In recent years, a certain logical fallacy has appeared in the writings of various prominent Libertarian and/or Objectivist theoreticians. It is the purpose of this essay[1] to explore the nature of this fallacy, "the fallacy of the frozen abstraction," to identify and analyze several instances of this fallacy, and to identify and validate the epistemological principle which this fallacy violates.

Phase I: Identifying the Nature of the Genus-Species "Freeze."

As defined by Ayn Rand, the fallacy of the frozen abstraction (or, in the language of Nixonomics[2], "the genus-species freeze") is a fallacy "which consists of substituting some one particular concrete for the wider abstract class to which it belongs."[3]

In other words, this fallacy entails the refusal to include certain members of a class in the wider class to which they belong, and instead limiting the class to one or a select few of its members.

The example used by Rand in introducing this fallacy is that of many people who have been taught to view morality strictly from the altruist standpoint. They have learned to equate altruism--which is one specific ethic--with the wider, more general abstraction of "ethics."

As a consequence, they refuse to regard egoism, hedonism, etc., as being alternative ethical systems or theories. Their concept of "morality," in other words, is frozen on the level of one of the species of morality, rather than being integrated to the higher, genus level, so as to include all of the species of morality.

As one might gather, this fallacy is singularly well-suited for propagating subtle (and not-so-subtle) untruths, particularly in the realm of normative (i.e., value) considerations. In committing the frozen abstraction fallacy, a given speaker substitutes his view of what a given thing ideally should be, for the wider class of what that thing has been, is, and can or should or will be. He then defines his concept of that thing so as to exclude all non-ideal, imperfect, or bad (evil and/or harmful) examples of that thing from the concept.

Perhaps the most fascinating historical example of this fallacy is Plato's theory of the nature of abstract ideas (Forms) themselves. Plato maintained that abstractions, or abstract ideas, actually exist apart from the concrete things in which they appear to be embodied and from the mind which seems to discover them. Abstract ideas, or Forms, exist in another, transcendental realm, separate from the world of our experience. They serve as models or patterns for the actual world and are somehow present in it. The world of our experience is merely the pale, imperfect reflection or "image" of the realities in the realm of the Forms.

Regarding these Forms (abstractions), Plato seemed torn between two quite different views. On the one hand, he felt that there must be Forms for all general terms. There must be Forms to serve as the model for every different kind of thing. There must be a perfect, ideal exemplar for each of the different types of thing existing in this imperfect, actual world.

On the other hand, it was very disturbing to Plato to entertain the possibility that there might very well be Forms for such "vile and paltry" things as hair, mud, dirt, etc.[4] These are undesirable--non-ideal, in an ethical or esthetic sense--as well as being merely imperfect, as are all material existents (non-ideal in a metaphysical sense). From this, Plato concludes (in a non sequitur) that "ideal" models of the undesirable could not possibly exist. This latter view seems to be the one that prevails in Plato's writings. It is the one that exemplifies the fallacy of the frozen abstraction.[5]

Plato's basic error is a confusion of metaphysics with ethics--more precisely, of imposing ethics upon metaphysics. He denies real existence (i.e., he denies that there is a real Form that corresponds) to those things of which he disapproves, according to his ethical or esthetic standards. He denies that there are metaphysically ideal (non-material or essential) Forms that are not also normatively ideal (good or beautiful). To admit that there were, after all, might besmirch or contaminate the "perfect, ideal" world of the Forms.

Regardless of the motive involved, this is the basic pattern and premise of all instances of the frozen abstraction fallacy as it occurs in a normative context. The normative ideal (the good) is being equated with the epistemological ideal (the essential). In other words, what remains after one has abstracted away the evil or ugly is being equated with what remains after one has abstracted away the non-essential. In order to illustrate this point, I will now examine several instances of the fallacy as it has appeared in certain Libertarian and Objectivist writings.

Phase II: Thawing-Out Selected Frozen Abstractions.

