Disputation and Burden of Proof: A Round with Professor Novak

I proposed for consideration a bit of dialog:

A: The law of noncontradiction (LNC) is a law of thought merely.

B: I dispute your claim. LNC is not a law of thought merely; it is also a law of extramental reality.

In this example, B disputes what A says by making a counter-claim, a counter-assertion. Both are asserting. It strikes me as foolish to ask who has the burden of proof (BOP). How decide such a question?  I assume that in a dialectical situation like the  above, if BOP considerations are relevant at all, then the BOP is on one side or the other, but not on both, and not on neither.  But there is no non-arbitrary way to place the onus probandi on one side or the other.  Therefore, BOP considerations are a useless detour.  Why not go straight to the question and evaluate the arguments pro et contra?

Suppose you say that the BOP rests on the one who opposes the received or traditional view. Then the BOP would be on A. But if you say that the BOP rests on the one who makes the stronger claim, the more committal claim, then the BOP would be on B.  I don't see how there could be a non-arbitrary assignment of BOP in a dialectical situation like this.  Correlatively, I don't see how it could be non-arbitrarily claimed that there is a defeasible presumption (DP) in favor of A's assertion or of B's. So I suggest we drop the BOP talk!

Lukas Novak's response:

Concerning your dialogue: In my opinion, both A and B bear a burden of proof here. For that reason, it is an unlucky start of a dispute - because it is in fact the  start of two disputes at once, and a dialectical confusion is likely to arise. In order that the dialogue be fruitful, B should not have put forward a negation of A's claim as his own claim, but simply refuse to accept A's claim until proved (this is the meaning of the rule Necessitas probandi incumbit ei qui dicit non ei qui negat - "negare" here has the technical meaning of "to refuse to concede until proved", according to the rules of disputation). If A failed to produce a proof, his case would fail. If he produced one, his case would succeed unless and until B attacked that proof, thus prompting  another argument to "restore" the former one. And so on, until one of the parties failed to do their duty. Only if A was the one who so failed, would it be in place for B to state his opposite meaning as a claim, if he wishes, with the burden of proof incumbent on him

There are three, not two possible dialectical states of a proposition: (i) proved (ii) disproved (iii) neither. The "burden of proof" just means that the default state is (iii).

Perhaps our difference boils down to this: you think that a dispute is about truth or falsity of a proposition, whereas I think that it is about validity or invalidity of rational support of a proposition. Whereas from the former point of view the dialectical situation comes out as symmetric, in the latter view it is inherently asymmetric.

Reply to Novak

Part of our difference here may be due to a different understanding of 'dispute.'  I think Lukas may be using it is a technical way similarly as he uses negare in a technical way.  And perhaps these technical meanings are the same.  When I used 'dispute' in the little dialog above I was using it to mean 'disagree with.'  Lukas seems to be using it to mean 'refuse to concede until argument is provided.' 

Lukas seems just to be assuming that the BOP rests on A who must "produce a proof" otherwise his "case would fail." I take that to mean that A is obliged to give an argument for the claim he has made.  (In my book, an argument is not the same as a proof, although every proof is an argument.)  But, by my lights, if so, then the same goes for B: he too must give an argument for his counterclaim.  B cannot just cross his arms across his chest and say, "I don't have to give an argument for my assertion; it suffices for me to poke holes in your argument.  The BOP is on you, not on me."  This is precisely what I reject.  Otherwise, there would be a presumption in favor of B's claim. But there isn't.  And to insist that there is, is to beg a philosophical question.

I think Lukas is right when he says that, for me, the dialectical situation is symmetric, at least in the example given above, while for him it is asymmetric.

Lukas is also right when he says that, for me, the dispute (disagreement) is about the truth-value of a proposition: Is it true or is it false that LNC is a law of thought merely?  He says that, for him, the disagreement is "about validity or invalidity of rational support of a proposition."

