110-126 problem solution by monkeys following bilateral removal of the prefrontal areas. v. spatial...

7/28/2019 110-126 Problem Solution by Monkeys Following Bilateral Removal of the Prefrontal Areas. v. Spatial Delayed Re

1/17

PROBLEM SOLUTION BY MONKEYS FOLLOWING BILAT-ERAL REMOVAL OF THE P RE F RO N T A L A RE A S .V. SPATIAL DELAYED REACTIONSBY R. J. CAMPBELL AND H. F. HARLOW *

University of WisconsinI. INTRODUCTION

The use and importance of the delayed reaction tes t as a techniquefor investigation of the effect of large lesions of the prefrontal areason subhuman primates was first emphasized by the work of Jacobsenand his co-workers (21, 22, 25). These investigators found thatmonkeys and chimpanzees failed to solve spatial delayed reactionstested by a direct method after bilateral removal of the prefrontalareas. They interpreted this deficit as a result of a loss of thecapacity for ' imm ediate memory.' This type of behavioral adapta-tion appeared to be specifically dependent upon the prefrontal areas,since extensive bilateral lesions to the parietal, temporal, or motorand premotor areas occasioned no loss.Jacobsen and Nissen (24) subsequently reported that the delayedalternation habit was also abolished by bilateral ablation of thefrontal association areas . These results were 'in harmony with theeffects of frontal injury on performance of delayed response.'Finan (2) demonstrated that bilateral prefrontal monkeys couldsolve temporal d iscrimination problems, a shuttle box and a temporaldiscrimination maze. A ttemp ting to relate these data to those ofJacobsen, he assumed that "temporal organization . . . can not beconsidered the mechanism of immediate memory."In a second experiment (3), Finan found, as had Jacobsen, acomplete loss of delayed response in two monkeys following bilateralablation of the frontal association areas. Upon introducing pre-delayreinforcement, in which before each trial the animal received a rewardin the food cup subsequently correct, the same two Ss attained alevel of correct response on 15-sec. delays significantly better thanchance. These results led Finan to the assumption that the primarycontribution of the prefrontal areas was tha t of a 'derived overtreinforcing agency' and not the retention of imm ediate experiences.The problem of the role of the prefrontal areas was furthercomplicated by the finding of Malmo (29) that monkeys could solve

* This research was supported in part by a grant from the Special Research Fund of theUniversity of Wisconsin for 1943-44.110


2/17

PROBLEM SOLUTION BY MONK EYS. V 11 1spatial delayed reactions run by an indirect technique offeringmaximal opportunity for the formation of positive after-images.Successful response was inversely related to the degree of illuminationof the cage during the delay interval. The hypothesis was advancedthat bilateral frontal lobectomy renders the animal more susceptibleto 'retroactive inhibition.'

I I . PURPOSEThe purpose of the present investigation was to study the effectof the bilateral ablation of the prefrontal areas on the performanceof rhesus monkeys on spatial delayed reactions using a direct methodof testing . To appraise the effect of emotional disturbances, graded,stepwise procedures were used throughout. In order to determinethe time interval necessary for recovery, the postoperative periodbefore the initiation of the present tests ranged in the various Ssfrom slightly less than three months to over three years.

I I I . SUBJECTSSix operated and six normal rhesus monkeys were used as Ss in the present experiment.

The previous test history of these animals is indicated in Table I, in which the numbers in theparentheses refer to the references given at the end of the paper.TABLE la

PREOPERATIVE AND POSTOPERATIVE TEST EXPERIENCE OF THE OPERATED SSAnimalnumber

54$5652 257

Animalnumber737475762 156

(7),(7),(6),(6),

(8 )(8)(8)(8)

Preoper

, (14 ) ,, (14),, (35),, ( 9 ) ,

ative

. (33). (33), (38). (35)

TEST

( I O) ,(10),(10),(10),( 6 ) ,( 6 ) ,

, (39)

1 'est experience( 1 2 ) ,( 1 2 ) ,( 1 4 )(14)(14)(14)

TABLE ItEXPERIENCE

( I I ) ,( I I ) ,( I I ) ,( I I ) ,( 8),( 8 ) ,

(13),(13),Postoperative

(15), (16), (17),(15), (16), (17),

OF T H E NOR MAL S STest experience

(18)(18)(18)(18)( 1 8 ) ,( 9 ) , (35),(18), (38),, (35),. (39)(38), (39)

(19),(19), (34)(34)

IV . APPARATUSThe general test situation used in this investigation, save for the fact that a mirror wasplaced at the side to make possible continuous observation of the animal, is the same as thatdescribed by Spaet and Harlow (34). Th e stimulus tray was the discrimination tray commonlyused in earlier investigations and had two food wells set 12 in. apart, center to center. Theobjects used to cover the food wells during the delay intervals were two unpainted 1 }4 in. woodencubes. The essential characteristics of the test situation, the stimulus tray and the stimulus

objects are illustrated in Fig. 1.


3/17

112 R . J . C A MP B ELL A N D H . F . H A R LO W

la . Food is shown to mo nkey pre pa rato ry to baiting correct food well.

ib . Monkey displaces correct object after delay interval and obtains reward.FIG. I. Solution of spatial delayed response by rhesus monkey 54


4/17

PROBLEM SOLUTION BY MONKEYS. V 113V. METHOD

Operative proceduresAll operations were carried out in a single stage using full aseptic precautions.1 The extentof the lesions is indicated in Table II, the numbers of the prefrontal areas being those describedby W alker (36).

