Maudlin's Computation and Consciousness

[Evgenii Rudnyi]

Dear Bruno,

Long time ago you have discussed Maudlin's paper. At that time I somehow
did not get interested. Yet, other day I have got strong feeling that I
must read Maudlin's paper right now. I guess this could be explained by
peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's
Sandman, and once more Maudlin's paper. I have enjoyed reading, the
paper is nicely written. I guess I have understood the argument. Thank you.

Best wishes,

Evgenii

[Bruno Marchal]

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7, or I could add a simpler step 8. 

Kind regards,

Bruno

Best wishes,

Evgenii

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

[David Nyman]

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

, or I could add a simpler step 8. 

And here. 

David 

Kind regards,

Bruno

Best wishes,

Evgenii

--
You received this message because you are subscribed to the Google Groups "Everything List" group.

To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscribe@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.

Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

--
You received this message because you are subscribed to the Google Groups "Everything List" group.

To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscribe@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.

[Bruno Marchal]

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

Bruno

David 

Kind regards,

Bruno

Best wishes,

Evgenii

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscribe@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

--
You received this message because you are subscribed to the Google Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to everything-list+unsubscribe@googlegroups.com.
To post to this group, send email to everything-list@googlegroups.com.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Everything List" group.

To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.

Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

[David Nyman]

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

David 

[Bruno Marchal]

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

To be sure, this assumes Mechanism, in some strong sense. When the argument is translated in mathematics, we get a more constructive view of what physics can be (the logic of Bp & Dt (& p) with p semi-computable (sigma_1, partial computable), and that can be tested. And if the test violates the logic Bp & Dt  (& p), it would mean that we have the following disjunction:

CTM is refuted OR we are in a malevolent simulation OR there is some magic at play.

Then, what the MGA shows, is that the third disjunct is basically an element of the first or second disjunct. It is not really necessary, except for those who don't know really what is a computation and like to cut the air. This makes also Maudlin's contribution more interesting, because it relates the difficulty of defining what is a physical instantiation of a computation. But with CTM, it is the whole "physical" idea which can no more be instantiated by any computation or even non-computation: it really becomes a view from inside arithmetic.

Note that in the original long text in french (Conscience & Mécanisme) the UDA and the MGA were used only to motivate the "& Dt" addition to Bp, to get a probability notion. When Brent asks us to take the environment into account, he is using a similar intuition. The logic G fails on knowledge, because it lacks Bp -> p, and it fails on "probability" because it lacks Bp -> Dp. Adding Dt avoids the cul-de-sac world/state, and provides sense to the idea of betting on alternative results possible.

Bruno

To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.

Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

[David Nyman]

On 5 Oct 2017 16:06, "Bruno Marchal" <mar...@ulb.ac.be> wrote:

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

To be sure, this assumes Mechanism, in some strong sense. When the argument is translated in mathematics, we get a more constructive view of what physics can be (the logic of Bp & Dt (& p) with p semi-computable (sigma_1, partial computable), and that can be tested. And if the test violates the logic Bp & Dt  (& p), it would mean that we have the following disjunction:

CTM is refuted OR we are in a malevolent simulation OR there is some magic at play.

Then, what the MGA shows, is that the third disjunct is basically an element of the first or second disjunct. It is not really necessary, except for those who don't know really what is a computation and like to cut the air. This makes also Maudlin's contribution more interesting, because it relates the difficulty of defining what is a physical instantiation of a computation. But with CTM, it is the whole "physical" idea which can no more be instantiated by any computation or even non-computation: it really becomes a view from inside arithmetic.

Note that in the original long text in french (Conscience & Mécanisme) the UDA and the MGA were used only to motivate the "& Dt" addition to Bp, to get a probability notion. When Brent asks us to take the environment into account, he is using a similar intuition. The logic G fails on knowledge, because it lacks Bp -> p, and it fails on "probability" because it lacks Bp -> Dp

Did you mean Bp -> Dt here, or am I missing another nuance? 

