1 / 54

THE CODAM MODEL OF CONSCIOUSNESS

THE CODAM MODEL OF CONSCIOUSNESS. John Taylor  King’s College London emails:john.g.taylor@kcl.ac.uk. HOW IS CONSCIOUSNESS CREATED IN THE BRAIN ?. MANY CONFLICTING THEORIES ABOUT CONSCIOUSNESS FOCUS ON ATTENTION AS THE GATEWAY TO CONSCIOUSNESS (necessary, but not sufficient)

emera
Télécharger la présentation

THE CODAM MODEL OF CONSCIOUSNESS

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. THE CODAM MODEL OF CONSCIOUSNESS John Taylor  King’s College London emails:john.g.taylor@kcl.ac.uk

  2. HOW IS CONSCIOUSNESS CREATED IN THE BRAIN ? • MANY CONFLICTING THEORIES ABOUT CONSCIOUSNESS • FOCUS ON ATTENTION AS THE GATEWAY TO CONSCIOUSNESS (necessary, but not sufficient) • ALSO PROGRESS IN *Western Phenomenology *Physiology of Meditation • USE ALL THREE TO SUGGEST ANSWER BASIC QUESTION ABOVE

  3. Consciousness Requires Attention • Neglect: Inability to direct attention focus (loss of parietal (TPJ/IPL) by stroke) • Attentional Blink: Loss of ability to move attention from T1 to T2 in RSVP stream (worst at 300msecs after T1) • Inattentional Blindness: Inability to detect unatended change in environment (but semantic brain activations) • “The further function of attention is to allow selected perceptual information a foothold in consciousness” (Shapiro et al TICS, ‘97) Must search in Attention for Consciousness

  4. ROUTE MAP • PROGRESS OCCURRING IN BRAIN SCIENCE • ALSO PROGRESS IN MENTAL ILLNESS/BRAIN *Schizophrenia (+ve) as lack of agency (C Frith) *OCD as incorrect (& correctible) AC/BG circuits * Schizophrenia as a deficit-sometimes hyperreflexivity-in the ‘self (Parnas, Sass, Zahavi) • MAIN QUESTION BECOMES: WHAT ATTENTION FRAMEWORK LEADS TO CONSCIOUS EXPERIENCE IN BRAIN? • NEED TO RELATE CLOSELY TO ITS DISEASES (schizophrenia, depression, OCD, autism, ADHD, AD,..)

  5. CONTENTS 1) ATTENTION AS CONTROL 2) CONTROL MODELS FOR ATTENTION 3) THE NEW FEATURES OF CONSCIOUSNESS 4) CREATING THE PRE-REFLECTIVE SELF BY CODAM 5) DESTRUCTIONS OF THE SELF 6) PROGRAM TO FOLLOW 7) OVERALL CONCLUSIONS

  6. 1. ATTENTION AS CONTROL • ATTENTION = SELECTION OF PART OF SCENE FOR ANALYSIS (acts as ‘filter’ on input) • AMPLIFICATION OF ATTENDED + INHIBITION OF DISTRACTORS (in sensory & motor cortices, & higher sites) • DETECT ATTENTION CONTROL SIGNAL IN NETWORK OF CORTICAL REGIONS

  7. OVERALL: ATTENTION MOVEMENT INVOLVES BRAIN SITES WITH 2 DIFFERENT FUNCTIONS:  AMPLIFY/DECREASE SENSORY INPUT (in sensory cortices)  CREATE CONTROL SIGNALS FOR THIS (in parietal & frontal cortices): “Attention-related activity in frontal and parietal areas does not reflect attentional modulation of visually evoked responses, but rather the attentional operations themselves.”(Kastner & Ungerleider, Neuropsychologia, 39: 1263-1276, 2001) • EXPECT SITES WITH SPECIFIC FUNCTIONS OF CONTROL (goals, monitors/errors, feedback signals)

  8. CONCLUSIONS ON ATTENTION • ATTENTION MOVEMENT = ‘ACTION’ • DEVELOP CONTROL FRAMEWORK FOR IT •  2 SORTS OF ATTENTION AS CONTROL: sensory motor •  VARIOUS CONTROL MODULES SUPPORTED BY DATA (attention control, goals, buffer/forward model, monitor) • TO EXPLORE THESE MODULES FURTHER BY SIMULATION

  9. 2. CONTROL MODELS FOR ATTENTION • ENGINEERING CONTROL THEORY WELL DEVELOPED - • USE AS ‘FUNCTIONAL MODEL’ (neglect problems of binding, detailed representations, oscillations, etc) • FOR VISUAL ATTENTION CONTROL MODEL:

  10. CONTROL MODEL COMPONENTS: Inverse model controller (IMC) state(t) control(t) desired state(t) Forward model/ observer state(t) state(t+1) control(t)

