Neuropsychologia 50 (2012) 3070–3079

Contents lists available at SciVerse ScienceDirect

Neuropsychologia

0028-39

http://d

$Deg

procedun Corr

E-m

journal homepage: www.elsevier.com/locate/neuropsychologia

Item memory, context memory and the hippocampus: fMRI evidence$

Michael D. Rugg a,n, Kaia L. Vilberg a, Julia T. Mattson a,b, Sarah S. Yu a, Jeffrey D. Johnson c, Maki Suzuki d

a Center for Vital Longevity and School of Behavioral and Brain Sciences, 1600 Viceroy Drive, Suite 800, Dallas, TX 75235, United Statesb UT Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75390, United Statesc Department of Psychological Science, 210 McAlester Hall, University of Missouri, Columbia, MO 65211, United Statesd Khoyama Center for Neuroscience, Department of Intelligent Systems, Faculty of Computer Science and Engineering, Kyoto Sangyo University, Kamigamo-Motoyama, Kita-Ku,

Kyoto 603-8555, Japan

a r t i c l e i n f o

Available online 23 June 2012

Keywords:

Episodic memory

Familiarity

Memory strength

Recollection

Remember-know

32/$ - see front matter & 2012 Elsevier Ltd. A

x.doi.org/10.1016/j.neuropsychologia.2012.06

rees of freedom corrected for non-sphericity w

re.

esponding author. Tel.: þ1 972 883 3725; fax

ail address: mrugg@utdallas.edu (M.D. Rugg).

a b s t r a c t

Dual-process models of recognition memory distinguish between the retrieval of qualitative informa-

tion about a prior event (recollection), and judgments of prior occurrence based on an acontextual

sense of familiarity. fMRI studies investigating the neural correlates of memory encoding and retrieval

conducted within the dual-process framework have frequently reported findings consistent with the

view that the hippocampus selectively supports recollection, and has little or no role in familiarity-

based recognition. An alternative interpretation of these findings has been proposed, however, in which

it is argued that the hippocampus supports the encoding and retrieval of ‘strong’ memories, regardless

of whether the memories are recollection- or familiarity-based. Here, we describe the findings of eight

fMRI studies from our laboratory: one study of source memory encoding, four studies of the retrieval of

contextual information, and three studies of continuous recognition. Together, the findings support the

proposal that hippocampal activity co-varies with the amount of contextual information about a study

episode that is encoded or retrieved, and not with the strength of an undifferentiated memory signal.

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction

There is a broad consensus that recognition memory issupported by two different processes, usually referred to asrecollection and familiarity (Mandler, 1980; Yonelinas, 2002;Wixted & Mickes, 2010). Recollection occurs when a recognitiontest item elicits retrieval of qualitative information about thestudy episode. This information includes not only the identity ofthe studied item, but also details about the context in which itwas studied. By contrast, familiarity supports a sense of pastoccurrence that is devoid of contextual information. Whereasboth recollection and familiarity can support simple ‘old/new’recognition judgments, judgments based on the content of anepisode – source memory or ‘Remember’ judgments for example(see below) – depend largely on recollection.

The functional characteristics of recollection and familiarity,and their neural substrates, are currently matters of debate.Contentious issues include the questions of whether the memorysignal associated with recollection is better modeled as a thre-sholded or a continuous process (Wixted & Mickes, 2010;Yonelinas, Aly, Wang, & Koen, 2010), and whether recollection

ll rights reserved.

.004

ith the Geisser–Greenhouse

: þ1 972 883 3250.

and familiarity are differentially dependent upon the hippocam-pus (Eichenbaum, Yonelinas, & Ranganath, 2007; Squire, Wixted,& Clark, 2007). Here, we focus on the second of these issues as it isinformed by functional neuroimaging studies, although we touchupon the first issue also.

The proposal that recollection, but not familiarity, is dependentupon the hippocampus has been advanced by numerous authors.Several lines of evidence have been interpreted in favor of theproposal, although none has gone unchallenged. For example,whereas some studies of patients with lesions restricted to thehippocampus have reported disproportionate deficits in estimatesof recollection (e.g., Aggleton et al., 2005; Holdstock et al., 2002;Mayes et al., 2004) studies of other, seemingly similar patientshave reported that recollection and familiarity are impaired to anequivalent extent (e.g., Cipolotti et al., 2006; Jeneson, Kirwan,Hopkins, Wixted, & Squire, 2010; Manns, Hopkins, Reed,Kitchener, & Squire, 2003; Wais, Wixted, Hopkins, & Squire,2006). And whereas there is compelling evidence from studies ofexperimental animals that hippocampal lesions can have little orno impact on recognition memory of single items as assessed bytasks such as delayed non-match to sample (for review, see Brown& Aggleton, 2001) or spontaneous exploration (e.g., Good, Barnes,Staal, McGregor, & Honey, 2007) it is unclear how directly thesefindings relate to the constructs of recollection and familiarity,which are difficult to operationalize in experimental animals(although see Fortin, Wright, & Eichenbaum, 2004).

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–3079 3071

An important line of evidence in support of the proposal thatrecollection is disproportionately dependent on the hippocampuscomes from functional neuroimaging studies in healthy humans thatemployed event-related fMRI. Such studies have investigated both theneural correlates of the encoding processes associated with successfulrecollection on a later memory test, and the correlates of successfulrecollection at the time of retrieval. In these studies, recollection wasalmost invariably operationalized either in terms of successful versusunsuccessful memory for a contextual feature of the study episode(source memory), or as the difference in neural activity whenrecognition of a test item was accompanied by a phenomenal senseof recollection compared with when it was accompanied only by asense of familiarity (‘Remember vs. Know’, Tulving, 1985). In studiesof both encoding (e.g., Davachi, Mitchell, and Wagner, 2003; Duarte,Henson, and Graham, 2011; Kensinger and Schacter, 2006; Otten,2007, Ranganath et al., 2004; Uncapher and Rugg, 2005) and retrieval(e.g., Cohn, Moscovitch, Lahat, and McAndrews, 2009; Diana,Yonelinas, and Ranganath, 2010; Eldridge, Knowlton, Furmanski,Bookheimer, and Engel, 2000; Montaldi, Spencer, Roberts, andMayes, 2006; Yonelinas, Otten, Shaw, and Rugg, 2005) it has beenreported that, relative to familiarity-driven recognition, recollection isassociated with enhanced hippocampal activity. These findings havebeen taken as evidence that the hippocampus plays a selective role insupporting recollection, possibly because of its unique ability tobind the different components of a study episode into a cohesivememory representation (Diana, Yonelinas, & Ranganath, 2007; Mayes,Montaldi, & Migo, 2007).

Recently, an alternative account of these fMRI findings has beenadvanced (Squire et al., 2007; Wixted, Mickes, & Squire, 2010). By thisaccount, encoding- and retrieval-related hippocampal activity areneural correlates not of recollection, but of ‘strong’ memory – asoperationalized by the accuracy and confidence of recognition judg-ments – regardless of whether the strength of the memory is due torecollection, high familiarity, or a combination of the two memorysignals. In support of this account, Wixted et al. (2010) argued thatcontrasts intended to identify the neural correlates of recollection (forexample, contrasts between items attracting correct versus incorrectsource judgments, or items endorsed ’Remember’ versus ‘Know’), areinvariably confounded with differences in item recognition accuracyand hence memory strength. Wixted et al. (2010) noted that even instudies in which confidence judgments were employed to segregate‘unrecollected’ test items according to their level of familiarity (Cohnet al., 2009; Montaldi et al., 2006; Ranganath et al., 2004; Yonelinaset al., 2005), recognition accuracy was invariably higher for itemsendorsed as recollected than it was for highly familiar items (but seeMontaldi & Mayes, 2010). Wixted et al. (2010) argued that it was thisdifference in item accuracy, and not the distinction between recollec-tion and familiarity, that was responsible for the differential hippo-campal activity reported in those studies.

