Quick viewing(Text Mode)

Consciousness, Metacognition, and the Unconscious

Consciousness, Metacognition, and the Unconscious

41 , Metacognition, and the Unconscious

Piotr Winkielman & Jonathan W. Schooler

What are we aware of? What do we know about the growing interest in psychology in questions of our own and feelings? These questions consciousness and metacognition. We then review about consciousness and metacognition lie at the selected findings on consciousness and metacog- heart of social – a field that relies on nition, using an important distinction between and methods of , mental states that are simply conscious and mental cognitive science, and neuroscience to understand states that are accompanied by rich metacognitive how people think about others and themselves. representations of those states. We then move to a Importantly, these questions need answers not discussion of cognition and emotion in the absence only because they are scientifically fascinating but of consciousness. Throughout, we address neural also because it is practically important to examine correlates of consciousness and metacognition how in daily life people manage to (sometimes) and touch on questions concerning the relation- answer questions like: “What are you thinking ship between and self-cognition. about right now?”, “How do you feel right now?”, “How much do you want to drink?”, “Do you want to smoke?”, “Do you dislike this social group?”, “Do you find this person attractive?”, “Did you enjoy this piece of music?”, “Did you HISTORICAL PERSPECTIVE notice this object?”, “Do you understand this pas- sage of text?”, and “Were you influenced by this Just like the actual phenomena, the history of message?” These standard questions require the thinking about consciousness and metacognition ability of the person to access, evaluate, and undergoes a pattern of waxing and waning. express the cognitive and affective contents of her Descartes famously argued that consciousness is own mind. And though these answers may often the only undeniable reality. Everything else could come to our mind effortlessly, giving the illusion be created by a malicious demon, but our very of “self-transparency,” the process of reaching own thoughts and feelings are necessarily true. them is actually tricky and may be fraught with For Descartes, consciousness also comes with error. After all, not all mental states are conscious. metacognitive transparency, making our own Furthermore, those states that are conscious can mental states self-apparent. It is worth noting this be so in various forms and to different degrees. because the Cartesian view still underlies much of Finally, translating our consciousness into a report everyday intuitions about the mind and, as we may introduce a variety of distortions. describe later, is being challenged by modern Our chapter reviews theorizing and empirical social-cognitive research that argues that even research on consciousness and metacognition, one’s own thoughts may be subject to mis- using the following organizational structure. We representation, in the form of temporal and trans- start with a few historical remarks, highlighting lational dissociations, or may even be completely

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5544 22/13/2012/13/2012 9:54:239:54:23 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 55

obscured, as in the case of unconscious goals and biological vocabulary. Elements of this view are unconscious affect. also visible in modern social cognition when Following some 19th-century philosophers researchers argue that as psychological science (e.g., Husserl, Brentano), early psychologists had identifies more and more simple, low-level, and focused on the nature of intentionality and the thus unconscious mechanisms it will no longer structure of experience (Boring, 1953). Their need complex, high-level concepts, like “con- method – − assumed that the critical sciousness” (e.g., Bargh, 2007). This view is constituents of mental life (thoughts, feelings, related to a position in philosophy called “elimi- volitions) are in principle consciously accessible. native materialism,” which proposes replacing They also argued psychology should be funda- high-level mental concepts with references to low- mentally interested in content-bearing intentional level biological substrates (Churchland, 1981). states (thoughts, intentions, goals), but not so For example, instead of talking about “seeing,” much in non-conscious events (e.g., associative science should talk about specific visual computa- chains, reflexes, physiology), which are best left to tions in the occipital lobe. However, note that this biologists. As we discuss in detail later, these foun- reductionist argument is logically problematic. dational notions are being challenged by modern For one, it confuses the identification of low-level researchers in social cognition who argue that mechanisms with a satisfying explanation of states, such as goals, can operate unconsciously. high-level concepts (Fodor, 1968). But, low-level From the perspective of modern social cogni- explanations often offer an uninteresting or even tion, another interesting historical aspect was obscuring reduction. For example, providing a the early interest in social foundations of self- complete low-level explanation of physical mate- . This was particularly pronounced in rials that money is made out of (paper, metal) is the writing of symbolic interactionists (e.g., not interesting, and does not in any way make the Mead, 1934) as well as some developmental explanations of money in terms of its functions psychologists (e.g., Vygotsky, 1962). For them, superfluous. In fact, it is counter-productive as it self-consciousness is partly created by the neces- occludes relevant properties (e.g., that a $1 coin sity to function in a social community. After all, functions exactly the same as $1 paper bill). social coordination requires the individual to be Furthermore, the argument that consciousness can able to adjust to others, see things from their per- be reduced to smaller, mechanistic, unconscious spective, anticipate their reactions, self-regulate, parts is also problematic because consciousness is and, critically, learn to adapt towards oneself the an emergent phenomenon that requires an interac- stance of a “generalized other.” As we discuss tion of many (possibly dumb) parts, and thus by later, these historical themes are reflected in definition not reducible to any single one of them. contemporary social-cognitive research on self- As an analogy, it is impossible to understand monitoring, self-regulation, and self-awareness. the of “driving,” or how a car works by Though consciousness was of central impor- considering its parts separately. tance for the early psychologists, still they admit- In the general experimental psychology, the ted some role for unconscious processes. For anti-consciousness behaviorist stance started to example, Helmholtz proposed that vision involves relax in the mid-1970s when researchers began unconscious inferences, whereas James wrote to tackle issues like controlled and automatic about habits and the subconscious (Kihlstrom, processing (Shiffrin & Schneider, 1977), atten- 2007). Behaviorism, and the ambition to make tional selection (Kahneman, 1973), and uncon- psychology “objective” and equal to other natural scious (Marcel, 1983), leading to sciences, brought disfavor to the methods of recognition that consciousness is “respectable, introspection and to mentalistic concepts like useful, and probably necessary” (Mandler, 1975). consciousness. Along with this came the belief Now, the legitimacy of the topic is fully reestab- that behavior is ultimately under the control of lished and “everyone who is conscious, is study- the environment, rather than mysterious “internal ing consciousness” (Churchland, 2005). forces.” This behaviorist credo still guides Social cognition researchers have always been some modern researchers in social cognition who in the game of understanding “consciousness,” tend to highlight how much of social behavior is though, often with the goal of demonstrating its under “environmental control,” at least in the limits. Thus, a pioneering study showed that a sense of “power of the situation” (e.g., Bargh, stimulus’ value can be enhanced via its uncon- 2007). Behaviorists also assumed that providing scious mere exposure (Kunst-Wilson & Zajonc, mechanistic explanations of behavior will make 1980). Another pioneering study showed that peo- concepts like “consciousness” and “volition” dis- ple’s conscious beliefs about the causes of their appear from the psychological vocabulary, the own behavior can be at odds with actual causes way concepts like “phlogiston’ or “life force” (Nisbett & Wilson, 1977). These early studies disappeared from modern scientific physical and opened the way to a wealth of demonstrations of

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5555 22/13/2012/13/2012 9:54:239:54:23 AMAM 56 THE SAGE HANDBOOK OF SOCIAL COGNITION

unconscious influences on social perception, emotion, ), and how they can improve affect, reasoning, decision, and behavior (e.g., them to make their cognitive functioning more Bargh, 1989; Wegner, 2002; Wilson, 2002). Some efficient. Accordingly, research on the develop- researchers became so excited about these find- mental and educational aspect of metacognition ings that they began to argue that “most of a has a long history (Flavell, 1979). This history person’s everyday life is determined not by their continues to have resonance today. For example, conscious intentions and deliberate choices but by there has recently been some enthusiasm about mental processes that are put into motion by fea- a Vygotsky-inspired curriculum (“tools-of-the tures of the environment and that operate outside mind”) which, among other things, teaches chil- of conscious awareness and guidance” (Bargh dren self-knowledge and self-regulation skills in & Chartrand, 1999, p. 462). Along with all the a social setting (Diamond et al., 2007). excitement about the unconscious in social cogni- Somewhat trickier problems arise when - tion came fascination with all things “implicit” – cognition invokes concepts such as “a sense of perception, , attitudes, self-esteem, uncertainty,” “feeling-of-knowing,” “tip-of-the- self-concepts, stereotypes, partisanship, goals, etc. tongue,” or “sense of agency” (Koriat, 2006; (Greenwald et al., 2002). Perhaps capturing all Nelson, 1996). These require stronger commit- this, it is telling that the latest Handbook of Social ment to a view of metacognition as a particular Psychology has a chapter on “Automaticity and kind of inward-directed, self-reflective, metacon- the unconscious” but no chapter on consciousness scious capacity. This capacity is a necessarily (Fiske, Gilbert, & Lindzey, 2010). complex skill, as it involves re-representing one’s However, there have also been attempts to own current thoughts (i.e., in addition to thoughts, highlight the limits of the . In having thoughts about those thoughts). Also, this fact, some cognitive psychologists seem to relish capacity is necessarily introspective, implying their role as an empirical and theoretical “police” rich subjectivity, with the ability to discern subtle on more spectacular abilities attributed to the internal states (e.g., a feeling that one knows the unconscious in social cognition (Kihlstrom, 2008; answer, without being able to think of the answer Merikle & Reingold, 1998; Shanks, 2005, 2006). right now). These mentalistic and introspective Some critiques of the scientific as well as public elements of metacognition make some researchers misconceptions about the unconscious also came uncomfortable. But explanations of performance from within social psychology (e.g., Greenwald, on tests of “metacognitive abilities” that do not 1992). Interestingly, there has been fairly constant postulate any kind of access to introspective interest within social psychology in processes under- “private” information have trouble explaining lying conscious self-control (Metcalfe & Mischel, some basic data (e.g., why first-person observers 1999) and in the questions of self-regulation are so much better in making metacognitive judg- and, more recently, “free will” (Baumeister, 2008; ments than third-party observers; Jameson et al., Schooler, 2010). Recently, social psychologists 1993). Furthermore, as we will elaborate below, have also become more interested in the question it is hard to non-introspectively explain the main of the social origins of consciousness – an inter- metacognitive issues that we will discuss here, esting return to ideas of Mead and Vygotski (e.g., which have to do with a person waxing and waning Baumeister & Masicampo, 2010). out of “metacognitive” awareness as a function of her to the direction of her own thoughts (Schooler, 2002). As a result, the metacognitive perspective is increasingly gaining in popularity, A few historic remarks including in social cognition. Since about the mid- on metacognition 1990s there have been many conferences, books, and papers dedicated to social metacognition and The history of ideas about metacognition is a bit several comprehensive reviews are available (Jost, less oscillatory. Some of this is because the basic Kruglanski, & Nelson, 1998; Petty, Briñol, concept of metacognition carries less philosophi- Tormala, & Wegener, 2007; Schwarz, 2004). cal baggage. After all, few doubt that people have some capacity to think about their own mental states, skills, and capacities. Consequently, there is little controversy in the science of metacogni- Definition and distinctions tion, understood as an effort to uncover the rela- tion between people’s mental states and their But what exactly is “consciousness” – with beliefs about those mental states (Koriat, 2006). and without the “meta”? Some writers on Specifically, since the time of Piaget, psycholo- consciousness devote pages to the elucidation of gists have wanted to know how much children, different meanings (e.g., Zeeman, 2002), but we and adults, know about their own cognitive will highlight only two different senses in which processes and skills (comprehension, , the term appears in psychological literature.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5566 22/13/2012/13/2012 9:54:239:54:23 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 57