  11. CONTROL MODEL COMPONENTS: Forward Output Model state(t) estimated sensory feedback (t +1) Error Monitor Module state(t) error(t) desired state(t)

  12. POINTS TO NOTE:  Can apply to other sensory modalities  competition/combination between different sensory modalities  separate control of attention to object features (colour, shape,etc) & to space (most basic)  Learning by Monitor error signal  Sub-cortical sites involved in input representations & goals as well as Monitor & Forward models (such as Cb)  Sub -cortical activation through Ach (from NBM)  Also DA learning (in BG/PFC/Amyg/HC) * Include TH/NRT/CX complex (NRT as inhibitory sheet)

  13. Posner benefit task: how much does attention help?

  14. Posner benefit results

  15. Posner benefit simulation * • Use very simple architecture • Goal module: 3 nodes (L, R, & Central) • IMC module ditto, with lateral inhibition • Object module ditto • Architecture: IN→OBJ←IMC←GOAL←IN

  16. SIMULATION OF SENSORY ATTENTION MOVEMENT (with M Rogers: Neural Networks 15:309-326 (2002)) Figure of Invalid Cueing (Posner Benefit - exogenous)Figure of Invalid Cueing (Posner Benefit - endogenous) Figure of Validity Benefit as function of CTOA

  17. SENSORY-MOTOR ATTENTION CONTROL MODELS • ATTENTION CRUCIAL TO MOTOR LEARNING •  SEPARATION OF  spatial attention in R hemisphere:  ICMAV  motor attention in L hemisphere:  ICMAM • USE ATTENTION TO LEARN RULES (in DLPFC) & IMC (in PPC) • MOTOR CONTROLLER CODES FOR MOTOR INTENTIONS: object->action • LEARN MOTOR ATTENTION CONTROL & GOAL MODULES STRUCTURES (by DA/ACh/NA error-based from CX/subCX) • NUMBER OF CONTROL MODULES STILL TO IDENTIFY

  18. Left-Hemisphere Dominance for Actions (Schluter et al, ;01)

  19. FOR JOINT VISUAL/MOTOR CONTROL TAKE (joint Visual attention/motor attention/motor control)

  20. SIMULATION OF VARIOUS PARADIGMS Simulink architecture for joint visual attention/motor attention/motor control

  21. Simulation of Schluter et al (Neuropsychologia 39:105-113 ’01; JGT & NF, IJCNN’03) • Choice reaction time: Respond with 1st finger if one object (of two) & with second finger if 2nd object (of two) • Simple RT:respond with 1st finger to any input • Learn correct response by error-based learning • Agreement OK

  22. Motor Preparation Paradigm (Motor Posner Benefit) * • Determine benefit of motor attention (intention) by pre-cueing required response • Two hexagons, one on either side of fixation point • Border of one brightens as a pre-cue • Then centre of same (valid) or other (invalid) hexagon brightens for response to separate buttons • RT measured in valid and invalid cases

  23. SIMULATING R-H & L-H DEFICITS IN RUSHWORTH et al. USING SIMULINK ARCHITECTURE Rushworth et al.Neuropsychologia, 35 (9), 1261-1273 Group mean RTs (Rushworth et al.) Group mean RTs (simulated)

  24. Siting of Brain Modules for Engineering Control Model • Plant: Sensory CX & Temporal lobes • IMC: LIP/SPL • WM: IPL • Monitor: Parietal Lobes/ Pulvinar/ Superior Colliculus/Ant Cingulate • Goals: PFC • Observer/ WMcd: TPJ/ Parietal CX/ PFC

  25. CONCLUSIONS ON ATTENTION CONTROL MODELS • TREAT ATTENTION MOVEMENT AS AN ‘ACTION’ • DEVELOP CONTROL FRAMEWORK FOR IT •  2 SORTS OF ATTENTION: sensory motor •  VARIOUS CONTROL MODULES SUPPORTED BY DATA (attention control, goals, buffer/forward model, monitor) • APPLIED TO SIMULATE *visual attention control *joint visual/motor attention (M Rogers & JGT) control learning (N Fragopanagos & JGT)

  26. 3. NEW FEATURES OF CONSCIOUSNESS • NEED NEW GUIDANCE TO CROSS MIND/BRAIN GAP (EVEN USING ATTENTION TO HELP) • ‘INNER SELF’ NOT PART OF MAINSTREAM ‘CONSCIOUSNESS’ STUDIES (Rejected by Descartes, Hume, Western Cognitive science) • ‘NAÏVE’ EXLORATIONS OF SELF ‘GAPPY’ NATURE OF AWARENESS (G Strawson, JCS 2000) • RECOGNISED IN EASTERN THOUGHT (Like ‘beads on a string’) • ALSO FROM WESTERN PHENOMOENOLOGY (Husserl, Sartre, Merlau-Ponty, Franck, Zahavi, Parnas, …)

  27. ACCEPT 2 PARTS TO CONSCIOUSNESS: 1) Purely intrinsic, non-relational, Pre-Reflective Self (PRS) 2) Relational, contentful consciousness of External World • LEADS TO PROBLEM OF INTERACTION Pre-reflective ‘Inner’ Self  Outer World • PRS ≠ Body (Proprioception) • NEW STATES OBSERVED UNDER MEDITATION PCE (pure conscious experience); DMS (dual mystical state); OMS (oceanic mystical state • PCE HAS NO CONTENT ‘Reports of pure consciousness suggest that, despite the absence of mental content, the subjects were somehow aware that they remained aware throughout the period of pure consciousness (R Forman, 1999)

  28. ‘…awareness to recognize itself, without the mediation of conceptual objects’ (Andreason, 2000, on the practice of ‘Nregs cgod) ‘the phenomenological reduction of the mind and its contents to their intertwined nature of ‘emptiness’ and the mind’s indwelling clarity (ibid) • PCE HAS DISTINCT PHYSIOLOGICAL CORRELATES, SEPARATING IT FROM OTHER ‘ALTERED’ STATES (sleep, drug induced, OOB, hypnosis…) • THE FOLLOWING FEATURES OF PCE HAVE BEEN OBSERVED BY NUMEROUS EXPERIMENTS * wave synchronisation *skin conduction *respiratory rate

  29. *brain imaging in PCE have shown PFC/Parietal  Sensory areas  CONCLUSION TO SECTION 3: TWO COMPONENTS TO CONSCIOUSNESS: 1) Contentful 2) Pre-reflective (PRS): no content TIMING NON-TRIVIAL (gappy) EXTEND PRS TO PCE, THEN DMS, THEN OMS?

  30. 4. CREATING THE PRE-REFLECTIVE SELF BY CODAM • EXTEND CONSCIOUSNESS AS gappy/discontinuous   pre-reflective self (PRS)   altered mystical states (PCE) • BRIDGE GAP BETWEEN PRS & CONTENT BY TEMPORAL DYNAMICS OF ATTENTION MOVEMENT CONTROL SYSTEM • EXPLAIN PRS AS

  31. CODAM (Corollary Discharge of Attention Movement) MODEL (JGT: CODAM: TICS 6:206-210, 2002; JCS 9:3-22, 2002; NSci Abstr 26:2231, #839.3, 2000) • SPLIT WM (forward model) INTO CD & INPUT COMPONENTS • USES FAST RESPONSE FROM CD=‘OWNER’

  32. USE TO EXPLORE MENTAL STATES BY ANALOGY WITH THEIR KNOWN FEATURES (eg light as EM radiation by Maxwell) • LOOKING AT QUALITATIVE FEATURES ONLY Transparency, Closeness, Temporality, ... (‘Race for Consciousness’, MIT Press, ‘99, C&C, ‘01)→ • FURTHER TESTS NEEDED • CODAM MODEL TEMPORAL DYNAMICS PREDICTIONS • WMcd CREATED EARLY (at 150-200 msecs SOA) ON ATTENTION MOVEMENT

  33. IDENTIFY WITH N2/P2? (under attention control: lost in AB) • LATER PRODUCTION OF THE SENSORY BUFFER SIGNAL = P3 (expected at about 350-500 msecs) • ALSO RELATED TO EARLY (100 msecs SOA) EVOKED OSCILLATIONS SUGGESTED FOR CAUSE OF AB? • DETAILED MODEL OF AB BEING DEVELOPED (N Fragopanagos, S Kockelkoren & JGT)

  34. MEG Measurement of N2 (conjunction search, Hopf et al 02)

  35. Nature of AB: Dependence on Lag Variations

  36. Simulation network architecture for AB:(S Kockelkoren, N Fragopanagos & JGT)

  37. Simulation of AB (S Kockelkoren, N Fragopanagos & JG Taylor): measurement of WM Sens potentials.Uses WMcd as prior amplifier, & WMSens inhibits it

  38. EXTINCTION WITH SIMULTANEOUS STIMULATION: WHAT IS SIMULTANEOUS (Baylis et al, 2002)? • SIMULTANEOUS  MAXIMUM EXTINCTION • LAG BY 200-300 msecs  PERCEPTION SIMULTANEOUS • UNDERSTAND BY CONTROL MODEL:

  39. MAX INHIBITION OF (IMCLIMCR) AT t=0 INRGOAL(exog)RMONRIMCVRINR(amp) WMR‘awareness’  slow by damaged WMR/MONR/IMCR, by 200-300 msecs PREDICT WMR SIGNAL ONLY IF 200msecs AHEAD

