Subjectivity and the limits of scientific enquiry
Jordi Fernandez (Jorge_Fernandez@Brown.edu)
Philosophy Department Box 1918
Providence, RI 02912
In 'The view from nowhere', Thomas Nagel argues for the thesis that no objective description of the world can be complete. An objective view of the world, Nagel argues, cannot account for who its subject is and it is consequently incomplete. The fact to be explained is presumably the fact that a particular person who is described in the objective view is the subject of it as well. Given that a scientific picture of the world is meant to be objective, an important point involved in this discussion is that subjectivity cannot be scientifically accounted for. I will try to reconstruct Nagel's argument for the incompatibility of objectivity and completeness and raise some objections to it. My intention is to show that Nagel's thesis owes its philosophical significance to an ambiguous notion of completeness Nagel is working with. I shall distinguish two plausible senses in which a picture of the world can be said to be incomplete, that I will call an "epistemic" and "metaphysical" sense. My suggestion will then be that either Nagel's reasons in support of his thesis are insufficient (under its metaphysical reading), or the thesis in question lacks most of the significance it initially seems to posses (if it is epistemologically read).
On the Search for the Neural Correlate of Consciousness: Some Caveats
GÃ¼ven GÃ¼zeldere (email@example.com)
Department of Philosophy &
Center for Cognitive Neuroscience
201 West Duke Building
Durham, NC 27708
I argue that the search for the neural correlate of consciousness (NCC) may be a misguided research effort. I argue that if the NCC is construed on the basis of the non-modularist hypothesis, it is too general to constitute a genuine research program, and if it is based on the modularist hypothesis, the debate on how to identify the real NCC will never successfully get resolved because not only is there no NCC, in this specific sense, but there cannot be. Regarding the second horn, my claim is that the only sense in which search for the NCC can constitute a well-defined research program is committed to a misguided concept of consciousness. What can make this question meaningful is a "modular" conceptualization of consciousness such that it can be isolated in its function and singled out in its anatomical underpinnings. This is a misconceptualization, because consciousness is not dissociable from other mental functions in this way. I argue, therefore, that there exists no piece of localizable tissue in the brain the lesioning of which will destroy only consciousness, and leave the non-conscious organism otherwise indistinguishable in its other mental functions and behavior, e.g., memory and language.
Redcar Rocks: Strong AI and Panpsychism
J.M. Bishop (J.M.Bishop@Reading.ac.uk)
Deptartment of Cybernetics
University of Reading
The claimed observer relativity of computational states forms the basis of both Putnams 1988 attack on Functionalism and Searle's 1990 attack on Computationalism. The argument to be presented herein, being a simple extension of that originally given by Putnam, is not significantly original but appears to foil the main criticisms of Putnam and Searle's approach (see Chalmers et al., 1994), and hence has critical implications for our understanding of consciousness. In this paper, instead of seeking to emulate Putnam's claim that, everything implements every Finite State Automata (FSA), I will simply establish the weaker result that everything implements the specific FSA [Q], when executing program (p) on input (x). Then, equating Q(p,x) to an AI program passing the Turing Test, I will show that conceding the Strong AI thesis for Q (crediting it with mental states and consciousness) opens the door to a vicious form of panpsychism whereby all open systems, even rocks and cups of tea, have conscious experience.
Free will and the readiness potential
Gilberto Gomes (firstname.lastname@example.org)
R. Lopes Quintas 100-605-I
22460-010 Rio de Janeiro
The readiness potential precedes voluntary acts by about half a second. According to Libet, free will does not initiate the neural process that leads to action but is able to control it. While disagreeing with many points of his interpretation of results, we should agree that voluntary acts are nonconsciously initiated. Voluntary acts are felt to have been determined by a conscious decision. This seems to conflict with the idea that all physical events are caused by other physical events. However, choice, decision and action can be considered as part of the natural world. All we need to assume is a decision system that can represent actions before their performance and select them according to its internal state. Free will is not an illusion because free acts are not caused by external factors. From the first-person perspective, I am the cause of my actions. But what am I? According to compatibilism, the free agent is a brain system capable of choice, decision and action. The readiness potential will be seen as an expression of it. We should distinguish the intention to act in the future, the intention to act now and the irrevocable decision to act now. This causes the action before we become conscious of it. A distinction is proposed between deliberate and non-deliberate voluntary acts. A testable prediction is that the RP should be longer in the case of deliberate actions. Non-deliberate voluntary acts manifest an intermediate degree of free will, since they and the possibility of doing otherwise were not consciously considered before starting their performance.
The role of binding in the brain and of correspondences in theorizing
P.H. de Vries
P.H. de Vries (email@example.com)
Deptartment of Psychology
University of Groningen
Grote Kruisstraat 2/1
9712 TS Groningen
At the structural or neural level, binding refers to processes of the creation of temporary connections between different populations of neurons. At the functional level these temporary connections provide in the system an integrated representation of its environment. At the level of subjective experience the process of binding is likely to be a necessary condition for the occurrence of a conscious experience. The fundamental problem in the study of consciousness is then to find the correspondences between these levels of description. A general architecture for cognitive brain functioning in terms of a conceptual network will be presented which is based on these correspondences. The process of binding is conceived as the effectuation of temporary connections in a network of cell- assemblies, which allows for the formation of the concept of, e.g., 'object A at position P'. Binding is also a necessary mechanism for relating procedural knowledge and declarative knowledge. The functioning of the network is based on the selective, context- dependent propagation of excitation loops along permanent and temporary connections. A computer simulation of the network will be discussed in relation to new experiments on object-identification and illusory conjuctions.
Temporal Synchronization: A Possible Mechanism for the Binding Together of the Conscious Self
Logan Trujillo (firstname.lastname@example.org)
Department of Psychology
University of Arizona
Tucson, Arizona, 85721
Current neurobiological approaches to consciousness have proposed that rudimentary self-consciousness may arise out of the binding together of neural maps with first-order representational properties to form second-order mappings. However, the mechanism for such binding, both at the first- and second-order level, is unclear. It has been suggested that temporally synchronized oscillatory neural activity may play an important role in the binding together of cognitive-perceptual events; an idea for which there is an increasing amount of empirical evidence. This presentation will address the possibility that synchronous oscillatory behavior may subserve the binding together of neural representations associated with the conscious self. It will be proposed that short- and long-range intramodal synchronies bind together the neural elements constituting first-order mappings. The binding together of these maps into the second-order representations underlying self-consciousness would then be achieved by short- and long-range intermodal synchronies. This framework has implications for the search for the neural correlates of consciousness (NCCs) in that such second-order representations may or may not be localized to particular cortical regions. In addition, this presentation will explore the potential usefulness of the temporal synchronization hypothesis towards understanding the etiology of, and providing a means of clinical assessment for, psychological dissociative disorders where the conscious self is fragmented in a maladaptive manner.
The temporal binding problem: what it is and how it might be solved
David M. Eagleman & Terrence J. Sejnowski
David M. Eagleman (email@example.com)
Computational Neurobiology Lab
The Salk Institute
10010 N. Torrey Pines Rd.
La Jolla, CA 92037
While it is clear that different features of stimuli are processed in different areas of the brain, it is underappreciated that information is processed in these areas at widely varying times. Yet somehow perception retains very exact temporal information about outside events. This leaves us with the surprising result that physiologically-measured latency differences do not generally translate into perceptual time differences. This is what we define as the temporal binding problem. In other words, how do the widely varying stimulus-evoked latencies in the neural tissue become temporally aligned to yield the coherence of perception? Concentrating on the visual system, we employ physiological and psychophysical data to argue that the features of awareness necessitate a window of delay and postdiction, the act of retrospectively attributing an interpretation to events in the past. We demonstrate that postdiction is the only framework that provides a unified explanation for many psychophysical phenomenon.
Cortical MEG activity dissociates coherence and meaning
Marina Pavlova, Werner Lutzenberger, Alexander Sokolov & Niels Birbaumer
Marina Pavlova (firstname.lastname@example.org)
Institute of Medical Psychology and Neurobiology
Although current research views the high-frequency cortical oscillations as subserving binding mechanisms for processing of coherent stimuli (Bertrand and Tallon-Baudry, 1999; Singer, 1999), it remains unclear whether brain responses in the gamma frequency range vary with stimulus coherence by itself or also with a meaningful representation of a coherent structure. By manipulating task demands, we have demonstrated that only an attended pattern of visual motion leads to an increase in gamma MEG activity (Sokolov et al., 1999). Most recently, however, we showed that when a task requires attention to both a coherent stimulus and a similar incoherent noise, only coherent patterns elicit enhancements in the gamma response. Furthermore, it appears that the early MEG gamma response (80-100 ms) over the primary visual cortex exhibits a sensitivity to the stimulus coherence regardless of perceptual awareness, while later consecutive enhancements over parietal and temporal areas reflect the meaningful representation of a coherent structure from motion. Such evidence helps to clarify the ongoing debate about the neural correlates of consciousness (Crick and Koch, 1998; Logothetis, 1998; Singer, in press).
