These notes are intended primarily as a resource for students working in my lab or taking my courses. I have attempted to provide concise summaries of theoretical concepts and themes important to my thinking. More detail on everything here can be found in my book, Experienced Cognition (Erlbaum, 1997; for a review of this book, see Hendrick, 1998). Others are welcome to use these notes for scholarly or scientific purposes, given appropriate citation. References to this page should use the following format:

        Carlson, R.A. (2000). Theoretical notes (July 2000 revision). Available: http://gandalf.la.psu.edu/rich/theory

Please address comments or questions to racarlson@psu.edu . There's a lot to scroll through here, so you can use these contents links to find particular sections. You can also go to a summary of current projects , where you will find some of the ideas expressed more straightforwardly in terms of their theoretical implications.

Note that this web site is not organized as a continuous document, though I've mostly kept everything on a single page for ease of printing. So don't expect logical flow between sections!

Thanks to the following for their comments (but don't blame them for anything here!): Marios Avraamides, Jodi Forlizzi, Nathan Lauffer, Andrew Peck, and Lisa Stevenson

What's here:

The cospecification hypothesis
Conscious mental states
Three aspects of consciousness:
    Subjectivity
    Agency
    Reflection
Emotion and experienced cognition
Frames of reference
Cognitive skill
Goal instantiation and conscious intentions
Temporal tuning
The ecological approach to perception
Normative models
Implicit learning
Is there an "unconscious mind?"
Recommended readings

The cospecification hypothesis

        The core idea of the experienced cognition framework is the cospecification hypothesis -- the idea that the experiencing self and the objects of experience are simultaneously specified in the arrays of information processed by the brain and nervous system. This hypothesis is an implication of the ecological view of perception advocated by James Gibson (e.g., 1966).

       The cospecification hypothesis in its most basic form is quite simple: The basic structure of consciousness consists in the fact that the information available from perception or memory specifies both objects of cognition and the point of view from which those objects are experienced. Take, for example, the information captured in a photograph or perspective drawing -- both the layout of the scene and the point of observation are conveyed by the information in the image. And it is the same array of information that conveys both -- self and object are cospecified. The static case suggested by this example is, of course, an idealization that we never experience in reality; our points of view are constantly changing, specifying different views of the environment. Even when standing still, eye movements ensure that our points of view change at a typical rate of several times per second.

        What makes the cospecification hypothesis general is the assumption that the cospecification of self and object is a general characteristic of cognition, regardless of the modality (visual, auditory, etc.) or source (environment or memory) of the information considered. Conscious experience occurs in a manifold of informational arrays, and always in the context of activity that results in constantly changing points of view.

        In the simplest case, the cospecific nature of conscious experience implies no self-awareness; that is, it is not essential to the basic idea of cospecification that any process discriminate between self-specific and object-specific aspects of the informational manifold. Typically, of course, experience does entail such a distinction, and some (e.g., Dennett, 1991) have argued that discriminating self and other is fundamental to consciousness. I call the consequence of such discrimination egolocation, locating the self. Egolocation is indeed central to conscious experience, though it is a matter of degree in several ways discussed in Experienced Cognition. To get a sense of how egolocation can be a matter of degree, contrast your conscious experience in circumstances in which you are extremely "self conscious" (perhaps because whatever you're doing is not going well) with your experience when you are "lost in" whatever you're doing (likely because that activity is proceeding very smoothly).

Conscious mental states

        An individual's conscious experience can be theoretically described as a sequence of conscious mental states that make up the stream of consciousness. In order to be scientifically useful and linked with contemporary theory in cognitive science, this theoretical description must be in terms of variables (Dulany, 1968, 1974, 1991), and in informational terms that allow the possibility of computational accounts of consciousness.

        Disclaimer: It is important to note that talking about the structure of a single conscious state in isolation is a theoretical abstraction. Consciousness is actually a systemic property of an organism with distal senses (such as vision) engaged in purposive activity in an information-rich environment. There may be a "quantum" character to conscious experience, such that individual states are of determinate duration and temporal boundaries. But due to the persistence of information availability in the environment, from one's body, and in memory, and the ecological grain size of conscious experience (and of psychologically appropriate descriptions of the environment), it makes sense to treat consciousness as a more-or-less continuous stream of experience, analyzed as a sequence of more-or-less discrete states, each lasting between approximately 50 milliseconds and 5 seconds.

FIGURE 1: The stream of consciousness (graphic by Nathan Lauffer)

        As the figure shows, the stream of consciousness might be described as path through the world, described for purposes of psychological theory in terms of affordances . In order to better understand the structure of individual conscious states, consider an expansion of a portion of the stream of consciousness showing the structure of an individual mental state. The diagonal lines in this image are simply meant as a reminder that we are looking at a portion of a stream of activity and experience that is extended in time. If you would like to see the two images side-by-side, click here .