  40. RELATION OF CODAM TO OTHER MODELS OF CONSCIOUSNESS • Only NN/Functional model of PRS • Only model using Attention as ‘Gateway’ to Consciousness • Extends various other models (GW of Baars, GW of Dehaene et al, Hopfield nets of Aleksander) using WM/ continued activity as basis of content (qualia) • Also extends complexity of processing (eg Edleman) • No free will: deconstructs homunculus • Thought conscious by attention to neural activities (so completes motor approach of Feinberg & Frith)

  41. CONCLUSIONS TO SECTION 4 • SPLIT WM BUFFER INTO *Sensory buffer (WM Sens) * Corollary discharge buffer (WMcd) • WM Sens = ‘content of consciousness’ • WMcd = ‘ownership of that content’ = pre-reflective self • WMcd signal = N2/P2 at180-240ms? • WMcd signal = ‘echo’ of attention move command signal

  42. 5. DESTRUCTIONS OF THE SELF • USE CODAM TO ATTACK PROBLEMS OF DISEASES OF THE MIND • SCHIZOPHRENIA BASIC HERE: “The libido of the schizophrenic withdraws from the outer world onto its own ego” (Freud) “His pesonal inner drive does not seek to identify with reality, but becomes like an empty slate.” (Minkowski, 1927) “In truth, I am closer to the soul, to Dante’s Paradise, in that world, but I feel removed from life, deviod of emotion and detached from everything” (Elena, a patient of Morselli, 1930) 

  43. “Such hyperattentional dysfunction might represent the primary cognitive abnormality seen in schizophrenia” (Sarter & Bruno, 1999) “The patient does not feel being fully existing or alive, fully awake or alert, or fully present and affected” (Parnas, 2000) • DAMAGES TO SELF AS OBJECT (damaged access/ representation/ output) • ALSO DAMAGE TO OTHER COMPONENTS (Goals/ monitors/ attention controller:hyperreflexivity) • ALSO TO WHAT CAN ACCESS CONSCIOUSNESS (semantic level maps intrude onto sensory buffers)

  44. “A terrible cold/ An atrocious abstinence/ The limbo of a nightmare of bone and muscles, with the sensation of stomach functions snapping like a flag in the phosphorescence of the storm/ Larval images that are pushed as if by a finger and have no material thing” (Artaud, 1965) • DAMAGES OF PRS (=WMcd):

  45. NUMEROUS POSSIBILITIES OF DAMAGE: * WMcd ownership/hyperreflexivity “shrinkage of ipseity” (Parnas) * WMcd increased ownership/decreased content “my mind is empty” (Artaud) • SUPPORTED BY HYPERACTIVATION OF PARIETAL IN PET STUDY ON VOLUNTARY MOVEMENTS (Spence et al, ’97) • UNCLEAR IF THESE CAN SWITCH (As for Artaud?) • NEED TO INVESTIGATE NA/ACH MODIFICATIONS (cause of some part of the effects on WMcd)

  46. IF WMcd to Semantic ِtoo big  WM Sens Overactive Awareness but no GoalAlien thought as hallucination (Corrects Frith et al) • RELATION TO PCE: COULD BE IDENTICAL (but no goal to achieve PCE set up in schizophrenia) • LEADS TO FEELING ‘OUT OF CONTROL’ IN SCHIZOPHRENICS • FEAR CAN ALSO ARISE IN MEDITATORS “To put aside the self is a premature laying down of our weapons before we know for sure what lies ahead. It’s all an insane risk” (Bernadette Roberts, ‘The Experience of No Self’) • OTHER BRAIN DEFICITS OBSERVED: * ACG x PFC ON STL (list lrng; Fletcher et al, Brain’98) * DLPFC(L)/IFG/PL(m)(verbal fluency,Curtis et al, IOP, ‘99) * ACG ERN (Stroop, Alain et al, ‘02) * DLPFC, IPL, SPL, ACG,STL (2-back WM; Menonet al. ‘01)

  47. OTHER DEFICITS AS PART OF OVERALL CIRCUITRY (Frith, Blakemore & Wolpert, 2000, for motor control deficits) *Attention Controller PRS/Content (neglect-like) * Attention Controller ADHD-like * Goals planning/rules following (frontal) * Goals Obsessive behaviour (OCD-like) * Episodics/Semantics  Memory loss (AD) * Input  Sensory buffer  (blindsight-like) * Motor attention controllerhemiplegia-like * Plus motor goal/monitor anosognosia-like

  48. Parietal activity in schizophrenics v control

  49. Deficits for Motor CX v PCX Patients in Response Control

More Related