The Octopus Analogy for Thought

The same process that gives the octopus mobility and stability with its arms,

gives us mobility, stability and continuity in our thought.

Thinking about how thought changes within the brain through time, I came to the conclusion that groups of neurons interact in a very specific, repeating pattern over time. I tried to come up with an analogy to explain this pattern. I  likened it to the pattern of footsteps taken by an octopus that is “seafloor walking.” I feel the analogy works in a few different ways, but the main point is that while much activation in the brain is often rapid and transient, some activation is relatively more long lasting. I think that the long lasting activations lend structure and continuity to thought and contribute to the phenomenon of sentience.

Octopuses occasionally walk along the seafloor and when they do, they often exhibit a particular stride. As they move forward, they plant some of their arms on the sand below them and replant arms that they have left behind. Imagine yourself climbing up a ladder. You hold on to each rung until the rung is so low that you let go and reach for a higher one. Now imagine climbing a ladder with many arms. They wouldn’t all move at once, only the lowest ones would move. This is very similar to how an octopus walks. The key idea is that, at any one point in time, some arms retract from their positions, some arms are relocated to new positions but yet most remain planted firmly where they are. This is very much like the pattern of activation in the cerebral cortex.

Individual nodes (groups of cells that are wired together) in the brain correspond to specific mental representations. Each thought that we have necessitates large numbers of these nodes to represent the various things going on in our mind. The octopus analogy uniquely describes a system where certain nodes are conserved through time as others come and go. The stride of an octopus that plants the majority of its arms temporarily while actively repositioning arms that have let go of their footholds, represents the uninterrupted, nonlinear, spatio-temporal pattern of node activation, deactivation and coactivation in the brain.

Some nodes can probably be retained even after the transitions between a number of thoughts. This happens when your thoughts cycle and change but hold a common element or theme constant. When we attempt to solve a novel and complex problem we try to keep the majority of our octopus arms firmly planted so that we can keep the problem set in mind. The fact that some nodes (within association areas) remain active for prolonged periods (i.e. the octopus arms remain planted), during reciprocal top-down to bottom-up communications, accounts for the continuity found between successive brain states. The longer nodes in association areas can be continuously activated - over a series of states - the longer they can influence sequences of bottom-up imagery in a sustained and consistent way allowing modeling, planning and working memory in general. The result is a stream of consciousness where each thought is quantitatively different from the ones preceding, as newly relevant nodes are added and the least relevant ones are removed.

The fact that the placement of some of its arms are conserved, over sequential moments, gives the octopus balance and stability and it also provides the physical basis for the continuity of thought. The nodes that are conserved reside mostly in association areas whereas nodes in early sensory areas are activated much more transiently. This is as if the arms in back of the octopus (corresponding to posterior sensory cortices) move much more quickly than the forelimbs (which find firm, reliable footholds in anterior association cortices and the PFC). The octopus analogy has the advantage of demonstrating how several interacting elements combine to allow thought. Also because these elements remain active for different durations, thought does not stop and go in discrete steps but is continually “carried along” by those elements that endure through time. All elements, or neural nodes, will deactivate within several seconds, but the intermingling of nodes of some temporal stability with those of more fleeting persistence sustains the associative bridges that allow the thematic and unifying consistency that is a hallmark of cognition.

The routine of node activation and deactivation is very similar to “polypedal locomotion” or movement in animals with many legs. It is not much like the locomotion of an insect such as a millipede or a centipede though because these animals move their legs in stereotypical, repetitive ways where the placement of each leg is not actively influenced by the placements of other legs or of the qualities of the footholds. The pattern of activations in the brain is more like the polypedal locomotion of an octopus that is seafloor walking because it is asymmetrical, dynamic and the placement of the next legs is influenced by the octopus’ stance, posture and the characteristics of the footholds themselves. Most importantly, this model can accommodate nonlinear aspects of neurodynamics. One neural node does not activate the next in sequence. Several nodes are coactivated together and they pool their activation energy to determine which nodes will be activated next.

