Below is a listing of the 36 concurrent talks being presented at ASSC-XIII. Talks are listed in chronological order of presentation. Concurrent sessions will be held at Humboldt-Universität zu Berlin on the Saturday and Sunday afternoons of the meeting.
To Think or Not to Think? a Critique and Reappraisal of "Unconscious Thought Theory"
Axel Cleeremans, Universitè Libre de Bruxelles
Laurent Waroquier, Universitè Libre de Bruxelles
Marchiori David, Universitè Libre de Bruxelles
Olivier Klein, UPsyS
Saturday, June 6th, 14:00–14:30: Room 3096, Humboldt-Universität zu Berlin
Recent findings (i.e.,Dijksterhuis et al., Science, 2006) that complex normative decisions (e.g., choosing a car or an apartment) are best made without conscious deliberation have led to the idea that ''unconscious thought'' not only exists, but also often results in superior information processing, specifically when the required decisions involve the processing of many attributes. Here, we take issue both with the theoretical claims that underpin Dijksterhuis' ''Unconscious Thought Theory'' and with the relevant empirical findings. We report on five experiments (n = 529) inspired from the original design, in which participants were asked (1) to process information about cars by learning about their different attributes (e.g.,"The Hatsun has a powerful engine"), and (2) to choose the best car after given a chance to engage in deliberate, conscious thinking about the cars (''conscious thought'' condition) or after being distracted through performing an anagram solving task (''unconscious thought'' condition). Experiments 1, 2, and 3 respectively offered conceptual, identical, and methodologically improved replications of Dijksterhuis & al. (2006). We failed to find any evidence that decisions made after a period of distraction are better than after a period of conscious consideration. Experiment 4 showed that a majority of participants had in fact determined their attitudes towards each car before they engaged in the deliberation or distraction tasks; a finding that may explain the previous null results. Experiments 5 shed further light on the role that decision timing plays in decision quality, and suggested that conscious, deliberate decisions are more efficient than decisions made immediately or after a period of distraction when attitudes are not determined before engaging in the deliberation or distraction tasks. Taken together, these results simply suggest, contrary to previous claims, that conscious thinking tends to produce better decisions, a conclusion that should come as no surprise. While not denying that complex unconscious information processing exists, we conclude that it is not as powerful as previously claimed and that there is in fact no evidence for the idea that one can ''think'' without awareness.
On the Relationship Between the IIT of Consciousness and the Search for the NCC
Christof Koch, California Institute of Technology
Giulio Tononi, University of Wisconsin-Madison
Saturday, June 6th, 14:00–14:30: Room 2097, Humboldt-Universität zu Berlin
Integrated Information Theory (IIT) (Tononi 2004; Balduzzi & Tononi 2008) posits that consciousness is the irreducible property of complex networks of causally interacted components. IIT defines a fundamental quantity, integrated information Φ (expressed in bits), which captures precisely to what extent a system is more than a collection of parts. To that extent, it constitutes a complex, which by definition cannot be reduced to a set of independent modules. Φ can also be thought of as a measure of interconnectedness of a system. IIT provides a principled way of characterizing a quale, a particular conscious experience: the set of informational relationships that are simultaneously available within a complex when its mechanism is in a given state. Over the last two decades, brain scientists have initiated an empirical program based on searching for the Neuronal Correlates of Consciousness (NCC), the minimal set of neuronal mechanisms jointly sufficient for any one conscious experience (Crick & Koch 1995; Koch 2004; Tononi & Koch 2008). IIT emphasizes global properties of the brain while the NCC stresses local properties, such as particular neuro-anatomical connections or firing states. In this talk, we emphasize that these two different approaches are complementary rather than conflicting. In particular, we discuss the relationship between activity in specific brain regions (e.g. visual cortex) and the quale as defined by IIT, and how the meaning of any one consciously perceived stimulus will be affected by the loss of specific neuronal populations. Finally, we will discuss certain extensions to the way Φ can be computed and approximated for realistic biological networks.
A New Theoretical Framework for Agency, Ownership and Responsibility
Albert Newen, Ruhr-Universität Bochum
Saturday, June 6th, 14:00–14:30: Room 3075, Humboldt-Universität zu Berlin
The structure of the acting self is still perplexing and remains an often confounded issue in the recent debates. This talk provides a new systematic theory of self-consciousness in general and of two of its main features, the sense of agency and the sense of ownership. Furthermore, the phenomenon of responsibility can be separated from both phenomena. The theoretical framework will be shown to be fruitful in the context of the recent experiments in neuroscience including own studies worked out in cooperation with Vogeley. An important background of the new approach is a theory of levels of mental representation. It will be argued that we have to distinguish at least nonconceptual representations, conceptual representations and meta-representations. On the basis of clearly defined levels of representation it will be argued that one has to differentiate (i) an individual-orientated cognitive dimension of agency and ownership from (ii) a socio-normative dimension of responsibility. Gallagher introduced the distinction between agency and ownership. We need this distinction to account for passive movement of my arm since I still have the feeling of ownership in such a case but no feeling of agency. Furthermore, it is shown that we have to distinguish the feeling of agency and the judgment of agency. I can develop a feeling of agency in everyday automatic doings without explicitly judging that I am the agent. The feeling of agency is realized by nonconceptual representations. I may also develop a judgment of agency without any feeling of agency: The judgment of agency is realized by conceptual representations. Furthermore, it is shown that responsibility is a separate dimension from both aspects. I can judge that I am the agent of an action but deny responsibility by arguing that I just followed a strict order. This indicates that the ascription of responsibility is presupposes a theory of social interaction. Responsibility is relying on meta-representations which are typically involved in the so-called theory-of-mind ability. Analogous distinctions have to be made concerning the phenomenon of ownership. The proposed theory of self-consciousness is shown to be very fruitful from the perspectives of psychology, psychiatry and neuroscience.
Knowledge Applied to New Domains: The Unconscious Succeeds Where the Conscious Fails
Ryan Scott, University of Sussex
Zoltan Dienes, University of Sussex
Saturday, June 6th, 14:30–15:00: Room 3096, Humboldt-Universität zu Berlin
There is a common view that consciousness is needed for knowledge acquired in one domain to be applied to a novel domain. We present evidence of precisely the opposite; where the transfer of knowledge from one domain to another is achieved only in the absence of conscious awareness. Adopting higher order thought theory and exploiting subjective measures of consciousness we examine knowledge transfer using the artificial grammar learning (AGL) paradigm. In the standard non-transfer AGL task participants are initially exposed to strings of letters which, unbeknown to them, conform to complex grammar rules. Subsequently participants are informed of the existence of rules and required to distinguish the grammaticality of new letter strings. In this context participants reveal both conscious and unconscious knowledge; judgments made with confidence and those attributed to random selection both show above chance accuracy, with conscious knowledge being more reliable. Knowledge in all categories appears to derive from familiarity (Scott & Dienes, 2008). In the transfer variant of AGL, training strings are presented in one alphabet or modality and test strings presented in another. We examined three types of transfer: between one letter-set and another, between tone sequences and letter sequences, and between tone sequences and symbol sequences. For each test string participants reported whether they thought the string was grammatical, how familiar it felt, how confident they were in their grammaticality judgment, and their perceived basis for that judgment i.e. random selection, intuition, familiarity, rules, or recollection. An objective measure of fluency was also obtained using a timed perceptual clarification task. Results were consistent across all three conditions. Responses attributed to random selection showed above chance accuracy (60%) while those attributed to other categories did not (52%). Familiarity ratings were predicted by consistencies in the repetition structure of training and test strings and were hence related to grammaticality. Fluency, though increasing familiarity, was unrelated to grammaticality. While all judgments were predicted by familiarity ratings only those attributed to random selection showed a significant additional contribution of grammaticality. It appears that in knowledge transfer, as in visual perception (Marcel, 1993), the unconscious may outperform the conscious.