FIGURE 2: The intentional structure of mental states

        This figure illustrates some of the intentional structure of conscious mental states. The experiencing self (the "I") and the content of awareness are informationally cospecified in the course of purposive action in the environment. Some of this action comprises resampling and exploration of the available information. An individual experiences a mental content in some psychological mode -- for example, believe , desire, see, intend, and so on. A mental content is about some object -- perhaps an object in the environment, perhaps a remembered or anticipated experience supported by memory. The mode and content of a mental state relate the individual to an object from some particular, momentary point of view, presenting some aspect of the object to awareness. To take a very simple example, if you see a tree, you do so from some point of view, seeing the aspect of the tree visible from that point of view.

        An important feature of this description of conscious mental states is that it allows description in terms of variables which can participate in lawful theoretical relations. For example, psychological mode may be characterized by degree of commitment -- how strongly, for example, we believe or desire something. Carlson and Dulany (1988) describes research on how conscious contents and degrees of belief are related in inferential processes for reasoning with circumstantial evidence. Another example of a variable is what I call intentional distance -- roughly speaking, the possibility of distinguishing content and object. A content represented as a verbal description has large intentional distance; it can be considered and manipulated in the absence of its object, and easily distinguished from the object. In contrast, in typical cases of visual perception, the content of awareness cannot be considered or manipulated in the absence of the visible environment, and there is little or no experienced distinction between content and object.

        In thinking about the properties of conscious mental states in terms of variables, it is important to note that only some of those variables describe the content (and might thus be called "conscious" in the sense that the values of the variables are experienced). Others, such as degree of commitment, are noncontent properties that characterize the intentional structure of a mental state. The values of these variables are experienced only in reflection on the mental states they describe. And it is often important to consider variables that do not describe the intentional structure of mental states directly at all; such implementation properties might include the power of a content to evoke associated contents, or the duration of the state. Failure to distinguish among these variables has led to a great deal of confusion in the literature on consciousness.

Three aspects of consciousness

Subjectivity

        The subjectivity of consciousness refers to its personal, idiosyncratic, and (largely) private nature. Each conscious individual has experience characterized by a unique sequence of points of view. "Points of view" here may be taken quite literally; as a number of authors have noted, one characteristic of an individual's stream of experience is that it comprises a temporally and spatially (more or less) continuous viewpath. Gibson (1966, 1979) and Neisser (1988, 1993) have noted that this viewpath is specified by the information available to perception; this is what Neisser has labeled the "ecological self," and is the basis for the cospecification hypothesis. The uniqueness of each individual's ecological self -- a straightforward consequence of the unique viewpath experienced by each individual -- is one aspect of the subjectivity of consciousness. The information that specifies one's ecological self is directly available only to oneself.

        A second aspect of the subjectivity of consciousness is the private nature of much of the information available to an individual. The perceptually-available information that specifies the ecological self is, in a sense, public -- someone can occupy a particular location occupied at another time by someone else, and thus infer much of what that other individual's conscious experience was (or would be) like. However, much of the informational support for one's conscious experience comes from memory, and we have relatively direct experience on the basis of only our own memories. Knowledge of someone else's memory is necessarily indirect and impoverished. Because all conscious experience is (at least as a reasonable hypothesis) informed by memory, consciousness is subjective in this sense. It is also important to note that a very similar observation can be made about the role of goals or intentions in conscious experience. Most of the time, we are doing something "on purpose," however trivial that purpose might be. Our intentions serve to structure our activity (again, that is at least a reasonable hypothesis), but can only be inferred by others. There is more to say about this in discussing agency, but it should be clear that our intentions are often private and subjective, as are their consequences for the structure and content of our conscious experience.

        A third aspect of the subjectivity of consciousness is often labeled by the philosopher's term qualia. Qualia are supposed to be the "raw feels" of conscious experience, "what it is like" to have experience with a particular content. For example, what it is like to see something red is said to be subjective and in principle private -- we have no way of knowing whether the experience that you and I both label as "red" is actually the same "from the inside."

Agency

        Agency is the experience of oneself as an active source of control of experience and activity. Conscious agency is realized through instantiated goals (conscious intentions). An instantiated goal is a mental state in the mode intend, with a content that specifies that some outcome is to be achieved by the self performing some action. That is, the state has the structure:

I intend that {x be accomplished by (my) doing y }

where the content of the state is the portion in {brackets}, the psychological mode is intend, and the self is specified as engaged in intentional action. Goal instantiation is discussed in more detail below.

        A second way of thinking about agency is to think about the acting self as specified in the course of action. All intentional activity is organized in terms of  frames of reference that organize the relations of actions to objects and their affordances. As discussed below , psychological reference frames, including familiar spatial frames, are procedural in the sense that what they organize is actions, or objects in relation to actions, rather than objects considered apart from the individual. The agent, or active self, is simply the origin of the current reference frame -- the "here and now" that serves as a reference point for coordinating activity and affordances. This idea shows us how the self can be seen as a causal locus, an origin of action, without a mysterious appeal to "free will" that assumes intentions are "uncaused causes." Harre and Gillett (1994) discuss a similar idea, though not at the level of detail required for cognitive theory. I am currently (summer 2000) working on a manuscript that elaborates this view of agency.

