Criticism of Epistemological Ideas in Peikoff's Logic Course - Part 2
More criticism of Peikoff's epistemology
Table of Contents
I've been going through Peikoff's Introduction to Logic Course (the videos for which are available on YouTube. The handout with the exercises is freely available on the Ayn Rand Institute's website for the course. I'll be posting some exercises I've done in the course soon, but my purpose for this post is to criticize Peikoff's advocacy of mistaken methods of dealing with ideas. In doing so, I will be drawing very heavily on ideas and arguments from Elliot Temple's Yes or No Philosophy course, along with my background knowledge of other relevant thinkers like Karl Popper. This post draws in particular on the "Support Contradicts Logic" and "Burden of Proof" videos in Yes or No Philosophy. However, the analysis is my own, as are any errors. This is part 2 of a series (which I've given a more general name in this course because I think disagreements with Peikoff's epistemology as presented in this course are going to be a running theme that comes up repeatedly and which I'll want to write about repeatedly).
Example of Criticizing an Arbitrary Claim: Diseases
Continuing on from the previous post, let's hear more from Peikoff on the supposed fallacy of Argumentum ad Ignorantiam (my transcript throughout):
The structure of [Argumentum Ad Ignorantiam] would be: X is true (in this case, you committed a murder) because you can't disprove it. You get this name, the appeal to ignorance, because you are appealing to a person's ignorance, his inability to disprove something, his ignorance of how to disprove it, as the means of proving it.
Now you should be able to see easily enough, that if this were permitted, the most bizarre fantasies would be established simply by virtue of the impossibility of refuting them if they're put forth arbitrarily. I can point to someone and say "I feel terribly sorry for you because of the awful disease which you have." And the person says, "Well, what do you mean, I've just been to a doctor, got a clean bill of health" and I say "Yes, I know, but the terrible thing about this disease is that it hasn't been discovered by medical science yet. The worst thing about it is you feel perfectly healthy and then you go just like that. Prove that you don't have that disease. Now if I give you no argument as to why you do have it it is impossible to prove that you do not have such a disease. It cannot be done.
Peikoff wants such a claim to not be permitted – by both dismissing it as arbitrary and by rejecting subsidiary claims as fallacious. As in my previous post, I think there are better grounds for criticizing the claim than this (even though the claim is very underspecified, which, by the way, would itself would be a criticism of the claim):
- One could ask the person why medical science is unaware of this supposed disease and why they think they have bested all of medical science. This would involve a discussion of things like method and resources. E.g. they should explain why medical science's approach to discovering diseases has failed and what methodological improvements they have made, what their background knowledge and resources for understanding and analyzing diseases is, etc. If they can't answer these questions or have no explanation, you can ask why they think they've bested all of medical science in identifying some disease. (Many responses to this will involve some sort of conspiracy theory, so you can use standard arguments for addressing conspiracy theories).
- One could point out that by a similar method one can claim that anyone has any arbitrary disease. But it's clear that people have specific diseases and not just whatever stuff people can imagine. This failure of the method to correspond to the apparent reality of the situation is a major problem that needs to be addressed.
- One could apply a similar method to the person making the claims. E.g. you could claim that they have a mental disease which causes them to imagine that other people have unknown diseases. The fact that applying the same method to the person's own claims leads to a negation of the claims illustrates a major problem with the person's method: that it cannot be used to decisively settle disputes and reach the truth.
- One could ask them for details about things like how the disease works, why they think you have it, and how they ruled out competing explanations (like not having the disease, the disease not existing, someone having a cold instead of the disease, etc.)
Peikoff sees a real problem here – how to deal with certain classes of problematic claims – but his methods are flawed. He's worried that some absurd claims would be "established" if we don't prohibit them via means like categorizing them as arbitrary and rejecting claims that follow from them as fallacious. But I think the better approach is to take a step back and a step up to a higher level of discussion – to focus on methods and flaws in the approach of the person you're criticizing, rather than (IMHO irrationally) dismissing their claims on as not meeting a (misconceived) burden.
