Find Enlightenment

OPAR, Chapter Five
by David Ross

Date: 3 Feb 1993
Forum: Moderated Discussion of Objectivist Philosophy
Copyright: David Ross

Peikoff begins this chapter with some definitions of reason, e.g. Ayn Rand's (p. 152):

"the faculty that identifies and integrates the material provided by man's senses"

and one of his own (p. 152):

"the faculty that organizes perceptual units in conceptual terms by following the principles of logic"

He explains that one cannot prove that reason is valid; since reason is the faculty of proof, any proof presumes the validity of reason. However, one can validate reason by identifying its relation to reality. The essential issues in this validation are the validity of the senses, of concepts, and of logic; he has covered these issues in earlier chapters.

In this chapter, he will cover two remaining questions: Is reason man's _only_ means of knowledge? and Is certainty possible? In answering these questions, he will differentiate reason and emotion, the logical and the arbitrary, and knowledge and omniscience.


Peikoff begins by telling us that an emotion is "a response to an object one perceives (or imagines), such as a man, an animal, or an event" (p.154). Objects cannot induce reactions themselves; the subject must provide two cognitive elements - identification and evaluation. In order to have an emotional response to something, one must know what the thing is (or at least have some idea or think that one knows what it is), and one must have some idea about its implications for one's life.

Emotions have intellectual causes. This, Peikoff explains, differentiates emotions from sensations. Like emotions, sensations are automatic. But unlike emotions, sensations do not depend on one's ideas.


Peikoff supports the premise advertised in this section's title with two examples. In the first, six men look at some medical slides projected on a screen. Each has a different emotional response. The differences are based on their different understandings of what the slides are (One man understands fully their scientific meaning, another is reminded of canvases by Kandinsky) and by the implications of the slides for each (two of the men understand what the slides mean scientifically, yet respond differently from one another; one recognizes that the slides signal the imminent death of a friend, the other recognizes them to provide proof of a biological theory that he has proposed.)

In the second example, Peikoff recounts springing a surprise quiz on his students; most students were distressed, because they expected to fail such a quiz. Auditors were indifferent.


There are four steps in the experience of an emotion: perception, identification, evaluation, and response. The middle two steps are performed by automatized cognitive functions, and so are experienced unconsciously. Nevertheless, these evaluations are based on one's view of the world, on one's implicit philosophy. If the ideas that one holds unconsciously contradict one's conscious beliefs, one may experience emotions that conflict with one's conscious evaluations.

Mind-Body Dichotomy

Thus, all conflicts between one's emotions and ideas are really conflicts between ideas - conscious and unconscious. The opposite view - that emotions are fundamentally independent of ideas and that conflict between the two is virtually inevitable - is prevalent in our culture. Its source is the philosophical doctrine of the mind-body dichotomy, the view that the mind (ideas) and the body (emotions) are radically different aspects of humans. According to Peikoff, the historical source of this idea is Plato's distinction between the world of forms and this world. More generally, however, its root may be any sundering of reason and reality.


In this section, Peikoff discusses the fact that emotions are not tools of cognition. He considers that this fact supports the claim that reason is man's only means of knowledge because emotion is the only other means ever proposed.

Peikoff's argument here is that a means of cognition must be self-regulating, it must be able to direct its own actions, and it must have independent access to facts, it must be able to check its conclusions. Reason qualifies - it is volitional, and it is precisely the faculty that identifies and integrates sense data - whereas emotion does not.


In this section, Peikoff discusses the Objectivist opposition to arbitrary propositions. I find Peikoff's discussion somewhat confused, so this section of my essay will be a criticism of this part of his book.

I would say that an arbitrary claim is one that is put forth without evidence. To make such a claim, and to regard it as viable unless evidence against it is produced, is to commit the fallacy of argumentum ad ignorantium. For example, consider the claim "The fellow we know as Eyal Mozes is actually a criminal hiding from the FBI, living under an alias and pretending to be a grad student". I have no evidence for this claim, so if I were to make it, it would be arbitrary. My making the claim places the burden of proof on me, and I must discharge this burden to place the claim "into play" cognitively.

Consider now, another 'claim': "Grimps of various heights are influencing my reading ability". This 'claim' commits what I shall call - following the usage of certain logicians [see, e.g., Hodges, (1977)] - referential failure. The essential problem with this would-be claim is that it contains the symbol 'Grimps', which has some of the non-essential attributes of words - it is composed of letters, which we can translate into sounds, and these sounds fit rather convincingly into the structure of an otherwise valid sentence, in the place where grammar dictates that a subject belongs - but it lacks the essential attribute, that of being associated with a concept. "Grimp" does not denote anything, it does not refer to anything. My "claim" is not a claim at all.

