Possibility vs. Conceivability (pre-Chalmer notion ;p)
There is a great discussion on the relation of the concepts of possibility and conceivability. How you might wonder, don’t they mean the same (especially as a non-native speaker like me :D)? Before I dive into Chalmer’s Does Conceivability Entail Possibility? I want to ponder a bit on this curious relation.
If the two concepts are different from each other than there should be things / entities that are conceivable but not possible or (though more rarely a case?) possible but not conceivable.
What about the claim that Pegasus exists? Pegasus existing is certainly prima facie conceivable (ignoring the problem of the sentence’s truth value). However, is it possible for Pegasus to exist? This depends on your notion of possibility. If for something to be possible it needs to be able to exist in this actual world then many people might disagree with Pegasus possibly existing. Another understanding of possibility might allow Pegasus to exist in a possible world (an imaginary universe different from ours in minuscule or grand ways) and take this as a sufficient proof of Pegasus’s existence as being possible.
‘What is the difference of Pegasus being conceivable then?’, you might ask. It appears as if the notions of conceivability and possibility have merged. However, this only happens if you give a definition of possibility that depends on conceivability (or the other way around). So what are the definitions of the terms then?
Also, which term has a stronger bearing on Pegasus’s ontological status? Probably possibility is the stronger ontological status of an entity. The oxfordlearnersdictionary gives this definition for the adjective ‘conceivable’: ‘that you can imagine or believe’. Consequently something is inconceivable if it is not possible for you to imagine or believe in it.
You see what I did there? It seems so intuitive to define conceivability through possibility (for me). How can we separate them? The corresponding notions of the adjective ‘possible’ in the above mentioned dictionary are these: ‘that might exist or happen but is not certain to’ / or ‘reasonable or acceptable in a particular situation’.
So conceivability is concerned with the imagination and possibility with the actual ontological reality of something. Consider these examples:
Zombies (living dead) are conceivable -> Zombies (living dead) are possible
My own death is conceivable -> My own death is possible
Round squares are inconceivable -> Round squares are not possible.
Artificial consciousness is inconceivable -> Artificial consciousness is not possible
Artificial consciousness is conceivable -> Artificial consciousness is possible.
Suppose that I think of possibility as ‘possible in the actual world’. Accepting 1. then would likely pose a problem for me. Despite many things being conceivable (Pegasus, Zombies, Superman etc.) I would not think of all of them being possible. Zombies (living dead) are not possible in my understanding of the actual world if I prescribe to a reliable source of ‘knowledge’ such as modern science. Of course it could be that in the future science finds out that zombies (living dead) are actually possible. However, this may be a way to beg the question of asking if zombies (living dead) are possible. Consider this argument:
p1) Zombies are not possible as we understand them now
p2) Scientific advances can lead to new discoveries which we held to be impossible before
c1) Zombies may be possible in the future
c2) Zombies were possible after all.
I am not sure if this is valid argument. c2) is certainly a leap of some kind. Just because Zombies may be possible in the future it does not mean that they were possible after all. Zombies being possible may prove to be wrong no matter the scientific advances. How is it then, that I feel like many people are taking c1) to be a proof of c2)? The possible future argument does not yield much for me. If our method of gaining new beliefs is a reliable way to gather knowledge now then we should be justified to say that c2) is unlikely.
2. and 3. seem to be true prima facie. However, the same applies here, future could falsify these statements. An ideal reasoner that has evolved beyond human rationality may find that round squares are possible. Again I wonder what the future argument actually does for me. Consider:
q1) In the past people believed in the possibility of Pegasus’s existence
q2) Our scientific knowledge is sometimes (although rarely) wrong about substantial      matters of the world
c3) There is a possibility that Pegasus existed in the past or does exist right now
c4)Â It is possible for Pegasus to exist after all
r1) By our understanding right now it is impossible for Pegasus to exist
r2)Â Scientific advances can lead to new discoveries which we held to be impossible before
c5) It may be possible for Pegasus to exist in the future
c6) Pegasus existence is possible after all
Again c4) and c6) are leaps. However, with Pegasus as the thing that we ask to be possible c3) and c5) seem weird. How is it that the Zombies should be different? Let us leave it at that.
What about 4. and 5.? There are many people that believe either of these two statements are true. If we believe in the law of non-contradiction (and I suppose we do) then we have a problem here. Where lies the mistake?
