This thought is a spinoff from a freewill issue in the AI for beginners thread…
I have heard it said that perfect reason would not have a real choice when faced with more than one option (in any decision), as one option would always be better than the rest and therefore perfect reason would choose it every time.
Is this right?
Isn’t it logically possible to be faced with two (or more) choices that are different but exactly equal in merit, benefit etc?
Nevertheless, I’d say yes, that’s possible, why wouldn’t it be?.
I don’t readily see why that would be a problem for any theory of the mind, though, as long as it isn’t a theory that makes the mind a very, very rigid thing.
It’s certainly logically possible. Practically possible? No, for the same reasons it’s not practically possible to create a block that weighs precisely 1kg. Perfect precision is pretty hard to come by.
Fortunately, though, perfect reason also doesn’t exist - reason is always tempered by emotion, which is pretty fly-by-the-seat-of-your-pants sorta deal.
Jeff
Sure…
do you go with the blond or the brunette?
They are different because both have attractive personalities.
They are equal because there is no blond.
Presumably only if the person making the subjective judgement is equally divided him/herself. That would seem to be the issue; ‘we’ are very long way from understand all our drives, subconscious, emotional and other, so my initial view would be that without that foundation of knowledge it’s impossible to factor eveything involved in decision-making.
Certainly there are trivial examples of such a decision, e.g. “Which would you rather have, a quarter or two dimes and a nickel.” But I don’t think that’s the question that you are asking. I think you are asking “What does a perfect reasoner do when presented with two equivalent choices?”.
Two answers:
a. If you could have a perfect reasoner, it would reason “It doesn’t matter which one I choose” and would select one alternative randomly.
b. If you were in the real world, you can’t have a perfect reasoner for a bunch of reasons, including lack of complete knowledge (both of the present and of the future), and lack of enough computational resources to evaluate all possible alternatives before the universe burns out. So what happens is that you spend as much time dithering between the two options as time and energy allow, and then you make your best guess.
A response to the practically possible contradiction.
If there is no way to tell the difference between two things in a particular measurement are they equal in that measurement?
So that even though perfect measurement is impossible, due to quantum mechanics for instance, there can exist a best possible measurement.
So even with perfect reasoning you could not determine the heaviest of two objects both with mass 1kg as measured to the best possible measurement. Thus both would be equal choices if choosing based purely on mass.
As to Goodbeem’s cite.
Has a donkey ever died for lack of decision between two means of escape from death? The cite seems to not consider that randomness in decision cannot exist. But randomness in decision is necessary for individual decision to be optimized for a group.
A group which always makes the best possible decisions individually will be predictable, and will use up resources faster than a group whose individual decision making processes are at least partially random. This is IMHO where I am considering such simplified groups as are present in * Simulated Annealing * or * Genetic Algorithms * and applying these to a more complex group of living entities.
The movie “Sophie’s Choice” had such a choice: The title character had to choose which of her two children to save from death. Horrible.
Sure it is, at least for trivial things. If you have to pick between chocolate ice cream and vanilla ice cream, I think either choice has about the same moral value.
That’s precisely why our nervous system has noise.
Oh, and it also makes it easier to detect weak patterns.
Bippy, I mentioned the cite because Mangetout’s question reminded me of Buridan’s Ass.
If I recall correctly, this “paradox” was first used as an argument in a medieval debate over free will (and I have the impression that is were Mangetouts question derives from also, but I could be wrong). Without free will, the argument goes, the donkey (having the relevant perfect knowledge and rationality) would not be able to choose between either side. Enter free will, and the donkey lives!
I’m not saying this is a valid argument (it confuses randomness or noise with free will, to start with). It does try to combine rationality with free will, which is kind of interesting to me. And as far as I know, no donkeys have ever died in these experiments.
Thanks Goodbeem I wasn’t so much attacking your cite, which was very appropriate. I was wondering if there are cases where living organisms (maybe much simpler than Donkeys, Bacteria for instance) can be placed between two food sources and yet starve to death because they fail to move to one or the other. I could imagine a bacteria with chemical concentration sense and cilia trying to swim on two directions at the same time and failing to get anywhere. Could this happen with more complex creatures? At what point does random noise or free will start to manifest itself?
At a very basic level, I’d presume.
I think that it might be possible for a bacterium to oscillate endlessly between two attractive locations, but the circumstances required for that to happen are so unlikely that it virtually never happens. (Once the bacterium heads towards a source, it’ll tend to be more attracted to it and less than the other.)
Of course it’s possible! Watch:
Mrs. Beaker: Which blouse should I wear? The blue one or the red one?
Grim_Beaker: I don’t care!
Interesting…
A search in Phil. Abstracts, nor PsycInfo, nor Zoological Record, yields an experiment of the sort Bippy wonders about. So I guess it hasn’t been done. It has been tried with animats, though. Most mobots seem to have no problems with equal options.
Yes. It has happened to me. Two enzymes, identical activity and concentration, identical price, identical means of ordering, identical lab history with each different company (none at all).
So I flipped a coin.
That’s the frustrating thing about armchair philosophy – someone smarter has thought longer and better about all the questions that I come up with. So, what’s the point? 
It’s fun! it’s fun!
Not quite; the qeuation is more along the lines “is it actually possible to offer the perfect reasoner two different choices which are exactly equivalent?” - or does the fact that they are different make it impossible that they are equivalent at every level?
Two things cannot be equal in every way and yet distinguishable. Since distinguishability requires difference. But a choice can exist between two things that even a perfect reasoner cannot measure one choice as greater than the other. For instance
If a perfect reasoner, has perfect ability at reasoning, but is not omniscient, they cannot chose between two unknown items (the contents of two otherwise identical boxes, where the boxes cannot be examined before the choice).
If a perfect reasoner would have to be omniscient to be able to distinguish between two apparently identical things, but what then in this scenario.
Mrs. Omniscient knows all and will post a letter today and a letter tomorrow. She also has two stamps. She knows both letters will arrive on time no matter which stamp is put on which letter.
Maybe she knows that no matter which stamp she uses when there will be no difference in the Universe other than the stamp she chooses.
So though everything is different, the ability to tell the difference between any two things is omniscience, and that ability will lead to the omniscient being knowing that two things can have identical value.
OK I’m probably off key somewhere here, but I hope this is of some interest.
Cheers, Bippy
