Menu directory status & updates copyrights help

Howell: What is consciousness: from historical to Grossberg

What is consciousness? - from historical to Grossberg

"... Many theories of consciousness do not [address, emerge from] lower-level [neuron, module, brain, pyschology, neuroscience] cart [architecture, process]s which can [model, predict] data. Are those theories now almost irrelevant, except as inspirations and the rare gems scattered in their verbiage? After all, most of the core concepts of consciousness have been around for at least a hundred years. ..." (Howell Apr2023)


Consciousness: table of comparisons

Very few concepts of consciousness are listed here, compared to what is available in the literature. My intent is just to provide an idea of several [historic, popular, interesting (to me)] concepts, and their diversity, for the most part relying on Wikipedia as a "somewhat unbiased popular mainstream source" rather than reviews of the subject. More detailed descriptions of each concept in the table can be viewed via the links.
concept
[name, authors]
consciousness components neural architecture, computational models
Pre-historic concepts related to consciousness [consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, social [cooperative-competitive, inhibition-excitation, consciousness]]

Tongue in cheek : It seems inconcevable to me that prehistoric people would NOT have had a very [solid, broad, practical] understanding of many of the components of consciousness and some idea of how they work together, even if they didn't formally define consciousness, or language had not yet even been developed? That may sound strange, but I distinctly remember facing harsh critiques when I started working, and didn't [focus, pay attention, anticipate outcomes, etc, etc]. Furthermore, the social aspects such as [team, cooperation, competition, top-down, bottom-up] (many are the same as for the individual) became immediately clear from kindergarten on. I've highlighted in red font some of the basic concepts related to consiousness in the list below. There isn't much that the pre-[historic, language] humans missed, I suspect, and some of those that aren't highlighted below are usually missed in modern theories of conscipousness too.
neural architecture, computational models :
no computational models - there were no computers
[Ancient, medieval] philosophy [consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

philosopher concept
Aristotle r2c2
Rene Descartes r4c2
Thomas Hobbs? r3c2
??? r5c2
neural architecture, computational models :
no computational models - as with prehistoric, there were no computers, but conscious processes in the brain are more computationally powerful than our computer models anyways.
Social Consciousness
(Sigmund Freud ?early 1900s?, Carl Jung ?mid 1900s?)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

Modern concepts on Social Consciousness are perhaps [elaboration, formalization]s of :
  • pre-language (probably pre-human) [knowledge, practice] (... here I should look into studies of possible consciousness in animals (especially primates) ...)
  • verbal (pre-written) [behaviour, myth, religion]
  • perhaps early written [math, science] as part of religion etc?
However, modern attempts have the [resources, training, knowledge] to apply formal [hypothesis, data collection, analysis, modelling]
neural architecture, computational models :
I haven't looked for these - but there are good computational models in psychology, perhaps for social consciousness too?
Heredity
(20Jul2023 list later...)
Howell 20Jul2023 : This is a placeholder for one of my key interests, along with neural [architecture, function, process, etc]. neural architecture, computational models :
Still thinking... but this topic is a search for direct [, epi-]genetic links to neural [architecture, function, process, etc], and presupposes models for the latter. Consciousness is NOT an initial target, as progress is needed at more basic levels first.
conscious ART (cART), etc
(Stephen Grossberg & colleagues 1971-2021) - link is missing
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness] and many more.

Grossberg's models aren't just for consciousness - many were initially developed for subconscious modelling.

Unlike almost all other models of consciousness:
  • Grossberg etal's models tightly tie specific neural architectures to [function, process, learning, evolution]
  • Critically important issues are addressed, a few of which include :
    • lifelong learning without catastrophic forgetting
    • autonomous adaptation in a changing environment
    • explain and simulate many key [vision, auditory] perceptual [process, illusion]s explain synchronization issues and how the brain handles them (or doesn't with brain dysfunction)
    • tie the subconscious to the conscious, and explain when something becomes conscious
  • A surprisingly small number of neural architectures can simulate [extensive, diverse] [neuro, pyscho]logical data at BOTH the [sub, ]conscious levels, and for [perception, action] of [sight, auditory, touch, language, cognition, emotion, etc]. This is similar to what we see in physics.
  • In a sense, conscious [mechanism, model]s emerge from earlier work on subconscious processes, and are largely the same.
  • Grossberg's [bio, neuro, psycho]logical basis and neural [microcircuits, cellular assemblies, macrocircuits (like laminar cortex), brain regions, etc] and their [function, process, interaction]s greatly exceed any of the other concepts listed on this webPage, to the best of my limited knowledge.
  • [extensive, diverse] ex-bio applications have been successfully [developed, applied], based on Grossberg etal's computational models.
many [neural architecture, computational model]s :
CogEM - Cognitive-Emotional Motor
CLEARS - Consciousness, Learning, Expectation, Attention, Resonance, Synchrony
ART - Adaptive Resonance Theory, and many extensions, including :
  • LAMINART - laminar ART, explains cortical layer functions, etc
  • cART - conscious ART
  • see simple grepStr search results : 'ART|cART|pART|ARTMAP|ARTSTREAM|ARTPHONE|ARTSCAN|dARTSCAN|pARTSCAN|ARTSCENE|ARTSTREAM|ARTWORD|cARTWORD|LAMINART|PARSE|SMART|START|nSTART'
Grossberg and colleagues have produced many other models! However, these have not, as far as I can see, been gathered together into a large library for easy public use. This would very much limit their use, so I need to check that is the case.

