Is this a valid argument?

Use this philosophy forum to discuss and debate general philosophy topics that don't fit into one of the other categories.

This forum is NOT for factual, informational or scientific questions about philosophy (e.g. "What year was Socrates born?"); such homework-help-style questions can be asked and answered on PhiloPedia: The Philosophy Wiki. If your question is not already answered on the appropriate PhiloPedia page, then see How to Request Content on PhiloPedia to see how to ask your informational question using the wiki.
Kaz_1983
Posts: 391
Joined: May 26th, 2019, 6:52 am

Is this a valid argument?

Post by Kaz_1983 » September 13th, 2020, 10:04 am

P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.

User avatar
Terrapin Station
Posts: 4083
Joined: August 23rd, 2016, 3:00 pm
Favorite Philosopher: Bertrand Russell and WVO Quine
Location: NYC Man

Re: Is this a valid argument?

Post by Terrapin Station » September 13th, 2020, 6:50 pm

It's not really valid as stated, partially because it's stated rather carelessly. There are a large number of different terms in the way you're stating this, and the relationship between the terms is rather unclear, which makes ambiguous with respect to determining validity.

More broadly, it seems like you're saying that if a person isn't deciding actions, then we're not talking about morality, and then you're noting that computer simulations don't involve people doing things . . . But that doesn't seem to get the gist of what computer simulations are. It just needs to bear some observable resemblance to how people actually make decisions about what to order.

Morality is probably only going to confuse the issue here. It would be better to start with this: is there any way to create a computer simulation of any sort of decision-making? Couldn't we create a computer simulation of, say, 50 different people deciding what to order for lunch? And couldn't we do that in a more sophisticated way than simply using a random number generator, by defining our simulated persons' preferences, defining the probability that they'll simply choose their normal order (in line with their preferences), etc., where we could use empirically-gained statistics as a guide?

Maybe the computer wouldn't literally be making decisions, but it doesn't need to. It's merely simulating this.

User avatar
Terrapin Station
Posts: 4083
Joined: August 23rd, 2016, 3:00 pm
Favorite Philosopher: Bertrand Russell and WVO Quine
Location: NYC Man

Re: Is this a valid argument?

Post by Terrapin Station » September 13th, 2020, 6:51 pm

Oops, this sentence was supposed to be at the end, rather than in the middle of my post: "It just needs to bear some observable resemblance to how people actually make decisions about what to order."

Gertie
Posts: 1047
Joined: January 7th, 2015, 7:09 am

Re: Is this a valid argument?

Post by Gertie » September 13th, 2020, 7:44 pm

Kaz_1983 wrote:
September 13th, 2020, 10:04 am
P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.
P1 - I'm not sure if this is necessarily, or always will be, true. if it is, fine.

P2 - You might consider the objection that there could be an inherently morally responsible person or machine, but they do not have the opportunity to act morally, as in P1.

P3 - I'm not sure what you're saying here, or even if you need it.

User avatar
Marvin_Edwards
Posts: 607
Joined: April 14th, 2020, 9:34 pm
Favorite Philosopher: William James
Contact:

Re: Is this a valid argument?

Post by Marvin_Edwards » September 13th, 2020, 11:36 pm

Kaz_1983 wrote:
September 13th, 2020, 10:04 am
P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.
The broader model is this: 1) Something causes unnecessary harm. 2) We desire to prevent that harm from continuing or repeating. 3) We seek to identify the cause(s) of the harm. 4) We correct the causes. 5) That harm is no longer repeated and everyone is better off. That's the goal of morality.

A computer may have a programming error that causes harm. The programmer applies debugging techniques to identify and correct the error resulting in less future harm. The moral model here is that the programmer holds the program responsible for the harm and uses the appropriate method to correct it.

A driver accidentally runs over a child. Accident investigators seek the causes of the accident. Was the driver intoxicated? Was the driver distracted by his cell phone? Was the child unsupervised? Did the child run into the street making the accident unavoidable? Were there issues with the traffic signals? Was the speed limit appropriate for that road?

The key is to correct any and all of the significant causes of the accident. Each cause is effectively "held responsible" for its role in the accident and corrective measures are applied. Addressing all of the significant causes minimizes the risk of accidents in the future.

However, the method of correction will be whatever is most effective and appropriate for each cause.

A schizophrenic kills someone because he is convinced they are an alien from outer space intent on killing him. The mental illness is held responsible for the murder and subjected to psychiatric treatment for correction.

A person deliberately robs a convenience store because he's short on cash. The deliberation process is the cause. The person is held responsible and treated in a correctional facility using punishment, counseling and other methods of rehabilitation designed to correct his thinking process.

