Reductionism

[Brainee]

What is reductionism?

[joigus]

In a nutshell, it is the contention that you can understand a system by analysing its parts and the relationships between them. Assuming that it makes sense to consider it as made up of distinct parts, that is.

https://en.wikipedia.org/wiki/Reductionism

[iNow]

Taking larger concepts down to their barest essence, often at the cost of oversimplifying and losing utility. 

[Phi for All]

>

[Boltzmannbrain]

An implication of reductionism is that you can always make correct predictions about a system by knowing its parts.

An example of this is for science.  Most of the time science can make predictions by knowing the parts well enough.

But where hard reductionism fails is for the consciousness.  The parts that come together are thought not to be conscious.  So we have the opposite of reductionism, emergentism, or, another way to put it, irreducible.      

[Genady]

23 minutes ago, Boltzmannbrain said:

But where hard reductionism fails is for the consciousness.

How do we know this?

 

23 minutes ago, Boltzmannbrain said:

The parts that come together are thought not to be conscious.

This does not contradict reductionism. Why wouldn't it be possible to predict consciousness based on a knowledge how the unconscious parts are arranged?

[Sensei]

To exist or not to exist.

It is a binary question.

 

[Boltzmannbrain]

29 minutes ago, Genady said:

How do we know this?

If we are assuming that things like atoms are not conscious (by consciousness I mean awareness) then we pretty much know.  Although there is a theory IIT trying to reduce consciousness that appears to be just panpsychism, but it is still pretty radical.

Quote

This does not contradict reductionism. Why wouldn't it be possible to predict consciousness based on a knowledge how the unconscious parts are arranged?

Maybe we are talking about 2 different definitions of consciousness. 

Edited February 26, 2023 by Boltzmannbrain

[Genady]

2 minutes ago, Boltzmannbrain said:

Maybe we are talking about 2 different definitions of consciousness. 

Not necessarily. It looks quite straightforward to me: reductionism allows for a whole to have a property which the parts lack. For example, a composition of non-white colors makes white color. Compositions of sinusoidal functions make functions which are not sinusoidal. Molecules have properties which atoms lack. Similarly, some composition of some unconscious parts makes something conscious, why not?

[Boltzmannbrain]

10 minutes ago, Genady said:

Not necessarily. It looks quite straightforward to me: reductionism allows for a whole to have a property which the parts lack. For example, a composition of non-white colors makes white color. Compositions of sinusoidal functions make functions which are not sinusoidal. Molecules have properties which atoms lack.  Similarly, some composition of some unconscious parts makes something conscious, why not?

An implication of reductionism is that the whole should be predicted by its parts.  We can predict systems like the ones you mention.  And if the system is too complex to predict, we usually do not observe anything as different as a conscious.  The properties that arise are normal physical types of properties that have not yet been modeled.  But the consciousness is not anything like the unexplained physical properties that are in complex systems.  

 

[Genady]

4 hours ago, Boltzmannbrain said:

But the consciousness is not anything like the unexplained physical properties that are in complex systems. 

I don't see what makes it so different. Do you?

I rather think that it is another yet unexplained property of a system and that it is a consequence of the system's structure.

Another analogy might be a programming language. It consists of a finite number of simple instructions. When put together in a certain way they calculate number pi.

[dimreepr]

9 hours ago, Boltzmannbrain said:

An implication of reductionism is that you can always make correct predictions about a system by knowing its parts.

An example of this is for science.  Most of the time science can make predictions by knowing the parts well enough.

But where hard reductionism fails is for the consciousness.  The parts that come together are thought not to be conscious.  So we have the opposite of reductionism, emergentism, or, another way to put it, irreducible.      

Reductionism fails because of tomorrow and becomes less and less relevant with every succeeding tomorrow.

You can't reduce consciousness, mostly because we can't understand the concept beyond ourselves; but also it's binary, because at the middle the spectrum consciousness is a switch and therefore it's a binary question; how do you reduce a binary question?

[joigus]

Reductionism has been very heavily criticised in science in the last decades. One good reason for this is that there are emergent aspects of natural laws that seem impossible to fathom by simply looking at the basic law and its constituent 'parts.'