The total number of cortical nodes that can be coactivated must be somewhat stable given known limitations on things like neural excitability, cortical hemodynamics and working memory. In our analogy the number of available octopus arms is very stable and this represents our fixed, innate capacity for working memory. Even though the number of chunks (psychologically perceptible units of perception and meaning) that can be held in working memory, 7 plus or minus 2, coincidentally coincides with the number of arms that a living octopus has (8), this is not a reliable indication of the number of nodes that can be coactivated in association cortex. This is true because even though chunks and nodes may be relatively congruent, the exact relationship between them is currently unclear. It seems clear though that the octopus has a relatively invariant number of arms and that perhaps, in order to bring a new node into the train of thought it must first let go of some other node. Surely the number of activatable nodes differs from area to area and from task to task, but it probably remains relatively constant within tasks.

It is not always the case that the majority of nodes are conserved from one thought to another. Most nodes can be dropped or abandoned at the same time, i.e. when they become a lower priority. This readily happens when we are exposed to a new, salient, perhaps emotionally laden, stimulus. When this occurs, the octopus “jumps,” taking all of its arms with it, and reorients to the new stimulus and its accompanying set of features. Such a jump would constitute a disruption of mental continuity. So clearly mental continuity can be viewed on a continuum where a high proportion of nodes are conserved between brain states on one end of the continuum and a low proportion are conserved on the other end. Disruptions in continuity might occur due to a distracting stimulus in the environment, or from an internally generated stimulus. Evolution has probably programmed the octopus to jump and reposition its arms quickly in order to respond to important sensory stimuli, so that mammals react to them with all of their cognitive resources. Mental continuity is less easily disrupted in humans than it is in other mammals, although perhaps more easily disrupted in people with habituation deficits. Attention and distraction must be intimately related to the temporal conservation of nodes. In fact, the extent of attention deficit and distractibility should be inversely related to the neurological capacity to conserve nodes in association areas from second to second.

Another analog of this analogy is the idea that the octopus will “topple” if it loses its grip on a sufficient number of nodes. This makes the body of the octopus analogous to consciousness because brains become unconscious once coactivation (especially in the frontal and parietal fields) is sufficiently diminished. Nodes in early sensory areas are often active during unconscious states, but nodes in association areas are less active and out of sync with those in sensory areas. Thus anterior-posterior balance and coordination are important for our allegorical octopus.

Unlike subcortical areas, strictly one-to-one, linear activation is probably rare in the cortex. Also, unlike subcortical areas, information processing in the cortex is not compartmentalized into individual nuclei that are relatively isolated from processing occuring elsewhere. Rather, cortical nodes coactivate together to spread the activation energy necessary to recruit, or converge upon, the next set of nodes that will be coactivated with the remaining nodes from the previous cycle. An node is released from coactivation when it no longer receives sufficient activation energy from its inputs i.e. its relevance to the processing demands diminishes. When a node is deactivated, the perceptual or conceptual element corresponding to it becomes deactivated and can no longer impact present experience. Whatever new node is introduced will inform the present sum of coactivates in a unique and informative way.

The nodes that are coactivated, at any one time, sum their component features together to create mental imagery and this occurs in both sensory and association areas.  Mental imagery changes plastically as nodes that continue to be useful are maintained, nodes that are rendered less useful are released from activation, and nodes that are newly recognized as useful are activated and incorporated into the remaining amalgamation of useful coactivations.

Table 1: Definitions of Terms

	Psychological Aspects	Neurological Aspects	Octopus Analogy Analog
Neuron	Variable if not negligible	A single cell	A grain of sand on the cortical seafloor that the octopus stands on
Neural Assembly	Element, feature, or fragment of a construct in long-term memory	A cortical minicolumn or a collection of cells with very similar receptive fields	A patch of sand that is currently in contact with a suction cup on an octopus arm
Neural Ensemble	A psychologically perceptible construct of long-term memory that can serve as a feature of a current thought	A collection of coactivated assemblies that are bound in a Hebbian manner	A region of sand that is currently in contact with a single octopus arm and its suction cups
A Thought	A composite of several representations that combine to create mental imagery	A set of coactivated ensembles	The set of all octopus arms that are currently in contact with the cortical seafloor
Thinking / Consciousness	A progression of related imagery formed through reciprocating transformations between association and sensory cortex	A sequence of related sets of coactivated ensembles where some remain active over the duration	The locomotive behavior or past and present footsteps of the octopus
Unconscious Processes	Implicit mental behavior unavailable to psychological introspection	The connectivity responsible for the selection of assemblies and ensembles	The automatic processes corresponding to the selection of arm placements

Read the full article that I wrote on this topic here:

http://www.sciencedirect.com/science/article/pii/S0031938416308289

If you found this interesting, please also consider visiting aithought.com. The site delves into my model of working memory and its application to AI, illustrating how human thought patterns can be emulated to achieve machine consciousness and superintelligence. Featuring over 50 detailed figures, the article provides a visually engaging exploration of how bridging the gap between psychology and neuroscience can unlock the future of intelligent machines.