Example A: The simplest of the cases analyzed here is the claim by Adam Reed that one should not refer to such pull-seekers as the "Big Four" of California railroad history or James Taggart and Orren Boyle (in Atlas Shrugged) as being "businessmen." In criticizing R. A. Childs' essay on big business-promoted statist government policies[6], Reed says:

Childs talks about organized crime, but insists on calling it "Big Business" throughout the article. Socialist slogans to the contrary, business and crime are not synonymous. A businessman is a man who earns wealth by organizing the production, distribution, and voluntary exchange of values. A man who obtains things by initiating force, either directly or through hired thugs, is properly called a criminal. A criminal is no less a criminal when he lets the taxpayers hire his thugs for him...No one who has read Atlas Shrugged would call either Taggart or Boyle a businessman. Instead, Taggart and Boyle, like the "Big Four" and others, are criminals who were afraid of competition from (real) businessmen.[7]

Childs' reply to this criticism is right on target:

By Mr. Reed's definition, there can be no such thing as a dishonest businessman (one who accepts favors from the government). this use of the term would, I submit, rob it of what modern logicians call "existential import," i.e., it might very well have no referents. I prefer to use the term as it is used by Rand and innumerable other thinkers and then to qualify the concept with adjectives like "honest," "dishonest," and so forth.[8]

The question that remains is this: Who are "those who are conventionally called 'businessmen'"? What is the definition of the term "businessman," "as it is used by Rand and innumerable other thinkers"? I would propose the following: a businessman is a man who engages in the activity of organizing the production, distribution, and voluntary exchange of values.

With such a value-neutral definition, both dishonest or criminal businessmen and honest or non-criminal businessmen are included in the entire set of referents of the term. We then make our moral and legal distinctions between the two types. We form these distinctions by identifying why, whether, and to what extent businessmen carried out that activity coercively (by obtaining government favors, for instance) or in a laissez-faire manner.

On this basis, we can sub-categorize our concept of "businessman" and qualify it with adjectives, as Childs correctly observes. In so doing, we can avoid freezing our abstraction of "businessman" on the lower level of one of the concretes that is subsumed within it (namely, "honest businessman"). We can keep our abstraction of "businessman" thawed out, without compromising our moral condemnation or disapproval of certain of its units.

Example B: In her highly controversial essay, "The Nature of Government," Ayn Rand presented a concept of government and of what government should be that are still the focus of considerable debate. Leaving aside the question of the validity of her discussion of this subject (merely noting that I have cited her concept of 'government' as not being an example of the present fallacy), let us note what Rand says in that same essay about the concept of 'society':

...these very benefits [knowledge and trade] indicate, delimit and define what kind of men can be of value to one another and in what kind of society: only rational, productive, independent men in a rational, productive, free society. A society that robs an individual of the product of his effort, or enslaves him, or attempts to limit the freedom of his mind, or compels him to act against his own rational judgment--a society that sets up a conflict between its edicts and the requirements of man's nature--is not, strictly speaking, a society, but a mob held together by institutionalized gang-rule. Such a society destroys all the values of human coexistence, has no possible justification and represents, not a source of benefits, but the deadliest threat to man's survival.[9][emphasis added]

Observe the thinly disguised switch of definition of the concept of 'society' from social-environments-in-general to those social environments where physical force is barred from relationships among men. Rand starts with the commonly accepted meaning of 'society' that includes all social environments, whether good or evil, rational or irrational, slave or free, peaceful or warring, civilized or primitive, moral or immoral. Then she shifts to a position holding, in effect, that 'society' is synonymous with "moral, rational, productive, free society"! A slave society "is not, strictly speaking, a society," Rand maintains.

Thus, Rand, because of her understandable and justifiable hatred toward slavery and mob-rule, is drawn into freezing her abstraction of 'society' to the level of one of its species: 'moral society.' Or rather, she has shrunken it down and then frozen it. For she first allows that a slave society is a society long enough to condemn it, and on that basis then denies that it is a society. Because a slave society does not fulfill the proper, moral function of society--viz., the facilitation of knowledge and trading goods--Rand denies that it is a society at all. Thus, she freezes slave societies out of her abstraction of 'society.'

Of course, the much simpler, much more rationally desirable way of mentally pigeon-holing 'slave social environment' is as one subcategory of 'society in general,' itself a value-neutral class including all social environments. We can then distinguish moral and immoral, good and evil societies by identifying whether and to what extent they uphold or fail to uphold individual rights--and we can qualify them accordingly, with the appropriate adjectives. In no way does this entail the conceptual ostracism of undesirable members of a class for what it is: an inappropriate, unnecessary way to mentally deal with them.

