Matter and mind:
Over, and under…
That it is is a necessary assumption for those who accept that minds exist, but equate it with either the central nervous system or its encapsulating body.
Much of the issue around physicalism vs. the question of minds’ existences may have less to do with the nature of reality than with what we mean by certain words. If anything and everything is physical, then “physical” is a unifying rather than polarizing or distinguishing concept; there is no contrast class to it (that is: nothing non-physical exists).
The project for physicalist cognitive scientists and philosophers of mind–those who, in short, believe that everything (including the mind) is physical–is, perhaps, simply to extend the language of physics to also accommodate “mind”. (And physicalism could well have to change, itself, in order to be able to do so!)
It may be convenient for a field like cognitive neuroscience especially to have a common framework for its studies of neural and cognitive (mental) systems, which have traditionally–a la Descartes and his successors–been regarded as more ontologically separate than they are, these days. Still, people’s mileages vary with respect to whether physicalism is the right unifying ideology for reality even aside from this possibility.
[The discussion from which this post arose took place on INTJforum, at: https://intjforum.com/topic/175287-if-physicalism-how-is-it-that-we-can-trust-our-cognitive-abilities/ ]
“…How are people to become aware of their more destructive acts to begin with [e.g., of environmental disregard]—the fact that they, themselves, might be committing them regularly?
“The [phenomenological-critical] method can be performed mentally or verbally. Where individual privacy is of concern, the former would be preferred, and where other people can help one become aware of something less subjectively-accessible the latter would be. The step-by-step logical form I have chosen may dovetail neatly with efforts in schools to teach algorithmic thinking (e.g., through the instruction of computer programming), and it would likely prove most prudent to encourage such a habit of critical awareness from a younger age. However, the general method may just as well be practiced by any capable adult in any other realm or mode of existence.
“Naturally, people who think less linearly may find the method disagreeable, though it has been deliberately kept simple for early and general use. The method’s efficacy will also covary with individual differences in temperament, personality factors, attention-distributive tendencies, motivation, and health conditions, and would require appropriate adaptation for non-English speakers” (Sood, 2016).
T: I feel like psychology is too weak on its own…
SS: Ironically, actually, it is psychology’s over-reliance on the other, more established sciences (viz., biology) that has done it a disservice in the past century. Its dominating attempts to imitate precisely traditional empiricism and its methods have left it theoretically underdeveloped and disunited, so that there are a number of sub-disciplines making up ‘psychology’, without any coherent framework making their connections–and, thus, the status of psychology as a grounded and self-consistent science–clear, to either its insiders or those looking from without.
T: I don’t see over-reliance being an issue. Fields separate because of expertise and imaginary divisors of what is connected to what (it’s all connected). The brain is small, so it’s easier to understand by partitioning.
SS: It’s fine that studies of the brain be partitioned, and additionally for disciplines to look toward others in figuring out how to best proceed themselves. But over-reliance, for psychology’s case, has made it such that it has no clear idea of what mind’s ontological status is, which is highly relevant to the extent that it should want to precisely characterize and explore the phenomenon (or, perhaps more accurately: its sub-phenomena) on its own terms.
Mind is not–in any obvious, or hitherto successfully argued-for way–‘physical’ in the same sense that every other physical object that has been studied scientifically has been shown to be. On the other hand, if its operations and contents actually are directly (i.e., not relying solely on measures of brain activity) detectable via physical methods, then it must be admitted that we currently lack the technology with which to achieve such. (Whereas for formerly-discovered phenomena of at least a somewhat similar nature, w/r/t their raising for us measurability problems–e.g., the invisible parts of the electromagnetic spectrum–further technological advancement was needed in order to actually establish their physical reality.)
Gravity, of course, would be another example of a postulated force with causal potency in relation to the known-about physical realm, but that is still undetectable in a way that has prevented us from knowing exactly what kind of physical substance it is or is made up of. (Assuming that it is, indeed, ‘physical’ in the traditional sense of possessing measurable energy or discrete constituents of matter.)