This will also allow us to briefly comment on in subjective experience – it “feels like some- some current debates in social cognition. thing” to be in a conscious state of, say, seeing red (as opposed to just unconsciously reacting to red). Second, being conscious of x means having x Conscious as “awake and mindful” potentially available to report and to use in inten- The word “conscious” can refer to a global state of tional control of behavior. It is in that second an individual. One use of this word is similar to sense that psychologists are interested in whether “awake” or “vigilant,” as opposed to “asleep” or there are unconscious , , “comatose.” The sleepy−vigilant dimension is goals, attitudes, or emotions (Bargh, 1989; typically investigated by neurologists, though Greenwald, 1992; Winkielman & Berridge, 2004). some interesting social cognition studies have shown that anesthetized patients form implicit, but not explicit, memory for events during surgery Metaconsciousness (Kihlstrom et al., 1990). A more relevant meaning The interest in what makes certain mental content of “conscious” as a description of a global state “conscious,” and what makes it available for refers to a “mindful,” as opposed to a “robot-like,” report and control, binds together the research on dimension. In that sense, being conscious is the consciousness with research on metacognition. As ability to have subjective experiences, wishes, mentioned earlier, much of the work under the desires, and complex thoughts, and to perform heading of metacognition simply asks about a flexible, self-initiated, purposeful behaviors. For person’s beliefs about her own comprehension, example, patients in a persistent vegetative state memory, intelligence, etc. However, other (PVS) maintain regular sleep−wake cycles, researchers, including one of us in particular respond to simple stimulation (e.g., withdraw their (J.W.S.), are interested in metacognition as it hand from sharp objects), yet are not considered reveals the ability (and failures) of conscious and possessive of “personhood” by to represent itself. Specifically, about a decade medical experts. This is because PVS patients are ago, Schooler (2001, 2002, Schooler & Schreiber, unable to make choices, process complex infor- 2004) proposed a distinction between conscious mation, show flexible behavior, and initiate purpo- thoughts that occur without the additional element sive actions (Laureys et al., 2002). Interesingly, of explicit self-reflection, and thoughts that are some recent work that investigated everyday crite- accompanied by the explicit representation of ria for attribution of a “mind” found that college having a thought. That is, mental content could be students basically use two dimensions in their “experientially conscious,” a constituent of ongo- decisions – capacity for agency and capacity for ing experience, without being explicitly reflected experience – and that these dimension predict how upon. One example of this is the experience of much the “organism” is valued (Gray, Gray, & mind-wandering while where people can Wegner, 2007). temporarily fail to notice that their eyes are It is interesting to contrast the above medical, moving across the page but their mind is com- legal, and everyday view that conscious experi- pletely elsewhere. But, mental content can also be ence and purposeful action is essential for “per- “metaconscious” (or “meta-aware”) and explicitly sonhood” with the dominant view in social represented as a content of one’s own conscious- cognition that minimizes the role of consciousness ness. As for example, when one suddenly realizes in complex thought, choice, and purposive behav- that one has been mind-wandering instead of ior. We actually doubt that, despite some radical paying attention to what was being read. It is this “anti-consciousness” declarations in articles and type of consciousness that is typically assessed chapters, modern social cognition researchers when an experimenter asks participants questions privately believe there is little distinction between like, “What are you thinking about now?”, “What people and complex robots. We also noticed that goal are you currently pursuing?” or “How happy in informal discussion those researchers readily do you feel right now?” An example of metacon- concede that only some thoughts, attitudes, goals scious affect would be feeling happy and at the or decisions are unconscious, and that only some- same time having an articulated thought: “I am times a sense of voluntary control is illusory. We happy now” (Schooler, Ariely, & Loewenstein, will return to this issue later. 2003; Schooler & Mauss, 2010). The above distinction between consciousness and metaconsciousness is allied with some Conscious as “subjectively experienced” and terms that have been introduced over the years. “available for report and intentional use” For example, it roughly corresponds to the distinc- A major cluster of meanings for “conscious” cent- tion between “first-order” and “second-order” ers on the subjective status of a particular mental consciousness. For example, Lambie and Marcel content (perception, thought, or feeling). First, (2002) argued that individuals with alexithymia being conscious of x means having x represented have a first-order experience of emotions but lack

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5577 22/13/2012/13/2012 9:54:239:54:23 AMAM 58 THE SAGE HANDBOOK OF SOCIAL COGNITION

a second-order awareness of the fact that they that consciousness functions to allow communica- are experiencing the emotions. However, it is tion and coordination between the many isolated, also worth noting also that others use the term parallel sub-processors in the human mind “second-order consciousness” in a manner that (Baars, 1988). Consciousness constitutes a “global does not directly map on to the notion of metacon- workspace” where various local processors can sciousness as used by Schooler and colleagues. “broadcast” their outputs and talk to each other in For example, philosopher Ned Block (1995) a common internal code. As a result, the previ- introduced a distinction between phenomenal ously independent and isolated local processes consciousness (first order) and access conscious- can coordinate, sequence, and structure their ness (second order), which resembles a distinction actions, thus helping the organism achieving its between “perceptual vs conceptual” content. This goals. For example, by representing tactile, visual, diverges from the present view, in which both and auditory processes in a common matrix, the perceptual and conceptual content can become putative global workspace allows for the identifi- metaconscious. Similarly, a popular distinction in cation of novel cross-modal and cross-temporal social cognition concerns “construal levels” (see connections (e.g., “The sequence of musical notes Chapter 12), but again we propose that informa- I just heard has the same pattern as the sequence tion from any construal level can become of colored lights I saw before.”). Of course, like “metaconscious.” Finally, Rosenthal (1986) has almost anything else, cross-modal integration can emphasized a philosophical distinction between be automatized into unconscious, suggesting that first-order and higher-order mental states. In his access in the global workspace might be only view, consciousness (of any kind) only occurs needed to initially connect novel sensations and when a mental representation is accompanied by a responses. More importantly, the global accessi- higher-order state explicitly articulating the con- bility of conscious representations makes them tent of a first-order thought. This is different from available for verbal report and for high-level proc- our view that first-order mental states can be con- esses such as conscious judgment, reasoning, and scious, even when they are not accompanied by the planning and guiding of action. explicit knowledge of their occurrence. For fur- But what gives representations conscious or ther discussion of these issues, see Winkielman “global” access? Cognitive researchers often and Schooler (2011). emphasize the role of representation “strength” (Cleermans, 2005). The notion of “strength” cap- tures the idea that representations require a certain stability and quality before they can enter working Functions of consciousness memory, where they can be actively maintained, and metaconsciousness and become accessible for potential report. One determinant of “strength” is activation, which A central assumption in social cognition is that in turn is determined by many factors, such as mental information is represented on several stimulus energy (longer presented items are more levels. Accordingly, much research attention likely to become conscious than briefly presented focuses on understanding how these different items) and recency (more recent items are more levels relate to each other (Smith & DeCoster, likely to become conscious than older items), and 2000; Strack & Deutsch, 2004). Here, we ask so forth. Representational strength is also influ- what distinguishes unconscious, conscious, and enced by focused attention – a perceptual ampli- metaconscious representations. This question fier and selector of events (conscious and touches on a more general problem of the purpose non-conscious) that fall within its scope. Thus, an of consciousness – a problem that has received a objectively very weak stimulus can reach con- variety of functional and mechanistic answers in sciousness, if it receives attentional processing the psychological literature. In general, research- and there is little perceptual competition ers have emphasized the idea that consciousness (Breitmeyer & Ogmen, 2006). Interestingly, recent is associated with (i) special access to mental research shows, somewhat paradoxically, that content, and (ii) special functions that can be focused conscious attention may be necessary for performed on this content. some unconscious processes (Koch & Tsuchiya, 2007). For example, subliminal priming is enhanced by attentional cuing of location (Sumner Conscious access et al., 2006) and limbic responses are stronger if Several theories posit that consciousness is a brief affective stimuli fall in the scope of focused representational system characterized by special attention (Pessoa et al., 2002). These observations access to mental content. One useful framework may explain why so many successful subliminal is the Global Workspace Theory, which proposes priming paradigms in social cognition require

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5588 22/13/2012/13/2012 9:54:239:54:23 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 59

that the subject is paying attention to a specific Conscious thinking area on the screen (even if the prime remains Some argue that consciousness enables higher- invisible). It may also explain why many social- order, meaning-based, logically well-structured cognitive studies on unconscious processes use processing of information (Block, 1995; Searle, “unobtrusive” rather than subliminal priming. In 1997). In contrast, the unconscious is restricted those studies, participants are exposed to stimuli to simpler, associative forms of processing. This in a definitely conscious, attended, and prolonged distinction resembles, but does not completely fashion (e.g., as a part of a sentence-unscrambling overlap with, “dual process” theories in social task or a crossword puzzle), with the “uncon- cognition. For example, Strack and Deutsch (2004) scious” element being the relevance of the task suggest that social cognition is carried out by two to subsequent judgment or the importance of a systems: a reflective system that relies on knowl- particular stimulus dimension. In short, focused edge about facts and values, and an impulsive attention might be a precondition for many system based on associative links and motivational unconscious effects. orientations. The differential information base Another factor that modulates whether or not upon which the two systems rely determines the mental content is conscious has to do with ana- types of responses they engender. The reflective tomical and functional disconnection. Thus, a system, drawing on propositions about the world, visual representation in blindsight patients can be leads to responses based on rational considera- strong (e.g., it can drive pointing behavior), but tions. In contrast, the impulsive system, drawing remains unconscious because it is restricted on associations and impulses, leads to non- to lower visual pathways (Weiskrantz, 1986). reasoned actions. Similarly, habits (e.g., biking) may involve repre- Does processing of meaning require conscious- sentations that are very robust, but unconscious, ness? This question is a subject of long debate, because they are only instantiated in the motor which touches on tricky issues of the relation system (Cleermans, 2005). A functional discon- between semantic cognition and associationism nection may occur when input fails to cohere with (McClelland & Rogers, 2003). It is now widely currently processed information. For example, a accepted that subliminally presented pictures and distinct, prolonged, unusual, and dynamic event words can activate related semantic and affective (e.g., a gorilla slowly walking through a room of categories (Greenwald, Draine, & Abrams, 1996; people passing balls to each other) can remain Marcel, 1983). Even subliminally presented single unconscious, when participants “look” at the digits can activate magnitude information scene and closely attend to its more familiar (Dehaene et al., 2006). Thus, there is no doubt that features (Simons & Chabris, 1999). One explana- complex content can be unconsciously activated tion of this “blindness” is that the event is incom- across meaning dimension. However, the evidence patible with the current mental model (i.e., for unconscious semantic processing, rather than generalized schema) of the situation or with par- automatic activation, is sparse. For example, ticipants’ current perceptual goals. unconscious priming responds to partial- rather Finally, there is some exciting neuroscientific than whole-word information (e.g., fragments of research on the mechanisms of conscious access. affectively negative words “smut” and “bile” Some evidence suggests that consciousness repre- prime the affectively positive word “smile”), is sents a form of multiregional activation, which is not sensitive to basic operations like negations perhaps integrated by oscillatory activity (Tononi, (“not,” “un-,” or “dis-”), and cannot process two- 2004). For example, conscious perception of a digit numbers (Abrahms & Greenwald, 2000). stimulus is associated with synchronous activation One may wonder whether these limitations arise of higher associative cortices, particularly pari- because subliminal presentations afford very weak etal, prefrontal, and anterior cingulate areas, stimulus input. However, similar results hold whereas unconscious perception is associated when the input is conscious and only conscious only with a local activation (Dehaene et al., 2006). processing capacity is reduced. Thus, processing Synchronized oscillations may serve as a mecha- relational information such as negation (“no dis- nism for binding information in the “global work- ease”) or causality (“smoke causes fire”) requires space” discussed earlier. Consistent with these conscious capacity, whereas processing informa- ideas, clinical work has shown that the previously tion about association does not (Deutsch, mentioned patients in a PVS (awake but uncon- Gawronski, & Strack, 2006; Hummel & Holyoak, scious) show only localized, modality-specific 2003). In a straightforward but telling demonstra- responses to stimuli, whereas patients in a tion of this point, DeWall et al. (2008) presented minimally conscious state show coherent responses participants with a standard set of graduate record across multiple sensory and associative systems examination (GRE) analytical problems, and (Laureys et al., 2002). asked them to solve these problems under typical

