In: The Oxford Handbook of Memor ~ .E. Tulving & F.l.M. Craik (E s.).New York: Oxford University ress.2000.

Nonconscious Forms of Human Memor'( J

JEFFREY P. TOTH

s~med that consciousness did not playa largero~e in the production ,of thought and behav-iot. And even if this vieiw was not universallyh~ld (by psychoanalysts, for example), fewm~thods existed for enlpirically approachingth~ problem.

! This all changed in the early 1980s, how-ever, when memory and consciousness cametogether in such a wa~r as to fundamentallyalter the landscape of cognitive research. Ineslsence, it was disco"ered that a person'sthought and behavior could be influenced byprior events of which that person was notaware; and that the nature of such noncon-scious mnemonic influlences-the principlesby which they operated-differed significant-ly from the more conscious forms of memorythat researchers had traditionally investigated.These findings were So important that todayany serious account of memory must acknowl-edge the dramatic difference that obtains be-tween "memory with awareness" and "mem-ory without awareness." So, too, observeddifferences in these two' kinds of memory havebeen influential in making consciousness andawareness respectable topics of inquiry inpsychology and neuros(~ience (Barrs, 1988; Co-hen & Schooler, 1997; Picton & Stuss, 1994).

The purpose of this chapter is to providean overview of nonconscious forms of humanIn,emory. The first seclion ("Phenomena andPhenomenology") provides more background

"The relationship betwet3n memory and con-sciousness is one of Ithe most fascinatingaspects of cognitive research. However, it isonly recently that students of psychology havetruly appreciated the intimate connection be-tween these two aspects of mental life,

Systematic research on memory began withEbbinghaus's (1885/1964) pioneering work onmemory for nonsense syllables and has sinceprogressed through a D.umber of phases in-cluding studies of the Jlearning and memorycapabilities of nonhuman animals (e.g" Hull,1943; Tolman, 1932), operant-behaviorist con-ceptualizations of memory in terms of stim-ulus control (Skinner, 1969), and neo-behav-iorist studies of word-list learning and themechanisms of forgetting (Underwood, 1957).These ideas set the stage for modern cognitiveconceptions of memory whereby prior experi-ences are viewed as mental representations,encoded, stored, and retrieved in a human in-

formation-processing sy:stem.In contrast to research on memory, research

on consciousness and awareness, although nev-er completely absent (e,g" Adams, 1957), hasbeen relatively sparse s:ince the early specu-lations and introspective studies of psychol-ogists such as Wundt, Titchener, and James,Although it is probably safe to say that psy-chologists have never doubted the existence(or at least the problem) of consciousness,much of twentieth-century psychology as-

'1245

MEMORY IN THE LABORATOFtY246

information and examples, essentially firmingup the case that the qualitative distinction ~e-tween conscious Imd nonconscious memory lisa necessary one. This first section also intrp-duces some of the terminology and researc!:hdesigns used in the field. The second secti~n("Paradigms") describes research methods Inmore detail and also provides an overview ofmajor empirical findings. The final secti<jln("The Future of Nonconscious Memo~")briefly notes CritilGal issues for future rese;J;hand describes ho'w ideas about nonconscio~smnemonic proces:ses are beginning to be in~e-grated with other aspects of cognition, as w,llas with the study of the brain.

The reader may note the above organizati~ndoes not include sections entitled "Finding~"or "Theories." The reason for this is that thereare simply too many relevant empirical res~tsand theoretical positions to be includedwithin this short chapter. More extensivereviews of the ,empirical literature can befound in Kelley and Lindsay (1996); Roediserand McDermott (1993); Moscovitch, Vriez~n,and Goshen-GottsteiD (1993); and Richardson-Klavehn and Bjork (1988). Readers interest~din large-scale theoretical accounts of nonco~-scious processes in memory are directed '

Ito the following sources: Jacoby, Yonelinas, &

Jennings, 1997; 1vfoscovitch, 1994; Ratcliff &McKoon, 1996; F~oediger, Weldon, & Challls,1989; Schacter & Tulving, 1994a; Toth & H~t,1999. I

Phenomena and

Phenomenolog)'

How might one el;tablish that consciousness i isa necessary conc:ept in understanding me~-ory? Perhaps the single most important sourceof evidence for allswering this question comesfrom research with amnesic patients. Damageto the medial teDJ.porallobes, usually involv-ing the hippoc8Inpus and related structur,s,produces a profound inability to consciou~lyrecollect one's past. Patients with such d$-age can participa1:e in common events, such lasreading a list oj: words or conversing withtheir doctor, yet be totally unable to later te-call or recognize those prior events. IndeEId,for the majority of this century it was widelybelieved that suc:h patients simply could qotcreate {encode or store) a long-term record loftheir experience. But starting with studies byWarrington and Weiskrantz {1968; see also

~I,

Rozin, 1976) this view was shown to be incor-rect. If amnesic ]patients are given a task thatis related to their prior experiences, but whichmakes no reference to those experiences-such as simply rereading the list of words orcompleting fragDlented forms of those words(me -o -y)-these patients show incontro-vertible evidence that their prior experienceswere indeed recorded and subsequently usedto influence theiJr performance.

Amnesia, then, provides a crucial piece ofevidence that memory can occur withoutawareness: the aJnnesic patient fails to experi-ence the phenoJ[Ilenology of remembering-she believes she is not remembering-but herperformance shows that memory is indeed be-ingused.

The flip side lof this memory-awareness re-lation is also av~rllable. Patients with damageto frontal lobes, often including the basal fore-brain, demonstrate a phenomenon known asconfabulation in which they relate events oftheir past, and steadfastly assert their reality,when in fact tho~;e events never happened (seeKopelman, 1987; Moscovitch, 1989). The con-fabulating patient thus exhibits the converseof the amnesic patient's unaware form ofmemory: he experiences the phenomenologyof remembering-believes he is remember-ing-but historic:al records, such as those pro-vided by relatives, show that no memory ex-ists.

Amnesic and confabulating patients thusindicate a complex relation between, and dis-sociation of, the retention and expression ofprior experience on the one hand, and a phe-nomenal awareness of that experience on theother. Such patilents are fascinating becausethey suggest that to understand memory onemust also under:;tand the phenomenal aware-ness of remembering. But two more issuesmust be addressed before the relation betweenmemory and awllreness can be viewed as fun-damental. First, it must be shown that simi-lar phenomena (i.e., memory/awareness dis-sociations) can occur in persons without braindamage. And :;econd, that memory withawareness and memory without awareness op-erate according to different rules or principles.As described nE~xt, there is strong evidencesupporting both of these points.

Imagine that ~l normal (non-brain-damaged)individual is prlesented with a list of words,half visually, h~uf aurally. Moreover, withineach presentation modality, the person makesa superficial judlgment for half of the words,such as countin~: the number of vowels, and a

IIIIII~

1111111

NONCONsCIOUs FORMS pF HUMAN MEMORY 247

~ more meaning-based judgment for the otherhalf, such as decidinl~ whether the word refersto a pleasant or unpleasant concept. If onewere to assess this person's memory with an"explicit" (conscious] test of memory such asrecognition or recall, it is highly likely that theindividual would show superior memory forthe meaningfully processed words, but thatthe modality of ori~:inal presentation wouldhave little impact on performance (figure 16.1).

Consider now a different test in which thesubject is shown ~rord stems (e.g., tru-),some of which can be completed with pre-viously encounterecl words, but is simplyasked to complete E~ach stem with the firstword that comes to IJllind. When tested in this"implicit" way (figurE~ 16.2), performance looksdramatically different from that found withexplicit tests. In particular, the meaningful-ness with which words were encoded has noinfluence on performance, but the modality ofinitial presentation has a large effect. Note thatperformance in such implicit tests is assessedrelative to a baselinE~ condition for which nostudy items were presetJlted; the increase in

performance for studied over nonstudied(baseline) items is clllled "priming."

Dissociations such as that shown in figures16.1 and 16.2 have now been firmly estab-lished for a large number of experimental ma-nipulations (see Roediger & McDermott, 1993)and thus make a stIong case that consciousand nonconscious forms of memory operateaccording to different principles.

