Agrippa's Trilemma

[The Black Jester]

In Epistomelogical circles, Agrippa's Trilemma refers to the difficulty in establishing a rational foundation for any give piece of knowledge. 

Any reason we give for a particular 'truth' must itself be justified, and so on, ad infinitum.  Is there any way out of this infinite regress?

Three responses have been proposed:

Infinitism: There is indeed an infinite regress of 'reasons,' but this is not a problem.  At any point, a proper reason can be given for any link in the infinite chain when called for.

Foundationalism: There are truths that are simply 'obvious' and 'self-justifying,' in need of no further rationalization, and upon which all knowledge can be built.

Coherentism: the chain of rationalizations of any particular fact is not a linear chain but a circular one.  Something is true because of how well it 'coheres' with other facts and theories to which it is tied.  All elements in this circular chain of reasoning support one another.

Which response is most persuasive to you, and why?  Or do you have another response?

[Davin]

I'd say all three, but the explanation will take some time so I'll do it after work if I remember, otherwise I will definitely do it on Tuesday.

I normally don't use the forums when I'm not at work. This is just a place holder so that I don't have to search for it in case it doesn't pop up in recent posts.

[Recusant]

My take on this is probably closest to what you call Foundationalism.  I wouldn't necessarily call any axioms 'obvious' and 'self-justifying,' but one must proceed from some basis. Axioms from which it's possible to produce working results are justified to some extent by those results, however I think it's good practice to consider such axioms provisional rather then final; being aware that new information may modify them.

[The Black Jester]

I'd say all three, but the explanation will take some time so I'll do it after work if I remember, otherwise I will definitely do it on Tuesday.

I normally don't use the forums when I'm not at work. This is just a place holder so that I don't have to search for it in case it doesn't pop up in recent posts.

Awesome!  I appreciate you engaging!

[The Black Jester]

My take on this is probably closest to what you call Foundationalism.  I wouldn't necessarily call any axioms 'obvious' and 'self-justifying,' but one must proceed from some basis. Axioms from which it's possible to produce working results are justified to some extent by those results, however I think it's good practice to consider such axioms provisional rather then final; being aware that new information may modify them.

Obvious in the sense that they are immediately apprehensible and apodictic.  But there is definite disagreement about whether or not there is knowledge of this type.  The problem with axioms is that it can be argued that if they are not intuitively obvious, then they must be arbitary.  But I like your idea that they can be adopted provisionally.  But then they are not truly foundational in the sense that they are only provisionally true.

[xSilverPhinx]

I'd say that it's a mix between foundationalism and coherentism. I don't particulary like infinitism because it really doesn't take long for responses to get purely speculative and so rather pointless in my point of view.

[Twentythree]

Foundationalism doesn't seem right. In a topic I posted a while back I brought up the idea that truth of any kind is subjective depending on your personal view of reality. Atheist and theist believe vastly different things when it comes to the nature of reality therefore anything that looks obvious or self justifying will depend entirely on the view of reality held by the individual.

Coherentism doesn't seem right either as there could be many rational causes for any particular fact. Following the chain of cause and effect for any fact would seem to expand in ever more complex webs rather than neat coils. That would leave me to think Infinitism makes the most sense. I experienced something like this  recently in a discussion with my good pal Davin. I tried to make a general point about the hypocrisy of Christianity and eventually ended up realizing that my actual point was that everyone is responsible for everything. I'm no expert but sometimes I gotta go with what personal experience has taught me.

[The Black Jester]

Foundationalism doesn't seem right. In a topic I posted a while back I brought up the idea that truth of any kind is subjective depending on your personal view of reality. Atheist and theist believe vastly different things when it comes to the nature of reality therefore anything that looks obvious or self justifying will depend entirely on the view of reality held by the individual.

Coherentism doesn't seem right either as there could be many rational causes for any particular fact. Following the chain of cause and effect for any fact would seem to expand in ever more complex webs rather than neat coils. That would leave me to think Infinitism makes the most sense. I experienced something like this  recently in a discussion with my good pal Davin. I tried to make a general point about the hypocrisy of Christianity and eventually ended up realizing that my actual point was that everyone is responsible for everything. I’m no expert but sometimes I gotta go with what personal experience has taught me.

This is a very interesting point regarding 'intuitive truths.'  It is something that I have often remarked on with regard to a priori reasoning and necessary truth.  Those philosophical precepts rest heavily on what is intuitively conceivable, but we have no way of knowing if such conceivability is the result of privaleged access to a metaphysical reality or merely habitual patterns of thought (my strong suspicion is the latter, in many cases).

The problem I have with infinitism is that, while theoretically it is possible to give an account of any step in the chain of reasons upon demand, it is not possible to justify the entire chain of beliefs itself, since the chain is infinite, there will always be more beliefs than we can practically account for.

[Whitney]

I think that those ideas which are thought to be obviously true (foundationalism) were actually figured out via trial and error and simply appear to be matter of fact to us because of their tried and true nature.

Kinda like how life looks like it could be designed but those who really understand know it evolved...we just don't know through what steps the foundational concepts evolved to what they are.

[penfold]

On Brute Facts by G E M Anscombe [Analysis 18, 3 (1958)]

Essentially,

A is true if brute fact(s) B hold and circumstances are normal

eg the 'water in the kettle is hot' if 'the kettle has just been boiled' and circumstances are normal.

We can say that the fact of 'the kettle having just been boiled' stands brute relative to the fact of 'the water being hot'.

The fact of 'the kettle have just been boiled' will itself be true reliant upon further brute facts (such as me plugging it in and switching it on) in normal circumstances (ie the fuse has not blown, there wasn't a powercut, I was not hallucinating etc...)

This approach is certainly not foundational. One could claim it to be infinitist or coherentist, it would be a monumental challenge to try and discern which.


The advantages of Anscombe's approach as I see it is are as follows:

(1) It matches are experience; when asked how we know a fact to be true we tend to refer to other facts.

(2) I can know things without having to work from first principles. This means that I can know things without having to do all the leg-work. So I can know that the force of gravity between two objected as being inversely proportional to the square of their distance apart; without taking any measurements. Essentially I know this relative to the brute fact of being told so by scientists, who, in normal circumstances, are correct about such things (after all they have taken measurements).

(3) The criterion truth is loosened. this brings seemingly normative statements as "It is wrong to do x" into the epistemic arena (conta Hume). eg "It is wrong to punish someone" is true relative to the brute facts "they are innocent" and "their innocence is known", in normal circumstances. No foundational ethical discussion is required.

The obvious disadvantage of Anscombe's approach is that truth is relative. Ah well, can't have everything.