Example C: Lest anyone regard this as an isolated instance, attributable to "the early Rand" (c. 1963), I now present a more recent case in point, from Rand's essay, "The Age of Envy." This essay is an elaboration upon her claim that the emotional atmosphere of today's culture is one of envy or, more precisely, "hatred of the good for being the good."[10]

The experience of this emotion is possible only to a person who has sabotaged his/her cognitive development by avoiding mental effort and understanding. Such a person is instead pursuing whims and deception of others (thus freezing his/her mental functioning to the concrete level appropriate to childhood).[11] Anyone who experiences this emotion as a characteristic response to the sight of his/her values, is referred to by Rand in bitterly caustic terms as a "hater," an "inhuman object," a "creature," "it," a "hating creature," an "envious hater," a "monster."[12]

In other words, if one's basic, typical response to the sight of one's real values is hatred, one is not human, one is not a man. Yet, curiously enough, even though this assertion is stated or implied numerous times in Rand's essay, there are also certain passages in which she relents and temporarily admits these "haters" back to the human race:

The hater of the good is the man who did not make this transition [from the perceptual level to the conceptual level]...[The hater has] as stagnant a mentality as a human being can sustain on the edge of the borderline separating passivity from psychosis...How does a human descend to such a state?[13][emphasis added]

Similarly to her treatment of slave societies, Rand first relents long enough to condemn those human beings who are haters of the good. She then denies that they are human beings, but later lapses back into referring to them as human beings (or men), after seeming to have firmly ostracized them from the human race with such epithets as "creature," "monster," "inhuman object," and "it." (!)

As with her abstraction of 'society,' Rand has frozen her abstraction of 'man' ('human being'). She excludes from it certain men whom she considers as possessing "a quality of abysmal evil."[14] Then she fails to integrate her frozen abstraction consistently --which would be impossible anyway, with her knowledge of man's nature--instead allowing it to thaw out and expand again. (Coincidentally, this happens as her most intense expressions of moral wrath subside and scientific curiosity takes over.)[15]

Unless we choose to indulge in psychologizing and to speculate as to Rand's possible motives, we are left with a sense of confusion and uncertainty. Why does she present such a grossly inconsistent discussion of the concepts of 'man' and 'society'? Surely it would not be out of place to suggest that there is some carelessness here--a subconscious confusion of conceptualization with evaluation. It certainly appears that Rand has on occasion allowed her value-responses (i.e., her emotions) to control the way she sets up and uses her abstraction.

What, then, is the preferable policy? To conceive of and define 'man' as: the rational animal. This, of course, means not that man characteristically acts in accordance with reason, but that man has the volitional capacity to act rationally. Therefore, unless one contents that haters (and appeasers, who are even worse!) are metaphysically irredeemable, one must limit oneself to classifying them as (abysmally) evil men. Such a policy results in mental clarity, precision and objectivity--with no compromise of one's moral principles.

Example D: Next let us consider an instance of the fallacy that is considerably more complex. In his essay "Man's Standard of Value," Morris Tannehill challenges and rejects Rand's concept of "value" as "that which one acts to gain and/or keep."[16] To him, a value instead is:

...anything which is actually beneficial to human well-being, whether a man "acts to gain and/or keep" it or not. If it's conducive to human well-being, it's a value, and a man's evaluation has nothing to do with its being a value...A value, because it is necessary for human well-being, is objectively beneficial to men (e.g., fresh air is a value to man), and even an objective evaluation has nothing to do with making it a value. It's a value precisely because it objectively contributes to human well being.[17][emphasis added]

Thus, if one believes that a water supply is safe to drink and acts so as to drink it, but the water in fact is poisonous, then the water was not actually a value. One evaluated it wrongly as being a value; instead it was actually a dis-value. There is no such thing as a harmful value; there are only (to be redundant) beneficial values. So Tannehill maintains, anyway.

It is apparent from Tannehill's claim that "a man's evaluation has nothing to do with its being a value," that he regards value as being independent of consciousness. Instead something must only be actually beneficial to (objectively contribute to) human well-being, Tannehill believes. One's conscious recognition that a thing is actually beneficial to one is irrelevant to that thing's being valuable, he holds.

