Saturday, July 9, 2011

Is consciousness search?

Search is all the rage in apps and the internet. What if search is consciousness as well?

After reading Antonio Damasio's book "Self comes to mind", about consciousness, I am little wiser than before. One of its few significant points is to reiterate that explicit memory is a matter of replaying data stored entirely unconsciously in certain areas of the brain (like the frontal lobes) in other areas regularly involved in presenting sensory information. Thus, if we remember a day at the ocean, a brain scanner would see our visual, hearing, and smell processing areas light up with a muted version of the actual experience, presumably transmitted there as a pattern from memory storage areas that, as we learned, basically dump out in reverse what previously played into them.

The book's second significant point is that consciousness is largely about learning, by using long-term, large scale context drawn from unconscious sources to train unconscious modules elsewhere in the brain. Consciousness doesn't directly execute anything, as anyone learning a new skill can attest. If we are learning to play a musical instrument, a very painful process of setting a conscious goal alternates with halting execution as we consciously try to get our body to adopt new actions and habits. Then we consciously perceive the woeful result, and the learning cycle continues. Once the relevant unconscious processes form, conscious direction can be devoted to ever-greater levels of abstraction and indirect control.
"In the words of William James, ‘‘consciousness’’ appears as ‘‘an organ added for the sake of steering a nervous system grown too complex to regulate itself’’ (James, 1890, chapter 5)."   (All quotes taken from this article, not from Damasio)

Third and last is that Damasio is convinced that consciousness involves very low levels of the brain- the stem and thalamus, in addition to higher cortical regions. In one of the few items of data presented, he claims that hydrancephalic patients, born with with no cortex at all, (and who can live into their 20's), still have a rudimentary and responsive form of consciousness. They like some people and not others, like some foods and not others, etc. He also notes that only very few lesions, such as some in the higher brain stem, can radically impair consciousness, indicating that consciousness is a rather broad and evolutionarily primitive function.

So much for the book review. Thankfully, I ran across a much more lucid and up-to-date review of the field, which seems to be freely available, and which supplies relevant quotes below.



What gets me going are databases, so I would naturally see consciousness as a search function over a database. Are feelings involved? Sure, they are just another datatype in the database! Specifically, I'd propose that consciousness is the integration of connections drawn from a vast database which instantly (that is to say, within about 0.3 to 0.5 second) inbue perceptual or recollected data with context, meaning, valance, and feeling. That, I would argue, is the solution of the consciousness problem... the hard problem that is posed as: what makes the "redness" of red?

On its face, red is just a brain-encoded category plucked out of the electromagnetic spectrum more or less arbitrarily. What makes it conscious and rich for us, unlike what a video camera might experience of the same visible phenomenon, are the feelings and knowledge associated with red in our internal constellation. Perhaps we "like" red, and have special associations with various shades of red. Perhaps we know a little about interior design, or graphic design, and see the power that red can lend to projects in those areas. To a baby, red is just flat data, though it may also be intrinsically attractive, based on biological programming of our perceptual apparatus and feelings. Red is the color of blood, after all, and of ripe fruits- doubtless a powerful color on purely inborn biology alone. But to adults, it can be far more meaningful, in qualitative terms of association, which leads to the term "qualia".

Our brains are connection machines, (and also modelling machines, at a higher level). All data that pours into our heads are automatically linked to a web of ambient data- how a cookie smelled as it was being dunked in an aromatic tea by our fingers which felt its slightly greasy, crumbly surface, as we fumbled through a conversation and thought about the morning ahead. Stream-of-consciousness is old-hat by now in literature, but it is important to recognize how significant it is, relative to what is possible with current computers. In database technology, we struggle to relate individual pieces of data, and have to frame them into arbitrary classifications, numerically indexed and neatly tucked into bins. The technology is utterly unlike the freewheeling automatically everything-connects-with-everthing-else analog methods of brain data storage.

Once the data is in, it is ready to come back out, whether implicitly as we go about our lives and keep associating whatever is new with all that has gone before, or explicitly via reverie and replay in our sensory brain regions. One can have more or less data, thus more or less consciousness. Without training, one might be oblivious to the subtleties of classical music, or baseball, in effect being less conscious. Animals have all sorts of levels of consciousness, depending on how much information they can muster, and how much of it their brains keep in close and immediate connection, rather than in the vast unconscious troves of senstivities and implicit memories that far outnumber what is conscious.

What differentiates unconscious from conscious data? Unconscious data is surely usable in specialized systems, and through the mysterious processes of the dynamic unconscious, may create truly novel linkages (ideas) that pop into consciousness from time to time. Unconscious processing is also parallel, in contrast to consciousness, which, while maybe disjoint, still exhibits a linear "flow". This makes unconscious processing far, far more efficient and powerful than consciousness. But it is not continuously integrated and available, perhaps via the high-frequency gamma-wave attention system and long-range axons that seem to correlate with consciousness. The cerebellum is an example- a module of the brain devoted mostly to fine motor control, which processes information without contributing one iota to consciousness, probably because it isn't wired into the top-level consciousness system that exchanges information all over the brain, if in limited amounts.
"Human ERP [event-rrelated potential] and MEG [magnetoencephalography] recordings also revealed that conscious perception is also accompanied, during a similar time window, by increases in the power of high-frequency fluctuations, primarily in the gamma band (>30 Hz), as well as their phase synchronization across distant cortical sites (Doesburg et al., 2009; Melloni et al., 2007; Rodriguez et al., 1999; Schurger et al., 2006; Wyart and Tallon-Baudry, 2009)."
"Nonconscious stimuli can be quickly and efficiently processed along automatized or preinstructed processing routes before quickly decaying within a few seconds. By contrast, conscious stimuli would be distinguished by their lack of ‘‘encapsulation’’ in specialized processes and their flexible circulation to various processes of verbal report, evaluation, memory, planning, and intentional action, many seconds after their disappearance (Baars, 1989; Dehaene and Naccache, 2001). Dehaene and Naccache (2001) postulate that ‘‘this global availability of information (...) is what we subjectively experience as a conscious state.’’"