Reflection

        Many authors have argued (or perhaps more often, assumed) that reflection -- having thoughts about our thoughts and experiences -- is essential to consciousness. This assumption has often taken the form of equating verbal report with conscious awareness, and the inability to verbalize with a lack of conscious awareness. One current philosophical theory of consciousness, the HOT (Higher-Order Thought) theory (Rosenthal, 1993), asserts that the criterion of whether a mental state is conscious is whether or not one has a higher-order thought about that state. I agree that the capacity for reflection is important to a full account of human consciousness. However, I disagree with theories that identify either actual reflection or the possibility of reflection with the conscious status of a mental state. As discussed in Experienced Cognition, the foundational sense of consciousness is the experience of an organism engaged in purposive activity in an information-rich environment, experience that need not involve reflection. This is similar to Damasio's (1999) more recent concept of "core consciousness," which he argues is preserved even in individuals with severe brain damage that eliminates all linguistic competence or autobiographical memory.

        Of course, we often do reflect on our experience, constructing conscious states whose objects are other conscious states. Understanding our capacity for reflection, and using this capacity as the basis for experimental techniques for assessing consciousness (as, for example, in studies of implicit learning), requires that we consider both the bases for and constraints on that capacity. I assume that reflection is required for deliberate reports of awareness, so that everything in this section applies to both private reflection on and public report of the contents of awareness.  In particular, the following aspects of mental states are critical to the possibility of reflection (and thus of communication about conscious states):

Representational format and vocabulary

        We often think in words, literally if covertly talking to ourselves. When we do so, we can often tell others what we have told ourselves. Furthermore, we can often remember more or less precisely exactly what we have said to ourselves. However, much of our conscious experience is not represented verbally, and may be impossible to translate into, and extremely difficult to describe in, natural language. To use a striking example I've heard used by philosophers, what is the verbal translation of the experience of orgasm? Ericsson and Simon (1983) considered this issue years ago in the context of using verbal protocols ("talk aloud" and "think aloud" procedures) as data on cognition. They noted that verbal protocols are most likely to accurately reflect cognition when the content is already in verbal format (as when we talk to ourselves while doing mental arithmetic). If it is not, then the availability of appropriate vocabulary, and the possibility of reflection while formulating a verbal report, will be critical in determining the validity of the verbal protocol. And these considerations are just as important in talking to ourselves as in talking for the benefit of an experimenter.

        As argued in Experienced Cognition, I think that mental representation is performatory -- that is, the "vehicles" of content are performances such as private speech. Other performances include visual exploration (looking), gesture, and so on. Different performance modalities support different kinds of experienced contents, and skill level is critical in determining the content that can be represented in a particular kind of performance. In brief, we experience what we do, and what we do is far more than talk to ourselves.

Memory and capacity limitations

        Two kinds of observations are commonplace in the research literature: Limits on our ability to remember, and limits on the amount of information that can be considered in limited periods of time. If we make the simple and reasonable assumption that the mental processes involved in reflection are identical with or similar to mental processes that serve other purposes, it follows that these capacity limits constrain reflection. Consider, for example, the common practice of asking someone (or oneself), "What were you thinking while you did that?" Answering this question accurately requires accurate memory, and sufficient capacity to hold in mind what we've recalled while formulating an adequate description of it.

        The point here is really very simple -- conscious experience is informationally rich, and limits on memory and on processing capacity constrain both our ability to reflect on and our ability to report to others our conscious experience. Although there is a sense in which "unconscious" is sometimes used simply to mean "not recallable," we should be cautious about claims that this or that process is unconscious in some more interesting sense when those claims are supported by failures of retrospection.

The relativity of content ascription

        An important observation about the content of mental states and the information available to support them is that contents that refer to the same object may do so in a variety of ways. One example is the phenomenon of referential opacity in language. The idea here is just that two ways of talking about the same thing may not be interchangeable. For example, although one of the computers in my lab might be the oldest computer in the department, I might talk about "the computer in Room 633" and not realize that a member of the computer staff was referring to the same object when he mentioned "the oldest computer in the department" (and vice versa).

        A second example is what Searle (1992) has called the "aspectual shape" of mental contents. This simply means that the content of a mental state represents or describes only certain aspects of its object. For example, as I look at my computer monitor, I see only one possible visual aspect, the front of the monitor (the side which "faces" me). In the case of functional fixedness in problem solving, a problem solver sees only one possible function (one affordance) of an available object.

        A third type of example is discussed by Vallacher and Wegner (1987) in the context of their Action Identification Theory. The idea is that any activity can be described in multiple ways; for example, the behavior of pushing a doorbell button might be described as "seeing if someone is home" or "pressing a small metal button." They argue that individuals flexibly move among hierarchically organized alternative "identifications" of their actions, controlling activity on the basis of the highest level (most abstract and global) identity that works, moving down the hierarchy when necessary to control an activity that is encountering difficulty.

        What all of these examples point to is that ascribing content to an individual's conscious experience must be relative to that individual's moment-to-moment perspective. When you and I say we are "thinking about the same thing," that generally means (assuming we're correct at all!) that we're thinking about different aspects of the same object or event.