T: I’m hoping computational cognition becomes a field in the future, once we have cracked the puzzle, which will be accomplished in the next few years.
SS: Cognitive science is already computational.
T: That’s true, but it’s not authentic, nor is there even a basic mind that’s capable of meaningful use by means of artificial self-awareness.
SS: I’m not really sure what you mean by cognitive science’s treatment of mind not being “authentic”, though it certainly is incomplete at present. Perhaps that we haven’t (yet) established sentience/self-awareness in artificial/computational models of mind? If this is what you are getting at, the intellectual or empirical proof of any entity being sentient/self-aware is a highly complex problem, particularly the more we look at species (or, perhaps and for some, things like computers) that are not human.
T: Deep learning is about as sophisticated as it gets, and it’s not really even that impressive. The brain is highly parallel and we’ve only just begun to get the hardware that is matching it. Graphics cards follow this structure.
SS: Yes, and neural nets. Though those have been around for quite a while, and are extremely simplistic when compared to the actual human brain.
Again, though, the mind’s computational aspects have been overemphasized (since at least the cognitive revolution of the ’70s–though the story may well have started earlier, solidifying most definitively during psychology’s behaviorist phase) at the expense of a rigorous and fully-general understanding of what mind actually is. For the most part, cognitive psychologists don’t study and have not studied philosophy, which has also meant that they are largely unaware of 1) just how relevant the mind-body problem is to their work, and 2) how multi-faceted and complex mind truly is.
T: We are going to have to agree to disagree. I see nothing special about the mind that differs from any model of computation. Are you not in favor of the Turing Machine model that’s theorized to compute anything that a human brain can?
SS: Well, independent of my own views on mind, that a Turing machine can compute “anything” a human brain can is so far promissory note, at most. (How do we come up with an exhaustive list of all of the computations the human mind is capable of performing? At what point or how could we know whether such would be complete?) As is the general assumption that mental phenomena are, or will be found to be governed by the same natural laws that physical ones are.
T: As far as what is computable by the human brain, it is mostly assumed at this time.
T: Philosophy or philosophical approaches is/are useful in any discipline. Consciousness is a problem for metaphysicians. The “study of behavior” is not deep enough, but it’s a good starting point.
SS: Consciousness has become a problem for quite a few more groups than metaphysicians, hence its currently being a highly interdisciplinary field (consisting of researchers from A.I.; phenomenology; cognitive neuroscience; philosophy; physics; computer science…). As metaphysics has for so long proven incapable of solving the hard problem of consciousness (or even characterizing it in a way that would make it more amenable for other disciplines), it makes more sense now than ever that the problem should be treated in such a multi-plural manner.
T: All of those prior to the precomputational era are exempt. Metaphyisicans up from the 1900s are the only ones that count. To date, none of them are particularly impressive. As philosophy has been phased out somewhat, it’s slowly coming back, for good reason.
SS: No cracking the puzzle of mind without understanding just what it is and how it works, which may hinge on the extent to which progress is made on the question of how subjective experience (which has to be central to any complete science of mind) and the body are related: in addition to how the former is even possible, and what its nature (history; universal scope; dynamics…) is. We are not close to an answer to this ‘hard problem’, at present, and so the hope that it will be answered successfully “in the next few years” seems preposterous to the extreme.
T: I refuse to believe it’s as difficult as people make it out to be. Humans are emotionally driven, it’s not that hard. Emotionally Oriented Programming as a humorous analogy. After that, it’s simply an amalgam of networks that are extremely, extremely complex.
SS: You can refuse it all you like, but the fact is that it really has been so difficult for the vast majority of people who have studied and are studying mind.
Emotion is a part of the puzzle (and one that has been pretty neglected by psychology, and especially by cognitive science), but it is not by any means the only or necessarily biggest part of it.