User avatar
Steve3007
Posts: 7922
Joined: June 15th, 2011, 5:53 pm
Favorite Philosopher: Eratosthenes of Cyrene
Location: UK

Re: Is this a valid argument?

Post by Steve3007 » September 14th, 2020, 3:08 am

Kaz_1983 wrote:P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.
It seems to me that this is simply a restatement of the free will versus determinism argument in which it is imagined that we could rewind the universe to a prior state and it is proposed that for a given prior state the universe will always evolve into the same later state. Your version of the argument uses the fact that computer simulations are deterministic and that it is possible to re-run them with exactly the same initial state.

The argument says that since our actions would be exactly the same each time, this somehow means that we're not choosing those action. (Your version further implicitly says that lack of choice removes moral responsibility). I think the flaw in the argument is in the fact that, unlike with a computer simulation, the idea of rewinding the universe to precisely the same state is an abstract concept. It's not real. It's an idealization existing only in our minds, like mathematics. The whole concept of absolute 100% precision is an abstract concept in the same sense that (and for the same reason that) such things as dimensionless points are abstract concepts. So the idea that we lack free will is an abstract concept.

User avatar
Marvin_Edwards
Posts: 607
Joined: April 14th, 2020, 9:34 pm
Favorite Philosopher: William James
Contact:

Re: Is this a valid argument?

Post by Marvin_Edwards » September 14th, 2020, 5:47 pm

Steve3007 wrote:
September 14th, 2020, 3:08 am
Kaz_1983 wrote:P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.
It seems to me that this is simply a restatement of the free will versus determinism argument in which it is imagined that we could rewind the universe to a prior state and it is proposed that for a given prior state the universe will always evolve into the same later state. Your version of the argument uses the fact that computer simulations are deterministic and that it is possible to re-run them with exactly the same initial state.

The argument says that since our actions would be exactly the same each time, this somehow means that we're not choosing those action. (Your version further implicitly says that lack of choice removes moral responsibility). I think the flaw in the argument is in the fact that, unlike with a computer simulation, the idea of rewinding the universe to precisely the same state is an abstract concept. It's not real. It's an idealization existing only in our minds, like mathematics. The whole concept of absolute 100% precision is an abstract concept in the same sense that (and for the same reason that) such things as dimensionless points are abstract concepts. So the idea that we lack free will is an abstract concept.
Ironically, the problem rest upon the simple distinction between "could have" and "would have". The "rule" is that, whenever choosing happens in a universe of perfectly reliable causation, "I could have done otherwise" will always be true while "I would have done otherwise" will always be false.

Choosing inputs two or more options, applies some criteria of comparative evaluation, and outputs a single choice. Each option is a "can do". Each choice is a "will do". Choosing requires, by logical necessity, at least two real possibilities to choose from. We will always have at least two "can do's" whenever choosing is involved. The past tense of "can" is "could have". The past tense of "will" is "would have".

At the end of the choosing operation between two options, we'll have exactly one "will do" and one "could have done" for the option that was not selected.

The hard determinist simply needs to stop asserting that we "could not have done otherwise" and replace it with the assertion that we "would not have done otherwise".

And no one would argue that they would have made a different choice given the same person, the same problem, and the same circumstances. After all, if I've thought the problem through, and A was a better choice than B, then why would I ever select B rather than A?

But telling someone that they "could not" have made any other selection collides with the logic of the choosing operation, and people will insist that they could have made a different choice, because they had two real possibilities and the ability to choose either one.

Alias
Posts: 2826
Joined: November 26th, 2011, 8:10 pm
Favorite Philosopher: Terry Pratchett

Re: Is this a valid argument?

Post by Alias » September 15th, 2020, 12:45 am

Maybe, within its parameters - whatever they are. What it is is unclear.
Kaz_1983 wrote:
September 13th, 2020, 10:04 am
P1) A computer simulation can’t have any uncontrollable variables.
A computer simulation of what? What does 'can't have' mean - that it is not possible for such a variable to get into a computer simulation, or that if such a variable were introduced, the simulation would become invalid? Why not? By what law?
P2) Moral responsibility needs personal control over your actions.
What's moral responsibility got to do with computer simulations? Whose personal control over whose actions?
P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.
Who took personal control of what from the simulation? Who is "yourself"? What makes "yourself" "uncontrollable"? How did "yourself" get into the computer simulation?
C) Moral responsibility within a computer simulation cannot exist.
That's probably true. Why not just propose it as question, without the non-premises?

User avatar
Steve3007
Posts: 7922
Joined: June 15th, 2011, 5:53 pm
Favorite Philosopher: Eratosthenes of Cyrene
Location: UK

Re: Is this a valid argument?

