intelligence and the cognitive...

P1: PHB Trim: 7in × 10in Top: 0.5in Gutter: 0.875inCUUS1280-22 cuus1280/Sternberg ISBN: 978 0 521 51806 2 February 11, 2011 17:57

CHAPTER 22

Intelligence and the CognitiveUnconscious

Scott Barry Kaufman

The definition of genius is that it actsunconsciously; and those who haveproduced immortal works, have done sowithout knowing how or why. The greatestpower operates unseen, and executes itsappointed task with as little ostentation asdifficulty.

– William Hazlitt1

Intelligence tests were originally createdwith the practical goal of identifying stu-dents in need of alternative education(Binet & Simon, 1916). Because intelligencetests were originally devised to predictschool grades, the items were intention-ally designed to measure a general abilityto profit from explicit instruction, concen-trate on a task, and engage in intellectualmaterial. Indeed, research shows that sucha general ability does seem to exist. Overa century ago, Spearman (1904) discoveredthat when a wide range of cognitive tests

1 William Hazlitt (1846), “Essay IV. Whether GeniusIs Conscious of Its Powers?” in Table Talk: Opinionson Books, Men, and Things, Second Series, Part I(pp. 37–49). New York, NY: Wiley & Putnam.

that have explicit instructions and requireeffortful concentration is administered to adiverse group of people, all of the tests tendto be positively correlated with one another,a finding often referred to as a “positive man-ifold.” Spearman labeled the factor on whichall individual tests loaded g, for general intel-ligence.

Over the past 100 years, the existence ofg as a statistical phenomenon is one of themost replicable findings in all of psychology(Carroll, 1993; Chabris, 2007; Jensen, 1998).Nonetheless, there is still work to be done todetermine what explains the positive man-ifold (see Maas et al., 2006), the cognitivemechanisms that support g (see Chapter 20,Working Memory and Intelligence, this vol-ume; Kaufman, DeYoung, Gray, Brown, &Mackintosh, 2009; Sternberg & Pretz, 2005),and whether there are other forms of cog-nition that display meaningful individualdifferences and predict intelligent behaviorabove and beyond g and the cognitive mech-anisms that support g.

This chapter presents evidence thatmechanisms relating to the cognitive uncon-scious – “mental structures, processes, and

442


INTELLIGENCE AND THE COGNITIVE UNCONSCIOUS 443

states2 that can influence experience,thought, and actions outside phenomenalawareness and voluntary control” (Dorf-man, Shames, & Kihlstrom, 1996, p. 259)also make an important contribution tointelligent behavior. Although intelligencetesters have done a remarkable job devel-oping tests that measure individual dif-ferences in explicit, controlled cognitiveprocesses, the investigation of individual dif-ferences in implicit, nonconscious processeshas not received nearly as much attention(Kaufman, 2009a, b).

Furthermore, researchers have createdclever experiments to probe the natureof the cognitive unconscious by lookingat implicit memory, implicit perception,and other forms of implicit cognition andthought3 (for reviews, see Kihlstrom, 1987,and Litman & Reber, 2005), but theyhave focused primarily on group-level data,ignoring individual differences (see Cron-bach, 1957). Additionally, some researchershave downplayed the existence of continu-ous individual differences in the cognitiveunconscious that are meaningfully relatedto important life outcomes (Reber, 1993;Stanovich, 2009).

There have been some recent studies,however, that look at individual differencesin the cognitive unconscious. This chap-ter focuses on individual differences andreviews recent empirical work on relationsamong the cognitive processes underlyingpsychometric intelligence and the cognitiveprocesses underlying the cognitive uncon-scious, attempting to bridge two majorresearch programs that, until recently, havetraveled on separate but parallel paths.

2 I include “implicit thought” in this definition as well,although Kihlstrom tends to refer to “implicit cog-nition” differently from the “cognitive unconscious”(Dorfman, Shames, & Kihlstrom, 1996).

3 I assume in this chapter that intelligent “thought”can operate either with or without awareness ofthat thought. As Dorfman, Shames, and Kihlstrom(1996) astutely note, the idea of “implicit thought”is a difficult concept because the notion of think-ing has traditionally been equated with notions ofconsciousness. For instance, William James (1890)thought the notion of “unconscious thought” was acontradiction in terms!

Integrating Two Research Traditions

The 20th century witnessed at least twomajor paradigm shifts within psychologicalscience. One major shift was from behav-iorism to the “cognitive revolution,” whichbrought along with it a shift in focus fromlearning and conditioning toward investigat-ing the mental processes involved in con-scious thought, including memory, think-ing, and problem solving (Miller, 2003). Thisshift has had an enduring effect on concep-tualizations of human intelligence as wellas research methodology. Indeed, one ofthe earliest investigators of the develop-ment of intelligence in children was JeanPiaget (1952), whose focus was on conscioushigher order reasoning and how childrenat different ages think. This emphasis onage differences in thought as well as thenotion that intelligence involves conscious,deliberate reasoning also underlies the logicbehind the first widely administered intel-ligence test, the Binet-Simon Scale (Binet& Simon, 1916). Furthermore, the discoverythat performances on diverse tests of explicitcognitive ability tend to correlate with oneanother – Spearman’s (1904) so-called pos-itive manifold – further supported the ideathat intelligence tests are tapping into a “gen-eral cognitive ability.”

Around the same time the shift frombehaviorism to the cognitive revolution wastaking place, another dramatic shift in psy-chology was occurring. The conceptual-ization of the unconscious that was pre-dominant with psychodynamic theories ofpersonality was slowly being transformedinto an unconscious recognized to servemany adaptive functions among bothmodern-day humans and our evolutionaryancestry (Epstein, 1991; Hassin, Uleman, &Bargh, 2005; Wilson, 2004). Over 30 yearsof research in cognitive science reveals thata considerable amount of information pro-cessing takes place on a daily basis automat-ically – without our intent, awareness, anddeliberate encoding – and plays an impor-tant role in structuring our skills, percep-tions, and behavior (Epstein, 1991; Hassinet al., 2005; Kihlstrom, 1987; Lewicki & Hill,


444 SCOTT BARRY KAUFMAN

1987; Reber, 1993; Stadler & Frensch, 1997)as well as facilitating problem solving andcreativity (Dijksterhuis & Nordgren, 2006;Dorfman, Shames, & Kihlstrom, 1996; Lit-man & Reber, 2005).

Kihlstrom (1987) distinguishes betweenthree types of nonconscious mental struc-tures that together constitute the domainof the “cognitive unconscious.” Unconsciousrepresentations fit within the domain of pro-cedural knowledge and are inaccessible tointrospection under any circumstances. “Byvirtue of routinization (or perhaps becausethey are innate), such procedures operateon declarative knowledge without eitherconscious intent or conscious awareness,in order to construct the person’s ongoingexperience, thought, and action” (p. 1450;also see Anderson, 1982). Subliminal percep-tion, implicit memory, and implicit learn-ing fit the category of preconscious declar-ative knowledge structures. In contrast tounconscious representations, preconsciousstructures can be available to phenomenalawareness and can be introspected upon, butthey can also influence ongoing experience,thought, and action without ever enteringinto working memory. Finally, Kihlstromdescribes subconscious declarative knowl-edge mental representations such as thoseactivated during hypnosis, which can bequite available to introspection but inacces-sible to phenomenal awareness.4

Note that even though some noncon-scious representations have such high lev-els of activation that they enter workingmemory, they still might not meet thecriteria of conscious awareness. As notedby Kihlstrom, William James (1890) sug-gested over a century ago in his Principles of

4 Note that only Kihlstrom’s (1987) notion of “uncon-scious” mental structures meets all four of Bargh’s(2004) horsemen of automaticity: lack of aware-ness, lack of intention, high efficiency, and inabilityto control. Kihlstrom’s notion of the preconsciouslacks intention, but only under some circumstancesis efficient, lacks awareness, and can’t be controlled.Kihlstrom’s notion of the subconscious can be inten-tional and efficient, and even can be controlled, butthe key to defining the subconscious according toKihlstrom is the lack of phenomenal awareness.

Psychology that the key to consciousness isself-reference:

In order for ongoing experience, thought,and action to become conscious, a linkmust be made between its mental repre-sentation and some mental representationof the self as agent or experiencer – as well,perhaps, as some representation of the envi-ronment in which these events take place.These episodic representations of the selfand context reside in working memory, butapparently the links in question are nei-ther automatic nor permanent, and mustbe actively forged . . . without such linkagescertain aspects of mental life are dissoci-ated from awareness, and are not accom-panied by the experience of consciousness.(Kihlstrom, 1987, p. 1451)

A great deal of research has demonstratedthe sophisticated and intelligent nature ofthe cognitive unconscious (Epstein, 2001;Lewicki, Hill, & Czyzewska, 1992; Loftus &Klinger, 1992). For instance, after reviewingthe literature on the nonconscious acqui-sition of information, Lewicki, Hill, andCzyzewska (1992) asked, “Is the noncon-scious information-processing system ‘intel-ligent’?” – to which they concluded:

The answer to the question about intelli-gence would be affirmative if intelligence isunderstood as “equipped to efficiently pro-cess complex information.” In this sense,our nonconscious information-processingsystem appears to be incomparably moreable to process formally complex knowl-edge structures, faster and “smarter” over-all than our ability to think and identifymeanings of stimuli in a consciously con-trolled manner. (p. 801)

The idea that the unconscious can besmart is also illustrated by the title of arecent popular summary of the fast-and-frugal heuristics literature: Gut Feelings: TheIntelligence of the Unconscious (Gigerenzer,2007).5 Today there is a strong consensus

5 But note that Gigerenzer (2007; Gigerenzer &Brighton, 2009), in contrast to those who view thecognitive unconscious as able to process complexinformation, views the cognitive unconscious asoperating by the principle “less is more,” selectingthe right rule of thumb for the right situation.



among contemporary researchers in cog-nitive science, philosophy, cognitive psy-chology, social psychology, reasoning, andmorality that humans possess two quite dis-tinct modes of thought – one controlledand the other more automatic (Epstein,2003; Evans & Frankish, 2009; Stanovich &West, 2002). Indeed, dual-process theoriesof cognition are becoming increasingly nec-essary for explaining a wide variety of cogni-tive, personality, social developmental, andcross-cultural phenomena (Evans & Frank-ish, 2009). For instance, Klaczynski (2009)makes a case for adopting and develop-ing a comprehensive dual-process theory ofdevelopment, reviewing studies from suchdiverse research topics as memory, judg-ments and decisions, reasoning, motivatedreasoning, stereotypes, and magical reason-ing to support his argument.

Dual-Process Theories of Cognition

Type 1 processes6 are thought to com-prise a set of autonomous subsystems(Stanovich, 2004) that include both innateinput modules (Fodor, 1983) and domain-specific knowledge acquired by domain-general learning mechanisms that operateautomatically and efficiently (Reber, 1993).Type 1 processes process information fast(relative to type 2 processes); are heavilyinfluenced by context, biology, and pastexperience; and aid humans in mapping andassimilating newly acquired stimuli into pre-existing knowledge structures.

An advantage of type 1 processes overtype 2 processes is that the former require lit-tle conscious cognitive effort and free atten-tional resources for computationally com-plex reasoning. According to Lewicki, Hill,and Czyzewska (1992),

6 Many dual-process theorists refer to two “systems”(see Kahneman & Frederick, 2002). In recent years,however, critics of dual-system theorists have calledfor the use of a different name, arguing that “sys-tem” carries with it a lot of conceptual baggage(see Evans, 2008; Keren & Schul, 2009). In line withEvans’s (2008) suggestion, I refer here to “types” ofthought processes instead of “systems.”