T: Emotions are an evolutionary advantage, a synthetic approach would be a network that precedes and influences the logical network and determines how well it functions based on the load (whether too high or low) in its various sub-networks put on this artificial emotion network.
SS: I haven’t studied the question of interactive emotional-logical computational systems much myself, but here is something related in case you are unaware of it: https://en.wikipedia.org/wiki/Affective_computing
T: Subjective and objective are really not any different. It’s an elaborate illusion.
I can’t believe no ones figured that out yet.
SS: Believe it.
The subjective/objective question is still more in the domain of philosophy, and less within consciousness studies and cognitive science (though it is significantly more present in the former of these two than the latter).
Re. the subject-object split being an “illusion”, Eastern traditions (e.g., Buddhism) have held this view for centuries, and Western thought has recently begun to catch on to it, as evidence continues to mount in disfavor of Descartes’ original material body-immaterial soul(/mind) separation.
It is still an open question, however, whether or not subjective and objective may still be meaningfully distinguished in certain ways (or for certain reasons).
T: Dualism is just [stupid]. I won’t dismiss it though when dealing with whatever a “soul” can be, but for everything else, it’s unnecessary. Not including it won’t even compromise computation.
You can give the credit to any previous schools of thought, but the fact is, they’re not necessary for solving this puzzle. Plenty of people have said wise things, yet those wise things on their own are meaningless until given proper application.
T: Subjectivity is not much more than an internalized world shaped by language processing of that mind/body and feeling of mind/body. Due to diversity of parameters, each of us differs.
SS: If subjectivity really were “not any different” from objectivity, as you asserted above, then it would be much more than a mere “internalized world”. In consciousness studies, the problem of subjectivity is the ‘hard problem of consciousness’:
“The really hard problem of consciousness is the problem of experience. When we think and perceive, there is a whir of information-processing, but there is also a subjective aspect. As Nagel (1974) has put it, there is something it is like to be a conscious organism. This subjective aspect is experience. [emphasis added]“ (Chalmers, 1995: Facing Up to the Problem of Consciousness)
T: Of course, all of this is meaningless without a proper model for a basic mind, which is achieved after stripping the senses. Adding them back on simply adds potential for freedom and further learning.
SS: How could the senses possibly be 100%-stripped in a way that would leave what would presumably be left of the mind intact? It seems dubious that the (self-aware) mind as we are familiar with it could exist at all in this purely ‘senseless’ kind of manner.
T: A mind would be intact, but if the mind in this example would be mirrored from birth, it’s not a worthwhile exercise. A baby with no senses would lead to much at all. Not much would be learned.
It’s simply a process akin to reverse engineering, by trial-and-error. It’s kind of cheating, since adults have already accumulated lots of data (through experience/time).
The major issue is mostly “what is reality?”, not “what is subjectiveness?”. “What is mind?” is what naturally follows thereafter. Tying the t[w]o together, “what is complexity?”. “What is emerging complexity?” Although they are related, it’s clear to me at least, that subjectivity isn’t the right starting point, but rather a red herring.
SS: Personally, I think cognitive neuroscience (which has largely come to subsume psychology’s studies of mind) and cognitive-behaviorism (psychology’s current ‘paradigm’, if one may roughly speak of one) will have a lot to learn in particular from quantum physics; philosophy; and consciousness studies (e.g., phenomenology), if psychology and cognitive science are to have any hope of meaningfully contributing toward solving the mind-body problem. Computational theories have helped and conceivably will continue to do so, but their limits–which one need only look at artificial intelligence’s history from the past century to appreciate–will have to be acknowledged, before mind can be defined as accurately, and characterized as faithfully as it will ultimately need to be for our genuine understanding.
T: Quantum computers offer specialized processing for different areas of computation. Whether that relates to the mind or not remains to be determined. I suspect that the processes it would help are nonetheless not important for our starting goal, but essential for reaching higher peaks of intelligence.