David 

[David Nyman]

On 5 October 2017 at 16:06, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

​A small additional point. It strikes me that insistence on an ultimately 'physical' ​basis for computation is related to the intuitionist tendency in mathematics, i.e. the notion that mathematics is in fact secondary to certain primary relations between physical objects. If so, computation in this view would similarly have to be seen as deriving from, or more simply *being*, such relations between particularised physical objects contrived for the purpose.

Trouble is, were that to be the case, the same arguments thus used to dismiss any 'existence', independent of physics, for computation would also dispose of any analogous existence for consciousness as its presumed consequence. IOW, if computation need be considered nothing more or less than the implied relations between certain physical objects then consciousness may be conceived in exactly the same way. But then we would be left with a bare identity theory: physical relations=consciousness, with no elementary connection to computation per se.

Also since 'existence' here essentially equates, as I remarked, to explanatory power, what will have occurred, on the foregoing assumptions, is the withdrawal of any such power from either computation or consciousness not already attributable to a primitively physical causality. Since explanatory entities should not needlessly be proliferated, the assumption of primitive physicality should equally entail the dismissal of such supernumerary hypotheses. AFAICS this is what Dennett, obfuscatory language apart, actually seeks to do. Trouble is, these 'supernumerary' hypotheses subsume the entire spectrum of phenomenal reality, or what we are pleased to call the ('actual') world. Ah well, so much the worse for something.

David

[Bruno Marchal]

On 09 Oct 2017, at 15:29, David Nyman wrote:

On 5 October 2017 at 16:06, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

​A small additional point. It strikes me that insistence on an ultimately 'physical' ​basis for computation is related to the intuitionist tendency in mathematics, i.e. the notion that mathematics is in fact secondary to certain primary relations between physical objects.

The common naturalism, I would say? usually intuitionism in mathematics is seen as an an idealism, even a solipisme: it is: "the first person view *only*". (at the depend of the other).

But I cut the air. I continue the distinction between Bp & p (S4grz, the soul, the first person singular) and Bp & Dt & p (the feeler, sensible matter).

If so, computation in this view would similarly have to be seen as deriving from, or more simply *being*, such relations between particularised physical objects contrived for the purpose.

What would that mean? We can define computation assuming elementary arithmetic. To define computation in physics, without using a subsystem of physics provably equivalent to the universal number/machine (the arithmetical notion) I am not sure the idea of defining computation in physical sense makes sense. You might say, with David Deutsch,  "unitary transformation", but that will only mean "computation" for the reason that you will prove that a function from N to N is computable iff it can be coded in that "programming" language, actually implemented also by arithmetic. It is selecting one universal number among all the other.

When actually with mechanism we must justify the "physical universal number" (the Gödel number of a first order specification of unitary transformation theory) from the sum on all computations.

Trouble is, were that to be the case, the same arguments thus used to dismiss any 'existence', independent of physics, for computation would also dispose of any analogous existence for consciousness as its presumed consequence.

Yes? Arguably so when keeping the mechanist assumption in the mind. To remain consistent here, it is important at some point to see that if true, the Mechanist hypothesis cannot be claimed to be true. The most we can get is a first person confirmation, but it proves nothing (even to oneself as the anosognosia illustrates).

IOW, if computation need be considered nothing more or less than the implied relations between certain physical objects then consciousness may be conceived in exactly the same way. But then we would be left with a bare identity theory: physical relations=consciousness, with no elementary connection to computation per se.

It is the belief that one universal machine won against all the other. 

With mechanism, we have the fight between a finite number of universal machine (from the colleagues to some bacteria ...) above our substitution level, and infinitely many universal numbers below our substitution level (the results of the global FPI on the infinitely many relative computations).

The chance is that this is structred by the logic of self-reference.

Also since 'existence' here essentially equates, as I remarked, to explanatory power, what will have occurred, on the foregoing assumptions, is the withdrawal of any such power from either computation or consciousness not already attributable to a primitively physical causality.

That could have made sense, but by assuming the digitalness mechanism (and most of its weakening with oracle) just made the invocation of the physical like an invalid use of an ontological commitment to avoid a problem.

If the physical play a role, it needs some magic ability to make consciousness, if consciousness is not in the trueness of the relation defining this or that computation. mechanism makes it into an inoperative god or a selector of computation, ignoring the first person indexical selection.

