Falsification of Scientific Theories of Consciousness

[AshvinP]

So every scientific paper that has ever been written on consciousness is not science. And every researcher doing any research in the area of consciousness is deluded.

What's more likely every scientist doing research on consciousness is deluded or you are?

Every paper and researcher trying to explain the existence of consciousness from a 'material' substrate is not scientific. Rather it is a statement of religious cult-like dogma. I hope my emphasis above was adequate to prevent continued misrepresentations of what I am claiming.

You are not understanding any of these theories. Where do any of them make any sort of metaphysical claim? Perhaps one does but I'm not aware of it. You seem to be making a straw man argument. Nobody that I know of is claiming a 'material' substrate as the ultimate reality in any of these theories.

I didn't say they were claiming a 'material substrate' is fundamental reality. I said they were trying to explain consciousness from material substrate. It doesn't matter what we call the latter - just call it non-consciousness. Trying to explain consciousness from non-consciousness is not scientific. It is no more scientific than someone trying to explain Nature using a mish-mash of increasingly convoluted theological doctrines.

[Martin_]

Did anyone here actually edit:readreally understand the article?

[Eugene I]

Did anyone here actually edit:readreally understand the article?

Well, I read it but can't claim I followed all the math. But I think it's basically a math proof of Chalmers proposal that the states of conscious experience (qualia) are irreducible and indescribable by physical theories. This paper particularly states that, if an independence of the observable and reported data about conscious system is postulated, it if by default falsified (in other words, postulation of such independence leads to contradiction). Or, if a strict dependence is postulated, any non-contradictory theory of consciousness (TC) is not falsifiable.

Here is a simple illustration of the second scenario (strict dependence). Let's say we have one variant of theory TC1 that makes a prediction of the experience of "blackness" based on neural measurements in the neurons of the visual area of the brain that react to the black color on the picture. And the same theory also contains an inference of the same quale of "blackness" based on the verbal report "I see a black spot" by the person under test. But then we can formulate another theory TC2 that would predict that the patient would actually experience a quale of "whiteness" from seeing a "black" spot, with the same results of neural measurements and the same verbal report. Now, how can we prove TC1 true and TC2 wrong? The paper proves mathematically that such falsification of TC2 is impossible, but it is actually intuitively obvious, and that is what Chalmers was saying two decades ago with his "inverted color" arguments. In simple words, if I claim that dolphins actually see "white" when people see "black", there is no way to falsify my claim by any physical measurements.

[Simon Adams]

Yes, being able to ‘demonstrated’ the theory in some unambiguous way is maybe better than just looking for falsification.

I would add that it should add an explanatory dimension. There should be something it explains better than other theories to make it a good scientific theory.

All that said, it does seem like consciousness is almost by definition beyond the scientific process. It’s always going to be a ‘black box’ in terms of scientific methods of investigation...

I don't think it makes it beyond the scientific process at all. We have self reports and measurable behaviors that can be matched with brain activity. Also, some interesting experiments can be done that only require observable behavior.

Here is an article about an experiment showing monkeys have meta-cognition.

https://www.scientificamerican.com/arti ... nking-too/

Here's another similar one with rhesus monkeys:

https://yaledailynews.com/blog/2021/04/ ... s-monkeys/

All of this is easily observed behavior.

Of course, science can't get inside the head of the monkey and see what the monkey is seeing like they could in Brainstorm. Yet.

https://en.wikipedia.org/wiki/Brainstorm_(1983_film)

All the examples you mention are still ‘black box’. Something goes in and something comes out, but you have no access to what happens inside.

This is very different from, say, a computer where every aspect of what the computer does could be reverse engineered just by careful physical analysis. Even the software. There is no reason to believe that a machine/instrument will ever be able to access human memories for example, only physicalist and sci fi assumptions...

[Simon Adams]

There is no reason to assume that. It is only when we self-impose a materialist or dualist framework does there arise some fundamental limit to what can be empirically and rigorously studied. Goethe was a great example of someone who began to investigate the dynamics of consciousness in relation to the phenomenal world and that was centuries ago. Just think where we would be if his method of scientific investigation had become the gold standard over that of Newton, Descartes, etc. and Kant. Steiner presents elaborations in his book on Goethean Science. Here is an excerpt which draws conclusions and relates them to Goethe's color theory, but it is best to visit the link and read through that entire chapter at least:

Yes fair point, but you’re crossing an interesting boundary here, you could call it the boundary between ‘hard science’ and ‘soft science’. The hard sciences in theory require a type of objectivity that does seem difficult to apply to questions on the nature of consciousness. The questions around the behaviour of consciousness are slightly different, but even then I would still argue that it’s still essentially a ‘black box’ scenario in standard scientific terms (unless you make physicalist assumptions about the nature of consciousness).

[Martin_]

Did anyone here actually edit:readreally understand the article?

Well, I read it but can't claim I followed all the math. But I think it's basically a math proof of Chalmers proposal that the states of conscious experience (qualia) are irreducible and indescribable by physical theories. This paper particularly states that, if an independence of the observable and reported data about conscious system is postulated, it if by default falsified (in other words, postulation of such independence leads to contradiction). Or, if a strict dependence is postulated, any non-contradictory theory of consciousness (TC) is not falsifiable.