SS: Intelligence is a problem I won’t go into, here; but, it will be interesting to see what potential new possibilities quantum computing will open up for cognitive science, as the former becomes more developed and as the two fields are synthesized.
T: Intelligence is one of the most key factors in solving this issue, but what the word encapsulates is clearly its own problem.
[To be continued…?]
[Original post: http://intjforum.com/showpost.php?p=5159026&postcount=2%5D
My thoughts (I’m ‘S’)—
Video: King’s Wise Men/NAO demonstration
S: Impressive. I’d be interested to see what future directions the involved researchers take things.
Prof. Bringsjord: “By passing many tests of this kind—however narrow—robots will build up and collect a repertoire of abilities that start to become useful when put together.”
S: Didn’t touch on how said abilities could be “added up” into a potential singular (human-like) robot. Though Bringsjord doesn’t explicitly seem to be committing to such, the video’s narrator himself claims to see it as “much like a child learning individual lessons about its actual existence”, and then “putting what it learns all together”; which is the same sort of reductionist optimism that drove and characterized artificial intelligence’s first few decades of work (before the field realized its understandings of mind and humanness were sorely needing). So the narrator lverbally) endorses the view that adding up robotic abilities is possible within a single unit, which I have yet to see proof or sufficient reason to be confident of.
(Being light, for a moment: Bringsjord could well have a capitalistic, division-of-labor sort of robotic-societal scenario ready-at-mind in espousing statements like this…)
Narrator: “The robot talked about in this video is not the first robot to seem to display a sense of self.”
S: ‘Self’ is a much trickier and more abstract notion to handle, especially in this context. No one in the video defines it, or tries to say whether or how it’s related to sentience or consciousness (defined in the two ways the narrator points to), and few philosophers and psychologists have done a good job with it as of yet, either. See Stan Klein’s work for the best modern treatment of self that I’ve yet come across.
Video: Guy with the synthetic brain
S: Huh…alright–neat, provided that’s actually real. Sort of creepy (uncanny valley, anyone?), but at least he can talk Descartes…not that I know why anyone would usefully care to do so, mind, at this specific point of time in cognitive science’s trajectory.
Dr. Hart: “The idea requires that there is something beyond the physical mechanisms of thought that experiences the sunrise, which robots would lack.”
S: Well, yeah: the “physical mechanisms of thought” don’t equal the whole, sum-total experiencer. Also, I’m not sure what he means by something being “beyond” the physical mechanisms of thought…sort of hits my ears as naive dualism, though that might only be me tripping on semantics.
Prof. Hart (?): “The ability of any entity to have subjective perceptual experiences…is distinct from other aspects of the mind, such as consciousness, creativity, intelligence, or self-awareness.”
S: Not much a fan of treating creativity and intelligence as “aspects of the mind”…same goes for consciousness, for hopefully more-obvious reasons. Maurice Merleau-Ponty is the one to look into with respect to “subjective perceptual experiences”, specifically his Phenomenology of Perception.
Narrator: “No artificial object has sentience.”
Well, naturally it’s hard to say w/r/t their status of having/not having “subjective perceptual experience”, but feelings are currently being worked on in the subfield of affective computing. (There’s still much work re. emotion to be done in psychology and the harder sciences before said subfield can *really* be considered in the context of robotic sentience, though.)
Narrator: “Sentience is the only aspect of consciousness that cannot be explained…many [scientists] go as far as to say it will never be explained by science.”
S: They may think so, and perhaps for good philosophical reasons; but that won’t stop, and indeed isn’t stopping some researchers from trying.
Narrator: “Before this [NAO robot speaking in the King’s Wise Men], nobody knew if robbots could ever be aware of themselves; and this experiment proves that they can be.”
Aware of themselves again leads to the problem briefly alluded to above, regarding the philosophical and scientific impoverishment of the notion of ‘self’. I know what the narrator is attempting to get at, but I still believe this point deserves pushing.
