Bernard Baars, In the Theater of Consciousness. Oxford U. P. 1997. Chapters 3-4. [Comments in square brackets are my own thoughts.] Ch.3: "O N S T A G E" (p.62-94) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ This chapter is about the vision system, mental images, inner speech, and abstract concepts and ideas; and the differences among these with respect to how they register consciously. The "sensory bias" of consciousness (p. 62-64) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Baars notes that sensory experience (vision, touch, hearing, smell, taste) seems much more vivid than our inner re-creations of them (imagining sights, sounds, smells, etc). Mental images are "faint copies of sensations" (Aristotle). Also we are not conscious of our muscular activity, except through perception of the resulting output (speech or motion). At the same time sensory experience seems to combine multiple modalities -- e.g., color, texture, shape, and classification of an apple; e.g., the words, cadence, and meaning of a poem. By contrast, we are only dimly aware of abstractions or plans we might think about, or our control of ourselves in action -- these are neocortical functions. [This may explain why some theorists -- e.g., Nicholas Humphreys (and to some some extent Baars himself) -- lean towards the view that thinking is largely imagistic: we simply are more *aware* of imagistic aspects of our thinking than abstract ones. Certainly these imagistic components are important in problem solving, and when we focus on certain concepts, such as a "giraffe", we virtually can't help conjure up a mental image. But from an information processing and language understanding/ production perspective, it would be extremely inefficient to rely mostly on image manipulation rather than more abstract processing at a symbolic level. Consider this paragraph that you just read: apart from the "giraffe" example, what images did it evoke?] Vision (p.65-73) ~~~~~~~~~~~~~~~ The successive stages/regions of the visual cortex: V1 (edge, point, line, moving line detectors); V2 (orientation and color); V3 (shape); V4 (color); V5 (motion detection). Projection from V4, V5 to frontal cortex. 2-way loopy pathways connecting V1 & V2, V2 & V3, etc., V1 and thalamus; V & areas in lower part of temporal lobe (object recognition); V & upper part of parietal lobe (object location). Blindsight (p.66-71) ~~~~~~~~~~~~~~~~~~~ Recognition without visual consciousness [as in Blackmore]. [One point that thoroughly puzzles me is how blindsight patients can (unconsciously) recognize object shapes, locations, etc., without area V1 -- don't you HAVE to detect edges, etc., to make out the shape and object type? What is known about those other, secondary vision pathways that such patients make use of?] Why is V1 key to consciousness? Even to conscious awareness of abstract properties like shape or object type? Is V1 the "spatial coordinating map"? Binocular rivalry again (Logothetis & Schall, upward and downward-moving gratings -- macaque monkeys); V5 firing reflected what the monkeys were experiencing, as indicated by their button-pushing. Corroborative evidence from human brain imaging. [Blackmore said the *temporal* lobe showed alternation.] The reticular nucleus atop the thalamus may "control the gates" of the other major thalamic nuclei (relay stations for sight, hearing, touch, etc.). Conscious experience requires open gates. How is all this coordinated? -- the 40 hertz hypothesis of Llinas, Crick, Ribary and others: oscillations (running backward and forward in the brain) produced by the reticular nucleus and intralaminar nuclei create these oscillations, temporally binding the activity of whatever neural systems are currently jointly responsible for conscious experience. Visual imagery (p.73-4) ~~~~~~~~~~~~~~~~~~~~~~~ Stephen Kosslyn's work showing that "the mind's eye" is quite a realistic figure of speech -- the inner visual field has the same shape as the actual one, and mental rotation takes time proportional to angle, etc. The brain areas active in "imagining" are the same ones active in actually seeing (similarly for hearing, touch, etc.) Inner speech (p.75-7) ~~~~~~~~~~~~~~~~~~~~~ This seems to be our way of bringing thoughts to consciousness. Both Broca's area (speaking) and Wernicke's are (speech recog.) are active -- but only the latter seems to correspond to consciousness of inner speech (as in real speech understanding). Gary Dell's work (originally at UR!) on speech errors -- they occur in the sequencing in the brain (hence also in inner speech), not in the articulatory execution. Phobias (in image or voice form) can be combatted by practicing control of mental imagery and inner speech. The contrast between mental images and abstractions (p.77-84) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Baars' interest in this theme: contrastive phenomenology. Mental images are always particular, not abstract (Plato, Berkeley). "Bird" or "vegetable" evoke (faint but) rather specific images (e.g., a carrot, in a particular orientation). The semantic content of language also seems accessible only in combination with particular verbalizations. (In reading or hearing a sentence, we may be aware of a "click of comprehension".) We quickly forget specifics of verbalizations, but can recover the content (more or less), using a potentially rather different verbalization. [However, I think many theorists go too far in concluding that thought is nothing like language; I think language "compacts" thought, but still reflects much of its essential character.] P.79-84 discusses prototypes, with reference to Lakoff & Johnson. Even words like "democracy" tend to evoke specific images, and this raises the question of what the utility of these "prototypes" is in thinking. [To reiterate, for me this is somewhat backward. I think what gets lost in this discussion -- as is often the case with cognitive scientists who do not specialize in language -- is the enormous power of SYMBOLIC THOUGHT. I think THIS is what launched homo sapiens on its dominant, communicative, cultural trajectory (something Stephen Pinker, among others, argues.) I had to work rather hard to come up with images for "democracy", which indicates to me that I don't think about democracy primarily in images; in fact, while we can ATTACH images to almost anything, and this can be helful in certain cases (e.g., imagining whether an open umbrella will fit into the trunk of a standard sedan), I think symbolic thought has enormous advantages in generality and speed in much of our thinking. It is the relative vividness of images, compared to the seemingly unconscious symbols, that can mislead us into thinking that imagery is the main vehicle of thought. Note that Baars does say, on p.86, that beliefs and ideas do not seem to have any distinct perceptual quality, and though they may have "fleeting fragments of images associated with them, they are not at all easy to retrieve".] P.82 mentions Raymond Gibbs' "love as a nutrient" -- thriving on love, starved for love, etc. [But I would point out that the metaphor is more general than that -- witness "thriving on the Bohemian lifestyle, starved for some light entertainment, etc. It's not surprising that much of our abstract vocabulary derives from analogy with our concrete physical experience; but we shouldn't conclude from this that abstract thinking actually consists predominantly of physical imagery.] Baars does mention that metaphors can be misleading; e.g., (p.84) "blowing off steam" turns out to aggravate a tendency towards angry outbursts, rather than ameliorating it (Seymour Feshbach); so the popular advice, "Don't bottle it up", based on the metaphor, may be bad advice. Similarities between sensations & ideas (p 86-88) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Meanings seem constrained to be unambiguous: you have trouble keeping continuously in mind different meanings of the same word ("bank", "fix"), or different interpretations of the same image (a Necker cube, or just a "bookend"). The impossibility of a single coherent interpretation makes the "devil's pitchfork" so fascinating/irritating. Related facts are the breakdown of binocular fusion if the two images sent to the two eyes are too different, or the breakdown of the sense of continuous vision for a film with too-large pauses & changes between successive images. Baars' conclusion (p. 90) is that conscious experience is multilayered, having both sensory and abstract aspects. The sensory aspect can itself be multilayered -- e.g., outlines, colors, shapes, shadows, reflections, texture, on a coffee cup. The abstract aspect also seems multilayered -- certain stroke patients lose just noun meanings, or just function-word meanings. "Semantic satiation" (p.92) indicates that specific neural pathways are involved in semantic interpretation, and these can become fatigued, and need a few seconds to recover. He reiterates p.93 that abstract thinking "may ride on the older functioning of the sensory cortex" [but I think, again, that this notion can be taken too far, blinding us to that most powerful of recent evolutionary inventions, symbolic thought]. Ch. 4: T H E S P O T L I G H T (O F A T T E N T I O N) (p. 95-108) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ [Stephen Wright: "I was trying to daydream, but my mind kept wandering."] This chapter is concerned with how we construct our particular reality by spinning a coherent (but often inaccurate) story out of the fragmentary experiences we derive from the flitting about of our attention (e.g., visual saccades & fixation). The sensory experiences (plus interpersonal feedback) provide the reality check in our day-to-day