There seem to be two possibilities. Firstly it could be that it is simply not conceivable whether it is conceivable or inconceivable that artificial consciousness will exists. How can this be though? Consider a):
a) It is inconceivable to find out if artificial consciousness is conceivable -> It is not possible to find out if artificial consciousness is conceivable.
Would not the truth of a) mean a dead end for us? Also, would not most people disagree with the conclusion of a) (in different ways though)? What is our other option?
Secondly, entailment of possibility by conceivability may simply be wrong. Even if it is not possible to conceive of artificial consciousness (now) its existence may be possible (in the future) or the other way around. Now, how come the future argument works better for 5. and 6. than for 1.?
b) Habitable planets similar to the earth are conceivable -> Habitable planets similar to the earth are possible.
c) Life beyond death is conceivable -> Life beyond death is possible.
It seems to me that here we have either two different notions of conceivability or possibility. That is because I would think that b) appears to be more secure than c). How come? Also c) seems to differ from 1. and we might follow Popper and say that c) cannot be falsified and consequently can neither be proven to be right or wrong (because the idea of life beyond death is historically beyond human reason) while 1. has at least a chance at being false (or true).
Anyways, what do we need to deny if we do not want zombies (living dead) to wander the earth? We need to deny i):
i) Conceivability entails metaphysical possibility.
ii) Conceivability entails mere possibility.
With mere possibility I mean that it certainly is possible for zombies (living dead) to exist and that there is a life beyond death but that this does not tell us anything about whether these things could really be actualized in our world / understanding of the human, universe etc. Metaphysical possibility instead implies that whatever is metaphysically possible can actually be realized / true in our actual world.
Do we also need a similar distinction for conceivability? Consider d) and e):
d) It is conceivable that tomorrow I will meet my girlfriend.
e) It is conceivable that tomorrow I will skip the bus and teleport to university.
Again d) and e) seem to play on a different notion of conceivability. d) is just more likely to actually be true. Consequently, we may speak of mere and metaphysical conceivability. What do we get from this insight?
iii) Metaphysical conceivability entails metaphysical possibility.
iv) Metaphysical conceivability entails mere possibility
v) Mere conceivability entails metaphysical possibility.
iv) Mere conceivability entails mere possibility.
iv) and v) seem unlikely to be correct by intuition. If something is metaphysically conceivable like in the case of d) than it seems unreasonable to claim that d) only possesses mere possibility. Similarly, if something is merely conceivable such as in e) then it would be unreasonable to claim metaphysical possibility for e) at the same time.
We are left with iii) and iv) which seem to imply that if something is not metaphysically conceivable then it will never possess metaphysical possibility. Sounds obvious to you?
Now, after we have (frantically) tried to find a difference in conceivability and possibility let us consider if they could be identical in meaning after all.
A (otherwise circular) definition of possibility might also be: ‘that what is conceivable’. However, how do we deal with the cases in which something is possible but not conceivable? Chalmers talks about a mathematician who holds that a specific outcome to a mathematical problem is inconceivable. Nevertheless, once the problem is solved this specific answer turns out to be true and was thus possible after all.
An objection to this example might be that the outcome of the problem was conceivable after all. The mathematician simply did not see that. However, by this reasoning we may again think of the ideal reasoner who finds many things conceivable that we find inconceivable because he is much smarter / rational than we are.
In the past there were many things that were inconceivable to people. Suppose that back in ancient Greece (not the best example for a culture that did not conceive of many things ;)) nobody would have thought of sound waves travelling on the medium of air. Was it inconceivable for them to think of this though?
How about that the idea that humans could be connected over a system of mysterious energy that manifests itself in complex devices that allow communication over distance? Not that anyone in ancient Greece would have thought of something so specific. However, is it right to say that it was inconceivable for them? Is it inconceivable that the ancient Greeks could have thought of something like this? I would say, why not? It is not like by a priori conditions they could not think of a system like this.
So if it is true that the internet was not a priori inconceivable despite them not being able to think of it (because they were still before the advent of electricity etc.) then why should we think that the ideal reasoner would find anything that we find a priori conceivable to be a priori inconceivable or anything that we find a priori inconceivable to be a priori conceivable?
An advocate of difference between conceivability and possibility may claim that this is because the ideal reasoner simply is much more rational than we are and that we cannot compare our a priori reasoning to hers. How much power can this argument claim though?
I need some more time to think about this.