Perceptual Filling-in of DarknessMichael E. Rudd
Michael E. Rudd (email@example.com)
University of Washington
Department of Psychology
Seattle, WA 98195-1525
A growing body of evidence suggests that the visual brain computes surface color in a multistage process that includes 1) an early neural encoding of color contrast at the locations of borders between regions of homogeneous luminance within the retinal image and 2) a subsequent filling-in of the colors belonging to regions lying within or between the borders. I will present a computational model of brightness (achromatic color) based on this type of multistage neural algorithm. The model differs significantly from earlier filling-in models by emphasizing the importance of the filling-in of darkness signals, as opposed to lightness signals. The behavior of the model will be illustrated by showing how it accounts quantitatively for the magnitudes of darkness induction effects exerted by regions of high luminance within the visual scene on the brightness of a test region of lower luminance, as measured by a psychophysical brightness matching technique (Rudd & Arrington, ARVO, 2000).
Priming the kinetics of pointing movements: Online-control by barely visible isoluminant color stimuli
Thomas Schmidt (firstname.lastname@example.org)
University of Goettingen
Institute of Psychology
Two-visual-systems theory (Milner & Goodale, 1995) states that color processing is generally separated from motor control in the primate brain. In order to use color information in a motor task, one would have to rely on color as consciously experienced, which would be very difficult if stimuli are efficiently masked. Here, a motor priming paradigm (Vorberg, Mattler, Heinecke, Schmidt, & Schwarzbach, in prep.) was used where participants were required to make a speeded pointing response toward the one of two target stimuli having a prespecified color. Targets were preceded by primes at the same positions having the same (consistent) or reversed (inconsistent) colors as the targets. Due to masking by the targets, discriminability of primes was uniformly low independent of prime-target delay. In spite of this, pointing movements started off in the direction specified by prime rather than target color, with a time course closely locked to prime onset. When the target arrived, this movement was either maintained when primes and targets were consistent or had to be reversed when they were inconsistent, with a time course now locked to target onset. These results strongly suggest a direct dynamic link between color processing and response control.
Can Synesthesia be explained by 40Hz Oscillations?
Lucy J. Troup
Lucy J. Troup (email@example.com)
Department of Psychology
University of the West of England, St Matthias Campus
Oldbury Court Road, Fishponds,
Bristol. BS16 2JP
Synesthesia has been explained as a break down of modularity in perceptual processes. What is unclear is the nature of the processes that elicit synesthetic experiences. Interestingly research in developmental psychology suggests that neonates are in fact synesthetes and modularity of perceptual processing evolves over time. In the last 10 years there has been a great deal of research implicating 40hz oscillations as a possible mechanism for perceptual processing. It has more controversially been implicated as a possible mechanism for visual binding. Computer modeling work has suggested that 40hz oscillations are mediated by feedback connections in artificial oscillatory neural networks. Interestingly there is research that suggests that visual system in the neonate is far from fully developed. In fact it is not until the neonate is a year old that its feedback pathways in the brain begin to develop. Is it possible that this, coupled with the notion that cognitive processing, namely perceptual processing, is in fact explained by 40hz oscillations, could be a possible mechanism for synesthesia? Is it that synesthetes have some kind of dysfunctional or inappropriate feedback pathways causing synchrony of 40hz oscillations in inappropriate "perceptual modules"?
Digit-Colour Synaesthesia: An Investigation of Extraordinary Conscious Experiences
Daniel Smilek, Mike J. Dixon, Cera Cudahy & Philip M. Merikle
Daniel Smilek (firstname.lastname@example.org)
Department of Psychology
University of Waterloo
Digit-colour synaesthesia occurs when a black digit elicits a conscious experience of a highly specific colour (a photism). It has been suggested that photisms differ from imagery in their consistency, their automaticity and their reliance on externally presented stimuli for elicitation. We tested C, a digit-colour synaesthete, to evaluate these claims. The consistency of C's photism colours for each digit (0-9) was evaluated by having her name the colour of each digit. C named the same colour for each digit over 10 repetitions. To assess the automaticity of C's photims, digits were displayed in colours congruent or incongruent with her photism colours. C's colour naming reaction times were significantly slower for incongruent trials than congruent trials. Finally, to evaluate whether an external stimulus was necessary to elicit C's photisms, C was shown arithmetic problems (e.g., 5+2) followed by a colour patch that she had to name. Naming times were slower when colours were incongruent with C's photisms for the answer to problems than when the colour patches were congruent. We conclude that the photisms experienced by C are consistent, automatic and can be induced simply by activating the concept of a digit in the absence of an external stimulus.
When are we conscious? Some thoughts on a seemingly uncontentious topic
Rimas Cuplinskas (email@example.com)
University of Bonn
Am Hof 1
Precisely what scientists from the different fields studying the human mind mean by the term "consciousness" is less clear than it may seem. Although philosophers are in general agreement that being conscious involves a subject being in a phenomenal state, such states are strictly speaking not directly accessible from a third-person perspective. For this reason the empirical sciences have had to work with other operational definitions according to which one might infer whether or not a subject is conscious of a given stimulus. Bernard Baars (1997), in attempting to define more clearly the differences between our concepts of consciousness, attention, perception and working memory, has suggested "accurate reportability" as an operational definition for consciousness. However, if we posit the existence of a short-term sensory buffer for visual and acoustic stimuli, the contents of which may if necessary be consciously reviewed, then it would both be true that (i) one was not conscious of a given stimulus at the time of perceiving it, and that (ii) one is nevertheless able to accurately report on it when cued within the buffer's decay period. Recent findings concerning so-called "change blindness" as well as Ned Block's distinction between phenomenal and access consciousness will also be reviewed within this context.
The Categorical Fluency Effect: Nondeclarative Memory in Visual Category Learning
Paul J. Reber
Paul J. Reber (firstname.lastname@example.org)
Department of Psychology
2029 Sheridan Road
Evanston, IL 60201
Nondeclarative memory affects behavior without conscious retrieval of specific facts and events. One mechanism by which nondeclarative memory has been shown to affect processing is by affording perceptual fluency for previously presented stimuli (priming). The role of fluency in complex cognitive processing such as categorization has not been explored. Four experiments are reviewed that examine the neural basis of a visual category learning task that is supported by nondeclarative memory. These results suggest that fluency for categorical stimuli plays an important role in categorization. The hypothesis is advanced that learning a new visual category results in a change in early visual processing that leads to increased fluency for subsequent processing of categorical stimuli. This fluency effect is contrasted with the neural correlates of similar declarative memory tasks such as recognition or categorization supported by a conscious, explicit strategy. The patterns of activity in the visual system are shown to consistently distinguish between the operation of nonconscious, nondeclarative memory and conscious, declarative memory. This contrast suggests that fluency supports more complex cognition than simple repetition priming effects and is distinct from the mechanisms support conscious memory retrieval.
Conscious and nonconscious memory across saccadic eye movements
Karl Verfaillie, Peter De Graef & Veerle Gysen
Karl Verfaillie (Karl.Verfaillie@psy.kuleuven.ac.be)
Department of Psychology
University of Leuven
In a transsaccadic integration paradigm, observers make a saccadic eye movement to a visually presented target object and they have to detect object changes that occur during their saccade. Because of saccadic suppression, the transient that normally accompanies such a change is not perceptible and change detection is only possible by the integration of presaccadic and postsaccadic information. We previously observed that object displacements are very hard to detect while depth rotations of the object are readily noted, suggesting that transsaccadic memory for object position is inaccurate whereas memory for depth orientation is good. However, it has been suggested that, in contrast to what our findings suggest, transsaccadic coding of object position is in fact accurate, but the information is unavailable to conscious perception, because the visual system assumes the world remains stable. Deubel et al. (Vision Research, 1996) demonstrated that, by briefly blanking the saccade target during and just after the saccade, the default assumption of a stable visual world is invalidated and saccade- contingent displacements are relatively easy to detect. We report a series of experiments in which we use this blanking technique to unravel the nature of conscious and nonconscious transsaccadic memory.
Electrophysiological measures of conscious and nonconscious memory
Ken Paller (email@example.com)
In contemporary memory research, the subjective experience of remembering a fact or event, conscious recollection, has frequently been dissociated from a variety of other forms of memory that tend to occur without any recollective experience. In particular, patterns of memory impairment in patients with amnesia suggest that memory for facts and episodes depends on a process of neocortical consolidation that is not required for other types of memory such as perceptual priming. Electrophysiological and hemodynamic measures of brain activity can provide additional insights into the processes responsible for remembering. In this presentation I will describe distinct brain potentials that have been specifically associated with recollection and priming. These electrophysiological results add to the growing body of evidence supporting neurobiological conceptualizations of the distinction between recollection and priming. Moreover, these results hold promise for measuring and conceptualizing the brain events responsible for the conscious experience of remembering.
Does consciousness achieve binding?
Max Velmans (firstname.lastname@example.org)
Department of Psychology
University of London
In 1980, the dualist, John Eccles proposed that the "self-conscious mind" selects, attends to and integrates information displayed on the neocortex. Adopting a form of functionalism, Baars (1997) makes the related claim that consciousness carries out system-wide integration and dissemination of information, forms new links between unconscious processors, and so on. In this paper, I suggest that these claims are false, whether they are couched in dualist or functionalist terms. Neural binding is likely to be a necessary condition for having an integrated conscious experience. However conscious experience does not carry out neural binding, nor does it disseminate information throughout the brain. Such claims confound correlation and causation with ontological identity, and they confuse the relation of conscious phenomenology to the information processing that supports it. Neural binding and information dissemination are both achieved non-consciously.