An Invisible Green Gremlin
And I can multiply fantasies truly without end. I can say at the back of the room there is at present a green gremlin, taking notes, listening, swallowing tutti-frutti ice cream. Prove that there isn't. You say you don't see it. I say this is a very small gremlin beyond the most powerful microscopes so you can't see it. You say I don't hear anything. I say it's an inaudible gremlin. You say how do you know that it exists. I say I don't have any proof that he does exist but you don't have any proof that he does not exist. Now prove that he doesn't. You cannot do it. If you then say I can't disprove it, therefore there is a gremlin at the back of the room, that is argumentum ad ignorantiam.
Criticizing the Gremlin
My questions/criticisms regarding the gremlin: (I am building up quite a library for these sorts of claims):
- Is there a reason for every particular fact being claimed? For instance, why a green gremlin and not a blue gremlin? Why tutti-frutti ice cream and not Dipping Dots? Why a gremlin and not a troll? If so, what are the reasons? If not, it would seem that things are being claimed according to something other than a process of reason (e.g. according to whim or a flight of fancy), in which case it's not clear why the claims should be taken seriously (given that anything can be claimed according to whim or a flight of fancy).
- How are the apparent internal contradictions resolved? E.g. How was it it determined that the gremlin is green and eating tutti-frutti ice cream if it is invisible? And is not just the gremlin invisible but the ice cream as well? How the hell does that work?
- Given that the gremlin appears to have no effect on reality and explains no phenomena, what problem does it solve? Why should I care about something with no effect or consequences?
- I could claim that the gremlin advocate has a brain disease which is causing him to hallucinate a gremlin eating tutti-frutti ice cream and ask him to prove that he doesn't have such a disease. This illustrates the futility of this method (since the same method used to put forth the gremlin claims can be used to negate it, and thus the method is useless for truth-seeking).
Positive & Negative Statements
Now some years ago, when I used to give the green gremlin example, I thought to myself, well, this is a little far-fetched, but it illustrates the point. Today this is a naturalistic report of the method by which people argue with the profusion of claims for every kind of occult supernaturalist influence, parapsychology, poltergeists, demons, and a national wave of enthusiasm for exorcism, demonology and demoniacal possession etc., all of which is in the same category as the gremlin, logically.
Now what is the essential error here? The first thing to observe is that it is impossible to disprove an arbitrary assertion that something exists. Let me repeat that. It is impossible to disprove an arbitrary assertion that something exists. Now the assertion that something exists is called a logically positive statement. A positive statement simply means the claim that something exists. The corresponding negative would be the statement that it does not exist. Now the point to see here is that it is impossible to disprove an arbitrary positive statement. You cannot directly prove that X, whatever it is, doesn't exist unless evidence is put forth to prove that it does.
I think Peikoff is fundamentally wrong here. Both positive and negative statements imply that things do and don't exist, and so you can't treat them as being fundamentally different epistemologically.
For example, suppose you say that "A chair exists directly in front of my desk." Peikoff would classify that as a positive statement – let's call it statement X. But X also implies the statement "An expanse of air unobstructed by any furniture does not exist directly in front of my desk", which i will call statement Y. Y is, per Peikoff's categorizations, a claim about something that does not exist. But Y was implied by X. Peikoff wants to treat one of these statements, Y, as an epistemological second class citizen. He wants to do this even though Y was implied by X and even though it's doing, fundamentally, the same thing as X (i.e. describing a particular state of affairs in reality). Why? Because he thinks that if you don't do this, you open the door to all sorts of arbitrary crap. As I've been trying to illustrate in going carefully through Peikoff's examples, I think there is a better way to address the problem Peikoff is concerned about.
Insufficiency of Burden of Proof to Deal With Ad Hoc Claims
Take the gremlin case. Suppose the following: the person puts forth evidence, no matter how lame, but he tries to put forth evidence. He says, well, look, can't you see that green shimmering, in the air back there. That's my argument for the gremlin. Well, now he gives you something to go on. And you are then free to offer a different interpretation and by that means to refute his statement. You refute it by showing that the evidence he claims doesn't in fact prove it. So you say for instance, well the reason is that there's a radiator there, and it's the heat waves, and the reason it looks green is because the light bulbs have a greenish cast, and therefore I can explain the phenomena without reference to the gremlin, and then i've refuted the gremlin. But I did not refute it directly, I did not prove there is no gremlin out of a void. I simply attached myself to the positive evidence that the person claimed and showed that it had been misinterpreted and in that way you can refute a statement.