Oliver Sacks’ Confabulating Butcher

I found a lego minifigure the other day who is dressed as a butcher and it reminded me of the the delicatessen owner in Oliver Sacks’ book, The Man Who Mistook His Wife For A Hat. In the chapter titled “A Matter of Identity,” Sacks enters a clinic room where he is greeted by a delusional Mr. Thompson. Mr. Thompson is a former deli owner who, because of Korsakoff’s dementia, struggles to remember anything new. The disorder is often due to severe alcoholism and is associated with anterograde amnesia, retrograde amnesia, lack of insight, apathy and confabulation. Confabulation is a verbal statement that contains a memory disturbance that causes the person to inaccurately describe the past or present. It is often referred to as honest lying.   I think that confabulation, as a natural phenomenon, is fascinating – mostly because I catch myself doing it on small scales all the time. We all confabulate when we tell stories or relate past events. We add small bits that don’t belong as our memory tries its best to fill in gaps. Mr. Thompson’s degree of confabulation though, taught me a lesson that I’ll never forget. The following passage showed me that when memory fails utterly, humans have a tendency to compensate, by inventing reality:

'What'll it be today?' he says, rubbing his hands. 'Half a pound of Virginia, a nice piece of Nova?' (Evidently he saw me as a customer—he would often pick up the phone on the ward, and say 'Thompson's Delicatessen'.)

'Oh Mr Thompson!' I exclaim. 'And who do you think I am?'

'Good heavens, the light's bad—I took you for a customer. As if it isn't my old friend Tom Pitkins ... Me and Tom' (he whispers in an aside to the nurse) 'was always going to the races together.'

'Mr Thompson, you are mistaken again.'

'So I am,' he rejoins, not put out for a moment. 'Why would you be wearing a white coat if you were Tom? You're Hymie, the kosher butcher next door. No bloodstains on your coat though. Business bad today? You'll look like a slaughterhouse by the end of the week!'

Feeling a bit swept away myself in this whirlpool of identities, I finger the stethoscope dangling from my neck.

'A stethoscope!' he exploded. 'And you pretending to be Hymie! You mechanics are all starting to fancy yourselves to be doctors, what with your white coats and stethoscopes—as if you need a stethoscope to listen to a car! So, you're my old friend Manners from the Mobil station up the block, come in to get your boloney-and-rye ...'

William Thompson rubbed his hands again, in his salesman-grocer's gesture, and looked for the counter. Not finding it, he looked at me strangely again.

'Where am I?' he said, with a sudden scared look. 'I thought I was in my shop, doctor. My mind must have wandered ... You'll be wanting my shirt off, to sound me as usual?'

'No, not the usual. I'm not your usual doctor.'

'Indeed you're not. I could see that straightaway! You're not my usual chest-thumping doctor. And, by God, you've a beard! You look like Sigmund Freud—have I gone bonkers, round the bend?'

'No, Mr Thompson. Not round the bend. Just a little trouble with your memory—difficulties remembering and recognising people.'

'My memory has been playing me some tricks,' he admitted. 'Sometimes I make mistakes—I take somebody for somebody else ... What'll it be now—Nova or Virginia?'

So it would happen, with variations, every time—with improvisations, always prompt, often funny, sometimes brilliant, and ul¬timately tragic. Mr Thompson would identify me—misidentify, pseudo-identify me—as a dozen different people in the course of five minutes. He would whirl, fluently, from one guess, one hypothesis, one belief, to the next, without any appearance of un¬certainty at any point—he never knew who I was, or what and where he was, an ex-grocer, with severe Korsakov's, in a neurological institution.

