Journal of Experimental Psychology: Animal

Learning and Cognition

Control of Working Memory in Rhesus Monkeys (Macaca
mulatta)
Hsao-We Tu and Robert R. Hampton
Onne Frst Pubcaton, Apr 21, 2014. http://dx.do.org/10.1037/xan0000030
CITATION
Tu, H.-W., & Hampton, R. R. (2014, Apr 21). Contro of Workng Memory n Rhesus Monkeys
(Macaca muatta). journal of Experimental Psychology: Animal Learning and Cognition.
Advance onne pubcaton. http://dx.do.org/10.1037/xan0000030
Control of Working Memory in Rhesus Monkeys (Macaca mulatta)
Hsiao-Wei Tu and Robert R. Hampton
Emory University
Cognitive control is critical for efficiently using the limited resources in working memory. It is well
establishedthat humansuserehearsal toincreasetheprobabilityof rememberingneededinformation, but
littleisknowninnonhumans, withsomestudiesreportingtheabsenceof activecontrol andotherssubject
to alternative explanations. We trained monkeys in a visual matching-to-sample paradigm with a
post-sample memory cue. Monkeys either saw a remember cue that predicted the occurrence of a
matchingtestthatrequiredmemoryfor thesample, or aforgetcuethatpredictedadiscriminationtestthat
didnot requirememory of thesample. Infrequent probetrialsonwhichmonkeysweregiventestsof the
type not cued on that trial were used to assess whether memory was under cognitive control. Our
procedures controlled for reward expectation and for the surprising nature of the probes. Monkeys
matchedless accurately after forget cues, whereas discrimination accuracy was equivalent in the2cue
conditions. Wealsotestedmonkeyswithlistsof 2consecutivesampleimagesthat sharedthesamecue.
Again, memory for expected memory tests was superior to that on unexpected tests. Together these
results showthat monkeys cognitively control their working memory.
Keywords: memory control, activeprocessing, directed forgetting, rehearsal, substitution procedure
Workingmemoryisasysteminwhicharelativelysmall amount
of information is temporarily maintained for active manipulation
and rapid access (Baddeley, 1986; Cowan, 2005). Because the
capacity of workingmemory is highly constrained(Cowan, 2005;
Cowan, Morey, Chen, & Bunting, 2007; Williams & Woodman,
2012), cognitivecontrol iscritical for efficientlymanaginglimited
resources for processing themost relevant information. It is well
established that humans selectively rehearseappropriateinforma-
tion to keep working memory updated and to minimize interfer-
ence fromirrelevant information (Cowan, 2005; Sheard & Ma-
cLeod, 2005). For example, wemay maintain thenameof anew
acquaintanceinmemory by repeatingit subvocally sothat wecan
access it immediately when necessary.
The human ability to actively control memory resources has
beenstudiedover thepastfewdecadesusingdifferentvariationsof
thedirected-forgetting paradigmdeveloped by Bjork, LaBerge,
and LeGrand (1968). In the itemmethod, participants are in-
structedtoremember or forgetindividual stimuli immediatelyafter
eachpresentation. Inthelist method, theinstructionto remem-
ber or forget occursinthemiddleof alist of stimuli andappliesto
all thestimuli precedingtheinstruction. Whenaskedinsubsequent
tests to report as many stimuli as possibleregardless of previous
instructions, subjects report more of the stimuli they were in-
structedto remember (e.g., Hauswald& Kissler, 2008; MacLeod,
1999; Minnema & Knowlton, 2008; Titz & Verhaeghen, 2010).
Converging evidence shows that humans are able to adaptively
allocatememory resources to maintaintheavailability of specific
itemsinworkingmemory(Hourihan, Ozubko, & MacLeod, 2009;
Sheard & MacLeod, 2005; Williams & Woodman, 2012).
Applyingthesedirected-forgettingparadigms to nonhumanan-
imalsischallengingandhasyieldedcontroversial results(Roper &
Zentall, 1993; Zentall, Roper, Kaiser, & Sherburne, 1998). As a
result, studies of directedforgettinghaveyet to establishwhether
nonhuman working memory involves cognitively effortful pro-
cessing, such as rehearsal, although other kinds of evidence are
suggestive(Basile& Hampton, 2013).
Because we cannot directly instruct nonhuman animals to re-
member somestimuli and not others, weinstead haveto arrange
contingenciesof reinforcement that favor activeuseof memory in
someconditions but not others. Typically, subjects arepresented
with a sample stimulus followed by either a remember cue (R
cuehereafter), indicatingthatatestof memoryfor thesamplewill
follow, or a forget cue (F cue hereafter), indicating that a
memory test for thesamplewill not follow. On F-cued trials the
memory test iseither omitted(e.g., Roberts, Mazmanian, & Krae-
mer, 1984; Roper, Chaponis, & Blaisdell, 2005; Washburn &
Astur, 1998) or adifferent test unrelatedto memory for thegiven
sample stimulus is substituted in place of the memory test (e.g.,
Grant & Barnet, 1991; Kendrick, Rilling, & Stonebraker, 1981;
Hsiao-Wei TuandRobert R. Hampton, Department of Psychology and
Yerkes National PrimateResearch Center, Emory University.
Hsiao-Wei Tuisnowat thePsychologyDepartment, VirginiaPolytech-
nic Instituteand StateUniversity.
This work was supportedby theUnitedStates Department of Health
& Human Services, National Institutes of Health (NIH), National In-
stituteof Mental HealthGrant R01MH082819andby National Science
Foundation Grants BCS-0745573 and IOS-1146316. This project was
funded by the NIH Division of Nutrition Research Coordination, Na-
tional Center for ResearchResources Grant P51RR165andis currently
supportedby theNIH Officeof ResearchInfrastructureProgramsGrant
OD P51OD11132. We thank Emily K. Brown, J essica A. J oiner, and
Tara A. Dove-VanWormer for assistance with testing animals and
animal care.
Correspondence concerning this article should be addressed to Hsiao-
Wei Tu, 230 Williams Hall, Virginia Polytechnic Institute and State
University, 890 Drillfield Drive, Blacksburg, VA 24061. E-mail:
hsiaowei@vt.edu
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
J ournal of Experimental Psychology: Animal Learning and Cognition 2014 American Psychological Association
2014, Vol. 40, No. 3, 000 2329-8456/14/$12.00 http://dx.doi.org/10.1037/xan0000030
1
Maki, Olson, & Rego, 1981; Washburn & Astur, 1998). After
training, probe trials are administered, in which a memory test
unexpectedly follows anF cue, toassess whether memory perfor-
manceis contingent onthecuegivenfollowingthesample. Con-
trol of memory by theR and F cues is inferred to theextent that
performance on F-cued probe trials is less accurate than that on
regular R-cued trials (Roper & Zentall, 1993).
Previousstudiesthat appliedthedirected-forgettingparadigmto
nonhuman animals were performed almost exclusively with pi-
geons, and the results are controversial. Alternative explanations
exist for the differences in accuracy observed between R-cued
trials and F-cued probes. For example, when subjects aretrained
usinganomissionprocedure, inwhichmemory tests after F cues
areomitted, subjects may associatetheF cuewiththeabsenceof
foodreward. Expectationof no rewardcoulddecreasemotivation
or increasetheprobability of behaviorsthat areincompatiblewith
accuratetest performance, suchasturningawayfromtheresponse
apparatus, resultingin poorer performanceon F-cuedprobetrials
(Maki &Hegvik, 1980; Zentall, Roper, &Sherburne, 1995). When
alternativerewardedtestsfollowF cuesinsubstitutionprocedures,
the interpretation of results may depend on the nature of the
substitutedtask andtherequiredresponse(Grant & Barnet, 1991;
Kendrick et al., 1981). In some designs (e.g., Grant & Barnet,
1991; Grant & Soldat, 1995), arandomstimulus was presentedat
test after F cuesandchoiceswererewardedindiscriminately. This
may train subjects to respond nondifferentially after seeing an F
cue, resulting in the performance discrepancy between regular
R-cuedtrialsandF-cuedprobes. Inother studies(e.g., Kendricket
al., 1981; Maki & Hegvik, 1980; Maki et al., 1981), the same
discrimination task was always presented after F cues. Because
subjects can anticipate the response required at test in such de-
signs, F cues may trigger prospectivememory for thediscrimina-
tion target, and this may passively displace any items already in
memory (seeWashburn & Astur, 1998, for arelated discussion).