Emotion and experienced cognition

        Emotion is an important aspect of conscious experience, and the integration of theories of emotion and of cognition is a long-standing challenge. The standard approach to this integration has been to conceive of emotion (or the "feelings" that are its basis) as a particular kind of experienced content. For example, Damasio (1994, p. 159) wrote "feelings are just as cognitive as any other perceptual image."

        The experienced cognition framework suggests an alternative way of thinking about emotion. The basis of emotion is information specifying the state of the body (a statement with which almost every researcher would agree). However, according to my perspective, this information contributes to conscious experience not primarily by providing a basis for experienced contents, but instead by providing self-specific information that contributes to point of view from which objects are experienced, and thus the aspectual shape of experienced content. For example, if an individual is angry, he or she experiences events and objects under their annoying aspects. Of course we do have mental states whose content specifies our emotions, but such states are instances of reflection rather than definitive of emotional experience. This view explains how it can be that an individual can (for example) be angry and not realize it -- as in the common situation in which we realize that not only are we angry, but we've been angry for a while.

        The experienced cognition view of emotion also provides an account of the function of emotion. As many authors (e.g., Damasio, 1994) have noted, emotion seems critical to sustained goal-directed activity. There are two ways in which emotion is relevant to goal-directed activity: First, by providing salient self-specific information, the onset of emotion serves as a context for changing or establishing goals (which specify the self in purposive action). Second, because emotion is slow relative to cognition -- for example, emotions last at least many seconds, while we shift our visual attention several times per second -- it serves as a source of stability in the specification of self, helping to support sustained goal-directed activity. That is, we typically experience many rapidly-changing mental states from the "perspective" of a single, slowly-changing emotional state.

Frames of reference

        A frame of reference is usually described as a system for organizing locations and spatial relations among locations. For example, a set of orthogonal axes, an origin, and a metric for distances together identify a Cartesian frame of reference. The relations between two alternative frames of reference in this sense can be described by describing the relation between their origins, the orientations of their axes, and their metrics in terms of some common standard frame. Two individuals will have different egocentric frames of reference in this sense: the origins of the frames correspond to the locations of the two individuals in some larger space, and the up-down, front-back, and left-right axes of each frame can be identified with respect to the orientations of their bodies (again, relative to some larger space that contains both). Often a conventional metric (e.g., feet or meters) is assumed to characterize these reference frames. An advantage of this way of thinking about reference frames is that individual frames and the relations among alternative frames can be given rigorous mathematical descriptions.

        It is important to note, however, that it is an assumption or empirical hypothesis, not an a priori fact, that such Cartesian description of reference frames is appropriate for psychological theory. For adult vision in nearby space, the assumption is probably warranted -- light travels in pretty much straight lines, and individuals' pragmatic scaling of their activity to the environment is fairly accurate and stable. And for many theoretical purposes, it is the description of  "objective" space that is important. Psychological frames of reference, however, need not fit this description. For example, Piaget regarded the ability to use a Cartesian reference frame as a developmental achievement.

Cognitive skill

        Practice makes perfect. Or at least, almost anything we do improves with practice. Cognitive skills, like "physical" skills become more fluent -- smoother, more reliable, less effortful, and usually faster -- with practice. There is a large literature on the acquisition of cognitive skills, and much of my own research has focused on this topic. Here, I will just describe a few highlights that are especially relevant to understanding experienced cognition:

• Cognitive or mental skills have symbolic inputs and outputs -- for example, the skill of arithmetic is not defined by whether the operands and answer are spoken, written, or typed, but instead by their roles as symbols. This contrasts with perceptual-motor skills.
• Cognitive skills are typically multiple-step and cascaded -- that is, performance requires weaving together a series of steps realized by component skills, and the result of one step provides at least part of the input to a subsequent step.
• The learning curve for cognitive skills is "negatively accelerated" -- there are large improvements in performance over the first few practice trials, and while improvement continues for a long time with additional practice, the amount of improvement per trial gets smaller and smaller. This pattern is often captured very well by a mathematical relationship known as a power function, though there is some debate about this (Heathcote, Brown, & Mewhort, 2000).
• Improvements in performance result from three kinds of mechanisms: restructuring, component speedup,  and temporal tuning. Restructuring means shifting to more efficient procedures, such as retrieving an addition fact rather than "counting on." Component speedup just means that individual steps become faster with practice. Temporal tuning is discussed below .
• Perhaps most important in the present context, practice changes the way that we experience and control our skilled performance, and increasing skill changes the ways we can experience the world -- the world affords different experiences for individuals with different skills. Consider, for example, the experience you are having as you gaze at this page; it is quite different than the experience of an individual who cannot read English! Or, think about any skill you've acquired -- for example, your experience of surfing the Web or using a word processor is quite different than it was the first time you tried it.