TABLE IIOPERATIVE DAT A CONCERNING THE SS *

Ani-malnum-ber

545564652 257

8b ,8b,8b ,8b,8b,8b ,

9,9,9 .9 ,9,9,

1 0 ,1 0 ,1 0 ,1 0 ,1 0 ,1 0 ,

TotalI I ,I I ,1 1 ,II, 12,II, 12,II, 12,

464646464613, 46

Extent of lesion

Severe12 , 2512 , 25' 2 . 2513. 25,13 . 25,14 , 25,

454545

8a,8a,8a,8a,8a,8a,

Partial

J J3 > 14.14,

2 42424

454545

Gm stissueexcisedL

2 .92.83.03-53-24 - i

R2-73.03.I3-53-53.8

M o.post-oper-re -covery

393964-534* Drawings showing the exact amount of tissue excised in every animal are on record andwill be sent upon request.

General test proceduresThe delayed reaction problem was divided into three parts: (1) the preliminary training,(2) a delayed response series, in which increasingly long delay periods were introduced usingstepwise procedures, and (3) a zero-sec, interspersed delayed response series, similar to thedelayed response series bu t with zero-sec, delays interspersed among the longer delayed responses.8A non-correction method (20) was used throughout and an inter-response food reward (31)was administered throughout except in case of failure on the zero-sec, trials.For animals 54, 57, 64 and 65 and for all normal animals the baiting of the correct stimulusposition was accomplished by showing a piece of food to the animal, dropping the food in the

correct food well and simultaneously placing the stimulus-objects over the two food wells.In the case of subjects 55 and 22, a stimulus-object was placed over the incorrect food welland the food reward shown to the monkey at the same time . The food was then dropped intothe other food well both hands were withdrawn and both hands brought forward while placingthe remaining stimulus-objects.During the trials in which no screen was utilized the E held his hands over the stimulus-objects during the delay interval.Preliminary training trials

Ten zero-sec, delayed reaction trials were run to the criterion of 9 correct out of 10 trialsfor two successive days, with the added provision that not less than five days of such trainingwere to be given (Step PT -i ).

1 Th e operations on animals 54, 55 and 64 were carried ou t by H . F. Harlow and the operationson animals 65, 22 and 57 were carried ou t by Paul H. Settlage.s This procedure was planned on the basis of personal communication with Dr. P. H. Settlage,who found that normal rhesus monkeys trained to solve spatial delayed reactions of long (45-60sec.) duration often showed severe loss on, or complete disruption of, delayed response of anyduration after repeated failure. The procedure employed was designed to reduce the effectsof emotional disturbances arising from repeated failure at any point of the series upon the mon-key's actual ability to solve problems involving delay.


5/17

114 R. J. CAMPBELL AND H. F. HARLOWSolution of this test was followed by 10 trials a day of alternate zero- and 5-sec. delayedreactions, again to the criterion of 9 correct responses a day for tw o successive days (Step PT -2) .An additional behavioral criterion was enforced in that the monkey was allowed circlingbehavior during the delay interval on not more than two trials of any day's run.* Whenevercircling appeared the animal was not pe rmitted to make a choice; instead, the tria l was rerun.

The delayed response seriesTw enty trials a day were run throughout the delayed response series. The criterion forsolution of any step was 17 correct responses out of 20 trials for each of two successive days,a level of response significantly bette r tha n chance a t the one percent confidence level. Theanimal was allowed four days to a ttain this level of success. Th e additional behavioral criterionused in the preliminary tra ining trials also was enforced. Circling responses appeared onlyduring the early part of this response series.The following Tests and Steps and their essential characteristics are given below:Test I : (D R-I) Neither the opaque screen nor the one-way-vision screen wereutilized.zero-sec, delays alternating with 5-sec. delays5-sec. delays5-sec. delays alternating with lo-sec. delaysThe opaque screen was interposed between the animal and thetest situation on all save the zero-sec, delay trials,zero-sec, delays alternating with 5-sec. delays5-sec. delays5-sec. delays alternating with 10-sec. delaysThe opaque screen was interposed between the animal and the

tes t situation on all save the zero-sec, delay trials. The one-way-vision screen was interposed between the test situation and the Eon all save the zero-sec, delay trials,zero-sec, delays altern ated with 5-sec. delays5-sec. delays5-sec. delays a lterna ted with 10-sec. delays10-sec. delays10-sec. delays alternated with 15-sec. delays15-sec. delays15-sec. delays alternated with 20-sec. delays20-sec. delays20-sec. delays a lternated with 30-sec. delays30-sec. delays30-sec. delays alternated with 45-sec. delays45-sec. delays45-sec. delays alternated with 60-sec. delays60-sec. delays

Regression following failure was always instituted and was thought to be an essential partof the test. In th e case of failure on any step of any test, the animal regressed one step withinthe same tes t (a par tial regression). If the monkey passed this step, he again advanced to thestep above; if the animal again failed he regressed to that step of Test I (DR-I) correspondingin length of delay interval to the step just failed or to DR-Ic if Test I offered no correspondingdelay interval. In th e case of further failure the S regressed through all the steps of DR-I.This was called a complete regression. If DR-Ia was failed, the animal was given preliminarytraining (zero-sec, delay), differing from the original preliminary training in that 20 trials a daywere run. This step was labeled PT-3 . If at any point during the regression steps the animalmet the criteria, he progressed forward in the stepwise order already outlined.As soon as any monkey solved te st D R -II In he was tested on a second seriesthe ZS series.All Ss who failed to solve any DR-III step were run on the ZS series after they had made two