Data indicate that as compared with con-sciously controlled cognition, the noncon-scious information-acquisition processesare not only much faster but are alsostructurally more sophisticated, in thatthey are capable of efficient processing ofmultidimensional and interactive relationsbetween variables. Those mechanisms ofnon-conscious acquisition of informationprovide a major channel for the develop-ment of procedural knowledge that is indis-pensable for such important aspects of cog-nitive functioning as encoding and inter-pretation of stimuli and the triggering ofemotional reactions. (p. 796)

The advantages of type 1 processes can alsobecome disadvantages under certain circum-stances. When thinking is dominated bytype 1 processes, task representations arehighly contextualized. This contextualiza-tion can lead to the thoughtless applica-tion of judgment and decision heuristics.According to Stanovich and West (2000),this mode of thought is in fact the “default”mode in humans. They refer to this ten-dency toward automatic contextualizationof problems as the “fundamental computa-tional bias” in human cognition (Stanovich& West, 2000). A similar idea can be found inChaiken’s (1987) heuristic systematic modelof persuasion, according to which peopleare guided in part by a “principle of leasteffort.” Because people have limited cog-nitive resources, and because heuristic pro-cessing is easy and adequate for most tasks,heuristic processing from type 1 is generallyused unless there is a special need to engagein systematic processing (see also Simon,1979). In line with this idea, Klaczynski andCottrell (2004) have argued that “metacog-nitive intercession” often occurs, wherebyresponses derived from intuition are avail-able in working memory, where reflectionis possible. However, according to Klaczyn-ski, most people do not take advantageof the opportunity to reflect on the con-tents of working memory, taking the con-tents from the experiential system as self-evidently valid. Finally, the view of type1 processes as the default mode of humancognition is also present in Haidt’s (2001)



social intuitionist model of moral reasoning,in which it is posited that intuitive process-ing is the default process, with deliberatereasoning called upon only when intuitionsconflict with reason (see also Stanovich &West, 2000).

In contrast, type 2 processes are typ-ically characterized by deliberately con-trolled, effortful, and intentional cognition.Individual differences in this system havebeen linked in the past to psychometricintelligence (see Stanovich, 2009). Accord-ing to Stanovich and West (1997), a hallmarkof this type of thought is the ability to decon-textualize task representations.7 Type 2 pro-cesses can deal with abstract content underconditions of awareness8 and are not domi-nated by the goal of attributing intentional-ity nor by the search for conversational rel-evance (Margolis, 1987). It has been positedthat type 2 processes are evolutionarily morerecent and uniquely developed in humansthan type 1 processes (Epstein, 2003; Evans,2008; Gabora & Kaufman, 2009).

Note that while some aspects are com-mon across most dual-process theories,there are also distinct differences (Evans,2008). Most dual-process theorists agreeon the automatic/controlled distinctionbetween the two modes of thought, aswell as the idea that type 2 processes areconstrained by a central working memorysystem whereas type 1 processes are uncon-strained by a central pool of resources. Dual-process theorists differ, however, in termsof other features they attribute to the twomodes of thought. For instance, some dual-process theorists emphasize the affectivenature of type 1 processes (Epstein, 1994;Metcalfe & Mischel, 1999; Zajonc, 1980),whereas emotions are not a key componentof other models of implicit cognition (e.g.,Reber, 1993).

Also, as Evans (2008) rightly points out,some of the distinctions between the two

7 Although note that this system can also deal withcontextualized content (see Cokely & Kelley, 2009;Cokely, Parpart, & Schooler, 2009).

8 Although note that some researchers have arguedthat aspects of System 1 (e.g., implicit learning) canalso deal with abstract material (see Reber, 1989).

modes of thought (e.g., abstract vs. contex-tualized, associative vs. rule-based, sharedwith other animals vs. unique to humans)are not as neat and clear-cut when one con-siders that type 1 isn’t a unitary system,but includes a set of autonomous systems,some of which are innately specified andsome of which come about through learn-ing and practice (Stanovich, 2004; but seeEpstein, 2010). Evans (2008) also points outthat “type 2” is most likely not a unitarysystem, suggesting that not all type 2 pro-cesses are consciously controlled. Addition-ally, Cokely and Kelley (2009) and Cokely,Parpart, and Schooler (2009) have noted thateven controlled processes may rely on auto-matic processes for processing, even at thestage of early attentional selection. Othercriticisms (see Aczel, 2009; Gigerenzer &Regier, 1996; Keren & Schul, 2009) have beenleveled against dual-system models, a signthat the study of the dual-process natureof the mind is an active area of researchand debate. In line with these criticisms, theremainder of this chapter will refer to “dual-process” theories instead of “dual-system”theories and will assume that the variousprocesses are not completely independentbut can interact with each other and facil-itate (or inhibit) each other in importantways.

Indeed, in his review of dual-processaccounts of reasoning, judgment, and socialcognition, Evans (2008) notes two distinctkinds of dual-process theories. One kind,which he refers to as “parallel-competitive”forms of dual-process theory, states thatthere are two forms of learning that leadto two forms of knowledge (explicit andimplicit) and each form competes for thecontrol of behavior. Evans refers to anothercategory of dual-process researchers as the“default-interventionists,” who assume thatrapid preconscious processes supply contentfor conscious processing and that the explicitsystem can intervene with the applicationof controlled processes. It should be notedthat not all dual-process theories fall neatlyinto one category or the other. For instance,Epstein (2003) assumes that the two systemsoperate in parallel and are bi-directionally



interactive. As the implicit system has afaster reaction time it is more likely to initi-ate an action sequence. Nonetheless, Evans(2008) does offer a useful classification ofdifferent dual-process theories.

There is evidence for both categories;fMRI evidence suggests that the type ofprocesses are independent – under process-ing conditions that favor automatic process-ing, automatic cognitive processes and thebrain regions supporting those processes aremore active than the brain regions support-ing controlled cognition. Conversely, underconditions that favor controlled processing,controlled cognitive processes and the brainregions supporting those processes (such asthe dorsolateral prefrontal cortex) are moreactive than the brain regions supportingautomatic cognitive processes (Lieberman,2007).

There is also support for the default-interventionists’ view in that humans onaverage have a tendency to contextual-ize information (i.e., automatic cognitionis the default mode in most humans)and that in some instances it is impor-tant for controlled cognition to reflect onthat contextualization and potentially over-ride the outputs of automatic cognition(Kahneman & Frederick, 2002; Chapter 22,Intelligence and Rationality, this volume).Nonetheless, in some situations the outputof the automatic system is beneficial forintelligent behavior, and controlled cogni-tion is not necessary, or can even get inthe way.

Interestingly, a number of neuroimag-ing studies in humans and lesion stud-ies on rodents have found that the basalganglia and medial temporal lobe (mTL)function competitively (Packard, Hirsh, &White, 1989; Poldrack & Packard, 2003). Inan interesting study, Packard, Hirsh, andWhite (1989) found that rats with basal gan-glia lesions performed better than normalrats on an mTL-specific task, and rats withmTL lesions performed better than nor-mal on the basal ganglia–specific task. Theseresults suggest that the presence of a nor-mally functioning medial temporal lobe mayinterfere with performance on tasks that

strongly recruit basal ganglia functions, andperformance is thus improved on these taskswhen the medial temporal lobe is removed(Lieberman, 2007).

Therefore, intelligence and the cognitiveunconscious mostly work in concert witheach other during our daily lives, but insome situations they may be competitive –and depending on the situation, either con-trolled or spontaneous cognitions will be themore important contributor to intelligentbehavior.

Interestingly, while various dual-processtheories of cognition have been proposedover the years, only two are explicitly theo-ries of human intelligence. Below I will reviewboth: Anderson’s (M. Anderson, 2005) theoryof the minimal cognitive architecture under-lying intelligence and development and therecent dual-process (DP) theory of humanintelligence (Kaufman, 2009a).

The Theory of the Minimal CognitiveArchitecture underlying Intelligenceand Development

Based on Fodor’s (1983) distinction betweencentral processes of thought and dedi-cated processing input modules, Anderson’s(2005) theory synthesizes the idea of gen-eral and specific abilities and incorporatesthe notion of development. Anderson arguesthat knowledge is acquired through two dif-ferent “processing routes,” with central pro-cesses (route 1) being tied to individual dif-ferences and input modules being tied tocognitive development (route 2). Accord-ing to Anderson, route 1 involves “thought-ful problem solving” and is constrained bythe speed of a basic processing mechanism.Anderson argues that “it is this constraintthat is the basis of general intelligence andthe reason why manifest specific abilities arecorrelated” (p. 280). Anderson’s basic pro-cessing mechanism comprises both a ver-bal and a spatial processor that are nor-mally distributed, uncorrelated with eachother, and each having their own predictivepowers.

In contrast, the second route for acquiringknowledge in Anderson’s model is tied to



dedicated information-processing modules,such as perception of three-dimensionalspace, syntactic parsing, phonologicalencoding, and theory of mind. Accordingto Anderson, this route is tied to cognitivedevelopment as these modules undergodevelopmental changes in cognitive com-petence across the life span. Andersonacknowledges that modular processes canbe acquired through extensive practice,but both are similar in that they operateautomatically and independently of the firstroute and are therefore unconstrained bythe speed of the basic processing mecha-nism.

Anderson makes the case that the modu-lar component of his cognitive theory allowsfor an integration of Gardner’s “multipleintelligences” and “general intelligence,” asthe theory includes domain-specific mod-ular functions as well as a basic process-ing mechanism. Anderson also argues thathis theory explains how low-IQ individu-als can be capable of remarkable cognitivefeats (e.g., “savant” abilities), including var-ious practical skills, such as the ability toacquire language or see in three dimen-sions that are considerably more compu-tationally complex than the abilities thatare tapped by IQ tests. Anderson arguesthat his theory also can explain how devel-opmental disabilities such as dyslexia andautism can exist in the presence of typi-cal or even above-average IQ (Anderson,2008).

Note that in Anderson’s model there islittle room for individual differences in route2. Furthermore, Anderson does not proposeany domain general learning mechanismsthat are part of route 2, focusing insteadon the Fodorian definition of modules. Bylimiting the cognitive mechanisms associ-ated with each “route,” the total amountof other research that could be broughtto bear on the cognitive processes under-lying the two information-processing routesbecomes unnecessarily restricted. Nonethe-less, Anderson’s model makes an importantcontribution to investigation of intelligenceby expanding modes of thought and incor-porating development.

Dual-Process (DP) Theory ofHuman Intelligence

The dual-process theory of human intelli-gence aims to integrate modern dual-processtheories of cognition (e.g., Evans & Frankish,2009) with research on intelligence (Kauf-man, 2009a). The theory is an organizingframework for various constructs relatingto human cognition that are at least par-tially separable and display individual differ-ences that are meaningfully related to a widerange of socially valued intelligent behav-iors. A main goal of the theory is to expandboth the range of methodologies and thedependent measures traditionally studied byintelligence researchers in order to moreclearly define the cognitive mechanismsunderlying each construct and to developinterventions to increase these abilities ineveryone.

According to the theory, performanceacross a wide range of intelligent behav-iors can be predicted through a hierarchicalstructure of controlled and spontaneous cog-nitive processes. Controlled cognitions aregoal directed and consume limited centralexecutive resources, whereas spontaneouscognitions aren’t constrained by the samelimited pool of attentional resources. Anassumption of the theory is that both con-trolled and spontaneous cognitive processesto some degree jointly determine all intelli-gent behaviors, although in varying degrees.For instance, prediction of performance onan IQ test will maximize the measurementof controlled cognitive processes whereasperformance on a test that requires the inci-dental learning of a complex pattern or per-formance in a domain in which someone hasacquired a large body of expertise will max-imize the measurement of spontaneous cog-nitive processes.

Echoes of this idea can be found inHammond, Hamm, Grassia, and Pearson(1987) when they argue that differentdecision-making situations will draw on dif-ferent strategies in a continuum betweenpure intuition and pure rational analysis.According to the dual-process theory, nei-ther component is more important than the



other, but what is important is the abilityto flexibly switch between modes of cog-nition depending on the task requirements(for applications of this idea to creativity,see Chapter 17, The Evolution of Intelli-gence, this volume; Gabora & Kaufman,2009; Howard-Jones & Murray, 2003;Martindale, 1995; Vartanian, 2009). Accord-ing to the theory, what has traditionallybeen labeled general intelligence (g) is pri-marily tapping into explicit cognitive abil-ity, and the theory predicts that individualdifferences in spontaneous cognition will pre-dict variance in a wide variety of intelligentbehaviors above and beyond the variabilityin g, which itself is thought to be only a partof controlled cognition.