In the following sections, we describe the findings from a series ofstudies from our laboratory that bear directly on the question ofwhether the construct of memory strength is sufficient to account formodulation of encoding- and retrieval-related hippocampal activity,as this activity is assessed with fMRI. On the basis of these findings,we argue that hippocampal activity is sensitive neither to variation inthe strength of an undifferentiated memory signal, nor to whether atest item elicits a subjective sense of recollection. Instead, hippocam-pal activity reflects the amount of contextual information that isencoded during a study episode, or that is retrieved in response to atest item.

2. Measurement of memory strength and recollection

Before turning to our empirical findings we briefly review thebehavioral measures that we and others have employed to

estimate memory strength and recollection in studies of recogni-tion memory. As the construct is employed in the literature citedabove (e.g., Squire et al., 2007; Wixted et al., 2010), strength isoperationalized in terms of the accuracy (pHit/(pHitþpFalseAlarm)) and confidence with which recognition memory testitems are correctly endorsed as studied, and we follow thatpractice below. We defer discussion of the value of the constructof memory strength in understanding the neural correlates ofmemory performance until later.

We have employed two different procedures to operationalizerecollection. The first is the ‘Remember/Know’ procedure, intro-duced by Tulving (1985). In its simplest form, the procedurerequires subjects to signal whether recognition of a test item isaccompanied (Remember judgment) or is not accompanied(Know judgment) by the retrieval of a contextual detail or detailsabout the item’s study presentation. Remember and Know judg-ments are assumed to map onto the constructs of recollection andfamiliarity, respectively. The procedure has been criticized on thegrounds that it merely distinguishes between relatively strongand relatively weak memories (e.g., Donaldson, 1996). Recentevidence suggests however that this is not necessarily the case.Rather, independently of differences in memory strength,Remember and Know judgments can indeed segregate test trialsaccording to whether or not contextual (source-specifying) infor-mation was retrieved (Ingram, Mickes, & Wixted, 2012; Wixted &Mickes, 2010). The second procedure adopted to operationalizerecollection requires subjects to make an explicit judgment abouta specific contextual feature of the study episode, for example,whether a test word was presented at study in a red or a greenfont. It is typically assumed that accurate retrieval of such ‘source’information is indicative of the recollection of the relevantcontextual feature, and hence of the retrieval of qualitativeinformation about the study episode. It is important to note,however, that failure to retrieve source information does notnecessarily mean that recollection failed; the possibility thatrecollection occurred, but that it did not include retrieval ofcontextual features relevant to the source judgment (‘non-criter-ial recollection’, see Yonelinas and Jacoby, 1996), cannot easily bediscounted.

In several of the experiments described below, we employedvariants of the Remember/Know and source memory proceduresto identify and characterize recollection-related hippocampalactivity. Additionally, we estimated the memory strengths of‘recollected’ and ‘unrecollected’ test items when this waspossible.

3. Hippocampal activity and the encoding of source and iteminformation

Below, we describe data drawn from a larger study of theeffects of age on the neural correlates of source memory encoding(unpublished data), focusing on findings from the 17 youngparticipants (ages 18–27, 6 male, all right-handed with noreported neurological or psychiatric histories). During thescanned study phase, subjects viewed a series of 180 colorpictures of objects, each of which was preceded by a cue thatsignaled whether the object should be judged as to whether itwould be more likely to be found indoors or outdoors, or wassmaller or larger than a shoebox. fMRI data were acquired andanalyzed according to our standard methods (e.g., Gottlieb andRugg, 2011). In an unscanned test phase that began approxi-mately 25 min after exiting the scanner, subjects viewed the 180critical study pictures intermixed with 90 new items. Therequirement was to first make an ‘old/new’ judgment on eachpicture and, for each item endorsed ‘old’, to then judge whether

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–30793072

the item had been subjected to an indoors/outdoors or a sizedecision at study. Both the item and source judgments were madeusing confidence ratings: ‘confident old’, ‘unconfident old’, ‘don’tknow’, ‘unconfident new’ and ‘confident new’ for item memory,and ‘confident location’, ‘unconfident location’, ‘don’t know’,‘unconfident size’ and ‘confident size’ for source memory.

Behavioral performance on the test task is summarized inTables 1 and 2. Recognition accuracy (and thus memory strength -see above) for items receiving a ‘confident old’ item memoryjudgment was at ceiling, whereas the accuracy of items accordedan ‘unconfident old’ judgment was significantly lower (accuraciesof .99 and .88 respectively, po .01). Additionally, as has beenreported previously (Kirwan, Wixted, and Squire, 2008; Wais,Squire, and Wixted, 2010), source accuracy was substantiallyhigher for items accorded confident than unconfident itemmemory judgments (.86 and .68 respectively, po .001). Therefore,to avoid any confound between memory strength and sourceaccuracy, only study trials that went on to receive a confident

Table 1Item memory performance for old and new trials by response type (unpublished

data).

Trial type CO UO DK UN CN

Old .76 (.02) .08 (.01) .03 (.01) .07 (.01) .06 (.01)

New .01 (.00) .01 (.00) .04 (.02) .15 (.04) .78 (.06)

Note: Response abbreviations correspond to the following: CO, confident old; UO,

unconfident old; DK, don’t know; UN, unconfident new; CN, confident new.

Standard errors are shown in parentheses.

Table 2Source memory performance for confidently recognized old items according to

confidence of source judgment (unpublished data).

Source confidence Source correct Source incorrect

Confident .67 (.03) .07 (.01)

Unconfident .17 (.03) .06 (.01)

Note: Performance values do not sum to 1.00 due to exclusion of don’t know

source responses. Standard errors are shown in parentheses.

Par

amet

er e

stim

ates

y = -19

L

-0.4

-0.1

0.2

0.5

0.8

Sourcehit

Sourcemiss

Itemmiss

Fig. 1. Findings from unpublished data. A: The top panel displays the outcome of the sou

the average across-subjects parameter estimates for each response category

B: The top panel displays the outcome of the source miss4 item miss contrast (th

parameter estimates for each response category in the left perirhinal peak voxel (coor

item judgment were employed to identify the neural correlates ofsource encoding.

The fMRI analyses focused on contrasts between three classesof study items: ‘source hits’, ‘source misses’, and ‘item misses’.Source hit items were those that went on to receive both aconfident old item judgment and a confident, accurate sourcejudgment on the subsequent memory test. Source misses refer tostudy items that also received a confident old item judgment onthe subsequent test, but for which the source judgment waseither inaccurate or not given (don’t know response). Item missesrefer to items that were either recognized with low confidence orwere misclassified as new on the subsequent test. Thus, thecontrast between source hits and source misses permitted iden-tification of the neural correlates of the encoding of ‘strong’source memories when item memory strength was high andequivalent for items associated with successful versus unsuccess-ful encoding of source information. The contrast between sourcemisses and item misses was employed to identify the neuralcorrelates of the encoding of strong item memories.