What form does this "information" take? That is a significant question, even if we accept the overall hypothesis about a global information exchange network or workspace that correlates closely with consciousness. Information all over the brain takes the form of action potentials, quite different from whatever we might imagine as qualia- cloudy whisps, movie images, compositions by Bach, etc. Consciousness would clearly take the same form, being a subset of the wider information flow in the brain, with the properties of being integrative, linear, and consistent in form while varying in content. Positing any other form that this information could take wouldn't make biological sense, nor would it help clarify what makes qualia special.
"A notable feature of the dynamic core hypothesis is the proposal of a quantitative mathematical measure of information integration called F, high values of which are achieved only through a hierarchical recurrent connectivity and would be necessary and sufficient to sustain conscious experience: ‘‘consciousness is integrated information’’ (Tononi, 2008)."

One special property of consciousness is that it can reach into many other areas, depending on attentional focus. Special long-range axons (pyramidal) are thought to provide some of these connections:
"The ‘‘special morphology’’ of the pyramidal cells from the cerebral cortex was already noted by Cajal (1899–1904), who mentioned their ‘‘long axons with multiple collaterals’’ and their ‘‘very numerous and complex dendrites.’’ ... Furthermore, quantitative analyses of the dendritic field morphology of layer III pyramidal neurons revealed a continuous increase of complexity of the basal dendrites from the occipital up to the prefrontal cortex within a given species (DeFelipe and Farin ̃ as, 1992; Elston and Rosa, 1997, 1998) and from lower species (owl monkey, marmoset) up to humans (Elston, 2003). ... These observations confirm that PFC [prefrontal cortex] cells exhibit the morphological adaptations needed for massive long-distance communication, information integration, and broadcasting postulated in the GNW [global neuronal workspace] model and suggest that this architecture is particularly developed in the human species."

Outside the brain, search has evolved rapidly over cultural history. First, we gave up our deep cultural memories, such as epic poems and stories, in favor of writing and books, such as scriptures. Now we give up our medium-term memories to the internet, not bothering to remember the blizzard of factlets that can so easily be looked up on Wikipedia. Perhaps the next step is supplementing consciousness itself- our moment-to-moment short term memories that make up the database that connects everything in our heads into meaningful constructs. While technically fanciful, or at any rate very far off, (cue the singularity people), if we could get faster and bigger analog memory storage from chips implanted into our heads, and connections between them made compatible with the existing consciousness network, then search would be internalized and create new levels of consciousness.

As an example, head-mounted displays are already in existence that automatically annotate ambient scenes, for instance for military pilots. So imagine looking out, but instead of a bare streetscape, it is covered with highlighting colors or text annotations that indicate properties of interest, like restaurants, or street signs, or dirt ... whatever you are concerned about. This would be a richer form of consciousness than we are normally used to, though still limited by what we can pack through our existing sensory apparatus. Suppose that these annotations were internally generated and spontaneously pop up in response to whatever we were thinking or looking at ... that would be consciousness itself.


Some more interesting papers and links on the topic:
  • Phase transitions and chaos.
  • Gamma and the coding of consciousness.
  • More on gamma.
  • Relations of gamma and theta waves.
  • Not all gammas are the same.
  • Another all-around review of the field.
  • The demented Stuart Hammeroff retreats from quantum consciousness, to only slightly more plausible dentritic gap junction consciousness.
"Interestingly, recent research also suggests that spontaneous brain activity, as assessed by resting-state EEG recordings, may be similarly parsed into a stochastic series of slow ‘‘microstates,’’ stable for at least 100 ms, each exclusive of the other, and separated by sharp transitions (Lehmann and Koenig, 1997; Van de Ville et al., 2010). These microstates have recently been related to some of the fMRI resting-state networks (Britz et al., 2010). Crucially, they are predictive of the thought contents reported by participants when they are suddenly interrupted (Lehmann et al., 1998, 2010). Thus, whether externally induced or internally generated, the ‘‘stream of consciousness’’ may consist in a series of slow, global, and transiently stable cortical states (Changeux and Michel, 2004)."
  • Unemployment is also a civil rights issue.
  • Stiglitz speaks.
  • Trickle down? Nope. Wages are declining.
  • Corporations to rule over us all.
  • Christians still hunting for Adam and Eve.
  • It's just a fish, so who cares
  • Another telling of the regulatory and free market debacle.
  • Bradley Manning, Hero.
  • Economics quote of the week- Warren Buffet:
"If Greece could print its own drachmas, it wouldn't have a debt problem."

No comments: