NOTE!: Ancient Origins of Consciousness by same authors in 2016!

See Also: Book Notes, (me), Notes on Consciousness, Consciousness and the Brain, Ancient Origins of Consciousness, Sapolsky: Behave, How Emotions Are Made, Consciousness and the Brain, Consciousness: Confessions, The Quest for Consciousness, Happiness Hypothesis, Blank Slate, Info Viz & Perception, On Intelligence, Neuroscience of Human Relationships, Human: Makes Us Unique, Thinking, Fast and Slow, Righteous Mind, Ravenous Brain

Consciousness Demystified

Todd E. Feinberg and Jon M. Mallatt

Septmber 2018 | MIT Press | ISBN: 9780262038812
122 pages of text, 40 Illustrations (10 on color plate), 14 page glossary, 18 pages of notes, 32 pages of references and 13 pages of index.

This book is a concise description of the brain structures and processes that give rise to Sensory, aka Pheonomal Consciousness, which consists of Exteroception (5 senses), Interoception (gut feelings) and Affect (emotions). For me, this provides a comprehensible working model of consciousness, especially the hardware and firmware aspects.

Consciousness Demystified extracts and refines the core ideas about consciousness from the fantastic book they wrote in 2016, Ancient Origins of Consciousness, which traced the evolution of the structures of the brain that enable consciousness. It also expands the discussion of closing the "Explanatory Gap", aka the "Hard Problem", which seems to be a modern form of Dualism. I was sold by the 2016 book and do not mention it here. I think the expanded treatment will make the book more powerful to new readers.

Please note! This is NOT an attempt to summarize this book. It goes well beyond what I put here. These are just a few notes to myself, and you.

The key concepts carry over from 2016: 1) NeuroOntologically Subjective Features of Consciousness, and 2) The Defining Features of Consciousness, although the articulation of the defining features is greatly expanded. The explanation of the the bridge of the explanatory gap is also greatly expanded. The new concepts are a gradation from image based consciousness to affective consciousness, with an emphasis on the maps. See Plate 1 to the right. There are fantatic illustrations, as with the 2016 book.

Isomorphic Maps (image-based consciousness)

Isomorphic, from American Heritiage, Biology: Having a similar structure or appearance but being of different ancestry.
This is key and Feinberg & Mallatt paint the picture better than anyone. Plate 1 is amazing. All visual information and probably audio and perhaps much more is represented in 2D maps that are overlayed one upon the other to integrate all sensory data. It is like an image with many layers that automatically binds the properties of the red Ferrari zooming by.

p.40 (2018):   To summarize the chapter, isomorphic maps are the corner stone of image-based sensory consciousness, these maps evolved in early vertebrates more than 520 million years ago, and this process was the natural result of the extraordinary innovations of the camera eye, neural crest, and placodes. These events led to the mental images that mark the creation of the mysterious explanatory gaps and the subjective features of consciousness outlined in chapter 2.

Box 4.1: NoSFC: NeuroOntologically Subjective Features of Consciousness

NSFCs are carried forward from 2016. A bit like the NCCs of Christof Koch - Neural Correlates of Consciousness. NoSFCs are NeuroOntologically Subjective Features of Consciousness. This phrase is important for closing the Explanatory Gaps. NoSFCs:

1) Mental Unity - we have one coherent notion of Reality. See: Unitary Nature of Consciousness
2) Qualia - subjective perceptrons of qualities.
3) Referral, aka Projection - Reality seems to be Out There, not in our heads.
4) Mental Causation, "Using our notion of Reality as a map to direct our actions to objects in the environment on which we cause affects." (imagining how to interact shapes what we see.)

p.49   The intercommunicating axons of affective pathways also branch a lot more than in the exteroceptive pathways, sending signals to many different parts of the system. Another difference is that affective circuits communicate less through short-distance neurotransmitter chemicals and more through far-diffusing neuromodulator chemicals than do exteroceptive circuits.'
  While these distinctions may seem fuzzy, recent studies in rodents provide more concrete information about affective structures. Individual neurons in the affective system encode distinct valences (fig. 4.2A) [jch: only plate 1 copied, sorry]. That is, one class of neurons only carries the positive signal of reward, and another class carries the negative signal of aversion. These valence neurons code their + or - signal both in their electrical firing patterns and in the kinds of chemicals they release to signal other neurons (such as dopamine, which signals an anticipated reward). Valence neurons have distinctive locations, targets, and networks within reward and aversion pathways. In some places, they form "valence maps" where, for example, + neurons produce pleasure "hot spots" in the brain (fig. 4.2B).