Narrator: “The question should be, ‘Are we nearing robotic phenomenological consciousness?’”
S: Yep! And indeed, you have people like Hubert Dreyfus arguing for “Heideggerian AI” as a remedy for AI’s current inability to exhibit “everyday coping”, i.e. operating with general intelligence and situational adaptability in the world (or “being-in-the-world”, a la Heidegger).
In cognitive science terms, this basically boils down to the main idea underlying embodied cognition, a big move away from the old Cartesian or “representational” view of mind.
Narrator: “When you take away the social constructs, categories, and classes that we all define ourselves and each other by, and just purely looking at what we are as humans and how incredibly complex we are as beings, and how remarkably well we function in a way, actually really amazing, and kind of beautiful, too…so smile, because being human means that you’re an incredible piece of work.”
S: I wish the narrator would have foregone the cheesy-but-necessary-for-his-documenting-purposes part about humans’ beauty and complexity, in favor of going a bit further into the obviously difficult and tricky territory of “social constructs, categories, and classes” that we “all define ourselves and each other by”: and how near or far robots can be said to be from having truly human-like socio-cultural sensibilities and competencies.
[Addressed to the INTJforum ‘MBTI and Personality Theories’ sub-forum]:
Easily the most prevalent complaint about the so-called Jungian or ‘cognitive’ functions is that they lack too much empirical support to warrant the frequently charitable assumptions made in discussions surrounding them. Many of the frustrations of people who remain skeptical of the functions thesis can be captured by the fact that, at least at present, the functions do not (easily) lend themselves or stand up to scientific scrutiny or testing (or funding).
The idea behind this post (and possibly future ones like it) represents a desire on my part to hopefully mitigate some of the above-named concerns, by highlighting just a few of the real connections between our accepted understandings in cognitive psychology, and how we (viz. casual and expert commentators, and typologists like Berens and Nardi) generally conceptualize the functions. As this can be approached in a number of ways, for the purpose of not rambling myself to death at one time, I will narrow the scope of this post to Ni, Si, (as much about them as we think we might know, in this preliminary stage) and how modern-day cognitive scientists understand memory.
The Jungian functions
It is commonly considered that Si has a memory component to it. Indeed, Dario Nardi’s own observations on the subject seem to validate this when he notes that:
“Si types may get ‘in the zone’ when reviewing past events…ISTJ and ISFJ easily enter an expert flow state while recalling, particularly if they close their eyes and take the time to immerse themselves in the memory, reliving it in rich detail” (Neuroscience of Personality, p. 94).
“[Si types] have a propensity for rote memorization, repetition, and in-depth reviews of daily events…Si types are highly capable at recalling information that has little or no context, such as lists of random words…” (NoP, p. 94)
But what most seem to leave out in their examination of Si is that, like Ni, it is predictive and allows users of it to “consider the future” (NoP, p. 95). Nardi notes that both Si and Ni types show moderate-high activity in a brain region that helps us do this, especially insofar as it is helping us plan our own actions ahead of time.
Conversely from Si, Ni is most commonly thought of as the ‘predictive’ function, or the one that most often and accurately allows us to predict what will occur in the future. Though Ni’s power to do this is clearly exaggerated in the mainstream typology culture, I will not attempt to dispel this misconception at the present time. Suffice it instead to point out what Nardi observes about Ni types, who “may easily show a zen state [overall brain pattern] when tasked to envision the future” (NoP, p. 102). And whether you want to call them Ni or NJ types, it is common for these types to self-report in confirmation of this observation made by Nardi in his MBTI-EEG studies.
But Nardi doesn’t mention how Ni looks back in time, or even how Si looks ahead. Probably our forum’s leading proponent of the functions model, whom we all know (to varying degrees of reverence) as […], stated it thusly:
“Ni can deduce the past from the present, and predict the future from the present, in terms of dynamics. Si instead sees things as mostly constant, and tends to be surprised by change. Both Si and Ni are predictive, but Ni types tend to impress others in terms of predicting things that were not ‘obvious’. (I.e., it’s obvious that if this is a rock, then it was a rock, and it will be a rock in the future; it’s not obvious that this is/was a meteorite that fell from the sky, and contains metals/isotopes that aren’t commonly found on Earth.)”