In contrast to the number of studies done with pigeons, there
havebeen fewefforts to test for directed forgetting in nonhuman
primates, despitethefact that similar mnemonic processingmight
be most likely in our close primate relatives. In several experi-
ments, Roberts et al. (1984) did not find any evidencefor active
memory control in squirrel monkeys (Saimiri sciureus) when a
largeset of stimuli wasused, but after reducingthestimulusset to
only twoimages, performanceonmemory tests was lower after F
cues thanafter R cues inafinal test. Becauseanomissionproce-
dure was used in which the F cue signaled the absence of rein-
forcement, theinferenceof memorycontrol isquestionablefor the
reasons cited above. Thedifficulty in interpreting theseresults is
highlighted by a later study using rhesus monkeys (Macaca mu-
latta; Washburn& Astur, 1998). Inthisstudy, accuracyof all tests
on which reinforcement was expected was comparatively high,
whether or not remembering was necessary, emphasizing theim-
portanceof controlling theexpectation of reward. This last study
isthemost recent publishedinvestigationof directedforgettingin
monkeysweknowof, andtheauthorsconcludedthat noevidence
wasfoundfor anactivemaintenanceprocess. . . Nonetheless, the
conclusion of Roberts et al. (1984) that directed forgetting may
occur only whenmemory istestedwithasmall set of imagesmay
haveforecast morerecent resultscontrastingmemoryfor largeand
small sets of images (Basile & Hampton, 2010, 2013; Eacott,
Gaffan, & Murray, 1994).
Wereport studies in rhesus monkeys using methods that elim-
inatesomeof thealternativeexplanationstowhichearlystudiesof
nonhumananimals weresubject. InExperiment 1, monkeys were
trained with an R cuethat predicted theoccurrenceof amemory
test andanF cuethat predictedadiscriminationtask that didnot
require memory of the sample, similar to the directed-forgetting
paradigmadoptedinmanypigeonstudies(Roper &Zentall, 1993).
Weused manipulations that address concerns about motivational
state, level of expectationof reward, andtheactivationof incom-
patiblememories onF-cuedtrials. InExperiment 2, wepresented
lists of two consecutivesampleimages followed by onecueand
tested memory for both of the studied images. This experiment
tested whether themonkeys would generalizeuseof theR and F
cues and also allowed a closer comparison to the list method
usedinhumanresearch. Bothexperimentstest thehypothesisthat
monkeys use effortful and limited cognitive control to maintain
sample images in working memory during the delay interval of
memory tests. According to this hypothesis, monkeys actively
maintainmemoriesonR-cuedtrialsandthiseffortful maintenance
is rewardedwhenmonkeys respondcorrectly inmemory tests. In
contrast, monkeys forgo active maintenance on F-cued trials be-
causetheeffortful maintenancehasnot beenreinforcedintraining.
Experiment 1A
The substitution procedure used in some pigeon directed-
forgetting research (Roper & Zentall, 1993) was used in Experi-
ment 1A. Monkeys were trained to expect a matching-to-sample
test when an R cuefollowedthesampleandto expect adiscrim-
inationtest whenanF cuewaspresented. Onlyonediscrimination
problemwas used and thespecific test that followed F cues was
therefore highly predictable. If monkeys actively maintain infor-
mation in working memory when they expect amemory test, but
donot dosowhenamemory test isnot expected, performanceon
thememory tests following R cues should bemoreaccuratethan
performanceontheF-cuedprobetrials. Wealsoevaluatedwhether
any decrement observed after F-cued probe trials is specific to
failure to maintain memory for the sample, or is rather due to a
general decrement that would be present in any unexpected test.
WedidthisbyincludingR-cuedprobetrials, inwhichRcueswere
followed by an unexpected discrimination test. If discrimination
performanceis worseontheseprobetrials thanonnormal F-cued
trials, this would indicate that performance is impaired in any
unexpected test.
Method
Subjects and apparatus. Six adult male rhesus monkeys
(Macaca mulatta) were used. All monkeys were pair-housed,
receivedfull rationsof foodeachday, andhadadlibitumaccessto
water at all times. Duringtesting, pairswereseparatedbydividers
suchthat eachmonkey couldaccess only its owntest equipment,
but still had limited social contact with the cage mate. Monkeys
hadprior experiencewithcomputerizedmatching-to-sampletests.
Monkeys weretested six days aweek with aportabletesting rig
attachedtothefront of thehomecage. Testingrigs consistedof a
15-inch color LCD touch-sensitive monitor (Elo TouchSystems,
MenloPark, CA) runningataresolutionof 1024768pixels, two
automatic fooddispensers (MedAssociates, Inc., St. Albans, VT)
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
2
TU AND HAMPTON
that delivered nutritionally balanced primate pellets (Bio-Serv,
Frenchtown, NJ ) into foodcups belowthescreen, andapersonal
computer that controlled the experiments with customprograms
written in Presentation (NeuroBehavioral Systems, Albany, CA).
Object-discrimination learning. Each object discrimination
problemconsistedof four color clip-art images, 160pixelshigh
200pixelswide, oneof whichwasrandomlyselectedasthetarget
at the beginning of training. Trials started when a green square
(100 100 pixels) appeared at the lower center of the screen.
Oncethesubject touched it twice(FR2), thegreen squaredisap-
peared and the four images making up the discrimination were
presented, each at a randomly determined corner of the screen.
Touching (FR2) the target image was rewarded with one pellet
together with an excellent! sound, whereas touching oneof the
other threeimagesresultedinadoh! soundandatimeout of 5s
duringwhichthescreenwas black. Consecutivetrials weresepa-
rated by an intertrial interval of 5 s. Monkeys learned four dis-
crimination problems concurrently, in sessions in which each
problemoccurred40timesinarandomorder. Oncetheproportion
correctwasatleast.90for eachproblemsimultaneouslyinasingle
session, oneof theproblems was randomly selectedto beusedin
Experiment 1A.
Cued matching to sample. Four color clip-art images, 160
pixels high 200 pixels wide, were used for matching. Two
shapes (triangle or oval) of two colors (blue or yellow) were
randomly assigned as R cues and F cues, respectively, for each
monkey. Eachtrial beganwhenagreensquare(100 100pixels)
appearedcenterednear thebottomof thescreen. Touchingit twice
(FR2) startedthestudyphase, inwhichoneof thefour imageswas
pseudorandomlyselectedasthesampleandpresentedinthecenter
of thescreen. Oncemonkeys touched thesampletwice(FR2), it
disappeared. After 200 ms, an R or F cue appeared in a random
location on the screen. Monkeys touched the cue twice (FR2) to
enter thedelay periodduringwhichthescreenwasblack. Thetest
phasebegan after theprogrammed memory delay had elapsed. If
anR cuehadappearedafter thesample, all four imagesappeared,
each randomly assigned to onecorner of thescreen. Selection of
thesampleimage(FR2) shownduringthestudy phaseresultedin
a food reward and an excellent sound. Selection of any of the
threedistractorsresultedinad=oh soundand5-stimeout, during
which the screen was black. If an F cue had appeared after the
sample, thefour imagesof thediscriminationproblemappearedon
the screen. Choosing the target image (FR2) resulted in a food
reward and an excellent! sound, whereas incorrect responses
generatedadoh! soundanda5-s timeout period. Half of trials
wereR-cuedandhalf F-cued. Consecutivetrialswereseparatedby
an intertrial interval of 5 s (Figure1A).