Grossberg's concepts are NOT normally listed in [compilations, reviews] of consciousness, which is a [puzzle, failure] that I address separately.
attention content-copy (expectations)
(John G. Taylor 2006)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

attention content-copy
context related to Paul Werbos's "Brain as a controller"? - Approximate Dynamic Programming (ADP)
neural architecture, computational models :
no computational models?, descriptive only
For me, the significance of Taylor's concepts for consciousness is that they have an [entirely different, well-[established, proven] track record in very advanced [system identification, control theory, real applications]. More than the consciousness concepts in the sections below, Taylor's consciousness nicely [complement, contrast]s Grossbergs concepts.
Transformer Neural Networks (TrNNs)
(Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, Illia Polosukhin 2017)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

Large Language Models (LLMs) contain information across much of human knowledge, and are able to understand user's questions, then [extract, analyze, report] autonomously perhaps better than many experts (or all experts if a question cuts across enough areas of expertise?).
As with any other topic, LLMs can respond to questions about consciousness, but could the LLM models themselves be used as a starting point to quickly build systems using Grossberg's models? Or could existing LLMs be "backfitted" with Grossberg's consciousness?
neural architecture, computational models :
TrNNs - "Attention is all you need" (Dec2017)
LLMs - Large Language Models

Perhaps LLMs can also be seen more generally to be Cognitive user [interface, API, operating system]s, as they are being tied to systems with specialized [capability, software, etc].
Confabulation Theory
(Robert Hecht-Nielson 2002,2007)
This concept is included because of its 8-15 year anticipation of TRNN-LLMs that came out after ~2017. Some of the results, and the current excitement, are uncannily similar.

Confabulation model of cognition (NOT consciousness!) in mammals, and perhaps reptiles. Confabulation is "the other side of the coin" from Bayes Theory, and likely explains many sucesses of Bayesian statistics. Under simplifying assumptions, it reduces to Aristotolean logic.

By the way, Grossberg specifically points out in his book how brain processes are, generally speaking, NOT Bayesian, even for some of the occasions when they are statistical in nature.
neural architecture, computational models :
emphasized Thalamo-cortical loops
[Huge-at-the-time, general-knowlege, high-quality] text from [newspapers, magazines, books] was used to generate knowledge. Hecht-Nielson was very excited about the quality of the results at even a very early stage. In a 2007 book, this system generated much the same basic [results, excitement] as chatGPT in Nov2022, although the latter excited the general public, and not just a few rare researchers. The two are so similar in so many ways it's almost scary. Except, of course, chatGPY is vastly larger, and was available to the public freely, so they could [bypass, ignore, out-think] most experts. The point that non-experts can get a great head-start without experts is one of the very interesting outcomes of LLMs. Bad results can also come from not-so-good [expert, amateur]s.
Body and Emotion in the making of consciousness
(Antonio Damasio 1999)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

Grossberg 2021 p048c2h0.90 "... Perhaps the theory of the highly influential Portugese-Amereican neuroscientist and author, Antonio Damasio, comes closest to theoretically linking brain to mind in his beautifully written 1999 book with the title The Feeling of What Happens: Body and Emotion in the making of consciousness (Damasio 1999). ... Damasio used the cliical data that he elegantly summarized in his book to guide him to what is, in effect, a heuristic derivation of the Cognitive-Emotional-Motor, or CogEM, model that I had first published in 1991. ..."
neural architecture, computational models :
not much on neural [architecture, function, process]
Recurrent thalamo-cortical resonance
(Rudolfo Llinas 1998)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

One comment by (Grossberg 2021 p046c1h0.05) :
"... Of particular interest are data supporting ART's predicted link between synchronous oscillations and consciousness that neuroscientists such as ;
  • (Victor Lamme 2006)
  • (Rudolfo Llinas, Ribrary, Contreras, Pedroarena 1998)
  • (Wolf Singer 1998)
  • and their colleagues
have reported. ..."
neural architecture, computational models :
must input information...
Thalamic reticular networking
(Byoung-Kyong Min 2010)
[consciousness, learning, expectation, attention, resonance, synchrony, top-down bottom-up, cooperative-competitive, inhibition-excitation, social consciousness]