But this needs explaining.  Validity and invalidity are technical concepts from formal logic.  Our present topic, however, is not formal logic, but dialectics.  Lukas seems to think that there are certain procedural rules that govern the conduct of a discussion, and that these rules induce certain rights and duties in the interlocutors.  Thus, he who makes an assertion puts himself under a dialectical obligation to support his assertion with one or more arguments, while the one to whom the assertion is made is under no obligation to support the negation of the asserted proposition: he has the right to do no more than find fault with the arguments for the asserted proposition.

I am skeptical of this entire adversarial model which has its provenience in the court-room situation and makes perfect sense there, but seems to me not appropriate in philosophy which, by my lights, is not a matter of debate or disputation but one of dialogue in which the interlocutors are not out to prove propositions they antecedently accept and do not question, but who aim at arriving at the truth together, a truth that they do not claim to possess, but are seeking.

See also:  Philosophy, Debate, and Dialog

Related articles

Are Burden-of-Proof Considerations Relevant in Philosophy?

Comments

[Lukas Novak]

Dear Bill,

that's an excellent summary of our respective positions, thanks. You say that you are skeptical of the adversary model, as

it has its provenience in the court-room situation and makes perfect sense there, but seems to me not appropriate in philosophy which, by my lights, is not a matter of debate or disputation but one of dialogue in which the interlocutors are not out to prove propositions they antecedently accept and do not question, but who aim at arriving at the truth together, a truth that they do not claim to possess, but are seeking.

Several points about that:

(1) I don't think the "adversarial model" (AM) originated in courtrooms. It is the backbone of scholastic disputation, which for several centuries was the main raitonal tool of teaching, research, and expression of ideas. This is the method that created the university.

(2) It seems to me that you confuse adversarial personal attitudes with adversarial nature of a problem. A philosophical problem is adversarial by its nature – truth and falsity are irreconcilable. So, IMO, the method correctly reflects this. It is entirely irrelevant to the method whether there are two interlocutors, or whether you just "argue with yourself", attempt a private analysis of an issue. Using this method does not imply any ardversariness on personal level, it does not exclude joint effort to attain truth. This is nicely witnessed by Leibniz in his letter to G. Wagner (ed. Gerhardt, VII, 522), who writes (abridged):

I once made this experience in a mathematical discussion with a very learned man: we both were seeking truth, and the letters we were exchanging were very polite. Still, each of us was constantly complaining that other party misrepresented his words. I suggested to employ syllogistic form, and my opponent agreed. We arrived at the twelfth prosyllogism, when suddenly all difficulties disappeared. We started to understand each other, to a great advantage of each of us. I have no doubt that if we did that more often – if we exchaged arguments and proper responses in forma – one could often get to the heart of the problem and remove many misconceptions...

(3) I disagree that a necessary condition of proper philsoophizing is abstinence from truth claims. If there was no hope to attain and certainly hold truth, at least in some rare cases, philosophy would be a ridiculous and absurd enterprise, unworthy of pursuing. So there is nothing un-philosophical on "holding a truth". And it does not, of course, exclude willingness to let oneself persuaded to the contrary - as long as one holds a truth simply because it is truth, he will have not motivation to stick to it when its verity gets seriously undermined in a discussion. Besides, I think it is impossible not to hold philosophical beliefs as true; so trying to pose onself as if will often lead to self-deception and philosophical blind-spots. Everyone has his antecedent truths, there is no greenfield site start in philosophy. The AM allows you to immediately handle these preconceptions and see whether they can be substantiated or not (you don't need a real adversary to do that, but it is always safer to have one - given the possibility of blind-spots etc.).

(4) That said, "claiming to posses a truth" is not required for the AM. The method is "adversarial" not because you can't help sticking to "your truth" so strongly, but in order that one could be sure, when the dispute arrives at certain result, that this result has been tested in the strongest possible way. So the adversariness is fact a service to truth, since the more unforgiving the parties (whether real interlocutors or just "parts of your mind") are during the dispute, the more solid the rational support of the result is. Truth matters!

[BV]

Thanks, Lukas.