A New Method Convenient for Studying NCC-dn as Different from NCC-n, As Disentangled from Attention
Talis Bachmann, University of Tartu
Saturday, June 6th, 14:30–15:00: Room 2097, Humboldt-Universität zu Berlin
The nasty obstacles for productive research on NCC include (i) confounds of attention and awareness effects, (ii) confounds of subjetive awareness variability with objective stimulus variability, (iii) anisochrony or temporal uncertainty of NCC signatures of awareness-necessary brain processes with regard to the corresponding direct phenomenal experiences in real time, (iv) spatial overlap between the target stimuli for measurement of their awareness and the interfering stimuli used to cause variation in target awareness (eg, as in masking, attentional blink, binocular rivalry). The best methods so far free from some of these shortcomings are continuous flash suppression in rivalry, standing wave illusion of invisibility, motion-induced blindness, and visual afterimages (e.g., Lou, Tsuchiya, Koch, Blake, Bonneh, et al.). Here, I present a new method that is free of all the above mentioned problems and demonstrate it on screen. Subjects adapt to different coloured objects; afterimages are created. At the onset of the otherwise blank display used as a background for afterimages experience, a word signifying a target colour is presented right near the continuously visible fixation dot. While continuing to keep visual fixation, subjects have to report whether the target-coloured afterimage object disappeared before other objects or not. Spatially attended target-couloured afterimages tended to disappear first more often than other afterimage-objects when compared to the theoretically expected chance frequency. Thus phenomenal experiences counteracting attention, free of direct presence of the objective sensory input, spatially separated for convenient selective brain-scanning of cortical activity from the corresponding receptive field, and clearly markeable in terms of phenomenal onset and offset, are brought about. Importantly, this method is a little step forward in helping to distinguish between NCC that could reflect necessary prerequisites for awareness, but uncertain in terms of direct timing of the emergence and disappearance of phenomenal experiences (NCC-n) and NCC-dn that could be a sign of necessary prerequisites for awareness and also bear one-to-one correspondence with phenomenology in real time.
Enactivism and Bodily Sensations
Frederique de Vignemont, CNRS/NYU
Saturday, June 6th, 14:30–15:00: Room 3075, Humboldt-Universität zu Berlin
The enactive approach aims at providing a unified account of perceptual experiences in terms of bodily activities. Most enactive arguments come from the analysis of visual experiences (No', 2004), but there is one domain of consciousness where the enactive theses seem to be less controversial, namely, bodily sensations. On this view, bodily sensations consist in knowledge of sensorimotor contingencies. For instance, feeling one’s arm touched would be equivalent to know that if one presses upon the touching object, the tactile signal will increase. One can contrast the enactive view of bodily sensations with the dominant neuropsychological theory that draws a sharp distinction between the body schema and the body image (Paillard, 1999; Gallagher, 2005). According to the body schema/body image theory, bodily information related to action does not participate to bodily experiences. There are two distinct systems, which can be dissociated in some rare pathological cases like deafferentation and numbsense. It was shown indeed that patients with numbsense do not feel tactile sensations but are still able to point where they have been touched (Rossetti et al., 1995). In contrast, deafferented patients can identify their body part, but have difficulties in reaching it (Paillard, 1999). Here I shall analyse the relationship between bodily experiences and action. I shall show what kind of role action can play in bodily experiences, and what kind of role action cannot play. In order to do so, I shall distinguish between action-orientated and action-grounded body representations.
The Role of Stimulus Awareness in the Neural Computations of Value for Simple Choices
Milica Milosavljevic, California Institute of Technology
Julien Dubois, California Institute of Technology
Antonio Rangel, California Institute of Technology
Christof Koch, California Institute of Technology
Saturday, June 6th, 15:00–15:30: Room 3096, Humboldt-Universität zu Berlin
Several human fMRI and monkey electrophysiology studies showed that neural activity in the orbitofrontal cortex (OFC) correlates with behavioral measures of subjective values, known as a goal-value (Roesch, & Olson, 2004; Padoa-Schioppa, & Assad, 2006). No previous studies have compared how the computation of goal values changes with awareness of the stimulus being evaluated. We use continuous flash suppression (CFS) - subjects’ dominant eye receives dynamic images consisting of randomly generated patches of flickering colors, while the non-dominant eye is shown the target stimulus that is to remain subliminal (Tsuchiya & Koch, 2005). fMRI scans are acquired using a 3T-Trio scanner with 8-channel head coil. Before scanning, subjects provide liking ratings for 30 images of food items. In the scanner, on each trial, hungry subjects make a Y/N decision about whether or not they want to eat the food shown for 2s. After the experiment, one trial is randomly selected and the subject is given the food chosen on that trial. The key manipulation is that on half of the 210 trials, subjects are conscious of the stimulus, but half of the time the food image is hidden using CFS. In control trials, no food image is shown, and only noise is perceived. Subjects are told that even if they do not see a food item they should make their best guess (choice) based on how they feel. The preliminary data (p<0.05 uncorrected, and an extent threshold of 5 voxels) suggests that activity in the medial OFC correlates with values during both conscious and unconscious trials. Interestingly, the locus of activation significantly overlaps with those correlating with goal values during conscious presentations found in previous studies. The pilot data further suggests that the value signal is stronger on average during the conscious trials than during the unconscious trials. A similar result holds for the difference between areas encoding for value in conscious and unconscious trials, but at a lower level of statistical significance. The results suggest that the brain may be able to compute value of choice options even when subjects are not aware of the choice options.
Functional Connectivity During Propofol-Induced Unconsciousness
Gabriele Lohmann, Max Planck Institute
Matthias Schroeter, MPI Human Cognitive and Brain Sciences
Christian Schwarzbauer, MRC, Cambridge
Daniel Margulies, Berlin School of Mind and Brain & Max Planck Institute, Leipzig
Wolfgang Heinke, University of Leipzig
Saturday, June 6th, 15:00–15:30: Room 2097, Humboldt-Universität zu Berlin
We present a functional magnetic resonance study with 12 normal human volunteers in which we compared functional correlations within a network of human brain areas before and during propofol-induced unconsciousness. We focus on functional connectivity between brain regions implicated in the 'default mode' network. Due to the association of these regions with self-representation, episodic memory processes and consciousness, we hypothesize this network to be strongly affected by anesthesia. The fMRI study was conducted at the Max-Planck-Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, in collaboration with the Department of Anaesthesia and Intensive Medicine of the University of Leipzig (Heinke et al. 2004, British Journal of Anaesthesia, 92:641-650). The experiment consisted of three blocks lasting 10 minutes each during which normal sentences and pseudo-word sentences were presented in random order. Subjects were instructed to give a yes/no button-press response to indicate whether a sentence was made up of pseudo-words or not. Block 1 was the awake state and served as the control state. During block 2, anaesthesia was induced with 3 mg kg-1 propofol administered over a period of 10 minutes in order to achieve a slow transition from wakefulness to unconsciousness. In block 3, anaesthesia was maintained with 3 mg kg-1h-1. During the second block, all subjects ceased responding, and remained unresponsive during the third block. We compared functional connectivities during the first and the third block of the experiment (before and during anaesthesia) in six manually defined regions of interest (ROIs), namely precuneus, retrosplenial cortex, temporoparietal junction, thalamus, hippocampus, and the anterior frontomedian cortex. We extracted low-frequency components from the fMRI data and computed pairwise correlations among the ROIs. We applied Wilcoxon's test for matched pairs to test for differences in correlations between blocks 1 and 3. We found that during unconsciousness functional correlations between areas of the association cortex were largely preserved while thalamocortical and hippocampal correlations were significantly reduced. Since the hippocampal network is known to be vital for memory processes, we conclude that it may be possible that subjects still maintain self-representation but are unable to integrate it into their episodic memory.
Towards a General Theory of Action Awareness
Myrto Mylopoulos, Graduate Center, CUNY
Saturday, June 6th, 15:00–15:30: Room 3075, Humboldt-Universität zu Berlin
In this paper, I take steps towards articulating and defending a general theory of bodily action awareness. This theory is sensitive to the view that bodily actions are typically multistage events unfolding in time. They involve two causally related components: intention and bodily movement. I argue that a proper appreciation of the dual component ontology of action yields a correspondingly multifaceted account of action awareness. According to this account, what I call 'general action awareness' may be broken down into two types: 'Anticipatory awareness' and 'concurrent awareness'. Anticipatory awareness is awareness of what one will do, and corresponds to the intention component of action. Concurrent awareness is awareness of what one is presently doing and corresponds to the bodily movement component of action. General action awareness is a function of the degrees of each of these two types of action awareness, which, I will argue, are integrated on the basis of an inferential mechanism. In the first part of the paper, I lay out the account sketched above in further detail. In the second part of the paper, I evaluate the account against competing accounts of action awareness. I argue that the account I defend has theoretical advantages over alternative accounts in that it (i) better explains pathological data from cognitive neuroscience and psychology, (ii) better explains salient experimental results in the literature, i.e., Marcel's vibro-tactile illusion and Haggard's intentional binding effect, (iii) better explains the sense of agency and its various components, i.e., the sense of ownership and sense of control that we have over our own actions, and (iv) is equally capable of accounting for the intuitions that action awareness is ''immediate'', ''authoritative'', and ''transparent'' (see O'Brien, 2003).