Byoung-Kyong Min 2010 "A Thalamic reticular networking model of consciousness"

"... The model suggests consciousness as a "mental state embodied through TRN-modulated synchronization of thalamocortical networks". In this model the thalamic reticular nucleus (TRN) is suggested as ideally suited for controlling the entire cerebral network, and responsible (via GABAergic networking) for synchronization of neural activity. ..." (Wiki2023)
neural architecture, computational models :
must input information...
[fast iconic, slow attentional] memory
(Crick, Koch 1990)
slow-attentional mechanism transiently binds together all those neurons whose activity relates to the relevant features of a single visual object.
this is a key historical paper that generated much modern interest in consciousness
neural architecture, computational models :
not much on neural [architecture, function, process]?...
global workspace
(Bernard Baars 1988)
Global workspace theory (GWT) is a simple cognitive architecture that has been developed to account qualitatively for a large set of matched pairs of conscious and unconscious processes. neural architecture, computational models :
not much on neural [architecture, function, process]?...
Dehaene–Changeux model
(Dehaene, Changeux 1986)
The Dehaene–Changeux model (DCM), also known as the global neuronal workspace or the global cognitive workspace model is a computer model of the neural correlates of consciousness programmed as a neural network. Stanislas Dehaene and Jean-Pierre Changeux introduced this model in 1986.[6] It is associated with Bernard Baars's Global workspace theory for consciousness. neural architecture, computational models :
neural network model, not much on neural [architecture, function, process]
neural global workspace model
(Stanislas Dehaene 2014)
"consciousness is global information broadcasting within the cortex [to achieve] massive sharing of pertinent information throughout the brain" neural architecture, computational models :
neural network model, not much on neural [architecture, function, process], but some good insights towards that
Neural correlates of consciousness
(authors)
Neural correlates of consciousness (NCC) formalism is used as a major step towards explaining consciousness. The NCC are defined to constitute the minimal set of neuronal events and mechanisms sufficient for a specific conscious percept, and consequently sufficient for consciousness. neural architecture, computational models :
not much on neural [architecture, function, process]?...
Integrated Information Theory
(Giulio Tononi 2004, 2012, 2015)
In IIT, a system's consciousness (what it is like subjectively) is conjectured to be identical to its causal properties (what it is like objectively). Therefore it should be possible to account for the conscious experience of a physical system by unfolding its complete causal powers (see Central identity).[4] ... Specifically, IIT moves from phenomenology to mechanism by attempting to identify the essential properties of conscious experience (dubbed "axioms") and, from there, the essential properties of conscious physical systems (dubbed "postulates").. neural architecture, computational models :
not much on neural [architecture, function, process]?...
Multiple drafts
(Daniel Dennett 1991)
The Multiple Drafts model makes [the procedure of] "writing it down" in memory criterial for consciousness: that is what it is for the "given" to be "taken" ... There is no reality of conscious experience independent of the effects of various vehicles of content on subsequent action (and hence, of course, on memory). neural architecture, computational models :
not much on neural [architecture, function, process]?...
Functionalism
(authors)
Functionalism is a view in the theory of the mind. It states that mental states (beliefs, desires, being in pain, etc.) are constituted solely by their functional role – that is, they have causal relations to other mental states, numerous sensory inputs, and behavioral outputs. neural architecture, computational models :
not much on neural [architecture, function, process]?...
Electromagnetic theories
(authors)
Electromagnetic theories of consciousness propose that consciousness can be understood as an electromagnetic phenomenon that occurs when a brain produces an electromagnetic field with specific characteristics.[7][8] Some electromagnetic theories are also quantum mind theories of consciousness.[9] neural architecture, computational models :
not much on neural [architecture, function, process]?...
quantum processes
(Roger Penrose, >name? Hameroff ?date)
Hamerof added speculations that quantum processes related to consciousness may occur in neuron microtubules? (eg Pribraum etal "Quantum Consciousness Proceedings" - but I've lost the proceedings) neural architecture, computational models :
not much on neural [architecture, function, process]?...

backgound colours in the table signify :
pinkspeculation (Howell) - just a reminder that people may always have thought of this
whitepriority focus of this project
lime greenconcepts that, to me, nicely [contrast, compare] to the [Grossberg, TrNN] concepts that are the priority focus of this website
light greyOft-cited theories of consciousness
(Grossberg 2021) "do not describe the underlying organizational principles, neural mechanisms, or brain representations that embody subjective conscious aspects of experience"
Howell - I agree with Grossberg. These lack the [heredity, architecture, function, process, etc] that I am looking for. But they may be sufficient for others, and remember that any of them may have "gems in the rough" that are useful inspiration.
medium greyOR[philosophical, no [neural architecture, computational model]]
dark greyalmost magic science