Peikoff does not seem to distinguish between these two types of utterances. He seems to lump them indiscriminately into one category, which he refers to as 'arbitrary'. According to Peikoff, such utterances have the following properties (among others):

They are based on emotion. (p. 163)

They have no relation to man's means of knowledge. (p. 164)

They are detached from the realm of evidence. (p.164)

No process of logic can assess them. (p. 164)

Previous knowledge is irrelevant to them. (p. 164)

None of the concepts used to described human knowledge can be applied to them (p. 165).

They are neither true nor false (p. 165).

Peikoff provides some examples:

Gremlins are studying Hegel's _Logic_ on Venus.

Your fate is determined by your birth on the cusp of Capricorn and Aquarius.

There is a God.

My first objection to Peikoff's presentation is that it seems to exclude my claim about Eyal's criminal career from the classification 'arbitrary'. (I'd also say that it excludes his 'fate' example, assuming that by 'fate' is meant the course of one's life with no mystical implications). My claim has no evidence for it, but it is not 'detached from the realm of evidence'. It is certainly something to which previous knowledge is relevant - anyone who knows Eyal personally knows that my claim is ridiculous, absolutely inconsistent with his character - and at least one reader of this essay - Eyal himself - knows with certainty that the claim is false.

Thus, Peikoff's classification seems to omit many of the claims that I would consider arbitrary. These are _important_ claims, e.g. they are the claims whose invalidity is behind the legal principle of 'innocent until proven guilty'. In connection with this, it seems worth noting Peikoff's queer formulation of the onus of proof principle (p. 167);

If a person asserts _that a certain entity exists_, he is required to adduce evidence supporting his claim.

This is a very restricted form of the principle. Claims about states of affairs, the occurence of actions, the possession of attributes - _all_ claims, not just those about the existence of entities - carry with them the onus of proof. It seems to me that Peikoff's odd emphasis on the existence of entities is consistent with his conflation of referential failure and the fallacy of ignorantium.

My second objection is that Peikoff's presentation seems to miss the mark regarding an individual's responsibility for the meaning of his or her utterances. The fundamental problem with utterances about grimps or gods or gremlins is not that they are unsupported, but that they are meaningless. An arbitrary claim does not warrant consideration, but it is at least considerable; a meaningless utterance is not considerable.

Peikoff seems to address this in saying that certain arbitrary claims can be 'transferred to a cognitive context', but his examples are confusing. He gives the example of the arithmetic sum mouthed by a savage (e.g., a chemist) versus the same sum stated by someone who understands the reason behind it (e.g., a mathematician). But there is a middle ground between these two, the person who understands the meaning of the sum without understanding the reasons behind it (e.g. an engineer). In the case of a sum, this is unlikely, but in the case of more complicated mathematical formulas, this is frequently the case. Unproved claims can certainly be true - I often provide engineers and chemists with formulas for which they do not know the proofs, and which they use successfully in their work. To regard these as _untrue_ seems ridiculous - and I'd say that in the limiting case, even a meaningful claim for which _no_ evidence has been adduced can be true.

Peikoff refers to arbitrary statements again in the next section, in connection with the evidential continuum. He explains that, in a murder case, if the objective criteria for proving guilt have been met, a defense attorney could not save the defendant by "uttering a string of _arbitrary_ 'maybe's' such as (p. 179)

Maybe the gun seller lied
maybe the fingerprints are a frameup ..."

If Peikoff intends the use of the word 'arbitrary' here literally, in the sense that he has been discussing, then these examples contradict his claim that such statements are neither true nor false. The gun seller - presumably a witness at the trial - either lied or he didn't. The fingerprints are either genuine or they are not.

In summary: Peikoff seems to equate arbitrary claims and meaningless utterances. As a result, he concludes that arbitrary claims are neither true nor false. I disagree with this equation, and I think that Peikoff's characterization of the arbitrary is at odds with standard usage. I think that certain of the utterances that he addresses commit referential failure, and are thus meaningless and neither true nor false. For these, evidence is not an issue. What I'd consider an arbitrary claim - one that's put forward without evidence - may be meaningful, and thus may be true or false.

There is more to be said here, but I expect we'd best leave it to the discussion.