For example, in recent decades there's been a lot of discussion about universality of certain power laws, which would occur no matter what constituent elements make up the 'stuff.' If such were the case, reductionism would take a big blow.

I would say the increasing relevance of this concept 'emergence' has a lot to do with why increasingly scientists are crossing out their names from the list of devotee reductionists.

[Boltzmannbrain]

7 hours ago, Genady said:

I don't see what makes it so different. Do you?

Yes, I definitely see what makes it different, its very nature.  It is not physical.  Physical properties affect the physical in one way or another.  The consciousness does not.  It shouldn't be there; it is unpredictable.  Also, a normal physical property is observed.  The consciousness observes; it does not get observed.  Those are two completely different things.  

Quote

I rather think that it is another yet unexplained property of a system and that it is a consequence of the system's structure.

And I agree that when these physical properties are put in the correct process, a consciousness emerges.  But given our understanding of fundamental physics, there is nothing known to this day that would predict such a property.  It is irreducible now.  I put "now" because I suppose maybe we just have to add consciousness to our models once we understand exactly what brings about consciousness.

Quote

Another analogy might be a programming language. It consists of a finite number of simple instructions. When put together in a certain way they calculate number pi.

I don't know enough about how reducibility applies to computer science.  From Wolfram,

"Computations that cannot be sped up by means of any shortcut are called computationally irreducible. The principle of computational irreducibility says that the only way to determine the answer to a computationally irreducible question is to perform, or simulate, the computation". 

Computational Irreducibility -- from Wolfram MathWorld

[Genady]

1 minute ago, joigus said:

universality of certain power laws, which would occur no matter what constituent elements make up the 'stuff

It happens thanks to mathematics, i.e., the same outcomes occur regardless of implementation when certain conditions are satisfied. It does not contradict that the outcomes occur because the constituent 'parts' satisfy the conditions.

I don't cross my name from that list, yet. Did not see yet an emergent property appearing from nowhere. 

[joigus]

5 minutes ago, Genady said:

It happens thanks to mathematics, i.e., the same outcomes occur regardless of implementation when certain conditions are satisfied. It does not contradict that the outcomes occur because the constituent 'parts' satisfy the conditions.

I don't cross my name from that list, yet. Did not see yet an emergent property appearing from nowhere. 

It's OK. There's a way to hold your own proudly in these anti-reductionist times. Just say you're a reductionist, only not just a naive reductionist.

 

 

[Genady]

10 minutes ago, Boltzmannbrain said:

Yes, I definitely see what makes it different, its very nature.  It is not physical.  Physical properties affect the physical in one way or another.  The consciousness does not.  It shouldn't be there; it is unpredictable.  Also, a normal physical property is observed.  The consciousness observes; it does not get observed.  Those are two completely different things.  

And I agree that when these physical properties are put in the correct process, a consciousness emerges.  But given our understanding of fundamental physics, there is nothing known to this day that would predict such a property.  It is irreducible now.  I put "now" because I suppose maybe we just have to add consciousness to our models once we understand exactly what brings about consciousness.

I don't know enough about how reducibility applies to computer science.  From Wolfram,

"Computations that cannot be sped up by means of any shortcut are called computationally irreducible. The principle of computational irreducibility says that the only way to determine the answer to a computationally irreducible question is to perform, or simulate, the computation". 

Computational Irreducibility -- from Wolfram MathWorld

I understand what you mean. But I think that this is unjustified. I see consciousness as physical. It might turn out equivalent to activation of a specific circuit in the brain, when consciousness is a subjective perception of this activation. Just like your subjective perception of a specific color is hidden from everyone else, but nevertheless is not unphysical and not irreducible to neuron processes.

Computational irreducibility is something completely different. Computational irreducibility is reduced by performing, or simulating, the computation.

16 minutes ago, joigus said:

It's OK. There's a way to hold your own proudly in these anti-reductionist times. Just say you're a reductionist, only not just a naive reductionist.

 

 

This is a good advice, and I will follow it. 

(As I've learned recently, during another discussion here, not to call myself an atheist, because of its narrow interpretation.)