The outcomes of these contrasts are illustrated in Fig. 1. Forthe contrast between source hits and source misses (heightthresholded at po .001 with a cluster extent threshold corrected(po .05) for a mask encompassing the MTL), a single clusterlocalized to the right hippocampus was identified (MNI coordi-nates (27, �19, �17); Z¼3.84; see Fig. 1A). The associatedparameter estimates are illustrated in Fig. 1A, where it can beseen that item misses elicited a level of activity similar to thatelicited by source misses. When the same statistical thresholdswere applied to the source miss4 item miss contrast, no voxelswere identified within the MTL at the corrected extent threshold(7 voxels). Reducing the height threshold to po .005 led to theemergence of a 19 voxel cluster located in left perirhinal cortex(MNI coordinates (�36, �10, �29); Z ¼ 3.37; see Fig. 1B). As isevident from the parameter estimates illustrated in Fig. 1B, thislatter effect was also evident for source hits.

To summarize, we found encoding-related activity in thehippocampus to be predictive of successful source memory,whereas activity in perirhinal cortex predicted successful itemmemory. These findings are consistent with those reported inseveral prior studies in which item memory strength was not

y = -10

L

0

0.2

0.4

0.6

0.8

1

Sourcehit

Sourcemiss

Itemmiss

Par

amet

er e

stim

ates

rce hit4source miss contrast (thresholded at po .001). The bottom panel displays

in the right hippocampal peak voxel (coordinate of 27, �19, �17).

resholded at po .005). The bottom panel displays the average across-subjects

dinate of �36, �10, �29).

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–3079 3073

equated between study items that attracted correct and incorrectsource judgments (e.g., Davachi et al., 2003; Ranganath et al.,2004; Kensinger and Schacter, 2006, Duarte et al., 2011). Thus, thepresent findings suggest that the prior findings were not aconsequence of confounding effects of memory strength. Onecaveat to this conclusion arises from the fact that the accuracy ofitem memory in the present experiment was at ceiling. Wecannot rule out the possibility that although the memorystrengths associated with accurate and inaccurate source judg-ments were both very high, a difference in strength nonethelessexisted between the two response categories.

The foregoing experiment is not the first to address thequestion whether encoding-related activity in the hippocampusis predictive of source memory performance on a later memorytest when item memory strength is equated. Kirwan et al. (2008)described an experiment similar in several respects to the onedescribed here. When these authors contrasted the activityelicited on study trials for which the subsequent source judgmentwas accurate or inaccurate, they found that study words thatwere both later recognized and for which the source (one of twoencoding tasks) was correctly retrieved elicited greater righthippocampal activity than did recognized words associated withinaccurate source memory (a result replicated in the study wereport here). No hippocampal effect was evident, however, whenthe analysis of subsequent source memory effects was restrictedto items that had been recognized with high confidence, and wasconducted by performing a linear trend analysis to identify studyactivity that co-varied with the accuracy and confidence of thesubsequent source memory judgments. Kirwan et al. (2008)proposed that their initial finding of a hippocampal subsequentsource memory effect was due to the confounding influence ofmemory strength (like us, they found that items attracting correctsource judgments on the later memory test were recognized withhigher confidence than items attracting inaccurate judgments). Atodds with this proposal, however, subsequent memory effectsassociated with differing levels of memory strength were foundnot in the right hippocampal region that demonstrated thesubsequent source memory effect, but rather, on the perimeterof the hippocampus bilaterally (see Song, Jeneson, and Squire,2011 for similar findings). An alternative possibility is that thedependence of the right hippocampal subsequent source memoryeffect on whether memory strength was or was not equatedreflects differences in the efficiency of the statistical methodsused to identify the effect in each case (linear trend analysis asopposed to a binary contrast).

Table 3Item recognition accuracy for old items classified as R-large, R-small, and K in the

experiments reported by Vilberg and Rugg (2007; 2009a).

Experiment R-large R-small K

Vilberg and Rugg (2007) .99 (.00) .98 (.01) .91 (.02)

Vilberg and Rugg (2009a) .99 (.00) .97 (.01) .91 (.02)

Note: Accuracy is defined as hits/(hitsþfalse alarms). Standard errors are shown in

parentheses.

4. Hippocampal activity and amount of recollectedinformation

In this section we describe previously unpublished analyses ofdata from three studies, focusing on the relationship betweenretrieval-related hippocampal activity and amount of informationrecollected in response to a recognition memory test item. Thefirst study (Vilberg & Rugg, 2007) comprised two experiments (Nsof 14 in each case) that employed almost identical procedures. Asin the original paper, the data are presented here collapsed acrossthe two experiments. In each experiment, subjects initiallystudied a series of picture pairs under the requirement to imaginethe two denoted objects interacting. In the scanned test phase,the items comprised a series of single pictures, two thirds ofwhich were studied and one third of which were new. One of fourresponses was required for each test item: if both the test pictureand its studied pairmate could be recollected a ‘Remember 2’ (R2)response was to be given, if the test picture could be recollectedalong with a detail or details about the study episode other than

its pairmate, a ‘Remember 1’ (R1) judgment was required, itemsthat were judged as studied but for which recollection was absentwere assigned a ‘Know’ (K) response, and items for which thestudy status was uncertain, or that were judged as unstudied,were given a ‘New’ response. We assumed that R2 judgmentswould, on average, be associated with the recollection of moreinformation than would R1 judgments. This assumption wasvalidated with a preliminary behavioral study and a follow-upmemory test conducted after subjects had completed the scannedtest phase.

The second study (Vilberg & Rugg, 2009a) employed a designthat was essentially identical to that just described (Vilberg &Rugg, 2007). However, word pairs rather than picture pairs werepresented at study (the task being to generate a sentenceincorporating the two words), and single words were employedas test items. As in the earlier study, the test requirement was tomake one of two different Remember responses (R2 or R1)depending on whether or not recollection included memory forthe test item’s studied pairmate, and to respond Know to itemsthat were judged old solely on the basis of a sense of familiarity.

In the third experiment (Vilberg & Rugg, 2009b) subjectsinitially studied a series of picture pairs superimposed on acomplex scene. As in the first experiment, the requirement wasto imagine the two objects interacting. To vary the amount ofinformation that would be encoded, and thus the amount thatwould be recollected on the later memory test, the study itemswere presented for either 6 or 1 s. Test items comprised studiedand unstudied single pictures, with the requirement to make anR, K, or New judgment to each item. We confirmed that themanipulation of study duration modulated the amount of infor-mation recollected with a preliminary behavioral study and apost-scan test on which, for each item endorsed ‘R’, subjects wererequired to verbally report the recollected content. Compared tothe 1 s study trials, the 6 s trials were associated with retrieval ofsignificantly more contextual details.

Trials in each of the three experiments associated with therecollection of a relatively large amount of information are here-after referred to as R-large trials (R2 hits in experiments 1 and 2;R hits for items studied for 6 s in experiment 3), whereas trialsassociated with recollection of a relatively small amount arereferred to as R-small (R1 hits in experiments 1 and 2; R hitsfor items studied for 1 s in experiment 3). Table 3 summarizesitem recognition accuracy for the R-large, R-small and K responsecategories in the first and second studies (accuracy cannot becalculated as a function of amount recollected in experiment3 because the different amounts were not associated with distinctresponse categories). As can be seen from the table, accuracyincreased as a function of amount of information recollected(R-large vs. R-small: F(1, 44)¼7.11, po.025; R-small vs. K: F(1, 44)¼19.89, po .001), reaching ceiling for the R-large condition. Impor-tantly, accuracy exceeded 90% for K responses, indicating thatitem memory was strong even in the absence of phenomenalrecollection. In the third study, accuracy (collapsed across thestudy duration manipulation) for R and K judgments was also

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–30793074

high, although significantly higher in the case of R judgments (.99and .89 respectively; F(1, 17)¼9.96, po .01).