Step IStep IIStep IIITest II:Step IStep IIStep IIITest III:

Step IStep IIStep IIIStep IVStepVStep VIStep VIIStep VII IStep IXStepXStep XIStep XIIStep XIIIStep XIV

(DR-Ia)(DR-Ib)(DR-Ic)(DR-II)(DR-IIa)(DR-IIb)(DR-IIc)(DR-III)

(DR-IIIa)(DR-IIIb)(DR-IIIc)(DR-IIId)(DR-VIIe)(DR-IIIf)(DR-IIIg)(DR-IIIh)(DR-IIIi)(DR-IIIj)(DR-IIIk)(DR-IIII)(DR-IIIm)(DR-IIIn)

1 Circling is not to be confused with pacing or mere hyperactivity.or from side to side did not constitute circling behavior. Pacing back and forth


6/17

PROBLEM SOLUTION BY MONKEYS. V 115Complete Regressions (regressed from any step of Test DR-II or DR-III to DR-Ia or PT-3) orfive Partial Regressions.

The Zero-Sec. Interspersion Series (ZS-series)In the ZS series 30 trials a day were presented, 20 of which were always zero-sec, delays and10 of which increased in length with each step in the series as follows: 5 sec, 10 sec, 15 sec,

20 sec, 30 sec, 45 sec, and 60 sec.The interspersion of the zero-sec delay trials followed a predetermined irregular order.

Using two basic zero-sec presentation series 2-3-3-2 and 1-4-4-1, four presentation series werepossible (2-3-3-2-2-3-3-2; 1-4-4-1-1-4-4-1; 2-3-3-2-1-4-4-1; and 1-4-4-1-2-3-3-2). The 10 longertrials were then placed after each zero-sec, trial group in such a manner that the daily seriesalways began with a zero-sec, trial rather than with a long delay. Since there were eight zero-sec,trial groups, there were also eight positions in which the 10 longer delay trials could be placed;consequently, at least two of these longer delay positions contained two longer delay trials.Of the many possible combinations, the following 15 were utilized, and in the following order:

a) 2* 3 ** 3 * 2 * 2 * 3 * 3 * * 2 *b) 1 * 4 ** 4 ** 4 ** 1 * 1 * 4 ** 4 1c) 2 * 3 * 3 * 2 * 1 * 4 * * 4** 1*d) 1 * 4 ** 4 * * 1 * 2 * 3 * 3 * 2 *e) 2 * 3 * * 3 * * 2 * 2 * 3 * 3 * * 2/ ) 1 * 4 ** 4 * 1 * 1 * 4 ** 4 ** 1g) 2 * 3 * 3 * * 2 * 1 * 4 * * 4 * 1 *h) 1 * 4 ** 4 ** 1 * 2 * 3 * 3 ** 2j) 1 * 4 * 4 ** 1 * 1 * 4 ** 4 * 1 *j) 2 * 3 ** 3 * 2 * 1 * 4 ** 4 ** 1ir) 1 * 4 * * 4 * 1 * 2 * 3 * 3 * * 2 */) 2 * 3 * * 3 * 2 * 2 * 3 * * 3 * * 2

m) 1 * 4 * 4 * 1 * 1 * 4 * * 4 * * 1 *n) 2 * 3 ** 3 ** 2 * 1 * 4 ** 4 * 10) I * 4 * 4 ** 1 * 2 * 3 ** 3 ** 2

(Note.Numbers here refer to the number of zero-sec, trials presented; each asteriskrepresents one longer delay trial.)

The criterion for the solution of any step of the ZS-series was 17 correct responses out ofany consecutive 20 of the longer delay trials (a two-days' run).

Following is the step wise procedure for the ZS-series:Test I: (ZS-I) Neither the opaque screen nor the one-way-vision screen was utilized.

Step I (ZS-Ia) zero-sec delays and 5-sec. delaysStep II (ZS-Ib) zero-sec, delays and 10-sec delays

Test II: (ZS-II) The opaque screen was interposed between the animal and the testsituation on all save the zero-sec, delay trials.

Step I (ZS-IIa) zero-sec, delays and 5-sec. delaysStep II (ZS-IIb) zero-sec delays and 10-sec delays

Test III: (ZS-III) The opaque screen was interposed between the animal and the testsituation on all save the zero-sec delay trials. The one-way-visionscreen was interposed between the test situation and the E on allsave the zero-sec, delay trials.

Step I (ZS-IIIa) zero-sec, delays and 5-sec. delaysStep II (ZS-IIIb) zero-sec, delays and 10-sec. delaysStep III (ZS-IIIc) zero-sec, delays and 15-sec. delaysStep IV (ZS-IId) zero-sec, delays and 20-sec. delaysStep V (ZS-IIIe) zero-sec, delays and 30-sec. delaysStep VI (ZS-IIIf) zero-sec, delays and 45-sec. delaysStep VII (ZS-IIIg) zero-sec, delays and 60-sec. delaysAs with the DR-series, regression was always instituted after failure in any step, and the

same procedure of regression was used. If Step ZS-Ia was failed, regression was to a step con-sisting of 30 trials of zero-sec, delays.