How do we know that the person tested im-plicitly is not const::iously aware that his orher primed responses came from the prior en-coding event (or tha:t such awareness is not anecessary condition f-or the pattern of perfor-mance obtained)? This is a difficult issue thatsets the stage for mllch of the recent researchin the field (see below). But despite this diffi-culty, the bulk of evidence shows that neitherawareness of a prior event nor an intent to re-member that event is a necessary condition forprior experience to exert a significant influ-ence on performan(::e. Much of the researchcited in the following sections shows this tobe the case. Here, however, two phenomenacan demonstrate the basic argument.

1.00

0.90'C~'ia 0.80()!0 0.70

E'CG) G)~ N 0.60'C-G) C-D)'C O 0.50~ ()...G)0 ..-..0.40o 0

C.2 0.30t:og. 0.20..Q.

0.10

0.00

"~eanlngfulSuperficial

Type of encoding

Figure 16.1 Idealized effects of semantic elaboration and study modality on an explicit test of

memory.

248 MEMORY IN THE LABORATORY

1.00

0.90'CQ)u~

'CO...0.

(I)EQ)

~

'CQ)

'C~

-(I)

-o

co

:e00.O...

Q.

0.80

0.70

0.60

0.50

0.40

0.30

0.20

0.10

0.00

" Superficial Meaningful

Type of encoding

Figure 16.2 Idealized effects of semantic elaboration and study modality on a visualceptual") implicit test of memory.

per

First, consider what is known as stochasticindependence between memory tests. In stud-ies examining stochastic independence, a sub-ject is, for example, shown a list of words andthen given two consecutive memory tests. Oneof these tests (e.g., word-fragment completion)is implicit such thclt, although some of thefragments can be completed with previouslypresented words, th'B subject is simply askedto complete each fragment without referenceto any prior experience (with "the first wordthat comes to mind"). In the other test, thesubject is asked tlD consciously recognizewords from the original list, some of which arethe same as those tested on the fragment test.By examining the relation between the twotests, one can ask whether performance is cor-related-that is, whl~ther recognizing a partic-ular word predicts one's ability to complete afragment with that word, or vice versa. In fact,results show that thl~re is often no relation be-tween performance on the two tests (i.e., thetests show stochastic independence); recog-nizing a word tells .LIS nothing about whetherthe subject will use that word to complete a

fragment; and completing a fragment with astudied word tells us nothing about whetherthe subject will recognize that word as beingon the study list Uacoby & Witherspoon, 1982;Tulving, Schacter, 8: Stark, 1982).

Stochastic independence thus shows thatconscious memory of a particular event is nota necessary condition for that event to influ-ence performance, l\n even more dramatic il-lustration of this point comes from studiesmanipulating the amount of attention avail-able for encoding an event. If subjects areasked to read a list of words while at the sametime performing a difficult secondary task(such as monitoring a series of digits for a par-ticular sequence), subsequent conscious mem-ory (e.g" recall) will be much impaired rela-tive to the case when the subject devotes fullattention to study. In contrast, implicit tests ofmemory such as stem and fragment comple-tion show little or no effect of such attentionalmanipulations (see .figure 16.3). Such dissocia-tions as a function of attention have been pro-duced with a variet~v of tests and measurementtechniques (Parkin, Reed, & Russo, 1990; Ja-

II

NONCONSCIOUS FORMS OF HUMAN MEMORY 249

1.00

Explicit

...0-cn 0.90G)-

."fi i' 0.80= G)Q,-E ~ 0.70

.-()-.-'CQ.G) ~ 0.60()-~'C'C0 G) 0.50...NQ,.- c~ D) 0.40

G)~-.0-~

0.30

0.20o

:e0c.0...c.

0.10

0.00

Full Divided

Level of attention at enco,ding

Figure 16.3 Idealized effects of full and divided attention on implic:it and explicit tests of

memory.

coby, Toth, & Yonelinas, 1993) and furtherstrengthen the case that memory can occur inthe absence of conscious awareness.

Stochastic independence and the effects ofdivided attention make a strong case that con-scious and nonconscious forms of memory aredistinct, qualitatively different ways in whichthe past can influence the present. A skepticmight further insist, however, that in bothcases, the subject is aware of the earlier eventand, more importaJlt, of the fact that the testbears some relation to that event. Even theseworries can be addIessed, however, as isshown in the following sections.

"

Paradigms

Mere-Exposure

How might one guaIantee that subjects cannotbecome aware of, or attempt intentional re-trieval of, a prior event? One method is topresent stimuli belo'w the threshold for aware-

ness. The notion hlere is that, if subjects areunaware that an item has been encoded, theycannot subsequently become aware of the rela-tion between a later memory test and thatprior encoding experience. This is exactly thestrategy employed in the mere-exposure para-digm in which subjects are exposed to stimuli,such as randomly gl3nerated geometric shapes,for a "mere" 1 millisecond (one one-thou-sandth of a seconcll). As you might expect,stimuli "shown" at jthis duration are very diffi-cult to see; and, in f;act, subjects in such exper-imental conditions claim to have not seenanything (except a l[)lank screen) and later areunable to recognize any of the briefly flashedshapes. Yet when later shown pairs of shapesand asked to choose the shape they "likebest," previously flashed shapes are chosensignificantly more often than shapes that werenot presented earlier (Kunst-Wilson & Zajonc,1980; Seamon et al., 1995).

Results from mere-exposure studies thusshow that subjects need not even be aware

I

250 MEMORY IN THE LABORATORY

that a prior event occurred for that event toinfluence their subsequent performance. How-ever, mere-exposure studies are limited be-cause, if subjects are unaware that an encod-ing event has occuned, they cannot act on itor process it in different ways. And gettingsubjects to process events in various ways iscrucial to determining the different operatingcharacteristics of conscious and nonconsciousforms of memory. Thus, paradigms discussedin the following sections use the more typicalprocedure of showlng items at easily perceiv-able levels, having subjects process thoseitems in different ways, and then employingvarious test conditions designed to elicit non-conscious forms of memory.

tasks, such as learning a new perceptual-mo-tor skill.

Most of what is ]mown about nonconsciousforms of memory is based on implicit tests. Asnoted earlier, this (;hapter will not attempt todescribe all of the relevant findings. However ,the major conclusions of this research can besummarized briefl~f: On perceptual implicittests, nonconsciouf; mnemonic processes ap-pear insensitive to conceptual manipulations(i.e., they appear "plresemantic") and yet showhigh sensitivity to changes in surface formfrom study to test (~iee figure 16.2). In contrast,conceptual tests rt3veal the opposite set ofcharacteristics, being insensitive to changes insurface form, but showing sensitivity to con-ceptual-encoding manipulations such as levelof processing (but see Vaidya et al., 1997).

The perceptual/1::onceptual theme runningthrough the above ,conclusions, as well as ta-ble 16.1, has becOmE3 central in theorizing aboutnonconscious mnemonic processes. Indeed,this distinction is the cornerstone of both"process" theories of memory (Roediger, 1990;see also Jacoby, 1983; Kolers & Roediger,1984) as well as more .'structural" Iilultiple-memory systems approaches (see Tulving &Schacter, 1990; Schlacter & Tulving, 1994a). Itis important to note, however, that the percep-tion/conception di:stinction, while undoubt-edly important, is c::learly insufficient to cap-ture the most cfllcial aspect of implicit/explicit memory d:issociations-namely, thatin one case, the sub,ject is aware of remember-ing a past event, w]i1ile in the other case suchawareness is absenjt (or at least unnecessary).This has led some researchers to deemphasizethe perception/conception distinction in fa-vor of distinctions more closely related to thephenomenological status of conscious andnonconscious mem.ory. For example, Jacoby(1991), noting relations between nonconsciousmemory effects and similar phenomena in thefield of attention, has advocated a distinctionbetween mnemonic: control and automaticity(see also Logan, 19190). Jacoby's view is alsosupplemented by a sophisticated analysis ofhow subjects infer jthe role of memory-thus,effectively constructing their subjective expe-rience of remembering-on the basis of theirperformance and relevant task demands (seeJacoby, Kelley, & Dywan, 1989). This attribu-tional approach is discussed more fully below.

Further questioning the usefulness of a per-ception/conception distinction, Masson andMcLeod (1992) hav,e conceptualized implicit!explicit dissociatioJlls in terms of the differ-

Implicit Tests

The most popular way of eliciting noncon-scious memory is with implicit tests such asthe stem- and fragment-completion tests de-scribed above. The defining feature of an im-plicit memory test is si1I1ply the instructions:subjects are told to respond to test stimuli(e.g., produce a word, classify an object, solvea puzzle) without reference to past experience.Although there has been a good deal of debateas to whether such instructions are a sufficientbasis on which to infer nonconscious memory(see Jacoby, 1991; Reingold & Merikle, 1990;Toth, Reingold, & Jacoby, 1994), there is noquestion that such tests show striking dissoci-ations (i.e., different patterns of performance)in comparison with more traditional (con-scious/explicit) tests of memory.