There is, in fact, a basic contradiction in his position, one which remains somewhat obscured by the monumental job of context-dropping he has done. The context he has dropped is precisely: how this projected benefit is to be conferred upon the recipient. That is, he has not deemed it necessary to consider whether I am to be permitted to choose an allegedly beneficial thing, or whether it is to be forced upon me. Tannehill presumes that value can be defined without reference to this question.

Actually, the opposite is the case: value cannot be defined without reference to the issue of force vs. free choice. If something is forced upon me, it cannot truly be said to be valuable for me at all. That which negates or circumvents my capacity to choose values--my rational-volitional mind--cannot be valuable for me, cannot be proper to my life, qua rational-volitional being.

Tannehill places much emphasis upon the end, or the effect, of some existent upon a projected beneficiary. This projected beneficent effect is the basis on which he deems a thing "valuable" or "a value" to man. But let us identify that which he overlooks, a very crucial aspect of the matter: the effect upon the beneficiary of the means of acquiring the allegedly beneficial thing. The means happens to be an integral part of the end in any given human causal action sequence. In the final analysis, the means affects the beneficiary's well-being just as surely as do the projected valuable thing's attributes do themselves.

Specifically, in this case, if the means is force, the object being conferred upon the beneficiary cannot be a value to him. If he chooses the object of his own volition, however, it may very well not be a value then either. We have not yet considered what effects are entailed by choice-by-whim vs. choice-by-reason. Still, as a minimum at least, we can see that value--qua actually beneficial or objectively contributory to human well-being--cannot be independent of the beneficiary's choice.

And entailed by his choice to gain and/or keep something are his antecedent belief (correct or incorrect) that it will benefit him, and his consequent act (present or future) of gaining and/or keeping it. So, to meet Tannehill's criterion that a thing be actually beneficial to human well-being, it must at the very least be true that a man "acts to gain and/or keep" it.

Here, then, is the contradiction: from the first half of Tannehill's definition of "value" (anything actually beneficial) can be inferred a conclusion that is in direct conflict with the second half of his definition (whether sought after or not). Assuming the first half to be true, it follows that the second half cannot be.

I would further contend that the first half of the definition is incorrect, as well. It is just too narrow. "Good" is already a serviceable concept for referring to a value chosen according to a rational standard of value. It would be a wasteful error to equate it with value-in-general. Also, "objective need" is already a serviceable concept for referring to that which one actually requires for one's survival or well-being. It would, therefore, be equally unparsimonious to limit "value" to this meaning.

It would be far more useful and far less confusing to conceive of "value" as Rand has done: to recognize that one's value may or may not be rationally chosen (good) and may or may not be in accord with one's survival requirements (objective needs). Tannehill has instead rendered the term "need" superfluous. He has frozen his abstraction of "value" to the level of those values that are actually beneficial to man. (He has done so, moreover, without even a proper understanding of the pre-condition of something's being actually beneficial.)

In so doing, Tannehill has made it necessary to add another term to apply to the specific case where one believes something to be valuable, chooses it as a value-goal, and acts to gain and/or keep it. The term his wife, Linda, proposed for this purpose--"evalue"--is defined as: "That which one believes, rightly or wrongly, to be a value."[18]

Note how the Tannehills have substituted their notions of "value" and "evalue" for Rand's concepts of "objective need" and "value," respectively:

value[Tannehill] = that which can actually be beneficial to one's well-being = objective need[Rand[

evalue[Tannehill] = that which one believes, rightly or wrongly, to be a value = that which one acts to gain and/or keep = value[Rand]

By this move, however, the Tannehill's have failed to explain more than--let alone to be as clear and precise as--Rand with her concepts of "objective need" and "value."

It would seem that part of the motivation behind their conceiving of "value" in this way is to deny that evil, irrational, life-destroying values (by Rand's definition of the term) are actually values. But it solves no epistemological problems and wins no moral battles to try to purge the unsavory units from one's concepts and to coin new concepts into which to dump them.

A subdivision of the original concept of "value," qualified by appropriate adjectives, would suffice--and it would avoid the unnecessary proliferation of concepts. To be somewhat rhetorical: surely anyone agreeing with Rand's epistemological "razor"--"concepts are not to be multiplied beyond necessity...nor are they to be integrated in disregard of necessity"[19]--can see that this is true.