[Boltzmannbrain]

19 minutes ago, Genady said:

I understand what you mean. But I think that this is unjustified. I see consciousness as physical. It might turn out equivalent to activation of a specific circuit in the brain, when consciousness is a subjective perception of this activation. Just like your subjective perception of a specific color is hidden from everyone else, but nevertheless is not unphysical and not irreducible to neuron processes.

I put equivalent in bold because now we are getting to the heart of the issue.  By "equivalent" are you saying that they are the same thing (interchangeable)?  If so, then this statement becomes an argument that the consciousness (mental aspect of it, not the physiological aspect) does not actually exist.  If yes, then two things are happening simultaneously: one, as you said, "an activation of a specific circuit"; two, subjective perception arises.  The former is a physical description of the system, and the latter is not a physical description of the system.

Please tell me your answer to my question above.

Quote

Computational irreducibility is something completely different. Computational irreducibility is reduced by performing, or simulating, the computation.

So computational irreducibility is reducible?

6 hours ago, dimreepr said:

Reductionism fails because of tomorrow and becomes less and less relevant with every succeeding tomorrow.

You can't reduce consciousness, mostly because we can't understand the concept beyond ourselves; but also it's binary, because at the middle the spectrum consciousness is a switch and therefore it's a binary question; how do you reduce a binary question?

Yes, that is a good point.

[EmDriver]

1 hour ago, Genady said:

I understand what you mean. But I think that this is unjustified. I see consciousness as physical. It might turn out equivalent to activation of a specific circuit in the brain, when consciousness is a subjective perception of this activation. Just like your subjective perception of a specific color is hidden from everyone else, but nevertheless is not unphysical and not irreducible to neuron processes.

Computational irreducibility is something completely different. Computational irreducibility is reduced by performing, or simulating, the computation.

This is a good advice, and I will follow it. 

(As I've learned recently, during another discussion here, not to call myself an atheist, because of its narrow interpretation.)

What do you think the role of having our pineal glands produce the worlds most powerful hallucinogenic by weight would be?

[sethoflagos]

17 hours ago, Boltzmannbrain said:

An implication of reductionism is that you can always make correct predictions about a system by knowing its parts.

An example of this is for science.  Most of the time science can make predictions by knowing the parts well enough.

But where hard reductionism fails is for the consciousness.  The parts that come together are thought not to be conscious.  So we have the opposite of reductionism, emergentism, or, another way to put it, irreducible.      

Consciousness is a really poor example on which to make definitive pronouncements. You are speculating and inviting others to speculate on a phenomenon that fails to generate a commonly agreed definition and for which there is no even partially accepted physical model.

Quote

A wide range of empirical theories of consciousness have been proposed.^[97][98][99] Adrian Doerig and colleagues list 13 notable theories,^[99] while Anil Seth and Tim Bayne list 22 notable theories.^[98]

Integrated information theory (IIT) postulates that consciousness resides in the information being processed and arises once the information reaches a certain level of complexity.^{[citation needed]}

Orchestrated objective reduction (Orch OR) postulates that consciousness originates at the quantum level inside neurons. The mechanism is held to be a quantum process called objective reduction that is orchestrated by cellular structures called microtubules. However the details of the mechanism would go beyond current quantum theory.^[100]

....https://en.wikipedia.org/wiki/Consciousness

A much better basis for discussion was offered by @joigus

1 hour ago, joigus said:

Reductionism has been very heavily criticised in science in the last decades. One good reason for this is that there are emergent aspects of natural laws that seem impossible to fathom by simply looking at the basic law and its constituent 'parts.'

For example, in recent decades there's been a lot of discussion about universality of certain power laws, which would occur no matter what constituent elements make up the 'stuff.' If such were the case, reductionism would take a big blow.

I would say the increasing relevance of this concept 'emergence' has a lot to do with why increasingly scientists are crossing out their names from the list of devotee reductionists.

But you seem to be ignoring this post.

 

[Genady]

40 minutes ago, Boltzmannbrain said:

The former is a physical description of the system, and the latter is not a physical description of the system.

Two descriptions of the same thing? This is fine with me. However, an external observer by observing the physical one is able to predict the mental one. 

 

49 minutes ago, Boltzmannbrain said:

So computational irreducibility is reducible?

Yes, it is.

 

50 minutes ago, Boltzmannbrain said:

  7 hours ago, dimreepr said:

how do you reduce a binary question?