Table 16.1 provides a relatively compre-hensive list of the many implicit tests cur-rently in use, although new tests are inventedevery year. Note tb.at the organization of thistable-based on a distinction among percep-tual, conceptual, and procedural processes-is not theoretically neutral. Nevertheless, itallows an extensive classification scheme thatcaptures most, if not all, of the tests available:perceptual tests, stlch as word identificationand fragment completion, are those that chal-lenge the various perceptual systems in someway-for example, with a fast presentation ora degraded/fragmented stimulus. Conceptualtests, in contrast, employ retrieval cues suchas category labels or general-knowledge ques-tions designed to elicit responses that aremeaningfully related to these cues. Finally,procedural implicit tests require subjects toperform extended sensory, motor, or cognitive

1:

Table 16.1 Implicit l:ests of memory (with representative references).

Jacoby & Dallas ('L 981)Graf et al. (1982)Tulving et al. (1982)Duchek & Neely 1:1989)Hashtroudi et al. (1988)Srivinas & Roedil~er (1990)Jacoby & Witherspoon (1982)Witherspoon & Allan (1985)

Shimamura & Squire (1984)Blaxton (1989)Kelley & Lindsay' (1993)Vaidya et al. (1997)Vaidya et al. (19'~7)Jacoby et al. (19~19b)Begg et al. (1992)

Adams et al. (19:B8)Knowlton & Squire (1994)Levy & Kirsner (1989)Kolers (1976)Reber (1989)

Verbal testsPerceptual

Word ("perceptual") identificationWord stem completionWord fragment completionLexical decisionWord namingAnagram solutionHomophone spellingTemporal (duration) judgments

ConceptualWord associationCategory-instance (exemplar) generationFact (general k31owledge) generationCategory verifi(::ationCategorization! classification judgmentsFame judgmentsTruth! credibility judgments

ProceduralWord-based problem solvingProbability judgmentsReading normal textReading altered (e.i-, mirror inverted text)Artificial grammar learning

Weldon & Roediger (1987)Mitchell & Bro~'ll (1988)Kroll & Potter (1984)Schacter et al. (~l990)Jacoby et al. (19B8)Nilsson et al. (1'992)Kunst-Wilson & Zajonc (1980)Benzing & Squire (1989)Bechara et al. (1995)

~

Smith & Bransc,omb (1988)no published studiesno published studies~

Non-verbal testsPerceptual

Identification of fragmented picturesPicture namingObject/non-object decisionPossible/impossible object decisionNoise judgmentsAction-decision test for dynamic events

Aesthetic/preference judgmentsSensory conditioningAffective conditioning

ConceptualPerson/trait attributions

Object categorizationCommon-uses decisions for objects

ProceduralSerial reaction timeControl of complex systemsMotor trackingMirror drawingJigsaw puzzlesTower of HanoiMaze learningMotor conditioning

~

Nissen & Bullelner (1987)Berry & Broadblent (1984)Corkin (1968)Milner (1962)Brooks & Baddeley (1976)Cohen et al. (1985)Milner (1965)Weiskrantz & Vvarrington (1979)

,\

MEMORY IN THE LABORATORY252

series of stimuli that embody a complex un-derlying rule structure that is not readily ap-parent. For examplt3, subjects may see stimuli(e.g., VTTRM) that ;appear to be a random col-lection of letters but which are actually gener-ated by a complex J:Ule system or grammar. Inlater transfer (implicit) tests, subjects are ableto discriminate "le!~al" from "nonlegal " letter

strings (i.e., those conforming vs. not conform-ing to the underlying grammar) despite beingunable to describe jthe basis for their decision.Similarly, when reacting to a series of eventsthat appear random (e.g., a sequential patternof lights) but which actually conform to com-plex transition ruLes, reaction times becomefaster despite the subjects' inability to de-scribe the underlying rule (e.g., Nissen & Bul-lemer, 1987). In general, implicit learningappears to be mosl: similar to procedural im-plicit memory test:~ and it is likely that bothphenomena draw on similar neurocognitivemechanisms (see Curran, 1998).

As noted above. one problem with studiesemploying implicit tests is the possibility thatsubjects are aware IJf, and potentially trying toretrieve, prior events. In the context of im-plicit/explicit test comparisons, the retrievalintentionality criterion is one way to addressthis problem. The next three sections describealternative approal::hes to solving the vexingissue of awareness and intent in the study ofnonconscious memory .

ence between an initial "interpretive encod-ing" that involves an integration of perceptualand conceptual components, and a second,more elaborative processing phase that is pre-dominantly conceptual. And Toth and Rein-gold (1996; see also Toth & Hunt, 1999) havestressed the importance of context-goal config-urations in understanding both conscious andnonconscious forms of memory. It is likelythat all of these ideas will figure in more ma-ture theories of memory.

A potential problem with comparisonsamong implicit and explicit tests concerns thedifferent cues they provide for access to mem-ory. Although implicit/explicit dissociationsare often taken to support a qualitative differ-ence in memory, and thus to reveal the operat-ing characteristics of nonconscious mnemonicprocesses, it is widely known that even differ-ent conscious tests of memory can show disso-ciations as a function of the retrieval cues theyprovide. As one classic example, recall andrecognition dissociate as a function of fre-quency in the language: high-frequency wordsare more likely to be re(ialled whereas low-fre-quency words are more likely to be recog-nized. In extending such observations to im-plicit memory , the worry is that implicit/explicit dissociations occurring between teststhat provide different retrieval cues (e.g., frag-ment completion and recognition) may not re-flect differences in awareness or intent butrather differences in the retrieval properties ofthe test cues (DUrul & Kirsner, 1989).

An important development in this regard isa method known as the retrieval intentionalitycriterion (Schacter, Bowers, & Booker, 1989),which recommends providing identical cuesat test (e.g., word stems) and varying only theinstructions given to the subject (implicit orexplicit). If one finds qualitative differences inperformance as a function of instructions, onecan then make a much stronger case thatawareness (or intent) was the crucial factor,rather than the nature of the retrieval cues. Al-though the logic underlying this strategy is notunassailable (see Toth et al., 1994; Reingold &Toth, 1996) there is no question that it repre-sents an important methodological advance inthe study of implicit memory.

Before leaving implicit tests, it is worthnoting another line of research-implicitlearning-that, while proceeding somewhatindependently, effectively employs the samestrategy as that underlying implicit memorytests (see Reber, 1989; Seger, 1994). In studiesof implicit learning, subjects are exposed to a

Self-Report

One strategy for il;olating nonconscious pro-cesses in memory is simply to ask subjectswhether they werle aware of, or tried to re-trieve, prior eventl; during performance of animplicit test. Altllough this strategy is notwidely utilized, the available research sug-gests two major (:onclusions: First, subjectsare often aware tllat items presented or pro-duced on an implicit test were previously en-countered in an ealflier task (e.g., Richardson-Klavehn, Lee, Joubran, & Bjork, 1994). Second,and as might be expected, subjects who areaware of the relation between an encodingevent and a subsequent test are more likely toattempt intentional retrieval of items from thatencoding event (Bowers & Schacter, 1990;Toth et al., 1994).

Given these pcltential problems, some re-searchers (e.g., Roediger & McDermott, 1993)have recommended disguising implicit teststhrough the use of long study and test lists,test lists with a hi~~h ratio of new to old items,

It NONCONSCIOUS FORMS OF HUMAN MEMORY 253

.

;

filler tasks presented before the implicit test,and instructions that encourage fast responsesto each test cue. Such techniques may indeedhelp disguise the nature of an implicit test, butthey do so at the expense of being able to inves-tigate variables that may turn out to be criti-cal for understanding nonconscious mnemonicprocesses (e.g., long vs. short lists, high vs. lownew/old ratios, fast vs. slow responses).