I'm a little confused as to Peikoff's discussion here. I think the method Peikoff describes – refuting the specific claims the person is raising – is perfectly valid and correct (at least earlier on in a discussion, before a pattern has emerged which indicates a problem). But Peikoff has been very concerned about the possibility of people raising arbitrary claims which they amend ad hoc to avoid criticism. So what prevents a person from making ad hoc changes to theories when they've brought up specific factual evidence? For example, conspiracy theorists bring up specific factual claims all the time and then just move on, unfazed, with minor changes if you call them out on their facts. And earlier, we saw Peikoff's example of someone claiming that you hired an actor to take your place in the hospital while you committed a murder. That's definitely a specific factual claim, and yet Peikoff seemed to think that you couldn't address that just by getting into the factual details (and I agreed with him, though I disagree with his attempted resolution). So I'm unclear why he thinks this case warrants different treatment and doesn't raise the same concerns about ad hoc changes that have been driving his whole analysis.
Essentially, my point is this: even if someone meets an initial factual "burden of proof", which is trivial to do, they can then switch to the mode of making ad hoc claims. So the burden of proof doesn't actually do a good job of policing the ad hoc claims issue. The whole point of the burden of proof is that it's supposed to be a shortcut way to deal with arbitrary crap. If it can't actually do that job, what's the point or advantage versus just engaging in critical discussion?
Evidence That Follows from Non-Existence & Resolving a Dispute
But if the person simply says, "There is a gremlin. Prove there isn't," without pointing to any facts in support of his statement, what he is saying is "Give me positive evidence of the non-existence of the gremlin. Give me positive data which follows from the non-existence of the gremlin." Now that is an assignment which cannot be fulfilled. It cannot be fulfilled for the most fundamental metaphysical reason that if the gremlin doesn't exist, it's nothing. Completely nothing. Zero. And as such it has no special effects, no consequences, it doesn't give off any negative aura, it doesn't give off any signs. There are no results in reality of a non gremlin. You couldn't point to a certain set of facts and say: these are the consequences of the absence of gremlins. Nothing is nothing. And that's all you can say: if it's not there, it's not there.
It's completely trivial to offer positive data which follows from the non-existence of a gremlin, in the sense of being consistent with a theory about the non-existence of a gremlin. For instance, the gremlin doesn't register in human vision, won't show up on regular or infrared cameras, can't be heard by humans, won't make sounds which can be recorded by iPhones, etc. This evidence all follows from the non-existence of the gremlin in the sense of being compatible with that theory. But then a person claims there is a green gremlin there eating tutti-frutti ice cream and the gremlin is just invisible. How do you resolve that? The evidence of your eyes (and iPhones) is compatible with both theories. The evidence doesn't solve the dispute. You need to use criticism, argument, creative thinking, and discussion in order to resolve that dispute. In criticizing the theories, everything is on the table, including issues of whether someone is arriving at the theories by rational or whimsical methods, is being biased, is vulnerable to the negation of their theory by the application of the same method they used, is internally inconsistent, has a specific problem to which their theory is the solution, and so on. You're not just limited to waiting for the other person to put forth claims which you then respond to in tit-for-tat fashion.
Onus of Proof Principle
Now, in such a case, what then would be evidence of the non gremlin? There is no positive evidence of the non-gremlin. The total absence of evidence for the gremlin is all that is required to say that it doesn't exist. The absence of any evidence for the gremlin is sufficient to justify the statement that there is no gremlin. Now if you understand, this you will understand the central principle that is involved in not committing this fallacy. That is known as the onus of proof prints and it states: the onus of proof is on him who asserts the positive. The onus of proof is on him who asserts the positive. Now, by positive in this statement, we are speaking logical in terms of the content of the meaning of the statement, not grammatically. For instance the statement "he is innocent" and the statement "he is guilty" are both grammatical positives, and yet obviously after those two which is the one that claimed something happened and which is the one that claims it did not happen? The statement "he is guilty" is the positive in this connection. He did commit a certain crime. The statement "he is innocent" is the negative. It is the statement he did not commit this particular crime.