Both forms of cognition involve the abil-ity and the tendency to engage in each modeof thought. The two are related becausepeople tend to engage in things they aregood at and avoid engaging in things theyaren’t good at. A key assumption of thedual-process theory is that abilities are notstatic entities but are constantly changingthrough the life span as the person contin-ually engages with the world. The more aperson engages in a mode of thought, themore that individual will develop skills inthat modality, which in turn increases thedesire for engaging with that skill. Indeed,research on expertise skill acquisition showsthat engagement in a domain through manyhours of deliberate practice contributes tothe generation of mental structures that cansurpass information-processing limitationswhen performing within that domain (Eric-sson & Charness, 1994; Ericsson & Kintsch,1995; Ericsson & Lehmann, 1996, but seeKaufman, 2007).

Controlled cognition is at the top of thehierarchy (alongside spontaneous cognition)because the capacity for goal-directed actionis an important component of human intel-ligence. Controlled cognition consists of aclass of cognitive processes that involvethe ability and tendency across situationsto think about thinking (i.e., “metacogni-tion” – see Dennett, 1992; Hertzog & Robin-son, 2005), reflect on prior behavior, anduse that information to modify behavior

and plan for the future.9 Constructs thatare part of the controlled cognition hier-archy include central executive functions(updating, cognitive inhibition, and mentalflexibility), reflective engagement, explicitcognitive ability (the skill sets that lie atthe heart of highly g-loaded tasks), intellec-tual engagement, and elementary cognitivetasks that support explicit cognitive ability.10

What links all of the processes together isthat they all draw on a limited capacity poolof attentional resources.

The second main component (alongsidecontrolled cognition) of the dual-process the-ory, and the component that contains pro-cesses relating to the cognitive unconscious,is spontaneous cognition. At the broadestlevel, individual differences in spontaneouscognition reflect the ability to acquire infor-mation automatically and the tendency toengage in spontaneous forms of cognition.For instance, whereas most people have theability to spontaneously experience emo-tions and daydream, there may be indi-vidual differences in the extent to whichpeople are willing to engage in their emo-tions and to daydream (see Pacini & Epstein,1999; Zhiyan & Singer, 1997).11 Constructsthat are part of the spontaneous cogni-tion hierarchy include spontaneous informa-tion acquisition abilities (implicit learning,reduced latent inhibition, etc.), spontaneousforms of engagement (affective engagement,

9 Note that other definitions of “controlled cognition”have been put forward (see Schneider & Shiffrin,1977).

10 It should be noted, however, that elementary cogni-tive tasks (ECTs) are not process pure, and motiva-tion, strategy use, and the allocation of attentionalresources play an important role in performance(see Chapter 38, Intelligence and Motivation, thisvolume; Cokely, Kelley, & Gilchrist 2006; Fox,Roring, & Mitchum, 2009).

11 Note that the distinction between controlled andspontaneous cognition is not always the sameas the distinction between conscious and uncon-scious modes of thought. Spontaneous cognitionscan be either conscious, such as when individualsare consciously aware of their daydreaming, fan-tasy, or mind wandering, or nonconscious such aswhen individuals are dreaming, daydreaming with-out conscious awareness, or implicitly learning theunderlying rule structure of the environment with-out awareness of how that tacit knowledge is affect-ing their behavior.



aesthetic engagement, and fantasy engage-ment), and various implicit domains of mindthat are universal human domains pertainingto knowledge of people, language, numbers,animals, music, visual images, aesthetics, orthe inanimate physical world (see Carey &Spelke, 1994; Feist, 2001; Hirschfeld & Gel-man, 1994).12

Other technical details about the the-ory, including the hierarchical nature of themodel can be found in Kaufman (2009a).Thus far, there is support for the theory fromdifferent branches of psychology and neu-ropsychology. The theory has not receivedmany criticisms, but it is still new; thus,the extent to which the dual-process the-ory of human intelligence advances the fieldby making new, testable predictions and theextent to which the theory more clearlydefines various constructs relating to intel-ligence is still to be determined.

The rest of this chapter reviews recentempirical work on linkages between the cog-nitive processes underlying psychometricintelligence and various aspects of the cog-nitive unconscious. First, relations betweenindividual differences in controlled cognitiveprocessing and individual differences in twoforms of preconscious processing, implicitlearning, and latent inhibition will be dis-cussed. Because intuitions and insights gen-erally follow preconscious processing, thenext section of this chapter reviews evi-dence on the relation between intelligenceand individual differences in both intuitionand insights. The following section will thenlook at the implications of intelligence andthe cognitive unconscious for two major

12 Implicit domains of mind are similar to group fac-tors in hierarchical models of intelligence. Indeed,research shows that group factors, such as mathe-matical, spatial, and verbal reasoning abilities pro-vide incremental validity for predicting associatedvocations above and beyond general intelligence(Achter, Lubinski, Benbow, & Eftekhari-Sanjani,1999; Humphreys, Lubinski, & Yao, 1993). Thesedomains of mind are also related to Howard Gard-ner’s “multiple intelligences” (Gardner, 1993, 1999),although the dual-process theory acknowledges thatthere are also more general forms of cognition thatcontribute to intelligent behavior, a criticism thatis often leveled against theories of multiple intelli-gences (see Lohman, 2001).

domains of human cognitive functioning:social cognition and creative cognition. Thechapter will then conclude with a call formore research. The review of studies in thischapter is by no means exhaustive but ismeant to highlight some of the latest think-ing and research on the relation betweenindividual differences in psychometric intel-ligence and individual differences in the cog-nitive unconscious.

Intelligence and PreconsciousProcessing

Intelligence and Implicit Learning

According to Reber (1993), implicit learn-ing is “a fundamental root process . . . thatlies at the very heart of the adaptive behav-ioral repertoire of every complex organism”and can be characterized as “the acquisitionof knowledge that takes place largely inde-pendent of conscious attempts to learn andlargely in the absence of explicit knowledgeabout what was acquired” (p. 5; for a sim-ilar view see Epstein & Meier, 1989). Wefrequently encounter many complex contin-gencies and patterns, and the ability to pre-consciously learn patterns and then use thatknowledge to recognize and detect patternsin the future is an important component ofintelligence (see Hawkins, 2005).

What is the link between psychometricintelligence and implicit learning? Accord-ing to Reber (1993) and Epstein and Meier(1989), individual differences in implicitlearning should be unrelated to individualdifferences in measures of explicit cognition.Applying principles of evolutionary biology,they argue that the capacity for explicitcognition arrived later on the evolutionaryscene than did implicit cognition. Nonethe-less, the older implicit learning mecha-nisms were unaffected by the emergence ofexplicit thought and continue to functionautonomously.

Thus far, the majority of the evidencesupports the notion that implicit learningability is independent of IQ. Some implicitlearning tasks have never demonstrateda relation with explicit cognitive ability



(e.g., artificial grammar learning; Gebauer& Mackintosh, 2007; McGeorge, Crawford,& Kelly, 1997; Reber, Walkenfeld, & Hern-stadt, 1991), whereas other tasks have notshown a significant association in the major-ity of the studies (e.g., serial reaction timelearning; Feldman, Kerr, & Streissguth, 1995;Kaufman, DeYoung, Gray, Jimenez, Brown,& Mackintosh, 2010; Pretz, Totz, & Kauf-man, 2010; Unsworth & Engle, 2005 – but seeSalthouse, McGuthry, & Hambrick, 1999).One other implicit learning task, whichinvolves unintentional exposure to pictures,did show an association once with explicitcognitive ability (Fletcher, Maybery, & Ben-nett, 2000). These results may be mixedas different implicit learning tasks are onlyweakly correlated with each other (Gebauer& Mackintosh, 2007, 2009; Salthouse et al.,1999). Further, some implicit learningparadigms may better capture implicit cog-nition than others, which may draw moreon explicit cognition (e.g., Seger, 1994). Animportant future line of research to betterunderstand the relation of implicit learn-ing to psychometric intelligence will be toconstruct reliable measures that more accu-rately assess implicit learning. Then, the fac-torial structure of implicit learning tasks canbe assessed and the convergent-discriminantvalidity can be compared to other measuresof psychometric intelligence.

Another methodology with which toinvestigate the link between implicit andexplicit cognition is to compare implicitand explicit versions of the same task.In one condition, experimenters instructedparticipants to find the pattern, whereasin another condition participants receivedno such instruction, thereby making learn-ing unintentional. When this methodol-ogy is employed, psychometric intelli-gence is more highly correlated with thetask under explicit instructions comparedwith the condition in which participantsare not instructed to intentionally searchfor the pattern (Gebauer & Mackintosh,2007; Unsworth & Engle, 2005). Usinga similar methodology, Feldman, Kerr,and Streissguth (1995) separated an inten-tional declarative component of an implicit

learning task from the procedural compo-nent using a sample of 455 adolescents;they found that while the declarative learn-ing component significantly correlated withexplicit cognition, the procedural compo-nent did not. In another line of research,using a population of individuals with autis-tic spectrum condition (ASC), Brown et al.(2010) found that matching for IQ, there wasstatistical equivalence between participantswith ASC and typically developing individ-uals on four implicit learning tasks. Further,this finding was not a consequence of com-pensation by explicit learning ability or IQ.Taken together, the research supports theseparation of explicit and implicit cognitionand the notion that individual differences inpsychometric intelligence are only weakly ifat all associated with individual differencesin implicit learning (e.g., McGeorge et al.,1997; Reber et al., 1991).

Recent research has found that individ-ual differences in implicit learning make anindependent contribution to complex cogni-tion above and beyond psychometric intel-ligence. Gebauer and Mackintosh (2009)administered a large battery of implicitlearning and intelligence tests to 195 Ger-man students. A factor analysis of all thetasks revealed two second-order principalcomponents: the first consisting primarilyof the intelligence measures and the secondconsisting of the measures of implicit learn-ing. Both factors were only weakly related toeach other. Additionally, the implicit learn-ing second-order factor was significantlyrelated to math and English grades, subjectsthat were foreign languages for the Germanstudents in the sample. Controlling for theintelligence second-order factor, the associ-ation between the implicit learning factorand English remained whereas the associa-tion with math was no longer significant.

Consistent with this finding, Pretz, Totz,and Kaufman (2010) found a relationbetween a probabilistic sequence learningtask and both the American College Test-ing (ACT) math and English scores, andthese effects were in the middle third ofeffect sizes reported in psychology (r = .2to .3; Hemphill, 2003). In another recent



study, Kaufman et al. (2009) investigatedthe association of individual differences inimplicit learning with a variety of cogni-tive and personality variables in a sampleof English 16- to 17-year-olds. Probabilisticsequence learning was related to intentionalassociative learning more strongly than psy-chometric intelligence, and it was not asso-ciated with working memory. Furthermore,structural equation modeling revealed thatindividual differences in implicit learningwere independently related to verbal ana-logical reasoning and processing speed, andimplicit learning was significantly corre-lated with academic performance on twoforeign language exams (French and Ger-man). Implicit learning also was positivelyrelated to self-report measures of personal-ity, including intuition, Openness to Expe-rience, and impulsivity. Also, a doubledissociation was found between a latentIntellect factor and a latent Openness toExperience factor – with Intellect relatingto working memory (.29) but not implicitlearning (.00) and Openness to Experiencerelating to implicit learning (.31) but notworking memory (.13).

This lack of association between implicitlearning and working memory is consistentwith other research on attention and execu-tive functioning. Research shows that thosehigh in working memory are better able tocontrol their attention and stay on task whenthere is interference (Kane, Bleckley, Con-way, & Engle, 2001) and this ability is asso-ciated with psychometric intelligence (seeChapter 20, Working Memory and Intel-ligence, this volume). There is an emerg-ing consensus that implicit learning requiresselective attention to the relevant stimulibut then learning about the selected stim-uli operates automatically, independent ofan intention to learn and without drawingon further central executive processing (e.g.,Baker, Olson, & Behrmann, 2004; Frensch &Miner, 1995; Jiang & Chun, 2001; Jimenez& Mendez, 1999; Turke-Browne, Junge, &Scholl, 2005).

Indeed, researchers have proposed thatcentral executive functions should beengaged only under intentional learning

conditions to aid in focusing attention,whereas only selective attention processesare necessary for learning stimuli inciden-tally (Cowan, 1988; Frensch & Miner, 1995,Johnson & Hirst, 1993). In support of thisview, Unsworth and Engle (2005) found thatvariations in working memory were asso-ciated with an implicit learning task onlywhen participants were instructed to explic-itly detect the covariation, but no associa-tion with working memory was found whenparticipants were not given that instruction.Feldman, Kerr, and Streissguth (1995) alsofound no relation between implicit learningand measures of working memory.