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 5599 22/13/2012/13/2012 9:54:239:54:23 AMAM 60 THE SAGE HANDBOOK OF SOCIAL COGNITION

conditions or under . Not surpris- effective in that particular case. In fact, when the ingly, loaded participants did much worse. “best” solution to a problem is to simply count the Note that even if some complex cognitive skills number of positive (+) vs negative (−) attributes, can ultimately be automatized, it does not follow engaging in deeper processing that focuses on the that the initial acquisition and mastery of these attribute meaning (cup holder vs safety) might skills is possible without consciousness. For lead to a suboptimal decision making (Gigerenzer example, most adults can do basic multiplication & Goldstein, 1996). Accordingly, the benefits of table automatically, via associative (2 × 2 “unconscious thought” may only apply to is 4). However, no one believes that the uncon- so-called “linear integration problems,” where the scious actually does multiplication. It is generally attribute content (cup holder vs safety) either does thought that highly trained operations become not matter, or can be consciously translated into automatic over time and can eventually be per- attribute weights before unconscious “thought.” formed by “dumb” associative retrieval (Logan, Also, note that it is strange that giving people 1998; Rickard, 2005; Smith & DeCoster, 2000). 4 minutes to think consciously about a simple As an example, solving a novel mathematical choice produces such a low response – only 20% problem, like 78 × 56, is not helped by “thinking correct in the conscious thinking group! After all, about it unconsciously,” under cognitive load, or it is not that complicated to figure out that a car by “sleeping on it.” with 12 (75%) positive attributes is better than a All this seems to suggest that the unconscious car with 4 (25%) positive attributes. This suggests is rather “dumb” (Loftus & Klinger, 1992). Yet, that the problem encountered by “conscious think- this image has recently been challenged by claims ers” may simply lie in greater confusion about the that complex decision are often better made by original attributes and their assignments to cars − “unconscious thought” (Dijksterhuis et al., 2006). perhaps because recall is susceptible to primacy or For example, in one study, participants were pre- recency effects and interference by the intermedi- sented with 12 attributes for each of four different ate task (Shanks, 2006). It may also suggest that cars (e.g., “car A has a cup holder,” “car B is conscious thinkers have no problem at all. After safe”). That is, participants learned about 48 all, there are some normative issues with this task, attributes total. One car had 75% positive attributes, as it is not clear whether a response that simply two had 50% positive attributes, and one had 25% maximizes + vs − is optimal for most subjects in positive attributes. Note that the presentation of these types of task (e.g., Why is it best to treat cup car−attribute pairs was all mixed, which makes it holders and safety as equivalent attributes?). In rather hard to keep track of all the information. short, while recent evidence does suggest some Participants were then asked to choose the best limitations to extensive deliberation, the degree to car. One group of participants (termed “conscious which this research implicates truly intelligent thinkers”) made their decision after 4 minutes of unconscious processing remains to be determined. deliberation, and another group (termed “uncon- Some doubts about the power of “unconscious scious thinkers”) after 4 minutes of engaging in a thought” are also raised by the fact that the finding distracting anagram-solving task. Interestingly, of superiority of the decisions in the distraction the results showed that 60% of “unconscious conditions, over control conditions, is empirically thinkers” chose the ostensibly “best” car (i.e., the fragile (Acker, 2008). Perhaps reflecting these one with the greatest number of positive attributes), critiques, more recent writings tone down the while only 20% of the “conscious thinkers” did radical earlier claims about the general superiority so. For the authors, these results show that the of unconscious processes for decision making unconscious thinking not only facilitates decisions (Bongers & Dijksterhuis, 2009). but also is actually better than conscious thinking. To quote: “It should benefit the individual to think consciously about simple matters and to delegate Conscious control thinking about more complex matters to the Consciousness is associated not only with special unconscious” (Dijksterhuis et al., 2006, p. 1007). access to mental content but also with special However, other interpretations are possible. operations that can be performed on this content. First, it is not clear why distraction by anagrams Several of these operations fall under the umbrella eliminates conscious thought, rather than reducing name “control,” thus linking consciousness to its amount. If so, perhaps distraction is advanta- what cognitive scientists call “executive func- geous because it helps to prevent overthinking tions” (Norman & Shallice, 1986). One aspect of (cf. Rey, Goldstein, & Perruchet, 2009). This is control is selection. Thus, conscious content can not unlike other “verbal overshadowing” effects, be preferentially attended to and maintained in to which we return later (Wilson & Schooler, working memory or discarded if not needed. 1991). Also, reduction in capacity may encourage Another aspect of control is intentionality. Action a reliance on simple heuristics that are more can be deliberately started and stopped, or can be

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6600 22/13/2012/13/2012 9:54:239:54:23 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 61

delayed until appropriate conditions materialize. different than in research on . Scheduling conflicts can be resolved, and new First, the “unconscious goal” paradigms rely on hierarchies can be established. Finally, with con- an unobtrusive activation of preformulated, stand- trol comes flexibility. Thus, extant mental content ard goals, rather than on the formulation of novel can be redeployed in adaptive, non-routine ways, goals. Second, those goals do not require partici- and old response chains can be broken up and pants to overcome stronger alternative behavior rearranged. This simple point was recently ele- (e.g., go against a prepotent tendency), but operate gantly demonstrated in a study where participants in situations where behavioral choices are already had to come up with novel titles, musical improv- predetermined (participants can either cooperate isations, or interesting drawings. Not surprisingly, or compete, with the likelihood of either action participants under cognitive load produced repeti- relatively equal). Third, the outcomes are fairly tive, inflexible, and uninspiring works (Baumeister, unimportant and do not require participants to Schmeichel, DeWall, & Vohs, 2007). reflect on the meaning or consequences of their One interesting aspect of conscious control is actions. Accordingly, we suspect that many effects its restricted capacity. Thus, only few elements attributed to unconscious goals simply reflect the can be consciously manipulated at any one time, influence of primes on the interpretation of a operations must be performed in a serial rather vague experimental situation, including giving than parallel fashion, and processing is subject to participants an idea of what and how much they severe bottlenecks (Pashler, 1998). In fact, it is are supposed to do (see Förster, Lieberman, & hard to be overwhelmed by the power of the Friedman, 2007 for discussion). Furthermore, in unconscious given how many accidents are caused some “unconscious goal” paradigms participants by ill-advised attempts to multitask (e.g., talking could also be actually conscious of the goal, but on the cell phone while driving; Levy, Pashler, & just confused about its source. As we discuss later, Boer, 2006). there are also multiple ways in which goals could Of course, not all forms of control are con- be conscious, but not verbally reported because of scious. The world is filled with very complex temporal and translation dissociations. mechanical devices that automatically monitor the Finally, while unconscious goal activation conditions of subordinate processes and adjust clearly operates under some conditions, uncon- their operation (Shinskey, 1979). The human body scious goals have yet to be shown to possess has many systems of complex control loops (e.g., anything approaching the potency or flexibility of homeostatic temperature and blood sugar mecha- conscious goals. To illustrate this, imagine a fol- nisms). Furthermore, several “mental” processes lowing experiment. Participants are in a room automatically adjust their operation based on con- with both food and drink. Participants are first told textual conditions (Carver & Scheier, 1990). Thus, once, consciously, that their goal is to eat. Next, people unconsciously regulate eye movements to participants are given an unconscious priming facilitate text processing (Reichle, Pollatsek, procedure with multiple words related to the goal Fisher, & Rayner, 1998) and unconsciously adjust of drinking. We predict that very few, if any par- hand movements to capture the desired object ticipants, would behave in accordance with the (Triesch et al., 2003). Finally, people are typically more recent, but unconscious “goal” to drink. unaware of several layers of control required for Indeed, in addition to offering a skewed per- coherent speaking and writing. spective on the role of consciousness in mediating Reflecting the preceding, and other considera- behaviors, the present trend towards attributing tions, Bargh (1989) pointed out that it should be the bulk of human action to unconscious mecha- possible to dissociate consciousness and control in nisms may have undesirable effects on people’s social cognition. Subsequently, various forms of self-regulatory ability. Evidence supporting this automatic control have been proposed (Fitzsimons concern comes from Vohs and Schooler (2008), & Bargh, 2004). One case is the pursuit of who asked some participants to read an excerpt “unconscious goals” (see also Chapter 5). Evidence from Francis Crick’s The Astonishing Hypothesis comes from studies where individuals who are that articulates the view that conscious control primed subliminally or unobtrusively with goal- is an epiphenomenon; i.e., that people lack any related words (e.g., “cooperate,” “achieve,” “mem- meaningful sort of free will. Compared to con- orize”) show corresponding adjustment in their trols, participants exposed to the message that behavior (e.g., show more helpful behavior, solve conscious control is illusory behaved more immor- more problems, or remember more details). ally on a passive cheating task. Moreover, their Furthermore, these adjustments appear to be sen- increased cheating was mediated by decreased sitive to conditions under which the goal is appro- belief in free will. In a second experiment, priate and to track success at goal pursuit. These exposure to deterministic statements led partici- findings are interesting, but note that the type pants to overpay themselves on a , of “control” explored in these studies is rather relative to participants who were exposed to

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6611 22/13/2012/13/2012 9:54:239:54:23 AMAM 62 THE SAGE HANDBOOK OF SOCIAL COGNITION

statements endorsing free will. Of course, such suggests that tacit monitoring systems failed to findings do not speak to the actual efficacy of catch the mind’s drifting, and that it takes a conscious control. Nevertheless, they do raise higher-level explicit monitoring process to take concerns about the impact that a scientific dis- stock of the specific contents of thought and alert missal of conscious control might have on the one to the fact that they have wandered off task. population at large, and thus further highlight the Over the last several years, Schooler and col- importance of not overstating the degree to which leagues have used the mind-wandering phenome- science has shown consciousness to be impotent non to examine the function of meta-awareness in (Schooler, 2010; Shariff, Vohs, & Schooler, 2008). a domain where mind-wandering is antithetical to success (see Smallwood & Schooler, 2006 for a review). Specifically, Schooler, Reichle, and Halpern (2005) developed a paradigm to identify META-CONSCIOUS MONITORING temporal lapses of meta-awareness during the attentionally demanding task of reading. In this research, participants read passages of text and As we have discussed, some forms of control indicated every time they caught their minds might be automatized and unconscious, but others zoning out. They were then asked whether they clearly involve consciousness. In fact, one form of had been aware that they had been zoning out control may require explicitly articulating the prior to reporting it. In a second condition, par- content of the conscious state to bring it into ticipants were additionally probed intermittently metaconsciousness. Thus, periodically, the mind and asked to indicate whether they had been encounters situations which require more resource- zoning out at that moment. The results revealed dependent conscious monitoring processes. In that participants: (1) frequently caught themselves effect, this occurs anytime one explicitly attempts zoning out during reading; (2) were still often to answer the question, “What am I thinking or caught zoning out by the probes; and (3) fre- feeling?” Given that the answer requires an explicit quently reported that they had been unaware that representation of one’s state, over and above the they had been zoning out, particularly when they state itself, it offers individuals the opportunity to were caught by the probes. These findings demon- step out of the situation, which may be critical strate that individuals frequently lack meta-aware- for many of the innovative behaviors of which ness of drifting off, even when they are specifically individuals are capable. However, it also raises instructed to be vigilant for such lapses. the possibility that in the re-description process Additional studies have elucidated the distinct individuals might get it wrong. processes associated with the occurrence of mind- There are two kinds of dissociations between wandering (as revealed by probe-caught episodes) levels of mental representation that follow from and meta-awareness of mind-wandering (as the claim that metaconsciousness involves the revealed by self-caught episodes). These studies intermittent re-representation of the contents of also examined the role that self-awareness may consciousness (Schooler, 2002). Temporal disso- play in the transition from consciousness to meta- ciations occur when metaconsciousness temporar- consciousness. For example, Sayette, Reichle and ily fails to take stock of the current contents of Schooler (2009) examined the impact of alcohol thought (e.g., failing to notice that one is mind- on participants’ self-caught and probe-caught wandering during reading). Translation dissocia- mind-wandering during reading. Alcohol was an tions occur if the meta-representation process excellent candidate as a variable likely to impact misrepresents the original experience. Such meta-awareness, given its well-established capac- dissociations are particularly likely when one ity to reduce people’s more general self-awareness verbally reflects on non-verbal experiences or (e.g. Hull, 1981). In this study, half of the partici- attempts to takes stock of ambiguous experiences. pants received a real alcoholic beverage, whereas Several interesting social-cognitive phenomena the other half of participants received a placebo illustrate these metacognitive dissociations. that they believed contained alcohol. Participants then engaged in a reading task, in which they read War and Peace while their mind-wandering was Temporal dissociations assessed using both the probe-caught and self- caught procedures. Results showed that alcohol Mind-wandering had different effects on self-caught vs probe- We suspect that all readers have had the experi- caught mind-wandering. Despite mind-wandering ence of suddenly realizing that, despite their best more than twice as often as participants in the intentions and the fact that their eyes have contin- placebo group (as revealed by the probe measure), ued to move across the page, they have no participants in the alcohol group were no more idea what they have been reading. Such examples likely (and, indeed, were slightly less likely) to