At our standard pre-experimental statistical threshold (po .001),separate voxel-wise analyses of each of these data sets revealedlittle evidence for differences in hippocampal activity across thedifferent conditions. Consistent effects were evident, however,when a region-of-interest approach was adopted. In our first study(Vilberg & Rugg, 2007) we had reported that exploratory analyses ata lowered threshold (po .01) identified a left hippocampal regionwhere activity elicited by R-large items exceeded the activityelicited by items endorsed K (peak coordinate �21, �21, �18;Z¼2.44). The parameter estimates associated with R-large, R-small,and K hit trials for a 3 mm sphere centered on that voxel are shownfor each experiment in Fig. 2A–C. Parameter estimates for studieditems misclassified as new (Misses) are also included in the figurefor the first and third experiments (there were too few Miss trials inthe second experiment to permit estimation). As is evident fromFig. 2, in all three studies retrieval-related activity elicited byR-large trials was greater than that for K trials. Moreover, in studies2 and 3, R-large activity also significantly exceeded the activityelicited on R-small trials. In no study did the activity elicited onR-small trials exceed that for K trials. According to Fisher’s proce-dure for combining significance levels of independent tests of thesame null hypothesis (Fisher, 1950), the conjoint one-tailed sig-nificance levels across the three studies for the R-large4R-smalland R-large4K contrasts were p o .001 and po .002 respectively.The conjoint probability of the R-small4K contrast was far fromsignificant (p4 .1).

The outcome of this retrospective analysis of these threestudies is clear: at least in the hippocampal region identified

-0.8

-0.4

0

0.4

0.8

1.2

R-large R-small K Miss

-1

-0.5

0

0.5

1

1.5

2

2.5

3

R-large R-small K Miss

Par

amet

er e

stim

ates

Par

amet

er e

stim

ates

**

***

Fig. 2. Parameter estimates associated with R-large, R-small, and K hit trials for a 3 mm

experiments conducted by Vilberg and Rugg (A: Vilberg and Rugg (2007), B: Vilberg and

on an individual subject’s T1-weighted anatomical image in panel D. Misses are displa

here, retrieval-related activity was greater when subjects recol-lected relatively large amounts of information about a studyevent (R-large trials) than when they recollected relatively smallamounts of information (R-small trials) or when phenomenalrecollection was absent (K trials). Moreover, R-small items eli-cited a level of hippocampal activity that was statistically indis-tinguishable from that elicited by items endorsed as K, despite themarked difference in memory strength between the two classesof item.

5. Hippocampal activity, phenomenal recollectionand source retrieval

The aim of the experiment described in this section (Yu,Johnson, and Rugg, 2012) was to investigate retrieval-relatedactivity in the hippocampus as a function of both phenomenalrecollection and the amount of source-specifying informationretrieved. Subjects studied a series of pictures that were pre-sented to the left or right of central fixation. While undergoingscanning, they viewed a mixture of studied and unstudied itemswith the requirement to make an initial R/K/New judgment toeach item and, for any item endorsed R or K, to then signal theitem’s study location, using a 6-point confidence scale (‘highconfidence left’, ‘moderate confidence left’, ‘low confidence left’,‘high confidence right’, ‘moderate confidence right’, ‘low confi-dence right’). We assumed that the greater the confidence withwhich an accurate source decision was made, the greater theamount of source-specifying information that was retrieved.

-1.2

-0.8

-0.4

0

0.4

0.8

1.2

R-large R-small K

L

y = -21

**

radius region of interest centered at �21, �21, �18 are displayed for the three

Rugg (2009a), C: Vilberg and Rugg (2009b)). The region of interest is superimposed

yed for illustrative purposes only. **¼po .01, *¼po .05.

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–3079 3075

Accuracy of item memory (memory strength) for R and Kjudgments is summarized in Table 4, broken down for items givenan R response according to the confidence/accuracy of theassociated source memory judgment. Highly and moderatelyconfident accurate source judgments are designated as ‘R-High’and ‘R-mod’ response categories. Low confidence and inaccuratejudgments make up the ‘R-weak’ category. K judgments werecollapsed over source accuracy and confidence to form a ‘K-all’category. Item accuracy was markedly greater for items endorsedR than K, and increased as a function of source accuracy.Importantly, accuracy for items endorsed R but for which thesource judgment was unconfident or inaccurate (‘R-weak’ items;92%), and accuracy for ‘K-all’ items (77%) spanned the rangeargued by Wixted et al. (2010) to be sufficient to give rise todifferential hippocampal activity (see Introduction).

The key fMRI findings are summarized in Fig. 3. In brief, thehippocampal activity elicited by K-all items and items receivingan R-weak response did not differ (Fig. 3B). Instead, activityelicited by items endorsed R co-varied with the confidence/accuracy of the associated source judgment. Therefore it appears

Table 4Item memory accuracy as a function of subsequent source memory judgments in

Yu, Johnson, and Rugg (2012).

Experiment R-High R-Mod R-Weak K-All

Yu et al. (2012) .99 (.02) .97 (.03) .92 (.09) .77 (.18)

Note: Accuracy is defined as hits/(hitsþfalse alarms). Standard errors are shown in

parentheses.

Par

amet

er e

stim

ates

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

R-high R-mod R-weak K-all

L

y = -16

p = 0.051

*

***

Fig. 3. A: Outcome of an omnibus ANOVA employed to identify response category

effects in the study of Yu, Johnson, and Rugg (2012) (po .001, 2-tailed).

B: Across subject parameter estimates from the peak voxel at �30, �16, �20.

***¼po .001, *¼po .05.

that hippocampal activity co-varied not with memory strength,but with the amount of contextual (source-specifying) informationthat was recollected. A subsidiary analysis investigated whether thefailure to find a difference in hippocampal activity for R-weak versusK-all trials was due to the inclusion in the K-all response category ofitems attracting highly or moderately confident source judgments.We addressed this question by forming a ‘K-weak’ response categoryanalogous to the R-Weak category described above. Like theR-weak4K-all contrast, the R-weak4K-weak contrast also failed toidentify hippocampal effects.

Not shown in Fig. 3 are the findings for studied items incorrectlyendorsed as new (Misses). These items elicited hippocampal responsesthat were reliably greater than those elicited on K trials, a finding thatechoes the results of the continuous recognition studies described in alater section. The finding should be interpreted cautiously, however,since items judged old were associated with two successive responses(R/K followed by a source judgment), whereas items judged newreceived only a single response, leading to a confound between judgedstudy status and response demands. The same confound is present inthe studies of Wais et al. (2010) and Smith, Wixted, and Squire (2011)(see below).

6. Recollection, familiarity, and retrieval-related hippocampalactivity

The findings described in the preceding two sections arestrongly convergent. They suggest that retrieval-related hippo-campal activity is sensitive neither to differences between testitems in their memory strengths, nor to whether or not an itemelicits a phenomenal sense of recollection (i.e., attracts an Rendorsement). Rather, hippocampal activity co-varies with theamount of information recollected about a study episode. Below,we discuss the implications of these findings for understandingthe role of the hippocampus in memory retrieval.