p.50   With both thirst neurons and danger neurons, it seems that neurons can code affective information that is far more sophisticated than just a simple liking of disliking.
  ... affective system seems diffuse, with its different intercommunicating parts overlapping one another in function, the parts do specialize for their own affective functions. These parts are shown in figures 4.1 and 4.2 [jch: not here], and they include the amygdala for fear and emotional learning, lateral habenular nucleus for coding punishment and disappointment, nucleus accumbens for motivated seeking of rewards, a part in the anterior reticular formation for arousal,' and the parts of the cerebral cortex that use reason to control one's core emotions and passions (e.g., the orbitofrontal cortex in fig. 4.2). Upon further study, this affective system could prove to have a well-organized and highly differentiated structure, after all, an organization that researchers are just starting to recognize.

p.66,67,68,69: The Defining Features of Consciousness. These are much more fleshed out than in 2016.

Box 6.1: The defining features of consciousness, Level 1: General biological features, which apply to all living things

Life, embodiment, and process
-- Life: use of energy to sustain self, growth, responsiveness, reproduction, and adaptiveness to change. All known life is cellular.
-- Embodiment: a body with an interior separated from the exterior world by a boundary.
-- Process: life functions are complex, dynamic processes, not material things.

System and self-organization
-- System: entity considered as a whole, in which arrangements and interactions between the parts are important.
-- Self-organization: interactions of the parts organize the patterns at global level of the whole system.

Hierarchies
-- Hierarchy: complex system with different interacting levels, organized from simpler to more complex. Levels may be nested within one another: e.g., macromolecules to cells to organs to the organism. New emergent features may naturally appear in the whole system by the addition of new levels and their interactions with lower levels.

Teleonomy and adaptation
-- Teleonomy: biological structures perform programmed, goal-directed functions.
-- Adaptation: a teleonomic structure or function as evolved by natural selection.

Box 6.2: The defining features of consciousness, Level 2: Neuronal reflexes and simple core brains

Rates
-- Reflexes and all neural communications are extremely fast relative to other large-scale physiological processes. Thus they can move a large body in response to a stimulus.

Connectivity
-- Simple reflex arcs are chains of several neurons connected at synapses. More complex arcs have more neurons in the chain (C) and in networks (N); they also have more sensory input (S), more neuronal interactions, (I) and capacity to process more information (P).

Increasing complexity
-- Further increase in CNSIP leads to complex nervous systems and consciousness.

Basic motor programs from central pattern generators
-- These nonconsciously control essential, repetitive behaviors

Core brain features
-- Modulatory, sensorimotor-integrating centers for arousal and directing attention.
-- Complex reflexes for inner-body homeostasis.
-- Rhythmic locomotion and other basic motor progams, [jch- brainstem pattern generators to sync...]
-- Automatic, not conscious.

Box 6.3: The defining features of consciousness, Level 3: Special neurobiological features, which apply to animals with primary consciousness

Neural complexity (more than in simpler, core brain)
-- A brain with many neurons.
-- Many subtypes of neurons.

Elaborated sensory organs
-- Image-forming eyes; multiple mechanoreceptors for touch; separate chemoreceptors for smell and taste.
-- High locomotory mobility, to gather the abundant sensory information.

Neural hierarchies with unique neural-neural interactions
-- Extensive reciprocal (reentrant, recurrent) communication occurs within and between the hierarchies for the different senses.
-- Synchronized communication by brain-wave oscillations may be required for sensory binding and generating mental images.
-- Higher levels of the hierarchy allow the complex processing and the unity of consciousness.
-- Hierarchies allow consciousness to predict events a fraction of a second ahead in time.

Neural pathways that create mapped mental images or affective states
-- Isomorphic representations: neurons are arranged in topographic maps of the world or body structures.
-- Affective states: from valence coding rather than from topographic mapping; the hierarchy and networks are more diffuse with more centers and more use of neuromodulators.
-- Both feed into premotor brain regions to motivate and guide behaviors in space.

Attention as a participant in consciousness
-- Selective attention mechanisms in brain: for focusing consciousness onto salient objects in the environment. Related feature of arousal is also present, adjusting the level of consciousness.