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6622 22/13/2012/13/2012 9:54:239:54:23 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 63

catch themselves in the act. One can argue that failing to explicitly notice that they are relapsing participants in the alcohol group should have had (Tiffany, 1990). Traditionally, the occurrence of many more opportunities to catch themselves, but absent-minded relapses was taken to suggest that they did not catch themselves more often than the cravings need not trigger relapse. However, the sober participants. Apparently, they were impaired view suggested by the present perspective is that in their ability to notice mind-wandering episodes, absent-minded relapses may be associated with whereas sober participants were more capable of craving states in which one lacks metaconscious- detecting mind-wandering when it occurred. ness of the cravings (see Cheyene, Carriere, & A related study by Sayette, Schooler, & Reichle Smilek, 2006). By simultaneously promoting (2010) further illustrates the value of the self- absent-mindedness and decreasing meta-aware- caught/probe-caught methodology for assessing ness, the unnoticed craving state may induce a the conditions that impact on meta-awareness. unique condition in which individuals are maxi- This study also explored the role of motivation in mally likely to engage in a relapse behavior and meta-awareness. Specifically, this study focused minimally likely to notice themselves doing so. on the effect of cigarette craving on people’s Another approach for investigating the role of mind-wandering and their meta-awareness thereof. meta-awareness in mind-wandering is simply Like alcohol consumption, craving is a factor that asking participants after probe-caught mind-wan- might reasonably be expected to impact on meta- dering episodes whether or not they had been awareness. This is because craving may engage previously aware that their minds had drifted. working memory and self-regulation resources Although it is not self-evident that participants that otherwise might be directed toward noticing would necessarily be able to accurately reflect on that one has become distracted, thereby simulta- the prior meta-awareness of their mental states, a neously increasing the occurrence of one’s mind- variety of studies suggest that they can accurately wandering and decreasing the probability of discern whether or not they were meta-aware of catching it. To explore this issue, Sayette et al. their mind-wandering, as evidenced by the con- employed the reading task described above with sistent differences between mind-wandering epi- smokers, but in this case half of the participants sodes characterized as occurring with vs without were induced into a craving state by refraining meta-awareness. For example, in one study par- from smoking for at least 6 hours prior to the ticipants read text one word at a time and were experiment. All participants were then given the warned that it would periodically stop making reading task with both self-caught and probe- sense – become gibberish (Schooler, McSpadden, caught assessment of mind-wandering. The results Reichle, & Smallwood, 2010). We found that revealed that craving significantly impacted on although participants typically noticed the occur- participants’ meta-awareness of their mind-wan- rence of gibberish as soon as it began, if they dering. Despite mind-wandering more than three failed to notice it right away, they often continued times as often as participants in the no-craving for several sentences before they did. Moreover, if condition, participants in the craving condition participants were probed following extended dura- were not more likely to spontaneously catch them- tions of missing gibberish, they were markedly selves mind-wandering. more likely to report mind-wandering without These findings suggest that craving, like alco- meta-awareness than if they were probed at hol, disrupts individuals’ meta-awareness of the random intervals. Similarly, Smallwood et al. current contents of thought. In so doing, they also (2008) found that if individuals reported mind- highlight possible why it is so difficult to wandering without meta-awareness prior to self-regulate during craving states. While conven- encountering clues in a Sherlock Holmes story, tional holds that individuals are fully they were markedly less likely to solve the who- aware of their cravings, some researchers suggest dunnit. Mind-wandering with meta-awareness that cravings can occur unconsciously (Robinson proved far less disruptive to the construction of a & Berridge, 1993). This debate assumes that crav- mental model. ings must be either conscious or unconscious. When mind-wandering episodes are character- The alternative framework suggested here, in ized as having proceeded without meta-awareness, which consciousness is divided into experiential they are not only more disruptive to task perform- consciousness (contents of experience) and meta- ance but also more neurocognitively distinct from consciousness (explicit awareness of the contents on-task performance. In a recent study (Christoff of consciousness) suggests that individuals can et al., 2009), we scanned participants during a be conscious of craving but lack metaconscious- non-demanding vigilance task in which they had ness of the fact that they are craving. Such a mental to respond to every digit that appeared on the state, in turn, might contribute to relapses. Con- screen with the exception of a specific rarely pre- siderable research suggests that relapse behaviors sented target item for which they had to withhold may occur absent-mindedly, with individuals a response. Because the targets are relatively rare,

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6633 22/13/2012/13/2012 9:54:239:54:23 AMAM 64 THE SAGE HANDBOOK OF SOCIAL COGNITION

it is common to mind-wander during this task and someone points them out to us. If we commonly then accidently respond when a target is pre- lack metaconsciousness of affective states, the sented. In this study, participants were periodi- induction of continuous metaconsciousness may cally probed and asked whether they had been alter the quality of affective experience. Schooler mind-wandering, and if so whether or not they had et al. (2003) explored this issue by asking partici- been aware of that fact. Two sets of regions were pants to report on-line happiness while listening to more active prior to periods in which individuals hedonically ambiguous music (Stravinsky’s Rite reported having been mind-wandering relative to of Spring). The results showed that continuous having been on-task. One was the default network hedonic monitoring reduced individuals’ post- (including the medial (PFC), music ratings of happiness, relative to a condition posterior cingulate cortex/precuneus region and in which participants listened to music without the temporoparietal junction) which is known to monitoring. The fact that hedonic monitoring become increasingly recruited with decreasing altered participants’ experience suggests that (by external task demands. This result is interesting, default) individuals are, at most, only intermit- as the default network is often implicated in tently metaconscious of their affective state. stimulus-independent thought (Mason et al., 2007). The second set of regions was the execu- tive network, including the dorsal anterior cingu- Automaticity late cortex (ACC) and the dorsolateral PFC Automatic behaviors are often assumed to be (DLPFC), which becomes activated when indi- unconscious (Bargh, 1997; Wood et al., 2002). viduals engage in demanding mental activity. Of However, there is a peculiarity to this designation. particular interest to the present discussion was Consider a person driving automatically while the further finding that the discrepancy between engaging in some secondary task (e.g., talking). areas of neural activity associated with on-task vs Although such driving is compromised, one still off-task thought was markedly more pronounced experiences the road at some level. Thus, a more when individuals lacked meta-awareness of the appropriate characterization of the consciousness fact that they were mind-wandering. of automatic behaviors may be that they are The fact that participants’ awareness of their experienced but lack metaconsciousness, the latter mind-wandering impacted on the magnitude of only taking hold when individuals run into discrepancy between on-task vs off-task thinking difficulty. suggests potentially important implications for conceptualizing metaconsciousness. First, it reveals the robustness of the distinction between Unwanted thoughts those mental states that are accompanied by meta- Wegner (1994) suggested that individuals possess awareness and those that are not. One might easily an implicit monitoring system that tracks unwanted have conjectured that participants would have a thoughts (e.g., of a white bear) in order to veer hard time reliably characterizing their mind- away from them. But what exactly is this system wandering as having been associated with meta- monitoring? Wegner suggests that it is monitoring awareness. However, these findings suggest that the contents of preconsciousness (i.e., thoughts this distinction was easy for participants to make. that are near, but below, the threshold of con- After all, mind-wandering episodes accompanied sciousness). However, another, and perhaps more by meta-awareness (vs no meta-awareness) were intuitive, possibility is that these systems actually associated with different patterns of brain activa- monitor the contents of consciousness itself: that tion. Second, one might have thought that when is, perhaps individuals can consciously think participants were mind-wandering without realiz- about a white bear, without explicitly realizing ing it, that their brains would show relatively little that they are doing so. In this case, the monitoring activity compared to being on-task. To the con- system can catch the unwanted thought and raise trary, it was during periods of mind-wandering it to the level of meta-awareness, in effect saying: without meta-awareness that individuals showed “There you go again, thinking about that unwanted the maximum distinctive regions of brain activa- thought.” Recent evidence for this account comes tion. Apparently, while mind-wandering without from a study in which participants were asked meta-awareness may be especially counterproduc- to try not to think about a previous romantic tive to ongoing task performance, the mind is relationship while reading or while simply far from being empty and rather is engaging in sitting quietly (Fishman, Smallwood, & Schooler, vigorous brain activity. 2006). As in standard unwanted-thought para- digms, participants were asked to self-report every time they noticed an unwanted thought coming Well-being appraisals to mind. In addition, however, they were periodi- We often fail to explicitly notice our own emo- cally randomly asked whether at that particular tional states (e.g., sullenness, cheerfulness) until moment they were having the unwanted thought.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6644 22/13/2012/13/2012 9:54:249:54:24 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 65