It might be argued that our conclusion that retrieval-relatedhippocampal activity is insensitive to memory strength is at oddswith our own data. In each experiment where the analysis waspossible, we found that recognition accuracy increased in tandemwith the amount of information recollected (Tables 3 and 4 andFigs. 2 and 3). Is it possible, then, that the hippocampus wasresponding to the strength of item memory rather than to theamount of information recollected? We think this possibility isunlikely for two reasons. First, item memory accuracy differedreliably between the R response category associated with theleast amount of recollected information and the K category, but inno case was there a concomitant difference in hippocampalactivity. This was so even in the study of Yu et al. (2012), wherethe recognition accuracies for R-weak and K-all trials (.92 and .77,respectively) spanned the range argued by Wixted et al. (2010) tobe sufficient to account for prior reports of recollection-relatedenhancement of hippocampal activity. The second reason forrejecting the possibility that our findings can be attributed todifferences in memory strength comes from a subsidiary analysisreported in Yu et al. (2012). Two sub-groups of subjects wereformed for whom recognition accuracy was equated between theR-high and R-mod response categories, and between the R-modand R-weak categories. In both sub-groups, hippocampal activitywas significantly greater when it was elicited by items from thecategory associated with the more confident source judgment.

The findings reported above are at variance with those reported intwo recent studies: Wais et al. (2010), and Smith et al. (2011). In Waiset al. (2010), equivalent levels of hippocampal activity were elicitedby test items attracting correct and incorrect source memory judg-ments when the items were equated for memory strength. Compar-ison of the levels of source accuracy in that study and the study of

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–30793076

Yu et al. (2012) described above suggest a possible explanation forthese divergent findings. In Wais et al. (2010), mean source accuracyfor test items recognized with high confidence was .67, and accuracyfor items recognized at the adjacent confidence level was .63. Thesetwo item classes were combined for the fMRI analyses, meaning thatsource accuracy would have fallen somewhere between these twovalues. Source accuracy levels in the study of Yu et al. (2012) wereconsiderably higher (.88 for R-high items). It is possible therefore thatthe failure of Wais et al. (2010) to detect effects of source accuracy inthe hippocampus reflected the relatively low levels of source recol-lection in that study. Smith et al. (2011) combined recognitionconfidence ratings on a 20-point scale with the R/K procedure. Theyreported that the hippocampal activity elicited by items endorsed Rexceeded the activity elicited by items given a K response whenmemory strength (operationalized by recognition confidence) wasnot equated, but that this difference was abolished when the R vs. Kcontrast was limited to items attracting the highest confidenceratings. Both classes of items did however elicit hippocampalresponses that exceeded the responses elicited by missed items.Smith et al. (2011) interpreted their findings as evidence for theinvolvement of the hippocampus in the retrieval of strong memories,regardless of whether the memory is supported by recollection orfamiliarity. An intriguing alternative possibility is that item memoryjudgments made with very high accuracy and confidence (andsubjectively experienced as strong familiarity) are supported by‘item-specific’ recollection rather than the familiarity signal thatsupports item memory more generally (J. Wixted, pers. comm. 12/08/2011; see also Wixted & Mickes, 2011). By this account, suchjudgments might be expected to depend on some of the same neuralsubstrates – perhaps including the hippocampus – as the process ofcontextual recollection that supports successful source memory. Tothe extent this account is valid, our proposal that encoding- andretrieval-related hippocampal activity index the amount of contextualinformation encoded in association with a study item will requirequalification.

As already noted, a consistent finding across the four experi-ments described above was that hippocampal activity did not differbetween items endorsed as R or K when the R judgment wasassociated with the retrieval of relatively little information about thestudy episode (the R-small and R-weak trials in Figs. 2 and 3). Thus,the endorsement of a test item as ‘recollected’ is not necessarilyassociated with enhanced activity relative to the activity elicited byitems judged as familiar only (at least at the sensitivity afforded byconventional fMRI). It is unclear how this null finding should beinterpreted. One possibility is that it is a reflection of the ‘processimpurity’ of the Remember/Know procedure. By this account, Rjudgments are made when memory strength (whether based uponrecollection, familiarity, or a combination of the two signals) exceedsa criterion level (Donaldson, 1996; Rotello, Macmillan, Reeder, &Wong, 2005). Thus, on average, there may be little or no differencebetween items endorsed as R or K in respect to the amount ofqualitative information retrieved. The finding that, when equated formemory strength, items endorsed as R are markedly more likely tobe associated with accurate source memory judgments than itemsendorsed as K (Wixted & Mickes, 2010; see also Ingram et al., 2012)indicates that this account does not hold for R judgments in general.It remains possible, however, that ‘near criterion’ R judgmentsreflect a combination of a weak recollection signal little differentfrom that associated with many items endorsed as K (though seebelow) and strong familiarity. Such judgments would be associatedwith relatively high memory strength and the retrieval of littlequalitative information, the characteristics associated with R-smalland R-weak judgments in the four experiments described above.

Alternatively, the failure of the hippocampus to discriminatebetween R-small/R-weak and K items may merely reflect theinability of the fMRI BOLD signal to detect differences in

hippocampal activity associated with recollection of only a smallamount of information. By this argument, subjects are capable ofdistinguishing accurately between trials on which recollection issuccessful or fails, responding R and K accordingly. However,whereas the retrieval of even a small amount of such informationis sufficient to boost recognition accuracy above that for items forwhich recollection fails, it is insufficient to elicit a detectablehippocampal response (cf. Squire et al., 2007). The currentlyavailable data do not allow adjudication between this and theforegoing possibility.

As was noted in the Introduction, the question whetherrecollection should be characterized as a thresholded or contin-uous process is a debated issue. According to one influentialmodel (Yonelinas, 2001), recollection is thresholded: unstudiedtest items almost never elicit a recollection signal, and studieditems either elicit a signal or fail to do so. An alternative model(Mickes, Wais, & Wixted, 2009; Wixted & Mickes, 2010; Ingramet al., 2012) argues that, like the signal supporting familiarity, therecollection signal is continuous: all test items elicit recollectionto some degree, and whether an item is endorsed as recollected(R) depends on the setting of a response criterion. At first sight,the behavioral and fMRI data from the four experiments describedabove appear to support the second of these two positions. Thebehavioral data clearly indicate that recollection is graded. Asdemonstrated in the experiments of Vilberg and Rugg (2007;2009a, 2009b), subjects can reliably report recollecting differingamounts of information across a series of test trials. And in thestudy of Yu et al. (2012), the accuracy and confidence of sourcememory judgments co-varied, as would be expected if thejudgments were supported by a continuously varying memorysignal (see also Mickes et al. (2009)). Moreover, as was describedabove, retrieval-related hippocampal activity co-varies with theamount of information recollected, a finding strongly indicative ofa graded memory signal. These findings are incompatible with an‘all or none’ model of recollection in which it is assumed thatrecollection either fails or is complete. As has been notedpreviously, however (e.g., Yonelinas et al., 2010), evidence thatrecollection is graded does not rule out threshold models ingeneral, since these models are not predicated on the all-or-noneassumption. Rather, the key assumption of threshold models isthat unstudied test items, along with some proportion of studieditems, fail to elicit a discriminating signal. An example of a modelof memory retrieval for which this assumption is a good approx-imation is provided by Norman and O’Reilly (2003); see alsoNorman (2010).