Post by Steve3007 » September 15th, 2020, 2:51 am

Marvin_Edwards wrote:The hard determinist simply needs to stop asserting that we "could not have done otherwise" and replace it with the assertion that we "would not have done otherwise".
In my view, he/she also needs to bear in mind that "would [not] have done otherwise" is an untestable proposition because, as I said, the notion of the universe being rewound to precisely the same state is an entirely abstract one, because 100% precision is an entirely abstract concept. And vanishingly small differences in early conditions result in macroscopic differences in later conditions.
And no one would argue that they would have made a different choice given the same person, the same problem, and the same circumstances. After all, if I've thought the problem through, and A was a better choice than B, then why would I ever select B rather than A?
I bet somebody would. There would probably be ambiguity as to what exactly is meant by "the same person, the same problem, and the same circumstances". And even if that were cleared up, I know of at least one person who would say that the essence of choice is that "the same person, the same problem, and the same circumstances" can indeed yield a different decision. That person would call anyone who says differently a fascist, or possibly a Marxist.

User avatar
Sculptor1
Posts: 2532
Joined: May 16th, 2019, 5:35 am

Re: Is this a valid argument?

Post by Sculptor1 » September 15th, 2020, 6:45 am

Kaz_1983 wrote:
September 13th, 2020, 10:04 am
P1) A computer simulation can’t have any uncontrollable variables.

P2) Moral responsibility needs personal control over your actions.

P3) Taking person control away from the simulation means adding yourself as an uncontrollable variable into the computer simulation.

C) Moral responsibility within a computer simulation cannot exist.
P1. Computers can simulate lack of control with the random number function.
P2 "moral responsibility" is not a thing that can "need" any thing. Moral Responsibility is an abstract concept with no needs or volition. If you were to restate this proposition, things might become clearer.
P3 Not sure where you are going with this. What simulation? If you are saying that a player is equivalent to an uncontrollable variable in a computer simulation, then fine. From the POV of the game designer, s/he will be adding an unknown moral agent into the game.
C Does not follow, since morality can be easily included in a game and the player can bring morality to bear in the game. Game designers can easily provide rewards for moral behaviour and demerits for behaviours they do not want to encourage. For example a game often awards points for killing. Who and what you kill is making a moral point, be they simulated innocents or monsters.

User avatar
Marvin_Edwards
Posts: 607
Joined: April 14th, 2020, 9:34 pm
Favorite Philosopher: William James
Contact:

Re: Is this a valid argument?

Post by Marvin_Edwards » September 15th, 2020, 8:06 am

Steve3007 wrote:
September 15th, 2020, 2:51 am
Marvin_Edwards wrote:The hard determinist simply needs to stop asserting that we "could not have done otherwise" and replace it with the assertion that we "would not have done otherwise".
In my view, he/she also needs to bear in mind that "would [not] have done otherwise" is an untestable proposition because, as I said, the notion of the universe being rewound to precisely the same state is an entirely abstract one, because 100% precision is an entirely abstract concept. And vanishingly small differences in early conditions result in macroscopic differences in later conditions.
Right. It is untestable. But it is just a "thought problem", so one might think it through, and see it as a movie that we rewind and replay.

I think that chaotic behavior (small divergences quickly ballooning into big differences) is a problem of prediction rather than of causation. But, solving the problem of prediction is key to any practical application of the concept.
Marvin wrote:And no one would argue that they would have made a different choice given the same person, the same problem, and the same circumstances. After all, if I've thought the problem through, and A was a better choice than B, then why would I ever select B rather than A?
Steve3007 wrote:
September 15th, 2020, 2:51 am
I bet somebody would. There would probably be ambiguity as to what exactly is meant by "the same person, the same problem, and the same circumstances". And even if that were cleared up, I know of at least one person who would say that the essence of choice is that "the same person, the same problem, and the same circumstances" can indeed yield a different decision.
The response to that claim would be "why did he choose differently this time?" And the answer would be a cause that reliably brought about the result. That's how the determinist would make his case.
Steve3007 wrote:
September 15th, 2020, 2:51 am
That person would call anyone who says differently a fascist, or possibly a Marxist.
I suppose that person might as well call them a "kangaroo"or an "aardvark" if name-calling is the point. But that should not be the point.

User avatar
Steve3007
Posts: 7922
Joined: June 15th, 2011, 5:53 pm
Favorite Philosopher: Eratosthenes of Cyrene
Location: UK

Re: Is this a valid argument?

Post by Steve3007 » September 15th, 2020, 9:04 am

Marvin_Edwards wrote:Right. It is untestable. But it is just a "thought problem", so one might think it through, and see it as a movie that we rewind and replay.