Ad (1). But surely Roman law antedates medieval disputation by quite a few centuries. And as I understand it, the notion of *onus probandi* is at least as old as Roman law. There are some interesting historical questions here.

Ad (2). The adjective 'adversarial' derives from the noun 'adversary,' and there must be at least two, both of them persons. A conversation can be adversarial, but I would deny that a proposition and its logical contradictory stand in an adversarial relationship.

But I agree that there is a sense in which a person can argue with himself. But I don't agree that a person can assign a burden of proof to himself. I may ask myself: what arguments can be adduced in support of p, and what arguments can be adduced in support of ~p. But it would make no sense to put the BOP on myself.

As for the rest of what you say, I don't see its relevance to my main claim. My main claim is that, with respect to many philosophical questions, it makes no sense to assign a burden of proof to one side or the other. Thus, e.g., it makes no sense to say that the theist (atheist) bears the burden of proof, or that, correlatively, that there is a presumption in favor of atheism (theism).

[Lukáš Novák]

Dear Bill,

well, I think I agree that it makes no sense to assign a burden of proof to the theist and not the atheist, or vice versa. I agree that insofar as the burden of proof is correlated with a presumption, in most philosophical questions there is no difference in terms of burden of proof between the sides.

What I wanted to convey is that this does not imply that the burden of proof does not apply at all to most of philosophy - for I consider this presumptiom-related application of the notion secondary, invoked only by rare special circumstances (viz. the existence of a legitimate presumption). Primarily, it seems to me, the question of burden of proof arises from the analysis of any philosophical question not in terms of the possible opposite answers, but in terms possible epistemic attitudes to those answers: for whereas positive maintaining or believing or claiming any of these answers invokes a burden of proof - a dialectic or epistemic duty to provide support, of which the sanction is falling into dogmatism -, mere refusal to concede or believe etc. does not. All I want to say is that this distinction is relevant and in fact omnipresent in philosophical enquiry, and also that it is easily overlooked, and therefore the burden-of-proof disputes are not trivial: because it is quite easy to confuse lack of concession with positive denial.

Isn't the very circumstance that you insist on the bipartite optics ("theism or atheism") and fail to perceive the relevance of the quadripartite analysis ("should theism be maintained - yes or no? Should atheism be maintained - yes or no?"), where the question of burden of proof surfaces, evidence that correct assignment of the burden of proof is not a trivial question?

The presumption-related application of burden of proof within the bi-partite optics is only possible, meseems, because the rare existence of the legitimate presumption results in epistemic and dialectic equivalence of the otherwise to be distinguished pairs of alternatives (given that there is a legitimate presumption that X is true, it is as dialectically and epistemically legitimate to positively maintain X, as it is merely to refuse to maintain non-X), and so, exceptionally, the question of burden of proof gains meaning even on the level of the bipartite analysis. But the fact that mostly, on that analysis, burden-of-proof question is senseless, is not a mark of its merely exceptional validity in philosophy, but a mark of the insufficiency of the bipartite analysis with regard to that question.

[BV]

Lukas,

It looks as if you are using 'burden of proof' in a different way than I am. I use it as strictly correlative to 'defeasible presumption.' In the law it is very clear how 'BOP' is used, and my claim is that, with respect to many of the deepest phil. questions, it makes no sense to bring in considerations of BOP.

For example, it might occur to someone to claim that in a dispute between Meinongians and anti-Meinongians the BOP lies on the Meinongians. I would have no trouble at all refuting that notion.

[Lukas Novak]

Bill,

yes, it seems that we operate different notions of "burden of proof". Is that a mere quaestio de nomine then? I am not sure; I would like to claim that (i) my notion of BOP is systematically more basic, and (ii) (this is a tentative claim) it is the notion primarily employed in the philosophical tradition, underlying such traditional principles as the one you cited, or "quod gratis asseritur, gratis negatur" etc.

Perhaps your position would be better stated as "defeasible presumptions are rarely legitimate in philosophy"? It seems that it would be both less ambiguous and more to-the-root (as the existence of a defeasible presumption seems to explain the existence of BOP, not vice versa).