Self-reports themselves are not withoutproblems. For one, subjects cannot be asked toreport on their level of awareness (or intent)for each test item, at least in experiments thatattempt to disguise the relation between studyand a later implicit test. This means that self-reports of awareness must be assessed retro-spectively (after the test is completed), a pro-cedure that depends on the accuracy ofmemory for prior states of awareness and in-tent. Perhaps even more problematic is thequestionable validity of self-reports becausetest demands and subsequent questions abouttest awareness may interact in complex ways.Implicit test instructions effectively tell sub-jects not to intentionally.retrieve prior studyitems. If subjects are later asked whether theyused intentional memory , answering yes istantamount to saying that they ignored priorinstructions. Subjects may be reluctant to dothis, thus falsely confirming an experimenter's

expectations.Despite these problems, the available stud-

ies using self-reports suggest that test aware-ness does not fundamentally alter the basic ex-pression of nonconscious (or unintentional)memory (Bowers & Schacter, 1990; Richard-son-Klavehn et al., 1994). Additional researchusing self-report methods, as well as varia-tions on these methods-such as the remem-ber/know technique developed by Tulving(1985)-may provide important insights intothe nature of nonconscious forms of memory,as well as how the influences of these forms ofmemory are "interpreted" by more consciousprocesses (see chapter 15; and Jacoby, 1998).

stead of asking subjects to complete stemswith studied words, or to respond with thefirst words that come to mind, subjects aretold to complete the stems only with wordsthat were not studied. If, in contrast to instruc-tions (and thus conscious intent), subjects com-plete stems with previously studied words, onehas gained strong evicLence for the operation ofnonconscious memory because consciousmemory would have resulted in the word be-ing excluded as a test response.

Using the oppositJlon procedure as a start-ing point, Jacoby (1991) developed a moreelaborate process-dissociation procedure de-signed to quantify tile strength of consciousand nonconscious foJms of memory .The pro-cedure involves combining results from theopposition (or exclusj'on) condition, with thosefrom an inclusion condition in which subjectsare told to use old words to complete teststems. By casting these two conditions intoequations that embocLy the logical relation be-tween conscious ancL nonconscious memory,one should be able to combine performance inthe two conditions so as to analytically esti-mate the separate contributions of the twokinds of memory .AD.d, indeed, mnemonic es-timates gained from this procedure have pro-duced a number of interesting dissociations.For example, divided attention Uacoby, Tothet al., 1993), speed of response (Toth, 1996a;Yonelinas & Jacoby, 1995), elaborateness ofstudy processing (Toth et al., 1994), agingUennings & Jacoby, 1993, 1997), and traumaticbrain injury (Ste-Malie, Jennings & Finlayson,1996; Toth, 1996b) a:ll have been shown to af-fect estimates of conscious memory , but toleave estimates of nonconscious memory un-changed. Variations on the procedure haveshown the opposite pattern-effects on non-conscious, but not conscious, memory (Hay &Jacoby, 1996). Importantly, many of these re-sults agree with those from implicit/explicittest comparisons (e.g:., Toth et al., 1994; Rein-gold & Goshen-Gotts1tein, 1996).

Subjects in opposition and process-dissoci-ation experiments 8J'e aware that their mem-ory is being tested. If this is the case, how canone speak of nonconscious processes in mem-ory? Indeed, research based on the process-dissociation procedllIe breaks with traditionin this regard by assuming that nonconsciousmemory rarely occUJrs in isolation, but ratherworks in combinatio:tl with conscious forms ofmemory to produce a variety of mnemonicphenomena, That is, it assumes that most actsof memory reflect a l)lend or co-action of con-

.

Opposition and theProcess-OissociationProcedure

One of the most important advances in thestudy of nonconscious influences of memoryis the opposition procedure developed by Ja-coby and colleagues (e.g..Jacoby. Woloshyn. &Kelley, 1989). For example, subjects mightstudy a list of words and then later be given atest of word-stem completion. However. in-

I,t

~II ' , .

! i-: ;1 j

:

~

I.

254 MEMORY IN THE LABORATORY

scious and nonconscious processes and thusattempts to separate the two kinds of memorywithin a single test. This contrasts with the im-plicit/explicit test approach, which attemptsto create tests that selectively measure eitherconscious or nonconscious forms of memory .Correspondence between the two approaches,including the absolute magnitude of estimatedmnemonic influences, provides some evi-dence that they are assessing the same under-lying constructs. Yet the two approaches areobviously different, a difference that has led tonumerous debates in the literature (e.g., Graf &Komatsu, 1994; Jacoby, 1998; Reingold & Toth,1996; Toth, Reingold, & Jacoby, 1995). Irre-spective of these debates, most researchersagree that nonconscious processes playa pow-erful role in conscious memory judgements.Thus, we next consider research that exploresthe complex interplay between conscious andnonconscious memory .

with the mere exposure paradigm (Seamon etal., 1995) and are relevant, not only to under-standing noncon~;cious memory and the na-ture of subjective experience but also to real-world decisions ;and actions because peopleoften act on the basis of their subjective expe-rience (see Jacoby', Kelley et al., 1989; Jacoby,Bjork, & Kelley, 1994).

In addition to influencing a person's inter-pretation of events in the present, fluent pro-cessing has been shown to influence inter-pretation of the past (i.e., conscious memoryjudgments). That~ by manipulating how flu-ently a stimulus Is processed on an expli-cit memory test, one can increase subjects' be-liefs that they are remembering, irrespective ofwhether the test stimulus was actually pre-sented earlier. For example, Whittlesea (1993;see also Jacoby & Whitehouse, 1989) manipu-lated the visual clarity or semantic context inwhich recognition-test items were presentedand found that mtJre fluently processed items(i.e., those presented more clearly or in morecongruent semantJlc contexts) were more likelyto be judged "oldl." Lindsay & Kelley (1996)produced similar "illusions of memory" bymanipulating the ease with which subjectscompleted word fragments. In addition to ex-tending the range of nonconscious memory ef-fects, these results may have significant impli-cations for the interpretation of self-reports ofawareness and thEi intention to remember. Forexample, fluent Jleprocessing may help ac-count for the phl~nomenon of confabulationdescribed earlier: even though a particularmental event may not have occurred in thepast, if that event is processed fluently in thepresent, people nlay incorrectly infer that itmust have occurrEid earlier in their life.

Taken together, the results cited in this sec-tion p ovide stro:l1ger evidence for a disso-ciation between m,emory and the phenomenalexperi nce of reIJrlembering. Moreover, theyreveal the operal:ion of nonconscious pro-cesses, not only in the use of memory but alsoin the attribution processes that subjects useto mak~ judgment:; about the past and present(Whittlesea, 1993).

Fluency, MemoryAttributions, and

Subjective ExperienceOn the basis of their experiments with the im-plicit word-identification test, Jacoby and Dal-las (1981) suggested that initial experiencewith a stimulus may result in more fluent(faster or more efficient) processing of thatstimulus when it is encountered at a latertime. They also suggested that such fluent re-processing may be the basis for implicit mem-ory effects (see also Masson, 1989) and that,under some conditions, subjects might use dif-ferences in fluency as the basis for consciousmemory decisions. These claims have turnedout to be very fruitful, and have led to a num-ber of experiments that reveal the complex na-ture of nonconscious memory processes.

Research has shown that fluent processingas a function of prior experience can changehow stimuli are processed in the present, andthis change can have interesting, nonconsciousinfluences on a person's subjective experi-ence. For example, fluent reprocessing as afunction of prior presentation can increase theapparent fame of nonfamous names Oacoby,Woloshyn et al., 1989), can lengthen the ap-parent exposure duration of a briefly flashedword (Witherspoon & Allan, 1985), can lowerthe apparent loudness of background noiseOacoby, Allan, Collins, & Larwill, 1988), andcan increase the apparent truth of false state-ments (Begg, Anas, & Farinacci, 1992). Theseeffects are conceptually similar to those found

The Future of N4~nconsciousMemorY: Problell11s and Prospects

\:,.,,;

:t

~~;;,It

-'",

The reader may halve noticed a change in tonethat h~ occurred throughout this review. Inparticu[ar, despite a "good start" in showingthe existence of nonconscious processes in

,NONCONSCIOUS FORMS OF HUMAN MEMORY 255

memory (based mainly on amnesic patientsand implicit/explicit dissociations in neuro-logically intact subjects), more recent researchhas been characterized by increasing complex-ity and debate about the nature of such influ-ences and the most appropriate way to mea-sure them. This is to be expected, perhaps,given the relatively short time in which re-searchers have investigated the relationshipbetween memory and consciousness. In thisfinal section, a set of issues are noted that mayhelp researchers develop more precise formsof measurement and more comprehensiveconceptualizations of nonconscious mnemon-ic processes.

