CHAPTER 11 Metacognition and Consciousness Asher Koriat Abstract reviews research addressing these questions, emphasizing its implications for issues con- The study of metacognition can shedlight on cerning consciousness; in particular, the gen- some fundamental issues about conscious- esis of subjective experience, the function of ness and its role in behavior. Metacognition self-reflective consciousness, and the cause- research concerns the processes by which and-effect relation between subjective expe- people self-reflect on their own cognitive rience and behavior. and memory processes (monitoring) and how they put their rnetaknowledge to use in regulating their information processing and Introduction behavior (control). Experimental research on metacognition has addressed the follow- There has been a surge of interest in ing questions. First, what are the bases of metacognitive processes in recent years, metacognitive judgments that people make with the topic of metacognition pulling in monitoring their learning, remembering, under one roof researchers from tradition- and performance? Second, how valid are ally disparate areas of investigation. These such judgments and what are the factors areas include memory research (Kelley that affect the correspondence between sub- & Jacoby, 1998; Metcalfe & Shimamura, jective and objective indexes of knowing? 1994; Nelson & Narens, 1990; Reder, 1996), Third, what are the processes that underlie developmental psychology (Schneider & the accuracy and inaccuracy of metacogni- Pressley, igq7), social psychology (Bless tive judgments? Fourth, how does the out- & Forgas, 2000; Jost, Kruglanski, & Nel- put of metacognitive monitoring contribute son, 1998; Schwarz, 2004), judgment and to the strategic regulation of learning and decision making (Gilovich, Griffin, & remembering? Finally, how do the metacog- Kahneman, 2002; Winman & Juslin, 2005), nitive processes of monitoring and con- neuropsychology (Shimamura, zooo), trol affect actual performance? This chapter forensic psychology (e.g., Pansky, Koriat, ZYO THO C,\MRRIDGb HANDBOOK OF CONSCIOUCNFCS & Goldsmith, 2005; Perfect, zooz), edu- trol). The object level, in contrast, has no cational psychology (Hacker, Dunlosky, & control over the meta level and no access Graesser, 1998), and problem solving and to it. For example, the study of new mate- creativity (Davidson & Sternberg, 1998; rial involves a variety of basic, object-level Metcalfe, 1998a). The establishment of operations, such as text processing, compre- metacognition as a topic of interest in its hending, rehearsing, and so on. At the same own right is already producing synergy time, metacognitive processes are engaged among different areas of investigation con- in planning how to study, in devising and cerned with monitoring and self-regulation implementing learning strategies, in moni- (e.g. Fernandez-Duque, Baird, & Posner, toring the course and success of object-level 2000). Furthermore, because some of the processes, in modifying them when neces- questions discussed touch upon tradition- sary, and in orchestrating their operation. In ally ostracized issues in psychology, such the course of studying new material, learners as the issues of consciousness and free are assumed to monitor their degree of com- will (see Nelson, 1996)~a lively debate prehension online and then decide whether has been going on between metacognitive to go over the studied material once again, researchers and philosophers (see Nelson how to allocate time and effort to different & Rey, 2000). In fact, it appears that the segments, and when to end studying. increased interest in metacognition research We should note, however, that the derives in part from the feeling that perhaps distinction between cognitive and metacog- this research can bring us closer to dealing nitive processes is not sharp because the with (certainly not resolving) some of the same type of cognitive operation may occur metatheoretical issues that have been the at the object level or at the meta level, and province of philosophers of thc mind. in some cases it is unclear to which level a particular operation belongs (Brown, 1987). Definition Research Traditions Metacognition concerns the study of what people know about cognition in general, Historically, there have been two main and about their own cognitive and mem- lines of research on metacognition that ory processes, in particular, and how they proceeded almost independently of each put that knowledge to use in regulating their other, one within developmental psychology information processing and behavior. Flavell and the other within experimental memory (1~71)introduced the term "metamemory," research. The work within developmen- which concerns specifically the