Sensitivity and Subjective Awareness Increase with Practice in Metacontrast Masking
Caspar Schwiedrzik, Max Planck Institute for Brain Research
Wolf Singer, Max Planck Institute for Brain Research
Lucia Melloni, Max Planck Institute for Brain Research
Saturday, June 6th, 16:00–16:30: Room 3096, Humboldt-Universität zu Berlin
Can practice effects on unconscious stimuli lead to awareness? Can we learn to see? Recent evidence suggests that blindsight patients trained for an extensive period of time can learn to discriminate and consciously perceive stimuli in their blind hemifield. So far, it is unknown whether such learning effects generalize to normal observers. Here we investigated practice effects in metacontrast masking. Objective thresholds were assessed individually as a function of stimulus onset asynchrony (SOA). Subjects were then trained for five consecutive days on the SOA that resulted in chance performance. Our results show a linear increase in sensitivity (d') but no change in bias (c) for the trained SOA. This practice effect on sensitivity spreads to all tested SOAs, but only partially to another stimulus location. The effects are still detectable after several months without exposure to the stimuli. Additionally, we show that subjects rate their perceptual awareness of the target stimuli differently before and after training, exhibiting not only an increase in sensitivity, but also in the subjective awareness of the percept. Preliminary data indicates an important role of feedback in the development of subjective perceptual awareness over the course of learning. We conclude that subjects can indeed learn to see.
Resting State fMRI Connectivity Reflects the Level of Consciousness During Anesthesia
Melanie Boly, University of Liège
Audrey Vanhaudenhuyse, University of Liège
Andrea Soddu, University of Liège
Vincent Bonhomme, CHU Liège
Christope Phillips, University of Liège
Pierre Maquet, University of Liège
Jean François Brichant, Anesthesiology, CHU Liège
Michael Greicius, Stanford University
Steven Steven, Cyclotron Research Centre, Liège
Pierre Boveroux, Cyclotron Research Centre
Saturday, June 6th, 16:00–16:30: Room 2097, Humboldt-Universität zu Berlin
Introduction: Mechanisms of anesthesia-induced loss of consciousness remain poorly understood. Our hypothesis was that anesthesia-induced loss of consciousness would be reflected in resting state fMRI connectivity impairment in higher order associative networks (the default mode network or DMN, and the paired left and right executive control networks or ECNs), compared to a relative preservation of connectivity in low level sensory cortices. Methods: Here we measured low-frequency spontaneous BOLD fluctuations during wakefulness, mild sedation, deep sedation with clinical unconsciousness, and subsequent recovery of consciousness in 12 healthy subjects. Data analysis investigated resting state fMRI connectivity in the DMN, the bilateral ECNs, and the auditory and visual networks using two different types of analyses, i.e. correlation analysis and independent component analysis. Group-level comparisons between normal wakefulness and sedation stages were performed using random effects analyses. Results were thresholded at p<0.05 corrected for multiple comparisons using false discovery rate. Results: We show that anesthesia-induced loss of consciousness correlates with a global decrease in connectivity in fronto-parietal networks - i.e. the DMN and ECNs - linked with a loss of thalamo-cortical connectivity in these networks. For both the DMN and ECNs, the most consistent finding linking connectivity and consciousness was the involvement of the thalamus in these networks. Anticorrelations between the DMN and lateral fronto-parietal cortices also decreased proportionally with the loss of consciousness across sedation stages. In contrast, even during deep sedation, we found globally preserved connectivity in low-level sensory cortices – the auditory and visual networks - along with preserved thalamo-cortical connectivity in these networks. Waning consciousness was, however, associated with a loss of cross-modal interactions between the visual and auditory networks - suggesting a loss of top-down control. Conclusions: Our results shed light on the functional significance of fluctuations in spontaneous brain activity observed in fMRI. They suggest that anesthesia-induced loss of consciousness could be linked to a breakdown of cerebral temporal architecture, which modifies both within- and between-network connectivity. They also emphasize the importance of thalamo-cortical connectivity in higher order associative networks in the genesis of conscious perception.
Somatoparaphrenia and Immunity to Error Through Misidentification
Timothy Lane, National Chung Cheng University
Caleb Liang, National Taiwan University
Saturday, June 6th, 16:00–16:30: Room 3075, Humboldt-Universität zu Berlin
In reflecting on self-consciousness, many philosophers have argued that there is an important distinction to be made between self-as-object and self-as-subject. The idea is that the distinctive character of self-consciousness is associated with the latter, not the former. It is also widely accepted that the notion of self-as-subject is substantially elucidated by Shoemaker’s immunity principle (IEM): when ''I'' is used as a subject we are ''immune to error through misidentification relative to the first-person pronoun'' (1968 and 1996). According to IEM, it would be impossible for a person who introspectively knows someone is in pain to be mistaken as to whether it is he who is in pain. In this paper we examine IEM in light of a pathological phenomenon, somatoparaphrenia (e.g. Baier and Karnath 2008; Bottini et al. 2002; Moro et al. 2004, and Vallar and Ronchi 2009). First, we argue that somatoparaphrenia, at least in some cases, involves more than just problems pertaining to ownership of body parts. Adequate description of the pathological phenomenology requires distinguishing between feeling that one owns a body part and feeling that one owns an experience. Second, we argue that some cases of somatoparaphrenia constitute genuine counterexamples to IEM, because they show that the presence of an experience and the sense of ownership of that experience are dissociable. Since the two are dissociable, it would be possible for a person who introspectively knows someone is in pain to be mistaken as to whether it is he who is in pain. Third we assess prior empirical attempts at refuting and defending IEM. Among other things, we argue that the distinction between agency and ownership proposed by Gallagher (2000, 2004) does not help to save IEM, and that the dual-component view of ownership proposed by Campbell (2002, 2003) is unable to accommodate the experience of somatoparaphrenia. Finally, we consider possibilities for revising IEM so that it might be treated as a substantive empirical hypothesis, one that can actively guide research.
When Seeing Outweighs Feeling: Integration of Somatic and Visual Information in Affective Blindsight
Silke Anders, University of Lübeck
Saturday, June 6th, 16:30–17:00: Room 3096, Humboldt-Universität zu Berlin
Affective neuroscience has been strongly influenced by the view that a feeling is the perception of somatic changes and has often neglected the neural mechanisms that underlie the integration of somatic and other information in affective experience. We investigated affective processing in nine cortically blind patients. In these patients, unilateral postgeniculate lesions prevent primary cortical visual processing in one part of the visual field which, as a result, becomes subjectively blind. Residual subcortical processing of visual information, however, is assumed to take place in the entire visual field. These patients show significant startle reflex potentiation when a threat-related visual stimulus is shown in their blind visual field. This is associated with an increase of brain activity in somatosensory-related areas, and an increase in self-reported negative affect. However, when the visual stimulus is presented in their sighted visual field the patients show a remarkable dissociation of somatic response and phenomenal experience of affect: Despite the fact that startle reflex potentiation was similar when the visual stimulus was shown in the blind or sighted visual field, patients reported significant less negative affect during stimulation of the sighted visual field. This decoupling of phenomenal affective experience and somatic changes was associated with an increase of activity in left ventrolateral prefrontal cortex (VLPFC) and a decrease of affect-related somatosensory activity. These findings demonstrate that similar affective somatic changes can be associated with different phenomenal experiences of affect depending on the level of cortical processing. Tentatively, we suggest that this decoupling of somatic responses and experienced affect and the reduction of negative phenomenal experience observed can be explained by a context-dependent VLPFC-mediated inhibition of affect-related somatosensory activity.