lives, but when it comes to notions about which we get no direct and immediate perceptual feedback, we spin fantastical stories -- and believe them. Metacognition (thinking about what we experience, think, and do) also provides critical oversight over our thinking and interpretation of what we perceive. But metacognition is kept out of consciousness (working memory) if we are absorbed in something, or under someone's "hypnotic" influence, or dreaming (except in lucid dreaming). Our inventiveness and gullibility, p.96-7: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ "But wherever we cannot obtain clear reality feedback we seem to spin beliefs that are contradictory, idiosyncratic, and fantastical. Any situation that does not provide rapid and accurate reality monitoring is an invitation to fantasy" "We are a gullible species whenever we cannot get rapid reality feedback" -- he goes on eloquently, mentioning advertising, politics, utopianism, cultural self-flattery (including the term "homo sapiens"), and our ways of coping with death, transitoriness and suffering, ethics, and the uncertainties of modern civilized life. [I think science rests on the recognition that discovering larger truths, beyond those about which our senses directly inform us, requires that we apply "reality checks" and logical soundness checks to all our hypotheses before accepting them, however difficult and roundabout this may be. Here metacognition -- necessary for skepticism -- is also crucial.] Attention (p.98-102) ~~~~~~~~~~~~~~~~~~~~ He notes that we don't *experience* attention, in our normal functioning; rather, attention selects what we experience. So attention is not the same as experience (contra many interpreters of William James). However, there is *voluntary attention*, a metacognitive function (deciding on what to attend to, and then enjoining oneself to attend to it). E.g., study vs. entertainment. p. 100 -- attentional network: working memory (space and verbal representation in frontal cortex) -> executive attention center in cingulate gyrus (gaze control), and visual features (in visual cortex). Again, only visual cortex is associated with conscious visual awareness. Metacognition -- required for voluntary attention (p.101) Decision-making takes time & occupies working memory space. [Oddly, he does not discuss or even cite Libet, though Libet's delay and "veto" theory are from the 70's and 80's. Maybe until later corroborations, and books like Wegner's The Illusion of Conscious Will, or Norretranders' The User Illusion, Libet's results were considered questionable or of uncertain significance.] Absorption -- displaces metacognition from working memory; ~~~~~~~~~~ So, one loses track of time, does not actively decide what to attend to, loses self-awareness, and suspends disbelief -- all metacognitive functions. "Hypnotic" states seem to be just states of absorption, or at least suspension of metacognition and hence deliberate decision- making, and self-skepticism. For the most suggestible 20% of the population, anything they *believe* to be capable of inducing a hypnotic state will do so. He mentions the placebo effect (e.g., capsules "better" than bare pills! Medical students' "disease of the week" syndrome), and the dependence of pain level in injuries on suggestion. [So do pain qualia depend on metacognition? Does reflecting on one's pain worsen it? This ties in with the informal observation that children cry more about a small injury if someone pays sympathetic attention.] William James' "ideomotor theory" -- acting on images in working memory by default, without exercising critical judgement. [Well, I don't find it easier to resist an unhealthy snack just because I'm thinking, "that would be unwise to eat"!] p. 105-6, dreams as virtual reality; p.107: orienting response to surprising events --preparing to flee, fight, or react in some other way. Based on "mismatch detection" between expectations and perception in hippocampus? (Tulving). p.107 dreams again -- this time as arising from unpatterned neural activity (J Allan Hobson); but he asks, what shapes this activity into "movies" of sorts? Neural activity in REM sleep much as in awake state. Lucid dreams (p.109) -- metacognition comes into play. [I would have expected some mention of the "rehearsal" theory, i.e., strengthening memories of the most recent, most striking events or newly gained knowledge; and the theory that dreams may "deliberately" create anxieties about potentially life-threatening contingencies in the kind of setting we evolved in -- encountering large predators or dangerous humans, falling of cliffs, drowning, etc. -- so that we will later think of ways to avoid or deal with such contingencies, should they actually arise. See my remarks on Blackmore's ch.7 -- in particular refer to Antti Revonsuo, http://philpapers.org/s/Antti Revonsuo.]