;

The Relation between

A wareness and Intent

Schacter (1987) defined implicit memory as afacilitation of task performance "that does notrequire conscious or intentional recollection "

(p. 507). However, recent work suggests thatconsciousness and intentionality can be disso-ciated and may even re1lect different neuralmechanisms (Schacter, Alpert, Savage, Rauch,& Albert, 1997). Richardson-Klavehn et al.(1994) argued that many of the mnemonicphenomena labeled as "unconscious" or "un-aware" are better conceptualized as uninten-tional or involuntary. That is, subjects can beaware that the items they encounter or pro-duce on a memory test were in fact encoun-tered earlier, yet their performance can still re-veal the types of dissociations normallyattributed to implicit memory. However, al-though it seems clear that awareness and in-tentionality are distinct concepts, it is lessclear that they are completely separate fromone another, as some have suggested (e.g.,Richardson-Klavehn, Gardiner, & Java., 1996).The problem is that awareness is a prerequi-site for intentional control (see Jacoby, Toth,Lindsay, & Debner, 1992; Toth, Lindsay, & Ja-coby, 1992) and, indeed, may sometimes evenencourage such control. Important goals forfuture research are therefore to determinewhether awareness per se can change the ex-pression of memory , and the degree to whichawareness and intent are dissociable.

ences were "context free" in comparison tomore explicit, episo(Lic forms of memory thatwere thought to be ,closely tied to a specificpast context (e.g., Mandler, Graf, & Kraft,1986). Recent resear(:h, however, suggests thatnonconscious memory is much more contextbound than previously thought, and that suchcontextual specificit;'{ may even be a definingcharacteristic of nonconscious memory. Asone striking example, Hayman and Tulving(1989) presented subjects with a list of wordsand then gave them Itwo consecutive fragmentcompletion tests. T.h~ found that when frag-ments on the two -re,Sfs had little or no over-lap (e.g., a- a -in and -ss- ss -) subjectsshowed stochastic independence between thetwo tests-that is, colmpletion of one fragmentdid not predict a sul)ject's ability to solve theother fragment, despite the fact that the targetword was the same. Another example of con-text specificity is prolvided by Oliphant (1983),who showed that words presented as part ofthe instructions for an implicit test may pro-duce no priming on that test (see also Levy &Kirsner, 1989; MacLeod, 1989).

Related to the isslle of context is a subject'sgoals in performing some act for which non-conscious influence:s of memory may be op-erating. This is inlportant because mentalevents make up a large part of the context forprocessing; and goaJls, in turn, are arguably acentral aspect of mental life (Bargh, 1997; Toth& Reingold, 1996). F~esearch on memory attri-butions, described iabove, suggest that goalsmay playa major role in how nonconsciousmemory effects are clDnsciously interpreted. Onthe basis of this, arld other research, Jacoby,Ste-Marie, and Toth 1:1993) argued that noncon-scious memory processes may be relative to thegoals set by cOnSCiOlllS intentions. Research byGoschke and Kuhl (1996; see also Marsh,Hicks, & Bink, 1998) seems to confirm this: sub-jects who expected Ito perfonn a future actionshowed greater implicit memory for words de-scribing that action, as compared to words de-scribing similar actions that the subject did notexpect to perfonn.

Similar to concllllsions from the previoussection, then, results reviewed here recom-mend more research. into the relation betweennonconscious memory and a subject's goalsand intentions. As discussed next, such re-search would be useful, not only in helping usbetter understand tile nature of nonconsciousmemory but also in bridging the gap betweenmemory phenomen;a occurring in the labora-tory and those occUJrring in the "real world."

The Role of Contextand Goals

Early research and theory on nonconscious in-fluences of memory suggested that such influ-

~

256 MEMORY IN THE LABORA TOR)'

Nonconscious Memoryin Real-World Thoughtand Behavior

The majority of research discussed so far hasbeen based on the processing of rather simplestimuli, such as isolated words, presented inthe context of rather simple tasks, such asidentifying briefly flashed stimuli, If, how-ever, nonconscious memory processes aretruly a critical aspect of everyday cognition,research must link laboratory demonstrationswith phenomena observed outside of the lab.Fortunately, this enterprise has begun and theinitial results, mainly reported in the social-cognition literature, are exciting (see Bargh,1997; Greenwald & Banaji, 1995). Already,strong links have been drawn between non-conscious processes in memory and sociallyrelevant phenomena such as impression forma-tion, stereotyping, and prejudice (see Banaji &Greenwald, 1994; Devine, 1989). Moreover,techniques for the measurement of noncon-scious forms of memory are being developedfor personality assessment in the applied (e.g.,job-related) sector (e.g., Dovidio & Fazio, 1992),

The role of nonconscious memory pro-cesses in clinical assessment is also maturing,Most memory-impaired populations (e.g" am-nesics, the elderly) lose only conscious mem-ory; nonconscious forms of memory appear toremain relatively intact throughout the lifespan. However, because people only have con-scious access to the forms of memory that de-cline, their subjective experience may often beof a complete loss of memory. This may leadmemory-impaired individuals to give up onadequately performing memory-related tasks,despite the fact that a proper arrangement ofcontextual cues could be used to maintain ad-equate levels of performance, This possibilityspeaks again to the complex interplay betweenconscious and nonconscious memory and de-serves further scrutiny in the future,

a triple dissociation between conscious mem-ory and two forml; of nonconscious memory.That is, KC showed no conscious memory forlearning episodes in which he was taught tomake correct responses in a semantic knowl-edge task, and hi;s semantic responses werefound to be stochastically independent of hisperformance in aD. implicit fragment-comple-tion test. Perhaps I~ven more impressive, bothforms of nonconscious memory were found tobe largely intact 1:~ months after initiallearn-ing. Thus, detailed studies of amnesic patientscontinue to proviwUJnportant insights into thefunctional (cogniffiiej organization of memory,a strategy that is l:ikely to continue in the fu-ture.

Insights into the neuroanatomical organiza-tion of nonconscious forms of memory havealso advanced in recent years, based largelyon our ability to use modern neuroimagingtechniques such a~i PET and fMRI to measurethe neural activity of normal patients whilethey perform memory tasks (Buckner, chap-ter 39). These shldies support lesion-basedresearch in showing that the neural systemsunderlying nonconscious memory are anatom-ically distinct from the medial-temporal struc-tures underlying conscious memory .In partic-ular, perceptual and conceptual priming isthought to occur in the very same neocorticalstructures that mediate initial perception andmeaning-based thought, while more proce-dural forms of nonconscious memory appearto rely on subcortical structures such as thebasal ganglia (Gabrieli, 1998). An obviouslyimportant goal for future neurocognitive re-search will be to bl~tter understand the detailsof these brain systf~ms, including their neuro-anatomy and chemistry , as well as their pointsof interaction.

Summary and Co,nclusions

Although research on nonconscious forms ofhuman memory can be traced back at least ahundred years (sel3 Schacter, 1987; Schacter& Tulving, 1994b; Toth & Reingold, 1996),modern research on the subject began with thediscovery that amnesic patients, previouslythought Unable to store their experience,could show memoJ:y for the past when testedimplicitly. Subseqtlent research showed simi-lar nonconscious memory phenomena in nor-mal subjects. Since these initial demonstra-tions, research on nonconscious mnemonicprocesses has been directed toward the de-

Nonconscious Memoryand the Brain

A final area of interest concerns the neuralsubstrates of nonconscious memory. As wasnoted in the introduction, much of the mod-ern interest in nonconscious memory wasbased on research with amnesic patients, andexploration of the memory capabilities ofthese patients continues. For example, in anextended case study of one patient (KC), Tulv-ing, Hayman, & MacDonald (1991) uncovered

~

I

NONCONSCIOUS FORMS OF HUMAN MEMORY 257

velopment of more precise methodologies forrevealing and measuring these processes.These methodologies include the mere-expo-sure paradigm, implicit tests, and self-reportsof awareness, intent, and the phenomenologyof remembering. More recent methods includethose that put conscious and nonconsciousforms of memory in opposition, and those thatattempt to quantitatively estimate their sepa-rate strengths (e.g., the process-dissociationprocedure). A central theme in many of theseapproaches is the distinction between per-ception and conception (meaning-based pro-cesses), although contextual and goal-relatedprocesses are also starting to be recognized asfundamental.