In sum, while the literature is not large,the evidence that does exist suggests thatimplicit learning is often unrelated to psy-chometric intelligence or working memorybut is independently associated with spe-cific forms of complex cognition, academicachievement, and particular aspects of per-sonality related to Openness to Experienceand impulsivity. Future research on thetopic is needed to clarify and extend thesefindings.

Intelligence and Latent Inhibition

It can be important in our everyday lives tobe able to automatically distinguish relevantfrom irrelevant stimuli and to filter outinformation irrelevant to the task at hand.For instance, when trying to concentrate onwriting poetry, it’s important to filter outthe rattle of the radiator. Such a mechanismhas been investigated and is called latentinhibition (Lubow, 1989). Latent inhibitionis often characterized as a preconsciousgating mechanism that screens from currentfocus those stimuli that have previouslybeen regarded as irrelevant (Lubow, 1989).Those with increased latent inhibitionshow higher levels of this form of inhibition(Peterson, Smith, & Carson, 2002). Variationin latent inhibition has been documentedacross a variety of mammalian species and,at least in other animals, has known biolog-ical substrates (Lubow & Gewirtz, 1995).Prior research has shown a relation betweendecreased latent inhibition and acute-phase



schizophrenia (Baruch, Hemsley, & Gray,1988a, 1988b; Lubow, Ingberg-Sachs,Zalstein-Orda, & Gewirtz, 1992). Peoplewith schizophrenia also tend to havereduced ability for central executivefunctioning (Barch, 2005).

Recent research suggests that reducedlatent inhibition can also have its advan-tages. In students with a high IQ (and pre-sumably a high level of central executivefunctioning), decreased latent inhibition isassociated with higher scores on a self-reportmeasure of creative achievement (Carson,Peterson, & Higgins, 2003). Interestingly,the researchers did not find a correlationbetween fluid intelligence and latent inhi-bition. Kaufman (2009a) also did not findan association between variations in g andvariations in latent inhibition. Additionally,Kaufman (2009a, b) examined the relation-ship between latent inhibition and individ-ual differences in the tendency to rely onintuition to make decisions. Indeed, latentinhibition is conceptually related to intu-ition: Jung’s original conception of intuitionis “perception via the unconscious” (Jung,1921/1971, p. 538). Kaufman hypothesizedthat an intuitive cognitive style would berelated to reduced latent inhibition. Resultsshowed that those with higher scores on afaith in intuition factor (consisting of intu-ition items related to affect) tended to havereduced latent inhibition. Further, latentinhibition was not associated with an intu-ition factor consisting of items having to dowith holistic processing of information or arational cognitive style. There was also a ten-dency for those scoring high (as comparedto medium or low) on the faith in intu-ition factor to benefit more from a preexpo-sure condition where participants receivedthe relevant stimuli in the first part of thetask. Therefore, current research suggeststhat decreased latent inhibition is unrelatedto general intelligence or a rational cognitivestyle. Since decreased latent inhibition maymake an individual more likely to perceiveand make connections that others do not see,this ability in combination with high psycho-metric intelligence can lead to the highestlevels of creative achievement.

Intelligence, Intuition, and Insight

Various researchers have come to the con-clusion that in many naturalistic situations,such as decision making in groups, very lit-tle controlled cognition is required (Klein,1999; also see Gladwell, 2007, for a summaryof relevant research). Instead, they note thatexpertise seems to be related to recognitionof a situation that had been encounteredpreviously and the retrieval of schemas thatmatch the situation.13 They argue that whilecontrolled cognition is sometimes impor-tant, the key to intelligent behavior is theautomatic retrieval process.

Similarly, Reyna (2004) argued thatexperts acquire knowledge that allows themto make fast, intuitive, and effective deci-sions whereas novices need to rely ondeliberate, effortful reasoning. Reyna noted,however, that automatic processes can leadto bias and error when experts are presentedwith novel problems (also see Chabris &Simons, 2010, for a summary of researchshowing the potential perils of relying onintuition when making expert as well asnovel decisions). Wilson and Schooler (1991)also showed the importance of automaticprocessing in decision making – they demon-strated that when making a decision thatis complex and multi-attributed, people dobetter when conscious deliberation is inten-tionally prevented. This idea is also a majortenet of the unconscious thought theory(UTT), in which it is argued that decisionsabout simple issues can be better tackled byconscious thought, whereas decisions aboutcomplex matter can be better approachedwith unconscious thought (Dijksterhuis &Nordgren, 2006, but see Aczel, 2009; Newell,Wong, & Cheung, 2009; Payne, Samper,Bettman, & Luce, 2008; Thorsteinson &Withrow, 2009).

13 For more on the relations between intelligenceand the acquisition of expertise more generally,see Ackerman (Chapter 41, Intelligence and Exper-tise, this volume). In this section I focus insteadon the relation between intelligence and intuition,particularly from an individual differences perspec-tive.



Along similar lines, Hogarth (2005) distin-guished between deliberate and tacit cog-nitive processes. According to Hogarth,complex decisions will benefit from tacitprocessing whereas less complex decisionswill benefit from deliberate processing. Anadditional component in Hogarth’s modelis the degree of bias in the original learn-ing environment. If the feedback presentedin the original learning environment regard-ing decision accuracy is clear and immedi-ate, the environment is considered “kind,”and accurate causal relationships can belearned. Environments in which feedback isunclear and not available in a timely man-ner are considered “wicked” and are consid-ered highly biased. In wicked learning envi-ronments, the intuitive system is prone toerrors. According to Hogarth, intentional,deliberate thought is best suited to biasedlearning environments where the complex-ity of the task is low, whereas intuitive pro-cessing is best suited to learning environ-ments in which bias is low and complexity ofthe task is high (see Epstein, 2003, and Kah-neman, 2009, for related ideas, including thenotion that the quality of an intuitive judg-ment is dependent upon the predictabilityof the environment in which the judgmentis made and the individual’s opportunity tolearn the regularities in that environment).

Recently, researchers have investigatedthe role of individual differences in the useof intuition. With the aim of integrating thepsychodynamic focus on unconscious pro-cessing with the cognitive focus on ratio-nal conscious thinking, Seymour Epstein putforth the cognitive-experiential self-theory(CEST; Epstein, 1994), which was an out-growth of ideas presented in Epstein (1973).The theory posits that humans have two par-allel but interacting modes of informationprocessing. The rational system is analytic,logical, abstract, experienced actively andconsciously, is slower to process informa-tion, and requires justification via logic andevidence. In contrast, the experiential systemis holistic, affective, concrete, experiencedpassively, processes information automati-cally, and is self-evidently valid (experiencealone is enough for belief).

Epstein’s experiential system is related tointuition in the sense of “gut-feelings” thatguide behavior. Based on his theory, Epsteindeveloped the Rational-Experiential Inven-tory (REI; Pacini & Epstein, 1999), whichmeasures individual differences in the ten-dency to rely on each mode of thought.His research program has discovered thatthe intelligence of each system is indepen-dent of, or very weakly correlated with,the intelligence of the other (Epstein &Meier, 1989), and each subscale (analyti-cal and experiential) has unique predictivevalidity for a wide range of intelligent behav-iors (see Epstein, 2003, for a review). Ingeneral, the rational scale is more stronglypositively related to measures of intellectualperformance such as scores on the Scholas-tic Aptitude Test (SAT) and grade pointaverages (GPA) than is the experientialscale, whereas the experiential scale is morestrongly positively related to extroversion,agreeableness, favorable interpersonal rela-tionships, empathy, creativity, emotionality,sense of humor, and art appreciation than isthe rational scale. The rational scale is morestrongly negatively associated with neuroti-cism, depression, anxiety, stress in collegelife, subtle racism, extreme conservatism,alcohol abuse, and naıve optimism than isthe experiential scale, whereas the experien-tial scale is more strongly negatively associ-ated with distrust and intolerance than is therational scale. Many of these relations heldeven after controlling for the NEO Five Fac-tor Inventory (NEO-FFI; Costa & McCrae,1989), which measures the Big Five fac-tors of personality. Other researchers haveused the REI to investigate human cogni-tion. For instance, Klaczynski (2009) reviewsa number of studies he and his collaboratorsconducted using the REI to investigate thedevelopment of dual processes across the lifespan.

Pretz (2008) has extended both the exper-imental work on intuition and the cognitivestyles approach by looking at the effects ofindividual differences in an analytical ver-sus intuitive strategy and level of experienceon practical problem solving. Pretz reasonedthat the more experienced an individual is



with a task, the less complex the task andthe more decomposable the problem willappear to that individual. Pretz noted thatthe relevant knowledge associated with aneveryday problem-solving task is likely to beacquired through informal experience, andindividuals with more experience will there-fore have more tacit knowledge but will alsobe able to better articulate that knowledge.As a result, the expert can use metacognitiveskills to explicitly identify the main prob-lem, identify the most relevant information,and identify the consequences of variouscourses of action (Antonakis et al., 2002).

In Pretz’s study, college students wereinstructed to use either holistic intuition(bringing to mind all relevant informationand trusting hunches) or analysis (defin-ing the problem, distinguishing the relevantfrom irrelevant information, and monitoringthe problem carefully) when solving vari-ous practical problems dealing with collegelife. Pretz found that the effectiveness of thestrategy on task performance interacted withthe participant’s level of experience: analysisworked better for more experienced individ-uals whereas novices were slightly more suc-cessful when they employed a holistic, intu-itive strategy. A similar pattern was foundlooking at existing individual differences instrategy preference.

Pretz’s study suggests that among indi-viduals with an intermediate level of exper-tise, analytical problem solving can be help-ful in perceiving the logic and structureof the problem, and intuition can distractthe expert from this critical information. Incontrast, intuitive, holistic thought may bebest suited for novices in a domain who seethe task as ill-defined and need to bring tomind the relevant information. An impli-cation of Pretz’s study is that intermediateexperts should rely on an analytical strategywhen solving complex, practical problems.Full-blown experts who have fully automa-tized their task may benefit from an intuitivemode of thought.

This distinction between holistic intu-ition (of the sort studied in Pretz’s study)and inferential intuition (full automatiza-tion) was made by Hill (1987–1988); the ideas

are consistent with Baylor’s (Baylor, 2001) U-shaped model of expertise and intuition andresearch showing the facilitation of intuitionfor complex, high-stakes decision making(Klein, 1999). Indeed, Pretz and Totz (2007)have developed a scale to measure individ-ual differences in the tendency to rely onthree different forms of intuition: affective,heuristic, and holistic. Another implicationof Pretz’s study is that many social prob-lems may be better suited to the cognitiveunconscious, as they may be more complexthan nonsocial problems. Whereas individ-ual differences in the cognitive unconsciouscan be adaptive for some social problems,there may be instances of social cognitionin which the cognitive unconscious can leadto undesirable outcomes (see Implicit SocialCognition section).

Another line of research has investigatedthe intimate connection between intuitionand insight. Anecdotally, insight has playeda crucial role in the generation of cre-ative ideas. The great French mathemati-cian Henri Poincare (1921) described inci-dents in which an answer came to him onlyafter his conscious attention was directedaway from the problem and he wasn’tconsciously deliberating on the problem.Poincare argued that these moments of sud-den inspiration are the result of unconsciousthinking. Based on reflections of his creativethought process, he argued that the cre-ative process starts with conscious work ona problem, followed by unconscious work,and then, if insight is successful, anotherstage of conscious work to verify that theideas makes sense and to work out the impli-cations of the idea. Indeed, insight is consid-ered an important component of the cre-ative process (Wallas, 1926).

Empirical work supports these anecdotes.In reviewing a number of experimentsrelating to implicit thought, intuition, andinsights, Kihlstrom, Shames, and Dorfman(1996) have this to say about the nature ofintuition:

From the experiments described in thischapter, it appears that the processesunderlying intuitions closely resemble those



which underlie implicit memory. In recog-nition, people’s intuitions about the past –the feeling of familiarity, in the absence offull recollection – seems to be based on theperceptual fluency that comes with prim-ing. . . . We actually think of these men-tal states as implicit thoughts: instances inwhich an idea or image influences experi-ence, thought, or action in the absence ofconscious awareness of what that idea orimage is.