Since explanatory entities should not needlessly be proliferated, the assumption of primitive physicality should equally entail the dismissal of such supernumerary hypotheses. AFAICS this is what Dennett, obfuscatory language apart, actually seeks to do. Trouble is, these 'supernumerary' hypotheses subsume the entire spectrum of phenomenal reality, or what we are pleased to call the ('actual') world. Ah well, so much the worse for something.

It cannot work. This avoids (implicitly) the measure problem. With mechanism, the "hard problem of consciousness" is more easy: it is solved or meta-solved by the proof that any machine which observe itself is led to an undoubtable & non-rationally-justifiable truth (not necessarily having a unique form/content). But the "matter" problem becomes harder, as we can no more point to one big universal number (the physical world)  and invoke it, we have to justify it by computing the relative sum on all histories which brought our relative states. 

Dennett seems to take Matter for granted, and then explain Consciousness away. He is inconsistent, like all people believing in weak materialism and mechanism. That works locally, but makes no sense in the possible "big picture".

Note that "primary matter" is never assumed in physics. It is an assumption in metaphysics/theology. 

Bruno

To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.

Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

[David Nyman]

​I'm not sure that I can wholly agree with you about this. Or rather I agree that it's a metaphysical or theological assumption but it's nevertheless one that is probably assumed, even if unreflectively, as a physical one by many practitioners. For me, 'primary' implies the assumption of irreducibility in terms of the relevant theory. I suspect that most physicists - Deutsch for one with the notion of 'computable' as ultimately coterminous with physically-transformable - ​consider physics to be in a rather strong sense the 'rock bottom' of reality, not further reducible to something 'not-yet-physical'. Krauss comes to mind as someone who proposes physical law as the ultimate 'reason' there is something rather than nothing. Tegmark may be an exception, albeit a not-always-consistent one.

David

[Brent Meeker]

On 10/9/2017 6:29 AM, David Nyman wrote:

On 5 October 2017 at 16:06, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

​A small additional point. It strikes me that insistence on an ultimately 'physical' ​basis for computation is related to the intuitionist tendency in mathematics, i.e. the notion that mathematics is in fact secondary to certain primary relations between physical objects. If so, computation in this view would similarly have to be seen as deriving from, or more simply *being*, such relations between particularised physical objects contrived for the purpose.

Trouble is, were that to be the case, the same arguments thus used to dismiss any 'existence', independent of physics, for computation would also dispose of any analogous existence for consciousness as its presumed consequence. IOW, if computation need be considered nothing more or less than the implied relations between certain physical objects then consciousness may be conceived in exactly the same way. But then we would be left with a bare identity theory: physical relations=consciousness, with no elementary connection to computation per se.

Which is quite plausible.  It explains the dependence of consciousness on brain chemistry.  It would explain how there can be degrees and varieties of consciousness/awareness/perception/reflexion/etc.  The connection to computation would be no different that the connection of your computer to arithmetic.  I'm not sure what you mean by "elementary".

Also since 'existence' here essentially equates, as I remarked, to explanatory power, what will have occurred, on the foregoing assumptions, is the withdrawal of any such power from either computation or consciousness not already attributable to a primitively physical causality. Since explanatory entities should not needlessly be proliferated, the assumption of primitive physicality should equally entail the dismissal of such supernumerary hypotheses. AFAICS this is what Dennett, obfuscatory language apart, actually seeks to do. Trouble is, these 'supernumerary' hypotheses subsume the entire spectrum of phenomenal reality, or what we are pleased to call the ('actual') world. Ah well, so much the worse for something.

I don't think that is right at all.  You seem to be making an argument that if the physical is necessary for consciousness then consciousness doesn't exist...an argument you wouldn't make for insurance or war not existing.  Dennett's book is "Consciousness Explained"  not "Consciousness Explained Away".  Does it explain away arithmetic to note that it is always done by something physical, i.e. existing.

Brent

To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.

[Bruno Marchal]

I agree. But I would say that the "matter = primary matter" is the distinctive Aristotelian trait that we have inherited since Aristotle. We can still use it as a simplifying assumption, as long as we don't work on the mind-body problem, or don't dig too much on the meaning of the schroedinger equation. 