He remembered nothing for more than a few seconds. He was continually disoriented. Abysses of amnesia continually opened beneath him, but he would bridge them, nimbly, by fluent confabulations and fictions of all kinds. For him they were not fictions, but how he suddenly saw, or interpreted,the world. Its radical flux and incoherence could not be tolerated, acknowledged, for an instant—there was, instead, this strange, delirious, quasi-coherence, as Mr Thompson, with his ceaseless, unconscious, quick-fire inventions, continually improvised a world around him—an Arabian Nights world, a phantasmagoria, a dream, of ever-changing people, figures, situations—continual, kaleidoscopic mutations and transformations. For Mr Thompson, however, it was not a tissue of ever-changing, evanescent fancies and illusion, but a wholly normal, stable and factual world. So far as he was concerned, there was nothing the matter…


From: The Man Who Mistook His Wife for a Hat. 1970. Touchstone. Chapter 12.

To read more about my own memory failures that I have noticed click here:

My Personal Experience with Chronic Stress: Compromised Cognition


Here are a few of my favorite books by Oliver Sacks:

How Thought Propagates Part 2

I believe that consciousness and working memory are driven by the reciprocating cross-talk between fleeting bottom-up imagery in early sensory cortex and lasting top-down priming in late association cortices. The images below attempt to explore how this reciprocal interaction works. This model depicts how the central executive (the association areas) directs progressive sequences of imagery in the visuospatial sketchpad and the phonological loop (sensory areas).

Diagram 1: Each concept held tonically active in association cortex is represented by a letter. The order these letters are pulled into consciousness arbitrarily follows alphabetical order. 1) Shows that concepts A,B,C,D and E all diverge onto posterior visual sensory cortex. 2) Shows that neurons involved in the retinotopic imagery from figure 1 converges on the association neurons responsible for holding concept F. It also shows that concept B drops out of activation, and that A, C, D, E and F remain active and diverge back onto visual cortex. 3) Shows that this same process leads to G being activated and A being deactivated in association cortex. 4) Shows that H is activated and that E is deactivated.

Diagram 2: Sensory areas can only create one sensory image at a time, whereas association areas are capable of holding the salient or goal-relevant features of several sequential images at the same time. In other words, neurons in association cortices can remain tonic for a number of seconds, making activity there progressive because it can reflect the most important aspects of several successive instances of sensory activity. Dopamine is key to this progressive process because it helps to determine which features are relevant/motivating. This process can be described by the following list:


1)  Sensory imagery is made in early sensory cortex. This is either a product of feedforward sensory information from sense receptors or from downstream retroactivation.


2)  Salient features are extracted and communicated with higher association areas


3)  Salient features that cohere with already active features in association areas (that are consistent with the present set of concerns) are added to the active features there, whereas the least relevant features in association areas are dropped from activation.


4)  The important features of the last few images are maintained active in association areas.


5)  The new set of features is handed down to early sensory areas.


6)  This new set of features is used by early sensory areas to construct topographic imagery. In reality, sensory cortex from each sense modality feeds their inputs to association cortex (and to each other).


7)  The process starts over.


A few months ago I thought to myself that if the association cortex could be protagonized what it has to say to earlier sensory cortex would be along these lines:

“Ok, I see the imagery that you have just generated and I have used it to update my own list of relevant concerns. Now, I believe that this updated list of concerns will change our perspective on the last few thoughts. Take this new concern, along with these other concerns that we have recently been handing back-and-forth and do your best to portray them all simultaneously in a single image. Then the image that you build will pass up our hierarchically-structured interface updating me of the new, goal-relevant concerns that your imagery was able to automatically identify”

This reasoning frames consciousness as a  polyconceptual, "partially-conserved, progressive process,” that performs its high-level computations through “reciprocating transformations between buffers.” More specifically, it involves “reciprocating transformations between a partially conserved store of multiple conceptual specifiers and another nonconserved store that integrates these specifiers into veridical, topographic representations.”

In other words, the association cortex is always saying the following to sensory cortex: “Ok gotcha, well then what about this, can you rapidly create an informative topographic image with these specifications?” And the early sensory cortices play along by cooking up the best image they can to corepresent the constructs handed down to them. In my opinion, these active nodes in frontal areas create the executive, and its top-down cognitive control. The diagram above makes this process seem somewhat random, but its not. The selection of features in association areas has been programmed by personal history and thus they have the capacity for creating "us." For more information see: "How Thought Propagates Part 1."