Here is a simple illustration of the second scenario (strict dependence). Let's say we have one variant of theory TC1 that makes a prediction of the experience of "blackness" based on neural measurements in the neurons of the visual area of the brain that react to the black color on the picture. And the same theory also contains an inference of the same quale of "blackness" based on the verbal report "I see a black spot" by the person under test. But then we can formulate another theory TC2 that would predict that the patient would actually experience a quale of "whiteness" from seeing a "black" spot, with the same results of neural measurements and the same verbal report. Now, how can we prove TC1 true and TC2 wrong? The paper proves mathematically that such falsification of TC2 is impossible, but it is actually intuitively obvious, and that is what Chalmers was saying two decades ago with his "inverted color" arguments. In simple words, if I claim that dolphins actually see "white" when people see "black", there is no way to falsify my claim by any physical measurements.

The part I have a problem with is their definition of independency.
My paraphrasing of their definition : "If 2 different observable datasets can lead to the same report, they are independent. " I'm quite sure what I just paraphrased is a misunderstanding, because if it's not, that's a very strong claim and their results don't make sense.

Also, the article doesn't say that we can't falsify theories of consciusness in general. All it says that all the currently prominent theories suffer from either

• Being falsified a-priori - due to the above independency being assumed in the theories

• or being unfalsifiable

They do point out that there is theoretical room for a middle ground - theories that are neither assuming super-depedency nor independency - that would not suffer from the problems discussed in the article.


TLDR: I can't tell if the article is deeply profound or entirely pointless.

[Jim Cross]

Did anyone here actually edit:readreally understand the article?

The article is sort of irrelevant.

The Idea That a Scientific Theory Can Be ‘Falsified’ Is a Myth

Fortunately, falsification—or any other philosophy of science—is not necessary for the actual practice of science. The physicist Paul Dirac was right when he said, "Philosophy will never lead to important discoveries. It is just a way of talking about discoveries which have already been made.” Actual scientific history reveals that scientists break all the rules all the time, including falsification. As philosopher of science Thomas Kuhn noted, Newton's laws were retained despite the fact that they were contradicted for decades by the motions of the perihelion of Mercury and the perigee of the moon. It is the single-minded focus on finding what works that gives science its strength, not any philosophy. Albert Einstein said that scientists are not, and should not be, driven by any single perspective but should be willing to go wherever experiment dictates and adopt whatever works.

https://www.scientificamerican.com/arti ... iscoveries.

[Jim Cross]

All the examples you mention are still ‘black box’. Something goes in and something comes out, but you have no access to what happens inside.

This is very different from, say, a computer where every aspect of what the computer does could be reverse engineered just by careful physical analysis. Even the software. There is no reason to believe that a machine/instrument will ever be able to access human memories for example, only physicalist and sci fi assumptions...

Actually there have been white room reengineering of computers and software so not knowing what goes on inside is not fatal.

Aside from that, how do you know what you think or perceive is going on inside really is what is going on inside or really gives you any insight into what is going on? You seem to think that because you experience it then it explains itself but perceptions we know are frequently wrong.

At any rate, my reaction to this article is "so what". Even if the authors are correct that theories of consciousness can't be falsified, that doesn't mean no science can be done.

[Eugene I]

Also, the article doesn't say that we can't falsify theories of consciusness in general. All it says that all the currently prominent theories suffer from either

• Being falsified a-priori - due to the above independency being assumed in the theories

• or being unfalsifiable

They do point out that there is theoretical room for a middle ground - theories that are neither assuming super-depedency nor independency - that would not suffer from the problems discussed in the article.

In their "Objections" and "Conclusions" sections they do discuss several options, including the "middle-ground" dependencies, dropping falsifiability as a truth criterium for scientific models etc.

One of the points they make is that the theories of consciousness require a different treatment and methodological approach compared to other areas of science, because the phenomena of conscious experience that such theories try to include, account for and explain have fundamentally different nature from the experimental point of view - they are only 1-st-person observable and non-objective.

[Simon Adams]

Aside from that, how do you know what you think or perceive is going on inside really is what is going on inside or really gives you any insight into what is going on? You seem to think that because you experience it then it explains itself but perceptions we know are frequently wrong.

At any rate, my reaction to this article is "so what". Even if the authors are correct that theories of consciousness can't be falsified, that doesn't mean no science can be done.

I probably wasn’t clear that I’m purely talking about the way hard science is done now. I’m in no way at all claiming that experience isn’t valid. There is a good reason why science looks for empirical, “objective”, repeatable evidence, but the problem comes when anyone then claims that the picture you get from this can ever be a complete picture. The hard science approach is a valid but extremely limited epistemology, and consciousness just happens to be one of the areas which is fundamentally outside of it’s scope.

Although I agree that falsification shouldn’t be the only criteria for a good scientific theory, there is a wider conceptual problem of the way science currently works. It’s a bit like quantum physics, there is no third person perspective from which you can theorise, it just doesn’t exist. That means the whole ‘outside looking in’ approach will never get to the fundamentals.