Example E: The fifth case I wish to discuss is actually not an instance of the frozen abstraction fallacy. It is so widely misunderstood among Libertarians and Objectivists, however, that it may readily be inferred as being one. I refer to the interpretation given by Tibor R. Machan to Rand's concept of "government."

In her essay "The Nature of Government," Rand makes two statements about government that give the appearance of being definitions:

A government is an institution that holds the exclusive power to enforce certain rules of social conduct in a given geographical area.

A government is the means of placing the retaliatory use of physical force under objective control--i.e. under objectively defined laws. [20][emphasis in original]

There are three distinct, different interpretations given to this latter, emphasized statement. First, there is the interpretation that Machan and I hold, which is strictly a literal one.[21] We hold that Rand is making a statement about all government and that the phrase, "the means of placing" literally means: "can be used to place" and not: "is always used to place." We hold that Rand is asserting a truth about all governments, in terms of their fundamental capacity as a human tool or instrumentality--i.e. in terms of the original, standard, proper function and purpose for which they were created.

Secondly, there is the radically non-literal interpretation given to it by some limited governmentalists (e.g., Charles Jackson Wheeler).[22] They hold that Rand is actually talking only about actually proper governments, not all governments. They interpret the phrase "the means of placing" non-literally as: is always or for the most part used to place.

Both the first and second interpretations are based upon what their proponents believe to be the context of Rand's italicized statement. The dispute between them has not yet been resolved, but its ultimate outcome is not crucial to the present discussion. I merely make note of these alternative views in order to contrast them with the third view of Rand's statement.

The third interpretation is the moderately non-literal view held by the anarchists Ronn Neff and Louis A. Rollins. In criticizing Machan's view, they raised objections that seem, at least prima facie, to be unanswerable. For instance, Rollins has maintained that:

...an implication of Machan's view, in the context of today's world, is that there are no governments in existence...the institutions generally called "governments" (such as "the U.S. federal government") cannot realistically be viewed as means of placing the retaliatory use of physical force under objective controls since they act so as to prohibit any protection from and retaliation against their own coercive laws.[23][emphasis added]

Neff has similarly stated that:

Nothing called a government has ever been a means of placing the use of retaliatory force under objective control. On the contrary, every government has always been a means by which some have oppressed others. It does not matter what governments may have been established to do, or were hoped to do; every government has been a means of oppression, not protection.[24][emphasis added]

As serious as these objections appear to be, however, they both contain a subtle but crucially important flaw. Although Rollins and Neff take Rand's statement as applying to all governments, they completely fail to grasp the interpretation of the phrase "the means of placing" that such a view necessitates. Instead, they regard it as meaning "is always or for the most part used to place." The consequently view Rand's italicized statement as inapplicable to all cases.

Rollins and Neff fail to recognize that there is an ambiguity in the phrase "the means of placing," just as there is in the phrase "the rational animal." Thus, their position is like that of someone who claims that the statement, "Man is a rational animal" cannot apply to all human beings, because not all human beings have actually been rational ("always or for the most part").

Clearly, this position is untenable. The phrase "the rational animal" can be interpreted to make that statement applicable to all human beings, if by "rational" we designate the capacity to be rational (which may or may not be actualized, and to a greater or lesser extent, by all human beings). The same is true for Rand's statement about government and the particular phrase in question.

But what if we were to assume that Rollins' and Neff's view of this statement was correct? We would then have to believe that Rand (or at least Machan, in his interpretation of her) is committing the frozen abstraction fallacy. That is, Rand would appear to be excluding all historically existing governments from the concept of "government," since not one of them has always (or for the most part) placed retaliatory force under objective control.

Since this view is invalid, of course, Rand and Machan do not actually commit the frozen abstraction fallacy. Yet, due to the fact that Neff's and Rollins' arguments have (until now) not been effectively refuted, there is some likelihood that other astute readers of this essay might have drawn the conclusion that a fallacy was committed. It is because of the current widespread confusion and misunderstanding on this issue that I have included the foregoing along with the bona fide instances of the fallacy that preceded it. Now no one will be able to confuse matters further in this issue by wrongly applying the frozen abstraction argument to Rand's view of government.