7. Hippocampal activity during continuous recognition

The great majority of studies of retrieval-related hippocampalactivity employed separate study and test phases. A few studieshowever have employed the continuous recognition procedure, whennew and old items are interleaved within a single list (Brozinsky,Yonelinas, Kroll, & Ranganath, 2005; Huijbers, Pennartz, & Daselaar,2010; Johnson, Muftuler, & Rugg, 2008; Suzuki, Johnson & Rugg,2011a, 2011b). In the simplest case, items are presented twice, withthe requirement to discriminate between first and second presenta-tions (see, for example, Brozinsky et al., 2005).

Below, we describe findings from three experiments from ourlaboratory that employed different variants of the continuousrecognition procedure. The experiments allowed investigation ofthe neural correlates of differences in the memory strength oftest items and of individual differences in the ability to retrievecontextual information associated with a repeated test item. Thefindings converge with those reported above from our experi-ments employing separate study and test phases: retrieval-related

L

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–3079 3077

enhancement of hippocampal activity is not a neural correlate ofincreased memory strength, but of successful contextual retrieval.

In the first two experiments (Johnson, Muftuler, & Rugg, 2008;Suzuki, Johnson, & Rugg, 2011a), items were presented a total offour times. The designs of the two studies were very similar,differing principally with respect to their response requirementsand stimulus materials. Specifically, whereas in Johnson et al.(2008) the task was simply to discriminate between old and newitems, in Suzuki et al. (2011a), the requirement was to discrimi-nate between test items on the basis of the number of presenta-tions (responding to first and third presentations on one key, andsecond and fourth presentations on another). The findings fromthe two studies were very similar. Despite the substantial‘strengthening’ of memory across the successive presentations(hit rates in Johnson et al. (2008) were .80, .94 and .98 for first,second and third repetitions respectively, against a false alarmrate of .06), no hippocampal voxels could be identified in eitherexperiment where activity increased as function of strength (incontrast to what was observed in cortical regions such as theprecuneus). Rather, as is illustrated in Fig. 4, hippocampal activitydecremented as a function of repetition.

We conjectured that the negative relationship between hippo-campal activity and memory strength observed in these studieswas a consequence of the failure of repeated items to engagerecollection. By this argument, the subjects in these studies reliedprimarily on differences in the familiarity of the test items toselect a response. In the absence of recollection, hippocampalactivity indexed the relative novelty of the eliciting items,reflecting the extent to which the items engaged processessupporting episodic encoding.

We assessed this proposal by employing a continuous recogni-tion procedure in which accurate performance depended onrecollection (Suzuki, Johnson, & Rugg, 2011b). Items comprisedpictures surrounded by a colored frame (orange, blue or gray). Thetask requirement was to respond ‘new’ to first presentations, and‘old’ to repeated items that were surrounded either by a grayframe, or a frame of the same color as when the item was firstpresented (‘targets’; i.e. orange–orange or blue–blue). Crucially,repeated items that were surrounded by a differently coloredframe (‘non-targets’; orange–blue or blue–orange) required a‘new’ rather than an ‘old’ response. Therefore, to avoid making a

Par

amet

er e

stim

ates

1st 2nd 3rd 4th

y = -20

L

-6

-4

-2

0

2

4

Fig. 4. The effects of item repetition on hippocampal activity in the study of

Suzuki et al. (2011a). Upper: Repetition-sensitive hippocampal regions. Lower:

across-subject mean parameter estimates for correctly endorsed items as a

function of number of presentations.

false alarm to non-target items it was necessary to use recollec-tion to ‘oppose’ their familiarity (Jacoby, 1991). Hence, thecontrast between the activity elicited by non-targets attractingcorrect versus incorrect responses should identify neural corre-lates of recollection. Subjects’ ability to perform the task variedwidely, with recollection performance (operationalized by thedifference between the probabilities of correct responses totargets and incorrect responses to non-targets; Jacoby, 1991)varying between .13 and .81. As is illustrated in Fig. 5, weidentified a right hippocampal cluster where recollection-relatedactivity co-varied across subjects with the behavioral estimate ofrecollection. Thus, consistent with the findings reported fromstudies employing separate study and test phases, successfulrecollection in continuous recognition is associated withenhanced hippocampal activity. Nonetheless, the different classesof studied item all elicited lower levels of hippocampal activitythan did unstudied items, in keeping with the findings of the twopreceding studies.

Clearly, the findings described in this section are incompatiblewith the idea that retrieval-related hippocampal activity is alwaysa positive function of memory strength. At least during continuousrecognition, increased memory strength can be associated withdecreased hippocampal activity. Similar results were reported inone other continuous recognition study (Brozinsky et al., 2005),and an analogous finding has been reported in several studies thatemployed separate study and test phases. In these studies, itemsattracting ‘new’ judgments (correct rejections or misses) elicitedlarger hippocampal responses than did correctly detected studieditems (e.g., Daselaar, Fleck, and Cabeza, 2006; Rugg, Henson, andRobb, 2003; Vilberg and Rugg, 2009c; Yu et al., 2012; see alsoFig. 2C). These findings have usually been interpreted as evidencefor the novelty-induced engagement of hippocampally-mediatedencoding processes (e.g., Duzel et al., 2003; Nyberg, 2005), con-sistent with the long-standing proposal that there is a bias favoring

Rec

olle

ctio

n-re

late

d ac

tivity

(con

trast

est

imat

es)

Probability of successful recollection

-10-8-6-4-202468

10

0.0 0.2 0.4 0.6 0.8 1.0

y = -10.5

Fig. 5. Hippocampal recollection effect in the study of Suzuki et al. (2011b).

Upper: Right hippocampal region where recollection-related activity co-varied

across-subjects with recollection performance. Lower: scatter plot illustrating the

relationship between the neural and behavioral indices of recollection. Each dot

represents a separate subject.

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–30793078

the encoding of relatively novel over relatively familiar items(Tulving & Kroll, 1995). Direct evidence in support of this inter-pretation of hippocampal ‘novelty effects’ was provided by Starkand Okado (2003), who reported that unstudied recognitionmemory test items that went on to be retrieved on a later memorytest elicited greater hippocampal activity than forgotten items.

The findings from our continuous recognition studies indicatethat the strengthening of the familiarity signal is associated with aqualitatively different hippocampal response than the response thatco-varies with the ‘strength’ of contextual recollection. The dissocia-tion in the direction of the two responses (a negative correlationwith familiarity, but a positive one with recollection) suggests thatthe construct of memory strength – as operationalized by accuracyof item memory – is of little utility in understanding the functionalsignificance of retrieval-related hippocampal BOLD activity.

As was just noted, findings of greater hippocampal activity forunstudied than for studied items are usually interpreted asevidence for the engagement of encoding processes by relativelynovel (unfamiliar) items. It is important to note that such anaccount does not rule out the possibility that the differentialhippocampal responses elicited by familiar and unfamiliar itemscould, in principle, provide a signal capable of supporting recog-nition memory judgments (though see Johnson et al., 2008) forreasons why this is unlikely). Were this to be the case, it wouldcall into question the proposal (e.g., Brown & Aggleton, 2001;Eichenbaum et al., 2007) that the hippocampus does not supportfamiliarity-based recognition. It would not, however, license theconclusion that the hippocampus is necessary for this form ofrecognition.