For me, 'primary' implies the assumption of irreducibility in terms of the relevant theory. I suspect that most physicists - Deutsch for one with the notion of 'computable' as ultimately coterminous with physically-transformable - ​consider physics to be in a rather strong sense the 'rock bottom' of reality, not further reducible to something 'not-yet-physical'. Krauss comes to mind as someone who proposes physical law as the ultimate 'reason' there is something rather than nothing. Tegmark may be an exception, albeit a not-always-consistent one.

Yes, the idea that the fundamental science might be mathematics instead of physics appeared with the greek Platonists. It has mainly the doctrine of those who called themselves Mathematician which might have meant originally those who reduce Plato worlds of Idea to the mathematical reality *only* (skeptical about primary matter), like Xeusippes (who asked Plato to fire Aristotle!). 

Plato did not follow them because he was wisest than that, and aware that the fundamental reality needs some relation with the soul experience, making him more "theologicalist" than mathematicalist (and I think he is right:  mathematics by construction remains in the realm of the 3p communicable, even when using axioms of infinity: there is still something more). 

Tegmark seems to ignore the mind-body problem in general, not just the computationalist version.  He is still not listening to the universal machine, nor to those listening to it, like the logicians Gödel, Löb, etc. 

Penrose did not realized, unlike already Emil Post, but also Judson Webb, the double-edge nature of using Gödel to prove-oneself superior or just different from the machine. When precise enough, we can see that the (Löbian) machine already found that argument, and refute it. Few people are aware of the importance of what Gödel saw already in 1931, the formalizability of its first incompleteness theorem. machines get it. 

The science of today is blinded on the theology/metaphysics issue: we do live in the Aristotelian Era, most people do assume, even unreflectively, that Reality is What we see, measure, observe, etc. It is normal in a history with predators and preys.

But if we come back to the scientific attitude in the field, the first exercise consists in succeeding in doubting the feeling of certainty that we are instinctively associating to our experiences. But with mechanism, and as the dream illustrates, we can believe and associate wrong interpretation on realities level, and the UDA is supposed to show that mechanism is either wrong or saves the "primary matter" appearance by exhibiting a coherent measure on all relative (to a Löbian machine) computational continuations.

That set of continuations is  structured in different ways in which the machine can see itself, leading to different logics and mathematics of true, provable, knowable, observable, sensible, and their splitting on proof and truth.

That is the whole point of the Aristotelian: they want the physical reality being conceptually simple, WYSIWYG: what you see is what you get.

Aristotle was aware it cannot work, and that is why they still add a "God", from its own First Mover, to a sort of Great Architect (more or less present even in some text of Plato like the Demiurge of the Timeaeus. Then by separating theology from science, that "god" becomes handy as a political tool, and the exploitation of the religious feeling/intuition.

Plato is often described as a dualist of some sort, because it is not alway clear if the sensible reality exists or not. He sees it as a pale imititation of the real, like a concrete circle drawn with a pen is an approximation of the "real platonic" circle. In my opinion, he was open to the abandon of the ontology on those approximations, notably in the Theaetetus and the Parmenides, but that move toward Monism (and their internal pluralism) has been mainly made by the Neopythagoreans, and the Neoplatonists.

Dennett and the Churchland, and others like Malcolm are coherent: if we want keep physicalism, we have to make consciousness inexistant, we have to declare ourself zombies! (or to abandon Mechanism, which they do not even suggest (they follow apparently Diderot equating Mechanism and Rationalism (that can make sense)).

If we keep Mechanism, Physicalism has to be abandoned, until we have some evidence of "primary matter", and I provide the way to look at such an evidence, like some serious departure from the logic of the observable inferred from observation (quantum logics) and the logic of the observable defined internally (S4Grz1, Z1*, X1*).

