~ prilllJle fronlal corlex-executive or Ihe brain. In K. It Pribram andA. R. Luria (Eds.) PsydIOJiltys/%gy 0/ till! Frontal Lobl!J. 197j, 293·314.

Pribram, K. II. 1I0w is II Ih:ll ~ensinlt so much we can do so 1l1l1e11n Tlte NeurosciencesStl/dy ProgruIII. III, 1974a, 249-261. Cumbridl:e MllSS.: MIT J'ress.

I'riur;lIlI. K. II. Ne.... lIimemiolls In Ihe functions of Ihe basalcanglia.l'roaedi/lgs 0/ thl!Americal/ Psychopatltologirall'lSJociatlon, 1977.

1', iUIlIIII, K. H.• & U:lc~ha\V, r..1. Further analysis of Ihe Icmporallobe syndrome ulilizingfrolllo·lemporalablalions. J. COlliI'. Neural., 1953,99,347·375.

Prihralll. K. II., \Jay. R., &. Johnsloll, V. S. Selcclive allenlion: disllnclive brain eleclricalp:lllcrns produced b)' dirrcrcnlial reinforcement In monkey llnd mun, In D. I. Moslof.~ky (Ed.) /Jehavior COllrrol und Modi/kutioll ofPh)'siological Activit~, 1917.

Pribr:Il11, K. II., & Gill, M. 1\1. f;erlCf's "Project" Re·Asse.rsed. New York: Daslc Dooks,1976.

Pribralll, K. II. & Isaacson. It. L. Summary chapler In The J/ippocampus Vol. Il, 1976,ol2Y·4·11. R. L. lsa'H:~ulI and K. II. Pribr:lIl1.

!'ribram, K, II. & MclCes. F, T, Emotion: The search for conlrol. In P. J. Vinken and a.\V. Druyn (Eu>.) /Iumtbook 01 Clil/ical Nf!/Irologjo. 1969, 3 [6-342. Amslerdam, NorihHolland PublishinC Co.

Pribr:III\, K. H. &. McGulllll~SS, D. Arousal. aClivation and cHorl in Ihe COlli rot of allen­lion. PJ)'clr. Rev/elv, 1975,81(2), [16-149.

rribram. K, [I.. NUlVer. 1\1., &. Baron, R. The holographic hypotheses of memory slruclurein br:lin funclion :lI\d pefceplion. In R. C. Alkinson, D..... Kranlz, R. C. Luce and P.Suppes (Eds.). COl/le/llpory Del't!lo//II'el/ts ill Matlle/llutical I's)'cllolvgy. San Fran­cisco: W. H. Freeman & Co., 197-1.416-467.

Pribram, K. II., Spinelli. \J,' N., & Kamb:lck. M. C. ,Eleclrical correlalcs of stimulusfc,pUlISe and rcinforcclnenl. Scil!/Ice, 1967, 157: 94-96.

I'ribram, K. 1-1., & Tubbs. W. E. shOrl·lerm memory, paning and Ihe primale fronlal cor­Iell. Sri':I/ce, 1967./J6, 1765-1967.

Quilliall. M. R. Word cOllcepls: a Iheor)' simulation of some basic semanlic capabililics.IJLoI/Uv. Sci.• 1967. 12, 410-430.

Rueh. T. C. 1\I0ior syslems. In Slevens, S. S. (Ed.) lIalldbook vf Expcri1lll!IItalPJ)'clloloCY. New York: John Wiley & Sons, 1951, 154·201l.

Schlldder, \V, ,f" Shirrrin. n. M. CUlIlrulled and aulumalic Il\lIn~1\ inform:llion procc»inl::I. \Jell'Cliun. search alld allemion. I'sycll. Nev. 1917, Bot: 1·66.

ShcrrinCIOII. Charles, 71,,· lI/r('J:falivl! Acliol/ of tile Nervous System. 1947, 433, (FirslJlublishcd in 19(6) New Ilaven: Yale Univ.l'ress.

_ ......-~~_.-.-. :-.,.*_......._._... * •• ~..••---~•• _-*.._--------

:;.

".~

: \ ~

<~,- \

~ .i

~~'I.

Wilson. M. Infcrolcmporal cortex and the proccssin!: of visual informalion in monkeys.

Nl!uropsychol.• 1968.6,135·140.

Wallllch, H., & Avcrback. E. On mcmory modalities. Am. J. Psychof.• 19S5, 68. No.2.

249-257.

Ungcr}cider, l., Ganz, l., &. Prlbraln. K. H. the effects of inferolemporal and foveal.preslriate resccliom on visual conslancy In Ihe monkc:y. £Xp/f. Drain Res. 1971.

Modes of Central ProceHing in HUl11an Lcarnins and Rcmcmberinc/ 10]

Tuly!n!, E. Episodic lind sClllantic melllOry, In E. Tulvinll and W. Donaldson (Eds.)Organlza/ion ofMemory. New Yor\(: Academic Press. 1972,382-403.

•Uncersh:dt, U. Drain dopamine neurons and behavior. In Tile Neuroscie;lce.s Third Study. Program. Ill; 1974, 695.7~. Cambridge, Mass.: MIT Press.

Tulvinc, E., & Oold, C. Stimulus Infonn~lion and contclllual information as dctcrmlnantsof tachistoscopic recognition of words. III R. N. Habcr (Ed.) CUlllclllpory Tlreory "ndResearch In Visuul Percrp/ion. New York: Holt. Rinehart &. Winslon. 1968.

Thompson, R. F. The scarch for the engram. Amcrleoll Psychologisl, 1916.

Tcuber, I-I. L; Perception. In J. Field, H. W. Magaun. &. V. E. Hall (Eds.) Handbook ofPhysIology. Sec. J: Neuroplryslo/ogy, Vul. Ill, 1960, 1595·1668. WashinGton:

American Physiological Society.

S'tevens, S. S. Mathcmatics, measurcmcnt and psychophysics. In S. S. Slevcns (Ed.) Uanc/­book of Experlmenlal PsyrlJology, 19~I, Chpl. J, 31·32. New York: John Witey &

Sons. Inc. Aho london: Chapman and Iiall, lid.)

Spinelli, D. N., Slarr, A., ~ Darret. T. Auditory specificitY In unll recording from cal's

'ihuat cortex. Exp. Ncural. 1968,22,75-84.

Sperry, R. W. Neurology and lhe mlnd·braln problcm. American Scilmtist, 1931, olD, No.

2,291-)12.

ShUfrln, R. M. anti Schneider W. Conllolletlantlaulomatlc human Information process­ing: II. Perccplllalicarnins, automalic Illlcndinll, llnu a Ilcneraltheory- p"ycll. Re .....

1977,84: 127-190.