A Mentalistic View of Conscious Unity and Dissociation
Donelson E. Dulany
Donelson E. Dulany (email@example.com)
Department of Psychology
University of Illinois, Urbana-Champaign
603 East Daniel Street
Champaign, IL 61820
Mentalism is a metatheory of mind that specifies roles for conscious states and nonconscious mental operations in a way that can provide constraints for theories of a range of phenomena (Dulany, 1997). On this view, symbols are mental contents carried exclusively by conscious states, a function that gives consciousness the adaptive significance of coping with a world beyond ourselves. Mental activity consists exclusively of mental episodes in which conscious mental states and contents are transformed by nonconscious mental operations, operations that are brain processes interrelating those conscious states and contents. Contents come in literal and identity codes that can be focal or non-focal, and remembrance and inference operations can yield higher order states of consciousness and a sense of agency. Dissociations occur within consciousness, when mental operations fail, not between consciousness and an unconscious. They can occur (a) between different forms of mental episodes, the evocative ("implicit") and the deliberative ("explicit"), (b) between literal and identity codes of awareness, (c) between first- order and higher-order consciousness, and (d) between conscious contents and a sense of agency (possession). The view is applied to several phenomena, ranging from imperfect unity in normality to more dramatic dissociations with brain pathology.
Integration, Phenomenal Unity, and Self-consciousness
Robert Van Gulick
Robert Van Gulick (RNVANGUL@syr.edu)
Syracuse, New York 13244-1170
The problem of unification and binding occurs in many forms:
- some of which apply only to phenomenal processes,
- some only to nonphenomenal processes and
- others of which apply to both.
Diverse versions of the problem focus on different types of integration or coherence including : temporal (at a time), cross-temporal, causal, dynamic, representational, intentional, experiential, intra-modal, cross-modal, and intuitional (in the Kantian sense of intuitions as the forms of sensible awareness) The general space of problems provides an opportunity for exploring detailed relationships between consciousness and its nonconscious correlates or substrates. What sorts of correspondences, isomorphisms and dependencies can one articulate between the types of unity and coherence that apply respectively at the conscious (phenomenal) level and nonconscious levels? Are such correspondences explanatory? In particular, might they help us understand how phenomenal consciousness could arise from or be wholly constituted by nonconscious process? That is, can understanding cross-level relations between the types of unity and coherence significantly help us close "the explanatory gap" and reduce the residual mystery that surrounds the so called "hard problem". The answer is yes, as long as one does not set an unjustifiably high standard of what counts as an explanatory correspondence. The notion of self-consciousness, if understood in a sufficiently broad and general way, may provide one useful bridge to link the diverse types of unity that apply at different levels.
Binding by synchrony and the transparency of consciousness
Markus Werning (firstname.lastname@example.org)
Department of Philosophy
Free University of Berlin
Am Schiessberg 6
Being conscious of something Â say, a red apple Â does not only imply having a representation of a red apple, but also implies that it is for the subject as if she were in a world with a red apple. This feature of consciousness is called transparency because the subject seems to access the object represented without prior accession of any intermediate representational state such as a mental sentence or image. Transparency can only be explained if the subject mistakes the representation of an object for the object represented. I define a neuronal algebra N which accounts for the empirically rather well confirmed hypotheses that there are collections of property-indicative neurons and that properties indicated by different collections are bound into a representation of an object by synchronous activation. Algebra N turns out to be isomorphic to a further algebra W that comprises worldly objects, properties and facts. Because N and W are isomorphic, they are indistinguishable for certain systems. To explain the transparency of consciousness, we have to assume that the brain is such a system: It mistakes the neuronal states of its own for objects, properties and facts of a possible world.
When 'I think' doesn't accompany my thoughts
FrÃ©dÃ©rique de Vignemont
FrÃ©dÃ©rique de Vignemont (email@example.com, firstname.lastname@example.org, email@example.com)
1, rue Descartes
In this paper, I intend to develop on conceptual and empirical grounds the distinction between two kinds of underlying mechanisms of consciousness unity, that is, the notion of access, and the notion of self-attribution. And none of these mechanisms is infallible, not even self-attribution, contrary to Shoemaker's claim of immunity to error through misidentification relative to the first-person. Usually, split-brain phenomenon is considered as the paradigmatic trouble of consciousness unity. Many accounts emphasize the notion of access (to controlateral hemisphere, to consciousness, to verbalization and so on). Unity of consciousness implies having access to one's own mental states. But it is not sufficient, as it appears in some schizophrenic symptoms. In thought insertion, the patient believes that someone has put an alien thought in his mind. He has access to it, but there is still a problem of unity, because of the defective monitoring of the cause of the thought. Unity depends on the acknowledgment of ownership, and so, on sense of agency. So, we have to distinguish two senses of consciousness unity. In the weak sense, the sufficient condition is to have conscious access to mental states. In the strong sense, we need to attribute to the same subject the available states.
States of consciousness in schizophrenia : a metacognitive approach to semantic memory
Elisabeth Bacon & Jean-Marie Danion
Elisabeth Bacon (firstname.lastname@example.org)
INSERM Unit 405, Psychiatric clinic,
University Hospital, BP 426,
F-67091 STRASBOURG Cedex,
Recent studies suggest that schizophrenia is a disease affecting states of consciousness. The present study aimed at investigating metamemory, i.e. knowledge about ones own memory capabilities, in patients with schizophrenia. The accuracy of Confidence Level (CL) in the correctness of recall answers and the predictability of Feeling Of Knowing (FOK) towards recognition were measured using a task of general information questions assessing semantic memory. Nineteen outpatients were paired with 19 control subjects with respect with age, sex and education. Patients with schizophrenia exhibited an impaired semantic memory. CL ratings and CL and FOK accuracy were not significantly different in the schizophrenic and comparison groups. However, FOK ratings were significantly reduced in patients, and discordant FOK ratings for correct answers were observed more frequently among schizophrenics. These results indicate that FOK judgments are impaired in patients with schizophrenia. They provide support for the hypothesis that the impairment of semantic performance observed in schizophrenia is due to a defect of access to appropriate knowledge and confirm that schizophrenia is characterized by an impaired conscious awareness of ones own knowledge.
Attribution of action in schizophrenic patients
Chloe Farrer , Nicolas Franck, Nicolas Georgieff & Marc Jeannerod
Chloe Farrer (email@example.com, firstname.lastname@example.org)
Institute of Cognitive Sciences
67 Bd Pinel
69675 Bron cedex
One class of schizophrenic symptoms the so-called positive symptoms are suggestive of an alteration of the awareness of ones own action and of the recognition of actions performed by others. One-hypothesis postulates that self consciousness relies on discrimination between central representations activated from within and those activated by external agents. A first experiment by Daprati et al (1997) has shown that schizophrenic patients tend to overattribute to themselves actions produced by others. In order to analyse more precisely the influence of the perceptive information on this agency judgement we realised another experiment where the parameters of the visual information were controlled using an electronic device, which allowed modifying the apparent direction and/or velocity of the movement actually performed, by the subject. 29 normal and 29 schizophrenic subjects performed a manual motor task and an agency judgement about this task. We determined angular and temporal threshold values for which subjects were no longer to distinguish the movements they executed from those they saw. The results showed greater thresholds values and significant more self-attribution responses for the patients compared to controls. We explained those results by a deficit at the level of the consciousness of action. Compared to controls schizophrenic subjects have more difficulties in recognising their own actions.
Binding in Dreams
Antti Revonsuo (email@example.com)
Department of Philosophy
Center for Cognitive Neuroscience
University of Turku
Dreaming provides us with a unique view to the way phenomenal representation -- consciousness -- is organized in the brain. Dream images often contain deviations and peculiarities that in dream research are referred to by the term "bizarreness". Bizarreness (e.g. incongruity, discontinuity) can be reconceptualized as referring to aberrations in the binding of phenomenal dream images coherently together. Incongruous dream images have features or appear in contexts inconsistent with waking perception. Discontinuity in dreams manifests failures of binding across time: successive dream images do not retain or properly update phenomenal representations, which leads to sudden appearance, disappearance or transformation of dream elements. Studies of bizarreness could therefore provide us with detailed descriptions of how the binding of phenomenal representations succeeds or fails during dreaming, possibly illuminating the mechanisms that work beneath the surface of phenomenal organization. I present a study on the bizarreness of human characters in dreams, especially designed to chart the failures of different types of binding. The results show that in dreams certain types of aberrations in binding are much more common than others. Information contributed by single modules tends to be coherent, but the global integration of phenomenal features contributed by non-modular or a number of different modular processing systems tends to be incoherently combined when binding the phenomenal dream world together.
Experience is not something we feel but something we do: a principled way of explaining sensory phenomenology, with Change Blindness and other empirical consequences
J. Kevin O'Regan & Alva NoÃ«
J. Kevin O'Regan (firstname.lastname@example.org)
Laboratoire de Psychologie ExpÃ©rimentale
Institut de Psychologie
Centre Universitaire de Boulogne
71, avenue Edouard Vaillant
92774 Boulogne-Billancourt Cedex
Any theory of experience which postulates that brain mechanisms generate "raw feel" encounters the impassable "explanatory gap" separating physics from phenomenology.