Now, if all physicist, and basically everybody today believes in the Aristotelian theology, my point is that they do not use it in their work. You will not find any paper in physics assuming primary matter. Only Einstein seems to intuit that the feeling of its existence is a religious feeling, although his argument defends more the feeling of wonder, and the music of the sphere. Gödel seemed to have eventually open his mind that such a feeling occur also in pure mathematics. Only Gödel will defend the idea that theology could be a science (but then missed a bit the point by formalising Saint-Anselme proof of the existence of God in an Leibnizian "naive" (S5) metaphysics, actually the only one not appearing in any machine's hypostases.

When theology and science are separated, automatically a part of science is taken for granted as being the non doubtable theology, like more or less naturalism/materialism is the current one. That is why atheists, by mocking theology, are de facto allies of all those who use argument per authority in the filed, and actually proceeds in the same way.

The debate God/Not-God is de facto a tool hiding the original 

question Primary-Physical-Universe/Non--Primary-Physical-Universe

I could have written Universe/Non-Universe, because today "Universe" is almost always conceived as "primary Physical universe", by the implicit metaphysics (imposed by 1500 years of Argument-Per-Authority, as there are still no evidence for such a primary Universe, also a fact rarely acknowledged, or made non-acknowledgeable by the WYSIWYG instinct/desire).

Bruno

To unsubscribe from this group and stop receiving emails from it, send an email to everything-li...@googlegroups.com.
To post to this group, send email to everyth...@googlegroups.com.

Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.

http://iridia.ulb.ac.be/~marchal/

[Bruno Marchal]

Have you read Dennets' book Brent? Consciousness is literally explained away in his chapter 12 "Qualia Disqualified".

Similarly, in his chapter 13 "The Reality of Selves" , the more general first person is under siege, I would say. 

I guess he realized it as the last chapter "Consciousness Imagined" aggravates his case and ended by a section--the last of the book entitled "Consciousness Explained, or Explained Away?".

He is almost enforced by its reasoning and metaphors to arrive at this by his unwillingness to doubt the existence of primary matter, that he confuses with matter like all Aristotelians.

He made eventually the usual argument that you have done sometimes: temperature does not disappear because it has been shown determined by molecules cinetic energy. I do agree that a "scientific explanation" must rely on 3p notions, but his physicalism forced him to explain consciousness in term of what he believes in: some "matter" having some ontological primariness, and that is provably impossible (by UDA) so he has to explain it away. We could have guessed at the start, becaude in the introduction he asserts that there is no more conceptual problem in physics, only technical issues, like Planck said before the genesis of relativity and the quantum (despite his key role in that genesis).

With mechanism, consciousness does not need to be reduced to anything physical: it is defined by machines experience, themselves defined in term of machine self-reference and their relation with truth, and we get the main attribute (undoubtability, apparent unjustifiability at the level of oneself, and a full theory of qualia extending quanta) from the number relations, syntactical and semantical, with some price for the Aristotelian, as mechanism explain *away* PRIMARY matter and refutes physicalism. But then, we don't have any conceptual problem with this, as there has never been any evidence for primary matter until now. It is a purely metaphysical notion to which we have some habit, easily explained by evolution. Mechanism shows primary matter as useful than ether, phlogiston or a god invoked in an explanation.  Dennett's book is just the nth attempt to put consciousness under the "material" rug, which he does not seem conscious that this too has to be explained. It is basically an unconscious act of faith in Aristotle theology. As much I liked his "brainstorm" and his book with Hofstadter "Mind's I", I find his "Consciousness explained" rather disappointing.

Bruno

http://iridia.ulb.ac.be/~marchal/

[David Nyman]

On 9 October 2017 at 20:07, Brent Meeker <meek...@verizon.net> wrote:

On 10/9/2017 6:29 AM, David Nyman wrote:

On 5 October 2017 at 16:06, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 05 Oct 2017, at 16:02, David Nyman wrote:

On 5 October 2017 at 13:50, Bruno Marchal <mar...@ulb.ac.be> wrote:

On 02 Oct 2017, at 15:07, David Nyman wrote:

On 2 Oct 2017 1:58 p.m., "Bruno Marchal" <mar...@ulb.ac.be> wrote:

Dear Evgenii,

On 28 Sep 2017, at 21:30, Evgenii Rudnyi wrote:

Long time ago you have discussed Maudlin's paper. At that time I somehow did not get interested. Yet, other day I have got strong feeling that I must read Maudlin's paper right now. I guess this could be explained by peculiarities of the universal dovetailer.