7/17

116 R. J. CAMPBELL AND H. F. HARLOWV I . R E S U L T S

A . A daptation to the delayed reaction test situationIn Table III and in Fig. 2 are presented data indicating thedegree of success attained by the normal and the operated Ss in

T A B L E I I IH I G H E S T PE R C E N T A G E O F C O R R E C T R E S P O N S E S M A D E I N A N Y O N E D A Y O N T H E M A X I M U M

D E L A Y I N T ER V A L O F S T E P P T - I A N D E A C H O F T H E S T E P S O F S E R I E S D R - I

Operated animals

Number

5455646s2257M

StepsPT-I*

IO OIO OIO OIO OIO O9098

DR-IatIO OIO OIO O90806088

DR-IbtIO OIO O7595555079

DR-IcIO O95506055068

Normal animals

Number

7374s5621M

StepsPT-I*

IO OIO OIO OIO OIO OIO OIO O

DR-IatIO O9O95

IO OIO OIO O97-5

DR-IbtIO OIO OIO OIO OIO O9599

DR-Ic95

IO OIO O95IO O8596

Maximum delay interval of the var ious s teps* = zero sec.t = 5 sec.% = s s = c - = 10 sec.In those cases in which a monkey failed to reach a particular step an arbitrary score of50 percent was given.adap ting to the delayed reaction test situation. These are datataken from Step PT-i and the Dr-I series, maximum delay intervalsbeing 10 sec. Since no interfering screens were used in these tests,an extremely simple and direct test situation is presented to theanimal. Th e data show each monkey's best day's score for themaximum delay interval for Step PT-i and each of the steps of theDR-I series.All of the normal Ss adapted rapidly and successfully to thedelayed reaction situa tion . There is little difference between th escores of the worst and the best S, and these differences are notindicative of later performance. Only two of the operated anim als,54 and 55, successfully adapted to the delayed reaction tests. Thedifferences between the scores of these Ss and the worst Ss, 22 and 57,is a difference between complete success and almost complete failure.These data show th a t monkeys with extensive lesions of the prefrontalassociation areas frequently fail to adapt to the delayed reaction


8/17

P R O B L E M S O L U TI O N B Y M O N K E Y S . V 117te st situa tion. This failure is no t to be confused w ith a mere reduc-tion of any mythical maximal delay interval.

B. Ma ximum length of delay attained .A comparison of the maximal length of delay achieved by thenormal and the operated animals is given in Fig. 3. The tota lnumber of trials taken and the percentage of correct responses madeon each step of all tests is presented in Table IV . The da ta indicate

IMA

O2 ^S 70KOU 6 0

5 *L > _K FUi0-OELAYSTEP

PREFRONTALMONKEYS

\ \\ X

V -0" 5- S" KCPT-I PR -I. D-Ib DK-K

MAXIMUM L E N G T H OF

NORMALMONKEYS

> 3 * 5 ' 10-PT-I H - I Nt-Ik CHWt

DELAY

L E G E N D - B E S T S UB JE CTO O - AV E RA GE S C O R E - 6 - S U B J E C T S0 0 - W O R ST S U B JE C TT F A I L E D

F I G . 2. H ighe st percentage of correct responses mad e in any one day on the maxim um delayinterval of step PT-I and each of the steps of series DR-I

the difficulty that the normal Ss met in the solution of delayedreactions even of moderate length, particularly when the opaqueand the one-way vision screens were used. No attem pt was made totest any 'limit of ability' of the rhesus monkey in delayed reactionperformance. These da ta indicate, however, th a t normal rhesusmonkeys do not easily solve delayed reactions of moderate lengthwhen tested under rigid and predetermined test conditions, in testsituations of the type described.


9/17

118 R. J. CAMPBELL AND H. F. HARLOWTwo of the prefrontal Ss, 54 and 55, made scores comparable tothose of the best of the normal Ss. Only one other operated anim al,65, gave any evidence of mastering the delayed reaction problem.This S solved 10-sec. delayed reactions (Step ZS-IB), but failedeither to go beyond this point or even maintain this score in latertests. Subjects 22 and 57 encountered difficulty in delayed reactionsof any appreciable length.4

P T * N DR SERIESID C ZA ZB ZC iA 3B 3C 5' 3" 10* 5" ST ' 5' 5' 10' 10"

Z S SERIESO" S

IA. IB5 " K>"

Is8T

1fti{\

? *s iSos 8&s s\

M$ &$ m1

ZA5'

K J O (

i1sI f

ZBt o r

iin

3A5"

136 X 35 5t Jf K10' W '"' 60'ii

I1I 11Mmirm \ \ \

L E G E N D . P A S S Q]] F A I L E DSECONDS MAXIMUM DELAY INTERVAL OF PARTICULAR TEST

FIG. 3. Maximum length of delay by normal and prefrontal monkeysAll normal Ss solved io-sec. delayed reactions (Step DR-IC andStep ZS-IB). The two monkeys whose scores were the poorest(73 > 74) were given tests on an extended ZS-I Test at the conclusionof the regular experiment. Both Ss solved 30-sec. delayed reactionsin these supplementary tests.

C. Influence of lesion sizeThe two Ss of the prefrontal group to achieve the longest delays,54 and 55, were the animals with the smallest lesions as indicated byweight of the excised tissues and observation of the excised tissues.Subjects 22 and 57, with very large lesions (including extensive4 Subjects 64, 65, 22 and 57 had solved 15-sec. delayed reactions in brief preliminary pre-operative tests.