A final noteworthy point is that, at least at the spatial scaleafforded by fMRI, familiarity-driven decrements and recollection-related enhancements in hippocampal activity can co-exist in thesame region. For example, the hippocampal cluster illustrated inFig. 5, where activity co-varied with a behavioral index ofrecollection, also demonstrated robustly greater activity for newitems than for studied items. Analogously, Vilberg and Rugg(2009c) reported overlapping recollection and novelty effects inthe anterior hippocampus in a study that employed separatestudy and test phases (see also Fig. 2C). These findings raise thequestion of whether the overlap between hippocampal recollec-tion and novelty (or familiarity) effects exists at the level ofindividual neurons or neuronal circuits, and hence whether thesame neuronal populations support both the encoding andretrieval of episodic information.

8. Concluding comments

We have described data from a variety of studies relevant tothe question of whether the construct of memory strength issufficient to account for the enhanced encoding- and retrieval-related hippocampal activity that has typically been attributed tothe engagement of processes selectively supporting recollection.We have argued that the construct of strength is not sufficient;rather, hippocampal activity co-varies with the amount of con-textual or, more generally, episodic information that is encoded orretrieved, and not with the strength of an undifferentiatedmemory signal. This does not mean however that the constructof recollection provides the best way to understand hippocampalfunction. Indeed, our own findings consistently indicate that thephenomenal sense of recollection, as indexed by a Rememberjudgment, is not necessarily accompanied by a hippocampalresponse greater than the response elicited by items judged asfamiliar only (Know judgment). The role of the hippocampus inepisodic memory will likely be more completely understood inthe context of a computationally explicit model that specifies

when and how this structure contributes to recollection (similararguments have been made previously, e.g., Davachi, 2006; Dianaet al., 2007; Norman, 2010; Wixted and Squire, 2011). Among thefindings that any such model should accommodate is the sensi-tivity of hippocampal fMRI BOLD activity to the amount ofinformation that is encoded or retrieved about a study episode.

Acknowledgments

This research was supported by NIMH grants 2R01MH072966and 5R01MH074528.

References

Aggleton, J. P., Vann, S. D., Denby, C., Dix, S., Mayes, A. R., Roberts, N., et al. (2005).Sparing of the familiarity component of recognition memory in a patient withhippocampal pathology. Neuropsychologia, 43, 1810–1823.

Brown, M. W., & Aggleton, J. P. (2001). Recognition memory: what are the roles ofthe perirhinal cortex and hippocampus?. Nature Reviews Neuroscience, 2,51–61.

Brozinsky, C. J., Yonelinas, A. P., Kroll, N. E. A., & Ranganath, C. (2005). Lag-sensitiverepetition suppression effects in anterior parahippocampal gyrus. Hippocam-pus, 15, 557–561.

Cipolotti, L., Bird, C., Good, T., Macmanus, D., Rudge, P., & Shallice, T. (2006).Recollection and familiarity in dense hippocampal amnesia: a case study.Neuropsychologia, 44, 489–506.

Cohn, M., Moscovitch, M., Lahat, A., & McAndrews, M. P. (2009). Recollectionversus strength as the primary determinant of hippocampal engagement atretrieval. Proceedings of the National Academy of Sciences of the United States ofAmerica, 106, 22451–22455.

Daselaar, S. M., Fleck, M. S., & Cabeza, R. (2006). Triple dissociation in the medialtemporal lobes: recollection, familiarity, and novelty. Journal of Neurophysiol-ogy, 96, 1902–1911.

Davachi, L. (2006). Item, context, and relational episodic encoding in humans.Current Opinion in Neurobiology, 16, 693–700.

Davachi, L., Mitchell, J. P., & Wagner, A. D. (2003). Multiple routes to memory:distinct medial temporal lobe processes build item and source memories.Proceedings of the National Academy of Sciences of the United States of America,100, 2157–2162.

Diana, R. A., Yonelinas, A. P., & Ranganath, C. (2007). Imaging recollection andfamiliarity in the medial temporal lobe: a three-component model. Trends inCognitive Sciences, 11, 379–386.

Diana, R. A., Yonelinas, A. P., & Ranganath, C. (2010). Medial temporal lobe activityduring source retrieval reflects information type, not memory strength. Journalof Cognitive Neuroscience, 22, 1808–1818.

Donaldson, W. (1996). The role of decision processes in remembering andknowing. Memory and Cognition, 24, 523–533.

Duarte, A., Henson, R. N., & Graham, K. S. (2011). Stimulus content and the neuralcorrelates of source memory. Brain Research, 1373, 110–123.

Duzel, E., Habib, R., Rotte, M., Guderian, S., Tulving, E., & Heinze, H. J. (2003).Human hippocampal and parahippocampal activity during visual associativerecognition memory for spatial and nonspatial stimulus configurations. TheJournal of Neuroscience, 23, 9439–9444.

Eichenbaum, H., Yonelinas, A. P., & Ranganath, C. (2007). The medial temporal lobeand recognition memory. Annual Review of Neuroscience, 30, 123–152.

Eldridge, L. L., Knowlton, B. J., Furmanski, C. S., Bookheimer, S. Y., & Engel, S. A.(2000). Remembering episodes: a selective role for the hippocampus duringretrieval. Nature Neuroscience, 3, 1149–1152.

Fisher, R. A. (1950). Statistical methods for research workers (11th ed.). London:Oliver and Boyde.

Fortin, N. J., Wright, S. P., & Eichenbaum, H. (2004). Recollection-like memoryretrieval in rats is dependent on the hippocampus. Nature, 431, 188–191.

Good, M. A., Barnes, P., Staal, V., McGregor, A., & Honey, R. C. (2007). Context- butnot familiarity-dependent forms of object recognition are impaired followingexcitotoxic hippocampal lesions in rats. Behavioral Neuroscience, 121, 218–223.

Gottlieb, L. J., & Rugg, M. D. (2011). Effects of modality on the neural correlates ofencoding processes supporting recollection and familiarity. Learning andMemory, 18, 565–573.

Holdstock, J. S., Mayes, A. R., Roberts, N., Cezayirli, E., Isaac, C. L., O’Reilly, R. C., et al.(2002). Under what conditions is recognition spared relative to recall afterselective hippocampal damage in humans? Hippocampus, 12, 341–351.

Huijbers, W., Pennartz, C. M. A., & Daselaar, S. M. (2010). Dissociating the ‘‘retrievalsuccess’’ regions of the brain: Effects of retrieval delay. Neuropsychologia, 48,491–497.

Ingram, K. M., Mickes, L., & Wixted, J. T. (2012). Recollection can be weak andfamiliarity can be strong. Journal of Experimental Psychology Learning, Memory,and Cognition, 38, 325–339.

Jacoby, L. (1991). A process dissociation framework: separating automatic fromintentional uses of memory. Journal of Memory and Language, 30, 513–541.

M.D. Rugg et al. / Neuropsychologia 50 (2012) 3070–3079 3079

Jeneson, A., Kirwan, C. B., Hopkins, R. O., Wixted, J. T., & Squire, L. R. (2010).Recognition memory and the hippocampus: a test of the hippocampalcontribution to recollection and familiarity. Learning and Memory, 17, 63–70.

Johnson, J. D., Muftuler, L. T., & Rugg, M. D. (2008). Multiple repetitions revealfunctionally and anatomically distinct patterns of hippocampal activity duringcontinuous recognition memory. Hippocampus, 18, 975–980.

Kensinger, E. A., & Schacter, D. L. (2006). Amygdala activity is associated with thesuccessful encoding of item, but not source, information for positive andnegative stimuli. Journal of Neuroscience, 26, 2564–2570.