Anyway, I have read Maudlin's paper, then I have read Hoffmanm's Sandman, and once more Maudlin's paper. I have enjoyed reading, the paper is nicely written. I guess I have understood the argument. Thank you.

You are welcome. Don't hesitate to ask any further questions. There are slight differing nuances between the Movie Graph Argument (MGA) and Maudlin's argument. Also, I have found a way to cut the UDA at step 7

More details please. 

Maudlin focuses on the counterfactual issue, which the MGA somehow avoids. Both argument shows the inadequacy of materialism and mechanism, but the MGA assumes that we have already a good idea of what a (mathematical) computation is, and that it is a special logical relation, not something in need of any physical assumption. Then Maudlin's analysis can be used to answer the "counterfactual objection", instead of reminding implicitly the logical nature of what is a computation. In the context of the UDA, the rôle of the MGA is only to show that the move in step 7 consisting in assuming a small primary universe, is isomorphic to creationist last rebuttal when saying eventually that they agree with the evidences for evolution, but that God was needed to make that evolution possible. That was already clear with Peter Jones old objection to UDA that only a computation supported by primary matter  can be conscious. That is a sort of magic way of thinking, by rebutting a theory (experimentally testable) by invoking a god or a magic substance which a priori is not testable ... to avoid the search of an a posteriori test, given here by the theory.

, or I could add a simpler step 8. 

And here. 

So, it is enough just to NEVER assumes a *primary* physical universe to start with, like I do. Then, we can avoid the MGA by explaining directly (in step 7) that the arithmetical reality implements all computations, which follows from what we can find in all textbook on theoretical computer science, like Davis chapter 4 (Turing machine self-applied). I am currently using that material to explain that very points to my students, so I might make a try to explain a bit here, to give the taste of it. The advantage of the Turing machine formalism, is that it is close to "physical computer", and yet simple enough to make the theory not too much hidden in technical details. I will think how to do that.

In effect then one argues as follows. Beginning with the assumption of CTM, we can agree in principle that the existence of a computational device, instantiated in a primary physical reality capable of computing sufficient of the trace of the UD, would permit the UDA to go through. This is the initial assumption of Step 7 of the UDA. Then any objection that went to a presumed insufficiency of such a primary reality to implement such a computer would be a merely contingent supposition of its 'actual' non-existence. But the question of what is 'actual' with respect to the theory is precisely what is being asked. So any such "Show me the computer" type of objection begs that very question.

Absolutely.

This is really forced by the initial assumption of CTM, which puts any theory relying on it in the position of justifying the appearance of any possible physics, including physical computational devices, on the basis of the existence of an arithmetical, not physical, basis of computation. 'Arithmetic' here just stands for any theoretically irreducible and sufficient basis for computation. The relevant sense of 'existence', as in any fundamental theory, essentially equates to explanatory power. It would of course be open to anyone to additionally assume the existence of a putatively more 'explanatorily primitive' physical reality. But this could only weaken the theory by arbitrarily invoking the 'preselection' of undetectable mechanisms that then had no further explanatory role in what followed. Hence this move should be abandoned in favour of greater explanatory parsimony.

I think you see the point very well. It makes "primary or primitive matter" exactly like "invisible horse". It adds something "invisible" and gives to it magical abilities having virtually no sense in the frame of the hypothesis. 

​A small additional point. It strikes me that insistence on an ultimately 'physical' ​basis for computation is related to the intuitionist tendency in mathematics, i.e. the notion that mathematics is in fact secondary to certain primary relations between physical objects. If so, computation in this view would similarly have to be seen as deriving from, or more simply *being*, such relations between particularised physical objects contrived for the purpose.

Trouble is, were that to be the case, the same arguments thus used to dismiss any 'existence', independent of physics, for computation would also dispose of any analogous existence for consciousness as its presumed consequence. IOW, if computation need be considered nothing more or less than the implied relations between certain physical objects then consciousness may be conceived in exactly the same way. But then we would be left with a bare identity theory: physical relations=consciousness, with no elementary connection to computation per se.