10/17

TABLE IVTOTAL NUM BER OF DELAYED REACTION TRIALS AND THE PERCENTAGE OF CORRECT RESPONSES

TABLE IVaTotal Number of Trials

Subj.No .541 552 6S

O 225721

13 56g 75 76Z 7374

PT-io"ICO155IOOICO702456o7520308o6o

PT-2s"35403540IO751 0152 02 02O1 0

iAlB5"218319260150no3019017516095300250

i C1 0"

58402 0

I O1 060202 070

2A2BS"180287

1208037043090180

2C10 "4O30

3206030

3A3B66

8080

1 00

3C3D10 "

86

606020

iA5"30405043560So2 02020604040

IB1 0"

20802 040

2O2020IOO9060

2A5"2 07020

2 O206030

1 2050

2 B10 "

2 040

2O1 0802050

3As"2O

IOO

60204070

3B1 0"

2O80

40204070

3CI S"

2 0l6o

40540

3D20"2 030

408030

3E30"

2050

702020

3F45"20

6020

3G60"18

40

TABLE IVbPercentage of Correct Responses

Subj.No .543 552 6Sa 6+O 225721

3 56g 75S 76* 7374

PT-i0"658683728372

1 0077IOO979587

PT-35"698069758059

10080IOOIOO9090

1A1B5"988777656263888495938290

rC1 0"

989860

70IOO97859594

2A2Bs"8787

899980746758

2C1 0"

8567

801 007270

3A3B5"86

IOO9584

3C3D1 0"

84

I OO9365

lA5"87937269537095851 00839398

IB1 0"

95949050

1 0090IOO

IOO9697

2A5"958440

1 00I OO88767158

2B1 0"

1 0085

90IOO758058

3A

9587

871 009085

3B

9584

931 007876

3C

9093

90808263

3D

1 0080

906977

3E

8556

736S85

3F90

70So

3G60"56

58

Seconds (o" , 5", etc.) = Maximum delay interval for each step.The total number of trials and the percentage of correct responses are given only for the maximum delay interval.


11/17

12 0 R. J. CAMPBELL AND H . F . HARLOWdamage to areas 13, 14 and 45), failed completely on the ZS seriesand did little be tter on the D R series. Subjects 64 and 65 madeintermediate scores, but more closely approximated the worst thanthe best.In discussing the effect of lesion size it should be emphasizedthat the lesions were large in the case of all six Ss, and that in allanimals the greater portion of the prefrontal association areas wereremoved. Fu rthermore , the description of the lesion size was basedon observation of tissue cleanly excised, and the actual damage wasdoubtless somewhat greater than that described.On the basis of these results it can be stated that the findingattributed to Jacobsen by Fulton (4, p. 426), that areas 9a and 10are the areas whose destruction produced the deficit in delayedreactions, needs considerable qualification.

D. Influence of duration of postoperative recoveryThe monkeys with the longest time interval between operationsand testing, 54 and 55, achieved a level of success definitely superiorto th a t of the other Ss with bilateral prefrontal lobectomy. T he

time lapse here was over three years. All of the other Ss were testedwithin six months of operation and the tests were completed withina year. Of these Ss the one with the greates t time interva l sinceoperation, 64, was consistently inferior to subject 65; a six monthsas opposed to a four and a half months postoperative period wasinvolved.Other investigators (3, 22, 29) have commonly begun testingwithin two weeks to two months of operation and have completedtheir experiments within four to eight months of opera tion. Thu s,we suspect that in many cases the brains of the monkeys werehardened in formaldehyde before they had fully recovered from theeffects of th e operation. Our observations indicate th at recoveryfrom the effects of ablation of the frontal association is slow, andthese observations are supported by the data of operations on man.Our own data are at present incomplete in that the monkeyswith the smallest lesions are also those with the longest postoperativeperiod of recovery.

E. Behavioral changesGeneral postoperative behavioral changes.All the monkeys withbilateral destruction of the prefrontal association areas showedmarked hyperactivity, consisting of circling, pacing, and over-response to extraneous stimulation (see 26, 28). This behavior wasparticu larly exaggerated in subjects 22 and 57. If these animals


12/17

PROBLEM SOLUTION BY MO NKEY S. V 12 1were even casually observed by an experimenter they would fly intoviolent and prolonged outbursts of random activ ity. For over amonth it was thought that neither monkey could possibly ever betested . Their behavior was similar to th a t described by Ruch andSchenkin (32) for monkeys following excision of areas 13 (an areawithout question very badly damaged in their cases) and was moreuncontrolled than that of any other prefrontal monkey subjected topsychological tests . Bo th animals gradually recovered and adaptedsurprisingly well to the test cage. Subject 57, a year after opera tion,gave vent to these violent and uncontrolled outbursts of randomactivity when excited.

Subject 22 had left homonymous hemianopsia for about 10 days'duration after operation. Subject 57 gave evidence of a type ofbehavior described by Kliiver (27) for a monkey following destructionof the prefrontal areas. The animal would accept food as rapidly asproffered, often dropping already obtained handfuls to accept more.Subject 65 developed an exaggerated to-be-groomed-and-petted re-sponse (see Kennard, Spencer and Founta in, 26). As soon as shewas touched and stroked her restless behavior ceased, and she wouldsit perfectly still with head drawn back and eyes closed for consider-able periods of tim e. The stitches were removed from her head fivedays after operation while she sat quietly in a corner of her cagewithout being restrained in any manner whatsoever! This monkeyresponded positively and apparently pleasurably to the removal ofthe stitches. These exaggerated to-be-groomed responses persistedfor abou t 10 days and then faded. Transien t evidence of similarresponsiveness appeared in subject 57 in the first week after operationbefore its wild flights of unrestrained behavior fully developed.Behavioral changes in the test situation. The restlessness, pacing ,and circling noted as general behavioral changes following bilateralprefrontal lobectomy were also apparent in the test situation, wherethe pacing was often (especially 22 and 57) so compulsive andviolent as to cause complete inatten tion by the animal to th e stimulus-objects. Th e ina tten tion of the prefrontal animalsa conditionreported in human clinical studies (1, 5, 30)was apparen tly a resultnot only of this pacing and circling activity, but of a generallylowered threshold of excitation to all external and internal stimulias well.The experimental group as a whole was further distinguished bya relatively quick loss of motivation in the test situation, with thepossible exception of subject 22. Since this loss of motivation in allcases was preceded by periods of repeated failure, th is was taken as adefinite reaction to frustra tion and failure. None of the four subjects