Nonconscious memory plays a critical rolein everyday thought and behavior both di-rectly (e.g", by bringing a particular thought tomind) and indirectly (by influencing a per-son's subjective experience). These influencesextend beyond the "cold cognition " studied in

the lab, and into the "hot" areas of emotionaland social information processing. So, too,neuroscientists are beginning to uncover theneural substrates of nonconscious memory.This will no doubt be an exciting time to re-member.

References

1

A. R. (1995). Double dissociation of condi-tioning and declarative knowledge relativeto the amygdala and hippocampus in hu-mans. Science, 269,1115-1118.

Begg, I., Anas, A., & Frarinacci, S. (1992). Dis-sociation of processes in belief: Source rec-ollection, statement familiarity, and the il-lusion of truth. Journal of ExperimentalPsychology: General, 121,446-458.

Benzing, w. E., & Squire, L. R. (1989). Pre-served learning and memory in amnesia:Intact adaptation level effects and learningof stereoscopic depth. Behavioral Neurosci-ence, 103,538-:#4;!1:.

Berry, D. C., & Broadbent, D. E. (1984). Onthe relationship between task performanceand associated verbalisable knowledge.Quarterly Journal of Experimental Psychol-ogy, 36, 209-231.

Blaxton, T. A. (1989). Investigating dissocia-tions among memory measures: Supportfor a transfer appropriate processing frame-work. Journal of Experimental Psychology:Learning, MemoIy; and Cognition, 15,657-668.

Bonanno, G. A., & Stillings, N. A. (1986). Pref-erence, familiarity, and recognition after re-peated brief exposures to random geomet-ric shapes. American Journal ofPsychology, 99(8), 403-415.

Bornstein, R. F. (1989). Exposure and affect:Overview and meta-analysis of research,1968-1987. Psychological Bulletin, 106,265-289.

Bowers, I. S., & Schacter, D. L. (1990). Im-plicit memory and test awareness. Journalof Experimental Psychology: Learning,MemoIy; and Cognition, 16, 404~416.

Brooks, D. N., & Baddeley, A. D. (1976). Whatcan amnesic patients learn? Neuropsycho-logia, 14, 111-122.

Cohen, I., & Schooler, I. w. (Eds.), (1997). Sci-entific approaches to consciousness. Mah-wah, NI: Erlbaum.

Cohen, N. I., Eichenbaum, H., Deacedo, B., &Corkin, S. (1985). Different memory sys-tems underlying acquisition of proceduraland declarative knowledge. Annals of theNew York Academy of Sciences, 444,54-71.

Corkin, S. (1968). Acquisition of motor skillafter bilateral medial temporal lobe exci-sion. Neuropsychologia, 6, 225-265.

Curran, T. (1998). Implicit sequence learningfrom a cognitive neuroscience perspective:What, how, and where? In M. A. Stadler &P. A. Frensch (Eds.), Handbook of implicitlearning (pp. 365-400). Thousand Oaks,CA: Sage Publications, Inc.

~

Adams, I. K. (1957). Laboratory studies ofbe-havior without awareness. PsychologicalBulletin, 54, 383-405.

Adams, L. T., Kassennan, I. E., Yearwood,A. A., Perfetto, G. A., Bransford, I. D., &Franks, I. I. (1988). Memory access: The ef-fects of fact-oriented versus problem-ori-ented acquisition. Memory 8' Cognition,16, 167-175.

Banaji, M. R., & Greenwald, A. G. (1994). Im-plicit stereotyping and unconscious preju-dice. In M. P. Zanna & I. M. Olson (Eds.),The psychology of prejudice: The Ontariosymposium (pp. 55-76). Hillsdale, NI:Erlbaum.

Bargh, I. A. (1997). The automaticity of every-day life. In R. S. Wyer (Ed.), Advances insocial cognition (pp. 1-61). Mahwah, NJ:Erlbaum.

Barrs, B. (1988). A cognitive theory of con-sciousness. New York: Cambridge Univer-sity Press.

Bartlett, F. (1932). Remembering: An experi-mental and social study. Cambridge: Cam-bridge University Press.

Bechara, A., Thanel, D., Damasio, H.,Adolphs, R., Rockland, C., & Damasio,

lat.

258 MEMORY IN THE LABORATORY

a 'traceless' memory system. Journal of Ex-perimental P.sychology: Learning, MemoIy;and Cognition, 14,941-956.

Hull, C. L. (1943). Principles of behavior.New York: Appleton.

Iacoby, L. L. (1983). Remembering the data:Analyzing interactive processes in reading.Journal of Verbal Learning and Verbal Be-havior, 22,485-508.

Iacoby, L. L. (1998). Invariance in automaticinfluences of memory: Toward a user'sguide for the process-dissociation proce-dure. JournC1LC!f Experimental Psychology:Learning, Am'tiiOIy; and Cognition, 24,3-26.

Iacoby, L. L. (1991). A process dissociationframework: Separating automatic from in-tentional use,s of memory. Journal of Mem-ory and Language, 30, 513-541.

Iacoby, L. L., Allan, L. G., Collins, I. C., & Lar-will, L. K. (1988). Memory influences sub-jective experience: Noise judgments. Jour-nal of Experimental Psychology: Learning,MemoIy; and Cognition, 14,240-247.

Iacoby, L. L., Bjork, R. A., & Kelley, C. M.(1994). Illusions of comprehension, compe-tence, and remembering. In D. Druckman& R. A. Bjork (Eds.), Learning, remember-ing, and believing: Enhancing human per-formance (pp. 391-422). Washington, DC:National Academy Press.

Iacoby, L. L., & Dallas, M. (1981). On the rela-tionship between autobiographical mem-ory and perceptual learning. Journal of Ex-perimental Psychology: General, 3,306-340.

Iacoby, L. L., Kelley, C. M., & Dywan, I.(1989). Memory attributions. In H. L. Roe-diger & F. I. M. Craik (Eds.), Varieties ofmemory and consciousness (pp. 391-422).Hillsdale, NI: Erlbaum.

Iacoby, L. L., Ste-Marie, D., & Toth, I. P.(1993). Redefining automaticity: Uncon-scious influences, awareness and control.In A. D. Baddeley & L. Weiskrantz (Eds.),Attention: Selection, awareness and con-trQl. A tribute to Donald Broadbent (pp.261-282). Oxford, England: Oxford Univer-sity Press.

Iacoby, L. L., Toth, I. P., Lindsay, D. S., & Deb-ner, I. A. (1992). Lectures for a layperson:Methods for revealing unconscious pro-cesses. In R. F. Bomstein & T. S. Pittman(Eds.), Perception without awareness (pp.81-120). New York: Guilford.

Iacoby, L. L., Toth, I. P., & Yonelinas, A. P.(1993). Separating conscious and uncon-scious influe:[lces of memory: Measuring

Devine, P. G. (1989). Stereotypes and preju-dice: Their automatic and controlled com-ponents. Journal of Personality and SocialPsychology, 56, 5-18.

Dovidio, J. F., & Fazio, R. H. (1992). New tech-nologies for the direct and indirect assess-ment of attitudes. In J. Tanur (Eds.), Ques-tions about survey questions: Meaning,memory, attitudes, and social interaction(pp. 204-237). New York: Russell Sage.

Duchek, J. M., & Neely, J. H. (1989). A dissoci-ative word-frequency x levels-of-process-ing interaction in episodic recognition andlexical decision tasks. Memory 8' Cogni-tion, 17, 148-162.

Dunn, J. C., & Kirsner, K. (1989). Implicitmemory: Task or process? In S. Lewandow-sky, J. C. Dunn, & K. Kirsner (Eds.), Im-plicit memory: Theoretical issues Hills-dale, NJ: Erlbaum.

Ebbinghaus, H. (1885/1964). Memory: Acontribution to experimental psychology(H. A. Ruger & C. E. Bussenius, Trans.).New York: Dover Publications.

Gabrieli, J. D. E. (1998). Cognitive neurosci-ence of human memory. Annual Review ofPsychology; 49, 87-115.

Goschke, T., & Kuhl, J. (1996). Rememberingwhat to do: Explicit and implicit memoryfor intentions. In M. Brandimonte, G. 0.Einstein, & M. A. McDaniel (Eds.), Prospec-tive memory: Theory and applications.Mahwah, NJ: Erlbaum.

Graf, P., & Komatsu, S. (1994). Process dissoci-ation procedure: Handle with caution! Eu-ropean Journal of Cognitive Psychology, 6,113-129.