Accuracy titration. Thematching accuracy of each monkey
on R-cued trials was titrated to between .50 and .70 by manipu-
latingthedelayinterval. Reducingaccuracytothislevel wasdone
to encourage memory maintenance and also to avoid ceiling ef-
fects that might mask differences in performance on R- and F-
cuedtrials. Eachsessioncontained160trials, 80withanRcueand
80withanF cue. Westartedtitrationwitha3-sdelayinterval from
theoffset of theRor F cuetotheonset of thetest. If theproportion
correct onR-cuedtrialswashigher than.70inasinglesession, the
delay for bothR-cuedtrials andF-cuedtrials was doubled. If the
proportion correct on R-cued trials was lower than .50 for two
Figure 1. (A) Thedirected-forgettingparadigm. TheR andF cues werepresentedat randomlocations onthe
screen. AnR cuewasfollowedbyamatchingtest of memoryfor thesampleimage(upper panel); anF cuewas
followedbyapre-traineddiscriminationtest (lower panel). Theredcirclesindicatethecorrect responses. (B) An
example of a two-sample trial. If an R cue was presented after the two sample images (upper panel), two
matchingtests appearedinsequenceafter thedelay interval intheorder correspondingto samplepresentation.
If anF cuewaspresentedafter thetwosampleimages(lower panel), twodiscriminationtestsrandomly chosen
from10pre-traineddiscriminationsappearedoneafter theother followingthedelay interval. Thecolor version
of this figureappears in theonlinearticleonly.
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
3
MEMORY CONTROL IN MONKEYS
consecutive sessions, the delay was decreased by a quarter. Per-
formanceonF trialswasnot expectedtobeaffectedby thelength
of delay becausememory for thesamplewas not required. Mon-
keysproceededtothenext stage, probetesting, whentheaccuracy
on R-cued trials fell between .50 and .70 for each of three con-
secutivesessions.
Probe testing. Ten percent of thetrials in each session were
probes, 5%wereF-cuedprobesand5%R-cuedprobes. OnF-cued
probetrials thematchingtest was presentedinsteadof adiscrim-
inationtest, whereasonR-cuedprobetrialsthediscriminationtest
was presented instead of a matching test. All trials for a given
monkey had the same delay length, determined by the titration
procedure. Thesamediscriminationtest usedintitrationwasused
here. Regular trials were rewarded when the correct image was
selected, that is, thesampleonR-cuedtrialsandthediscrimination
target onF-cuedtrials. Probetrialswererewardednomatter which
imagewasselected. Therewere320trialsineachsession, includ-
ing 288 regular trials, half of which were R-cued and half were
F-cued, and32probetrials, half R-cuedandhalf F-cued. Thesame
discrimination test used in titration was used here. Each monkey
received five sessions, yielding 80 R-cued probe trials and 80
F-cued probetrials.
Data analysis. Memory performanceonregular R-cuedtrials
and F-cued probe trials was compared by paired t tests to deter-
mine whether the R and F cues had any effect on memory for
sample images. Discrimination performance on regular F-cued
trialsandR-cuedprobetrialswasalsocomparedwithtest whether
accuracy on all unexpected tests was attenuated, whether or not
memory was required. All accuracy dataweretransformedbefore
analysis by taking the arcsine of the square root of proportion
correct scores. This transformationis recommendedto makepro-
portionsbetter conformtothenormality assumptionof parametric
statistics (Kirk, 1982). All trials wereprogrammedwiththesame
delay interval, but monkeys responded at their own pace. Given
that memory declines with increasing delay, differences in delay
intervals experienced by monkeys under the two cue conditions
could cause differences in accuracy. To examine this possibility
and to better isolate the effect of the cues on memory, median
experienceddelay intervals, calculatedastheinterval betweenthe
offset of thesampleandtheregistrationof atest response(FR2),
were also compared between regular R-cued trials and F-cued
probe trials as well as between regular F-cued trials and R-cued
probe trials by paired t tests. For each t value larger than 1, we
reportedthestandardizedCohensd
z
for within-subjectsdesignsas
suggestedbyLakens(2013) aswell asthe95%confidenceinterval
for this measure of effect size calculated using a bootstrapping
procedurein R (Cumming, 2012).
Results and Discussion
Bytheendof thetitrationstage, eachmonkeyhadbeenassigned
an individually tailored delay, ranging from8.5 s to 24 s (M
14.42s). Theaverageperformanceonregular R-cuedtrials inthe
last threesessions of all six monkeys was .60 .04 and that on
regular F-cued trials was .99 .01.
Monkeys were significantly more accurate on regular R-cued
trials, when they expected a memory test, than they were on
F-cuedprobetrials, t(5) 2.88, p .04, d
z
1.18, 95%CI [0.08,
2.22]; seeblack barsinFigure2. Thisdiscrepancy wasnot dueto
different lengths of experienceddelay intervals, t(5) 1.33, p
.24, d
z
0.54, 95% CI [0.34, 1.38], suggesting that monkeys
actively maintained memory on R-cued trials. Monkeys could
potentially havelearnedtochooseindiscriminately onprobetrials
becauseall choices onprobetrials werereinforced. Suchlearning
seems unlikely, however, because probe trials only constituted
10% of the trials in each session. If the monkeys had learned to
treat probe trials differently, their performance should have de-
creasedas they learnedover thefivesessions. Wetestedfor such
learningby comparingaccuracy inthefirst 20F-cuedprobetrials
inthefirst sessionwiththat inthelast 20F-cuedprobetrialsinthe
last session and found no difference, t(5) 0.45, p .67.
Incontrast tothedifferencefoundbetweenregular andprobe
memory tests, performance on discrimination tests was so ac-
curate no matter which cue occurred that there was nearly no
variationfor conductingt tests(regular F-cuedtrials: M 1.00,
SD .00; R-cued probe trials: M 1.00, SD .01). This
ceilinglevel accuracy might havepreventedus frommeasuring
a decrease in discrimination ability that would be evident if
performance were not at ceiling. To provide additional, and
potentially more sensitive, assessments of whether monkeys
were disrupted or surprised by the R-cued probe trials, we
compared experienced delay intervals of these trials to other
trial types. We found no significant difference in experienced
delay intervals between regular F-cued trials and R-cued probe
trials, t(5) 0.46, p .67, indicating no disruption in perfor-
mance. In addition, experienced delay intervals were signifi-
cantly shorter on R-cued discrimination trials than on R-cued
memory trials, t(5) 2.92, p .03, d
z
1.19, 95%CI [0.09,
2.23], consistent with a lack of hesitation in responding on
R-cued discrimination trials. Together, these results suggest
that the R and F cues determined whether monkeys actively
remembered the sample images during the delay interval.
Figure 2. Accuracyof regular R-cuedtrialsandF-cuedprobetrialsinthe
threeconditions inExperiment 1. AnR cuepredictedamatchingtest that
requiredmemoryfor thesampleimage; anF cuepredictedadiscrimination
test unrelated to memory for the sample image. Memory accuracy on
F-cuedprobetrials was significantly poorer thanit was onregular R-cued
trials. Theresults wereconsistent across Experiment 1A (black bars), 1B
(gray bars), and 1C (open bar), showing that monkeys followed thecues
and allocated their working memories accordingly. Error bars indicate
standard errors.