13/17

12 2 R. J . CAMPBELL AND H . F . HARLOW(21, 22, 56, and 57) which had been given preliminary preoperativedelayed reactions had refused to test on the preoperative test trials.6All of the prefrontal animals showed refusals to work (a fact sub-stantiated also on many other tests) in opposition to reports ofcom plete indifference to failure now in the literatu re (24, 37). Fin an(3) has mentioned resistance to work on the part of his monkeys.The refusal to work in itself differentiated the operated from normalanimals, for the pattern of behavior was different from that commonlyfound in the normal monkey after repeated failure. The norm alanimal typically retreats to the rear of the testing cage and remainsindifferent to the test situation or gives vent to aggressive behavior.Frustration on the part of the prefrontal animals (especially 57 and65) was accompanied by violent outbursts of activity so intense asto preclude any possibility of tes ting. The monkey would circlerapidly about the cage, jump up and down, and pace back and forth;under such conditions the animal could not respond to the originalexhibition of the stimulus objects. If the trial was run, nevertheless,the Ss chose wildly (if he temporarily became quiet enough to chooseat all), and often removed the food from the well only to drop itindifferently on the floor of th e testing cage. I t was noted th a t insuch circumstances the monkey would frequently refuse to acceptfree food, a condition not found in the frustrated normal animal.These refusals or failures to work appeared in the operatedmonkeys after some semblance of adaptation to the delayed reactionproblem had been accomplishednot ordinarily on the zero-sec,delayed reactions.

DISCUSSIONComparison of related studiesThe results of our experiment, combined with the work of otherinvestigators, make it appear that the spatial delayed reaction testedby the direct method is drastically affected by bilateral removal ofthe prefrontal association areas. Such problems have never beensolved by a monkey within a year of operation without specialsubsidiary training fjsee Jacobsen (21, 22), Finan (3), and the presentresults].

The data of the present experiment indicate that after prolongedperiods of recovery and other postoperative test experience therhesus monkey may solve spatial delayed reactions tested by the* Two monkeys, 54 and 55, were trained to do spatial delayed reactions using a combinedmatching-non-matching from sample method. Th e technique was th at described by Spaet andHarlow (34) save that on the non-matching trials there was no food under the sample object.In spite of the fact that there was no opportunity for the monkeys to benefit by any pre-delayreinforcement on the non-matching trials the prefrontal monkeys solved io-sec. delayed reactionsand made scores superior to two normal control animals, 51 and S3, selected as the best test

animals in the laboratory.


14/17

PROBLEM SOLUTION BY MON KEYS. V 123direct m ethod. Supporting evidence is to be found in th e report byFinan (3, p . 208).It appears likely that less severe loss is occasioned if the testsare carried out using some indirect method of testing the spatialdelayed reaction see Malmo (29), Spaet and Harlow (34), and thispaper, footnote 5].There are some data to indicate that the severity of the deficitmay be related to the amount of tissue destroyed in the frontal lobes.Jacobsen (21) has indicated that partial destruction of the prefrontalareas occasions a deficit in delayed reaction performance less severethan that resulting from total destruction of these cortical regions.Our own data indicate that the severity and duration of deficit maybe related to the am ount of tissue destroyed . W hether these resultsarise from intellectual loss or from other and secondary behavioraldisturbances has not been demonstrated.The problem of the effects of frustration upon the monkeydeprived the prefrontal association areas remains unsolved. Differ-ences in the literature may result from differences in the problemstested and th e procedures used in testing these problems, from pretestdifferences in the tameness of the animals and the degree to whichthey had been adapted to the test situation, from differences in theperformance of the animals on the test problems after operation,and from differences in the post-operative recovery period. Ourown data indicate that thoroughly tamed rhesus monkeys followingrecovery from bilateral prefrontal lobectomy give as much or moreevidence of frustrationas indicated by refusal to continue work ona problem as do normal inta ct monkeys. Jacobsen and Nissen(24) and Warden, Barrera and Gait (37) have independentlydescribed marked reduction in the severity of frustration responsesfollowing bilateral prefrontal lobectomy.

Nature of the intellectual deficitThere is considerable evidence to suggest that failure on the partof monkeys with large lesions in the prefrontal areas to adapt readilyto the delayed reaction situation is in large part a failure to attend

effectively to the presentation of food under the stimulus correct onany one tria l. Thus, the Ss fail to fixate and differentiate the cuesessential for problem solution. As a result of this inab ility to a ttendand fixate effectively the monkey either fails to learn or fails toestablish an association strong enough to override the interferingeffects of the responses made to earlier trials or to the effects ofextraneous stimuli. Any factor which facilitates attention andfixation of the position of the correct stimulus will facilitate delayedreaction solution.


15/17

12 4 R. J. CAMPBELL AND H . F. HARLOWA pre-delay reinforcing trial obviously accomplishes this, andFinan (3) has found that it materially aids in the solution of delayedresponses by frontally ablated monkeys. However, pre-delay rein-

forcement is not necessary for such solution. Delayed reactions runby indirect methods are also solved by monkeys following extensivefrontal lesions. One may reasonably assume th a t the earlier pro-longed pre-delay discrimination training heightens the attentivevalue of the stim ulus. Th e results of M almo (29) and Spaet andHarlow (34) are in keeping with this assum ption. Unpublished databy Harlow (this paper, footnote 5) also support this theo ry. Anyonewho has trained normal rhesus monkeys to solve delayed reactionsusing a direct method will recognize the difficulty of getting themonkeys to learn to attend to the presented food, and this difficultyis greatly exaggerated following bilateral removal of the prefrontalareas, particularly after the S has made errors in any day's testing.The observation that the excited prefrontal rhesus monkey may noteven attend to proffered food suggests that the attentive functionsuffers widespread im pairm ent. Similar loss appeared in the tes tsof initiation of behavior reported by Harlow and Johnson (17).

It must be admitted that 'loss of attention' cannot be defined inany fully satisfactory m ann er. The monkeys have difficulty inorienting toward, and maintaining this orientation to, a limitedportion of the te st situation. 'A tten tion ' is no t lost bu t is impaired.The animal is easily distracted at the time that the correct cup isbeing baited . A monkey may respond in a satisfactory m anner forseveral trials and then completely fail to attend, particularly aftererror. T he macaque does be tter if any indirect method is used th anit a direct method is used. But even with the former, the a ttentiv eresponses of the S are far from stable, particularly in the recoveryperiod lasting for at least several months.In essence, it is our belief that the difficulty that monkeys havein solving spatial delayed reactions results more from failure toestablish firmly an association between the position of the correctstimulus and the implicit food reward, and to differentiate thesecues from the rest of the test situation, than it is a fading of anymnem onic trace or loss of mem ory. Our position is similar to t h a tof Finan (3) and is in no essential position at variance with anydetailed analysis made by Jacobsen.Bilateral destruction of the prefrontal areas in monkeys producesdeficits resulting in impaired performance on many different kindsof tests . W hethe r this deficit is th e result of loss on a single inte l-lectual function or many has not been clearly demonstrated and isbeyond the province of this paper.


16/17

PROBLEM SOLUTION BY MONKEYS. V 125VII. S UMMAR Y AND C O N C L U S I O N S

1 . Twelve rhesus monkeys , six normal an imals and six whichhad been subjected to bi la tera l prefronta l lobectomy, were tes ted onspat ia l delayed react ions , us ing a d i rec t me thod .2. Two of the operated monkeys solved 20-sec . (or longer)de layed reac t ions , per formances comparab le to the bes t of the n o r m a lan ima l s . The other four opera ted monkeys adapted poor ly to thedelayed reac t ion tes t s .3. The per formance of monkeys fol lowing extensive bi la tera llesions of the prefronta l associa t ion areas appears to be influenced bythe length of the postoperat ive recovery per iod, previous tes t h is tory,and ex ten t of the lesion within the prefronta l areas .4. It is suggested that monkeys, fol lowing destruct ion of theprefrontal association areas, suffer deficit in a t t e n d i n g and fixatingl imited aspects of a t es t s i tua t ion , and tha t th i s opera tes to reducethe effectiveness of the acquis i t ion of complex problems such as thedelayed reac t ion run by the d i rec t me thod .

(Manuscript received October 10, 1944)REFERENCES

1. BRICKNER, R. M. The intellectual functions of the frontal lobes. New York: The MacmillanCo., 1936, rvi + 354 pp.2. FINAN, J. L. Effects of frontal lobes on temporally organized behavior in monkeys. / .Neurophysiol., 1939, 2, 208-226.3. FINAN, J. L. Delayed response with pre-delay reenforcement in monkeys after the removalof the frontal lobes. Amer. J. Psychol., 1942, 55, 202-214.4. FULTON, J. F. Physiology of the nervous system, Ch. XII. Cerebral cortex: The frontalassociation areas, pp. 418-428. New York: Oxford Univ. Press, 1943, x + 614 pp.5. GOLDSTEIN, K. The modifications of behavior consequent to cerebral lesions. Psychiat.Quart., 1936, 10, 586-610.6. HARLOW, H. F. Studies in discrimination learning by monkeys. I. The learning of dis-crimination series and the reversal of discrimination series. / . gen. Psychol., 1944,30, 3-12.7. HARLOW, H. F. Studies in discrimination learning by monkeys. II. Discriminationlearning without primary reenforcement. / . gen. Psychol., 1944, 30, 13-21.8. HARLOW, H. F. Studies in discrimination learning by monkeys. III. Factors influencingthe facility of solution of discrimination problems by rhesus monkeys. / . gen. Psychol.(In press)9. HARLOW, H. F. Studies in discrimination learning by monkeys. IV. Relative difficultyof discriminations between stimulus-objects and between comparable patterns with

homogeneous and with heterogeneous backgrounds. / . gen. Psychol. (In press)10. HARLOW, H. F. Studies in discrimination learning by monkeys. V. Initial performance byexperimentally naive monkeys on stimulus-object and pattern discriminations. / . gen.Psychol. (In press)11. HARLOW, H. F. Studies in discrimination learning by monkeys. VI. Discriminationsbetween stimuli differing in both color and form, only in color and only in form. J. gen.Psychol. (In press)12. HARLOW, H. F. Shifting of set by rhesus monkeys between matching from sample anddiscrimination problems using the same stimuli. (Unpublished manuscript)13. HARLOW, H. F. Effect of removal of the prefrontal areas in rhesus monkeys on ambivalentcue discriminations problems. (Unpublished manuscript)


17/17

12 6 R. J. CAMPBELL AND H. F. HARLOW14. HARLOW, H. F. Effect of the removal of the prefrontal areas in rhesus monkeys on theretention of the matching-from-sample and the odd ity principles. (Unpublished manu-script)15. HARLOW, H. F. Solution by monkeys following removal of the prefrontal areas of spatial

delayed responses using a combined matching-non-matching from sample method. (Un-published manuscript).16. HARLOW, H. F., & DAGNON, J. Problem solution by monkeys following bilateral removalof the prefrontal areas. I. The discrimination and discrimination-reversal problems./ . exp. Psycho!., 1942, 32, 351-356.17. HARLOW, H. F., & JOHNSON, T. Problem solution by monkeys following bilateral removalof the prefrontal areas. III. Test of initiation of behavior. / . exp. Psychol., 1943, 32,495-500.18. HARLOW, H. F., & POCH, S. Generalization by macaque monkeys to unidimensional andmultidimensional stimuli. / . comp. Psychol. (In press)19. HARLOW, H. F., & SPAET, T. Problem solution by monkeys following bilateral removal ofthe prefrontal areas. IV. Responses to stimuli having multiple sign values. / . exp.Psychol, 1943, 33, 5-57-20. HULL, C. L., & SPENCE, K. W. 'Correction' vs. 'non-correction' method of trial-and-errorlearning in rats. / . comp. Psychol., 1938, 25, 127-145.21. JACOBSEN, C. F. Functions of frontal association areas in primates. Arch.Neurol.Psychiat.,1935, 33, 558-56o22. JACOBSEN, C. F. Studies of cerebral function in primates: I. The functions of the frontalassociation areas in monkeys. / . comp. Psychol., 1936, 13, 360.23. JACOBSEN, C. F. Unpublished study, quoted by Fulton, J. F. Physiology of the nervoussystem.24. JACOBSEN, C. F., & NISSEN, H. W. Studies of cerebral function in primates: IV. Theeffects of frontal lobe lesions on the delayed alternation habit in monkeys. / . comp.Psychol., 1937, 23 , 101-112.25. JACOBSEN, C. F., WOLF, J. B., & JACKSON, T. A. An experimental analysis of the functionsof the frontal association areas in primates. / . nerv. mint. Dis., 1935, 82, 1-14.26. KENNARD, M. A., SPENCER, S., & FOUNTAIN, G., JR. Hyperactivity in monkeys following,lesions of the frontal lobes. / . Neurophysiol., 1941, 4, 512-524.27. KLUVER, H. Behavior mechanisms in monkeys. Chicago U niv. Press, 1933, xvii + 387 pp.28. LANGWORTHY, O. R., & RICHTER, C. P. Increased spontaneous activity produced byfrontal lobes lesions in cats. Amer . J. Physiol., 1939, 126, 158161.29. MALMO, R. B. Interference factors in delayed response in monkeys after removal of frontallobes. / . Neurophysiol., 1942, 5, 295-308.30. NICHOLS, I. C, & HUNT, J. McV. A case of partial bilateral frontal lebectomy. Amer .J. Psychiat., 1940, 96, 1063-83.31. NOER, M. C, & HARLOW, H. F. Discrimination of ambivalent cue stimuli by macaquemonkeys. / . gen. Psychol. (In press)32. RUCH, T. C, & SHENKIN, H . A. The relation of area 13 of the orbital surface of the frontallobes to hyperactivity and hyperphagia in monkeys. Quoted by Fulton , J. F., Physiologyof the nervous system.33. SPAET, T. Solution by rhesus monkeys of problems involving second order sign behaviorwith inter-modal cues. / . gen. Psychol., 1944, 64, 175-184.34. SPAET, T., & HARLOW, H. F. Problem solution by monkeys following bilateral removalof the prefrontal areas. II. Delayed reaction problems involving use of the matching-from-sample method. / . exp. Psychol., 1942, 32, 424-434.35. SPAET, T., & HARLOW, H . F . Solution by rhesus monkeys of multiple-sign problems utilizingthe oddity technique. / . comp. Psychol., 1943, 35, 119-132.36. WALKER, A. E. A cytoarchitectural study of the prefrontal areas of the macaque monkey.J. comp. N eural, 1940, 73, 59-86.37. WARDEN, C. J., BARRERA, S. E., & GALT, W. The effect of unilateral and bilateral frontallobe extirpation on the behavior of monkeys. / . comp. Psychol, 1942, 34, 149-171.38. YOUNG, M. L., & HARLOW, H . F. Solution by rhesus monkeys of a problem involving theWeigl principle using the oddity method. / . comp. Psychol, 1943, 35, 205-217.39. YOUNG, M. L., & HARLOW, H . F . Generalization by rhesus monkeys of a problem involvingthe Weigl principle using the oddity method. / . comp. Psychol, 1943, 36, 201-216.

110-126 problem solution by monkeys following bilateral removal of the prefrontal areas. v. spatial...

Documents