Into the Scanner, Out of the Body: Neural Correlates of Ketamine-induced Alterations in Body Perception
Marco Benz, University Hospital of Psychiatry Zurich
David Andel, University Hospital of Psychiatry Zurich
Franz X. Vollenweider, University Hospital of Psychiatry Zurich
Saturday, June 6th, 16:30–17:00: Room 2097, Humboldt-Universität zu Berlin
In humans, the N-methyl-D-aspartate (NMDA) receptor antagonist ketamine produces anaesthesia at high doses, while at lower doses it causes profound alterations in consciousness. These may include changes of space and time perception, depersonalisation, derealisation, altered affect and cognition as well as hallucinatory phenomena. Alterations in body perception are also commonly reported; these may range from ''floating'' sensations to pronounced body distortions or the subjective experience of the self being located outside the body. Such alterations and their neural underpinnings were the focus of the present analysis. We examined the effects of an intravenous subanaesthetic dose of (S)-ketamine on subjective experience and regional cerebral blood flow (rCBF) in 30 healthy human volunteers. Positron-emission tomography (PET) was used to measure rCBF as an index of brain activity. The 5D-ABZ questionnaire (Dittrich, Lamparter and Maurer, 1999), a visual-analog rating scale for characterizing altered states of consciousness, was used to assess subjective experience. We extracted three questions from the 5D-ABZ which explicitly address alterations or loss of body perception and which refer to subjective events commonly reported during out-of-body experiences (OBEs): 1) the feeling of being body-less, 2) the feeling of being located outside of one's own body, 3) the feeling of floating. Scores on these three items were summed and correlated with rCBF. We found a significant correlation between cortical activity in the right temporoparietal junction (TPJ) and the magnitude of alterations of body perception. This finding is in line with previous reports implicating the TPJ in out-of-body experiences and suggests a central role for this region in mediating the experience of the self and body in space. Furthermore our result supports an involvement of NMDA receptors and glutaminergic neurotransmission in OBEs. A ''ketamine model of the OBE'' is well compatible with the hypothesis of ''disintegration at the TPJ'' (Blanke et al., 2002) and also the dissociational theory (Irwin, 2000) of out-of-body experiences.
A Defense of Perceptual Accounts of Pain
Verena Gottschling, York University
Saturday, June 6th, 16:30–17:00: Room 3075, Humboldt-Universität zu Berlin
Perceptual accounts of pain (PAP) are very popular these days. Nonetheless, they are also under heavy attack, because these accounts seem to be vulnerable to certain severe challenges and objections, and incompatible with our commonsense and scientific understanding of pain. Recently, nice versions of these challenges for PAP were presented (Aydede 2006, forthcoming), the initial and the labeling argument. These will be my focus. As I argue, both arguments against PAP are based on a number of premises, which are dispensable for PAP proponents. Therefore, these objections propose problems for only certain classes of perceptual accounts; as an argument against PAP in general, they fail.They also rely on unfortunate assumptions about the nature of perceptual accounts. These include (a) that perception necessarily ends in conceptual content and belief formation, (b) that genuine perception necessarily involves sensory conceptual components, and (c) that the objects of perception is the feeling of pain. I will end by proposing a strong perceptual account of pain that avoids the mentioned problems and is more compatible with several empirical data. I claim that to be in pain is to be in a complex state. The components of this state include certain sensational components (a number of related qualia), as well as affective attitudes. But these components are not simply additive components. Pain is seen as involving non-psychological as well as psychological components. When subjects experience pain they are in a certain sort of psychological state towards a bodily state or condition. Pain is constituted by the bodily state though, not by the psychological state towards this bodily state. So I argue for a perceptual theory that identifies pain states with parts of our internal representation of our body’s physiological condition, a representation that registers physiological imbalance, and potential tissue damage. A perceptual process creates a pain-quale, but the object of the perception is the pain (a bodily state), not the feeling of pain. In other words, I claim that the pain is the object of perception rather than that pain is the experience of the object of perception, in contrast to most PAP accounts.
Loading Working Memory Can Reduce Inattentional Blindness
Nilli Lavie, University College London
David Carmel, NYU
Jake Fairnie, University College London
Saturday, June 6th, 17:00–17:30: Room 3096, Humboldt-Universität zu Berlin
We present a series of experiments testing novel predictions about distractor awareness derived from Load Theory of attention and cognitive control (e.g. Lavie et al., 2004). Load Theory suggests that working memory (WM) exerts executive control over attention by actively maintaining the prioritisation of relevant information and thus preventing the processing of irrelevant distractors. Loading WM consequently results in greater processing of irrelevant distractors (e.g. DeFockert et al., 2001). This led to the prediction that increasing WM load should increase the rate of distractor intrusions into conscious awareness. Hence loading working memory can reduce (rather than increase, c.f. Marios, et al., 2007) inattentional blindness. We measured distractor awareness using a modified inattentional blindness paradigm. Subjects performed a famous-name classification task (singer/politician) under either high or low WM load. On the final critical trial, a famous response-competing distractor face was presented. Awareness for this distractor significantly increased under high WM load. Additional experiments generalised this result to cases where the distractor faces were entirely task-irrelevant (appearing following a tool-name classification task) and clarified that the effect was confined to faces, regardless of whether they were famous or anonymous (awareness of task-irrelevant familiar buildings was unaffected by the level of WM load). These findings support our hypothesis regarding the role of executive control in suppressing distractors from awareness. Further, the results resolve previous discrepancies regarding the effects of WM load on task irrelevant processing (e.g. Marois, et al., 2007; Yi et al., 2004) in terms of the distracting potency of different task-irrelevant stimuli.
Disentangling the Automatic from the Conscious Brain: fMRI of the Vegetative State
Martin Monti, MRC Cognition & Brain Sciences Unit
Martin Coleman, Cambridge Impaired Consciousness Study Group
Adrian Owen, MRC Cognition and Brain Sciences Unit
Saturday, June 6th, 17:00–17:30: Room 2097, Humboldt-Universität zu Berlin
Recent evidence has suggested that functional neuroimaging may play a crucial role in assessing residual cognition, and awareness, in brain injury survivors. In particular, brain insults that compromise the ability of patients to produce motor output may render standard clinical testing ineffective. It is in these situations that functional neuroimaging may provide a window on brain function without requiring any behavioral response by the patient. However, using brain activity data to assess residual cognition, and especially awareness, presents many of the same complexities faced by bedside testing when teasing apart reflexive from voluntary behavior. Can automatic brain responses be separated from willful mental effort? To address this problem, we propose and present data from a hierarchical approach to patient assessment, based on fMRI. Using this methodology, different sensory modalities are tested at increasing levels of complexity. At the lowest level, basic sensory perception is probed (e.g. sound perception, responses to the presence or absence of light). In an increasingly complex series of tasks we then follow neuro-cognitive systems along their processing stream. In the case of vision, for example, color perception, motion, and object recognition are sequentially tested. At the highest level of the hierarchy, this approach probes for the ability to willfully modulate brain activity (e.g. top-down allocation of attention). From a methodological standpoint, we argue that carefully designed experiments may allow the identification of volitional activity on the basis of brain activation alone. In particular, when 'task' and 'baseline' epochs are perceptually identical, and differ only in terms of their respective instructions, it is impossible to interpret differential brain activation without assuming a willfully different "mind-set" across the two conditions.
Representationalism and Hallucinatory Pain
Manolo Martinez, Universitat de Barcelona
Saturday, June 6th, 17:00–17:30: Room 3075, Humboldt-Universität zu Berlin
An important objection to representationalism about qualia is that, while we readily recognise the possibility of perceptual hallucination, we do not acknowledge cases where it seems that we are in pain but in fact we are not. A representationalist account of pain must explain the asymmetry between pain and perceptual experiences in this respect. A common-sense answer to this Asymmetry Problem is that pain, independently of its success at representing the external world, is unpleasant. This, the explanation goes, is why our concept ''pain'' tracks the experience itself instead of whatever it represents. It is an as yet unmet challenge for representationalism to explain how the unpleasant aspect of pain phenomenology is to be accounted for in terms of representational content. In particular, I show that a causal-informational account of representation such as Tye's will not do: under such an account there is no way to tell apart the purported two contents of pain representations (tissue damage and badness) given that they are necessarily coinstantiated in the conditions in which content is fixed. Imperative content provides a better answer to the Asymmetry Problem: the whole content of perceptions can be veridical or non-veridical, but not the whole content of pain can; the imperative part (upon which unpleasantness supervenes) is neither veridical nor non-veridical, just as desires and imperative sentences are neither true nor false. Imperative content is not supposed to correspond to the facts at all. Thus, it makes no sense to talk of hallucinating the negative phenomenology of pain, hallucinations being a special kind of non-veridical experiences. The unpleasant side of phantom-limb pain is as fine from the veridicality point of view as the unpleasant side of limb pain is. If we are to take representationalism seriously, this kind of content is our best shot for the representational content upon which the negative affective phenomenal character of pain is to supervene. Finally, I canvass an account of imperative representations that is naturalistically acceptable and, I hope, acceptable by Tye in particular.