Ad Hoc Claims
If I claimed that Peikoff's dinner (which, let's stipulate, appeared to be in front of him) didn't exist, would he think that he has the burden of proving that it does, as the person making a positive claim? Perhaps Peikoff would say yes and think it's trivial to meet that burden. But if I'm free to make arbitrary ad hoc changes to my claim that his dinner is illusory, then he's in a pickle, and his burden of proof idea doesn't protect him here.
Positive & Negative Statements
"He is innocent" positively implies that the purportedly innocent person did some range of activities other than e.g. kill someone at a particular time on a particular day. It's a claim compatible with situations in which the accused person was e.g. sleeping or eating or traveling at the time the crime happened. It's saying that the state of affairs in the world included only things in the set that involved the non-criminal activities and excluded (at least some) criminal activities. It implies that some stuff happened (such as the accused sleeping, eating, traveling, watching TV, whatever) instead of some other stuff (such as the accused shooting, stabbing, or bludgeoning the victim) in the context of the interactions between particular people. Both "he is innocent" and "he is guilty" are making claims about what reality was like (at least to the point of specifying that certain relevant facts fell within a certain range of activities and not within another range, as described above). They're statements that are talking about the world and saying what it was like. So there's no fundamental epistemological difference between the two statements and they should be analyzed with the same methods (i.e. critical discussion).
Rejecting Peikoff's Approach
Peikoff's idea of evidence is misconceived. A given piece of evidence (such as a video recording that shows no gremlin at a particular location) is compatible with various theories (such as no gremlin and an invisible gremlin). The logical relationship between evidence and theories which the evidence does not contradict is the same for all such theories. In other words, for each theory that a piece of evidence does not contradict, we can say that the theory is not contradicted by the evidence, and that's it. Any such non-contradicted theories, no matter how outlandish they may seem to our intuition, stand in the same logical relationship to the evidence. Both "no gremlin" and "invisible gremlin" are not contradicted by the evidence of an iPhone video that does not show a gremlin. So we need something else to settle the question of which theory we accept and which we reject regarding the evidence. My solution is to use open-ended critical discussion and to bring up issues of method like irrationality, bias, and various other issues.
But Peikoff disagrees. And so, proceeding on his premises about the nature of evidence, he argues that there are two classes of claims with different epistemological statuses. Positive arguments, which assert that things exist, require that you offer some evidence in "support". Negative arguments, which claim that something doesn't exist, hold the field unless positive support is offered. But both positive and negative arguments make some claims about the world. They both reject some states and affirm some states. The statement "There is a chair in front of my desk" does two things. First, it positively asserts the existence of a chair. Second, it implies that there isn't an unbroken expense of air in front of the desk (because an unbroken expanse of air would be incompatible with a chair being in the same position). Likewise, saying a chair isn't in front of your desk (a "negative" statement) has "positive" implications about the state of the space in front of your desk. The situation might be that there is no chair there in general (because you use the desk exclusively as a standing desk), or that you need to buy a chair, or that you use a medicine ball instead of a chair. Regardless of the details, the "negative statement" regarding the non-existence of the chair positively implies that the situation regarding the chair falls into the range of possibilities consistent with not having a chair there. So positive and negative statements can't be neatly separated and divided into epistemologically significant categories. They both make statements about the world and they both have implications in the "opposite" direction (positive implications for negative statements, and negative implications for positive statements). Given that fact, I don't see a reason to treat them as epistemologically significant categories which require different methods or burdens of proof. Instead, I think they should be dealt with using the same methods.
Given that preceding analysis, I tentatively reject Peikoff's approach to defining the fallacy of argumentum ad ignorantiam. I reject it not because I think "X is true ... because you can't disprove it." is a good argument, but because I think the main issue with it is the concept of proof/evidence it takes for granted.