Memory
-- Short-term, minimal sensory memory is needed for the continuity of experience in time. Longer-term, higher-capacity memory evolved soon after consciousness arose. However, we cannot rule out more memory being present from the start, before consciousness arose.

p.82   From a different perspective, however, consciousness is valuable because it enables prediction processes, for which neural hierarchies are essential. The predictive mechanism, as explained in figure 6.5, depends on up-and-down communication within the hierarchy. As long as the predictions are kept updated, they stay a fraction of a second ahead of what is happening in the world - a huge benefit for survival whenever the conscious animal successfully predicts the attack path of a predator [ or prey/threat/reward in near term ]

[jch: NOTE!!! Bringing near term predictions to consciousness is different from prediction processing, which seemed to be covered in the 2016 book. [add to 2016 and put in link] Figure 6.5 stands in stark contrast to Jill Gregory's usual rich and detailed and attractive illustrations. ]

p.86   Our approach to the first question is to ask: what is the minimal memory requirement for sensory consciousness? Christof Koch conservatively reasoned that image-based consciousness needs at least a fleeting type of information storage called iconic memory. Actually, a better name is sensory memory, with the term "iconic memory" serving only for visual image perception (lasting less than half a second), "echoic memory" for sound perception (lasting about three seconds), "tactile memory" for touch (lasting about ten seconds), and "odor memory" for smell (lasting longer). The term "sensory memory" includes all these functions.

[jch: in prediction processing, this short term memory would be a partial mental model, broadcast from the unified mental model, and then relayed thru the icon buffers to - region -- be compared against current input. Instead of iconic memory of the past, we have the refreshed prediction. ]

p.86   This neuron type is in the parabrachial nucleus of the interoceptive pathway (fig. 2.3), and its signals elicit fear responses from affective brain centers like the amygdala,' With both thirst neurons and danger neurons, it seems that neurons can code affective information that is far more sophisticated than just a simple liking or disliking.

p.110   Conscious unity at the higher level of the special features is directly linked to reciprocal neural interactions that bind coded sets of sensory information together into a unified image or affect (top of plate 10). However, down at the more basic levels, all physiological life processes are integrated and unified to achieve homeostasis, and the reflexes are genetically prewired to create linked programs that effect unified actions. In short, the unity and integration that result from the special features (box 6.3) stem from the unified systemic features of life and the reflexes at the lower levels. Again, we have derived a gap feature at the top of plate 10 ("unity") from the physical features lower down in the figure.

p.109   ..., we also found that the neurobiological basis of primary consciousness, along with its accompanying subjective features like qualia, is remarkably varied both within individual brains and across the brains of different species from vertebrates to arthropods to cephalopods. To reinforce how diverse the neural contributors to consciousness are, let us reconsider a part of the vertebrate brain that we have only touched on: the wide-ranging reticular activating system and thalamus that contribute to attention, alertness, arousal, and wakefulness, without which consciousness and therefore subjectivity would not be possible (chaps. 3 and 6; fig. 4.1).3 This immensely broad distribution of structures that contribute to consciousness - perhaps the broadest of all neural functions - ensures that the neural basis of Consciousness will be at best poorly localizable within the brain.

Color Plates!

As with the 2016 book, Jill Gregory's rich and detailed and attractive illustrations add tremendous value. In Demystifing Consciousness, there are 10 color plate illustrations. No list with names, but, they are so key, I list them here:
  Plate 1: Isomorphic (normalized) Maps of Sensory Inputs - See above *** Amazing ***
    [jch: image-based consciousness automatically binds color, object, motion, etc.]
  Plate 2: Subtypes of Consciousness: Exteroceptive, Interoceptive, Affective...
  Plate 3: Optic Tectum in Monkey & Cod with Layers of Cortex ** Great ! ! **
  Plate 4: Sensory Pathways - [jch: most road lead thru Thalamus
  Plate 5: Basic Valence Circuit -
  Plate 6: Comparison of Frog, Octopus, Insect brains ** Wicked Aweseom ! ! **
  Plate 7: Brain Regions showing Sensory and Affective structures
  Plate 8: Stages of Evolution of structures of Consciousness
  Plate 9: Hiakouichthys (fish)
  Plate 10: Summury Illustration.

NIT: The definition of consciousness is not as crisp as in 2016. The authors use a long quote as part of the definition, and at one point say "the capacity to have any experience at all" and at other points, the authors use the "What is it like to be a bat" notion of consciousness, that seems polular. It was used as a working defintion by Anil Seth / Sam Harris in a podcast. To me, pinning down the definition is not as big of a deal as it used to be. I still love the defintion they used in 2016:

p.111(2016):   "But to us, real consciousness is indicated by the (optic tectum) making a multisensory map of the world and then attending to the most important object in this map and then signalling behaviors"... based on the map.

GEEKy suggestion, put box 4.1 on page 41 and box 6.1 on page 61.
P.S. the 2016 book inspired some Sensory Integration Art

jch.com/jch/notes/FeinbergMallatt2018.html 2018.11.22   YON   Book Notes