A way around the problem is to postulate that experience is not something we feel, but something we do: a kind of give-and-take with the environment, analogous to the "feel" of driving a car. One consequence of such a "sensorimotor" theory of experience is that it provides a way of explaining the differences between seeing, hearing, touch, etc., which is more principled and has more explanatory power than MÃ¼ller's notion of "specific nerve energy" or its modern counterpart, the notion of sensory pathways or cortical areas. The feasibility of sensory substitution is an empirically verifiable implication of this approach.
As applied to visual perception, a consequence of the sensorimotor approach is the idea that seeing does not consist in the creation of a "re-"presentation of the world inside the brain, but rather in knowledge that the outside world is immediately accessible through a flick of the eye or of attention, like an "outside memory". The world-as-an-outside-memory idea has empirically verifiable consequences in the phenomenon of Change Blindness, among others.
Unintended cognitive processing in briefly attended locations
Maria Stone & Roger W. Remington
Maria Stone (email@example.com)
MS262-4, NASA-Ames Research Center
Moffett Field, CA 94035
Previous experiments by Remington, Folk & McLean, using spatial cueing paradigm, have demonstrated that involuntary shifts of attention lead to processing of nontarget stimulus identity when the goals of the observer do not include such identity processing. Current experiments using the same paradigm examined limits on such unintended processing with cognitive operations that are more complex (Experiment 1) and less related to the observer's task (Experiment 2). In Experiment 1, we examine whether or not subjects carry out unintended same/different comparison of briefly attended nontarget stimuli, and in Experiment 2 we examine whether or not bilingual Russian-English subjects process letter identity of briefly attended nontarget letters in Russian when asked to respond to identities of English target letters. The results have implications for the more general question of when voluntary control over cognitive operations fails and when it is possible.
Capacity limitations in the detection and identification of change in visual arrays
Patrick Wilken & Jason B. Mattingley
Patrick Wilken (firstname.lastname@example.org)
Department of Psychology
University of Melbourne
MELBOURNE VIC 3010
Observers typically have trouble reporting salient changes between two visual displays if these are presented in alternating sequence, with a blank interval of short duration interposed between them. This phenomenon, commonly known as 'change blindness' (CB), suggests that the visual system maintains a relatively sparse representation of the world. Work in a number of research areas -- transsaccadic memory, visual-tracking, and visual short-term memory -- suggests that visual processing capacity is limited in many tasks to approximately 4-6 items. Here we report on an experiment in which observers were asked to detect and identify change in an array of coloured forms. In separate blocks of trials subjects were asked to detect and identify either a colour change (e.g. 'red' to 'blue') or a form change (e.g. 'L' to 'T') to one of the items in the array. False alarm rates were estimated from responses to those 50% of trials in which no change occurred. As expected the probability of detecting or identifying change in the form condition was much poorer than that in the colour condition. However, a simple lawful relationship was found to exist between the detection and identification of change in both conditions. Our results are inconsistent with a model in which limitations in the identification and detection of change are the result of a single underlying process, operating on a limited number of coherent objects held in a high-level working memory store. Instead, we suggest that detection and identification of change are separate processes, that share a common informational bottleneck.
Representing change with and without awareness: imaging studies
Diego Fernandez-Duque, Giordana Grossi, Ian M. Thornton & Helen Neville
Diego Fernandez-Duque (email@example.com )
Rotman Research Institute
Baycrest Centre for Geriatric Care
3560 Bathurst Street
Toronto, ON, M6A 1E6
Observers are often unaware of changes in their visual environment until attention is drawn to the location of change. Focused attention mediates change perception by giving objects coherence across space and time. To study the relation between focused attention and aware/unaware perception of change, we recorded Event Related Potentials from subjects performing a change blindness task. A complex scene was repeatedly presented for 500 ms, separated by a 300 ms blank mask. After several cycles, a change was introduced in the scene. Original and modified versions alternated until the change was reported (unaware change, unattended location of change). During the subsequent 30 to 40 flickers, subjects attended to the location of change and reported when the change was removed (aware change, attended location of change). Next, subjects looked for a second change in the same scene, usually absent (no change, unattended location of original change). Finally, subjects focused attention at the location of the original change to report its re-occurrence (no change, attended location of change). Preliminary analysis indicates activation differences, both as a function of attention (attended versus unattended location) and level of awareness (aware versus unaware). We plan to further explore these effects using functional magnetic resonance imaging.
Priming capabilities of 'attention masked' words and pictures: ERP studies on the attentional blink
Michael Niedeggen & Petra Stoerig
Michael Niedeggen (firstname.lastname@example.org)
Institute of Physiological Psychology II
Heinrich-Heine University of Duesseldorf
The 'attentional blink' relies on the rapid serial visual presentation (RSVP, ~10 items/s) of stimuli, in which two relevant targets are embedded: When the first target is detected, explicit processing of the second target (probe) is suppressed if it is presented 200-400ms later. To learn whether these 'attention masked' items (Exp.1: words, Exp.2: pictures, each n=16) affect the processing of test words presented following the RSVP, we studied covert recognition by presenting either the masked probe or a new stimulus. To examine semantical priming, we presented test words related or unrelated to the probes (Exp.3, n=12). Processing of the test word was monitored by recording event-related brain potentials (ERPs). The absence of explicit probe recognition reflected in a diminished P3 amplitude, probably indicating a mismatch in episodic memory. Nevertheless, repeated stimuli are treated differently from new items which generated a more negative-going ERP waveform. A comparable reduction of ERP negativity was obtained for semantically related in contrast to unrelated test words (Exp.3). In sum, our results indicate that lexical meaning of 'attention masked' items seems to be encountered, but not upgraded in episodic memory.
Effects of Color in Implicit Memory
Andre Melzer & Werner Wippich
Andre Melzer (email@example.com)
University of Trier
In two sets of experiments we investigated (a) the effects of study-task and object-color association manipulations on implicit memory for color-to-object identity bindings, and (b) whether the representations underlying task performance are semantic or perceptual in nature. In contrast to the study phase, test items were presented without color information. In the implicit color-choice test, priming of color was shown when the target colors were chosen more frequently for old items than for new items. In the first set of experiments, priming was observed even when participants disregarded the color information in the color-pictures study condition. However, when color information had been presented on separate cards during encoding, priming depended on whether the study task required attention to both the object identity and its specific color. We conclude that the amount of focal attention needed for color-to-object identity bindings to be reflected in an implicit memory test depends on the spatial integration of the color and the object identity during encoding. As to the second major issue, results were inconclusive: While there was evidence for a semantic basis underlying task performance in one experiment, results of a second experiment suggested that mental representations may also be perceptual in nature.
Distinguishing conscious from non-conscious discrimination: Exploring functional analogs of blindsight in normals using visuo-motor responses to masked targets
Mark C. Price, Elisabeth Norman & Simon C. Duff
Mark C. Price (firstname.lastname@example.org)
Institutt for Samfunnspsykologi
University of Bergen
Christiesgt. 12, 5015, Bergen
"Blindsight" patients with lesions in primary visual cortex have impaired phenomenal experience of visual stimuli, but may still make above-chance forced-choice discriminations to such stimuli, e.g. by pointing to their location. We report attempts to model blindsight in brain-intact subjects by using masking paradigms (borrowed from research on non-conscious perception) to present visual targets subliminally. These studies reveal potential pitfalls with both introspective and behavioural/operational measures of consciousness, raising doubt over some previous claims to have modeled blindsight in normals. For example, although subjects may declaratively indicate (by key pressing) the location of peri-liminal masked targets more accurately than they can distinguish between their presence or absence, this seems to be a consciously mediated advantage rather than the "blindsight-like" effect previously claimed by some studies. Further studies investigate whether masking paradigms can be used to obtain functional analogs of blindsight in normals when subjects attempt to guess the location of masked visual targets using finger pointing (visuo-motor) responses. Our studies are based on converging behavioural controls for consciousness that exploit qualitative differences between conscious and non-conscious processing, and which may be of relevance to future studies which attempt to map the contents of visual consciousness in patients.
A neural correlate of visual awareness: exploring the N265 component
Ville Ojanen, Maria Wilenius-Emet & Antti Revonsuo
Ville Ojanen (email@example.com)
Centre for Cognitive Neuroscience
University of Turku
In a recent EEG study by Wilenius-Emet and Revonsuo (in preparation) visual awareness was examined during an object detection task where the target stimuli were briefly flashed between forward- and backward masks. A prominent negative ERP-component that correlated with conscious perception was observed at around 265 ms after target stimulus onset. This "N265" component was elicited by all consciously perceived stimuli (i.e. objects and scrambled objects presented at stimulus durations at or above the recognition threshold) but not by any stimuli that could not be consciously perceived (i.e. similar stimuli presented at stimulus durations below the recognition threshold). Thus, the N265 component seems to reflect the neural correlates of visual awareness and binding. The present EEG study is a modification of the previous study by Wilenius-Emet & Revonsuo and the MEG study by Vanni et al. (1996). The aim was to further examine the N265 component and its correlation to visual awareness. The original experiment was modified to see whether the N265 component primarily reflects the conscious detection of stimulus brightness changes during the stimulation sequence (the target stimulus had higher luminance than the forward and backward mask) or the sequential presentation of three stimuli (mask-stimulus-mask). The experiment was modified respectively by prolonging the stimulus duration and by excluding either the forward or the backward mask or both from the sequence. The results discussed in the talk show how these manipulations modulate the generation, latency and amplitude of the N265 component. This in turn reveals whether the component is the "vertex negative component" earlier identified and studied by Jeffreys (1989) and how exactly this ERP-component is related to visual binding and awareness.