As for the link between intuitions andinsight, they then go on to say:

. . . it is clear that problem solutions, likememories, are not discontinuous, all-or-none affairs, remaining entirely uncon-scious until they emerge full-blown intothe full light of consciousness. There is apoint, as they approach and cross whatWallas (1926), following William James(1890), called the “fringe” of consciousness,when we know they are coming, even whenwe do not know what they are. This isthe point, between preparation and insight,where intuitions occur. (p. 19)

Other researchers have investigated the con-trolled and spontaneous cognitive mecha-nisms that underlie insight (see Sternberg& Davidson, 1995, for a review of researchon insight). A methodology that is oftenemployed is the Accumulated Clues Task(ACT), in which participants must discovera word, but are given clues (e.g., words thatare associated with the answer) along theway. After each clue is presented, partic-ipants are required to provide an answer.The clues get increasingly helpful (are morerelated to the answer) and the answers givenby the participants get objectively closer tothe answer in an incremental fashion thatoccurs before their subjective ratings of feel-ing close to an answer, which they oftenreport occurring to them in a sudden flashof insight (Bowers, Farvolden, & Mermigis,1995; Dorfman, Shames, & Kihlstrom, 1996).Research has shown that individual differ-ences in how long it takes participants toarrive at the correct answer correlate withverbal intelligence.

Recent research, however, suggests thatdifferent components of the task maydifferentially relate to controlled cognition.Reber, Ruch-Monachon, and Perrig (2007)first replicated earlier research on the ACTby finding that participants often under-estimated their degree of closeness to theanswer; these subjective reports of close-ness exhibited a positive slope, suggestingthat participants possessed implicit knowl-edge about the task and indeed felt hunchesabout their progress that weren’t necessarilyaligned with objective incremental progress.The researchers then distinguished betweenperformance level, processing style, implicitknowledge, and subjective feeling of close-ness to the solution on the ACT. Whileperformance level correlated with verbalintelligence, processing style and implicitknowledge were not correlated with ver-bal intelligence. Further, a faith in intu-ition cognitive style and the Big Five per-sonality traits Openness to Experience andConscientiousness were all correlated withprocessing style, but not with implicitknowledge on the task. These results sug-gest that a promising research direction is todecompose problem-solving tasks into theirprocessing style and intuitive componentsand investigate relations between individualdifferences in these components and indi-vidual differences in various processes andthinking styles relating to intelligence andthe cognitive unconscious.

Domains

Implicit Social Cognition

There is an emerging consensus in thesocial cognition literature that many of oursocial behaviors and judgments are madeautomatically, without intention, effort, orawareness (Bargh & Chartrand, 1999; Bargh& Morsella, 2008). Research on automaticevaluation, impression formation, and auto-matic characterization all demonstrate theprevalence of automaticity in social life. Itis generally thought now that mere percep-tion of a stimulus can lead instantly and



automatically to a judgment without anyconscious reflection or reasoning. Indeed,until the 1980s, attitudes were mostlyassumed to rely on consciously availableinformation (Nosek, Greenwald, & Banaji,2007).

Recently, researchers have investigatedindividual differences in implicit socialcognition, using a variety of measures“that avoid requiring introspective access,decrease the mental control available to pro-duce the response, reduce the role of con-scious intention, and reduce the role of self-reflective, deliberative processes” (Nosek etal., 2007, p. 267). Greenwald and Banaji(1995) have been among the most activeresearchers investigating the role of implicitcognition in various social psychology con-structs such as attitudes, stereotypes, andself-esteem. In their research, they attemptto “reveal traces of past experience that peo-ple might explicitly reject because it con-flicts with values or beliefs, or might avoidrevealing because the expression could havenegative social consequences. Even morelikely, implicit cognition can reveal infor-mation that is not available to introspec-tive access even if people were motivated toretrieve and express it” (Nosek et al., 2007,p. 266; see Wilson, Lindsey, & Schooler,2000, for related ideas about attitudes).

One of the best-validated measures ofimplicit social cognition is the Implicit Asso-ciation Test (IAT; Greenwald, McGhee,& Schwartz, 1998). The IAT requires theparticipant to categorize various stimu-lus exemplars representing four concepts(e.g., men, women, good, bad) using tworesponse options. When concepts that sharea response are strongly associated, it isexpected that the sorting task will be eas-ier for the participant (as indexed by fasterresponses and fewer errors) than when theconcepts are weakly associated. Thus, theIAT affords insight into automatic associa-tive processes that are introspectively inac-cessible. Over the last decade, the IAT hasbeen adapted for use in various disciplines(see Nosek et al., 2007, for a review) and toassess implicit attitudes related to categories

such as race, gender, and even insects. Instudies that involve some measure of dis-crimination toward a social group, bothexplicit and IAT measures predict behav-ior, with the IAT offering superior pre-diction (Greenwald, Poehlman, Uhlmann,& Banaji, 2009). Furthermore, it has beendemonstrated that people with the strongestautomatic racial biases are most likely toengage in a wide variety of discriminatorybehavior, including overt behavior (Rudman& Ashmore, 2007, but see van Ravenzwaaij,van der Maas, & Wagenmakers for an alter-native account).

Therefore, research on how individualdifferences in intelligence and the cognitiveunconscious interact to produce stereotyp-ing and attitude formation is of both theoret-ical and practical interest. Recent researchutilizing fMRI techniques provides someclues. Chee, Sriram, Soon, and Lee, 2000)used fMRI to examine participants whilethese individuals were taking the IAT. Theresearchers found that the left dorsolateralprefrontal cortex and to a lesser degree theanterior cingulate were most active dur-ing conditions in which items from incon-gruent categories (e.g., insect + pleasant)shared a response key than when itemsfrom congruent categories (e.g., flower +pleasant) shared a key. According to theresearchers, this suggests that greater con-trolled cognition was required in condi-tions in which it was necessary to overcomethe prepotent tendency to map emotion-ally congruent items to the same responsekey. In another study, Phelps et al. (2000)had White participants view faces of unfa-miliar Black and White males. Participantswho showed greater activation of the amyg-dala (a region of the brain associated withfear and negative emotions) while viewingBlack faces relative to White faces tendedto score higher on two measures of uncon-scious race evaluation: the IAT and the eye-blink response. In a second experiment, theresearchers did not find the same patternof brain activation when the faces werefamiliar and the participants regarded theBlack and White individuals positively. In a



related study, Cunningham et al. (2004) hadparticipants view Black and White faceseither subliminally or supraliminally dur-ing fMRI. When presented subliminally,the amygdala was more active for Blackfaces relative to White faces. This effectwas reduced when the faces were presentedsupraliminally. Further, control regions inthe prefrontal cortex (which are also acti-vated during working memory and psycho-metric intelligence tests) showed greateractivation for Black faces than White faceswhen presented supraliminally. Race bias asassessed by the IAT was related to a greaterdifference in amygdala activation for Blackfaces relative to White faces, and activityin the prefrontal cortex predicted a reduc-tion in amygdala activation from the sublim-inal to the supraliminal condition. Accord-ing to the researchers, this provides evidencefor neural distinctions between automaticand controlled processing of social groups,suggesting that controlled processes (whichsupport performance on measure of psy-chometric intelligence) may modulate auto-matic evaluation.

These results suggest that individual dif-ferences in measures of controlled cogni-tion may predict the extent to which auto-matic evaluations influence behavior. Toexpand the range of individual differences inimplicit social cognition investigated, it maybe useful to construct new implicit learningtasks that consist of stimuli relating to thelearning of real-world contingencies in thesocial domain. Tasks that already exist thatcould be adapted include the task used byLewicki, Hill, and Sasaki (1989), in whichparticipants implicitly learn to judge theintelligence of individuals from brain scansor the adaptation of that task employed byWoolhouse and Bayne (2000), in which par-ticipants implicitly learn to judge the jobsuitability of job candidates based on theirpersonality profile. Such research can helpdistinguish between situations in which indi-vidual differences in the cognitive uncon-scious contribute to intelligent behavior (forexample, when a person is engaging in anarea of expertise or generating novel ideas),and situations in which controlled cognition

may be the better predictor of intelligentbehavior (since it helps override generaliza-tions that can lead to explicit prejudice andstereotyping). Such research would furtherillustrate the need for measuring individualdifferences in both controlled and automaticcognitive processes in order to predict vari-ous forms of intelligent behavior.

Creative Cognition

Creativity requires both novelty and useful-ness (Kaufman, 2007). The Creative Cog-nition Approach endeavors to identify andinvestigate the role of mental processes increative cognition at various stages in thecreative process (Finke, Ward, & Smith,1992, 1995; Ward, Smith, & Finke, 1999).Creative cognition researchers have identi-fied two main phases of creative inventionthat occur in a cyclical fashion in ordinaryindividuals. During the generative phase,the individual generates numerous candi-date ideas or solutions and forms a men-tal representation (referred to as a preinven-tive structure). Then during the exploratorystage, the individual examines the can-didate mental representations and ideasand consciously and sometimes painstak-ingly works out their implications. Cognitiveunconscious processes activated throughdefocused attention most likely play moreof a role during the generative stage,whereas controlled cognitive processes acti-vated through focused attention most likelyplay more of a role during the exploratorystage. The highest levels of creativity, how-ever, most likely require the ability for bothmodes of thought and the flexibility toswitch modes of thought throughout thecreative process.

On the one hand, behavioral and brainstudies suggest that creative people are char-acterized by a lack of inhibition (Eysenck,1995; Martindale, 1999), and case studiesrepeatedly show that creative people dodescribe the creative process as effortlessand lacking in deliberation (Csikzentmiha-lyi, 1996). However, studies also show thatcreative individuals defocus their attentionwhen approaching a creative task but they



are capable of focusing their attention whenit comes time to make the ideas practical(Martindale, 1999). In recent years, OshinVartian and colleagues have extended thisresearch by showing in a series of cleverexperiments that creative people are able toadjust their focus of attention, depending onthe demands of the task.

In one study, Vartanian, Martindale, andKwiatkowski (2007) found a negative corre-lation between creative potential (measuredby fluency scores) and speed of informa-tion processing on two tasks that did notinvolve interference or ambiguity, and a pos-itive correlation between creative potentialand speed of information processing on twotasks that did require the inhibition of inter-fering information. Therefore, subjects withgreater creative potential were better ableto slow down or speed up their informationprocessing, depending on the task demands.A follow-up study found similar results andextended the earlier results in a sample ofhigh school students in Russia (2008). Thesame pattern was found between creativepotential (as measured by fluency, flexibil-ity, and originality) and response latency asin the earlier study, and the findings held,correcting for IQ. In a third study, partici-pants were instructed to judge whether twoconcepts were related or unrelated (Varta-nian, Martindale, & Matthews, 2009). Therationale was that creativity is frequentlydefined as the novel and useful associationof concepts that are not traditionally related.Therefore, this important cognitive processrelies at least in part on a person’s abilityto quickly assess the degree of relationshipbetween concepts. The researchers manip-ulated the degree of association betweenword pairs. Participants with greater cre-ative potential (assessed by a measure ofdivergent thinking) exhibited a faster reac-tion time when judging the relatednessof the concepts. Psychometric intelligencedidn’t account for additional variance aboveand beyond divergent thinking scores in pre-dicting the variability in reaction time per-formance. The researchers conclude that theability of individuals with higher creativepotential to more quickly judge the degree

of association between words can lead toan advantage over time in the total numberof potentially relevant conceptual associa-tions that can be considered. The researchersargued that the task they used involvedunambiguous task instructions and associa-tions and that it is just these conditions inwhich those with better divergent thinkingskills focus their attention, which can resultin a faster reaction time.