Read the full article that I wrote on this topic here:

http://www.sciencedirect.com/science/article/pii/S0031938416308289

If you found this interesting, please visit aithought.com. The site delves into my model of working memory and its application to AI, illustrating how human thought patterns can be emulated to achieve machine consciousness and superintelligence. Featuring over 50 detailed figures, the article provides a visually engaging exploration of how bridging the gap between psychology and neuroscience can unlock the future of intelligent machines.

The way memories are recorded and reused is analogous to the way tracks of water form on a window pane

New droplets fall into the paths of old droplets

On a rainy bus ride in 2004 I was staring out of the window, hoping to push myself into thinking up a novel conceptualization of the neurological basis of memory. Looking up from an article on axonal migration, my attention turned to the tracks of water that had formed on the window. You know how they look. Water droplets leave trails of water behind them as gravity pulls them down the pane. Early drops form trails that subsequent drops fall into and travel within. Of course these are trails of water on glass, whereas memories are recorded in tracts made of neurons and their projections. But the analogy felt substantial to me, and I think that it can be carried at least a short way before it fails.

The qualities of the surface of the glass and the placement of the water droplets on the glass determine how the drops will travel. This is similar to how the current physical state of the brain and the locations of the activations therein determine how neural pathways are selected. During every perception, cognition and experience old neural pathways are being retraced (memories, familiar perceptions) and new pathways are being created.

Some water droplet's trails extend long distances and subsequent drops fall directly within their borders. Depending on dirt and smudges on the glass other trails dry up before they extend very far and when a subsequent drop reaches this point, it is free to delimit its own path. Similarly our memories and experiences involve the activation of neural trails some of which are only small snippets of the original encodings.

Altering synaptic weights among groups of neurons creates patterns in the nervous system that are stable and that serve as preexisting pathways that constrain the future propagation of nervous energy. In this way, previous experiences imprint stable conduits that act as barriers for subsequent stimulation, thus the activation patterns elicited by new stimuli will not be random, they will be constrained by past activity. This accounts for how humans respond to disparate situations in the same inflexible ways, it accounts for how stubborn and resilient our personalities are, and accounts for how most of our actions are simply pastiches of previously used techniques. Consciousness and higher-order thought really are composed of multiple stable states organized in different configurations. In this way, consciousness is akin to the entire window and its current streaming activity.

This perspective feels limiting and deterministic, but it is important to point out that our window is huge and that we should stimulate and push ourselves to increase the number and complexity of stable states in our repertoire ensuring that our pane of glass has many possible tracks. Also, the pane of glass on the bus sits passively as water streams down its side and creates the pathways. Humans, on the other hand do not have to have sit passively and be written upon. With the right environmental provocations we can intentionally seek out activities that will influence the creation of new pathways.

I had always wondered how memories, thoughts and behaviors were made possible by something physical. I wondered how alterations in the connections between brain cells allowed mental representations that could remain stable over days or years. Like the water droplets, neural energy finds the path of least resistance, and information in the form of action potentials is pushed down those avenues that are best facilitated by the synaptic weights in the network. The analogy here is very simple. Today it seems so obvious as to be trivial, but I remember it being powerful and instructive for me at first consideration.

How The PFC Leads to Behavioral Inhibition

After hitting snooze this morning I had a dream that I was reading a Wolverine comic written by Chris Claremont and drawn by John Byrne. In the comic, Wolverine was confronted with a villain he was in an argument with and he was being provoked. The dream’s narrator said something along the lines of: “Wolverine had an inclination to unsheathe his claws and cleave his foe in two that moment but another impulse, to remain composed and noncombative outlasted it. This second impulse subdued the first. It won out merely because it remained firing in his brain for just a moment longer than the first.”

The PFC has cells with higher-order receptive fields that can help the brain to conceptualize abstract things like long-term goals, cost-benefit analyses and the repercussions of risky acts. Often its instructions are the opposite of what our lower-order brain systems try to impel us to do. The PFC also has cells that fire for longer. Could it be that simple? In Wolverine’s case above, could that instinctive impulse to attack have been impelling and strong but temporally more fleeting than the one that stayed his hand? If that hypothetical inhibitory ensemble in his PFC would have stopped firing earlier would he have engaged?  In humanity’s case could our culture and history be due to the fact that the brain areas that hold the more complex representations fire for just a few milliseconds longer, enabling the more complex behavioral instructions to narrowly outlast the more base ones?