Which is quite plausible.  It explains the dependence of consciousness on brain chemistry.  It would explain how there can be degrees and varieties of consciousness/awareness/perception/reflexion/etc.

​It would explain in the sense of giving an account, at whatever level of analysis, of the *covariance* of these categorically different things.​ It would not explain any necessity of a knowing subject to such an account, nor would it explain any topic-specific relation between that account and the phenomenology of such subjects. That's the reason Dennett is forced to say that the phenomenology is an 'illusion'. Perhaps somehow he hasn't yet twigged that an illusion is a veridical experience about whose origin we are mistaken.

  The connection to computation would be no different that the connection of your computer to arithmetic.

​Correct. And the connection of my computer, considered as a physical object​, to arithmetic is entirely a matter of extrinsic interpretation, a point you have often made yourself about your view of mathematics in general.

  I'm not sure what you mean by "elementary".

​I mean elementary in the sense of irreducible. In this case, the reference is to computation as the assumptive elementary ontology in the context of CTM.

​

Also since 'existence' here essentially equates, as I remarked, to explanatory power, what will have occurred, on the foregoing assumptions, is the withdrawal of any such power from either computation or consciousness not already attributable to a primitively physical causality. Since explanatory entities should not needlessly be proliferated, the assumption of primitive physicality should equally entail the dismissal of such supernumerary hypotheses. AFAICS this is what Dennett, obfuscatory language apart, actually seeks to do. Trouble is, these 'supernumerary' hypotheses subsume the entire spectrum of phenomenal reality, or what we are pleased to call the ('actual') world. Ah well, so much the worse for something.

I don't think that is right at all.  You seem to be making an argument that if the physical is necessary for consciousness then consciousness doesn't exist...an argument you wouldn't make for insurance or war not existing.  Dennett's book is "Consciousness Explained"  not "Consciousness Explained Away".

That isn't the argument, as we have discussed on many occasions before this. The physical is of course necessary for consciousness whatever one's theory, else that theory would demonstrably be false. The question is ​rather whether it is *sufficient* to explain consciousness, given a starting assumption of CTM.

Actually, it occurs to me that there may be an ambiguity, or at least a nuance

​,​

here which it may be worth making explicit. Sometimes in these discussions (e.g. with Peter Jones) what seems to be intended by 'physical' is some sort of 'stuffiness' that must elude any purely computational characterisation. IOW in this view the physical is more than the account of the evolution of physical states, notwithstanding how closed and 'complete' this may be. Moreover, this 'more than' is essential to any related phenomenon such as consciousness. It seems pretty obvious that this view would rule out any computational theory of mind that wasn't a clandestine dismissal of the subject matter.

The second variety, which I think is the one actually under discussion, is the notion that the 'physical' is indeed entirely characterisable in terms of computable state evolutions, with nothing essential thereby omitted. This version of physics equates to some part of the Computational Library of Babel which comes 'pre-selected', as it were, as an ontological given from which we are to derive all other phenomena. This preselection sets up a critical distinction because it demotivates any role for the knowing subject as selector of its own consistent phenomenal - and hence physical - environment. Without such a role what remains is then the bare evolution of a physical state but this, after all, was what demanded a 'physical' explanation. Physics doesn't set out to explain subjects or observers, although it is nonetheless forced to assume them. An explanatory hierarchy built on that foundation inherits the same limitations and that is its weakness in providing any illuminating explanation for consciousness.

 

  Does it explain away arithmetic to note that it is always done by something physical, i.e. existing.

​Well, yes it does on materialist assumptions. Arithmetic is derived purely from mentality, and mentality is in turn purely material. I think that's essentially the intuitionist thrust in mathematics. Dennett might do well to reconsider his old mentor Ryle's analogy of the University. "Where is it?" demands the naive student when presented merely with the college buildings and other such physical paraphernalia. Well of course we're supposed to smile indulgently at such innocence, but indeed where is the bloody thing? It's just a figment of our inter-subjective imaginings and, on Ryle/Dennett's own terms, that can be well enough accounted for by neurocognition.

David