Example F: Finally, I will briefly note another non-instance of the frozen abstraction fallacy, which involves Rand's concept of "art." Her refusal to include abstract art within the class of art-in-general[25] gives at least a strong appearance of being an instance of the fallacy.

But once we look upon art (and language) as having been originated in order to serve a certain proper purpose and function--viz., as a symbolic tool of cognitive integration of certain aspects of reality--her stand on "abstract art" no longer seems fallacious. That is, if a given artistic (or linguistic) form does not and cannot fulfill that standard, proper function and purpose (viz., cognitive integration), then it is non- functional (viz., cognitively non-integrative).

And if one attempts to regard it as functional in that respect (or if one intends it to be regarded by another as thusly functional), then it has (or is intended to have) a dysfunctional effect--viz., a cognitively disintegrative effect. These facts I presume to be the basis of the labels "anti-art" and "anti-concept" that Rand has used in her esthetic and epistemological writings.[26]

In this connection, I refer the reader to my essay "To Catch a Thief," which deals with the stolen concept fallacy and the Cretan Liar Paradox.[27] There I claimed that anyone stating one of the meaningless reformulations of that Paradox was not (as Ronn Neff claimed)[28] himself stealing the concepts "true" and "false," unless he also explicitly attributed truth or falsity to that meaningless statement.

Subsequently, I have concluded[29] that such a reformulator was doing something much more subtle and much worse: by employing an anti-sentence or meaningless sentence, such as "This sentence is false," such a person is encouraging others to commit the stolen concept fallacy. Accepting in good faith his pretense at meaningful communication, they either agree that it is false (which it is not)--or they instead object that it is true (which it is not, either).

The principle that they all unsuspectingly fail to apply here is this: since both truth and falsity presuppose meaningfulness, an anti- sentence (i.e., a meaningless sentence) can be neither true nor false.

Yet, Neff maintained that such a meaningless utterance did admit of truth or falsity. All that he succeeded in establishing, however, is that if it does imply another statement about itself being true or false, then that statement--not the Liar's utterance--is the one which is true or false.

In other words, suppose we grant Neff's assumption that (a) "This sentence is false" implies (b) "Sentence (a) is either true or false." We must then logically conclude that: sentence (b) is obviously itself false, since sentence (a) is meaningless and can be neither true nor false.

It was Neff's failure to grasp this fact that led me to reply as I did. Suffice it to say in the present context that the utterance "This sentence is false," is actually not a sentence at all, since the apparent subject-term is not actually a subject-term (there being no actual sentence it refers to at the time when it tries to refer to one). Rather, it is an "anti-sentence," just as abstract art is rightly classified as "anti-art." And classifying them in this manner does not commit the frozen abstraction fallacy.

Phase III. Mental Economics and the Frozen Abstraction Fallacy.

As I stated earlier, there is a basic pattern present in all instances of the frozen abstraction fallacy in discussions of normative issues. It is the confusion of the desirable with the distinctive, i.e., the confusion of the good with the essential.

In the preceding examples I have just "thawed out", the writers attempted to substitute what remained after they had mentally separated the evil members of a given class for what remained after they had mentally separated those existents not possessing the essential characteristic of members of that class. That is, each of them abandoned a value-neutral epistemological criterion and substituted for it a normative (viz., a moral) criterion in setting up their various abstractions. Thus, they conceptually ostracized the evil, undesirable members of a broader class and formed a new abstraction with which to mentally retain them.

But what of the essential similarities between members of the newly abstracted class and the members of the class from which they were ostracized? Leaving aside their moral or normative differences, are they not all similar in one essential respect? If so, what of the wider class to which they all belong, by virtue of that respect?

In none of the examples above is such a wider class specified (at least, not non- ambiguously). Yet, clarity and integrability of mental contents would seem to demand one. And in fact, this wider class is the original class which was shrunken and frozen so as to contain only its good members.

Thus, nothing has been gained either epistemologically or morally from such a policy. On the contrary, the relation between the good and bad existents is rendered less clear and simple as when they are regarded as qualified instances of a common wider class.