Goal instantiation and conscious intentions

        Agency is realized conscious mental states that instantiate goals by representing the self as achieving outcomes by intentional action. One implication of this statement is that we need to distinguish two senses of the term goal. The first sense is that of a desired outcome or state of affairs; in this sense, one may have goals that are not currently functional in controlling activity. The second sense is more limited, referring to instantiated goals that are currently functional in controlling activity. Goals in this second sense always include means -- actions intended to achieve particular outcomes. A conscious state that instantiates a goal has the structure: I intend that {outcome x be accomplished by my action y} where the psychological mode is intend and the content of the state is {outcome x be accomplished by my action y}. This second sense of goal is fairly close to the sense of the term in John Anderson's (e.g., 1993) ACT-R theory of cognition. Note that an acting self is specified in the content of the mental state, as well as in the state as a whole; in this sense, conscious intentions are necessarily self-referential.

        My students and I have proposed that the control of activity by conscious intentions has a characteristic information-processing structure, summarized by the GOER model. The GOER model simply asserts that each "step" of mental activity begins with Goal instantiation, which evokes a procedural frame to which one or more Operands are assimilated. This is sufficient to initiate Execution of a mental process realized by nonconscious information processes such as retrieval. The Result of the mental process then becomes available to serve as part of the content of a subsequent mental state, perhaps as an operand for a subsequent mental operation. The GOER model is intended not as a computational model but as a mnemonic for a set of hypotheses about the operation of conscious intentions. One hypothesis that we have tested extensively (e.g., Carlson & Sohn, in press; Sohn & Carlson, 1998) is the procedural frame hypothesis, which asserts that individuals will perform mental activities more fluently when information about operators is available before information about operands, so that a goal can be instantiated and a procedural frame evoked in order to most effectively process those operands. A second hypothesis, less directly tested to date, is the overlap hypothesis : only one goal can be instantiated at a time (corresponding to the serial cognition assumption of ACT-R), but the next goal can be instantiated during the execution phase of the current step. A third hypothesis, for which we are beginning to accumulate evidence, is the temporal tuning hypothesis , discussed next.

Temporal tuning

        A principle of cognitive control implicit in both common sense and most psychological theory is the juxtaposition principle. The idea is pretty simple; cognitive activity results from the temporal juxtaposition -- "having in mind" more or less simultaneously -- of appropriate mental contents experienced in appropriate modes. Here's a very simple example: to do mental arithmetic, you must have in mind close in time the intention to perform a particular operation (e.g., to add) and the operands (e.g., "2" and "7") on which the operation is to be performed. Of course, the juxtaposition principle is not a discovery or unique feature of the experienced cognition framework; instead it is a label for one of those things that "everybody knows." However, making the juxtaposition principle explicit and thinking about its implications in conjunction with the analysis of conscious intentions leads to some interesting hypotheses, the procedural frame and temporal tuning hypotheses.

        The procedural frame hypothesis says this: an instantiated goal evokes a procedural frame to which operands are assimilated in order to carry out mental activity. For example, the goal "to add" evokes a procedural frame to which numbers are assimilated, resulting in retrieval or calculation (e.g., by counting) of a sum. The major prediction of the procedural frame hypothesis is that performance of mental skills will be more fluent when an individual has the opportunity to instantiate a goal before considering the operands to which that goal applies. Evidence for this hypothesis is presented in Sohn and Carlson (1998) and Carlson and Sohn (in press).

        The temporal tuning hypothesis follows from the procedural frame hypothesis and the juxtaposition principle. It asserts that in acquiring skills, individuals learn to adjust the timing of various aspects of their performance (e.g., the pickup of information from the environment) such that temporal juxtaposition becomes more precies -- that is, they learn to pick up information so that it is available "just in time." Carlson and Stevenson (in preparation) and Stevenson (in preparation) present evidence that individuals do in fact temporally tune their performance in learning multiple-step arithmetic tasks.

The ecological approach to perception

        The ecological approach refers to a set of insights and assertions developed by James Gibson (1966, 1979). Gibson argued that perception is direct, unmediated by mental activity, rather than inferential, and much discussion of the ecological approach has focussed on this contrast. However, several other aspects of Gibson's thinking are more important to the experienced cognition framework. First, Gibson (1960) recognized early on that the nature and level of description chosen for the information available to perception is crucial in determining the nature of a theory of perception. He argued that the features of the environment important for understanding perception are affordances, aspects of the environment that allow and support particular activities. Second, he argued that the information for perception must be considered not as symbolic (on the model of language) or in terms of resemblance (on the model of pictures), but in terms of specificity, invariant relations between the patterning of energy picked up by perception and features of the world (described as affordances). Third, he argued that the function of perception is the guidance of purposive activity, not the presentation of images or descriptions of the world to an observer. Finally, Gibson noted that the information available to perception specifies both surfaces and objects in the world and the viewpoint of the individual, an insight that forms the basis of the cospecification hypothesis.

        One important aspect of the ecological approach is that it provides principled guidance for choosing the grain size at which both behavior and the environment should be described for purposes of psychological theory. Although few if any cognitive scientists would argue that the "input" for perception should be described in terms drawn directly from theories of physics,

        There is, of course, much more to say about the ecological approach, and it has generated a large literature.