I think that chaotic behavior (small divergences quickly ballooning into big differences) is a problem of prediction rather than of causation. But, solving the problem of prediction is key to any practical application of the concept.
Quite, but in my view there is a strong connection between problems of prediction/practical application and the question of whether an idea is physically meaningful. If an abstract idea is entirely abstract, in the sense that it has no applicability to the physical world, then I don't think it's coherent to make propositions about the physical world based on that idea. And I think a proposition like "humans have free will" or its opposite is a proposition about some things that exist in the physical world (humans).

So, to use an example I mentioned earlier, the concept of a dimensionless point is not applicable to the physical world, and therefore the concept of a 100% precise measurement is similarly inapplicable. Of course, we can still consider real, physical systems to approach arbitrarily close to being of zero (or infinite) size. And we can still consider real, physical measurements to approach arbitrarily close to being 100% accurate (notwithstanding QM). But that's not the same as saying that they can actually reach those limiting states. It just means (in the latter case) that for any given measurement accuracy, it's physically meaningful to contemplate a more accurate one.

User avatar
Steve3007
Posts: 7922
Joined: June 15th, 2011, 5:53 pm
Favorite Philosopher: Eratosthenes of Cyrene
Location: UK

Re: Is this a valid argument?

Post by Steve3007 » September 15th, 2020, 9:13 am

Marvin_Edwards wrote:The response to that claim would be "why did he choose differently this time?" And the answer would be a cause that reliably brought about the result. That's how the determinist would make his case.
And some others would take the position that human actions are acausal.
I suppose that person might as well call them a "kangaroo"or an "aardvark" if name-calling is the point. But that should not be the point.
Yes. He was a bit batty. Hang around for a bit and he'll be back. His name on here is Syamsu.

User avatar
chewybrian
Posts: 748
Joined: May 9th, 2018, 7:17 pm
Favorite Philosopher: Epictetus
Location: Florida man

Re: Is this a valid argument?

Post by chewybrian » September 15th, 2020, 10:03 am

Steve3007 wrote:
September 15th, 2020, 2:51 am
I know of at least one person who would say that the essence of choice is that "the same person, the same problem, and the same circumstances" can indeed yield a different decision. That person would call anyone who says differently a fascist, or possibly a Marxist.
Well, I don't agree with the facist/Marxist labels, but I do agree with the description of choices. In fact, I would say this is a fair description of the essence of being human. It's the only thing that gives meaning to being alive. This version of true choice is before us at every moment. It is not a theory. Rather the denial of it is the theory.

So, maybe you know at least two.
"If determinism holds, then past events have conspired to cause me to hold this view--it is out of my control. Either I am right about free will, or it is not my fault that I am wrong."

User avatar
Marvin_Edwards
Posts: 607
Joined: April 14th, 2020, 9:34 pm
Favorite Philosopher: William James
Contact:

Re: Is this a valid argument?

Post by Marvin_Edwards » September 15th, 2020, 11:07 am

Steve3007 wrote:
September 15th, 2020, 9:04 am
Marvin_Edwards wrote:Right. It is untestable. But it is just a "thought problem", so one might think it through, and see it as a movie that we rewind and replay.

I think that chaotic behavior (small divergences quickly ballooning into big differences) is a problem of prediction rather than of causation. But, solving the problem of prediction is key to any practical application of the concept.
Quite, but in my view there is a strong connection between problems of prediction/practical application and the question of whether an idea is physically meaningful. If an abstract idea is entirely abstract, in the sense that it has no applicability to the physical world, then I don't think it's coherent to make propositions about the physical world based on that idea. And I think a proposition like "humans have free will" or its opposite is a proposition about some things that exist in the physical world (humans).

So, to use an example I mentioned earlier, the concept of a dimensionless point is not applicable to the physical world, and therefore the concept of a 100% precise measurement is similarly inapplicable. Of course, we can still consider real, physical systems to approach arbitrarily close to being of zero (or infinite) size. And we can still consider real, physical measurements to approach arbitrarily close to being 100% accurate (notwithstanding QM). But that's not the same as saying that they can actually reach those limiting states. It just means (in the latter case) that for any given measurement accuracy, it's physically meaningful to contemplate a more accurate one.
As to relevance, I often assert that universal causal necessity/inevitability makes itself irrelevant by its own ubiquity. It's like a constant that appears on both sides of every equation and can be subtracted from both sides without affecting the result. It has no practical implications to any real world scenarios.

One can acknowledge it as a logical fact, but having acknowledged it there is nothing more to do with it than to ignore it. And that's what the reasonable mind does automatically.

Post Reply