An interesting question raised by Var-tanian, Martindale, and Matthews (2009)is whether the mechanism that regulatesthe focus of attention is itself automaticor requires self-control. They have arguedthat the unambiguous nature of their taskled to automatic regulation of attention.They point to evidence that in other circum-stances, top-down processing can also playan important role in creative cognition. Var-tanian, Martindale, and Kwiatkowski (2003)investigated the role of strategic flexibilityin creative problem solving. They adminis-tered a rule discovery task and found thatparticipants with higher creative potential(as measured by fluency scores) were betterat discovering the rules. Further, the strat-egy of generating disconfirmatory hypothe-ses played an important role for successfulparticipants in the later stages of hypothe-sis testing after the first feedback was given.Having already formed a representation ofthe problem space after feedback, successfulparticipants were flexibly able to switch toa more successful strategy following initialfeedback. Similar results have been foundby Gilhooly, Fiortou, Anthony, and Wynn(2007), who found that using think-aloudprotocols that alternative uses for a task hadgenerated earlier in the course of the taskdrew primarily on memory-based strategies,whereas uses generated later drew on a morelimited range of strategies requiring exec-utive processes, such as imagining the dis-assembly of the object and using the partsor recombining the parts into other objectsthat could be applied in other ways. Simi-lar to the results of the Vartanian, Martin-dale, and Kwiatkowki (2003) study, noveltyof responses was affected by the ability touse a specific strategy later in the course of



problem solving, supporting the view thatcreative people switch strategies during thecourse of a task but also suggesting that top-down processing can play an important rolein creative problem solving. Vartanian, Mar-tindale, and Kwiatkowski (2003) suggested abi-directional model of creativity in whichthe focus of attention is modulated accord-ing to top-down as well as bottom-up pro-cesses, with the use of bottom-up process-ing determined by the stage of the problem(bottom-up processing primarily during theearlier stages, and top-down processing pri-marily during the later stages). Both Var-tanian, Martindale, and Kwiatkowski (2003)and Vartanian (2009) mentioned that animportant future line of research will beto investigate the underlying mechanism(s)that enable the modulation of information-processing strategies during the course ofcreative problem solving.

Drawing more on the memory and brainliterature, Bristol and Viskontas (2006) cameto similar conclusions. They proposed thatcreative individuals are good at modulat-ing inhibitory processes, so that they haveboth the capability for cognitive control andthe capacity for disinhibition and can switchfluidly from one mode to another. In par-ticular, they argue that creative individu-als can defocus their attention at the earlystages of creative cognition so that theygrasp the whole set of potential covaria-tions; then, during the retrieval and elabo-ration stage, they can control attention sothat they can inhibit prepotent responsesand thereby allow remote associations toenter into consciousness without intrusions.Therefore, the researchers argue that cre-ative individuals are both able to overcomecognitive inhibition and are capable of sup-pressing undesired responses. They claimthat this skill requires the ability to acti-vate the dorsolateral prefrontal cortex andinhibit retrieval-related processes that mayinterfere with accessing remote associations,as well as to deactivate the dorsolateral pre-frontal cortex, depending on the contextof the task and the goals of the individual.They also left as an interesting open questiondetermining the precise brain mechanisms

that can modulate between the differentbrain activations and deactivations depend-ing on the demands of the task.

Conclusion

In his 1957 presidential address to the Amer-ican Psychological Association, Lee Cron-bach pleaded his case for uniting the bur-geoning field of cognitive psychology, withits focus on the experimental psychology ofhigher order information processing, withthe study of individual differences in Spear-man’s g. Cronbach’s call set off a greatdeal of research that would demonstratethat the newer theories regarding the natureof intelligence and the burgeoning fieldof information-processing psychology wereindeed quite compatible. The work by Hunt(Hunt, Frost, & Lunneborg, 1973) and Stern-berg (1977) helped lay the foundation for theexperimental study of intelligent reasoningprocesses that are deliberate and effortful.Subsequent research has tended to focus onboth lower level as well as higher level cor-relates of general intelligence.

One particular set of cognitive processesthat has not been investigated as thoroughlyas the others from an individual differencesperspective is the set related to the cogni-tive unconscious. This situation of mutualneglect has had the unfortunate conse-quence of limiting our picture of the natureof both human intelligence and the cognitiveunconscious, thus potentially limiting ourunderstanding of the role of individual dif-ferences in information processing in com-plex cognition more generally. The studyof individual differences in the cognitiveunconscious can increase our understandingof the nature of intelligence by helping usfind boundary conditions for so-called gen-eral intelligence (g) and by doing so, dis-covering where g breaks down. Similarly,the study of individual differences in gen-eral intelligence and its associated cognitivemechanisms can elucidate the nature of thecognitive unconscious by helping to clarifyand delineate automatic, spontaneous, andrapid information-processing mechanisms.



By charting new terrains, researchers canincrease understanding of the determinantsof intelligent behavior. A potentially fruit-ful line of research is to adapt already exist-ing experimental paradigms and constructnew tests that tap the cognitive uncon-scious. Individual differences in such tasksmay not be strongly related to psycho-metric intelligence but may still explainintelligent behavior independent of psycho-metric intelligence. Researchers can theninvestigate the precise cognitive and neuralmechanisms that underlie measures of thecognitive unconscious and develop interven-tions to raise these skills in everyone. Byfostering collaborations across the variousareas of psychology and related disciplines,and incorporating dual-process theory intoour thinking, we should be able to cometo a fuller, more complete understanding ofhuman intelligence.

Acknowledgments

I would like to thank Edward Cokely,Seymour Epstein, Jean Pretz, and RobertSternberg for their insightful suggestions onan earlier draft.

References

Achter, J. A., Lubinski, D., Benbow, C. P., &Eftekhari-Sanjani, H. (1999). Assessing voca-tional preferences among gifted adolescentsadds incremental validity to abilities: A dis-criminant analysis of educational outcomesover a 10-year interval. Journal of EducationalPsychology, 91, 777–786.

Aczel, B. (2009). Attention and awareness inhuman learning and decision making (Unpub-lished doctoral dissertation). University ofCambridge, Cambridge, UK.

Anderson, J. R. (1982). Acquisition of cognitiveskill. Psychological Review, 89, 369–406.

Anderson, M. (2005). Marrying intelligence andcognition: A developmental review. In R. J.Sternberg & J. E. Pretz (Eds.), Cognition &intelligence: Identifying the mechanisms of themind (pp. 268–288). Cambridge, UK: Cam-bridge University Press.

Anderson, M. (2008). What can autism anddyslexia tell us about intelligence? QuarterlyJournal of Experimental Psychology, 61, 116–128.

Antonakis, J., Hedlund, J., Pretz, J., & Stern-berg, R. J. (2002). Exploring the nature andacquisition of tacit knowledge for military lead-ership (Research Note 2002–04). Alexandria,VA: Army Research Institute for the Behav-ioral and Social Sciences.

Baker, C. I., Olson, C. R., & Behrmann, M.(2004). Role of attention and perceptualgrouping in visual statistical learning. Psycho-logical Science, 15, 460–466.

Barch, D. M. (2005). The cognitive neuroscienceof schizophrenia. Annual Review of ClinicalPsychology, 1, 321–353.

Bargh, J. A. (2004). The four horsemen of auto-maticity: Awareness, intention, efficiency, andcontrol in social cognition. In R. S. Wyer &T. K. Srull (Eds.), Handbook of social cognition(pp. 1–41). Hillsdale, NJ: Erlbaum.

Bargh, J. A., & Chartrand, T. L. (1999). Theunbearable automaticity of being. AmericanPsychologist, 54, 462–479.

Bargh, J. A., & Morsella, E. (2008). The uncon-scious mind. Perspectives on Psychological Sci-ence, 3, 73–79.

Baruch, I., Hemsley, D. R., & Gray, J. A. (1988a).Latent inhibition and “psychotic proneness”in normal subjects. Personality and IndividualDifferences, 9, 777–783.

Baruch, I., Hemsley, D. R., & Gray, J. A. (1988b).Differential performance of acute and chronicschizophrenics in a latent inhibition task. Jour-nal of Nervous and Mental Disease, 176, 598–606.

Baylor, A. L. (2001). A U-shaped model for thedevelopment of intuition by level of expertise.New Ideas in Psychology, 19, 237–244.

Binet, A., & Simon, T. (1916). The developmentof intelligence in children (E. S. Kite, Trans.).Baltimore, MD: Williams & Wilkens.

Bowers, K. S., Farvolden, P., & Mermigis, L.(1995). Intuitive antecedents of insight. InS. M. Smith, T. B. Ward & R. A. Finke (Eds.),The creative cognition approach (pp. 27–51).Cambridge, MA: MIT Press.

Bristol, A. S., & Viskontas, I. V. (2006). Dynamicprocesses within associative memory stores:Piecing together the neural basis of creativecognition. In J. C. Kaufman & J. Baer (Eds.),Creativity and reason in cognitive development(pp. 60–80). New York, NY: Cambridge Uni-versity Press.



Brown, J. B., Aczel, B., Jimenez, L., Kaufman,S. B., Mackintosh, N., & Plaisted, K. (2010).Intact implicit learning in autism spectrumconditions. Quarterly Journal of ExperimentalPsychology, 1, 1–24.

Carey, S., & Spelke, E. (1994). Domain-specificknowledge and conceptual change. In L. A.Hirschfeld & S. A. Gelman (Eds.), Mappingthe mind: Domain specificity in cognition andculture (pp. 169–200). New York, NY: Cam-bridge University Press.

Carroll, J. B. (1993). Human cognitive abilities:A survey of factor-analytic studies. Cambridge,UK: Cambridge University Press.

Carson, S. H., Peterson, J. B., & Higgins, D. M.(2003). Decreased latent inhibition is associ-ated with increased creative achievement inhigh-functioning individuals. Journal of Per-sonality and Social Psychology, 85, 499–506.

Chabris, C. F. (2007). Cognitive and neurobiolog-ical mechanisms of the Law of General Intel-ligence. In M. J. Roberts (Ed.), Integrating themind: Domain general vs. domain specific pro-cesses in higher cognition (pp. 449–491). NewYork, NY: Psychology Press.

Chabris, C.F., & Simons, D. (2010). The invisiblegorilla: And other ways our intuitions deceive us.New York, NY: Crown Archetype.

Chaiken, S. (1987). The heuristic model of per-suasion. In M. P. Zanna, J. M. Olson, & C. P.Herman (Eds.), Social influence: The Ontariosymposium (vol. 5, pp. 3–39). Hillsdale, NJ:Erlbaum.

Chee, M. W. L., Sriram, N., Soon, C. S., & Lee,K. M. (2000). Dorsolateral prefrontal cortexand the implicit association of concepts andattributes. Neuroreport: For Rapid Communi-cation of Neuroscience Research, 11, 135–140.

Cokely, E. T., & Kelley, C. M. (2009). Cognitiveabilities and superior decision making underrisk: A protocol analysis and process modelevaluation. Judgment and Decision Making, 4,20–33.

Cokely, E. T., Parpart, P., & Schooler, L.J. (2009).On the link between cognitive control andheuristic processes. In N. A. Taatgen & H.v. Rijn (Eds.), Proceedings of the 31st AnnualConference of the Cognitive Science Society(pp. 2926–2931). Austin, TX: Cognitive Sci-ence Society.

Cokely, K., Kelley, C. M., & Gilchrist, A. L.(2006). Sources of individual differences inworking memory: Contributions of strategyto capacity. Psychonomic Bulletin & Review, 13,991–997.

Costa, P. T., & McCrae, R. R. (1989). The NEO-PI/NEO-FFI manual supplement. Odessa, FL:Psychological Assessment Resources.

Cowan, N. (1988). Evolving conceptions ofmemory storage, selective attention, andtheir mutual constraints within the humaninformation-processing system. PsychologicalBulletin, 104, 163–191.

Cronbach, L. J. (1957). The two disciplines of sci-entific psychology. American Psychologist, 12,671–684.

Csikzentmihalyi, M. (1996). Creativity: Flow andthe psychology of discovery and invention. NewYork, NY: Harper Collins.

Cunningham, W. A., Johnson, M. K., Raye,C. L., Gatenby, J. C., Gore, J. C., & Banaji,M. R. (2004). Separable neural components inthe processing of black and white faces. Psy-chological Science, 15, 806–813.

Dennett, D. C. (1992). Consciousness explained.New York, NY: Back Bay Books.

Dijksterhuis, A., & Nordgren, L. F. (2006). Atheory of unconscious thought. Perspectives onPsychological Science, 1, 95–109.

Dorfman, J., Shames, V. A., & Kihlstrom,J. F. (1996). Intuition, incubation, and insight:Implicit cognition in problem solving. InG. D. M. Underwood (Ed.), Implicit cognition(pp. 257–296). New York, NY: Oxford Uni-versity Press.