Normative models

        An important, but underappreciated, aspect of the study of psychological processes is that it almost always depends on normative models of those processes. A normative model is simply a set of hypotheses about how a process ought to work, or what its outcome ought to be. For example, logic provides a normative model for some kinds of verbal reasoning, arithmetic provides a normative model for some kinds of numerical reasoning, and the several varieties of utility theory provide competing normative models for some kinds of decision making. Perhaps less obviously, much research on memory has depended on an implicit normative model of remembering that prescribes accurate, literal reproduction as "correct" memory.

       Two points about normative models are especially critical. First, the interpretation of research results almost always depends on a normative model, and making that model explicit can be important for evaluating specific interpretations. In many cases, both the appropriateness of particular models and the mapping of those models to specific situations can be challenged (e.g., Cohen, 1981; Glenberg, 1997; Oaksford & Chater, 1994). For example, a number of authors have recently argued that accurate, literal reproduction is not the appropriate normative model for memory (e.g., Glenberg, 1997). Second, it may be that some normative considerations are essential to any definition of mind or mental activity. For example, Harre' and Gillett (1994) argued that normative considerations are central to all psychological analysis, and Cherniak (1986) argued that "minimal rationality" is essential to the possession of mind. Although there are a number of subtleties and controversies here, one way to understand the point is this: The computational view of the mind that dominates current cognitive science proposes that mental activity be understood as the manipulation of mental representations, and such manipulation must be coherent and "truth preserving" if mental activity is to serve an adaptive function in the survival of organisms (Haugeland, 1985).

Implicit learning

        A very hot topic in recent cognitive psychology is implicit learning (and its near relative, implicit memory). Many authors have taken the phenomena studied in research on implicit learning as evidence of a powerful domain of unconscious cognitive processes; for example, Damasio (1999), in his otherwise careful analysis of consciousness, uncritically cites this literature as evidence of unconscious processing. Certainly the literature on implicit learning documents some important phenomena -- learning without an intention to do so, effects of learning and memory on current activity without a corresponding ability to attribute those effects to specific past experiences (for example, in the case of memory, absence of an experience of "recollection), and acquired skills that can be demonstrated in some performances (e.g., sequences of keypresses) but not others (e.g., verbalization of sequential regularities). Stadler and Frensch (1998) have edited an important volume that collects a variety of perspectives on these phenomena and their explanation.

        The debates over implicit learning echo an earlier literature on "learning without awareness" and the recurring topic of so-called "subliminal perception" (Brewer, 1974; Holender, 1986; Shanks & St. John, 1994). Some of my own earliest work is on this topic (Carlson & Dulany, 1985; Dulany, Carlson, & Dewey, 1984). Over the years, I have come to believe that although this literature contains some important observations about learning, it tells us little or nothing about consciousness (Carlson, 1991, 1994). Only very rarely have authors considered exactly what conscious content would, given reasonable process assumptions, be sufficient to produce the observed evidence of learning. I devoted a chapter of Experienced Cognition to this topic, but the bottom line for me is this: Consciousness is intimately linked with learning, and our moment-to-moment, "on line" experience while learning or performing acquired skills may be bound to specific situations or activities. However, the literature on implicit processes does not provide evidence that learning can occur without what Damasio (1999) has called "core consciousness" -- the immediate awareness of content that captures what is being learned.

Is there an "unconscious mind?"

        No. Many psychologists and laypeople believe that a large and powerful domain of mental activity is "unconscious." I reject this idea; indeed, I think that in an important sense, there are no unconscious mental processes . What does that mean, and how can I make such an assertion when the research literature contains so many demonstrations of so-called "unconscious" phenomena?

        First, we have to think about what it means for something to be "unconscious." I believe that the most common idea people have in mind when describing psychological phenomena as unconscious is this: there are mental states that are like conscious mental states but that do not participate in constituing the stream of subjective experience. That is, you have these states, but you don't experience them. This idea implies that we have unconscious beliefs, desires, intentions, and so on -- what is sometimes called the "deep unconscious." At least some of the time, this seems to be more or less what Freud had in mind when he described the unconscious (see Erdelyi, 1985, 1992, for an analysis of this sense of the unconscious in contemporary terms). There are also weaker senses of "unconscious" that are fairly commonly held. For example, some authors describe as "unconscious" any process that does not involve deliberate consideration of alternative courses of action. Others describe as unconscious any mental state that does not become the object of reflection or metacognition. And some describe any neural computational or information-processing activity that cannot be introspected as unconscious mental activity.

        Second, many of the empirical "demonstrations" of unconscious processes are methodologically flawed. These flaws have been discussed by many authors; Holender (1986) and Shanks and St. John (1994) and the accompanying commentaries provide useful overviews. The best of the efforts to demonstrate unconscious processes can be interpreted as effects of conscious contents that are experienced only fleetingly, and do not "hold still" long enough for the formation of autobiographical memories or appropriate attributions of the sources of conscious experiences or behavior.