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
4
TU AND HAMPTON
Experiment 1B
Themain result of Experiment 1A replicates thosefromsome
directed-forgettingstudies inpigeons. As was thecasewiththose
studies, thesamediscriminationproblemalways appearedafter F
cues, anditispossiblethattheF cuetriggeredprospectivememory
for thetarget of theupcomingdiscriminationproblem. Activation
of a memory for the target of the discrimination might displace
memory for the sample seen on that trial, leading to reduced
accuracyonF-cuedprobetrials. Suchdisplacement canbeviewed
as apassiveretroactiveinterferenceeffect, complicatinginterpre-
tation of the R cue as instructing active maintenance of memory
for thesample. To reducetheprobability that theF cueactivates
prospectivememory for thetarget of thediscrimination, wemade
it much more difficult for the monkeys to predict the specific
discriminationproblemthat wouldfollowF cues. Thediscrimina-
tion problemused on each particular F-cued trial was randomly
selected fromamong 10 well-learned discriminations in Experi-
ment 1B. Because the monkeys could no longer predict which
discriminationtarget wouldbeappropriatefor thecurrent trial, the
possibility that the F cues triggered prospective memory was
greatly reduced.
Method
Thesamesix monkeys andtestingrigs usedinExperiment 1A
wereused. All monkeys learnedanother two sets of four discrim-
inationproblems as describedabove. Droppingthediscrimination
problemused in Experiment 1A and another randomly selected
problemresultedinthepool of 10discriminationproblemsusedin
Experiment 1B. The accuracy on R-cued matching trials was
titrated again to between .50 and .70 by the same procedure as
before.
Trials proceededas inExperiment 1A, except that thediscrim-
ination task for each F-cued trial was pseudorandomly selected
fromthe10newdiscriminationproblems. Eachsessioncontained
800trials, half hadanRcueandhalf hadanF cue. Of the400trials
withR cuesandF cues, 360wereregular trialsand40wereprobe
trials. Sampleimages, cuetypes, anddiscriminationtaskswereall
counterbalanced in each session. Monkeys were tested for two
sessions, for atotal of 80probetrials of eachtype. Accuracy and
delay intervals were analyzed in the same way as in Experiment
1A.
Results and Discussion
After titrating the accuracy again with 10 possible discrimina-
tion problems following F cues, the delays ranged from930 s
(M 21.00s, SD 8.05s). Theaverageperformanceonregular
R-cued trials in thelast threesessions was .64 .02 and that on
regular F-cued trials was .99 .01.
The performance on regular R-cued trials was significantly
higher thanthat onF-cuedprobetrials, t(5) 3.20, p .02, d
z

1.31, 95% CI [0.15, 2.40]; see gray bars in Figure 2, and the
experienced delay intervals on F-cued probe trials and regular
R-cuedtrialsdidnot differ, t(5) 0.40, p .71. AsinExperiment
1A, accuracyonthefirst 20andlast 20F-cuedprobetrialsdidnot
differ significantly, t(5) 0.74, p .49, indicatingthat monkeys
did not learn to respond indiscriminately on probetrials. In con-
trast to thedifferenceinaccuracy betweenR andF-cuedmemory
trials, accuracy onR-cuedprobetrials remainedas highas that of
regular F-cued trials (regular F-cued trials: M .97, SD .06;
R-cuedprobetrials: M 1.00, SD .01). As inExperiment 1A,
therewas no significant differencein experienced delay intervals
betweenregular F-cuedtrialsandR-cuedprobetrials, t(5) 1.88,
p .12, d
z
0.77, 95% CI [0.18, 1.66], and monkeys were
significantly faster in responding on R-cued discrimination trials
than on R-cued memory trials, t(5) 2.61, p .05, d
z
1.06,
95% CI [0.01, 2.06], consistent with a lack of hesitation in re-
spondingonR-cueddiscriminationtrials. Theseresults, collected
under conditions that prevent prospectivememory for discrimina-
tiontargets, reproducethefindingsfromExperiment 1A indicating
activecontrol of memory on R-cued trials.
Experiment 1C
Thedisruptioninmemory accuracy followingF cues observed
inExperiments1A and1B providesstrongevidencefor theability
of monkeys to actively regulatethecontents of workingmemory.
However, monkeys performed almost perfectly on each discrimi-
nationproblem, earningarewardnearly every trial, whereas their
memoryaccuracyaveragedonly.64. Thisdiscrepancymayleadto
differences in expectation of food rewards after R and F cues. If
monkeys expect a food reward after seeing an F cue but then
receiveamemorytest that hasamuchlower probabilityof reward,
they may befrustratedor less motivated, providinganalternative
nonmnemonicexplanationfor thelower accuracyonF-cuedprobe
trials. ThisconcernwasaddressedinExperiment 1Cbyyokingthe
reinforcement onregular F-cuedtrialswiththat onregular R-cued
trials to equatetheexpectation of food reward.
Method
Thesamesix monkeys andtitrateddelays wereused. No addi-
tional trainingwasneeded. TrialsproceededasinExperiment 1B,
but theprobability of reinforcement on regular F-cued trials was
yoked with that on regular R-cued trials. Becausemonkeys were
nearly 100% correct on F-cued discrimination trials, this was
accomplished by calculating the proportion of R-cued trials re-
wardedintheprevioussessionandrewardingthesameproportion
of correct discrimination responses in the current session. Mon-
keysranonesessionof 400R-cuedmatchingtrialsand400F-cued
discriminationtrialsbeforewestartedyoking. Theperformanceon
theseR-cuedtrials was calculatedandusedfor yokingtheF-cued
trials in the first session of the current experiment. All settings,
including number of discrimination tasks used and trial numbers,
remainedthesameas inExperiment 1B. Probetrials wereimple-
mentedinthefirst yokedsession, andtwosessions wereobtained
fromeach monkey. The accuracy and response latencies of the
resulting80probetrials of each typewerecomparedin thesame
way as in Experiments 1A and 1B.
Results and Discussion
Monkeys were again more accurate on R-cue trials than on
F-cueprobetrials, evenafter theprobabilityof reinforcement onR
andF trialswasequated, t(5) 2.77, p .04, d
z
1.13, 95%CI
[0.05, 2.15]; see open bars in Figure 2. Further analysis showed
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
5
MEMORY CONTROL IN MONKEYS
that this differencewas not causedby adifferenceinexperienced
delayinterval, t(5) 0.06, p .96. AsinExperiments1A and1B,
performanceonF-cuedprobetrials didnot decreasesignificantly
betweenthefirst andlast 20trials, t(5) 1.36, p .23, d
z
0.56,
95%CI [0.33, 1.40], indicating that themonkeys did not learn
that probe trials were differently rewarded than regular trials.
Performanceinthediscriminationtests was high, andequivalent,
on F-cued and R-cued trials (regular F-cued trials: M 1.00,
SD .00; R-cued probe trials: M 1.00, SD .01). Again,
experienceddelayintervalsonregular F-cuedtrialsandonR-cued
probe trials did not differ, t(5) 0.57, p .59, suggesting that
there was no disruption in responding to R-cued probe trials. In
addition, experienceddelay interval onR-cueddiscriminationtri-
als was significantly faster than those on R-cued memory trials,
t(5) 2.86, p .04, d
z
1.17, 95%CI [0.07, 2.20], consistent
with a lack of hesitation in responding on R-cued discrimination
trials.
It might takesomeexperiencewiththeyokingprocedurefor the
monkeys to learn about the new reinforcement rate for the dis-
criminationtests. If thiswerethecase, wemight not expect tosee
aneffect of yokinguntil thesecondsessionof testing. Toevaluate
whether a yoking effect might emerge only after experience,
accuracies in the first and second session of testing were com-
pared. Theperformanceinthefirst andthesecondyokedsessions
didnot differ significantlyonregular R-cuedtrials(M.64inthe
1st session, M .66 in the 2nd session), t(5) 0.83, p .44,
regular F-cuedtrials (M .99, SD .01inthe1st session, M
1.00, SD .00inthe2ndsession), R-cuedprobetrials(M1.00,
SD .00inthe1stsession, M.99, SD .01inthe2ndsession),
or F-cuedprobetrials(M .46inthe1st session, M .50inthe
2nd session), t(5) 1.13, p .31, d
z
0.46, 95% CI [0.41,
1.29]. Theseresults indicatethat thepoorer memory performance
on F-cued probetrials was not dueto lower motivation or lower
expectation of reward.