Noise Masks Noise: Masking by Illusory Objects Reveals the Dynamics of Metacontrast Masking
Jerome Sackur, Ecole Normale Supèrieure
Sunday, June 7th, 14:00–14:30: Room 3096, Humboldt-Universität zu Berlin
Metacontrast is a form of visual backward masking whereby conscious perception of a target is reduced when it is followed by a surrounding mask. Here I first show that metacontrast can occur between objects defined by higher order properties, such as texture and movement. Second, I show that Metacontrast is even possible between two illusory objects: When a portion of a random texture is replaced by a new version of the same texture, brief illusory contours are seen, without any luminance or chromatic transients. I show that transient objects defined in such a way are subject to, and can induce, metacontrast masking. An illusory notched square is masked by an illusory square annulus that abuts the central square without overlap. At short Stimulus Onset Asynchronies (SOAs), discrimination of the position of the notch is nearly perfect. Performance drops dramatically at intermediate SOAs (40–60ms), but then increases again and plateaus at longer SOAs. Such a U-shaped function is typical of metacontrast. This new and greatly simplified stimulus permits an analysis of the dynamics of metacontrast. The two parts of the metacontrast U-shaped function result from the combination of two conflicting forms of integration: static integration which creates a single static object and dynamic integration (apparent movement) which is detrimental to discrimination performance on the target. At short SOAs static integration is powerful, perception of the discriminative feature of the target is good. As SOA increases, apparent movement progressively takes precedence over static integration, and perception of the target as such is impaired. Finally, at even longer SOAs, there is no integration, and the target and the mask are clearly seen as two separate objects.
"I Am a Cursor": First and Third Person Accounts of Operating Brain-computer Interfaces
Gabriel Curio, Charitè - Universitätsmedizin Berlin
Benjamin Blankertz, Computer Science, TU Berlin
Klaus-Robert Mueller, Computer Science, TU Berlin
Sunday, June 7th, 14:00–14:30: Room 2097, Humboldt-Universität zu Berlin
'To make matter move just by thinking about it' Ð yesterday's fiction has been turned into today's science through brain-computer interfaces (BCIs). Here, we survey our experiences of operating the non-invasive EEG-based Berlin BCI (www.bbci.de) which combines human cortical neurophysiology with Machine Learning algorithms to enable effective first-session performances of BCI-na•ve subjects. The BBCI principle works in three steps: (1) EEG activity is recorded during a series of pre-defined intended movements. (2) Computer programs extract user- specific 'thought-related' EEG patterns; these algorithms use diverse EEG signs of intended movements, such as slow 'readiness potentials' or the movement-related attenuation of sensorimotor 'idling rhythms'. Notably, studies of long-term arm amputees show that such EEG signs are usually preserved when patients engage to move their 'phantom hand'. (3) These EEG patterns are categorized to control devices like computer cursors, 'mental typewriters', gaming applications, and virtual prostheses. Importantly, the BBCI permits also the single-trial real-time detection of non-motor mental states, e.g., mental rotation (as required when playing a BCI- version of Tetris). While these findings were quantified rigorously constituting a classical scientific 'Third Person Account' of operating a BCI, an elementary collection of qualitative aspects should not be missed representing 'First Person Accounts' reported by subjects who experienced sessions with long periods of excelling BCI classification accuracies. First, the initial 'limb-focus' of the motor intention could be fading and gradually replaced by the impression to mind-control the particular effector device 'directly'. Second, if such direct, reliable and 'effortless' BCI control was maintained for several minutes, a temporary sense of bodily ownership could develop such that the effector (e.g., a screen cursor) was perceived as belonging to the I-domain of the BCI operator ('I am a cursor'), eventually waning along with EEG nonstationarities. Finally, a BCI operational state could ensue in which a wish to move might be executed prior to its mental completion. In conclusion, pursuing the line of Libet's legacy at this novel mind-brain-computer interface we posit that upcoming, possibly perfect BCIs operating at 'thought-speed' in real-time will require a refined understanding of the neurophysiological underpinnings of will, veto and responsible acting.
The Higher-order Global State (HOGS) Model of Consciousness
Robert Van Gulick, Syracuse University
Sunday, June 7th, 14:00–14:30: Room 3075, Humboldt-Universität zu Berlin
Higher-order theories of consciousness (Rosenthal, Lycan and Armstrong) and global workspace or integration theories (Barrs) are normally regarded as distinct and competitive alternatives. Moreover, both are typically regarded as more plausible as accounts of access consciousness rather than as accounts of phenomenal consciousness. Access consciousness (Block) is defined in terms of availability to report and wide inferential application. Information broadcast in the global workspace can be accessed by many mental system and modules, while the presence of a higher-order thought or perception directed at a first-order state makes it accessible for report. However, it is less clear why presence in the global workspace or the addition of a higher-order thought should give rise to the sort of ''what-it is-likeness'' that is regarded as definitive of phenomenal consciousness. I propose an alternative model that combines key aspects of the higher-order and global integration theories. The Higher-Order Global State (or HOGS) model is not a mere conjunction, but involves re-conceptualizing both of the prior theories in a way that reveals their underlying connections and provides some insight into the phenomenal aspect of consciousness. On the HOGS model the higher order aspect of consciousness is not located in explicit meta-states that are separate from their lower-order mental objects. The higher-order aspect is instead an implicit reflexive feature of the relevant globally integrated states. The transformation of a non-conscious mental state M into a conscious state is accomplished not by intentionally directing a separate and explicit meta-state upon M, but by recruiting M into a globally integrated state that implicitly incorporates a significant meta-intentional aspect in its content and organization. The meta-intentional aspect is directly linked to the structure of phenomenal experience, especially to the duality of subject and object. The structure of experience is that of a world of objects experienced from the perspective of the experiencing subject or self. The experiential self in turn is present in experience not as another object but as the unified point of view to which objects are present.
Cued Masks in the Change Blindness Experiment: Evidence for Overflow or Hyperillusion?
Tamás Bózsa, Technical University of Budapest
Zoltán Jakab, Technical University of Budapest
Sunday, June 7th, 14:30–15:00: Room 3096, Humboldt-Universität zu Berlin
In defending (and also qualifying) the idea that phenomenal consciousness is independent of access consciousness, Ned Block argues that mental states that are phenomenally conscious need not be present in the global workspace. Some information that is near the threshold of entering the workspace, but still requires amplification by mechanisms of attention may also be phenomenally conscious, Block argues. Thus in some experiments where the subjects have the impression that they see more than what they can report, a broader range of mental states is phenomenally conscious than what is present in the global workspace. In this sense, phenomenal consciousness overflows access consciousness. A rival hypothesis by S. Dehaene et al. is that in such cases there arises a hyperillusion: it seems to the subject that she is phenomenally conscious of certain objects and properties whereas in fact she is not. We ran a version of the change blindness experiment which included pairs of pictures differing in some detail, but the part making the difference in a given pair was also embedded in a Fourier mask that flashed between the members of the pair. Mask onset time and contrast of the embedded cue were varied. We took it that if the embedded cue helped subjects to spot the difference (by reducing the number of repetitions needed for the correct response) without the subjects noticing it (and subsequently being able to report it), then we would have preliminary evidence for phenomenal consciousness without the phenomenally conscious item being present in global workspace. However, we found that the embedded cues speeded up subjects' responses when and only when they indicated noticing them at the end of the experiment. Cues that went unnoticed did not help, which appears to support the hyperillusion hypothesis. Another possibility is that subjects did fixate on the critical part of some of the pictures and these parts also became phenomenally conscious, but they exited working memory before being compared to the corresponding part of the other picture. Following up this experiment by eye-movement tracking can help choosing between these hypotheses as we shall argue.
Instruction Following in Disorders of Consciousness: A Time-frequency EEG Approach
Manuel Schabus, Cyclotron Research Centre, Liège
Fabien Perrine, UMR520, CNRS-UCBL, Lyon, France
Caroline Schnakers, Cyclotron Research Centre, Liège
Steven Laureys, Cyclotron Research Centre, Liège
Sunday, June 7th, 14:30–15:00: Room 2097, Humboldt-Universität zu Berlin
One of the most important clinical challenges in patients with severe brain damage is to estimate their residual sensory awareness. We have previously shown that patients in a minimally conscious state (MCS) and some patients in a vegetative state (VS) exhibited larger P3 event-related potential (ERP) responses to their own first name, as compared to 7 unfamiliar equiprobable first names (Perrin et al., 2006). Although the data suggested that these patients were still able to process salient semantic information, the passive paradigm cannot give sufficient evidence for conscious brain activity as a P3 response to the own name can also be observed in sleep (Perrin et al., 1999) and anaesthesia (Davis et al., 2007). We therefore extended the previous paradigm with an additional - but crucial - condition in which subjects were asked to actively count one specified unfamiliar first name. Furthermore, we decided to not only rely on ERPs which reflect strictly phase-locked neuronal responses (i.e. evoked cerebral activities which appear with a constant latency after the stimulus). As cerebral functions of brain damaged patients are often distorted and delayed in time the analysis of induced responses (i.e. non phase-locked responses) may help to better characterize residual cerebral activity. Specifically we collected data from 22 patients diagnosed as VS (n= 8) or MCS (n= 14) as well as 12 age-matched controls. In a complementary time-frequency analysis to Schnakers et al. (2008) we found signs for EEG synchronization changes in MCS (but not VS) to ''to-be-counted'' target and own names, suggesting voluntary compliance to task instructions and therefore residual conscious processing. Yet, the observed EEG changes were rather untypical in nature as they appeared as desynchronization in a low frequency range (theta band), as opposed to classically seen alpha desynchronizations and delta/theta synchronizations in response to cognitive demands (for review see Klimesch, 1999). Given the present data it appears that MCS but not VS patients do show reliable residual cognitive processing. We believe that time-frequency analysis might offer an unexploited alternative to classical ERP analysis which in some cases might even be more sensitive for detecting awareness in disorders of consciousness.