Kirwan, C. B., Wixted, J. T., & Squire, L. R. (2008). Activity in the medial temporallobe predicts memory strength, whereas activity in the prefrontal cortexpredicts recollection. The Journal of Neuroscience, 28, 10541–10548.

Mandler, G. (1980). Recognizing: the judgment of previous occurrence. Psycholo-gical Review, 87, 252–271.

Manns, J. R., Hopkins, R. O., Reed, J. M., Kitchener, E. G., & Squire, L. R. (2003).Recognition memory and the human hippocampus. Neuron, 37, 171–180.

Mayes, A. R., Holdstock, J. S., Isaac, C. L., Montaldi, D., Grigor, J., Gummer, A., et al.(2004). Associative recognition in a patient with selective hippocampal lesionsand relatively normal item recognition. Hippocampus, 14, 763–784.

Mayes, A. R., Montaldi, D., & Migo, E. (2007). Associative memory and the medialtemporal lobes. Trends in Cognitive Sciences, 11, 126–135.

Mickes, L., Wais, P. E., & Wixted, J. T. (2009). Recollection is a continuous process:implications for dual-process theories of recognition memory. PsychologicalScience, 20, 509–515.

Montaldi, D., Spencer, T. J., Roberts, N., & Mayes, A. R. (2006). The neural systemthat mediates familiarity memory. Hippocampus, 16, 504–520.

Montaldi, D., & Mayes, A. R. (2010). The role of recollection and familiarity in thefunctional differentiation of the medial temporal lobes. Hippocampus, 20,1291–1314.

Norman, K. A., & O’Reilly, R. C. (2003). Modeling hippocampal and neocorticalcontributions to recognition memory: a complementary-learning-systemsapproach. Psychological Review, 110, 611–646.

Norman, K. A. (2010). How hippocampus and cortex contribute to recognitionmemory: revisiting the complementary learning systems model. Hippocampus,20, 1217–1227.

Nyberg, L. (2005). Any novelty in hippocampal formation and memory? CurrentOpinion in Neurology, 18, 424–428.

Otten, L. J. (2007). Fragments of a larger whole: retrieval cues constrain observedneural correlates of memory encoding. Cerebral Cortex, 17, 2030–2038.

Ranganath, C., Yonelinas, A. P., Cohen, M. X., Dy, C. J., Tom, S. M., & D’Esposito, M.(2004). Dissociable correlates of recollection and familiarity within the medialtemporal lobes. Neuropsychologia, 42, 2–13.

Rotello, C. M., Macmillan, N. A., Reeder, J. A., & Wong, M. (2005). The rememberresponse: subject to bias, graded, and not a process-pure indicator ofrecollection. Psychonomic Bulletin and Review, 12, 865–873.

Rugg, M. D., Henson, R. N. A., & Robb, W. G. (2003). Neural correlates of retrievalprocessing in the prefrontal cortex during recognition and exclusion tasks.Neuropsychologia, 41, 40–52.

Smith, C. N., Wixted, J. T., & Squire, L. R. (2011). The hippocampus supports bothrecollection and familiarity when memories are strong. The Journal of Neuroscience,44, 15693–15702.

Song, Z., Jeneson, A., & Squire, L. R. (2011). Medial temporal lobe function andrecognition memory: a novel approach to separating the contribution ofrecollection and familiarity. The Journal of Neuroscience, 31, 16026–16032.

Squire, L. R., Wixted, J. T., & Clark, R. E. (2007). Recognition memory and the medialtemporal lobe: a new perspective. Nature Reviews Neuroscience, 8, 872–883.

Stark, C. E. L., & Okado, Y. (2003). Making memories without trying: medialtemporal lobe activity associated with incidental memory formation duringrecognition. The Journal of Neuroscience, 23, 6748–6753.

Suzuki, M., Johnson, J. D., & Rugg, M. D. (2011a). Decrements in hippocampalactivity with item repetition during continuous recognition: an fMRI study.Journal of Cognitive Neuroscience, 23, 1522–1532.

Suzuki, M., Johnson, J. D., & Rugg, M. D. (2011b). Recollection-related hippocampalactivity during continuous recognition: a high-resolution fMRI study. Hippo-campus, 21, 575–583.

Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26, 1–12.Tulving, E., & Kroll, N. (1995). Novelty assessment in the brain and long-term

memory encoding. Psychonomic Bulletin and Review, 2, 387–390.Uncapher, M. R., & Rugg, M. D. (2005). Encoding and the durability of episodic

memory: a functional magnetic resonance imaging study. Journal of Neuroscience,25, 7260–7267.

Vilberg, K. L., & Rugg, M. D. (2007). Dissociation of the neural correlates ofrecognition memory according to familiarity, recollection, and amount ofrecollected information. Neuropsychologia, 45, 2216–2225.

Vilberg, K. L., & Rugg, M. D. (2009a). Left parietal cortex is modulated by amount ofrecollected verbal information. NeuroReport, 20, 1295–1299.

Vilberg, K. L., & Rugg, M. D. (2009b). Functional significance of retrieval-relatedactivity in lateral parietal cortex: Evidence from fMRI and ERPs. Human BrainMapping, 30, 1490–1501.

Vilberg, K. L., & Rugg, M. D. (2009c). An investigation of the effects of relativeprobability of old and new test items on the neural correlates of successful andunsuccessful source memory. Neuroimage, 45, 562–571.

Wais, P. E., Squire, L. R., & Wixted, J. T. (2010). In search of recollection andfamiliarity signals in the hippocampus. Journal of Cognitive Neuroscience, 22,109–123.

Wais, P. E., Wixted, J. T., Hopkins, R. O., & Squire, L. R. (2006). The hippocampussupports both the recollection and the familiarity components of recognitionmemory. Neuron, 49, 459–466.

Wixted, J. T., & Mickes, L. (2010). A continuous dual-process model of remember/know judgments. Psychological Review, 117, 1025–1054.

Wixted, J. T., & Mickes, L. (2011). Remember/know judgments in free recall.Abstracts of the Psychonomic Society, 16, 30.

Wixted, J. T., Mickes, L., & Squire, L. R. (2010). Measuring recollection andfamiliarity in the medial temporal lobe. Hippocampus, 20, 1195–1205.

Wixted, J. T., & Squire, L. R. (2011). The medial temporal lobe and the attributes ofmemory. Trends in Cognitive Science, 15, 210–217.

Yonelinas, A. P. (2001). Components of episodic memory: the contribution ofrecollection and familiarity. Philosophical Transactions of the Royal Society ofLondon Series B: Biological Sciences, 356, 1363–1374.

Yonelinas, A. P. (2002). The nature of recollection and familiarity: a review of 30years of research. Journal of Memory and Language, 46, 441–517.

Yonelinas, A. P., Aly, M., Wang, W. C., & Koen, J. D. (2010). Recollection andfamiliarity: examining controversial assumptions and new directions. Hippo-campus, 20, 1178–1194.

Yonelinas, A. P., Otten, L. J., Shaw, K. N., & Rugg, M. D. (2005). Separating the brainregions involved in recollection and familiarity in recognition memory. TheJournal of Neuroscience, 25, 3002–3008.

Yonelinas, A. P., & Jacoby, L. J. (1996). Noncriterial recollection: Familiarity asautomatic irrelevant recollection. Consciousness and Cognition, 5, 131–141.

Yu, S. S., Johnson, J. D., & Rugg, M. D. (2012). Hippocampal activity duringrecognition memory co-varies with the accuracy and confidence of sourcememory judgments. Hippocampus, 22, 1429–1437.