Takentogether, Experiments1A, 1B, and1Cdemonstrateactive
control of memorybymonkeysandeliminateconfoundingfactors
that commonly contaminate studies of active memory control in
nonhuman animals. Specifically, our data suggest that monkeys
areabletofollowtheR andF cuesandactively maintainmemory
when it is necessary to keep themost current and relevant infor-
mation availablein working memory.
Experiment 2
Theability to actively maintain working memory was demon-
strated in Experiment 1. Although this establishes aparallel with
the item method used in studies of human working memory,
human subjects areoften presented with lists of stimuli followed
by an instruction to remember or forget in a list method. To
determinewhether themonkeyswouldgeneralizetheRandF cues
to amodifiedtest situation andto provideacloser parallel to the
variety of methodsusedinhumans, wepresentedtwoconsecutive
sample images followed by a single cue and tested memory for
bothof thestudiedimages. If monkeysgeneralizeuseof theR and
F cuesfromthepreviousexperimentstothecurrenttestswithshort
lists of two images, weshouldcontinueto observesuperior accu-
racy on R-cued trials compared with F-cued probetrials.
Method
Subjects and apparatus. Thesamesix monkeys and testing
equipment wereused.
Cued matching to sample. Thefour images used on match-
ing trials in Experiment 1 were always grouped together as a
quad at test, withoneof themrandomlychosenasthesampleas
before. Four new color clip-art images, 160 pixels high 200
pixels wide, were added and grouped together to make a second
quad. Thus, one image from each quad could be randomly
selectedoneachtrial tocreateatwo-imagelist of samplesfor that
trial. Trials progressed as described in Experiment 1, except that
after touching (FR2) the first sample image, it disappeared and
another sampleimageappearedinthesamelocationafter 200ms.
Monkeys had to touch (FR2) the second sample image before
receivingtheR or F cueandthedelayinterval. At test, if anR cue
had followed the two sample images, two matching tests were
presented one-by-one in the same order as the corresponding
sampleimages hadbeenpresentedinthestudy phase. If anF cue
had followed the two sample images, two of the 10 pre-trained
discrimination tests were randomly selected and presented one
after the other (Figure 1B). Correct responses were immediately
rewardedwithafoodpellet andanexcellent! sound. Reinforce-
ment onF-cuedtrials was not yokedinthis experiment. Incorrect
responses resultedinadoh! sound, but no timeout period. The
timeout periodwasomittedtokeepthedelay interval betweenthe
offsetof eachsampleimageandtheonsetof thecorrespondingtest
constant, evenwhentherewereincorrect responses. Eachsession
contained192trials, half of whichwereR-cuedandtheother half
F-cued. The selection of samples for the two matching tests, the
order in which these two samples appeared in a trial, and the
discrimination problems used were counterbalanced in each ses-
sion. Thesamesymbols wereusedas R cues andF cues for each
individual.
Accuracy titration. Becausetherewerenowtwo, rather than
just one, to-be-remembered images, performance on the R-cued
trials was retitrated by the same procedure described in Experi-
ment 1. All trials initially hada3-s delay interval fromtheoffset
of thesampletotheonset of thecorrespondingtest. Monkeyswere
movedtothenext stagewhentheaverageproportioncorrect of the
twomatchingtrialsafter R cuesfell between.50and.70for three
consecutivesessions.
Probe testing. Eachsessionconsistedof 256R-cuedtrialsand
256 F-cued trials, among which 16 R-cued and 16 F-cued trials
wereselectedto beprobes. OnF-cuedprobetrials, bothdiscrim-
inations werereplacedby matchingtests intheorder correspond-
ingto samplepresentations, whereas onR-cuedprobetrials, both
matchingtestswerereplacedbydiscriminationtests. Regular trials
were rewarded when the correct image was selected, whereas
probe trials were rewarded no matter which image was selected.
Each monkey was tested for 8 sessions, yielding 128 probetrials
of each type.
Data analysis. Theeffects of cues and presentation order on
accuracy and experienced delay interval were both analyzed by
repeated-measuresANOVAs. All accuracy dataweretransformed
by takingthearsineof thesquareroot of correct proportionsprior
toanalysis(Kirk, 1982). Experienceddelayintervalsweredefined
astheinterval fromtheoffset of eachsampletotheregistrationof
aresponse(FR2) tothecorrespondingtest. Mediandelayintervals
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
6
TU AND HAMPTON
wereanalyzed. Significant main effects werefurther analyzed by
plannedpairedt tests. For eacht valuelarger than1, wereported
the standardized Cohens d
z
for within-subjects designs as sug-
gestedbyLakens(2013) aswell asthe95%confidenceinterval for
this measure of effect size calculated using a bootstrapping pro-
cedurein R (Cumming, 2012).
Results and Discussion
Accuracy titration. Titrateddelaysrangedfrom612s(M
8.50 s, SD 2.26 s). Thesedelays aresignificantly shorter than
those used in Experiments 1B and 1C, indicating that it is more
difficult for themonkeystoremember twoimagesthanone, t(5)
3.76, p .01, d
z
1.54, 95%CI [0.28, 2.73]. Accuracyinthelast
three sessions of titration averaged .62 .04 on the first test of
regular R-cued trials and .55 .04 on thesecond test of regular
R-cued trials. Monkeys were more accurate in first tests than in
secondtests, t(5) 3.96, p .01, d
z
1.62, 95%CI [0.33, 2.85],
averaged across thelast threesessions.
Probe testing. Whenthecuewasfollowedby matchingtests,
therewasasignificant maineffect of cue, F(1, 5) 15.30, MSE
0.01, p .01,
p
2
0.75. Further analysisshowedthatbothsample
imageswererememberedbetter followingRcuesthanfollowingF
cues: first tests: t(5) 3.99, p .01, d
z
1.63, 95%CI [0.34,
2.86]; secondtests: t(5) 3.43, p .02, d
z
1.40, 95%CI [0.21,
2.53]; see Figure 3. This discrepancy was not caused by differ-
ences in experienced delay intervals F(1, 5) 1.14, MSE
52,667.09, p .34,
p
2
0.19. The performance on the first 20
F-cuedprobetrialsdidnot differ significantlyfromthat onthelast
20 F-cued probe trials, t(5) 0.79, p .47, indicating the
monkeysdidnot learntorespondindiscriminately onprobetrials.
Performanceondiscriminationtrials was almost always at ceiling
without variation for conducting statistical analysis regardless of
whichcueprecededthetest(after Rcues: M.98, SD .02; after
F cues: M 1.00, SD .00), and no significant differencewas
foundinexperienceddelay intervals after F or R cues, F(1, 5)
0.12, p .74. These results reinforce the findings from the
preceding experiments and extend themto memory for multiple
sample images. As in Experiment 1, we compared the response
latencies on R-cued discrimination trials. Latencies for regular
F-cuedtrials andR-cuedprobetrials didnot differ either for first
tests, t(5) 0.61, p .57, or second test, t(5) 0.22, p .84,
indicatingthat performancewas not disruptedonR-cueddiscrim-
ination trials. In addition, response latencies for R-cued probe
trials were significantly shorter than those for regular R-cured
trialsonthefirst tests, t(5) 9.33, p .0001, d
z
3.81, 95%CI
[1.39, 6.22], and not significantly different than thosefor regular
R-cuedtrialsonthesecondtests, t(5) 1.91, p .11, d
z
0.78,
95% CI [0.17, 1.68], indicating that the monkeys were not
hesitant on R-cued probetrials.