Such a loss in mental economy and clarity is inevitably the result whenever people fail to define their concepts in terms of essentials. Using a moral criterion as one's differentia is merely one example of such a failure. The frozen abstraction fallacy, in other words, is a violation of the principle of unit-economy, which is the necessary basis of definition by essentials. In committing this fallacy, one has to evade, ignore, or omit the essential similarities between a group of existents. One then multiplies the number of abstractions or concepts that one must form in order to retain all those existents--when a nice, neat conceptual subdivision would have been so much more economical.

Thus, another way of stating the problem is this: the fallacy of the frozen abstraction entails a violation of one of Rand's epistemological razors: "...concepts are not to be multiplied beyond necessity...nor are they to be integrated in disregard of necessity."[30] As we can see, this is actually a double-edged razor, but only the first half of it applies in this context, the principle better known historically as Ockham's Razor or the Law of Parsimony.

The Law of Parsimony and its base, the principle of unit-economy, are not arbitrary, subjective criteria for conceptualization, true merely "because we never allow them to be otherwise," based purely upon traditional practice. On the contrary, they are objectively valid for all human beings, just as are corresponding normative principles in the domains of ethics, politics/law, and esthetics. And they are so because, as are the other principles, they are based upon the requirements of man's survival qua rational being.

More specifically, they are based upon the requirements of cognition (man's basic means of survival).[31] They are the guiding principles of man's conceptual faculty, helping him to condense his knowledge:  

...the range of what man can hold in the focus of his conscious awareness at any given moment is limited. The essence, therefore, of man's incomparable cognitive power is the ability to reduce a vast amount of information to a minimal number of units--which is the task performed by his conceptual faculty...Conceptualization is a method of expanding man's consciousness by reducing the number of its content's units.[32]

These principles are the proper principles of mental economy, of maximizing the cognitive gains of one's mental effort, of maximizing one's cognitive efficiency. They are the means of "freeing man's mind to pursue further, more complex knowledge."[33] The frozen abstraction fallacy, by contrast, is a very wasteful, unparsimonious policy of mental economy. It serves only to freeze the human mind.

To pursue the analogy further, the frozen abstraction fallacy entails a sort of moralistic "intervention" into one's conceptual hierarchy. One is allowing one's ethical considerations to distort the conceptual mechanism's workings. What is needed instead, in this context, is a careful "separation of epistemology and morality."

For it is only when one ceases the moralistic controls over one's genuses and species and allows them to function freely, according to the natural laws of mental economy that one can avoid the harmful effects of what we might call "conceptual inflation." In the combined words of Descartes and Legendre--which I have integrated in view of the requirements of epigrammaticism and in disregard of the protestations of linguistic purists (!)--the proper principle of mental economy is: Cogito, ergo... Laissez-nous faire!

To conclude: morality may properly offer the service of providing the standard for validating epistemological principles, but it may not presume to legislate reality in and out of existence. Reality is objective and exists independently of one's consciousness of it. It can neither be created nor destroyed merely by one's moral approval or condemnation. "Wishing will not make it so"--not even the special brand of wishful thinking known as freezing one's abstractions.[34]

Endnotes

[1] This essay, first published in 1973 in Equitas (a publication connected with a 1970s Midwest organization called Equitarian Associates) is organized around identifications I made during March and April of 1971. The identifications concern Ayn Rand's concept of 'society', Tibor Machan's view of government, and Morris Tannehill's view of value--all of which I believed at the time to be based upon a common fallacy. The essential nature of this fallacy became clear to me when I discovered the paradigm instance of it: Plato's theory of the Forms. This discovery was a byproduct of discussions I had with Douglas Rasmussen while we were attending a course on Plato at the University of Iowa. I then discovered it was Ayn Rand who gave a name and definition to this fallacy. It is because of this, as well as her identification of the principle of unit-economy (of which this fallacy is in violation), that I dedicate this essay to Miss Rand. Despite the irony of her committing the fallacy at least twice herself, her identification of it is a significant milestone in the understanding of how to properly form normative concepts.

[2] "Nixonomics" is the term that originated during the first presidential administration of Richard M. Nixon. It was used by economists, journalists, etc., to refer to his economic policy of wage-price controls, which included a temporary, so-called "freeze."