Memes and the malign user illusion
Department of Psychology
St Matthias College
University of the West of England,
Bristol BS16 2JP
Many authors have suggested that the apparent unity of consciousness is an illusion. Parfit contrasts ego theorists with bundle theorists, who believe that underlying the apparent continuity of self is only a series or bundle of experiences. Dennett, having rejected the audience in the Cartesian Theatre, suggests that the self is a benign user illusion. If we accept this, we must ask (1) How is the illusion constructed? (2) Why is it constructed, and (3) Is it really benign? Possible answers come from evolutionary, cognitive and social explanations. I shall argue that none of these is valid, for the illusion does not benefit either us or our genes, it benefits our memes. I shall explain the basic principles of memetics, pointing out sources of confusion over definitions and false analogies with genes, and stressing the importance of understanding memes as replicators. I shall explain how and why memes group together into co-adapted meme complexes, one of which is the self. This memeplex is constructed by the memes for their own propagation, not for our benefit. Indeed it is arguably the root of all human suffering. I conclude that both scientifically, and for living our lives, the illusion is malign.
Unity of Consciousness: What It Is and Where It Occurs
Andrew Brook (firstname.lastname@example.org)
Institute of Interdisciplinary Studies
Ottawa, Ontario, K1S 5B6
Unity of consciousness is our capacity to be conscious of a number of items all at once, in what could be called a single conscious act. Such unity is found in at least three 'places': consciousness of the world in general, consciousness of self in general, and paying focal attention to aspects of either. In all three, unified consciousness has both a synchronic and a diachronic dimension. That is to say, consciousness is unified both at a given moment and over time. Unified consciousness can break down by splitting (into two unified centres of consciousness, as in brain bisection operations) and by shattering (as in some severe schizophrenias and dysexecutive disorder). Studying it in its breakdown conditions is a good way to throw light on it. In this paper, we will delineate the unity of consciousness, explore some situations in which it breaks down, and relate it to some other mental unities.
The Appearance of Unity: A Higher-Order Interpretation of the Unity of Consciousness
Josh Weisberg (email@example.com)
The Graduate School and University Center of The City University of New York
Department of Philosophy, 7th floor
365 5th Ave. New York, NY 10016
Recent work in neuroscience and psychology has put pressure on the traditional philosophical idea of the unity of consciousness. Studies of split-brain and hemineglect subjects, as well as results from various priming and confabulation experiments, challenge the vision of unbroken subjective experience. Yet from the first-person point-of-view, things seem unified. Can this appearance of unity be accounted for in a physicalist theory of consciousness, or must we reject and eliminate the very notion of the unity of consciousness? In this paper, I will investigate the possibility that "higher-order" theories of consciousness can provide an explication of the appearance of unity, while respecting the relevant neuroscientific and psychological results. I will focus on David M. Rosenthal's higher-order-thought hypothesis (1986, 1997), which holds that higher-order thought determines what it is like for us to be conscious of our mental states. But higher-order thought can also have the effect of "cleaning up" the appearance of target states. We may represent things as unbroken and unified at the target level, when in fact they are not that way. The appearance of unity derives from the manner in which higher-order thought interprets lower-order states. I will evaluate the success of this proposal, focusing in particular on the idea that there may be an appearance/reality distinction within consciousness itself.
Rosenthal, D. M. (1986). "Two Concepts of Consciousness." Philosophical Studies 49, 329-359.
Rosenthal, D. M. (1997). "A Theory of Consciousness." in Block, N., Flanagan, O., and Guzeldere, G.,(eds.), The Nature of Consciousness: Philosophical Debates, Cambridge, MA: MIT Press, 729-753.
The continuity of self in collectivism and individualism
William P. Banks, Kris Y. Yi, Angela V. Lumanau & Nancy Chen
William P. Banks (Wbanks@pomona.edu)
Department of Psychology
Claremont, CA 91711
One important factor in the unity of consciousness is the self, which provides a resource for continuity, a reference for memories, and a coherent interpretation of reality. Cultural factors are essential to the construction of the self. The dimension of individualism and collectivism emerges as a universal classification of societies, and this is explicity a dimension that concerns the structure of the self. We found that individualistic respondents (American college students) and collectivistic respondents (college-age Japanese citizens) had, surprisingly, almost identical degrees of the self-reference effect in memory (the finding that a list of random trait adjectives, when judged as applicable to one's self, are easier to remember and recall than a list of the same adjectives learned using other methods). We also compared the two groups in the fundamental attribution error. We found the typical error for the individualistic participants, but a great reluctance by the collectivistic participants to make any attribution at all. Finally, we showed that a state of "war" (an intercollegiate game) causes American team members to become more collectivistic and to have simpler out-group stereotypes, and more complex and positive in-group attributions, than a non-war state. Implications for continuity and self cross-culturally will be discussed.
Alien Voices: An Event-Related fMRI Study of Overt Verbal Self-Monitoring
Cynthia H.Y. Fu, Edson Amaro Jr, Mick Brammer, Farooq Ahmad, Chris Andrew, Steve C.R. Williams, Nanda Vythelingum & Philip K. McGuire
Cynthia H.Y. Fu (firstname.lastname@example.org)
Division of Psychological Medicine
Institute of Psychiatry
de Crespigny Park
London SE5 8AF
A fundamental aspect of consciousness is the awareness of one's own thoughts, which requires monitoring of self-generated cognitive activity. The neural correlates of verbal self-monitoring have been investigated using positron emission tomography (PET) (McGuire et al., 1996). Functional magnetic resonance imaging (fMRI) has an advantage of greater temporal and spatial resolution than PET, but it has a disadvantage of the production of scanner noise during image acquisition. We have therefore employed an event-related fMRI design and scanner acquisition sequence to overcome this confound.
FMR images were acquired on a 1.5 Tesla GE Signa System. Seven healthy dextral male subjects read aloud adjectives and heard their voice which were either: (i) undistorted; (ii) pitch distorted; (iii) replaced by an "alien voice"; (iv) replaced by a distorted "alien voice".
The fMRI data revealed common areas of cerebral activations in an extensive network which includes the basal ganglia, insula, and inferior frontal, superior temporal, cingulate and cerebellar cortices.
Modified event-related sequences provide a means of avoiding the effects of scanner noise during fMRI. Verbal self-monitoring involves a network of areas implicated in the generation and perception of speech.
Mental states of oneself and others are distinctly implemented in the human brain
K. Vogeley, P. Bussfeld, A. Newen, S. Herrmann, F. HappÃ©, P. Falkai, J. Shah & K. Zilles
Kai Vogeley (email@example.com)
Department of Psychiatry
University of Bonn
The capacity for the meta-representation of oneÂ´s own mental states is a human cognitive capacity, closely related to theory of mind (TOM) paradigms in which mental states of others have to be modeled. However, it was unknown, whether the meta-representation of mental states of others and of oneself are differentially implemented in the brain. To empirically address this issue, a fMRI study was performed, employing a well-characterised "theory of mind" (TOM) paradigm, that was extended to include self perspectivity stimuli in a two-factorial design. Short stories with subsequent questions were presented to 8 right-handed normal male test persons in five different conditions (control condition, TOM- SELF-, TOM+ SELF-, TOM+ SELF+. TOM- SELF+). Activation associated with theory of mind-capacity was located in the anterior cingulate gyrus, as shown in previous studies (Fletcher et al., Cognition 57, 109-128, 1995). Activation associated with self perspectivity was mainly located in the right temporoparietal region, closely related to the body image region in the right parietal lobe. These results provide data on the cerebral implementation of a partial feature of human self-consciousness and suggest, that self perspectivity is involving the body axis as center of the ego-centered experiential space.
Stability phase transition in binocular rivalry
Yoshi Tamori & Ken Mogi
Yoshi Tamori (firstname.lastname@example.org)
Human Information System Laboratories
Kanazawa Institute of Technology
3-1, Yatsukaho, Matto
In binocular vision, there are two possible phases. Stable phase, when we get a stable visual image in our visual awareness, and unstable phase, when the percept we get changes with time. The latter phase is best studied in the binocular rivalry paradigm. By using partially low-path filtered (blurred) images of human face and natural objects, we studied the factors which contribute to the above transition between stable and unstable phases. Specifically, two images which have been low-path filtered in alternating patchy regions were presented binocularly. Under a certain range of parameters (e.g. cut-off frequency used in blurring and the size of the blurred patchy regions), the subject was able to see a stable and overall clear image. In this case, the brain has successfully picked up alternating "sharp" regions and constructed an overall sharp and stable image. Our analysis suggests that in order to have a stable percept, the overall structural coherence as well as a certain level of correlation between the binocularly presented images are necessary. When there is structural coherence but the correlation is low, the resulting percept becomes unstable. We discuss the factors contributing to the stability of visual awareness.
Evidence for multistability in visual perception of pigeons.