Dorfman, L., Martindale, C., Gassimova, V., &Vartanian, O. (2008). Creativity and speed ofinformation processing: A double dissociationinvolving elementary versus inhibitory cogni-tive tasks. Personality and Individual Differ-ences, 44, 1382–1390.

Epstein, S. (1973). The self-concept revisited or atheory of a theory. American Psychologist, 28,404–416.

Epstein, S. (1991). Cognitive-experiential self-theory: An integrative theory of personality.In R. Curtis (Ed.), The relational self: Conver-gences in psychoanalysis and social psychology(pp. 111–137). New York, NY: Guilford Press.

Epstein, S. (1994). Integration of the cognitiveand the psychodynamic unconscious. Ameri-can Psychologist, 49, 709–724.

Epstein, S. (2001). Manual for the ConstructiveThinking Inventory. Odessa, FL: PsychologicalAssessment Resources.

Epstein, S. (2003). Cognitive-experiential self-theory of personality. In T. Millon & M. J.Lerner (Eds.), Comprehensive handbook of psy-chology (Vol. 5, pp. 159–184). Personality andsocial psychology. Hoboken, NJ: Wiley.



Epstein, S. (2010). Demystifying intuition: Whatit is, what it does, and how it does it. Psycho-logical Inquiry, 21, 295–312.

Epstein, S., & Meier, P. (1989). Constructivethinking: A broad coping variable with spe-cific components. Journal of Personality andSocial Psychology, 57, 332–349.

Ericsson, K. A., & Charness, N. (1994). Expertperformance: Its structure and acquisition.American Psychologist, 49, 725–747.

Ericsson, K. A., & Kintsch, W. (1995). Long-termworking memory. Psychological Review, 102,211–245.

Ericsson, K. A., & Lehmann, A. C. (1996).Expert and exceptional performance: Evi-dence of maximal adaptation to task con-straints. Annual Review of Psychology, 47, 273–305.

Evans, J. S. B. T. (2008). Dual-processingaccounts of reasoning, judgment, and socialcognition. Annual Review of Psychology, 59,255–278.

Evans, J. S. B. T., & Frankish, K. (2009). In twominds: Dual processes and beyond. New York,NY: Oxford University Press.

Eysenck, H. J. (1995). Creativity as a product ofintelligence and personality. In D. Saklofske& M. Zeidner (Eds.), International handbookof personality and intelligence: Perspectives onindividual differences (pp. 231–247). New York,NY: Plenum Press.

Feist, G. J. (2001). Natural and sexual selection inthe evolution of creativity. Bulletin of Psychol-ogy and the Arts, 2, 11–16.

Feldman, J., Kerr, B., & Streissguth, A. P.(1995). Correlational analyses of proceduraland declarative learning performance. Intelli-gence, 20, 87–114.

Finke, R. A., Ward, T. B., & Smith, S. M. (1992).Creative cognition: Theory, research, and appli-cations. Cambridge, MA: MIT Press.

Fletcher, J., Maybery, M. T., & Bennett, S. (2000).Implicit learning differences: A question ofdevelopmental level? Journal of ExperimentalPsychology: Learning, Memory, and Cognition,26, 246–252.

Fodor, J. (1983). The modularity of mind. Boston,MA: MIT Press.

Fox, M. C., Roring, R. W., & Mitchum, A.L. (2009). Reversing the speed-IQ correla-tion: Intra-individual variability and atten-tional control in the inspection time paradigm.Intelligence, 37(76–80).

Frensch, P. A., & Miner, C. S. (1995). Zur Rolledes Arbeitsgedachtnisses beim impliziten

Sequenzlernen [The role of working mem-ory in implicit sequence learning]. Zeitschriftfur Experimentelle Psychologie, 42, 545–575.

Gabora, L., & Kaufman, S. B. (2009). Evolution-ary approaches to creativity. In J. C. Kauf-man & R. J. Sternberg (Eds.), Cambridgehandbook of creativity. Cambridge, UK: Cam-bridge University Press. Manuscript in pre-paration.

Gardner, H. (1993). Frames of mind: The theory ofmultiple intelligences (2nd ed.). New York, NY:Basic Books.

Gardner, H. (1999). Intelligence reframed: Multipleintelligences for the 21st century. New York, NY:Basic Books.

Gebauer, G. F., & Mackintosh, N. J. (2007). Psy-chometric intelligence dissociates implicit andexplicit learning. Journal of Experimental Psy-chology: Learning, Memory, and Cognition, 33,34–54.

Gebauer, G. F., & Mackintosh, N. J. (2009).Implicit learning and intelligence: A princi-pal component analysis. Manuscript submit-ted for publication.

Gigerenzer, G. (2007). Gut feelings: The intel-ligence of the unconscious. New York, NY:Viking.

Gigerenzer, G., & Brighton, H. (2009). Homoheuristicus: Why biased minds make betterinferences. Topics in Cognitive Science, 1, 107–143.

Gigerenzer, G., & Regier, T. (1996). How do wetell an association from a rule? Comment onSloman (1996). Psychological Bulletin, 119, 23–26.

Gilhooly, K. J., Fiortou, E., Anthony, S. H., &Wynn, V. (2007). Divergent thinking: Strate-gies and executive involvement in generatingnovel uses for familiar objects. British Journalof Psychology, 98, 611–625.

Gladwell, M. (2007). Blink: The power of think-ing without thinking. New York, NY: Back BayBooks.

Greenwald, A. G., & Banaji, M. R. (1995).Implicit social cognition: Attitudes, self-esteem, and stereotypes. Psychological Review,102, 4–27.

Greenwald, A. G., McGhee, D. E., & Schwartz,J. L. K. (1998). Measuring individual differ-ences in implicit cognition: The Implicit Asso-ciation Test. Journal of Personality and SocialPsychology, 74, 1464–1480.

Greenwald, A. G., Poehlman, T. A., Uhlmann,E., & Banaji, M. R. (2009). Understanding and



using the Implicit Association Test: III. Meta-analysis of predictive validity. Journal of Per-sonality and Social Psychology, 97, 17–41.

Haidt, J. (2001). The emotional dog and its ratio-nal tail: A social intuitionist approach to moraljudgment. Psychological Review, 108, 814–834.

Hammond, K. R., & Hamm, R. M., Grassia, J., &Pearson, T. (1987). Direct comparison of theefficacy of intuitive and analytical cognitionin expert judgment. IEEE Transactions on Sys-tems, Man, and Cybernetics, SMC, 17, 753–770.

Hassin, R. R., Uleman, J. S., & Bargh, J. A. (2005).The new unconscious. New York, NY: OxfordUniversity Press.

Hawkins, J. (2005). On intelligence. New York,NY: Holt.

Hemphill, J. F. (2003). Interpreting the mag-nitudes of correlation coefficients. AmericanPsychologist, 58, 78–79.

Hertzog, C., & Robinson, A. E. (2005). Metacog-nition and intelligence. In O. Wilhelm & R. W.Engle (Eds.), Handbook of understanding andmeasuring intelligence (pp. 101–123). ThousandOaks, CA: Sage.

Hill, O. W. (1987–1988). Intuition: Inferentialheuristic or epistemic mode? Imagination,Cognition and Personality, 7, 137–154.

Hirschfeld, L. A., & Gelman, S. A. (1994). Map-ping the mind: Domain specificity in cognitionand culture. New York, NY: Cambridge Uni-versity Press.

Hogarth, R. M. (2005). Deciding analyticallyor trusting your intuition? The advantagesand disadvantages of analytic and intuitivethought. In T. Betsch & S. Haberstroh(Eds.), Routines of decision making (pp. 67–82).Mahwah, NJ: Erlbaum.

Howard-Jones, P. A., & Murray, S. (2003).Ideational productivity, focus of attention,and context. Creativity Research Journal, 15,153–166.

Humphreys, L. G., Lubinski, D., & Yao, G.(1993). Utility of predicting group member-ship and the role of spatial visualization inbecoming an engineer, physical scientist, orartist. Journal of Applied Psychology, 78, 250–261.

Hunt, E., Frost, N., & Lunneborg, C. (1973).Individual differences in cognition: A newapproach to intelligence. In G. H. Bower(Ed.), The psychology of learning and motiva-tion: Advances in research and theory. Oxford,UK: Academic Press.

James, W. (1890). The principles of psychology.New, York, NY: Dover.

Jensen, A. R. (1998). The g factor: The science ofmental ability. Westport, CT: Praeger.

Jiang, Y., & Chun, M. M. (2001). Selective atten-tion modulates implicit learning. QuarterlyJournal of Experimental Psychology, 54A, 1105–1124.

Jimenez, L., & Mendez, C. (1999). Which atten-tion is needed for implicit sequence learning?Journal of Experimental Psychology: Learning,Memory, and Cognition, 25, 236–259.

Johnson, M. K., & Hirst, W. (1993). MEM: Mem-ory subsystems as processes. In A. F. Collins,S. E. Gathercole, M. A. Conway, & P. E.Morris (Eds.), Theories of memory (pp. 241–286). Hillsdale, NJ: Erlbaum.

Jung, C. G. (1921/1971). Psychological types (H. G.Baynes, Trans.). Princeton, NJ: Princeton Uni-versity Press. (revised by R.F.C. Hull).

Kahneman, D., & Klein, G. (2009). Conditionsfor intuitive expertise. American Psychologist,64, 515–526.

Kahneman, D., & Frederick, S. (2002). Represen-tativeness revisited: Attribute substitution inintuitive judgment. In T. Gilovich, D. Grif-fin, & D. Kahneman (Eds.), Heuristics andbiases: The psychology of intuitive judgment(pp. 49–81). New York, NY: Cambridge Uni-versity Press.

Kane, M. J., Bleckley, M. K., Conway, A. R. A.,& Engle, R. W. (2001). A controlled-attentionview of working-memory capacity. Journalof Experimental Psychology: General, 130, 169–183.

Kaufman, S. B. (2007). Creativity. In C. R.Reynolds & E. Fletcher-Janzen (Eds.), Encyclo-pedia of special education (3rd ed.). New York,NY: Wiley.

Kaufman, S. B. (2007). Investigating the role ofdomain general mechanisms in the acquisi-tion of domain specific expertise. High AbilityStudies, 18, 71–73.

Kaufman, S. B. (2009a). Beyond general intelli-gence: The dual-process theory of human intelli-gence (Doctoral dissertation). Yale University,New Haven, CT.

Kaufman, S. B. (2009b). Faith in intuition is asso-ciated with decreased latent inhibition in asample of high-achieving adolescents. Psychol-ogy of Aesthetics, Creativity, and the Arts, 3,28–34.

Kaufman, S. B., DeYoung, C. G., Gray, J. R.,Brown, J., & Mackintosh, N. (2009). Asso-ciative learning predicts intelligence aboveand beyond working memory and processingspeed. Intelligence, 37, 374–382.



Kaufman, S. B., DeYoung, C. G., Gray, J. R.,Jimenez, L., Brown, J. B., & Mackintosh, N.(2009). Implicit learning as an ability. Cogni-tion, 116, 321–340.

Keren, G., & Schul, Y. (2009). Two is not alwaysbetter than one: A critical evaluation of two-system theories. Perspectives on PsychologicalScience, 4, 533–550.

Kihlstrom, J. F. (1987). The cognitive uncon-scious. Science, 237(4821), 1445–1452.

Kihlstrom, J. F., Shames, V. A., & Dorfman, J.(1996). Intimations of memory and thought.In L. M. Reder (Ed.), Implicit memory andmetacognition. Hillsdale, N.J.: Erlbaum.

Klaczynski, P. A. (2009). Cognitive and socialcognitive development: Dual-process researchand theory. In S. B. T. Evans & K. Frank-ish (Eds.), In two minds: Dual processes andbeyond. New York, NY: Oxford UniversityPress.

Klaczynski, P. A., & Cottrell, J. M. (2004). A dual-process approach to cognitive development:The case of children’s understanding of sunkcost decisions. Thinking & Reasoning, 10, 147–174.

Klein, G. (1999). Sources of power: How peoplemake decisions. Cambridge, UK: CambridgeUniversity Press.

Lewicki, P., & Hill, T. (1987). Unconscious pro-cesses as explanations of behavior in cognitive,personality, and social psychology. Personalityand Social Psychology Bulletin, 13, 355–362.

Lewicki, P., Hill, T., & Czyzewska, M. (1992).Nonconscious acquisition of information.American Psychologist, 47, 796–801.