Graf, P., Mandler, G., & Hayden, P. E. (1982).Simulating amnesic symptoms in normalsubjects. Science, 218, 1243-1244.

Greenwald, A. G., & Banaji, M. R. (1995). Im-plicit social cognition: Attitudes, self-es-teem, and stereotypes. Psychological Re-view, 102, 4-27.

Hashtroudi, S., Ferguson, S. A., Rappold,v. A., & Chrosniak, L. D. (1988). Data-driven and conceptually driven processesin partial-word identification and recogni-tion. Journal of Experimental Psychology:Learning, Memory, and Cognition, 14,749-757.

Hay, J. F., & Jacoby, L. L. (1996). Separatinghabit and recollection: Memory slips, pro-cess dissociations, and probability match-ing. Journal of Experimental Psychology:Learning, Memory, and Cognition, 22,1323-1335.

Hayman, C. A. G., & 1\1lving, E. (1989). Ispriming in fragment completion based on

I

NONCONSCIOUS FORMS OF HUMAN MEMORY 259

~

recollection. Journal of Experimental Psy-chology: General, 122, 139-154.

Jacoby, L. L., & Whitehouse, K. (1989). An il-lusion of memory: False recognition influ-enced by unconscious perception. Journalof Experimental Psychology: General, 118,126-135.

Jacoby, L. L., & Witherspoon, D. (1982). Re-membering without awareness. CanadianJournal of Psychology, 36, 300-324.

Jacoby, L. L., Woloshyn, V., & Kelley, C. M.(1989). Becoming famous without beingrecognized: Unconscious influences ofmemory produced by dividing attention.Journal of Experimental Psychology: Gen-eral, 118,115-125.

Jacoby, L. L., Yonelinas, A. P., & Jennings,J. M. (1997). The relationship between con-scious and unconscious (automatic) influ-ences: A declaration of independence. InJ. Cohen & J. W. Schooler (Eds.), Scientificapproaches to consciousness Mahwah, NJ:Erlbaum.

Jennings, J. M., & Jacoby, L. L. (1993). Auto-matic versus intentional uses of memory:Aging, attention, and COR.trol. Psychologyand Aging, 8, 283-293.

Jennings, J. M., & Jacoby, L. L. (1997). Anopposition procedure for detecting age-related deficits in recollection: Telling ef-fects of repetition. Psychology and Aging,12, 352-361.

Kelley, C. M., & Lindsay, D. S. (1993). Remem-bering mistaken for knowing: Ease of re-trieval as a basis for confidence in answersto general knowledge questions. Journal ofMemory and Language, 32,1-24.

Kelley, C. M., & Lindsay, D. S. (1996). Con-scious and unconscious forms of memory.In E. L. Bjork & R. A. Bjork (Eds.), Memory(pp. 31-63). San Diego, CA: AcademicPress.

Knowlton, B., & Squire, L. R. (1994). Probabi-listic classification learning in amnesia.Learning and Memory, 1,106-120.

Kolers, P. A. (1976). Reading a year later. Jour-nal of Experimental Psychology: HumanLearning and Memory, 2,554-565.

Kolers, P. A., & Roediger, H. L. (1984). Proce-dures of mind. Journal of Verbal Learningand Verbal Behavior, 23, 425-449.

Kopelman, M. D. (1987). 1Wo types of confab-ulation. Journal of Neurology, Neurosur-ge~ and Psychiat~ 50,1482-1487.

Kroll, J. F., & Potter, M. C. (1984). Recogniz-ing words, pictures, and concepts: A com-parison of lexical, object, and reality deci-sions. Journal of Verbal Learning andVerbal Behavior, 23, 39-66.

Kunst-Wilson, W. R., & Zajonc, R. B. (1980).Affective discrimination of stimuli thatcannot be recognized. Science, 207, 557-558.

Levy, B. A., & Kirsner, K. (1989). Reprocess-ing text: Indirect measures of word andmessage level processes. Journal of Experi-mental Psychology: Learning, MemoIy; andCognition, 15,407-417.

Lindsay, D. S., & Kelley, C. M. (1996). Creat-ing illusions of familiarity in a cued-recallremember/know paradigm. Journal ofMemoIY and Language, 35, 197-211.

Logan, G. D. (1990). ~tition priming andautomaticity: Common underlying mecha-nisms? Cognitive Psychology, 22, 1-35.

MacLeod, C. M. (1989). Word Context duringinitial exposure influences degree of prim-ing in word fragment completion. Journalof Experimental Psychology: Learning,MemoIy; and Cognition, 15, 398-406.

Mandler, G., Graf, P., & Kraft, D. (1986). Acti-vation and elaboration effects in recogni-tion and word priming. Quarterly Journalof Experimental Psychology, 38A, 645-662.

Marsh, R. L., Hicks, I. L., & Bink, M. L. (1998).Activation of completed, uncompleted,and partially completed intentions. Jour-nal of Experimental Psychology: Learning,MemoIy; and Cognition, 24,350-361.

Masson, M. E. I. (1989). Fluent reprocessingas an implicit expression of memory for ex-perience. In S. Lewandowsky, I. C. Dunn,& K. Kirsner (Eds.), Implicit memoIY: Theo-retical issues. Hillsdale, NI: Erlbaurn.

Masson, M. E. I., & MacLeod, C. M. (1992).Reenacting the route to interpretation: En-hanced perceptual identification withoutprior perception. Journal of ExperimentalPsychology: General, 121, 145-:.176.

Milner, B. (1965). Visually-guided maze learn-ing in man: Effects of bilateral hippocam-pal, bilateral frontal, and unilateral cere-brallesions. Neuropsychologia, 3,317-338.

Mitchell, D. B., & Brown, A. S. (1988). Persis-tent repetition priming in picture namingand its dissociation from recognition mem-ory. Journal of Experimental Psychology:Learning, MemoIy; and Cognition, 14,213-222.

Moscovitch, M. (1989). Confabulation and thefrontal systems: Strategic versus associ~-tive retrieval in neuropsychological theo-ries of memory. In H. L. Roediger & F. I. M.Craik (Eds.), Varieties of memoIY and con-sciousness: Essays in honour of En del Thlv-ing (pp. 133-156). Hillsdale, NI: Erlbaurn.

260 MEMORY IN THE LABORATORY

Moscovitch, M. (1994). Memory and working-with-memory: Evaluation of a componentprocess model and comparison with othermodels. In D. L. Schacter & E. 1\1lving(Eds.), Memory Systems 1994 (pp. 269-310). Cambridge: MIT/Bradford Press.

Moscovitch, M., Vriezen, E., & Goshen-Gott-stein, G. (1993). Implicit tests of memoryin patients with focal lesions or degenera-tive brain disorders. In F. BoIler & I. Graf-man (Eds.), Handbook of neuropsychology(pp. 133-173). Amsterdam: Elsevier.

Murphy, S. T., & Zajonc, R. B. (1993). Affect,cognition, and awareness: Affective prim-ing with optimal and suboptimal stimulusexposures. Journal of Personality and So-cial Psychology, 64, 723-739.

Nissen, M. I., & Bullemer, P. (1987). Atten-tional requirements of learning: Evidencefrom performance measures. Cognitive Psy-chology, 19, 1-32.

Oliphant, G. w. (1983). Repetition and re-cency effects in word recognition. Austra-lian Journal of Psychology, 35, 393-403.

Parkin, A. I., Reid, T. K., & Russo, R. (1990).On the differen~al nature of implicit andexplicit memory. Memory and Cognition,18, 507-514.

Picton, T. W., & Stuss, D. T. (1994). Neurobiol-ogy of conscious experience. Current Opin-ion in Neurobiology, 4, 256-265.

Ratcliff, R., & McKoon, G. (1996). Bias effectsin implicit memory tasks. Journal of Exper-imental Psychology: General, 125, 403-421.

Reber, A. S. (1989). Implicit learning andtacit knowledge. Journal of ExperimentalPsychology: General, 118, 219-235.

Reingold, E. M., & Goshen-Gottstein, Y.(1996). Separating consciously controlledand automatic influences in memory fornew associations. Journal of ExperimentalPsychology: Learning, Memory; and Cogni-tion, 22, 397-406.

Reingold, E. M., & Merikle, P. M. (1990). Onthe inter-relatedness of theory and mea-surement in the study on unconscious pro-cesses. Mind and Language, 5, 9-28.