[3] Ayn Rand, "Collectivized Ethics," The Virtue of Selfishness, New York: Signet, 1964, p. 81.

[4] 1997 note: It is ironic, in this connection, that Rand, Peikoff, and others selectively deny the label "entity" to things such as clouds, rivers, and piles of dirt. E.g., Rand, in Introduction to Objectivist Epistemology (2nd ed.) said whereas a mountain was an entity, a pile of dirt would not be an entity, unless it had glue poured into it so that it was welded together and that there were actions possible to it as a whole. (pp. 268, 273). But as Rand herself said shortly thereafter (p. 277), materials (including dirt, let it be noted) do not belong to "a separate metaphysical category, because materials cannot exist except in the form of entities of some kind, nor can entities exist without materials. That is, physical entities. Matter is what all physical entities have in common, and "the things which we call physical entities are all made of some kind of material. But you can't consider one without the other." Thus, glued-welded or not, a pile of dirt must be an entity. It may not have all of the same actions possible to it, but surely some actions are possible to it. As Aristotle said--and as Rand, Peikoff, Kelley, and every sane person, Objectivist or otherwise, concurs--there is no such thing as an attribute apart from an entity that has that attribute. There is no cloud that does not have a shape, no river that does not have a length, no pile of sand that does not have a color. Ergo, these are all entities, and Rand et al, in denying that they are, are guilty of the fallacy of the frozen abstraction. Shades of Socrates!

[5] Plato, Parmenides. 1997 note: in a recent email, a college professor who for the time being will remain anonymous sent the following interesting comments. (Also see note 34 below.) (For several months prior to posting this essay on my website, I made repeated attempts to ask this gentleman whether I could refer to him by name. He finally said he had no recollection of the discussion, after which I emailed him a copy of this essay, with his appended remarks. As of May 19, 1998, that was over a month ago, and so far no further response. If he changes his mind after seeing this posting on the Internet, I will be happy to give replace the anonymous citation with his real name. In the meantime, I am happy to include his excellent comments.)

Professor Anonymous: I would like to say something to vindicate Plato because I think one really needs to be careful about what one asserts he believed there were Forms for. Roger cites Parmenides 130, which indicates to him that Plato did not have Forms for mud, hair and other "undignified and trivial objects." I think it is fairly clear that at least in one stage of his career, Plato did not believe in these. But the reasoning that Roger is using is Parmenides', not Socrates', who says something quite different: that he sees no reason for having Forms for mud and hair because "these things are just the things we see." His point is that mud is not an opposite, nor is hair. Plato believed in Forms for opposites throughout the Middle Dialogue period--large, small, beautiful, good, justice, etc. (His reasons for having mostly positive Forms have to do more with ontological parsimony than anything else, and he does not believe that genera defined negatively are legitimate--see the Phaedo 103 or so.) In the middle dialogues, you never see a Form for Horse. His reason is that you don't need a Form as paradigm example of horsehood, since prime examples are all around us. At Republic 523-5, he makes this very clear in the case of fingers--they are just what we see them to be. But large is not just what we see it to be--whatever we see to be large (in comparison with one thing) is also small (in comparison with something else). Without a paradigm for Largeness which is in no way small (the Form the Large), one would not be able to distinguish large from small, so a Form is required here. Roger also mentions Republic 595 or so where Socrates says they have been in the habit of postulating a Form for each general class, and then argues that the Form the Bed is one. One needs to be careful here since Plato notoriously believed that reality comes sorted and that there are Forms for only natural kinds. G.E.L. Owen and his followers would argue that Book X of the Republic belongs to a somewhat later development in the Theory of Forms--one in which the Timaeus was written. In the Timaeus, there seem to be Forms for horse and every other natural kind. The reason, Owen thought, is a new emphasis in Plato on the idea that everything is in a flux, and individual horses now fall short of perfect Horsehood because they are in time and move from non-horsehood to horsehood and vice-versa. That is, there is a move on Plato's part to assimilate horse with true blue opposite qualities like large or beautiful. In no part of Plato's dialogues does one find any evidence of a move from what ought to be to what is in support of the Theory of Forms. Undoubtedly, Plato was motivated to conceive of the world of the Forms as a perfect one, but this should not be construed as some kind of simple fallacy"