The Phenomenal Character and Epistemic Status of Unreported Perceptual Stimuli
Guven Guzeldere, Duke University
Sunday, June 7th, 14:30–15:00: Room 3075, Humboldt-Universität zu Berlin
The phenomenal character and epistemic status of unreported perceptual stimuli are controversial issues among both philosophers and cognitive psychologists/neuroscienists, though they often take distinctly different approaches to the dispute. There are several reasons why a perceptual stimulus may go unreported (if not unreportable): threshold conditions about the stimulus properties (e.g., low luminosity, brief exposure duration), processing constraints imposed on the perceiver (e.g., attentional load), and, possibly, subject's lack of relevant conceptual resources. These different reasons and conditions sometimes get confounded in the literature. The aim of this presentation is to show how and why such perceptual cases with different etiologies, even if they all yield a similar outcome (unreported stimuli), require different characterizations of phenomenal content and different explanations about their epistemic status. The central disagreement among philosophers is about whether unreported stimuli ever become part of the subject's phenomenal experience, even if they remain hidden from her cognitive access. For example, Dretske maintains that the content of phenomenal consciousness often outstrips what can be accessed or reported and this is a special case of a prevalent phenomenon he calls Ònon?epistemic seeingÓ, while Dennett claims such experiences, in the absence of cognitive upshot, are merely a Òtheorist's illusionÓ, with neither any epistemic significance nor scientific utility. In the empirical literature, Mack and Rock call the subject's inability to report such (visual) stimuli Òinattentional blindnessÓ, implying a failure of (conscious) perception, whereas Wolf claims that their data at best show a failure of recall, not perception, hence labeling the phenomenon Òinattentional amnesiaÓ. Here, the issue is on the functional role of attention vis?ˆ?vis conscious perception and memory. In addition to sorting out the different classes of reasons and conditions that require different explanations in these cases of unreported stimuli, I will also argue that there is differential epistemic significance to various such cases of Òunconscious perceptionÓ in the absence of phenomenal consciousness, but with important differences from conscious perception. My conclusion is that a detailed examination of what can be accomplished by perceptual processes lacking phenomenal content is essential in understanding the proper functional role of conscious perception, e.g. in the self?initiation, sustenance, and epistemic justification of perceptually?driven voluntary behavior.
Comparing the Updating of Conscious and Unconscious Perceptual Streams: A New Temporal Illusion
Chien-te Wu, Centre de Recherche Cerveau et Cognition
Niko Busch, Universitè Paul Sabatier, Toulouse
Michele Fabre-Thorpe, Centre de Recherche Cerveau et Cognition CNRS
Rufin VanRullen, Centre de Recherche Cerveau et Cognition CNRS
Sunday, June 7th, 15:00–15:30: Room 3096, Humboldt-Universität zu Berlin
An optimal correspondence of temporal information between the physical world and our perceptual world is an important factor for survival. Several recent studies revealed that specific discrepancies exist between physical time and time perception. However, the relationship between the perceived time of events and the dynamic state of their neuronal representations remains largely unexplored. In the current psychophysical study, we demonstrate a striking phenomenon in which the stimulus that systematically triggers the occurrence of another perceptual event is frequently perceived as lagging the event itself. We used a paradigm referred to as motion-induced blindness (MIB) in which a static visual stimulus presented on a constantly rotating background disappears and reappears from awareness periodically, with the dynamic characteristics of bistable perception. A sudden stimulus onset (e.g., a flash) presented during a period of perceptual suppression (i.e., during MIB) is known to trigger the almost instantaneous reappearance of the suppressed stimulus. Surprisingly however, we report here that although the sudden flash triggers the reappearance of the static target, it is systematically perceived as occurring after this reappearance. This temporal-reversal illusion may be explained by different time courses for updating the conscious representation of a temporarily suppressed "old stimulus", and for establishing conscious perception of a newly presented stimulus. In a second study, the color of the target was changed at different latencies (either before, simultaneously or after) relative to the onset of the sudden flash, and subjects reported the perceived color at target reappearance. Again, the target reappearance was frequently perceived as preceding the flash onset, but was generally associated with the color presented ~100 ms before the flash onset. In other words, a novel stimulus onset can reactivate a temporarily unconscious target representation with a shortened latency (resulting in the perceived temporal reversal) but not all the target features benefit equally from this reduced access time to consciousness - color apparently updating at a slower pace than the rest of the target representation. This illusion sheds important light on the distinct neural signatures of conscious and unconscious events in the brain.
Effects of Perceptual Expectations on Consciousness: Dissociation Between Long-range Synchronization and Gamma Oscillation
Lucia Melloni, Max Planck Institute for Brain Research
Notger Müller, Magdeburg-Universitätsklinik für Neurologie
Wolf Singer, Max Planck Institute for Brain Research
Eugenio Rodriguez, Max Planck Institute for Brain Research
Sunday, June 7th, 15:00–15:30: Room 2097, Humboldt-Universität zu Berlin
Conscious perception is not a passive process merely registering the information that impinges our senses. On the contrary, perception is a dynamic constructive phenomenon in which incoming sensory evidence is selected and compared to predictions on the basis of higher-order representations. Those predictive signals might be particularly relevant for stabilizing percepts and bringing them into consciousness when stimuli are ambiguous or constantly changing. It has been proposed that conscious perception is an emergent property resulting from the coordinated behavior of large populations of neurons, namely long-range synchronization. However, it is unknown how or whether neuronal synchronization is modulated by top-down processes. In this study we explore how higher-order representations affect visibility, and how this effect is reflected in brain activity. We hypothesized that long-range synchronization correlates with conscious perceptibility, whereas local synchronization in the form of local oscillation in the gamma frequency range correlates with the generation of top-down representations. We measured electroencephalographic activity in a visual hysteresis paradigm. The contrast of an initially hidden stimulus was first increased and then decreased, which led to continuous changes in visibility. Under these conditions, perceptual hypotheses are built up as soon as the subject perceives the stimuli, which in turn increases the visibility of subsequent lower contrast stimuli. Our behavioral results confirm this effect by demonstrating a shift in the visibility threshold: The same physical stimulus initially judged as invisible while the contrast was increased was judged as visible when the contrast was subsequently decreased. In addition, long-range synchronization correlates with conscious perception (seen vs. unseen stimuli). This effect is modulated by the generation of a perceptual hypothesis: Stimuli that are rated as invisible during the contrast increase phase elicit higher long-range synchronization when they become visible during the subsequent contrast decrease phase. In contrast, local middle frequency gamma oscillations (55-70 Hz) follow the visibility curves, whereas high frequency gamma oscillations (80–95 Hz) correlate with the formation of predictions. Our study provides further evidence for the hypothesis that long-range synchronization correlates with perceptual awareness, whereas top-down influence is reflected by an increase in local gamma oscillations.
Computationalism: Still Cool After All These Years
Marcin Milkowski, Polish Academy of Sciences
Sunday, June 7th, 15:00–15:30: Room 3075, Humboldt-Universität zu Berlin
In this talk, I want to systematically review the motivations for having a computational theory of consciousness to see if they turn out to be no longer plausible in the light of recent criticisms. These criticisms focus on the alleged inability of computational theories to deal with qualia, or qualities of experience (or objects of experience in some accounts), and with so-called symbol grounding on the other hand. Yet it seems that computationalism remains the best play in town when one wants to explain and predict the dynamics of information processing of cognitive systems. Conscious information processing does not seem to be explainable better within any other framework; computationalism regarding consciousness can only be discarded by supposing that consciousness is epiphenomenal in information processing. I will argue that recent theories of consciousness that are to deal with the so-called hard problem of consciousness remain in their core computational if they do not subscribe to epiphenomenalism. For example, the quantum theory as proposed by Stuart Hameroff remains openly computational; the same goes for pan(proto)panpsychist speculation of David Chalmers. The qualitative character of information processing that Chalmers takes to explain the existence of subjective experience piggy-backs, so to say, on the very fact that there is information processing that is best explained in a computationalist framework. I will also briefly show that other alternative accounts of consciousness (such as direct theories of consciousness) that were supposed to oppose computational and functionalist conceptions are not only compatible with them but require them to begin with. In short, to discard credentials of computationalism in consciousness research one would have to show that it's possible to explain conscious information-processing mechanisms sufficiently in a non-computational way. And this has not been done by any of the critics of computational accounts. This all doesn't suggest, though, that computational explanation is sufficient for building a complete theory of consciousness; it might however be necessary.