Lewicki, P., Hill, T., & Sasaki, I. (1989). Self-perpetuating development of encoding biases.[Empirical Study]. Journal of ExperimentalPsychology: General, 118, 323–337.

Lieberman, M. D. (2007). The X- and C-systems:The neural basis of automatic and controlledsocial cognition. Social neuroscience: Integrat-ing biological and psychological explanations ofsocial behavior, 290–315.

Litman, L., & Reber, A. S. (2005). Implicit cog-nition and thought. In K. J. Holyoak & R.G. Morrison (Eds.), The Cambridge handbookof thinking and reasoning (pp. 431–453). NewYork, NY: Cambridge University Press.

Loftus, E. F., & Klinger, M. R. (1992). Is theunconscious smart or dumb? American Psy-chologist, 47, 761–765.

Lohman, D. F. (2001). Fluid intelligence, induc-tive reasoning, and working memory: Wherethe theory of multiple intelligences falls short.

In N. Colangelo & S. G. Assouline (Eds.),Talent development IV: Proceedings from the1998 Henry B. and Jocelyn Wallace NationalResearch Symposium on Talent Development(pp. 219–227). Scottsdale, AZ: Gifted Psychol-ogy Press.

Lubow, R. E. (1989). Latent inhibition and condi-tioned attention theory. Cambridge, UK: Cam-bridge University Press.

Lubow, R. E., & Gewirtz, J. C. (1995). Latentinhibition in humans: Data, theory, and impli-cations for schizophrenia. Psychological Bul-letin, 117, 87–103.

Lubow, R. E., Ingberg-Sachs, Y., Zalstein-Orda,N., & Gewirtz, J. C. (1992). Latent inhibitionin low and high “psychotic-prone” normal sub-jects. Personality and Individual Differences, 13,563–572.

Maas, H. L. J., Dolan, C. V., Grasman, P. P.P., Wicherts, J. M., Huizenga, H. M., & Rai-jmakers, M. E. J. (2006). A dynamical modelof general intelligence: The positive mani-fold of intelligence by mutualism. Psycholog-ical Review, 113, 842–861.

Margolis, H. (1987). Patterns, thinking, and cogni-tion: A theory of judgment. Chicago, IL: Uni-versity of Chicago Press.

Martindale, C. (1995). Creativity and connection-ism. In S. M. Smith, T. B. Ward, & R. A.Finke (Eds.), The creative cognition approach(pp. 249–268). Cambridge, MA: MIT Press.

Martindale, C. (1999). Biological bases of creativ-ity. In R. J. Sternberg (Ed.), Handbook of cre-ativity (pp. 137–152). Cambridge, UK: Cam-bridge University Press.

McGeorge, P., Crawford, J. R., & Kelly, S. W.(1997). The relationships between psychome-tric intelligence and learning in an explicit andan implicit task. Journal of Experimental Psy-chology: Learning, Memory, and Cognition, 23,239–245.

Metcalfe, J., & Mischel, W. (1999). A hot/cool-system analysis of delay of gratification:Dynamics of willpower. Psychological Review,106, 3–19.

Miller, G. A. (2003). The cognitive revolution:A historical perspective. Trends in CognitiveSciences, 7, 141–144.

Newell, B. R., Wong, K.Y., & Cheung, J. C. H.(2009). Think, blink or sleep on it? The impactof modes of thought on complex decisionmaking. Quarterly Journal of Experimental Psy-chology, 62, 707–732.

Nosek, B. A., Greenwald, A. G., & Banaji, M. R.(2007). The Implicit Association Test at age



7: A methodological and conceptual review.In J. A. Bargh (Ed.), Social psychology and theunconscious: The automaticity of higher mentalprocesses (pp. 265–292). New York, NY: Psy-chology Press.

Pacini, R., & Epstein, S. (1999). The relation ofrational and experiential information process-ing styles to personality, basic beliefs, and theratio-bias phenomenon. Journal of Personalityand Social Psychology, 76, 972–987.

Packard, M. G., Hirsh, R., & White, N. M.(1989). Differential effects of fornix and cau-date nucleus lesions on two radial maze tasks:Evidence for multiple memory systems. Jour-nal of Neuroscience, 9, 1465–1472.

Payne, J. W., Samper, A., Bettman, J. R., & Luce,M. F. (2008). Boundary conditions on uncon-scious thought in complex decision making.Psychological Science, 19, 1118–1123.

Peterson, J. B., Smith, K. W., & Carson, S.(2002). Openness and extraversion are asso-ciated with reduced latent inhibition: Repli-cation and commentary. Personality and Indi-vidual Differences, 33, 1137–1147.

Phelps, E. A., O’Connor, K. J., Cunningham, W.A., Funayama, E. S., Gatenby, J. C., Gore,J. C., et al. (2000). Performance on indirectmeasures of race evaluation predicts amygdalaactivation. Journal of Cognitive Neuroscience,12, 729–738.

Piaget, J. (1952). The origins of intelligence in chil-dren. Madison, CT: International UniversitiesPress.

Poincare, H. (1921). The foundations of science (G.B. Halstead, Trans.). New York, NY: SciencePress.

Poldrack, R. A., & Packard, M. G. (2003). Com-petition among multiple memory systems:Converging evidence from animal and humanbrain studies. Neuropsychologia, 41, 245–251.

Pretz, J. E. (2008). Intuition versus analysis: Strat-egy and experience in complex everyday prob-lem solving. Memory & Cognition, 36, 554–566.

Pretz, J. E., & Totz, K. S. (2007). Measuring indi-vidual differences in affective, heuristic, andholistic intuition. Personality and IndividualDifferences, 43, 1247–1257.

Pretz, J. E., Totz, K. S., & Kaufman, S. B. (2010).The effects of mood, cognitive style, and cog-nitive ability on implicit learning. Learningand Individual Differences, 20, 215–219.

Reber, A. S. (1989). Implicit learning and tacitknowledge. Journal of Experimental Psychol-ogy: General, 118, 219–235.

Reber, A. S. (1993). Implicit learning and tacitknowledge: An essay on the cognitive uncon-scious. New York, NY: Oxford UniversityPress.

Reber, A. S., Walkenfeld, F. F., & Hernstadt, R.(1991). Implicit and explicit learning: Individ-ual differences and IQ. Journal of ExperimentalPsychology: Learning, Memory, and Cognition,17, 888–896.

Reber, R., Ruch-Monachon, M.-A., & Perrig,W. J. (2007). Decomposing intuitive compo-nents in a conceptual problem solving task.Consciousness and Cognition: An InternationalJournal, 16, 294–309.

Reyna, V. F. (2004). How people make decisionsthat involve risk: A dual-processes approach.Current Directions in Psychological Science, 13,60–66.

Rudman, L. A., & Ashmore, R. D. (2007). Dis-crimination and the Implicit Association Test.Group Processes & Intergroup Relations, 10,359–372.

Salthouse, T. A., McGuthry, K. E., & Hambrick,D. Z. (1999). A framework for analyzing andinterpreting differential aging patterns: Appli-cation to three measures of implicit learning.Aging, Neuropsychology, and Cognition, 6, 1–18.

Schneider, W., & Shiffrin, R. M. (1977). Con-trolled and automatic human informationprocessing: I. Detection, search, and attention.Psychological Review, 84, 1–66.

Seger, C. A. (1994). Implicit learning. Psychologi-cal Bulletin, 115, 163–196.

Simon, H. A. (1979). Models of thought. NewHaven, CT: Yale University Press.

Smith, S. M., Ward, T. B., & Finke, R. A. (1995).The creative cognition approach. Cambridge,MA: MIT Press.

Spearman, C. (1904). “General intelligence,”objectively determined and measured. Ameri-can Journal of Psychology, 15, 201–293.

Stadler, M. A., & Frensch, P. A. (1997). Handbookof implicit learning. Thousand Oaks, CA: Sage.

Stanovich, K. (1999). Who is rational? Studies ofindividual differences in reasoning. Mahwah,NJ: Erlbaum

Stanovich, K. E. (2004). The robot’s rebellion: Find-ing meaning in the age of Darwin. Chicago, IL:University of Chicago Press.

Stanovich, K. E. (2009). Distinguishing the reflec-tive, algorithmic, and autonomous minds: Isit time for a tri-process theory. In J. S. B. T.Evans & K. Frankish (Eds.), In two minds: Dualprocesses and beyond. Oxford, UK: OxfordUniversity Press.



Stanovich, K. E., & West, R. F. (1997). Reasoningindependently of prior belief and individualdifferences in actively open-minded thinking.Journal of Educational Psychology, 89, 342–357.

Stanovich, K. E., & West, R. F. (2000). Individ-ual differences in reasoning: Implications forthe rationality debate? Behavioral and BrainSciences, 23, 645–726.

Stanovich, K. E., & West, R. F. (2002). Individualdifferences in reasoning: Implications for therationality debate? In T. Gilovich, D. Griffin,& D. Kahneman (Eds.), Heuristics and biases:The psychology of intuitive judgment (pp. 421–440). New York, NY: Cambridge UniversityPress.

Sternberg, R. J. (1977). Component processes inanalogical reasoning. Psychological Review, 84,353–378.

Sternberg, R. J., & Davidson, J. E. (1995). Thenature of insight. Cambridge, MA: MIT Press.

Sternberg, R. J., & Pretz, J. E. (2005). Cognition& intelligence: Identifying the mechanisms of themind. New York, NY: Cambridge UniversityPress.

Thorsteinson, T. J., & Withrow, S. (2009). Doesunconscious thought outperform consciousthought on complex decisions? A furtherexamination. Judgment and Decision Making,4, 235–247.

Turke-Browne, N. B., Junge, J. A., & Scholl, B.J. (2005). The automaticity of visual statisticallearning. Journal of Experimental Psychology:General, 134, 552–564.

Unsworth, N., & Engle, R. W. (2005). Individ-ual differences in working memory capacityand learning: Evidence from the serial reactiontime task. Memory & Cognition, 33, 213–220.

van Ravenzwaaij, D., van der Maas, H.L.,& Wagenmakers, E.J. (in press). Does thename-race implicit association test measureracial prejudice? Experimental Psychology, doi:10.1027/1618-3169/a000093.

Vartanian, O. (2009). Variable attention facili-tates creative problem Psychology of Aesthetics,Creativity, and the Arts, 3, 57–59.

Vartanian, O., Martindale, C., & Kwiatkowski,J. (2003). Creativity and inductive reasoning:The relationship between divergent think-ing and performance on Wason’s 2–4–6 task.Quarterly Journal of Experimental Psychology,34, 1370–1380.

Vartanian, O., Martindale, C., & Kwiatkowski,J. (2007). Creative potential, attention, andspeed of information processing. Personalityand Individual Differences, 43, 1470–1480.

Vartanian, O., Martindale, C., & Matthews, J.(2009). Divergent thinking ability is relatedto faster relatedness judgments. Psychologyof Aesthetics, Creativity, and the Arts, 3, 99–103.

Wallas (1926). The art of thought. New York, NY:Harcourt Brace.

Ward, T. B., Smith, S. M., & Finke, R. A.(1999). Creative cognition. In R. J. Stern-berg (Ed.), Handbook of creativity (pp. 189–212). New York, NY: Cambridge UniversityPress.

Wilson, T. (2004). Strangers to ourselves: Discover-ing the adaptive unconscious. Boston, MA: Har-vard University Press.

Wilson, T. D., Lindsey, S., & Schooler, T. Y.(2000). A model of dual attitudes. Psychologi-cal Review, 107, 101–126.

Wilson, T. D., & Schooler, J. W. (1991). Think-ing too much: Introspection can reduce thequality of preferences and decisions. Journalof Personality and Social Psychology, 60, 181–192.

Woolhouse, L. S., & Bayne, R. (2000). Person-ality and the use of intuition: Individual dif-ferences in strategy and performance on animplicit learning task. European Journal of Per-sonality, 14, 157–169.

Zajonc, R. B. (1980). Feeling and thinking: Pref-erences need no inferences. American Psychol-ogist, 35, 151–175.

Zhiyan, T., & Singer, J. L. (1997). Daydreamingstyles, emotionality and the Big Five person-ality dimensions. Imagination, Cognition andPersonality, 16, 399–414.

intelligence and the cognitive...

Documents