Reingold, E. M., & Toth, I. P. (1996). Processdissociations versus task dissociations: Acontroversy in progress. In G. Underwood(Eds.), Implicit Cognition (pp. 159-202).Oxford: Oxford University Press.

Richardson-Klavehn, A., & Bjork, R. A.(1988). Measures of memory. Annual Re-view of Psychology; 39, 475-543.

Richardson-Klavehn, A., Gardiner, I. M., &lava, R. I. (1996). Memory: Task dissocia-tions, process dissociations, and dissocia-tions of consciousness. In G. Underwood

(Eds.), Implicit Cognition (pp. 85-158).New York: Oxford University Press.

Richardson-Klavehn, A., Lee, M. G., Joubran,R., & Bjork, R. A. (1994). Intention andawareness in perceptual identification prim-ing. MemoIY and Cognition, 22, 293-312.

Roediger, H. L. (1990). Implicit memory: Re-tention without remembering. AmericanPsychologist, 45, 1043-1056.

Roediger, H. L., & McDermott, K. B. (1993).Implicit memory in normal human sub-jects. In H. Spinnler & F. Boiler (Eds.),Handbook of neuropsychology (pp. 63-131). Am~am: Elsevier.

Roediger, H. L., Weldon, M. S., & Challis,B. H. (1989). Explaining dissociations be-tween implicit and explicit measures of re-tention; A processing account. In H. L. Roe-diger & F. I. M. Craik (Eds.), Varieties ofmemoIY and consciousness: Essays in hon-our of En del Thlving (pp. 3-41). Hillsdale,NJ: Erlbaum.

Rozin, P. (1976). The psychobiological ap-proach to human memory. In M. R. Rosenz-weig & E. L. Bennett (Eds.), Neural mecha-nisms of learning and memoIY (pp. 3-46).Cambridge, MA: MIT Press.

Schacter, D. L. (1987). Implicit memory: His-tory and current status. Journal of Experi-mental Psychology: Learning, MemoIy; andCognition, 13, 501-518.

Schacter, D. L., Alpert, N. M., Savage, C. R.,Rauch, S. L., & Albert, M. S. (1997). Con-scious recollection and the human hippo-campal formation: Evidence from positronemission tomography. Procedures of theNational Academy of Science, USA, 93,321-325.

Schacter, D. L., Bowers, J., & Booker, J.(1989). Intention, awareness, and implicitmemory: The retrieval intentionality crite-ria. In S. Lewandowsky, J. C. Dunn, & K.Kirsner (Eds.), Implicit memoIY (pp. 47-65). Hillsdale, NJ: Erlbaum.

Schacter, D. L., Cooper, L. A., & Delaney, S. M.(1990). Implicit memory for visual objectsand the structural description system. Bul-letin of the Psychonomic Society, 28, 367-372.

Schacter, D. L., & Thlving, E. (Eds.). (1994a).MemoIY systems 1994. Cambridge, MA:MIT Press.

Schacter, D. L., & Thlving, E. (1994b). Whatare the memory systems of 1994? In D. L.Schacter & E. Thlving (Eds.), MemoIY sys-tems 1994 (pp. 1-38). Cambridge, MA:MIT Press.

Seamon, J. G., Williams, P. C., Crowley, M. J.,Kim, I. I., Langer, S. A., Ome, P. J., & Wis-

NONCONSCIOUS FORMS OF HUMAN MEMORY 261

hengrad, D. L. (1995). The mere exposureeffect is based on implicit memory: Effectsof stimulus type, encoding conditions, andthe number of exposures on recognitionand affect judgments. Journal of Experi-mental Psychology: Learning, MemoIy; andCognition, 21, 711-721.

Seger, C. A. (1994). Implicit learning. Psycho-logical Bulletin, 115, 163-196.

Shimamura, A. P., & Squire, L. R. (1984).Paired-associate learning and priming inamnesia: A neuropsychological approach.Journal of Experimental Psychology: Gen-eral, 113,556-570.

Skinner, B. F. (1969). Contingencies of rein-forcement: A theoretical analysis. NewYork: Appleton.

Smith, E. R., & Branscombe, N. R. (1988). Cat-egory accessibility as implicit memory.Journal of Experimental Social Psychology;24, 490-504.

Srinivas, K., & Roediger, H. L. (1990). Classify-ing implicit memory tests: Category associ-ations and anagram solution. Journal ofMemory and Language, 29, 389-412.

Ste-Marie, D. M., Jennings, J. M., & Finlayson,A. J. (1996). The process dissociation pro-cedure: Memory testing in populationswith brain damage. The Clinical Neuropsy-chologist.

Tolman, E. C. (1932). Purposive behavior inanimals and man. New York: Appleton.

Toth, J. P. (1996a). Conceptual automaticity inrecognition memory: Levels of processingeffects on familiarity. Canadian Journal ofExperimental Psychology, 50, 123-138.

Toth, J. P. (1996b). Memory and attention fol-lowing traumatic brain injury: Process dis-sociations. Clinical Neuropsychologist, 10,350-351.

Toth, J. P., & Hunt, R. R. (1999). Not one ver-sus many but zero versus any: Structureand function in the context of the multi-ple-memory systems debate. In J. K. Fos-ter & M. Jelicic (Eds.), Memory: Structure,function, or process? (pp. 232-272). Ox-ford: Oxford University Press.

Toth, J. P., Lindsay, D. S., & Jacoby, L. L.(1992). Awareness, automaticity, and mem-ory dissociations. In L. R. Squire & N. But-ters (Eds.), Neuropsychology of memory(pp. 46-57). New York: Guilford.

Toth, J. P., & Reingold, E. M. (1996). Beyondperception: Conceptual contributions to un-conscious influences of memory. In G. Un-derwood (Eds.), Implicit cognition (pp. 41-84). New York: Oxford University Press.

Toth, J. P., Reingold, E. M., & Jacoby, L. L.(1994). Toward a redefinition of implicitmemory: Process dissociations following

elaborative processing and self-generation.lournal of Experimental Psychology: Learn-ing, Memory, and Cognition, 20, 290-303.

Toth, I. P., Reingold, E. M., & Iacoby, L. L.(1995). A response to Graf and Komatsu'scritique of the process dissociation proce-dure: When is caution necessary? Euro-pean lournal of Cognitive Psychology, 7 ,113-130.

Thlving, E. (1985). Memory and conscious-ness. Canadian Psychology; 26, 1-12.

Thlving, E., Hayman, C. A. G., & Macdonald,C. A. (1991). Long-lasting perceptual prim-ing and semaIi~eaming in amnesia: Acase experiment. lournal of ExperimentalPsychology: Learning, Memory, and cogni-tion, 17, 595-617.

Thlving, E. & Schacter, D. L. (1990). Primingand human memory systems. Science, 247,301-305.

Thlving, E., Schacter, D. L., & Stark, H. A.(1982). Priming effects in word-fragmentcompletion are independent of recognitionmemory. lournal of Experimental Psychol-ogy: Learning, Memory, and Cognition, 8,336-342.

Underwood, B. I. (1957). Interference and for-getting. Psychological Review, 64, 49-60.

Vaidya, C. I., Gabrieli, I. D. E., Keane, M. M.,Monti, L. A., Gutierrez-Rivas, H., & Zar-ella, M. M. (1997). Evidence of multiplemechanisms of conceptual priming on im -plicit tests. lournal of Experimental Psy-chology: Learning, Memory, and Cognition,23, 1324-1343.

Warrington, E. K., & Weiskrantz, L. (1968).New method of testing long-term retentionwith special reference to amnesic patients.Nature, 217,972-974.

Weiskrantz, L., & Warrington, E. K. (1979).Conditioning in amnesic patients. Neuro-psychologia, 17, 187-194.

Weldon, M. S., & Roediger, H. L. (1987). Alter-ing retrieval demands reverses the picturesuperiority effect. Memory & Cognition,15, 269-280.

Whittlesea, B. W. A. (1993). Illusions of famil-iarity. lournal of Experimental Psychology:Learning, Memory, and Cognition, 19,1235-1253.

Witherspoon, D., & Allan, L. G. (1985). The ef-fects of a prior presentation on temporaljudgments in a perceptual identificationtask. Memory and Cognition, 13, 101-111.

Yonelinas, A. P. , & Iacoby, L. L. (1995). Disso-ciations of processes in recognition mem-ory: Effects of interference and responsespeed. Canadian lournal of ExperimentalPsychology.