With Heaviside step function.

 

19 minutes ago, EmDriver said:

What do you think the role of having our pineal glands produce the worlds most powerful hallucinogenic by weight would be?

How does it relate to the topic here?

[EmDriver]

1 hour ago, Genady said:

How does it relate to the topic here?

I'm trying to think of anything that could potentially trigger a circuit in the brain to activate like you were talking about. But yea, it's speculation on top of a speculation so it probably does deviate too off topic.

[Genady]

1 hour ago, EmDriver said:

I'm trying to think of anything that could potentially trigger a circuit in the brain to activate like you were talking about. But yea, it's speculation on top of a speculation so it probably does deviate too off topic.

Yes, the topic is reductionism. Then consciousness has been introduced as an example of a failure of reductionism. This example has two interrelated problems: it is too ill-defined to be example of anything, and it is not clear at all that reductionism fails there, rather a matter of opinion. Now, a possible neuroscience behind consciousness is even more speculative and goes too far off the reductionism question, as you said.  

[joigus]

Yes, exactly, @Genady and @sethoflagos. Consciousness, we don't know what that is. In particular, we don't know whether it's an emergent phenomenon or there is perhaps basic physics we still don't know about involved in it. So it might be a bit adventurous to try and guess whether it makes a case for reductionism or not.

On the other hand, things like Zipf's law:

https://en.wikipedia.org/wiki/Zipf's_law

IMO make a good showcase --if not a robust case-- for non-reductionist elements having to do with patterns we see across different phenomena. The explanation, it seems, is purely statistical. If that's the case, one could argue that it doesn't matter too much what constituent elements the ensemble is made of, and regularities appear because of the clustering of data for which the underlying law can be very different in nature.

Very similar to the linear power law between internal energy and temperature in ideal gases, with specific gases having different specific heats depending on whether the molecule is monoatomic, diatomic, etc. => A statistical reasoning that gives you a pattern irrespective of the reductionistic model, but a concrete microscopic model that completes the parametrics of the problem --typically the constants.

The question is not an easy one in general. There are cases --like the power law between metabolic rate and mass of a multicellular organism-- where the power law can be guessed at from some kind of reductionist first principles. See, eg, https://www.science.org/doi/10.1126/science.276.5309.122

Of which there is criticism too.

[Genady]

17 minutes ago, joigus said:

Yes, exactly, @Genady and @sethoflagos. Consciousness, we don't know what that is. In particular, we don't know whether it's an emergent phenomenon or there is perhaps basic physics we still don't know about involved in it. So it might be a bit adventurous to try and guess whether it makes a case for reductionism or not.

On the other hand, things like Zipf's law:

https://en.wikipedia.org/wiki/Zipf's_law

IMO make a good showcase --if not a robust case-- for non-reductionist elements having to do with patterns we see across different phenomena. The explanation, it seems, is purely statistical. If that's the case, one could argue that it doesn't matter too much what constituent elements the ensemble is made of, and regularities appear because of the clustering of data for which the underlying law can be very different in nature.

Very similar to the linear power law between internal energy and temperature in ideal gases, with specific gases having different specific heats depending on whether the molecule is monoatomic, diatomic, etc. => A statistical reasoning that gives you a pattern irrespective of the reductionistic model, but a concrete microscopic model that completes the parametrics of the problem --typically the constants.

The question is not an easy one in general. There are cases --like the power law between metabolic rate and mass of a multicellular organism-- where the power law can be guessed at from some kind of reductionist first principles. See, eg, https://www.science.org/doi/10.1126/science.276.5309.122

Of which there is criticism too.

I believe that common patterns appear because of common properties in arrangement of elements in different phenomena. Any purely statistical explanation is reductionist because it is based on assumptions about properties / behavior of elements. For example, the following explanation for Zipf's law, regardless right or wrong, is reductionist:

Quote

The principle of least effort is another possible explanation: Zipf himself proposed that neither speakers nor hearers using a given language want to work any harder than necessary to reach understanding, and the process that results in approximately equal distribution of effort leads to the observed Zipf distribution.

If this is not reductionist, then I don't know what reductionist means, which is quite possible as I didn't know what atheist means and all my life thought that I am one.