When the cue was followed by matching tests, there was a
significant main effect of test order on performance, F(1, 5)
11.45, MSE 0.01, p .02,
p
2
0.70, andonexperienceddelay
interval, F(1, 5) 80.14, MSE 361,793.29, p .001,
p
2
0.94,
but theinteractionbetweencueandtest order wasonlysignificant
in performance, F(1, 5) 10.16, MSE 0.001, p .02,
p
2

0.67, not in experienced delay interval, F(1, 5) 4.31, MSE

22832.59, p .09,
p
2
0.46. Further analysis showed that on
regular R-cuedtrials, memoryfor thefirst imageseenduringstudy
(whichwasalsothefirstimagetested) wassignificantlybetter than
that for thesecondimage, t(5) 4.93, p .004, d
z
2.01, 95%
CI [0.54, 3.44] (seeFigure3), andtheexperienceddelay interval
was significantly shorter for thefirst test, t(5) 9.13, p .001,
d
z
3.73, 95%CI [1.35, 6.09]. However, onF-cuedprobetrials,
memory for first samples did not differ from that for second
samples, t(5) 1.68, p .15, d
z
0.69, 95%CI [0.24, 1.56]
(seeFigure3), but theexperienceddelay interval was still signif-
icantlyshorter for first thanfor secondtests, t(5) 8.40, p .001,
d
z
3.43, 95% CI [1.22, 5.62]. Superior memory for the first
imageisconsistent withthewell-knownprimacyeffect inmemory
for lists (Basile & Hampton, 2010; Sands, Urcuioli, Wright, &
Santiago, 1984; Wright, 1994). It may bethat higher performance
in tests with the first image reflects priority access of the first
imagestudiedtocognitiveresourcesnecessary for activememory
maintenance. Such an account is consistent with our finding that
this superior memory for thefirst imageis eliminated by presen-
tation of F cues.
Although it may be that the primacy effect found on R-cued
trials, but not onF-cuedprobetrials, isduetounequal distribution
of limited cognitive resources responsible for active memory
maintenance, thereareother factorsthat might producethepattern
of results observed. First, significantly shorter experienced delay
intervals for the first image could account for the higher perfor-
mance on first tests after R cues. However, even though similar
significant differences in experienced delay interval were found
after F cuesandR cues, accuracyonfirst andsecondtestsdiffered
only following R cues. This suggests that differences in experi-
enced delay are not the main reason for the different patterns of
accuracy after R and F cues. Second, taking the first test might
createvarious types of interferencethat couldlower performance
onthesecondtest. For example, retrievingthefoodrewardonthe
first test could interfere with accuracy in the subsequent test. If
retrieving a reward interfered with performance in a subsequent
test, wewouldexpect thedifferenceinaccuracy betweenthefirst
and second tests to begreater on F-cued probetrials, wherefirst
Figure 3. Sampleimages werebetter rememberedwhenmatchingtests,
not discrimination tests, wereexpected. In addition, when monkeys were
cuedto remember thesampleimages, thefirst matchingtests weresignif-
icantly moreaccuratethanthesecondone, suggestingthat activerehearsal
might occur duringthedelayinterval. Incontrast, whenmonkeysexpected
to receive discrimination tests on F-cued probed trials, accuracy was
equally poor for boththefirst andsecondmemory test. Error barsindicate
standard errors.
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
7
MEMORY CONTROL IN MONKEYS
responses wererewarded100%of thetimeindiscriminately, than
on regular R-cued trials, where first responses were correct and
rewarded only about 70% of the time. But that is not what we
found. Instead, wefoundthat onregular R-cuedtrials, secondtest
accuracywassignificantlyhigher whenfirsttestswerecorrectthan
when thefirst tests wereincorrect, t(5) 11.47, p .001, d
z

4.68, 95% CI [1.77, 7.60]. Third, absence of reward following
errors infirst tests might causeanemotional responsethat would
impair performanceinsecondtests. However, suchadecrement in
accuracy wouldonly beexpectedonregular R-cuedtrials but not
F-cuedprobetrialsbecauseall firstresponseonF-cuedprobetrials
were rewarded. In addition, if emotional responses caused by
errors and the absence of reward were the main cause of lower
accuracyinsecondtests, performanceonthesecondmemorytests
after F cues should beas high as that after R cues when thefirst
tests wererewardedbecauseF-cuedsecondtests always followed
reward. However, whenfirst memory testswerecorrect, monkeys
performedsignificantly better onthesecondmemory tests after R
cuesthanafter F cues, t(5) 12.05, p .001, d
z
4.92, 95%CI
[1.88, 7.97]. Thus, it appears that the R and F cues determined
accuracy in second tests.
Wewouldgenerally expect any effect of interferencefromfirst
tests to beconsistent, regardless of which cuewas presented and
whether monkeys actively rehearsed the sample images or not,
becausetherewerealways two tests on each trial. In contrast we
foundthat memoryfor thefirst imagewassignificantlybetter only
on R cued matching tests. Although the interpretation of these
results is not uncomplicated by the occurrence of two tests on
everytrial, theweight of theevidencefavorstheinterpretationthat
the first image studied gains priority access to active memory
resourcesandtheF-cueabolishesthisdifferencebyattenuatingthe
contribution of active memory. Nonetheless, a design that coun-
terbalancedtheorder inwhichmemorytestsaregivenafter Rcues,
or conducted only a single test on each trial, would provide a
stronger test and should beconducted in futurework.
General Discussion
In thepresent study, monkeys showed significantly lower per-
formance on memory tests that followed F than on tests that
followed R cues. The effect was consistent across a series of
experiments designed to eliminatecommon confounds. Although
it isstill possiblethat theR cuesreinforcedor taggedmemory for
thesampleimageinsomewaythat facilitatedrecognitionbut does
not necessitateactivemaintenance, our resultshaveshownthat the
contents of our samplemonkeys working memory was different
followingR cuesandF cues, eventhoughtheencodingconditions
wereidentical in both trial types. Theseresults parallel what has
been found in human-memory research (Allen & Vokey, 1998;
Sheard& MacLeod, 2005; Williams& Woodman, 2012), suggest-
ing that monkeys, likehumans, may beableto strategically allo-
cate limited working-memory resources in service of current
needs.
The original directed-forgetting paradigm used in humans
(Bjork et al., 1968) has beenmodifiedinvarious ways to accom-
modateanimal training. Howanimals areinstructed to forget is
critical tointerpretationof differencesbetweenRandF cuedtrials.
Omitting tests following F cues may create low motivation and
incompatible behaviors that confound mnemonic explanations
(Maki & Hegvik, 1980; Zentall et al., 1995). Instead, asubstitution
procedurewith asinglediscrimination test after F cues has been
usedmost commonly, withinconsistent results. Instudiesinwhich
nodeficit onF-cuedprobetrialswasfound(Kendrick et al., 1981;
Maki & Hegvik, 1980; Maki et al., 1981), it has beenarguedthat
the anticipation of making a choice might encourage animals to
continuerehearsingthesampleon F-cuedtrials, resultingin high
performanceonbothR-cuedtrialsandF-cuedprobes(Kendrick&
Rilling, 1986). Althoughpossible, thisisweak evidencefor active
memory, and at best shows poor discrimination of theconditions
under which rehearsal is useful. Other studies using thesubstitu-
tion procedure found poorer performance on F-cued probe trials
(Grant & Barnet, 1991; Roper, Kaiser, & Zentall, 1995), as we
foundwithour monkeysinExperiment 1A. However, theseresults
arevulnerableto thealternativeexplanation that memory for the
sample might have been passively displaced by the prospective
memory for the response required in the anticipated discrimina-
tion.
Given these concerns, it is difficult to determine conclusively
whether nonverbal animals showactivemaintenanceof memory.
Roper et al. (1995) argued that truedirected forgetting may only
occur when the original memory for the sample is updated with
newsample-independent informationinworkingmemory. Intheir
paradigm, memoryfor thesamplewasstrategicallydesignedtobe
displaced by memory for the F cues, and a significantly lower
accuracy for the original sample was observed on F-cued probe
trials. To further investigatewhether directed-forgettingeffects in
animals is due to cognitive control of active working memory
rather than passivedisplacement of memories with newinforma-
tion, wetrainedour monkeys onmultiplediscriminationtests and
randomizedwhichdiscriminationfollowedtheF cueoneachtrial
inExperiment 1B. Monkeysthereforecontinuedtoexpect tohave
tomakeachoiceandtoberewardedonbothR- andF-cuedtrials,
butwereunlikelytoprospectivelyremember thechoicetobemade
in thediscrimination tests. With this new variation of thesubsti-
tution procedure, wefound arobust directed-forgetting effect.
