Centre for Cognitive Science (COGS)

Summer term 2009

  • Week 2 (April 28th) Dr. Tim Bayne (St. Catherine's College, Oxford): The Unity of Consciousness as a Constraint on Theories of Consciousness
  • Week 3 (May 5th) Dr. Anil Seth (Informatics, Sussex): Measuring Consciousness: From Behaviour to Neurophysiology
  • Week 4 (May 12th) Dr. Romi Nijhawan (Psychology, Sussex): Predicting the Present: Visual-Motor Neural Delays and Compensation
  • Week 5 (May 19th) Dr. Nick Collins (Informatics, Sussex): Audiovisual Experiments 2002-2009
  • Week 6 (May 26th) Scott deLahunta (Research Fellow, Amsterdam School for the Arts): Choreographic Tools and Agents
  • Week 7 (June 2nd) Dr. Murali Ramachandran (Philosophy, Sussex): The Surprise Examination Paradox and Epistemic Closure
  • Week 8 (June 9th) Prof. Andy Clark (Chair in Logic and Metaphysics, Philosophy, Edinburgh): Spreading the Joy : Why Consciousness is (Probably) Still in the Head
  • Week 9 (June 16th) Dr. David Young (Informatics, Sussex): Searching Parameter Spaces by Mapping Likelihood
  • Week 10 (June 23rd) Dr. Michael Morris (Philosophy, Sussex): Language and Content

Week 2

Date: Tuesday, 28th April 2009
 Dr. Tim Bayne (St. Catherine's College, Oxford)

The Unity of Consciousness as a Constraint on Theories of Consciousness

This talk presents an overview of the unity of consciousness, with a focus on the question of how the unity of consciousness might constrain theories of consciousness. For the most part, theories of consciousness have not been constructed with the unity of consciousness in mind. This is unfortunate, for the unity of consciousness provides researchers with a rich set of constraints on theory-building. I examine those constraints, and explore the implications that they have for current discussions of the nature of consciousness itself.

Week 3

Date: Tuesday, 5th May 2009
 Dr. Anil Seth (Informatics, Sussex)

Measuring Consciousness: From Behaviour to Neurophysiology

How can we measure whether a particular sensory, motor, or cognitive event is consciously experienced or remains unconscious? Such measurements provide the essential data on which a science of consciousness depends, yet there is no clear consensus on how such measurements should be made. Measures of consciousness cannot exist independently of theory. Only by behaving sensibly in a theoretical context do proposed measures pick themselves up by the bootstraps, validating both themselves as measures of what they say they measure and also validating the theories involved. Much of what we know derives from subjective (introspective) verbal report, but on some theories such reports confound mechanisms of metacognitive access with mechanisms of consciousness and are also susceptible to biases. In response, there has been a growing emphasis on neurophysiological measures as well as on behavioral measures that do not rely on introspection. But for these 'objective' measures it can be hard to guarantee that they are measuring consciousness per se. I will review definitional, methodological, and conceptual issues surrounding the problem of measuring consciousness and describe some specific examples based on measures of complexity and causal density in neural dynamics. Advances in measuring consciousness have implications for basic cognitive science and neuroscience, for comparative studies of consciousness, and for clinical applications.

  • Seth, A.K., Dienes, Z., Cleeremans, A., Overgaard, M., and Pessoa, L. (2008). Measuring consciousness: Relating behavioural and neurophysiological approaches. Trends in Cognitive Sciences.12:314-321.
  • Seth, A.K. (2008). Post-decision wagering measures metacognitive content, not sensory consciousness. Consciousness and Cognition17:981-983.
  • Seth, A.K., Izhikevich, E.M., Reeke, G.N., & Edelman, G.M. (2006). Theories and measures of consciousness: An extended framework. Proc. Nat. Acad. Sci. USA103(28):10799-10804.

Week 4

Date: Tuesday, 12th May 2009
 Dr. Romi Nijhawan (Psychology, Sussex)

Predicting the Present: Visual-Motor Neural Delays and Compensation

For more than a hundred and fifty years, scientists have known that neural processes are relatively slow. This implies that the visually experienced world, which involves the activation of higher cortical areas several synapses removed from the visual input, and therefore subject to delays, is forever lagging the real one. Consequently, events should be perceived after a delay and moving objects should appear where they were in the recent (about 1/10th second) past. But why should it matter if this is so as long as actions of animals are not delayed? The reason is simple: Animals would not survive if neural delays remained uncompensated. I will make a case that delays, in addition to be being compensated by motor systems, are also compensated in the sensory pathways by 'visual prediction'. This compensation can be observed in several instances. Of these I will focus on a fascinating phenomenon (the flash-lag effect) in which a flashed object presented at the same position as a moving object appears in a position lagging the moving object. I will consider the various forms of this phenomenon and address the controversies raised by 'visual prediction'. Supportive data from other labs, for example prediction by retinal cells (rabbits and salamanders) and Head Direction cells (rats) will be mentioned.

Week 5

Date: Tuesday, 19th May 2009
 Dr. Nick Collins (Informatics, Sussex)

Audiovisual Experiments 2002-2009

A multimodal snapshot of audiovisual research will be presented. Solo work will appear alongside collaborations with Fredrik Olofsson in the live audiovisual laptop duo 'klipp av'. klipp av is Swedish for 'cut apart' and the duo have played extensively in VJ and live cinema contexts, exploring live algorithmic content combined with processed audio and video capture.

