Agrippa’s Trilemma posits that all attempts at justifying any given belief fall prey to one of three (informal) fallacies:
Circularity;
Dogma; or
Infinite regression.
All three, according to this argument, are unjustified. For circularity presupposes the truth of the conclusion as a premise; dogma assumes the truth of a conclusion whilst openly accepting that it is without reason; and an infinite regression of reasons rests on every reason lacking intrinsic justification to confer to the others.
Self-refutation Objection
A rather obvious objection to this argument is that, paradoxically, the truth of it would lead to it’s falsity (because that very claim would be justified without falling prey to its own trilemma) or to our inability to be justified it is true in the first place (because if it is true in the fullest sense then it also falls prey to the trilemma).
IMHO, when a thought experiment leads to self-refutation or paradox, it’s probably due to a flaw in our thinking and not some peculiar aspect of reality that we have discovered.
False Trilemma
It is clear, given its own self-refutation, that the trilemma is false; but what other option is there? I think it lies in the fact that assuming the truth is different than intellectually grasping it in an immanent way: certain truths are so simple that they demonstration rests almost within itself, but these truths are not justified in virtue of merely assuming them as true. For example, the principle of non-contradiction cannot be properly proven; however, it can be demonstrated to be true sufficiently by showing exactly what we would expect if it were true: namely, that we cannot but use it to argue against it. This is justification for believing it: we are not merely assuming it as ‘dogma’.
I think the only viable option is foundationalism; i.e., that justification for beliefs (1) terminates (2) at a non-circular terminus and these kinds of foundational beliefs are justified by demonstrating how fundamental they are to reasoning itself.
Perhaps the whole point of the exercise is to demonstrate the insufficiency of propositional knowledge when viewed in isolation from actual existence. After all, when I turn the handle on the door, I know it will open. Of course a sceptic may say, well it might be inoperable or the door might be locked — which is true! But the answer to that is not another argument, but actually opening the door.
Nor do any of the items in an infinite regress stand unsupported.
If we are to have a conversation, we must take something as granted. Even doubt presumes a background of certainty. That’s the gem that sits inside the self-refutation objection.
I went to the cited Wiki page and noticed that the trilemma is posed a little differently there, as dealing with the impossibility of proving any truth.
Would you agree that justifying a truth and proving a truth are not the same?
I encounter this issue in atheism-theism debates. Some atheists are adamant that reason and faith are opposed and that theists are guilty of irrationality as faithism is to believe a proposition without evidence. More experienced theists point out that reason is not as perfect as atheists assume, pointing to Agrippa’s trilemma. Foundationalism is synonymous with taking some propositions on faith. Euclid’s elements is basically a showpiece for foundationalism.
Circularity may not be an issue. No idea why, but there’s been talk of virtuous circles. In a loose sense 2 honest people may vouch for each other’s reliability.
Infinite regress is impossible to actualize, but it’s an inevitable consequence of the rules of logic. Nullifies all proofs/arguments.
Skepticism however distinguishes between the evident (they don’t question this) and the non-evident (dogma, which they question). It’s not that the skeptic hits a wall; they go by appearances (phantasiai).
The only thing I can say to those who cannot accept that we must use the given in order to move forward with our arguments and narratives, and not get paralyzed, is to live life as a psychological or phenomenological existence instead of a universal reality.
Premise: A proposition which is used to justify a conclusion, and which itself requires further justification
First principle: A proposition which is used to justify a conclusion, and which requires no further justification
When it comes to inferential reasoning the options seem to be either 1) That there are no first principles (i.e. infinite regress of premises); or 2) that there are first principles. Vicious circularity is one kind of infinite regress, and “dogma” in that pejorative sense is not really relevant to justificatory reasoning.
In the Aristotelian traditions there is a way of grasping first principles (“epagoge”) through experience (“empeiria”). In rationalist traditions there is a way of grasping first principles a priori.
In the case of “Agrippa’s Trilemma,” the crucial proposition is the one which says that the three options form an exhaustive division. If that proposition can be justified as a premise or shown to be a first principle then the trilemma could stand. Yet if it is justified then the trilemma-proof is self-refuting, and if it is shown to be a first principle then there is at least one more option than the three on offer. So as you say, it won’t hold even if we manage to support that proposition.
How do you think the difference between “justification” and “proof” bears on Agrippa’s Trilemma and the OP? Do you have some reason to offer for why the distinction you are making is relevant?
I think that not all circles are vicious circles, so it is not clear to me that a circular explanation is always deficient. In particular, if we come to understand things and their relations at an ever greater level of clarity, then we have less of a circle and more of an expanding (or ascending if you will) spiral. So we circle back on our original points, but now with a greater level of clarity.
I would say the same thing about any non-discursive faculty of knowledge, intellectus/noesis (or "intellectual intuition), your preferred solution. These are essential to the phenomenology of understanding. I think that if thought that was only discursive rule-following, only the lower intellectual faculties of ratio/dianoia, it would essentially be just a Chinese Room. It could produce complex outputs given inputs, but it would lack any act ofunderstanding and all phenomenal quiddity. (I mentioned this in the thread “The Harder Problem of Quiddity.”)
But, I don’t think the reduction of noesis to the mere grasping of completely simple first principles is quite right. I would reject foundationalism of this sort, for it has many problems well documented since Descartes. Unfortunately, philosophical pedagogy has long skipped directly from Aristotle to Descartes, and the few medieval speciality programs tend to focus on late-medieval nominalism where intellectus is already being corrupted into an exclusively supernatural sort of non-discursive knowledge, rather than an essential faculty involved in all thought. I much prefer the tradition that emerges in Middle Platonism and extends until Aquinas and High Scholasticism, were cognition involves a constant movement between its discusrive and receptive moments, so that noesis is involved in all knowledge, not just unexplainable foundations. This allows the “circle” to elucidate the principles, since they are not exhausted by our first understanding of them.
What this means is that we aren’t just invoking noesis for simple first principles or abstracting terms (the “first act of the mind”) but at every step of the way, with each act of understanding. Aquinas uses a good analogy here, likening ratio to motion and intellectus to rest at the destination moved to, or as the difference between acquisition and possession. Boethius, Dionysius the Areopagite, and Maximos the Confessor all use this example of motion, but bring in the Problem of the One and the Many and the structure of principles as a unifying one that explains the disparate many (unity in multiplicity), the key example here being a principal as the center of a circle, and then the circle itself, which encompasses the many, being the discursive motion that “goes out” to encompass the many before returning to the one in a more fully realized form (epistrophe).
In terms of understanding the principle → cause → particular relation, obviously Aristotle is foundational, but I really love Proclus here and the Liber De Causis. Also Radek Chlup on Proclus because Proclus is hard (but still much clearer than Dionysius and Maximos, who might be better, but are harder to extract a clear theory from).
On this view, theories, models, language, signs, are all discursive means of knowing, being primarily how we know and not what we know. This seems to diffuse a great number of modern epistemic problems while keeping the Augustinian insights re knowledge as participation alive. We get past the Trilemma by allowing a new horn (non-discursive knowledge) and accepting the circular formulation, now modified, in concert with this.
The “infinite regress” isn’t totally wrong either. A thing’s nature is know through its actions, which includes its context. But its full context in the horizontal range is ultimately the entire cosmos, while in the vertical range it is higher and higher principles, up to the First Principle, which is infinite. So, we do not have an infinite regress obviously, but we do have the inexhaustibility of creatures, which means that any “circle” is never finished. This, I do not take to be a problem, however, since it hardly seems that we must know everything in order to know anything.
I would say that, if truth is the adequacy of the mind to being, then our minds can be more or less adequate, a point Thomas makes more formally with the notion that truth and falsity are not related as negation and affirmation (as in modern logic) but as contraries (like light and dark)—although we can still maintain that Law of the Excluded Middle and treat them as binaries within the domain of logic.
I agree that the absurdity of the possible conclusions that one must accept indicates a flaw in the logic as opposed to an issue with reality; but I think in this case the trilemma is flawed because it falsely claims to be an exhaustive list.
Having an assumed starting point for a discussion is not the same as holding justification ends at assumptions. E.g., we may have to assume for the sake of discussion that the multiplication is repeated addition when disputing how the derivative of 2n is 2; but that’s entirely different than assuming 1=1 as a fundamental axiom which cannot be supported by any justification for the sake of doing math. One is grounding the view in unjustified beliefs; and the other is assuming it is justified for the sake of the discussion.
I agree with your post: can you elaborate more on the differences between how Aristotle viewed it and rationalist traditions viewed it? I’m just curious.
A long chain of justification that is circular is just a vicious circle with extra steps—no? It’s fundamentally the same issue: there’s no grounding for the reasons given because they mutually justify.
I agree; but I’m having a hard time grasping how this counts against foundationalism. Could you please elaborate?
What issues does mine have, though?
This seems like coherentism—is that the view you are forwarding here?
I’m sort of thinking out loud, but it seems to me that we can justify a belief, and justify why we call something “true,” and neither of these activities is the same as proving a truth. So it depends how high you want to set the bar, doesn’t it?
My belief, for instance, that the sun will rise tomorrow doesn’t fall prey to any of the horns of the trilemma, as far as I can see. That’s because I’m not asserting it as a piece of certain knowledge, only an extremely well-founded belief. If called upon, I can explain why it’s well-founded, which requires reasons, but not an infinite regress of them, since my belief is basically just a version of IBE.
So I would call that “justifying a belief.” Likewise, I can explain why “There is a tree outside my window” is called “true.” The explanation doesn’t require that the statement be true, only that we understand why it is claimed as true. This too doesn’t involve us in any trilemma horns.
Finally there is the matter of proving something to be true, which may well involve us in the trilemma. But I think that’s because it invokes the idea of certainty, which is a special kind of justification, and hard to come by, especially in the non-analytic realm.
As I say, these are idle thoughts; I’d be happy to hear another viewpoint.
There’s a worthwhile comparison to be made here with what the Foot article has to say about falling; we suppose there to be a compulsion in a proof, such that like falling, we have no choice.
But we do have a choice. Even given the tightest argument, there remains the opportunity to reject the conclusion - at a cost.
There’s a good case for a new thread on what it is for a thing to have been proven.