In this section, Peikoff discusses the contextual nature of knowledge. The basic point is that humans are not omniscient, so they cannot properly consider omniscience the standard of certainty; Just because you don't know everything, you can't conclude that you don't know anything. Moreover, since human knowledge is limited, awareness of the pertinent limits is an aspect of the validation of any idea; you must be careful not to apply your conclusions beyond the context in which the evidence justifies their application.

Peikoff gives his usual example of the blood types. I'll supply another standard example, Newtonian mechanics. If Newton had claimed that his laws would apply at any speed, no matter how great, he would have been unjustified (and, as we now know, wrong); having had no experience with near-luminal speeds, he had no basis on which to make such a claim. However, within the context of low speeds in which his laws were formulated, they were justified and they are true.

Peikoff goes on to discuss the evidential continuum, i.e. to discuss the use of the terms "possible" "probable" and "certain" to describe the logical status of ideas. We discussed this topic on the list not long ago, so I'll make this summary short. An idea is possible "if there is some, but not much evidence for it, and nothing known that contradicts it" (p. 176). An idea is probable "if the burden of a substantial body of evidence, although still inconclusive, supports it" (p. 178). An idea is certain "when the evidence in its favor is conclusive, i.e. when it has been logically validated" (p. 179).

These points are clear and important. Peikoff has presented them, however, only in the most general, sketchy way. He admits this at the end of the section, and he promises to treat certain of the relevant topics (e.g. induction) in more detail in another work. To understand the hard technical details behind Peikoff's sketch, we would have to understand his conception of "logic"; it seems that whenever we arrive at a point in the discussion where some hard theory is required, Peikoff tells us that this is provided by "logic". But he tells us little about just what it is that "logic" says on these topics. For example, his discussion of the evidential continuum _presumes_ that we have objectively defined standards of logical validation. But of course, the establishment of such standards is the difficult and interesting aspect of this topic - sorting out the terms "probable" "possible" and "certain" is a minor (but useful) contribution.

Statements such as this one, (p. 171):

"Logical processing of an idea within a specific context of knowledge is necessary and sufficient to establish an idea's truth"

make me all the more anxious to have Peikoff flesh out the details of his conception of logic. We might interpret Peikoff's use of the term "sufficient" here as stressing that no means other than logic are required. But this seems the less likely interpretation in the context, the more likely being that logic's being sufficient means that logical processing guarantees arrival at the truth. At this level of generality, I have no objection to his claim, though I am leery. However, in conjunction with his statement (p. 177):

"... we do know in principle how to demonstrate a murderer's guilt; in this case, the requirements of logical proof have been objectively defined"

(in the course of the example, he presents these standards), the earlier claim seems queer. Is Peikoff claiming that meeting this standard of proof guarantees the truth of the verdict? So it seems. But this is simply not the case, an innocent person _could_ be convicted by this standard; rationality is not infallibility. (In connection with this topic, see some of David Kelley's comments in his review of OPAR (1992)).

Here, it seems that Peikoff is using "truth" to denote satisfaction of some abstract logical criteria rather than correspondence with fact. (In connection with this, I must mention the notion of "truth relative to a mind", which Peikoff introduced in his essay Fact and Value (1989).) I'd like to analyze this more, in particular I'd like to look into its connection with his view on arbitrary claims. But this essay is already long overdue :-)


Peikoff concludes this chapter by discussing two broad classes of irrationalists, "mystics" and "skeptics". Mystics claim that there is some means other than reason for acquiring knowledge, e.g. faith or revelation. Since there is no such means, mysticism amounts to a type of emotionalism. Skeptics claim that knowledge is impossible to man by any means. This theory gives the skeptic license to reject any idea he feels like, and is thus again de facto emotionalism.

Mysticism is an expression of intrinsicism, skepticism an expression of subjectivism.

Both mystics and skeptics rely on the arbitrary, according to Peikoff. The mystics embrace ideas (e.g. God) arbitrarily, and the skeptics reject them arbitrarily. Just as a rational investigation of mystical phenomena would defeat the mystics' purpose, so a rational commitment to doubt would defeat the skeptics'; the mystics would find themselves deprived by such an endeavor of their omniscient, omnipotent God, and the skeptics would find themselves committed to a rational process of validation, and thus to the possibility of knowledge.


David Kelley, Peikoff's Summa, IOS Journal, Vol.1, #3, Spring 1992.

Wilfred Hodges, Logic, Penguin Books, Harmondsworth, Middlesex England 1977.

Leonard Peikoff, Fact and Value, The Intellectual Activist, Vol. 5, #1, 1989.

Rich Dempsey, Ilia Levy, and Carolyn Ray discussed these topics with me, and I thank them.

Find Enlightenment at