J.-D. Haynes, G. Vetter & S. Pfaff
John-Dylan Haynes (email@example.com)
Institute for Psychology and Cognition Research
University of Bremen
Grazer Str. 4
Perceptual multistability refers to cases where perception alternates between two or more interpretations of an unchanging sensory stimulus. We performed experiments with pigeons (Columba Livia) in order to seek for evidence for perceptual reversals. In a first experiment we trained 8 pigeons to discriminate horizontal and vertical apparent motion stimuli and then presented a multistable motion display. In 5 cases their behavior showed alternations similar to those known from human experiments. In a second experiment we varied the aspect ratio of the display in order to support the hypothesis of a percept-driven nature of the switching behavior. The pecking rates and mean phase durations varied as predicted: The animals responded significantly longer to the pecking key associated with the biased stimulus. Also the animals that did not reverse in experiment 1 now showed reversals. This is the first evidence of visual multistability in animals confronted with classical ambiguous figures. The data are evaluated and discussed in terms of classical models, such as satiation/adaptation, top-down processes and stochastic processes. Our results support a stochastic mechanism but with slightly different parameters than known from humans and from animal studies on binocular rivalry.
What is salient in binocular rivalry
Fumihiko Taya & Ken Mogi
Ken Mogi (firstname.lastname@example.org)
Sony Computer Science Laboratories
Takanawa Muse Bldg.
Shinagawa-ku, Tokyo, 141-0022
We studied what stimulus features count as salient in binocular rivalry. The stimuli were moving circles in a homogeneous background. There was a variable phase difference between the circles moving in the images presented to the right and left eyes. Under this condition, both circles were visible most of the time, indicating that the ocular dominance pattern is modulated in the spatio-temporal domain in such away that salient features (moving circles) were visible most of the time. Comparing this result with the control experiment where the circles were stationary, we conclude that motion is very effective in determining the ocular dominance pattern in binocular rivalry, suggesting a predominance of the dorsal visual pathway. A further experiment involving interactive computer graphics indicate that attention in general is dissociated from the process which correlates with the ocular dominance pattern in binocular rivalry. We thus arrive at a model of binocular rivalry involving three phenomenological layers. What we see in binocular rivalry is largely determined by the interaction between "pointers" and qualia, where the pointers represent salient features. Attention has only an indirect effect on visual awareness through its amplification of a particular pointer element.
Attentional guidance based on a preattentive analysis of emotional expression
John D. Eastwood, Daniel Smilek & Philip M. Merikle
John D. Eastwood (email@example.com)
Department of Psychology
University of Waterloo
To evaluate whether the focus of attention can be guided on the basis of a preattentive analysis of the emotion expressed in a face, participants searched displays of schematic faces for the location of a unique face expressing either a positive or negative emotion. On each trial the unique face was embedded among 6, 10, 14, or 18 distractor faces expressing a neutral emotion. The slopes of the search functions for locating the negative face were shallower than the slopes of the search functions for locating the positive face. However, when the faces were inverted to reduce holistic face perception, yet maintain feature differences, the slopes of the search functions for locating positive and negative faces did not differ. Taken together, the results suggest that emotion can be perceived outside the focus of attention and that focal attention can be guided on the basis of a preattentive analysis of the emotion expressed in a face.
Can emotions be dissociated from cognition?
Susanne Erk & Henrik Walter
Susanne Erk (firstname.lastname@example.org)
Department of Psychiatry
University of Ulm
The development of neuroimaging techniques has considerably contributed to recent progress in research on mental functions and emotions. Functional neuroimaging studies focus on localizing emotions to structures as amygdala, orbitofrontal cortex, basal ganglia, hypothalamus or brainstem. Several studies have tried to correlate specific emotions with specific activation patterns in these structures but the results have been more or less inconclusive. Apart from methodological problems we think that the reason is that most studies have tried to isolate pure emotions. But whether emotions can be isolated from cognitive processes at all naturally hinges on the question of what emotions really are - a question for which a rather huge number of answers exists. We propose a new concept of emotions which is based on computational models as well as neuromodulatory concepts of emotions. Emotions should be understood as modulators of behavior and cognition - the "how" of information processing. This approach has considerable implications for the role of emotions in constructing a theory of consciousness as well as for functional neuroimaging studies of emotions. We will present some of our own data using fMRI as an example for the plausibility of this approach.
Unconscious Emotions - Black Holes in the Cartesian Theatre?
Christoph Jaeger & Anne Bartsch
Christoph Jaeger (email@example.com)
Institut fÃ¼r Philosophie
In this paper we shall ask how recent psychological research on repression bears upon philosophical models of the self. Specifically, we shall discuss the consequences for the view that self-conscious subjects enjoy epistemically privileged access to their current thoughts and sensations. Although it is generally conceded that a tolerable formulation of this old "Cartesian" claim requires refinements and provisos, many philosophers (e.g. D. Davidson, S. Shoemaker) still hold that there is at least the following asymmetry: for a vast number of mental properties it holds that their ascriptions to others must rely on behavioral evidence, whereas self-ascriptions thereof need not rest on any such basis. However, recent psychological findings on emotional self-deception and repression (D.A. Weinberger, G.E. Schwartz, M.N. Davidson, K.W. Davidson, M. Mendolia,) seem to jeopardize even moderate versions of such tenets, according to which people are at least acquainted with their non-propositional, sensory mental states in a direct, non-inferential way, and corresponding reports enjoy a special first-person authority. Emotions seem to be paradigm examples of such states. But if, as current psychological investigations suggest, even emotions can be systematically ignored or misinterpreted, it seems that one of the last bastions of special access is ruined. In our talk, however, we shall argue that this conclusion is unwarranted. Relating the issue of repression to neurological theories of emotional information processing (LeDoux, Damasio), we shall argue, firstly, that there is a primary, subcortical level of emotion that cannot, by causal necessity, become conscious, and therefore can neither be a target of repression nor a realm of the mind where privileged access claims make sense. Repression must involve a representational level of potentially conscious emotion. Following Damasio, we shall, secondly, introduce a distinction between emotions and feelings, the latter being construed as experienced emotions. The advocate of privileged access can argue that emotional repression does not threaten his claim with regard to feelings, since they do have phenomenal content for their subjects. Thirdly, and perhaps most importantly, it will be shown that the very concept of repression only makes sense on the assumption that privileged access to (potentially conscious) emotions exists.
Sensory Qualities, Concsiousness, and Perception
David M. Rosenthal
David M. Rosenthal (firstname.lastname@example.org)
Ph.D. Program in Philosophy
CUNY Graduate Center
365 Fifth Avenue
New York, NY 10016-4309
For a mental state to count as a sensation of any sort, it must have qualitative properties. But there is a popular notion that a state's having qualitative properties means that there is something that it's like to be in that state--something that it's like _for the subject_. And, since there is something it's like for one to be in a state that isn't conscious, all sensations must, on this view, be conscious. One of my aims here will be to undermine this conception of what sensations are, which I believe is unfounded theoretically and unsupported by any reliable commonsense intuition. I'll argue that clarity about just what properties are essential to a state's being a sensation leaves no doubt that states can have those properties without being conscious. I'll then briefly sketch the higher-order-thought model of consciousness I've developed elsewhere, and argue that, contrary to what some critics have claimed, this model does justice to what's involved in there being something it's like for one to have conscious sensations. I'll conclude with an account of the qualitative properties of sensations which both fits with and sustains the earlier arguments.
Constrained inversions of sensations
Erik Myin (email@example.com)
AI Lab (WE Arti) & Wijsbegeerte (LW EMEP)
Vrije Universiteit Brussel
Inverted sensation arguments such as the inverted spectrum thought experiment are often criticised for relying on an unconstrained notion of 'qualia'. In reply to this criticism, 'qualia-free' arguments for inversion have been proposed, in which only physical changes happen: inversions in the world (e.g. the replacement of surface colors by their complements) and a rewiring of pheripheral input cables to more central areas in the nervous system. I show why such constrained inversion arguments won't work. The first problem is that the world lacks the symmetry to invert physical properties in the way required. The second problems concerns 'rewiring'. Empirical evidence indicates that the necessary rewirings are either impossible, or would not result in an inversion of sensations. This is illustrated by detailed examples involving both lightness and hue inversion. I propose the deeper reason for the failure of constrained inversion arguments lies in the fact that sensations are not properties of brain states, but spread into the world and the body.
Capturing Qualia: Higher-order Concepts and Connectionism
Bryon Cunningham (firstname.lastname@example.org)
Department of Philosophy
Bowden Hall, 214
561 S. Kilgo Circle
Atlanta, GA 30322
Antireductionist philosophers have argued for higher-order classifications of qualia that locate consciousness outside the scope of conventional scientific explanations, viz., by classifying qualia as intrinsic, basic, or subjective properties, antireductionists distinguish qualia from extrinsic, complex, and objective properties, and thereby distinguish conscious mental states from the possible explananda of functionalist or physicalist explanations. In this paper I argue that, in important respects, qualia are intrinsic, basic, and subjective properties of conscious mental states, and that, contrary to popular opinion, this higher-order classification is compatible with qualia reduction. This is shown by taking a closer look at the putative higher-order properties of qualia and comparing them to the higher-order properties characteristic of connectionist models of cognitive processes. It is my contention that the higher-order properties characteristic of connectionist networks approximate (in intertheoretic terms) the putative higher-order properties of qualia sufficiently well to conclude that qualia reductionism can accommodate: (1) claims that qualia are intrinsic, basic, and subjective properties; and (2) the motivating intuitions for those claims generated by inverted, absent, and alien qualia thought experiments. In this way I argue that (approximate versions of) the putative higher-order classifications of qualia not only fail to defeat qualia reduction but, ironically, turn out to support it.