The results revealed that participants frequently resulting from verbal reflection have also been experienced “unnoticed unwanted thoughts” about observed in various other domains hypothesized their previous relationship, which they experi- to rely on non-verbal cognition. Thinking aloud enced but failed to notice until they were probed. during can disrupt the intuitive Furthermore, these unnoticed unwanted thoughts processes associated with problem solving were detrimental to participants’ performance on a while having no effect on the logical processes test of the reading material, suggesting again that associated with analytical problem solving they were conscious. Intriguingly, participants for (Schooler, Ohlsson, & Brooks, 1993). Verbally whom the unwanted thoughts carried emotional reflecting on the basis of affective judgments can weight (i.e., they still wished they were in the interfere with the quality of affective decision relationship) were less likely than participants making, as assessed both by the opinions of who no longer wanted to be in the relationship to experts (Wilson & Schooler, 1991) and by post- notice the thoughts themselves, and more likely to choice satisfaction (Wilson, Lisle, Schooler, be caught having the thought. This suggests that Hodges, Klaaren, & LaFleur, 1993). Verbally cognitive defenses do not banish disturbing articulating the basis of the match between ana- thoughts to the unconscious, but rather prevent us logical stories can reduce people’s sensitivity to from reflecting on them (Schooler, 2001). meaningful deep-structure relationships, while increasing their emphasis on superficial surface−structure relationships (Sieck, Quinn, & Schooler, 1999). Of course, in many cases verbal Translation dissociations analysis can be helpful. This occurs when experi- ences are readily translated into words, due either The idea that metaconsciousness involves a to the nature of the task (e.g., logical problem re-representation of the contents of consciousness solving, Schooler et al., 1993) or to individuals’ suggests that, as with any recoding process, unique verbal expertise (e.g., wine experts, some information may get lost or distorted in Melcher & Schooler, 1996). However, the process translation. The likelihood of noise perturbing the of articulating experiences can sometimes result translation process may be particularly great when in translation dissociations, where meta-awareness individuals: (1) verbally reflect on inherently non- misrepresents conscious content. verbal experiences; (2) are motivated to misrepre- sent their experience; or (3) possess a lay theory that is inconsistent with their actual experience. Motivation In some situations, people may be motivated to misrepresent their experiences to themselves. For Verbal reflection example, homophobic individuals may not want Some experiences are inherently difficult to put to recognize when they are aroused by depictions into words: the structure of a face, the taste of a of homosexual acts (Adams, Wright, & Lohr, wine, the complex tonalities of Stravinsky, and the 1996): that is, individuals may consciously experi- intuitions leading to . If individuals attempt ence the arousal but, because of their motivation, to verbalize these inherently non-verbal and holis- fail to become meta-aware of it (see also Lambie tic experiences, the resulting re-representations & Marcel, 2002). Our perspective also suggests a may fail to do justice to the original experience. different view of repression. Freud argued that Schooler and Engstler-Schooler (1990) examined repression prevented unwanted feelings from the effects of describing faces, which, because of coming to consciousness, but we would say that it their holistic nature, are notoriously difficult to primarily prevents such feelings from reaching capture in words. Participants viewed a face and meta-awareness (Schooler, 2001; Schooler & subsequently either described it in detail or Schreiber, 2004). engaged in an unrelated verbal activity. When given a recognition test that included a different photograph of the target face, along with similar Stereotyping distractors, verbalization participants performed The distinction between conscious and metacon- substantially worse than controls. This effect of scious states also provides a way of potentially verbalization, termed “verbal overshadowing,” reconceptualizing existing findings in the domain has been found in variety of other domains of of stereotyping. For example, several researchers visual memory (Schooler, Fiore, & Brandimonte, work with the notion of “aversive racists,” defined 1997), including colors (Schooler & Engstler- as individuals who reveal evidence of implicit Schooler, 1990), shapes (Brandimonte, Schooler, racism but are not conscious of their racist & Gabbino, 1997), as well as other modalities tendencies (e.g., Gaertner & Dovidio, 1986; Son such as audition (Schooler et al., 1997) and taste Hing et al., 2008). This idea speaks directly to (Melcher & Schooler, 1996). Similar disruptions the disparities that can emerge when discrepant

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6655 22/13/2012/13/2012 9:54:249:54:24 AMAM 66 THE SAGE HANDBOOK OF SOCIAL COGNITION

motivations exist at different levels of conscious- to the standard view, attitudes are either con- ness. Aversive racists are identified empirically as sciously held (as revealed by standard attitudinal being those individuals who score high on racism measures) or else inaccessible to consciousness when gauged with implicit measures − i.e., the (as revealed by implicit measures such as the Implicit Association Test (IAT) − but low when IAT). However, the distinction between con- gauged with explicit measures. Evidence for the sciousness and metaconsciousness raises the pos- importance of this distinction comes from the sibility that some measures that have been examination of aversive racists’ evaluations of characterized as implicit might instead be con- stories depicting other-race target individuals, scious but lacking in metaconsciousness. who vary with respect to the degree to which low According to this view, there may be some situa- liking ratings can be attributed to something else tions in which individuals experience negative besides race. When aversive racists have no attitudes but do not acknowledge them as such, excuse for holding negative attitudes towards as was suggested as an interpretation of the other-race individuals (e.g., when the target person earlier described findings by Song Hing and is characterized as acting politely), then they colleagues. Some implicit measures may in fact behave very much like individuals with no racist tap such experienced but non-reported attitudes. tendencies. However, when there is an opportunity In contrast, other implicit measures may tap to justify their discriminatory behavior in a manner attitudes that are truly unconscious and never that does not necessarily invoke the label of experienced. This account may explain why “racist” (e.g., when the target individual behaves various tests of implicit attitudes such as the in a slightly unfriendly manner), these individuals IAT versus implicit priming can often be uncor- do act like racists. Son Hing et al. (2008) suggest related (Gawronski & Payne, 2010). Some tests that aversive racists behave in this fashion because may be measuring truly unconscious attitudes they hold non-conscious racist views that are whereas others may be measuring attitudes inconsistent with their conscious views and can that are experienced but occur without meta- only rely on their racist tendencies when they can awareness. avoid construing them as such. However, the dis- tinction between consciousness and metacon- sciousness raises another possibility: namely, that Faulty theories when individuals experience racist tendencies, Finally, translation dissociations can occur if indi- they simply do not recognize this experience due viduals have a faulty theory about what they to motivation not to take stock of racist reactions. should be feeling in a particular situation, which Accordingly, when confronted with the behaviors then colors their appraisal of their actual experi- of an individual towards whom they have racist ence. Individuals’ theories of how they think they attitudes, aversive racists experience negative should feel may also color their retrospective affect. If a justification for this affect exists that is appraisal of prior experiences. consistent with their views of themselves (i.e., that We have explored this possibility in the fasci- the individual behaved somewhat rudely), then nating but controversial domain of childhood they embrace this affect. However, when no such memory and childhood happiness. It is likely that outlet is available, they ignore it. Critical to this at least some aspects of one’s answer to the ques- account, however, is the notion that aversive rac- tion “How happy was your childhood?” are con- ists are actually experiencing the affect; it is structed using one’s naïve theories of how one simply a matter of whether or not they are pre- “must have felt” (Skurnik, Schwarz, & Winkielman, pared to allow themselves to take stock of it. 2000). One cue to this assessment comes from Thus, a reasonable alternative way to characterize cognitive outcomes, such as one’s memory for the aversive racists is to suggest that they experience period, with the bridge linking memory and hap- racism but lack explicit awareness of this piness provided by naïve theories. To investigate experience – or, in the terms of metaconscious- this process, Winkielman and Schwarz (2001) first ness theory, that they exhibit translation dissocia- asked participants to recall few (easy task) or tions due to a motivation not to acknowledge their many (difficult task) examples from their child- racist tendencies. hood. Participants were then asked to evaluate the quality of their childhood. After completion of the recall task, but prior to rating their childhood Implicit attitudes happiness, participants were provided with two The distinction between consciousness and different theories. Some participants were told metaconsciousness may also have important that psychologists have found that a poor child- implications for the study of implicit attitudes hood memory indicates an unhappy childhood, more generally (Greenwald et al., 2002). According because many unpleasant experiences are

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6666 22/13/2012/13/2012 9:54:249:54:24 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 67

purged from memory. Others were told that psy- COGNITION AND AFFECT WITHOUT chologists have found that a poor childhood CONSCIOUSNESS memory indicates a happy childhood, because many unpleasant experiences are ruminated upon. So far we have emphasized that many mental For both groups, it was emphasized that these are states are actually conscious, though lacking poorly supported hypotheses and that the relevant metacognitive representations. However, there are evidence is limited to small and unusual clinical clearly cases when a mental state has a demonstra- samples, making it worthwhile to test these ble influence on behavior, but cannot be directly hypotheses with a general college population. As accessed by consciousness. It is now almost uni- predicted, participants’ ratings of their childhood versally accepted that many perceptual and depended on the naïve theory offered to them cognitive operations can occur without individuals and the subjective difficulty of retrieving being conscious (Kihlstrom, 2007). One classic memories. When participants found it difficult to example comes from research on so-called retrieve childhood events, participants who were blindsight patients with damaged primary visual told that happy events fade from memory evalu- cortex (area V1 of striate cortex) but intact ated their childhood as happier than did partici- subcortical visual pathways. These patients can pants who were told that bad events fade from discriminate simple visual features (e.g., location, memory. The opposite happened when the retrieval or shape), as revealed in pointing and guessing was easy. behavior, while denying any awareness of the Distortions in retrospective meta-awareness discriminated features (Weiskranz, 1986). In that may also be one why individuals come case, the mental representation (e.g., “x is a to characterize childhood abuse experiences as square”) is genuinely unconscious − the patients having been previously forgotten (Schooler, 2001). truly do not know that they “know” what shape Recent evidence suggests that characterizations of was presented. Another classic example is patients abuse experiences as having been long forgotten with visual agnosia who cannot consciously and suddenly remembered may at least sometimes recognize visual features, but have largely intact be the result of metacognitive failures in which sensorimotor abilities, and can pursue actions in individuals overestimate the degree to which the response to these “unconscious” features (Milner memory was previously forgotten (Geraerts et al., & Goodale, 1995). As discussed earlier, the social 2008, Schooler, 2001). Factors that may contrib- cognition literature is rich with demonstrations of ute to such distortions are faulty theories in which unconscious processing in behavioral experiments individuals assume that they would surely recall (Bargh, 1997). having thought about past trauma, unless it was But what about affect? There is certainly actively repressed. Evidence in partial support of research suggesting that briefly presented affec- this view comes from a survey study (Joslyn & tive stimuli can work as unconscious triggers of Schooler, 2006) in which participants’ recollec- conscious affective states (Kihlstrom, 2007; tions of prior inappropriate sexual experiences Öhman, Flykt, & Lundqvist, 2000; Zajonc, 1994). were influenced by the manner in which questions But can affect itself be consciously inaccessible? about the experience were asked. Participants In recent years, one strand of our work has were asked to define what experiences constitute focused on exploring a dissociation which occurs sexual abuse either before or after they were asked when a person is in a demonstrable affective state if those experiences had ever happened to them. (as evidenced by its impact on behavior, physiol- Individuals who defined sexual abuse before ogy, and cognition) without having conscious considering whether they had been abused them- access to that state (Winkielman & Berridge, selves were more likely to identify themselves as 2004). The idea of “unconscious affect” may s having been abused, and reported being more eem initially strange – after all, how can there be upset at the time that the event occurred. feelings that are not felt? Note, though, that evolu- Apparently, thinking about the experience in the tionarily speaking, conscious representation of context of sexual abuse invoked theories about affect in the form of a “feeling” is a late achieve- how upsetting the experience must have been, ment compared with the ability to respond affec- which in turn colored the way they recalled feel- tively to relevant stimuli, which is present in ing at the time. If, as adults, individuals believe animals that extend deep into our evolutionary they were more traumatized by abuse than they ancestry, such as fish and reptiles. Accordingly, actually were as children, then the experience the basic affective neurocircuitry is contained may not have been as significant at the time, and in the subcortical brain, and can operate even in thus no special forgetting mechanisms may have the absence of cortex (Berridge, 2003). A reader been required for the experience to have been interested in a recent review of neuroscientific forgotten. evidence for “unconscious emotion” may want to

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6677 22/13/2012/13/2012 9:54:249:54:24 AMAM 68 THE SAGE HANDBOOK OF SOCIAL COGNITION

consult Winkielman, Berridge, and Sher (2011). value. Other participants rated various shades of However, evolutionary and neuroscientific consid- their current feelings on a 20-item scale. The erations can only be suggestive of unconscious results showed that the subliminal expressions affect in typical . We thus embarked on influenced the perceived value of the drink, with a program of psychological research to test this happy faces leading to higher ratings of willing- possibility using standard experimental paradigms ness to pay and the desire to drink. Again, no with normal college participants. changes in feelings were found. In sum, this study supported the idea that unconscious affect works via change in the perception of the desirability and value of presented options, without manifesting Impact of subliminal affective itself as a change in subjective experience. stimuli on behavior, but not on It is also worth highlighting that in both studies subjective experience described above, the effect of prime was amplified by thirst (Winkielman et al., 2005). This is con- One way of testing unconscious emotion involves sistent with other work from social psychology separating the impact of affective stimuli on suggesting that unconscious cues interact with behavior from their impact on conscious feelings. affective and motivational states in determining This was done in a series of studies by Winkielman, goal-oriented behavior (see Chapter 5; see also Berridge, and Wilbarger (2005). In Study 1 par- Custers & Aarts, 2010; Ferguson, 2007). However, ticipants were first presented with a series of our interpretation is that thirst does not represent subliminal emotional facial expressions – happy, an “unconscious goal,” but rather a low-level neutral, or angry. Immediately after this affect motivational amplifier of incoming affective cues induction, participants were given two counterbal- (Winkielman et al., 2011). anced tasks. One task required participants to self-report on conscious feelings of valence and arousal − a measure of introspective access to the current affective state. The other task was a meas- True dissociation or lack ure of behavioral impact of the current affective of meta-awareness? state and asked participants to take a pitcher of lemonade-like beverage and to pour into their cup One can ask whether participants in the above as much as they wanted and to drink as much as studies had no experience of their affective reac- they wanted. tion (true unconscious affect) or whether they The results of this study illustrate that sublimi- simply lacked meta-awareness of conscious affec- nal emotional expressions can influence people’s tive states (experienced but unrealized affect). actual consumption behavior. Subliminal happy After all, it is possible that participants were not facial expressions caused participants to pour attending on-line to their feelings, or did not more into their own cup, and to drink more than consider their subliminally biased feelings as a angry facial expressions. Importantly, participants potential impairment to their judgments and thus reported no conscious awareness of any interven- ignored them. Other studies from our lab have ing change in their subjective state, as measured addressed this possibility. by their reports of valence and arousal: that is, In one series of studies, participants were sub- they did not report feeling more pleasant (or liminally flashed facial expressions of happiness aroused) after happy facial expressions than angry and anger that were masked by to-be-rated Chinese expressions. ideographs (Winkielman, Zajonc, & Schwarz, This study suggests that consciously inaccessi- 1997). In addition, the studies employed various ble affective states can drive behavior. However, attributional manipulations in which some partici- how does unconscious affect accomplish this pants were informed about the possibility of result? After all, many steps of the consumption change in their affective experience and offered behavior are consciously mediated, in the sense possible causes of such change (irrelevant “other” that they require the ability to understand verbal pictures, irrelevant background music). If partici- instructions, form an intention, and execute com- pants’ feelings are indeed consciously accessible plex movements. To understand this process, we and form the basis of their judgments, such attri- tested the idea that unconscious affect directly butional manipulations should trigger corrective modifies the perceived value of presented options, processes, such as discounting and augmenting without changing conscious feelings. In Study 2, (Schwarz & Clore, 1983). However, the results we presented people with the same series of sub- of these studies showed no evidence of any dis- liminal happy or angry faces. Then some partici- counting or augmenting effects, as predicted by pants were given just a single sip of the fruit the attributional account (Winkielman et al., beverage, and were asked to rate its perceived 1997). Furthermore, there was also no evidence

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6688 22/13/2012/13/2012 9:54:249:54:24 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 69

for feelings in participants’ self-reports of experi- affective system, such as post-auricular startle ence, again consistent with the idea that the reflex (Starr, Lin, & Winkielman, 2007). facially triggered affect was unconscious. Participants showed more post-auricular startle Results of a recent study reaffirmed this con- reflex to unconscious happy rather than angry clusion with another paradigm (Bornemann, faces, suggesting genuine, albeit weak, activation Winkielman, & van der Meer, 2011). The study of the low-level affect system. Recently, we fol- investigated whether people can somehow lowed up on this finding and presented partici- “feel” their reactions to briefly presented emo- pants with affective pictures (faces or pictures) tional stimuli by deliberately focusing on their designed to induce an affective state, and evalua- internal subjective state. Specifically, participants tive words (adjectives and nouns) selected to were briefly flashed happy, neutral or angry faces activate relevant semantic dimensions. As and were asked to identify their valence. One expected, faces and pictures elicited stronger group of participants was instructed to do this task physiological responses than valence-matched while focusing on their feelings. One control words. Critically, changes in behavior were driven group was instructed to use a visual focus strategy, more strongly by pictures and faces, consistent and another group received no strategy instruc- with the notion that the behavioral changes reflect tions. The results showed no beneficial effect of genuine “hot” affect, rather than cold evaluation feeling-focus instruction on detection rates, sug- (Starr, Winkielman, Golgolushko, 2008). gesting that the affective responses to faces were In sum, a range of findings from behavioral and unconsciously unavailable, despite participants physiological experiments suggest that one can trying to use them. obtain genuine dissociation between an underly- ing affective process and its conscious, experien- tial awareness. As such, they give credence to the notion of “unconscious affect.” Importantly, Physiological consequences: Unfelt though, the idea of “unconscious emotion” does but genuine affect not imply that conscious feelings are an unneces- sary “icing on the emotional cake” (LeDoux, One issue often raised about these findings is the 1996). Conscious happiness, anxiety, anger, guilt, nature of the unconscious affective states. Are the and sadness are critical in people’s life. They may unconscious states elicited by subtle and brief as well be what makes life worth living. As an stimuli (like faces) simply “evaluative,” in the example, most people probably would not spend sense of changes in activation of value-related, but money on substances that make them only “uncon- cold concepts like “goodness” or “badness”? sciously happy,” but result in “happy” behavior. Or are they genuinely “hot” – in the sense of being In contrast, they are clearly willing to spend represented across multiple physiological and on substances, like alcohol or drugs, which psychological systems? This is a difficult ques- influence conscious states, without doing much tion, especially since unconscious affective states good to behavior. Besides recreational reasons, are likely to be weaker and less differentiated conscious emotions are actually useful in judg- (Clore, 1994). Still, some recent evidence from ments and decisions, giving the decision makers our lab suggests that unconscious affect involves valuable feedback that they might, but are not genuine physiological changes and is distinguish- forced to, exp licitly consider in making choices able from pure evaluative states. (Winkielman, Knutson, Paulus, & Trujillo, 2007). In the just-described study by Bornemann et al. (2011), in addition to behavioral responses, we monitored participants’ physiological activity using facial electromyography (EMG). The results Unconscious or not metaconscious? revealed distinct physiological responses for dif- ferent stimulus valences. Angry faces produced The preceding discussion argued for some cases the strongest reactions on the frown-generating where mental states, like affect, can remain genu- corrugator supercilii, and happy faces produced inely unconscious. However, this discussion also the lowest reactions. This suggests that briefly illustrates how hard it is to distinguish empirically presented and unfelt faces generate at least some between processes that are genuinely unconscious muscular reactions. or conscious but not meta-aware. This is tricky, However, one could argue that the facial EMG as a failure of verbal report can result from either responses to faces could simply represent motor an absence of experience or an absence of meta- mimicry. Thus, in other studies we assessed the awareness. However, future studies may help to reactions to unconsciously presented facial pic- adjudicate between them. For example, if unre- tures using physiological measures that serve ported states are indeed represented in conscious- as an index of activation of a low-level positive ness, then in principle they should be influenced

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 6699 22/13/2012/13/2012 9:54:249:54:24 AMAM 70 THE SAGE HANDBOOK OF SOCIAL COGNITION

by manipulations targeting consciousness, such as between conscious and metaconscious representa- cognitive load or explicit monitoring. Experiences tions, highlighting both temporal and translation in the absence of meta-awareness can also be dissociations. We also highlighted the possibility revealed retrospectively. For example, it is possi- that some seemingly unconscious states may ble to catch conscious, but not meta-aware, states in fact be conscious states lacking in meta- with the external probe procedure, which, as awareness, and showed how this possibility affects described earlier, was successfully employed in the interpretation of some major social-cognitive research on zoning out and unnoticed, unwanted phenomena. Finally, we considered some genu- thoughts. In principle, similar strategies could be inely unconscious mental events, including the used in other paradigms. For example, perhaps curious phenomenon of “unconscious affect.” individuals who fail to spontaneously report a What are some implications of all this for goal (e.g., competition) could be caught con- social cognition? Let’s start with some concrete sciously experiencing such goal states, if probed recommendations. First, as we have discussed at the right time. It may also be possible to refine throughout, researchers need to test more thor- individuals’ ability to carefully scrutinize their oughly if a mental content – a thought, a goal, or prior state. For example, if individuals are a feeling − is possibly experienced but not meta- experiencing something without concurrent meta- cognitively available before declaring it uncon- awareness, then in principle it may be possible to scious. Besides psychological tools, there are now have them later recall and metacognitively repre- plenty of physiological tools available that can sent their prior state, when some additional source facilitate such determination. Second, given that of self-insight (e.g., training) is pro- metacognition necessarily involves a process of vided or due to motivation are removed. translation, often into a verbal form, of partici- Thus, individuals going through the break-up of a pants’ thoughts and feelings, researchers should romantic relationship may retrospectively recog- use methods which capture the “first-order” nature nize past experiences of jealously or anger that had of these states . This may again involve behavioral previously escaped meta-awareness. Of course, techniques as well as psychophysiology. More retrospective analyses have their own pitfalls, as it theoretically, as we have emphasized throughout, is possible to infer states that may not have actually many of today’s popular frameworks depend (both been experienced at the time (Joslyn & Schooler, in the sense of conceptual foundations as well as 2006). However, if individuals are capable of retro- popularity) on the assumption that the process is spectively reporting states for which they lack a unconscious (e.g., unconscious goals, unconscious basis for inference (e.g., determining whether they thought, unconscious emotion, etc.). Challenging were subject to subliminally presented mood this assumption may require major reformulation manipulations), then the conclusion that the state of these theories. was experienced seems reasonable. Ultimately, determination of whether or not unreported states are genuinely unconscious, or experienced but not meta-aware, will come down to an assessment of CONCLUSION the preponderance of evidence in each case. Over the last several years there has been remark- able progress in the psychology and neuroscience of consciousness and metacognition. Clearly, SUMMARY AND IMPLICATIONS some believe that certain features of conscious- ness, like subjectivity, qualia, and intentionality, This chapter discussed the conscious, metacon- will never submit to scientific scrutiny (Searle, scious, and the unconscious, focusing on the 1997). Others believe that as research progresses, domain of social cognition. Our goal here was to “consciousness” will disappear from the scientific offer not only a review of theories and findings vocabulary (Rey, 1983). In contrast, we believe from our and related labs but also to problematize that future scientific advances will only increase some common assumptions and provide a fresh our appreciation of the marvels of consciousness perspective on classic issues. We began by empha- and metacognition. We hope that social cognitive sizing the resurgence of consciousness in contem- researchers will continue to contribute vigorously porary science, and contrasted this view with to this progress. eliminativism of mainstream social cognition. We then discussed what makes mental events con- scious, and highlighted the role of consciousness in complex thought and action. Following this, we ACKNOWLEDGMENT distinguished between conscious and metacon- scious states. We considered various disconnections We thank Shlomi Sher for his generous comments.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 7700 22/13/2012/13/2012 9:54:249:54:24 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 71

REFERENCES retrieval cues. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23, 915−931. Abrams, R. L., & Greenwald, A. G. (2000). Parts outweigh the Breitmeyer, B. G., & Ogmen, H. (2006). Visual masking: Time whole (word) in unconscious analysis of meaning. slices through conscious and unconscious Vision (2nd ed.). Psychological Science, 11, 118−124. New York: Oxford University Press. Acker, F. (2008). New findings on unconscious versus con- Carver, C. S., & Scheier, M. S. (1990). Origins and functions of scious thought in decision making: additional empirical positive and negative affect: A control-process view. data and meta-analysis. Judgment and Decision Making, 3, Psychological Review, 197, 19−35. 292–303. Cheyene J. A., Carriere J. S. A., & Smilek D. (2006). Absent- Adams, H. E., Wright, L. W., & Lohr, B. A. (1996). Is homopho- mindedness: Lapses of conscious awareness and everyday bia associated with homosexual arousal? Journal of cognitive failures. Consciousness & Cognition, 3, 578–592. Abnormal Psychology, 105, 440−445. Christoff, K., Gordon, A. M., Smallwood, J., Smith, R., & Baars, B. J. (1988). A cognitive theory of consciousness. Schooler, J. W. (2009). Experience sampling during fMRI New York: Cambridge University Press. reveals default network and executive system contributions Bargh, J. A. (1989). Conditional automaticity: Varieties of to mind wandering. Proceedings of the National Academy automatic influence in social perception and cognition. of Sciences, 106, 8719−8724. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought Churchland, P. M. (1981). Eliminative materialism and the (pp. 3–51). New York: Guilford Press. propositional attitudes. Journal of Philosophy, 78(2), Bargh, J. A. (1997). Advances in social cognition. In 67−90. R. S. Wyer, Jr. (Ed.), The automaticity of everyday life Churchland, P. S. (2005). Brain wide shut. New Scientist, 30 (pp. 1−61). Mahwah, NJ: Erlbaum. April, pp. 46–49. Bargh, J. A. (2007). Social psychological approaches to con- Cleeremans, A. (2005). Computational correlates of con- sciousness. In P. Zelazo, M. Moscovitch, & E. Thompson sciousness. Progress in Brain Research, 150, 81–98. (Eds.), The Cambridge handbook of consciousness. Clore, G. L. (1994). Why emotions are never unconscious. In New York: Cambridge University Press. P. Ekman & R. J. Davidson (Eds.), The nature of emotion: Bargh, J. A., & Chartrand, T. L. (1999). The unbearable auto- Fundamental questions (pp. 285−290). New York: Oxford maticity of being. American Psychologist, 54, 462–479. University Press. Baumeister, R. F. (2008). Free will in scientific psychology. Crick, F. (1994). The astonishing hypothesis. New York: Perspectives on Psychological Science, 3, 14−19. Scribner’s. Baumeister, R. F., & Masicampo, E. J. (2010). Conscious Custers, R., & Aarts, H. (2010). The unconscious will: How the thought is for facilitating social and cultural interactions: pursuit of goals operates outside of conscious awareness. How mental simulations serve the animal−culture inter- Science, 329, 47−50. face. Psychological Review, 117, 945−971. Dehaene, S., Changeux, J. P., Naccache, L., Sackur, J., & Baumeister, R. F., Schmeichel, B. J., DeWall, C. N., & Vohs, K. D. Sergent, C. (2006). Conscious, preconscious, and sublimi- (2007). Is the conscious self a help, a hindrance, or an nal processing: A testable taxonomy. Trends in Cognitive irrelevance to the creativity process? In A. M. Colombus Sciences, 10, 204−211. (Ed.), Advances in psychology research (Vol. 53, Deutsch, R., Gawronski, B., & Strack, F. (2006). At the bound- pp. 137−152). New York: Nova. aries of automaticity: Negation as reflective operation. Berridge, K. C. (2003). Comparing the emotional brain of Journal of Personality and Social Psychology, 91, humans and other animals. In: R. J. Davidson, 385−405. H. H. Goldsmith, & K. Scherer (Eds.), Handbook of affective DeWall, C. N., Baumeister, R. F., & Masicampo, E. J. (2008). sciences (pp. 25−51). New York: Oxford University Evidence that logical reasoning depends on conscious Press. processing, Consciousness and Cognition, 17, 628−645. Block, N. (1995). On a confusion about a function of con- Diamond, A., Barnett, W. S., Thomas, J., & Munro, S. (2007). sciousness. Behavioral and Brain Sciences 18, 27−287. Preschool program improves cognitive control. Science, Bongers, K. C. A., & Dijksterhuis, A. (2009). Consciousness as 318, 1387−1388. a trouble-shooting device? The role of consciousness in Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. goal-pursuit. In E. Morsella, J. A. Bargh, & P. Gollwitzer (2006). On making the right choice: The deliberation- (Eds.), The Oxford handbook of human action without-attention effect. Science, 311, 1005−1007. (pp. 589−604). New York: Oxford University Press. Ferguson, M. J. (2007). On the automatic evaluation of end- Boring, E. G. (1953). A history of introspection. Psychological states. Journal of Personality and Social Psychology, 92, Bulletin, 50, 169−189. 596−611. Bornemann, B., Winkielman, P., & van der Meer (2011). Can Fishman, D., Smallwood, J., & Schooler, J. W. (2006). you feel what you don’t see? Using bodily feedback to Unwanted and meta-unknown. Unpublished manuscript. detect briefly presented emotional stimuli. International Fiske, S. T., Gilbert, D. T., & Lindzey, G. (Eds.) (2010). Journal of Psychophysiology. doi:10.1016/j.ijpsycho. Handbook of social psychology (5th ed.). New York: 2011.04.007 Wiley. Brandimonte, M. A., Schooler, J. W., & Gabbino, P. (1997). Fitzsimons, G. M., & Bargh, J. A (2004). Automatic self- Attentuating verbal overshadowing through visual regulation. In Baumeister, R. F. & Vohs, K. D. (Eds.),

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 7711 22/13/2012/13/2012 9:54:249:54:24 AMAM 72 THE SAGE HANDBOOK OF SOCIAL COGNITION

Handbook of self-regulation: Research, theory and applica- Personality: Theory and Research (3rd ed.). New York: tions (pp. 151–170). New York: Guilford Press. Guilford Press. Flavell, J. H. (1979). Metacognition and cognitive monitoring: Kihlstrom, J. F. (2008). The automaticity juggernaut. In J. Baer, A new area of cognitive-developmental inquiry. American J. C. Kaufman, & R. F. Baumeister (Eds.), Are we free? Psychologist, 34, 906−911. Psychology and free will (pp. 155−180). New York: Oxford Fodor, J. (1968). Psychological explanation. New York: University Press. Random House. Kihlstrom, J. F., Schacter, D. L., Cork, R. L., Hurt, C. A., & Förster, J., Liberman, N., & Friedman, R. (2007). Seven prin- Behr, S. E. (1990). Implicit and explicit memory following ciples of automatic goal pursuit: A systematic approach surgical anesthesia. Psychological Science, 1, 303−306. to distinguishing goal priming from priming of non-goal Koch, C., & Tsuchiya, N. (2007). Attention and consciousness: constructs. Personality and Social Psychology Review, 11, Two distinct brain processes. Trends in Cognitive Science. 211−233. 11, 16−22. Gaertner, S. L., & J. F. Dovidio. (1986). The aversive form of Koriat, A. (2006). Metacognition and consciousness. In: racism. In: J. F. Dovidio & S. L. Gaertner (Eds.), Prejudice, Cambridge handbook of consciousness. New York: discrimination and racism: Theory and research Cambridge University Press. (pp. 61−89). Orlando, FL: Academic Press, Kunst-Wilson, W. R., & Zajonc, R. B. (1980). Affective dis- Gawronski, B., & Payne, B. K. (Eds.) (2010). Handbook of crimination of stimuli that cannot be recognized. Science, implicit social cognition: Measurement, theory, and appli- 207, 557−558. cations. New York: Guilford Press. Lambie, J. A., & Marcel, A. J. (2002). Consciousness and the Geraerts, E., Lindsay, D. S., Merckelbach, H., Jelicic, M., varieties of emotion experience: A theoretical framework. Raymaekers, L., Arnold, M. M., et al. (2008). Cognitive Psychological Review, 109, 219−259. mechanisms underlying recovered memory experiences of Laureys, S., Antoine, S., Boly, M., Elincx, S., Faymonville, M. E., childhood sexual abuse. Psychological Science, 20, 1. & J. Berre, J., et al. (2002). Brain function in the vegetative Gigerenzer, G., & Goldstein, D. (1996). Reasoning the fast and state. Acta Neurol Belg, 102, 177–185. frugal way: Models of bounded rationality. Psychological LeDoux, J. (1996). The emotional brain: The mysterious under- Review, 103, 650−669. pinnings of emotional life. New York: Simon & Schuster. Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of Levy, J., Pashler, H., & Boer, E. (2006). Central interference in mind perception. Science, 315, 619. driving: Is there any stopping the psychological refractory Greenwald, A. G. (1992). New Look 3: Reclaiming period? Psychological Science, 17, 228−235. unconscious cognition. American Psychologist, 47, Loftus, E. F., & Klinger, M. R. (1992). Is the unconscious smart 766−779. or dumb? American Psychologist, 47, 761−765. Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Logan, G. D. (1988). Toward an instance theory of automati- Nosek, B. A., & Mellott, D. S. (2002). A unified theory of zation. Psychological Review, 95, 492–527. implicit attitudes, stereotypes, self-esteem, and self-con- Mandler, G. (1975). Consciousness: Respectable, useful, and cept. Psychological Review, 109, 3−25. probably necessary. In R. Solso (Ed.), Information process- Greenwald, A. G., Draine, S. C., & Abrams, R. L. (1996). Three ing and cognition: The Loyola Symposium (pp. 229−254). cognitive markers of unconscious semantic activation. Hillsdale, NJ: Lawrence Erlbaum Associates. Science, 273, 1699−1702. Marcel, A. J. (1983). Conscious and unconscious perception: Hull, J. G. (1981). A self-awareness model of the causes and Experiments on visual masking and word recognition. effects of alcohol consumption. Journal of Abnormal Cognitive Psychology, 15, 197−237. Psychology, 90, 586−600. Mason, M. F., Norton, M. I., Van, J. D., Wegner, D. M., Hummel, J. E., & Holyoak, K. J. (2003). A symbolic-connection- Grafton, S. T., & Macrae, C. N. (2007). Wandering minds: ist theory of relational inference and generalization. The default network and stimulus independent thought. Psychological Review, 110, 220–264. Science, 315, 393−395. Jameson, A., Nelson, T. O., Leonesio, R. J., & Narens, L. McClelland, J. L., & Rogers, T. T. (2003). The parallel distrib- (1993). The feeling of another person’s knowing. Journal of uted processing approach to semantic cognition. Nature Memory and , 32, 320−335. Reviews Neuroscience, Joslyn, S., & Schooler, J.W. (2006). Influences of the present Mead, G. H. (1934). Mind, self, & society. Chicago, IL: on the past: The impact of interpretation on memory for University of Chicago Press. abuse. In L.G. Nilsson & N. Ohta (Eds.), Memory and Melcher, J., & Schooler, J. W. (1996). The misremembrance of society: Psychological perspectives. New York: Psychology wines past: Verbal and perceptual expertise differentially Press. mediate verbal overshadowing of taste. The Journal of Jost, J. T., Kruglanski, A. W., & Nelson, T. O. (1998). Social Memory and Language, 35, 231−245. metacognition: An expansionist review. Personality and Merikle, P. M., & Reingold, E. M. (1998). On demonstrating Social Psychology Review, 2, 137–154. unconscious perception. Journal of Experimental Kahneman, D. (1973). Attention and effort. Englewood Cliffs, Psychology: General, 127, 304−310. NJ: Prentice Hall. Metcalfe, J., & Mischel, W. (1999). A hot/cool system analysis Kihlstrom, J. F. (2007). The psychological unconscious. of delay of gratification: Dynamics of willpower. O. John, R. Robins, & L. Pervin (Eds.), Handbook of Psychological Review, 106, 3−19.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 7722 22/13/2012/13/2012 9:54:249:54:24 AMAM CONSCIOUSNESS, METACOGNITION, AND THE UNCONSCIOUS 73