One key technological edge has been to take advantage of machine listening and computer vision principles in feature extraction. Audio analysis has been used to drive the formation of audiovisual content databases on the fly. In 'audiovisual feedback', the data from audio and visuals is cross-mapped, informing the synthesis of future material; it is also possible to iterate the feature extraction through multiple generations of this process. In audiovisual concatenative synthesis, audiovisual feature vectors are created to mark-up a database of materials; a driver sequence can now be analyzed in turn, pulling close matches out of the database for resynthesis and mash-up effects.

Movie cut-ups will be illustrated using this technique, which however raises issues of respecting scene boundaries and congruence of meaning. It is hoped that the talk will promote some discussion of audiovisual work in a wider context of multimedia theory, audiovisual cognition and new technologically acute routes for live performance.

Week 6

Date: Tuesday, 26th May 2009
 Scott deLahunta (Research Fellow, Amsterdam School for the Arts)

Choreographic Tools and Agents

Contemporary choreographer Wayne McGregor has always drawn collaboratively on a wide range of perspectives in the creation process, and this has evolved toward a sustained research into the nature of dance making and the 21st Century body particularly its cognitive and biological/ technological aspects. This talk will feature two interdisciplinary projects McGregor and his collaborators are currently working on. The first is to make autonomous choreographic software agents that can generate novel solutions to choreographic problems alongside the choreographer's own creative decision-making. The second is to develop new choreographic ideation tools drawn from the field of cognitive science that will augment the creative process and, through this, increase our understanding of making dances.(http://www.randomdance.org/)

Week 7

Date: Tuesday, 2nd June 2009
 Dr. Murali Ramachandran (Philosophy, Sussex)

The Surprise Examination Paradox and Epistemic Closure

There are many, very diverse, attempts to resolve the surprise examination (or predictionparadox on the market. I wish to throw my own wacky proposal into the ring. It is only a vague, half-baked thought at present, but I am fairly confident that what little I have to say will already make it seem completely barmy and off track.

The paradox arises from the following announcements made by a teacher to her students-(to explain the labelling: 'E' for exam and 'S' for surprise):

(E) You will have an exam one morning next week at 10 am.
(S) But, you won't know which day you'll get it until you get it!

The students are apparently able to reason from (E) and (S) to the paradoxical conclusion that they will not be set the exam at all, contra (E). What gives?

I will consider some of solutions in the literature. Then I will show that (E) and (S) are not what generate the problem; what is under threat, rather, is the view that the students know (E) and (S) to be true-the paradox is really a paradox about knowledge by testimony. I am going to argue that while it may be true that they know (S) to be true, it is not legitimate for the students to use the fact that they know it in their reasoning! I will back up this view a consideration of epistemic closure and the recent transfer of warrant debate in philosophy (which pivots on the question of whether the warrant for the premises of a valid argument can 'fail to transfer' to the conclusion).

Week 8

Date: Tuesday, 9th June 2009
 Prof. Andy Clark (Chair in Logic and Metaphysics, Philosophy, Edinburgh)

Spreading the Joy : Why Consciousness is (Probably) Still in the Head

Is consciousness all in the head, or might the minimal physical substrate for some forms of conscious experience include goings on in the (rest of the) body and the world? Such a view might be dubbed (by analogy with Clark and Chalmers work on 'the extended mind') 'the extended conscious mind'. In this talk I review a variety of arguments for the extended conscious mind, and find them mostly flawed. One does better, but even it encounters problems. I then try to show why arguments for extended cognition do not generalize to arguments for an extended conscious mind.

Week 9

Date: Tuesday, 16th June 2009
Dr. David Young (Informatics, Sussex)

Searching Parameter Spaces by Mapping Likelihood

The efficient use of negative evidence in search problems has always been important: for example, string search algorithms such as Boyer-Moore make use of negative evidence to achieve greatly increased speed. However, it is not always clear how negative evidence can be exploited in a probabilistic framework. In this talk, I explore the accumulation of negative and positive statistical evidence by building a map of likelihood in parameter space, allowing a directed search of this space. I illustrate the approach with a simple image matching example, which emphasises the value of accurately modelling (or learning) image statistics. I discuss the conditions under which the method may be useful, and I propose that the correct framework for it is not a Bayesian one, but rather the likelihood method of A.W.F. Edwards.

Week 10

Date: Tuesday, 23rd June 2009
Speaker: Dr. Michael Morris (Philosophy, Sussex)

Language and Content

Almost all approaches to language assume (in effect) that the world, or thought, is already ready to be described, or expressed, in language: I call this the Readiness Assumption. This leads to a particular understanding of the sentences we use to say what people think and feel. Putting it roughly, we take the words in the 'that'-clauses of these sentences to record some antecedently existing structure (either in the mind or in the world) to which a person (the subject) stands in some determinate relation. But this falsifies our ordinary sense of what we are doing when we use these sentences, and makes it difficult to understand how we can say what non-human animals, or humans who speak different languages, or even humans who speak our language but are not now speaking, think or feel. It also appears to force us into a false choice between a language-of-thought kind of theory and an incoherent instrumentalism. I suggest that by denying the Readiness Assumption we can give a much more natural account of what we are doing when we say what people think and feel, which also avoids these problems.

Series organized by Joel Parthemore (jep25@sussex.ac.uk)