Continuity and Consciousness
William S. Robinson
William S. Robinson (email@example.com)
Department of Philosophy and Religious Studies
402 Catt Hall
Iowa State University
Ames, IA 50011-1306
It seems that many experiences exhibit continuity, e.g., foghorn sounds are spread out in time without interspersed silences, and ice cubes of the kind that Sellars called "homogeneously pink" are pink throughout, pink without gaps in the color expanse. A realist about continuity holds that apparently continuous spread of qualities of external things is possible because there actually is strictly continuous spread of phenomenal qualities in consciousness. The first part of the present paper defends realism about continuity against a highly developed critique by Austen Clark. Key distinctions here are between conditions on causes of appearances and conditions on appearances themselves, and between continuity of phenomenal qualities (or 'regionality' as contrasted with punctate character), and continuity of changes of phenomenal qualities. An implication of the first part is that continuity is a kind of unity (of phenomenal qualities with time and in some cases with both time and space). The second part of the paper articulates the nature of this unity, and investigates how it, and its causes, are similar and different to other kinds of unities, and their causes. This part of the paper thus clarifies the range of phenomena that a theory of causes of consciousness must explain.
The Dimensions of Conscious Experience: A Quantitative Analysis
Steven Lehar (firstname.lastname@example.org)
Schepens Eye Research Institute
14 Crooked Lane
Manchester MA, 01944
There are many fundamental properties of visual experience that are difficult to account for, not only in terms of contemporary theories of neural representation, but even in more general terms of computational theory. In the first place vision is vividly three-dimensional, consisting of solid volumes, bounded by colored surfaces, embedded in a spatial void. Every point on every visible surface produces a distinct experience of color. The perception of transparency, and the experience of empty space between the observer and a visible surface, demonstrate that multiple depth values can be experienced in every direction in visual space. Furthermore, perceived objects are observed to translate and rotate coherently through perceived space while maintaining their structural integrity and recognized identity, even through perspective and elastic distortion. These properties of perception have been so problematic for contemporary theories of neurocomputation that they have been essentially ignored. I propose a perceptual modeling approach, as opposed to a neural modeling approach, i.e. to model the subjective experience of vision in quantitative terms, rather than the neurophysiological mechanism by which that experience is subserved. This approach leads to new insights into the nature of the perceptual transformation, and the informational content of conscious experience.
Is Language Structure Accessible to Consciousness?
Maxim I. Stamenov
Maxim I. Stamenov (email@example.com)
Institute of the Bulgarian Language
Bulgarian Academy of Sciences
Shipchenski Prokhod St. 52, bl. 17
One of the great puzzles in cognitive sciences is the problem about the way of accessibility and representability of cognitive structure within and outside consciousness. Here the views expressed by different authors diverge radically. Searle (1992) made the point that for some mental content to be considered as such, it MUST be in principle accessible to consciousness. Those `contents' which are in principle inaccessible to consciousness actually do not belong to the province of the mental but are just `brute neurophysiological facts'. Chomsky (1965, 1993, 1995), on the other hand, in many publications repeatedly pointed out that language structure is in principle inaccessible to consciousness. Chomsky's argumentation possesses some specific features based on the methodology of the study of language. His points are sometimes not well appreciated outside linguistics. The aim of this paper will be to elucidate with appropriate examples why Chomsky found possible to maintain such an `extravagant' position regarding the relationship between language and consciousness.
On the unitary representation of the visual world: global criterion constraints in local visual awareness
Andrei Gorea & Dov Sagi
Andrei Gorea (firstname.lastname@example.org)
Laboratoire de Psychologie ExpÃ©rimentale
CNRS & RenÃ© Descartes University
71 Ave. Edouard Vaillant
92774 Boulogne-Billancourt, Cedex
Perceptual studies make a clear distinction between sensitivity and decision-criterion. The former is taken to characterise the processing efficiency of the underlying sensory system and it increases with stimulus strength. The latter is regarded as the manifestation of a subjective operation whereby individuals decide on (as opposed to react reflexively to) the occurrence of an event based on factors such as expectation and pay-off, in addition to its intensity. To do so, individuals need to have some knowledge of the internal response distribution evoked by this event. In a natural, multi-stimulus environment, observers must handle many such independent distributions in order to optimise their decision criteria. Here we show that they cannot do so. Instead observers adjust each criterion in relation to the internal distributions generated by the concurrent stimuli. The consequence of this global constraint is that less visible events are reported less frequently in the presence of more visible ones, whereas the latter are reported more frequently in the presence of the former. The specifics of this behaviour are in quantitative agreement with predictions based on the notion that observers represent a multi-stimulus environment as a unitary internal distribution to which each stimulus contributes proportionally to its probability of occurrence. This "unity" of the internal representation of the external world may underlie, or be tantamount to the unity of the (visual) world as experienced. It may account at least in part for a number of intriguing perceptual phenomena such as blindsight, hemineglect and extinction.
The tri-partite system of consciousness
Benny Shanon (email@example.com)
Department of Psychology
The Hebrew University
On the basis of a review of the literature coupled with a phenomenological analysis, it is proposed that consciousness be regarded as a unified system comprising of three states. The first, and fundamental, state is one of ill-defined, undifferentiated qualities; the second, which is the pivotal one, is that of well-defined, articulated patterns of expression and behavior; the third is a second-order state taking the patterns of the second state as objects for manipulation and reflection. These three states are encountered in all different facets of the system of consciousness: subjective experience, the self, meaning and time. For instance, in the case of subjective experience, the three states are manifested in the phenomena of sentience, internal mentation and reflective meta-cognition, respectively. It appears that there are regular relationships between the three states so that together they define one coherent and internally structured system. Furthermore, a dynamic perspective may be taken whereby the system is viewed in terms of the functional mappings between the states. These mappings may be regarded as "the acts of consciousness". Among them are internalization, differentiation, objectivization and world-building. The discussion is grounded in a pragmatic approach by which the basic capability of the human cognitive system is being-in- and acting-in-the-world. It is proposed that the systematic study of the acts of consciousness defines a new paradigm of psychological investigation. Theoretically, this paradigm conceptualizes psychology as the study of human experience; methodologically, it is based on systematic phenomenology.
The mechanism of action of hallucinogenic drugs
Hans Flohr (firstname.lastname@example.org)
Brain Research Institute
University of Bremen
P.O. Box 33 04 40
Hallucinogenic drugs cause altered states of consciousness, characterized by sensory illusions, hallucinations, disorganized thought and bizarre ego-disorders. According to their site of action one can distinguish three groups of hallucinogens:
- partial serotonin agonists, like LSD and mescaline,
- anticholinergic drugs, like scopolamine and atropine,
- non-competitive NMDA antagonists, like phencyclidine and ketamine.
In spite of their different targets the psychedelic symptoms caused by the different hallucinogens are remarkably similar. Here we propose a hypothesis on the mechanism of action of psychedelic drugs that (1.) leads to a realization hypothesis for altered states of consciousness and (2.) can explain this similarity:
(1.) Hallucinogenic effects are due to a common mechanism of action distal from the primary target. The ultimate target relevant for the occurence of altered states of consciousness is the formation of large-scale neuronal assemblies which is made possible by the cortical NMDA synapse.
(2.) All hallucinogenic drugs directly or indirectly inhibit the NMDA receptor and thereby the induction of activity-dependent, rapid plastic processes controlled by this receptor.
This hypothesis is in line with previous hypotheses on the key role of the NMDA synapse in the realization of conscious states (Flohr 1991, 1995a, b). According to these hypotheses the NMDA synapse implements the binding mechanism that the brain uses to produce large-scale neuronal assemblies which instantiate higher-order representations. Abnormal states of consciousness are produced, iff this binding mechanism is disturbed and deformed higher-order representations are generated.
Flohr, H. (1991) Theory and Psychology 1, 245-262
Flohr, H. (1995a) Neuropsychologia 33, 1169-1180
Flohr, H. (1995b) Behav.Brain Res. 71, 157-161
Interaction between precuneus and dorsolateral prefrontal cortex may play a unitary role in consciousness - A principal component analysis of rCBF
Troels W. Kjaer & Hans C. Lou
Troels W Kjaer (email@example.com)
The John F. Kennedy Institute
Gl. Landevej 7
A fundamental principle in the neural organization in all mammals is a division into two major moieties, a posterior for perception and an anterior for action. The dichotomy is present in the spinal cord, in the diencephalon and extends upwards into the cerebral cortex, where the regions posterior to the central sulcus are dedicated to perception, and those in front to action. The perception-action-cycle suggested by Fuster (1995), implies that consciousness of goal directed action is a function of interaction between polymodal cortices in the perceptory and executory domains. It is hypothesized that if neural activity here exceed some threshold in intensity and / or time consciousness occurs. We hypothesize that this axis has a mere general role in consciousness independent of action. To test this hypothesis we analyzed H2O15-PET data on regional cerebral blood flow from 42 PET scans of 7 yoga practitioners during attentive and meditation. The two states are characterized by specific EEG patterns. The data were analyzed by statistical parametric mapping and the principal component method. Statistical parametric mapping revealed region selectively activated in each of the two conditions. The first principal component explained 25% of the rCBF variability and consisted primarily of executory regions: Prefrontal cortex, striatum, thalamus, and cerebellum, regions also found to decrease during meditation. The second explained 18% and consisted of sensory posterior cortical regions, regions also found to increase during meditation. The precuneus, middle prefrontal cortex, and striatum contributed markedly to both the executory network and the sensory network. This combined action of precuneus and prefrontal cortex across conscious state is a confirmation of the hypothesis of an active perception-action axis in unitary consciousness independent of action.