Milner, D., & Goodale, M. (1995). The visual brain in action. Schooler, J. W. (2010). What science tells us about free will. Oxford: Oxford University Press. In R. F. Baumeister, A. R. Mele, & K. D. Vohs (Eds.), Free Nelson, T. O. (1996). Consciousness and metacognition. will and consciousness: How might they work? American Psychologist, 51(2), 102−116. (pp. 191−218). Oxford: Oxford University Press. Nisbett, R. E., & Wilson, D. S. (1977). Telling more than we Schooler, J. W., Ariely, D., & Loewenstein, G. (2003). The can know: Verbal reports on mental processes. Psychological pursuit and assessment of happiness can be self-defeating. Review, 84, 231−253. In I. Brocas & J. Carrillo (Eds.), The psychology of economic Norman, D. A., & Shallice, T. (1986). Attention to action: decisions. Oxford: Oxford University Press. Willed and automatic control of behaviour. In Schooler, J. W., & Engstler-Schooler, T. Y. (1990). Verbal R. J. Davidson, G. E. Schwartz, & D. Shapiro (Eds.), overshadowing of visual memories: Some things are better Consciousness and self-regulation: Advances in research left unsaid. Cognitive Psychology, 17, 36−71. and theory. New York: Plenum Press. Schooler, J. W., Fiore, S. M., & Brandimonte, M. A. (1997). At Öhman, A., Flykt, A., & Lundqvist, D. (2000). Unconscious a loss from words: Verbal overshadowing of perceptual emotion: Evolutionary perspectives, psychophysiological memories. In D. L. Medin (Ed.), The psychology of learning data and neuropsychological mechanisms. In R. D. Lane, and motivation (pp. 293−334). San Diego, CA: Academic L. Nadel & G. Ahern (Eds.), of Press.Schooler, J. W. & Mauss, I. B. (2010). To be happy emotion (pp. 296−327). New York: Oxford University Press. and to know it: The experience and meta-awareness of Pashler, H. E. (1998). The psychology of attention. Cambridge, pleasure. In K. Berridge & M. Kringlebach (Eds.), Pleasures MA: MIT Press. of the brain. Oxford: Oxford University Press. Petty, R. E., Briñol, P., Tormala, Z. L., & Wegener, D. T. (2007). Schooler J. W., McSpadden M., Reichle E. D., Smallwood J. The role of meta-cognition in social judgment. In (2010). Unnoticed nonsense: Mind-wandering can prevent E. T. Higgins & A. W. Kruglanski, (Eds.), Social psychology: people from realizing that they are reading gibberish. A handbook of basic principles (2nd ed., pp. 254−284). Manuscript submitted for publication. New York: Guilford Press. Schooler, J. W., Ohlsson, S., & Brooks, K. (1993). Thoughts Pessoa, L., McKenna M., Gutierrez E., & Ungerleider, L. G. beyond words: When language overshadows insight. (2002). Neural processing of emotional faces requires Journal of Experimental Psychology: General, 122, attention. Proc Natl Acad Sci USA, 99, 11458–11463. 166−183. Reichle, E., Pollatsek, A., Fisher, D.L., & Rayner, K. (1998). Schooler, J. W., Reichle, E. D., & Halpern, D. V. (2005). Toward a model of eye movement control in reading. Zoning-out during reading: Evidence for dissociations Psychological Review, 105, 125−157. between experience and meta-consciousness. In D. T. Levin Rey, A., Goldstein, R. M., & Perruchet, P. (2009). Does uncon- (Ed.), Thinking and seeing: Visual metacognition in adults scious thought improve complex decision making? and children (pp. 204–226). Cambridge, MA: MIT Press. Psychological Research, 73, 372−379. doi:10.1007/ Schooler, J., & Schreiber, C.A. (2004). Experience, meta- s00426-008-0156-4 consciousness, and the paradox of introspection. Journal Rey, G. (1983). A reason for doubting the existence of of Consciousness Studies, 11 (7–8), 17–39. consciousness. In R. Davidson, G. Schwartz, & D. Shapiro Schwarz, N. (2004). Meta-cognitive experiences in consumer (Eds.), Consciousness and self-regulation (Vol. 3, judgment and decision making. Journal of Consumer pp. 1−39). New York: Plenum Press. Psychology, 14, 332−348. Rickard, T. C. (2005). A revised identical elements model of Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and arithmetic fact representation. Journal of Experimental judgments of well-being: Informative and directive func- Psychology: Learning, Memory, and Cognition, 31, tions of affective states. Journal of Personality and Social 250−257. Psychology, 45, 513−523. Robinson, T. E., & Berridge, K. C. (1993). The neural basis of Searle, J. (1997). The mystery of consciousness. New York: drug craving: an incentive-sensitization theory of addiction. New York Review Press. Brain Research Reviews, 18, 247−291. Shanks, D. R. (2005). . In K. Lamberts & Rosenthal, D. (1986). Two concepts of consciousness. R. Goldstone (Eds.), Handbook of cognition (pp. 202−220). Philosophical Studies, 49, 329−359. London: Sage. Sayette, M. A., Reichle, E. D., & Schooler, J. W. (2009). Lost in Shanks, D. R. (2006). Are complex choices better made uncon- the sauce: The effects of alcohol on mind-wandering. sciously? Science, 313, 716. Psychological Science, 20, 747–752. Shariff, A., Schooler, J. W., & Vohs, K. (2008). The hazards of Sayette, M. A., Schooler, J. W., & Reichle, E. D. (2010). Out for claiming to have solved the hard problem of free will. a smoke: The impact of cigarette craving on zoning-out In J. Baer and R. Baumeister (Eds.), Psychology and free during reading. Psychological Science, 21, 26−30. will (pp. 181−204). Oxford: Oxford University Press. Schooler, J. W. (2001). Discovering memories in the light of Shiffrin, R. M., & Schneider, W. (1977). Controlled and auto- meta-awareness. The Journal of Aggression, Maltreatment matic human : II. Perceptual learn- and Trauma, 4, 105−136. ing, automatic attending and a general theory. Psychological Schooler, J. W. (2002). Re-representing consciousness: Review, 84, 127−190. Dissociations between consciousness and meta-conscious- Shinskey, F. G. (1979). Process control systems (2nd ed.). ness. Trends in Cognitive Science, 6, 339−344. New York: McGraw-Hill.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 7733 22/13/2012/13/2012 9:54:249:54:24 AMAM 74 THE SAGE HANDBOOK OF SOCIAL COGNITION

Skurnik, I., Schwarz, N., & Winkielman, P. (2000). Drawing Wegner, D. M. (1994). Ironic processes of mental control. inferences from feelings: The role of naive beliefs. Psychological Review, 101, 34−52. In H. Bless & J. P. Forgas (Eds.), The message within: Wegner, D. M. (2002). The illusion of conscious will. The role of subjective experience in social cognition and Cambridge, MA: MIT Press. behavior (pp. 162−175). Philadelphia: Psychology Press. Weiskrantz, L. (1986). Blindsight: A case study and its implica- Sieck, W. R., Quinn, C. N., & Schooler, J. W. (1999). tions. Oxford: Oxford University Press. Justification effects on the judgment of analogy. Memory Wilson, T. D. (2002). Strangers to Ourselves: Discovering the and Cognition, 27, 844−855. adaptive unconscious. Cambridge, MA: Belknap Press of Simons, D. J., & Chabris, C. F. (1999). Gorillas in our midst: Harvard University Press. Sustained inattentional blindness for dynamic events. Wilson, T. D., Lisle, D. J., Schooler, J. W., Hodges, S. D., Perception, 28, 1059−1074. Klaaren, K. J., & LaFleur, S. J. (1993). Introspecting about Smallwood, J., McSpadden, M. C., & Schooler, J. W. (2008). reasons can reduce post-choice satisfaction. Personality When attention matters: The curious incident of the and Social Psychology Bulletin, 19, 331−339. wandering mind. Memory & Cognition, 36, 1144−1150. Wilson, T. D., & Schooler, J. W. (1991). Thinking too much: Smallwood, J., & Schooler, J. W. (2006). The restless mind. Introspection can reduce the quality of preferences and Psychological Bulletin, 132, 946−958. decisions. Journal of Personality and Social Psychology, 60, Smith, E. R., & DeCoster, J. (2000). Dual process models in 181−192. social and cognitive psychology: Conceptual integration Winkielman, P. & Berridge, K. C. (2004). Unconscious and links to underlying memory systems. Personality and emotion. Current Directions in Psychological Science, 13, Social Psychology Review, 4, 108–131. 120−123. Son Hing, L. S., Chung-Yan, G. A., Hamilton, L. K., & Winkielman, P., Berridge, K., & Sher, S. (2011). Emotion, Zanna, M. P. (2008). A two-dimensional model that consciousness, and social behavior. In J. Decety & employs explicit and implicit attitudes to characterize J. T. Cacioppo (Eds.), Handbook of social neuroscience prejudice. Journal of Personality and Social Psychology, (pp 195−211). New York: Oxford University Press. 94, 971−987. Winkielman, P., Berridge, K. C., & Wilbarger, J. L. (2005). Starr, M. J., Lin, J., & Winkielman, P. (2007). The impact of Unconscious affective reactions to masked happy unconscious facial expressions on consumption behavior versus angry faces influence consumption behavior and involves changes in positive affect: Evidence from EMG judgments of value. Personality and Social Psychology and appetitive reflex-modulation. Poster presented at Bulletin, 1, 121−135. 47th Annual Meeting of Society for Psychophysiological Winkielman, P., Knutson, B., Paulus, M. P., & Trujillo, J. T. Research, Savannah, GA. (2007). Affective influence on decisions: Moving towards Starr, M. J., Winkielman, P., & Gogolushko, K. (2008). the core mechanisms. Review of General Psychology, 11, Influence of affective pictures and words on consumption 179−192. behavior and facial expressions. Poster presented at Winkielman, P. & Schooler, J. (2011). Splitting consciousness: Society for Psychophysiological Research, Austin, TX. Unconscious, conscious, and metaconscious processes in Strack, F., & Deutsch, R. (2004). Reflective and impulsive social cognition. European Review of Social Psychology, determinants of social behavior. Personality and Social 22, 1–35. Psychology Review, 8, 220−247. Winkielman, P., & Schwarz, N. (2001). How pleasant was your Sumner, P., Tsai, P-C., Yu, K., & Nachev, P. (2006). Attentional childhood? Beliefs about memory shape inferences from modulation of sensorimotor processes in the absence of experienced difficulty of recall. Psychological Science, 12, perceptual awareness. PNAS, 103, 10520−10525 176−179. Tiffany, S. T. (1990). A cognitive model of drug urges and Winkielman, P., Zajonc, R. B., & Schwarz, N. (1997). Subliminal drug-use behavior: Role of automatic and nonautomatic affective priming resists attributional interventions. processes. Psychological Review, 97, 147–168. Cognition and Emotion, 11, 433−465. Tononi, G. (2004). An information integration theory of con- Wood, W., Quinn, J., & Kashy, D. (2002). Habits in everyday sciousness. BMC Neuroscience, 5, 42. life: Thought, emotion, and action. Journal of Personality Triesch J, Ballard D. H., Hayhoe, M. M., Sullivan, B. T. (2003). and Social Psychology, 83, 1281−1297. What you see is what you need. Journal of Vision, 3, 86−94. Zajonc, R. B. (1994). Evidence for nonconscious emotions. In Vohs, K. D., & Schooler, J. W. (2008). The value of believing P. Ekman & R. J. Davidson (Eds.), The nature of emotion: in free will: Encouraging a belief in determinism increases Fundamental questions (pp. 293−297). New York: Oxford cheating. Psychological Science, 19, 49−54. University Press. Vygotsky, L. S. (1962). Thought and language. Cambridge, Zeeman, A. (2002). Consciousness: A user’s guide. New MA: MIT Press. Haven, CT: Yale University Press.

55698-Fiske-Ch04.indd698-Fiske-Ch04.indd 7744 22/13/2012/13/2012 9:54:249:54:24 AMAM

Top View