Awareness and Decision in Monkey with Blindsight
Masatoshi Yoshida, National Institute for Physiological Sciences
Kana Takaura, National Institite for Physiological Sciences
Tadashi Isa, National Institite for Physiological Sciences
Sunday, June 7th, 16:00–16:30: Room 3096, Humboldt-Universität zu Berlin
Macaque monkeys with a unilateral lesion in V1 have been used as an animal model of blindsight. Here we sought for the neural mechanisms responsible for their visual awareness and residual vision. First, we examined whether the monkeys showed behavior analogous to human blindsight patients. Two macaque monkeys with a unilateral V1 lesion were tested with two saccade tasks. 1) A forced-choice (FC) task, in which the saccadic target comes on one of two possible positions, requires discrimination of target positions. 2) A yes-no (YN) task, in which, in addition to the above condition, the monkeys have to maintain fixation when the saccadic target is absent, requires detection of the target. The d' for the FC task was significantly higher than that for the YN task, consistent with the studies of human blindsight patients. Next, we recorded neural activities from the superior colliculus (SC) of the monkeys performing the FC and YN tasks. We found that the transient response of the ipsilateral SC to the visual target was larger in the hit trials than the miss trials in the YN task. Such modulation was not found in the normal, contralateral SC. This suggests that the activity in SC comprises a part of neural correlates of reduced visual awareness specific to blindsight. Finally, we examined the decision threshold for initiation of saccades in the FC task. We modeled the distribution of saccadic reaction times by a modified diffusion model and obtained evidence that the decision threshold in the affected hemifield was lower than that in the normal hemifield (Yoshida et al. 2008). These results suggest that the geniculostriate pathway is crucial for decision processes. We propose that these results reflect deficits in deliberate control of visual-oculomotor processing after V1 lesions, which may parallel loss of visual awareness in human blindsight patients.
Neural Signatures of Body-ownership and Agency
Manos Tsakiris, Royal Holloway University of London
Matthew Longo, University College London
Patrick Haggard, University College London
Sunday, June 7th, 16:00–16:30: Room 2097, Humboldt-Universität zu Berlin
Body ownership can be easily confounded with the sense of controlling one's body because agency is a powerful cue to ownership: my body feels like mine because I can control it at will. As a result, interactions between body-ownership and agency are difficult to investigate. We developed an fMRI paradigm to investigate multisensory and sensorimotor aspects of body representation in the brain. Movements of the participant’s hand were either self-generated or externally-generated, and video-feedback was relayed in real-time or with a systematic delay. Analyses showed different activations in the right parietal lobe for intersensory and sensorimotor conflicts. Activity in the SMA was linked to a sense of agency over and above the sense of body-ownership, while activations in midline cortical structures were associated with a purely sensory-driven sense of body-ownership. The results are discussed in the light of recent neurocognitive models of self.
Qualia and Introspection
Michael Beaton, University of Sussex
Sunday, June 7th, 16:00–16:30: Room 3075, Humboldt-Universität zu Berlin
The claim that behaviourally undetectable inverted spectra are possible has been endorsed by many physicalists. I explain why this starting point rules out the standard form of scientific explanation, for qualia. The modern 'phenomenal concept strategy' is an updated way of defending such problematic intuitions as the inverted spectrum. However, I show that the phenomenal concept strategy cannot help to restore the possibility of such standard scientific explanation. I argue that it follows that Chalmers is right: we should accept the falsity of physicalism if we accept this problematic starting point. But we have a dubious basis, at best, for endorsing this costly starting point. This is because, I will argue, endorsing this starting point amounts to at least implicitly endorsing certain theoretical claims about the nature of introspection. I therefore suggest that we allow ourselves to be guided, in our quest to understand qualia, by whatever independently plausible theories of introspection we have. In order to achieve this, I propose that we adopt a more moderate definition of qualia, as those introspectible properties which cannot be fully specified simply by specifying (any number of) the non-controversially introspectible 'propositional attitude' mental states (including seeing x, experiencing x, and so on, where x is a specification of a potentially public state of affairs). I argue that properties fitting this definition may well exist, and be found within the properties introspectible on an independently plausible, naturalisable account of introspection. I argue that such properties have the potential to explain, rather than explain away, the problematic intuitions discussed in the earlier part of the paper. This more moderate approach holds out the hope of an integration of our understanding of qualia with the rest of science.
The Neuroanatomical Case Against Consciousness in Fish: Hard Evidence and Theoretical Speculations
Ilya Farber, Singapore Management University
Sunday, June 7th, 16:30–17:00: Room 3096, Humboldt-Universität zu Berlin
Can fish suffer? That is, can they consciously experience pain - or for that matter, any other sensation? In this paper, I argue that the neuroanatomical evidence is surprisingly unequivocal: fish brains lack anything like the right kind of structure for supporting the dynamic, multisensory feature binding that is essential for consciousness. In mammals, the diffusely-projecting structures of the intrathalamic nuclei and thalamic reticular nucleus are necessary for sensory awareness, and a lesion of either results in a permanent loss of consciousness. In fish these structures are absent, and importantly, no other structure is situated to perform the relevant functions. If we take the leading neural models of consciousness seriously, this alone should be sufficient to disqualify fish as bearers of consciousness - though there is also strong converging evidence from ablation studies and from comparative anatomy. My aim here is not to pursue any particular grudge against ichthyans, nor to make excuses for eating them (I don't). Rather, in previous work I have argued that taking ourselves seriously as theorists of consciousness must mean taking our theories seriously even - or especially - when they have consequences which are morally troubling. In my experience, many people (including neuroscientists) are particularly uncomfortable with the idea of using neural theories of consciousness to make claims about its absence. In this paper, then, my goal is to lay out the evidence and then focus on the questions surrounding this sort of judgment: Is it anthropocentric? Is it putting too much confidence in our theories? Are there alternative standards which are more suitable for making judgments about the moral status of nonhuman animals? I will argue that, in each case, the answer is no.
Out-of-Body Touch: Multisensory Mechanisms of Self-consciousness
Jane Aspell, Ecole Polytechnique Fèdèrale de Lausanne
Bigna Lenggenhager, Brain Mind Institute, EPFL, Switzerland
Olaf Blanke, Brain Mind Institute, EPFL, Switzerland
Sunday, June 7th, 16:30–17:00: Room 2097, Humboldt-Universität zu Berlin
The spatial unity between self and body can be disrupted by employing conflicting visual-somatosensory bodily input, thereby bringing neurological observations on bodily self-consciousness under scientific scrutiny. Here we designed a novel paradigm linking the study of bodily self-consciousness to the spatial representation of visuo-tactile stimuli by measuring crossmodal congruency effects (CCEs) for the full body. The crossmodal congruency task has been employed in a number of studies (e.g. Spence et al., 2004; Pavani et al., 2000) to investigate multisensory spatial representations, e.g. it has been used to assess the influence of mirrors, tools and rubber hands on the representation of visuo-tactile space. In the most commonly used design, foam blocks are held in the hands, and vibrator-light (LED) pairs are placed near the thumbs and index fingers of both hands. Subjects make speeded elevation (up/down) judgments of the vibrotactile stimuli while ignoring light stimuli. Subjects perform worse when the light is presented at an incongruent elevation to the vibration and this effect is larger when the stimuli are closer to each other in space. Here we measured full body CCEs by attaching the four vibrator-LED pairs to the trunks (backs) of subjects who viewed their bodies from behind via a camera and a head mounted display (HMD) (as in Lenggenhager et al., 2007). To modulate self-identification for the seen body subjects were stroked on their backs with a stick and the felt stroking was either synchronous or asynchronous with the stroking that could be seen via the HMD. We found that (1) tactile stimuli were mislocalized towards the seen body, (2) CCEs were modulated systematically during visual-somatosensory conflict when subjects viewed their body but not when they viewed a body-sized object, i.e. CCEs were larger during synchronous than during asynchronous stroking of the body and (3) these changes in the mapping of tactile stimuli were related to predictable changes in bodily self-consciousness. These data reveal that systematic alterations in the mapping of tactile stimuli occur in the full body illusion and thus establish CCE magnitude as an online performance proxy for subjective changes in global bodily self-consciousness.