And while his example regarding the rock, there, might receive mixed responses from the subforum community, the important point to focus on is that both Ni and Si are predictive and backward-looking functions, though they differ greatly in how they go about fulfilling those purposes.
Now, for those of you who have had enough Nardian ‘pseudoscience’ for one post, you can rest assured that from this point, we will be moving on to ‘actual’ cognitive psychology (though we will still be establishing its relations with Ni and Si in their primitive, abstract forms).
Mainstream cognitive psychology
In going forward, readers might find it helpful to keep this handy reference chart in view–but they should note that for the purposes of this post, we will be restricting our scope specifically to declarative (or “explicit”) memory:
(For more on long-term memory’s sub-systems, see here: http://en.wikipedia.org/wiki/Long-term_memory#Divisions_of_long_term_memory)
The important things to keep in mind (or commit to memory, as it were!) are that:
“As a result of implicit memory‘s functioning, we are able to learn without being aware that we are doing so (e.g., Graf & Schacter 1985), and we can retrieve or use that information without being aware that we have stored it in memory” (Revlin, C:T&P p. 153).
I believe that understanding the two types of declarative/explicit memory presented is key to understanding the memory components of Ni and Si. (For those interested in why I don’t consider implicit memory relevant to the present discussion, see the paragraph below and feel free to comment on its contents.)
[[[I don’t believe implicit memory is particularly important to understand, here, since it functions “semiautonomously”, meaning that its mental functions operate automatically and “in the background”. Treatments of the Jungian functions as unconscious processes are more apt to describe how each type’s tertiary and inferior functions work (in generally inopportune ways), whereas the dominant and auxiliary functions are those that we are conscious of (though it is true that we tend to take the dominant’s operation for granted, as it’s essentially the ‘water we swim in’ and we’re too used to it to take much ‘conscious’ notice). Further, the EEG technology which Nardi utilized only measured neocortical brain activity, meaning it could only be used to analyze the topmost (and newest) layer. As this layer corresponds most closely with conscious and observable thought processes, implicit memory’s mechanics are a little trickier to uncover without more sophisticated brain-imaging technology.]]]
Based on the quotes whose respective authors I’ve cited, the connections between Ni/Si and explicit memory should become clearer. Si thrives on reviewing past events in rich detail, which correlates strongly with our understanding of episodic memory. Both Ni and Si engage in prospective memory, and at least Si engages in retrospective memory (“reliving [past experiences] in rich detail”, as Nardi observed). Finally, Si certainly utilizes semantic memory, which serves as a “library where discrete facts are stored”.
The above seems to leave Ni a bit in the dark, however. Specifically, two questions are left unanswered: 1) Assuming it can equally well engage in retrospective memory, how does it do so in a manner distinct from Si?; and 2) Given that Ni is far more apt to store relations and abstract principles than “discrete facts”, what is Ni’s relation to semantic memory? Might it be that there is some other memory bank which has been either unexplored in cognitive psychology, or left out of the present discussion? For now, I will leave these questions to readers to examine, though I will do so myself in a (hopefully, though not necessarily) timely manner.
My point here hasn’t been to ‘prove’ or ‘disprove’ the functions. Rather, I went forward with the assumption that the functions are worthy of further refinement and scrutiny, and in this early stage of their treatment the best we can do is ensure that they be defined in terms as technically precise as possible. If this can be done, then perhaps the functions can someday be studied in a more rigorous and scientifically-respectable manner–and there are, for purposes of better understanding ourselves and others, very compelling reasons for the rich variety in cognitive modes across humans to be elucidated and properly accounted for.