        Third, as William James (1890, p. 107) argued in a frequently misread passage,

The distinction is that between the unconscious and conscious being of the mental state. It is the sovereign means for believing what one likes in psychology, and of turning what might become a science into a tumbling-ground for whimsies. It has numerous champions, and elaborate reasons to give for itself.  [italics in original]

James follows this passage by considering ten "proofs" of the unconscious, and reasons for rejecting them. Many of these arguments have contemporary counterparts, and many of James' objections hold against modern versions as well. The central point, I think, is this: postulating "unconscious" states whose explanatory force comes from analogies to conscious cognition (a) eliminates powerful sources of evidence and plausibility, (b) begs the importation of "extra" causality such as a homunculus, and thus (c) allows the survival of unsupported views of consciousness and its role (or lack thereof) in cognition.

    Why comment on this question? Two reasons: First, uncritical acceptance of the idea of an unconscious mind can distort psychological theory and research, inhibiting the careful and difficult examination of experienced cognition. Second, many students and others believe in such discredited phenomena as powerful "subliminal" influences, creating suspicion and distorted views of psychology.

References

This link provides the list of specific references cited on this page. Please let me know if you find anything missing!

Recommended readings

Students often ask me what they should read to learn about my work and related literatures. The best single source is probably my book, Experienced Cognition (not that I'm biased or anything :)). All of my publications are listed on my on-line vita ; the commentaries on other people's articles that I've published in American Journal of Psychology and Behavioral and Brain Sciences are probably the most concise statements of some of my theoretical views.

Here are some readings that I think are most useful on particular topics:

Consciousness

     Cohen, J. & Schooler, J. (Eds.), Scientific approaches to the study of consciousness . Mahwah, NJ: Erlbaum.
    Dennett, D.C. (1991). Consciousness explained. Boston: Little, Brown and Company.
    Dennett, D.C. (1996). Kinds of minds. New York:  BasicBooks.
    Dulany, D. E. (1968). Awareness, rules, and propositional control: A confrontation with S-R behavior theory. In T. Dixon & D. Horton (Eds.) Verbal behavior and general behavior theory.  (pp. 340-387). Englewood Cliffs, NJ: Prentice-Hall.
    Dulany, D. E. (1991). Conscious representation and thought systems. In R.S. Wyer, Jr.& T.K. Srull (Eds.) Advances in social cognition, Vol. 4 (pp. 97-117). Hillsdale, NJ, Erlbaum.
    Dulany, D. E. (1996). Consciousness in the explicit (deliberative) and implicit (evocative). In J. Cohen & J. Schooler (Eds.), Scientific approaches to the study of consciousness (pp. 179-212). Mahwah, NJ: Erlbaum.
    Dulany, D. E. (1974). On the support of cognitive theory in opposition to behavioral theory: A methodological problem. In W.B. Weimer & D.S. Palermo, (Eds.), Cognition and the symbolic processes (pp. 43-56). Hillsdale, NJ: Erlbaum.
    Ericsson, K. A., & Simon, H. A. (1980). Verbal reports as data. Psychological Review, 87, 215-247.
    Neisser, U. (1988). Five kinds of self-knowledge. Philosophical Psychology, 1, 35-59.
    Neisser, U. (1993). The self perceived. In U. Neisser (Ed.), The perceived self: Ecological and interpersonal sources of self-knowledge (pp. 3-21). Cambridge: Cambridge University Press.
    Searle, J. R. (1983). Intentionality: An essay in the philosophy of mind. Cambridge: Cambridge University Press.
    Searle, J. R. (1990). Consciousness, explanatory inversion, and cognitive science. Behavioral and Brain Sciences, 13, 585-642.
    Searle, J. R. (1992). The rediscovery of the mind. Cambridge, MA: The MIT Press.

Cognitive skill

    Anderson, J.R. (1982).  Acquisition of cognitive skill.  Psychological Review,  89, 369-406.
    Anderson, J.R. (1996).  ACT: A simple theory of complex cognition. American Psychologist, 51, 355-365.
    Anderson, J.R. (1993). Rules of the mind. Hillsdale, NJ:  Erlbaum.
    Anderson, J.R. (1987).  Skill acquisition: Compilation of weak-method problem solutions. Psychological Review,  94, 192-210.
    Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100 , 363-406.
    Ericsson, K. A., & Smith, J. (Eds.). (1991). Toward a general theory of expertise.  Cambridge: Cambridge University Press.
    Logan, G. D. (1988). Toward an instance theory of automatization. Psychological Review, 95, 492-527.
    MacKay, D. G. (1982). The problems of flexibility, fluency, and speed-accuracy trade-off in skilled behavior. Psychological Review, 89, 483-506.
    Neisser, U. (1983). Toward a skillful pychology. In D. R. Rogers, & J. A. Sloboda (Eds.), Acquisition of symbolic skill (pp. 1-17). New York: Plenum.
    Newell, K. M. (1991). Motor skill acquisition. Annual Review of Psychology, 42, 213-237.
    Norman, D. A. (1981). Categorization of action slips. Psychological Review, 88, 1-15.
    Proctor, R. W., & Dutta, A. (1995). Skill acquisition and human performance. Thousand Oaks, CA: Sage.
    Schneider, W., & Detweiler, M. (1988). The role of practice in dual-task performance: Toward workload modeling in a connectionist/control architecture. Human Factors, 30, 539-566.