Dreams as Imagination or Quasi-perception? A Reconciliatory Answer to a Classical Problem
Jennifer Windt, Johannes Gutenberg-Universität Mainz
Sunday, June 7th, 16:30–17:00: Room 3075, Humboldt-Universität zu Berlin
The question of how to conceptualize the conscious experience of dreaming dates back to antiquity. In On Dreams, one of the earliest philosophical treatments of dreaming, Aristotle described dreams as quasi-perceptual experiences involving the persistence of vision during sleep. This view was contradicted by Hobbes in the Leviathan, who defined dreams as the imaginations of them that sleep. The question of whether dreams are instances of quasi-perception or imagination continues to be a matter of philosophical controversy and was discussed in several recent publications. After briefly reviewing the main philosophical arguments, I will suggest a reconciliatory model of dreams as immersive, sequentially organized, visuospatial experiences during sleep. Unlike imagination models of dreaming, it takes seriously first-person reports of vivid, quasi-perceptual dreams characterized by the predominance of visual imagery and the subjective feeling of presence. It is also borne out by neuroimaging studies showing that REM sleep/dreaming is characterized by wake-like activation especially of cortical visual areas. On the other hand, phenomenological, functional and epistemological considerations suggest that dreams are phenomenally indeterminate with respect to the experience of bodily selfhood. While movement and vestibular sensations are common, other dimensions of bodily experience such as touch, proprioception and thermal sensations, which are related to full phenomenal embodiment in standard wakefulness, are absent in most dreams. I will argue that in dreams, the organization of vivid, quasi-perceptual visuospatial imagery around an internal, first-person perspective gives rise to a basic sense of immersion; because it arises independently from real-world inputs, visual imagery in the dream state can be described as hallucinatory. At the same time, most dreams do not seem to give rise to vivid, quasi-perceptual self-experience and full phenomenal embodiment. In terms of the phenomenal indeterminacy of the dream self, dreams are possibly more akin to offline wake states such as imagination and daydreaming. Further parallels between waking imagination and dreaming include frequent perspective changes, unstable location in the imaginary or dream world, and detachment from real-word and real-body inputs. Consequences of this view for the discussion on embodiment and the recently introduced concept of minimal phenomenal selfhood will be pointed out.
Probing the Phyletic Boundaries of Consciousness: Cephalopod Psychophysics
David Edelman, The Neurosciences Institute, San Diego
Graziano Fiorito, Stazione Zoologica Anton Dohrn
Sunday, June 7th, 17:00–17:30: Room 3096, Humboldt-Universität zu Berlin
Psychophysical experiments generally rely on the ability of conscious individuals to give accurate verbal reports of what they are experiencing, and therefore usually involve human subjects. In most non-human animals, percepts or sensations cannot be assessed directly through verbal report, and psychophysical experiments are therefore quite difficult to perform. Nevertheless, it is possible to exploit non-verbal, behavioral channels to investigate phenomenal experience in non-human animals. Here, we present a psychophysical experimental approach for assessing perceptual capacities of the cephalopod mollusc Octopus vulgaris that may be consistent with conscious states. In our experiments, sequences of visual data were shown to free behaving octopuses via a submersible organic light-emitting diode (OLED) video display. We presented rapid serial sequences of stimuli via the OLED display at rates impossible to achieve through manual presentation to explore the temporal properties of visual perception and sensory integration in O. vulgaris. We employed the 'attentional blink' paradigm, in which a target secondary visual stimulus is presented 200–500ms after a primary stimulus within a rapid serial stream of stimuli. When shown such a visual stream, human subjects remain unaware of the secondary 'blink' stimulus as long as the interval between presentation of the first and second stimulus does not exceed 500ms. If the interval exceeds 500ms, the human subject becomes aware of the blink stimulus. In these experiments, serial stimuli presented in video sequences were chosen for their particular salience for the octopuses. These stimuli included both images of objects with ecological salience i.e., crabs or other octopuses and images of objects immediately relevant to the octopuses' everyday laboratory experiences i.e., white plastic balls associated with positive experiences (fish) or red plastic balls associated with negative experiences (mild electrical shock). Changes in components of octopus body patterning (i.e., chromatic, textural, postural, locomotor) were considered to be forms of behavioral report. The absence of any change during presentation of a 'blink' stimulus known to otherwise induce specific patterning shifts indicated that a subject was unaware of that stimulus. We discuss our results and the possibility of functionally convergent properties of cephalopod and vertebrate perception.
Comfortably Numb: Fusion of Self and Other by Touch
Sebastian Dieguez, Swiss Federal Institute of Technology
Manuel Mercier, EPFL
Nate Newby, EPFL
Olaf Blanke, Brain Mind Institute, EPFL
Sunday, June 7th, 17:00–17:30: Room 2097, Humboldt-Universität zu Berlin
The experience that our body and its parts belong to us and are not those of other people is a key aspect of the self and has been called bodily ownership (BO). Here we investigated BO in a series of behavioural experiments with a tactile illusion that has the potential to disrupt BO instantly and repeatedly. We also report data from a somatosensory evoked potential (SEP) study during this illusion. The numbness illusion (NI) refers to a peculiar feeling of numbness occurring when one person holds his hand against another person's hand and simultaneously strokes the conjoined index fingers with his opposite hand. In study 1 we confirmed this informal observation using a 2x2 factorial design with naive subjects. Stroking of the fingers was carried out by the experimenter or by the subject (Stroking factor), either synchronously (2 fingers touched at the same time) or asynchronously (one finger after the other) (Synchrony factor). Subjects then rated different sensations on a questionnaire. A significant interaction between both factors was found (p<0.001), showing that the NI is maximal in the self-synchronous condition. Study 2 controlled for visual feedback of the hands and expanded the phenomenology of the NI. Viewing of the hands was not found to influence the NI (p>0.14). In study 3 and 4, we showed that the NI relies primarily on tactile signals rather than mere proximity of the hands or being the agent of the stroking. Finally, study 5 used SEPs to left median nerve stimulation during the NI, allowing to investigate the associated brain activation in sensorimotor cortex. Our SEP data shows that an early SEP component at 20ms recorded over the somatosensory cortex controlateral to the stroked hand, is modulated by stroker and synchrony and correlates with the intensity of the illusion. We argue that the NI is a promising paradigm for the study of BO and discuss its neural origin in primary somatosensory cortex. We compare the implication of this structure with respect to neural activations reported with other current methods to manipulate BO and bodily consciousness.
Part-time Zombies, First-person Privilege, and the So Called Explanatory Gap
Michael Pauen, Berlin School of Mind and Brain, Humboldt-Universität zu Berlin
Sunday, June 7th, 17:00–17:30: Room 3075, Humboldt-Universität zu Berlin
Zombie and Inverted Spectrum thought experiments are the bedrock of the explanatory gap or hard problem. Although skepticism against these thought experiments is quite old, none of the objections discussed so far has been found decisive. It will be demonstrated however that a basic flaw of these thought experiments can be made apparent by replacing the usual full-time zombies by part-time zombies. This leads to a more fundamental objection. According to basic assumption of the explanatory gap position, there is a severe asymmetry between the first and the third person perspective, such that there are first person facts (like spectrum inversions) that cannot be known in principle from the third person perspective. Given that scientific knowledge is third person knowledge, certain first person facts would be beyond the reach of scientific explanation. This view, however, is based on a confusion between experience and knowledge. Although there is a first-person privilege with respect to the experience of phenomenal qualities, there is no privileged first person knowledge of these states. Any fundamental restriction of third person epistemic access to a certain type of mental states will result in a similar restriction of first person access to this type of states. Given that most philosophers accept that we do have epistemic first person access to our own phenomenal states, we should also have epistemic third person access to these states. It would follow that there is no basic limitation for a scientific third person explanation of mental, particularly of phenomenal states within a physicalist framework. Even then the intuition might remain that the problem cannot be solved. But historical observation demonstrates that intuitions may change dramatically, particularly under the impression of scientific development. Findings from psychology of emotion and pain research show that this change is an ongoing process which is driven forward also by neuroscientific progress. It is, of course, unclear, how far this development will take us, both on the level of science and on the level of our intuitions. But this is a question of future scientific development, not a fundamental philosophical problem.