In addition to substituting F-cued tests with multiplediscrimi-
nationproblems, therewereseveral featuresof our designthatmay
differ fromthose used with pigeons. Typically, before the inclu-
sion of probe trials, animals were trained to associate the R cue
with a memory test and the F cue with another test that did not
require memory. The performance on R-cued and F-cued trials
might be high after training, but it did not indicate successful
learning of the cues. Because remembering was not particularly
effortful on these trials, subjects may simply have ignored the
post-samplecueand tried to remember on every trial to increase
theprobability of gettingareward. Wemodifiedour paradigmin
at least threewaystoemphasizetheimportanceof thecuesduring
trainingandencouragemonkeys to actively remember only when
necessary. First, thelocationof cuepresentationchangedoneach
trial, and the monkeys had to visually search for the cue before
touching it to proceed. This reduced thepossibility of inattentive
responses andstrengthenedtheassociationbetweeneachcueand
itscorrespondingtest. Second, theretentioninterval wastitratedto
bring accuracy into a moderate range. Because actively keeping
memory availableis an effortful process, it is possiblethat mon-
keys would not have adopted this strategy unless it had been
beneficial to do so. When the delay interval becomes longer,
memoryfor thesamplenaturallyweakens, andactivemaintenance
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
8
TU AND HAMPTON
becomes necessary to increase the chance of food reward com-
paredwithpassivelywaitingfor thetest. Third, thesmall imageset
used on R-cued trials created substantial interference, which re-
duced the possibility of using familiarity for matching and may
have enhanced the necessity of rehearsal (Basile & Hampton,
2013; Eacott et al., 1994; Roberts et al., 1984). A major property
of human working memory is that it is vulnerable to competing
cognitive demand (Logie, 1986; Phillips & Christie, 1977). In
monkeys, memory for images drawn froma small imageset has
been found to be especially vulnerable to competing cognitive
demands (Basile& Hampton, 2013), suggestingthat suchmemo-
ries are actively, rather than passively, maintained, a property
consistent withthememory control effectswehavereportedhere.
Experimental designs other than directed-forgetting paradigms
havebeendevelopedtostudy theactivecontrol of workingmem-
ory in nonhuman primates. For example, distractors, such as a
sample-independent stimulus (Artchakov et al., 2009; Takeda,
Naya, Fujimichi, Takeuchi, & Miyashita, 2005), amotor response
(Washburn & Astur, 1998), or a categorization task (Basile &
Hampton, 2013) have been inserted during retention intervals to
compete with memory for the sample image and the limited
cognitiveresources inworkingmemory. Whether activerehearsal
had been observed seemed to depend on the extent to which
distractor tasks increased cognitive load. List learning is another
commonmethodinwhichprimacyisoftenarguedtoreflectactive,
rehearsal-like process of items in memory (Basile & Hampton,
2010; Cook, Wright, & Sands, 1991), but alternative interpreta-
tions exist (Sands et al., 1984).
Our findings of significantly higher performance on the first
matching tests than on thesecond ones only after R cues but not
after F cues coincide with the results of human studies. For
example, clear primacy and recency effects have been reported
when participants were asked to recall items in the to-be-
remembered list, but only aslight primacy and no recency effect
wasfoundwhenrecallingitemsintheto-be-forgottenlist (Pastt-
ter, Kliegl, &Buml, 2012; Sheard&MacLeod, 2005). Sheardand
MacLeod(2005) further concludedthat selectiverehearsal maybe
themain explanation for directed forgetting. Although other fac-
tors may have also contributed to the primacy-like pattern we
observedonregular R-cuedtrials, this first attempt to incorporate
cuesinalist-learningparadigmofferedusanopportunity tostudy
activememory control in nonverbal animals in away that paral-
leled the paradigms used in human research, facilitating cross-
species comparisons of memory processes.
Here we eliminated memory-irrelevant factors that confound
manypreviousreportsof memorycontrol innonhumans, including
effects of rewardexpectancy andpotential interferencefrompro-
spective memory. The case for memory control was further sup-
portedby extensionof theeffects of theR andF cues to multiple
stimuli in lists. Given the fact that matching was more often
rewardedinthepresenceof theR cuethantheF cue, this pattern
of behavior would at one time have been described as better
stimuluscontrol of matchingbehavior bytheR cuethantheF cue.
Althoughsuchadescriptioncancorrectlypredict performanceand
well capturethecontingenciespresent inour experimental design,
itsayslittleaboutthecognitiveprocessesthatmightberesponsible
for thecontrol of memory retention by theR and F cues. Wedo
interpret our results as showing that matching performance is
under differential stimulus control by the R and F cues, but
additionally suggest that it is not just performance that is under
the control of these cues, but a memory process that facilitates
accuratematching. Futureworkwill evaluatewhether thismemory
process is an active one, vulnerable to competing cognitive de-
mand, as we have found to be the case with some, but not all,
memories (Basile & Hampton, 2013). The nonhuman primate
model of cognitivecontrol of memory establishedhereprovidesa
basis for further investigations of the underlying mechanisms by
which brains control memory.
References
Allen, S. W., & Vokey, J . R. (1998). Directedforgettingandrehearsal on
direct and indirect memory tests. In J . M. Golding & C. M. MacLeod
(Eds.), Intentional forgetting: Interdisciplinary approaches (pp. 173
195). Mahwah, NJ : Erlbaum.
Artchakov, D., Tikhonravov, D., Ma, Y., Neuvonen, T., Linnankoski, I., &
Carlson, S. (2009). Distracters impair and create working memory-
related neuronal activity in the prefrontal cortex. Cerebral Cortex, 19,
26802689. doi:10.1093/cercor/bhp037
Baddeley, A. D. (1986). Working memory. Oxford, UK: OxfordUniversity
Press.
Basile, B. M., & Hampton, R. R. (2010). Rhesus monkeys (Macaca
mulatta) show robust primacy and recency in memory for lists from
small, but not large, image sets. Behavioural Processes, 83, 183190.
doi:10.1016/j.beproc.2009.12.013
Basile, B. M., & Hampton, R. R. (2013). Dissociation of activeworking
memory and passive recognition in rhesus monkeys. Cognition, 126,
391396. doi:10.1016/j.cognition.2012.10.012
Bjork, R. A., LaBerge, D., & LeGrand, R. (1968). The modification of
short-termmemorythroughinstructionstoforget. Psychonomic Science,
10, 5556. doi:10.3758/BF03331404
Cook, R. G., Wright, A. A., & Sands, S. F. (1991). Interstimulus interval
and viewing time effects in monkey list memory. Animal Learning &
Behavior, 19, 153163. doi:10.3758/BF03197871
Cowan, N. (2005). Working memory capacity. NewYork, NY: Psychology
Press. doi:10.4324/9780203342398
Cowan, N., Morey, C. C., Chen, Z., & Bunting, M. (2007). What do
estimatesof workingmemorycapacitytell us?InN. Osaka, R. H. Logie,
& M. DEsposito(Eds.), The Cognitive Neuroscience of Working Mem-
ory (pp. 4358). NewYork, NY: OxfordUniversity Press. doi:10.1093/
acprof:oso/9780198570394.003.0003
Cumming, G. (2012). Understanding the new statistics: Effect sizes, con-
fidence intervals, and meta-analysis. NewYork, NY: Routledge.