Brain function during vegetative state and after recovery: a lesional approach to the study of human consciousness
S. Laureys, M.-E. Faymonville, N. Janssens, C. Degueldre, P. Peigneux, G. Del Fiore, P. Damas, B. Lambermont, A. Luxen, M. Lamy, G. Moonen & P. Maquet
Steven Laureys (firstname.lastname@example.org)
Cyclotron Research Centre
University of LiÃ¨ge, Sart Tilman B30
Vegetative state is a unique condition of preserved arousal with abolished awareness. Here, we identified the functional neuroanatomy during vegetative state and after restoration of consciousness. [18F]fluorodeoxyglucose PET was performed in 14 vegetative patients, 3 patients subsequently recovered consciousness and were re-scanned under the same experimental conditions. Using statistical parametric mapping (SPM99) we localized regions where metabolism was most impaired during vegetative state and returned to near-normal values after recovery of consciousness, compared with 35 age-matched controls. The analysis identified bilateral prefrontal cortices (Brodmanns area 8, 9 and 10), Brocas area, left superior temporal gyrus (area 22), bilateral inferior parietal lobule (area 39/40) and precuneus (area 7/31) (voxel-level-corrected p<0.01). We postulate that the resumption of long-range functional connectivity between these associative cortices and between some of them and the thalami constitute an important process in the reversibility of their functional integrity.1 Our observation suggests that conscious experience in part relies on functional properties of networks involving distant associative cortices, known to be involved in various consciousness-related functions (conscious perception, attention, working memory, episodic memory and language).
1. Laureys et al. Restoration of thalamo-cortical connectivity after recovery from persistent vegetative state. Lancet, in press.
Hypnosis and altered states of consciousness
Sakari Kallio & Antti Revonsuo
Sakari Kallio (email@example.com)
University of Turku
Department of Psychology
FIN 20014 Turku
The idea that hypnotic induction can lead to an altered state of consciousness (ASC) is highly controversial in hypnosis research. According to the "state"- theory, hypnosis evokes with high hypnotizable subjects essentially an ASC with its own characteristic neuropsychological and neurophysiological correlates. According to the rival "non-state" theory, all the psychological, physiological and behavioral changes associated with hypnosis can be explained without assuming any ASC. Recent attempts to resolve the controversy include the use of new brain sensing and imaging techniques that are expected to reveal the characteristic neurophysiological correlates of hypnosis. Although there is some evidence of such changes, the controversy has not been solved. The non-state theorists still refuse to admit that the reported neurophysiological changes should be interpreted as signs of an ASC. In the present paper, this controversy between the "state" and "nonstate" theories of hypnosis is analyzed from a theoretical and empirical viewpoint. We will argue that the disagreement largely concerns the definitions and criteria of the concept of "altered state of consciousness". If "ASC" is defined as a state in which subjective experience and objective behavior are different from the normal waking state, then there is no doubt that hypnosis can be classified as an ASC, regardless of what brought these changes about or what the associated neurophysiological changes may be. We will also present our own recent findings on changes in auditory event-related potentials (the mismatch negativity component) during hypnosis and during hypnotic hallucination of music. According to our findings hypnosis in some cases may involve an altered state of consciousness in which automatic auditory discrimination processes are enhanced but at the same time their ability to engage attentional orientation seems to be diminished. This disconnection between the automatic orienting mechanisms and the focus of attention may be the neurocognitive basis of the "absorption" considered to be characteristic of the "hypnotic state" of consciousness.
Neural correlates of explicit sequence knowledge: a novel application of the process dissociation procedure
A. Destrebecqz, P. Peigneux, P. Maquet, C. Degueldre, A. Luxen, M. Van der Linden & A. Cleeremans
Arnaud Destrebecqz (firstname.lastname@example.org)
Cognitive Science Research Unit
UniversitÃ© libre de Bruxelles
50 ave. F.-D. Roosevelt
A central yet controversial issue in the study of implicit learning and memory is to determine the extent to which newly acquired knowledge can be projected onto behaviour without being available to consciousness.
To address this issue, Jacoby (1991) proposed the process dissociation procedure (PDP). This framework takes into account the fact that tasks in general are not "process-pure", i.e. that any experimental setting simultaneously involves explicit and implicit influences. The PDP is based on comparing performance in two similar tasks in which explicit and implicit influences act either in conjunction (the inclusion task) or in opposition (the exclusion task) according to the instructions.
On a behavioural level, this method makes it possible to derive separate quantitative estimates of implicit and explicit influences. On a brain imaging level, the PDP can be used to identify the brain regions exclusively involved in supporting explicit knowledge by excluding those also involved in implicit processing.
In a new adaptation of the PDP to sequence learning, participants were first trained on a serial reaction time (SRT) task and subsequently asked to freely generate similar (inclusion condition) and different (exclusion condition) sequences of key presses. To identify the brain regions specifically involved in each condition, rCBF was recorded in a PET experiment while participants performed the two generation tasks. Behavioural results indicated that sequence knowledge was in fact available to conscious awareness. Imaging results, based on a Condition [Inclusion vs Exclusion] by Performance (generation scores) interaction, revealed, at the voxel-corrected level, significant correlated rCBF responses located in the depth of both anterior cingulate (BA 32) and medial frontal (BA 10) gyri.
Distinguishing explicit and implicit represented parts of event sequences: evidence from event-related brain potentials
Birgit Stuermer, Friederike Schlaghecken & Martin Eimer
Birgit Stuermer (email@example.com)
Humboldt-University of Berlin
Biological Psychology / Psychophysiology
One major research interest in the area of sequence learning concerns the question wether acquisition can occur without concurrent awareness of sequential structures. The claim of implicit learning is mainly based on dissociations between performance increments in a serial reaction task and knowledge reported in memory tasks performed afterwards. However, fragmentary explicit knowledge available during training may be forgotten before the memory test is administered. To circumvent these questionable backward inferences an on-line measure of explicit knowledge during acquisition would be helpful. We considered event-related brain potentials (ERPs) to serve as such. In 3 experiments reaction times (RTs) were slowed when sequential rules were interrupted by deviants. Moreover, we found an enhanced N2b and P3b component for deviant items. While ERP effects were obtained only for those sequence parts that were identified as explicitly represented, RT effects turned out to be independend of perfomance in subsequent memory tests. We conclude that implicit and explicit learning co-exist in serial RT tasks and ERPs serve as a valid measure of explicit knowledge.
The striatum is involved in the successful implicit learning of statistical higher-order knowledge
Philippe Peigneux, Pierre Maquet, Arnaud Destrebecqz, Christian Degueldre, AndrÃ© Luxen & Axel Cleeremans
Philippe Peigneux (Philippe.Peigneux@ulg.ac.be)
Cyclotron Research Centre
Bat B30, Sart Tilman
This H2150 activation PET study investigated the neural substrates of the implicit processing of higher-order sequential information. Fourteen subjects were scanned during practice of a serial reaction time (SRT) task, in which a finite-state grammar probabilistically determined the succession of stimuli. Hence, subjects acquired progressively a higher-order knowledge of the statistical relationships between serially-ordered elements. PET data were modelled by a random-effect analysis which embodies both between- and within- subjects variability of the SRT performance. This approach allows statistical inferences to be made at the population level. Using this approach, we showed a significant voxel-level blood flow increase in the striatum (and regional increase in the middle and inferior frontal gyri) when behavioural SRT performance demonstrates successful learning of higher-order sequential probabilities. Our results suggest that the striatum is not only involved in the implicit automation of serial information through its participation to the cortical-subcortical motor loop linking prefrontal and caudate areas, but also that it plays a significant role for the selection of the most appropriate responses in the context created by both the current and previous stimuli, leading to better efficiency and faster response preparation in the SRT task.
Knowledge consolidation and abstraction in sequence learning
Annemie Melis, Eric Soetens & Wim Notebaert
Annemie Melis (firstname.lastname@example.org)
Department of Cognitive and Physiological Psychology
University of Brussels (VUB)
In previous explicit learning studies, we found that information is consolidated in the period after termination of the acquisition phase. We also found evidence for consolidation in implicit learning tasks such as sequence learning. First, with successive days of practice with a patterned sequence of stimulus locations, participants display a reaction time (RT) decrease on day 2 relative to their performance in the ultimate training block of day 1. Second, in sequence learning tasks in which participants trained one sequence pattern and switch to another mirrored pattern, RT continued decreasing only when the introduction of the transfer sequence was delayed. Presumably, participants where unable to apply sequence knowledge acquired before transfer to the new sequence after transfer when there is not enough time to consolidate this information. We investigated the relation between the minimum consolidation time and abstraction of the to-be-learned knowledge by manipulating both the similarity between the initial sequence pattern and the transfer pattern, and the delay with which the transfer pattern is introduced. The results seem to affirm the corresponding role of consolidation in sequence learning and in explicit learning. They do not support the distinction of explicit and implicit learning as independent systems.