monitoring tal psychology was spurred by Flavell (see and control of one's learning and remem- Flavell, 1979; Flavell& Wellman, 1977), who bering. Metamemory is the most researched argued for the critical role that metacog- area in metacognition and is the focus of this nitive processes play in the development chapter. of memory functioning (see Flavell, 1999). Nelson and Narens (1990) proposed Within memory research, the study of a conceptual framework that has been metacognition was pioneered by Hart's adopted by most researchers. According to (1965) studies on the feeling-of-knowing them, cognitive processes may be divided (FOK), and Brown and McNeill's (1966) into those that occur at the object level work on the tip-of-the-tongue (TOT). and those that occur at the meta level: The There is a difference in goals and object level includes the basic operations methodological styles between these two traditionally subsumed undcr the rubric of research trahtions. The basic assumption information processing - encoding, rehears- among developmental students of mctacog- ing, retrieving, and so on. The meta level is nition is that learning and memory per- assumcd to oversee object-level operations formance depend heavily on monitoring (monitoring) and return signals to regulate and regulatory proficiency. This assump- them actively in a top-down fashion (con- tion has resulted in attempts to specify the METACOGNI'rION AND CONSCIOUSNESS 9l components of metacognitive abilities, to In addition to the developmental and trace their development with age, and to the experimental-memory lines of research, examine their contribution to memory func- there has been considerable work on tioning. Hence a great deal of the work metacognition in the areas of social psy- is descriptive and correlational (Schneider, chology and judgment and decision mak- 1985). The focus on age differences and indi- ing. Social psychologists have long been vidual differences in metacognitive skills has concerned with questions about metacog- also engendered interest in specifying "defi- nition, although their work has not been ciencies" that are characteristic of children at explicitly defined as metacognitive (see Jost different ages and in devising ways to rem- et al., 1998). In particular, social psychol- edy them. This work has expanded into the ogists share the basic tenets of metacog- educational domain: Because of the increas- nitive research (see below) regarding the ing awareness of the critical contribution of importance of subjective feelings and beliefs, metacognition to successful learning (Paris & as well as the role of top-down regula- Winograd, ~gqo),educational programs have tion of behavior. In recent years social psy- been developed (see Scheid, 1993) designed chologists have been addressing questions to make the learning process more "metacog- that are at the heart of current research in nitive." Several authors have stressed specif- metacognition (e.g., Winkielman, Schwarz, ically the importance of metacognition to Fazendeiro, & Reber, 2003; Yzerbyt, I,ories, transfer of learning (see De Corte, 2003). & Dardenne, 1998; see Metcalfe, 1998b). The conception of metacognition by Within the area of judgment and decision developmental psychologists is more com- malung, a great deal of the work concern- prehensive than that underlying much of ing the calibration of probability judgments the experimental work on metacognition. (Fischhoff; 1975; Lichtenstein, Fischhoff, & It includes a focus on what children know Phillips, 1982; Winman & Juslin, 2005) is about the functioning of memory and par- directly relevant to the issues raised in ticularly about one's own memory capac- metacognition. ities and limitations. Developmental work has also placed heavy emphasis on strate- Research Questions gies of learning and remembering (Bjorklund & Douglas, 1997; Brown, 1987; Pressley, This chapter emphasizes the work on Borkowslu, & Schneider, 1987). In addition, metacognition within the area of adult many of the issues addressed in the area of memory research. It is organized primarily theory of mind (Perner & Lang, 1999) con- around the five main questions that have cern metacognitive processes. These issues been addressed in experimental research on are, perhaps, particularly important for the metamemory. First, what are the bases of understanding of children's cognition. metacognitive judgments; that is, how do In contrast, the experimental-cognitive we know that we know (e.g., Koriat & Levy- study of metacognition has been driven Sadot, lqqq)? Second, how valid are subjec- more by an attempt to clarify basic questions tive intuitions about one's own knowledge; about the mechanisms underlying monitor- that is, how accurate are metacognitive ing and