Working memory

    Baddeley, A.D. (1986). Working memory. Oxford:  Oxford University Press.
    Cowan, N. (1993).  Activation, attention, and short-term memory. Memory & Cognition,  21, 162-167.
    Miyake, A. & Shah, P. (Eds.), Models of working memory: Mechanisms of active maintenance and executive control. Cambridge, UK: Cambridge University Press.
    Reisberg, D., Rappaport, I., & O'Shaughnessy, M. (1984). Limits of working memory: The digit-digit span. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 203-221.

Spatial cognition

    Eilan, N., McCarthy, R. A., & Brewer B. (Eds.), Spatial representation: Problems in philosophy and psychology.  Oxford: Blackwell.
    Bryant, D. J., Tversky, B., & Franklin, N. (1992). Internal and external spatial frameworks for representing described scenes. Journal of Memory and Language, 31, 74-98.
    Byrne, R. M., & Johnson-Laird, P. N. (1989). Spatial reasoning. Journal of Memory and Language, 28(5), 564-575.
    Carlson-Radvansky, L. A., & Irwin, D. E. (1993). Frames of reference in vision and language: Where is above? Cognition, 46, 223-244.
    Cutting, J. E. (1996). Wayfinding from multiple sources of local information in retinal flow. Journal of Experimental Psychology: Human Perception and Performance, 22, 1299-1313.
    Farah, M. J., Brunn, J. L., Wong, A. B., Wallace, M. A., & Carpenter, P. A. (1990). Frames of reference for allocating attention to space: Evidence from the neglect syndrome. Neuropsychologica, 28 , 335-347.
    Farrell, M. J., & Robertson, I. H. (1998). Mental rotation and the automatic updating of body-centered spatial relationships. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24, 227-233.
    Gibson, E. J., & Schmuckler, M. A. (1989). Going somewhere: An ecological and experimental approach to development of mobility. Ecological Psychology, 1, 3-25.
    Huttenlocher, J., & Presson, C. C. (1979). The coding and transformation of spatial information. Cognitive Psychology, 11, 375-394.
    Klatzky, R. L., Loomis, J. M., Beall, A. C., Chance, S. S., & Golledge, R. G. (1998). Spatial updating of self-position and orientation during real, imagined, and virtual locomotion. Psychological Science, 9, 293-298.
    Liben, L. S. (1988). Conceptual issues in the development of spatial cognition. In J. Stiles-Davis, M. Kritchevsky, & U. Bellugi (Eds.), Spatial cognition: Brain bases and development (pp. 167-194).  Hillsdale, NJ: Erlbaum.
    Logan, G. D. (1995). Linguistic and conceptual control of visual spatial attention. Cognitive Psychology, 28, 103-174.
    McNamara, T. P. (1986). Mental representations of spatial relations. Cognitive Psychology, 18, 87-121.
    Pick, H. L., Montello, D. R., & Somerville, S. C. (1988). Landmarks and coordination and integration of spatial information. British Journal of Developmental Psychology, 6, 372-375.
    Presson, C. (1982). Strategies in spatial reasoning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 8 , 243-251.
    Presson, C. C., & Montello, D. R. (1994). Updating after rotational and translational body movements: Coordinate structure of perspective space. Perception, 23, 1447-1455.
    Price, C. M., & Gilden, D. L. (2000). Representations of motion and direction. Journal of Experimental Psychology: Human Perception and Performance, 26, 18-30.
    Pylyshyn, Z. (1989). The role of  location indexes in spatial perception: A sketch of the FINST spatial-index model. Cognition, 32, 65-97.
    Rieser, J. J. (1989). Access to knowledge of spatial structure at novel points of observation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 1157-1165.
    Shelton, A. L., & McNamara, T. P. (1997). Multiple views of spatial memory. Psychonomic Bulletin & Review, 4, 102-106.
    Sholl, M. J. (1995). The representation and retrieval of map and environment knowledge. Geographical Systems, 2, 177-195.
    Taylor, H. A., & Tversky, B. (1992). Spatial mental models derived from survey and route descriptions. Journal of Memory and Language, 31, 261-292.
    Thorndyke, P. W., & Hayes-Roth, B. (1982). Differences in spatial knowledge acquired from maps and navigation. Cognitive Psychology, 14, 560-589.
    Tversky, B. (1991). Spatial mental models. In G. H. Bower (Ed.), The psychology of learning and motivation, Vol. 27 (pp. 109-145). New York: Academic Press.
    Warren, R. (1976). The perception of ego motion. Journal of Experimental Psychology: Human Perception and Performance, 2, 448-456.
    Wraga, M., Creem, S. H., & Proffitt, D. R. (2000). Updating displays after imagined object and viewer rotations. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 151-168.

Ecological approach to perception

    Gibson, J. J. (1960). The concept of the stimulus in psychology. American Psychologist, 15, 694-703.
    Gibson, J. J. (1966). The senses considered as perceptual systems. Boston: Houghton-Mifflin.
    Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton-Mifflin.
    Neisser, U. (1976). Cognition and reality. San Francisco: Freeman.