Eacott, M. J ., Gaffan, D., & Murray, E. A. (1994). Preserved recognition
memory for small sets, and impaired stimulus identification for large
sets, followingrhinal cortex ablationsinmonkeys. European Journal of
Neuroscience, 6, 14661478. doi:10.1111/j.1460-9568.1994.tb01008.x
Grant, D. S., & Barnet, R. C. (1991). Irrelevance of sample stimuli and
directed forgetting in pigeons. Journal of the Experimental Analysis of
Behavior, 55, 97108. doi:10.1901/jeab.1991.55-97
Grant, D. S., & Soldat, A. S. (1995). A postsample cue to forget does
initiateanactiveforgettingprocessinpigeons. Journal of Experimental
Psychology: Animal Behavior Processes, 21, 218228. doi:10.1037/
0097-7403.21.3.218
Hauswald, A., &Kissler, J . (2008). Directedforgettingof complexpictures
in an item method paradigm. Memory, 16, 797809. doi:10.1080/
09658210802169087
Hourihan, K. L., Ozubko, J . D., & MacLeod, C. M. (2009). Directed
forgettingof visual symbols: Evidencefor nonverbal selectiverehearsal.
Memory & Cognition, 37, 10591068. doi:10.3758/MC.37.8.1059
Kendrick, D. F., & Rilling, M. E. (1986). AIM: A theory of active and
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
9
MEMORY CONTROL IN MONKEYS
inactive memory. In D. F. Kendrick, M. E. Rilling, & M. R. Denny
(Eds.), Theories of animal memory (pp. 129152). Hillsdale, NJ : Erl-
baum.
Kendrick, D. F., Rilling, M., & Stonebraker, T. B. (1981). Stimuluscontrol
of delayed matching in pigeons: Directed forgetting. Journal of the
Experimental Analysis of Behavior, 36, 241251. doi:10.1901/jeab.1981
.36-241
Kirk, R. E. (1982). Experimental design: Procedures for the behavioral
sciences (4th ed.). Thousand Oaks, CA: Sage.
Lakens, D. (2013). Calculating and reporting effect sizes to facilitate
cumulativescience: A practical primer for t-tests and ANOVAs. Fron-
tiers in Psychology, 4, 112. doi:10.3389/fpsyg.2013.00863
Logie, R. H. (1986). Visuo-spatial processinginworkingmemory. Quar-
terly Journal of Experimental Psychology: Section A. Human Experi-
mental Psychology, 38, 229247. doi:10.1080/14640748608401596
MacLeod, C. M. (1999). Theitemandlist methods of directedforgetting:
Test differences and the role of demand characteristics. Psychonomic
Bulletin & Review, 6, 123129. doi:10.3758/BF03210819
Maki, W. S., & Hegvik, D. K. (1980). Directed forgetting in pigeons.
Animal Learning & Behavior, 8, 567574. doi:10.3758/BF03197771
Maki, W. S., Olson, D., & Rego, S. (1981). Directedforgettinginpigeons:
Analysis of cue functions. Animal Learning & Behavior, 9, 189195.
doi:10.3758/BF03197819
Minnema, M. T., & Knowlton, B. J . (2008). Directed forgetting of emo-
tional words. Emotion, 8, 643652. doi:10.1037/a0013441
Pasttter, B., Kliegl, O., & Buml, K.-H. T. (2012). List-method directed
forgetting: The forget cue improves both encoding and retrieval of
postcueinformation. Memory & Cognition, 40, 861873. doi:10.3758/
s13421-012-0206-4
Phillips, W. A., & Christie, D. F. M. (1977). Interferencewith visualiza-
tion. The Quarterly Journal of Experimental Psychology, 29, 637650.
doi:10.1080/14640747708400638
Roberts, W. A., Mazmanian, D. S., & Kraemer, P. J . (1984). Directed
forgetting in monkeys. Animal Learning & Behavior, 12, 2940. doi:
10.3758/BF03199810
Roper, K. L., Chaponis, D. M., & Blaisdell, A. P. (2005). Transfer of
directed-forgetting cues across discrimination tasks with pigeons. Psy-
chonomic Bulletin & Review, 12, 10051010. doi:10.3758/BF03206435
Roper, K. L., Kaiser, D. H., & Zentall, T. R. (1995). True directed
forgettinginpigeonsmayoccur onlywhenalternativeworkingmemory
is requiredonforget-cuetrials. Animal Learning & Behavior, 23, 280
285. doi:10.3758/BF03198924
Roper, K. L., & Zentall, T. R. (1993). Directed forgetting in animals.
Psychological Bulletin, 113, 513532. doi:10.1037/0033-2909.113.3
.513
Sands, S. F., Urcuioli, P. J ., Wright, A. A., &Santiago, H. C. (1984). Serial
position effects and rehearsal in primate visual memory. In H. L.
Roitbalt, T. G. Bever, & H. S. Terrace (Eds.), Animal Cognition (pp.
375388). Hillsdale, NJ : Erlbaum.
Sheard, E. D., & MacLeod, C. M. (2005). List methoddirectedforgetting:
Return of theselectiverehearsal account. In N. Ohta, C. M. MacLeod,
& B. Uttl (Eds.), Dynamic Cognitive Processes (pp. 219248). New
York, NY: Springer. doi:10.1007/4-431-27431-6_10
Takeda, M., Naya, Y., Fujimichi, R., Takeuchi, D., &Miyashita, Y. (2005).
Activemaintenanceof associativemnemonic signal inmonkey inferior
temporal cortex. Neuron, 48, 839848. doi:10.1016/j.neuron.2005.09
.028
Titz, C., & Verhaeghen, P. (2010). Aging and directed forgetting in
episodic memory: A meta-analysis. Psychology and Aging, 25, 405
411. doi:10.1037/a0017225
Washburn, D. A., & Astur, R. S. (1998). Nonverbal working memory of
humans and monkeys: Rehearsal in the sketchpad? Memory & Cogni-
tion, 26, 277286. doi:10.3758/BF03201139
Williams, M., & Woodman, G. F. (2012). Directedforgettinganddirected
rememberinginvisual workingmemory. Journal of Experimental Psy-
chology: Learning, Memory, and Cognition, 38, 12061220. doi:
10.1037/a0027389
Wright, A. A. (1994). Primacy effects in animal memory and human
nonverbal memory. Animal Learning & Behavior, 22, 219223. doi:
10.3758/BF03199923
Zentall, T. R., Roper, K. L., Kaiser, D. H., & Sherburne, L. M. (1998). A
critical analysis of directed-forgetting research in animals. In J . M.
Golding& C. M. MacLeod(Eds.), Intentional forgetting: Interdisciplin-
ary approaches (pp. 265287). Mahwah, NJ : Erlbaum.
Zentall, T. R., Roper, K. L., & Sherburne, L. M. (1995). Most directed
forgetting in pigeons can beattributed to theabsenceof reinforcement
onforget trialsduringtrainingor toother procedural artifacts. Journal of
the Experimental Analysis of Behavior, 63, 127137. doi:10.1901/jeab
.1995.63-127
Received August 7, 2013
Revision received March 14, 2014
Accepted March 19, 2014
T
h
i
s
d
o
c
u
m
e
n
t
i
s
c
o
p
y
r
i
g
h
t
e
d
b
y
t
h
e
A
m
e
r
i
c
a
n
P
s
y
c
h
o
l
o
g
i
c
a
l
A
s
s
o
c
i
a
t
i
o
n
o
r
o
n
e
o
f
i
t
s
a
l
l
i
e
d
p
u
b
l
i
s
h
e
r
s
.
T
h
i
s
a
r
t
i
c
l
e
i
s
i
n
t
e
n
d
e
d
s
o
l
e
l
y
f
o
r
t
h
e
p
e
r
s
o
n
a
l
u
s
e
o
f
t
h
e
i
n
d
i
v
i
d
u
a
l
u
s
e
r
a
n
d
i
s
n
o
t
t
o
b
e
d
i
s
s
e
m
i
n
a
t
e
d
b
r
o
a
d
l
y
.
10
TU AND HAMPTON