The critics of rationalism object to the practice of determing complex matters of fact through "logical" deductions from over-generalized descriptions of facts. Use of over-generalized facts is often a symptom of insufficient knowledge. People who lack mastery (i.e., relevant factual knowledge) of a given subject don't realize the extent of their ignorance. They are therefore incapable of appreciating why their conclusions are false. The problem with rationalism, therefore, is not that the rationalist derives conclusions without factual evidence, but that he derives conclusions without sufficient evidence. The rationalist suffers from empirical irresponsibility.
There is actually very little difference between the rationalist and the rationalizer. While it is possible to imagine a rationalist who is not a rationalizer, all rationalizers are rationalists. The method of the rationalist is, par excellence, the method of the rationalizer. The rationalizer seeks to justify a given belief. The conclusion is predetermined, and reasons are added merely to add a veneer of logical justification to the non-rational conclusion. The method of rationalism makes use of vague, empirically impoverished generalizations for three primary reasons: (1) to avoid grappling with pesky details which may upend the predetermined conclusions; (2) to hide a lack of expertise in the subject at hand; and (3) to provide ample wiggle room for equivocation.
Objectivism is officially and theoretically opposed to rationalism. Rationalists are guilty of using what Objectivists refer to as "floating" abstractions. A floating abstraction is a concept that doesn't stand for anything in reality. You might think that concepts such as griffen, phlogiston, wookie might qualify as floating abstractions; but this is not what Rand had in mind. The notion of floating abstraction only makes sense under the Objectivist assumption that concepts have "true" meanings which contain all the existents which they integrate. Floating abstractions are essentially "true" (or, rather, "valid") concepts which don't, however, contain all the existents which they integrate. The floating abstraction doesn't refer to everything that is subsumed under it, because the person who uses it hasn't formed it properly. The individual who makes use of floating abstractions fails to understand precisely what these abstractions refer to in reality. Like the parrot, he uses words without understanding their meanings.
If the notion of floating abstraction was introduced to describe and do battle against rationalism, it's a signal failure. The problem with rationalists is not that they don't understand the meanings of the terms they use, but that they resort to terms that are too general to bear the weight of their contentions. They use the vagueness of terms to conceal their empirical irresponsibility.
That rationalists are guilty of using the vagueness of overly general, "abstract" terms to defend their various positions should be obvious to anyone who has examined, with a critical eye, rationalism in practice. Even Objectivists have noticed this. "A rationalist is all in favor of abstractions," notes Leonard Peikoff. "He likes broad abstractions. And to that extent, he has great virtue." [Understanding Objectivism, Lecture 7] What is particularly interesting is that, even though Peikoff connects rationalism with "broad abstractions," he doesn't grasp why this is a problem. On the contrary, he describes liking broad abstractions as a "great virtue." What is going on here? Why would Objectivism, which declares itself an enemy of rationalism, be praising one of the chief defects of the rationalist?
The answer is actually very simple: Objectivism approves of broad abstractions because Objectivism, at its core, is a rationalist philosophy. It may oppose rationalism in theory; but in practice, it loves to use "broad abstractions" to reach predetermined conclusions. Curiously, in the same lecture, Peikoff trots out Leibniz's rationalist argument in favor of monads as an example of rationalism. The Leibniz argument essentially is of the same type that is familiar in the Objectivist metaphysics: it's an attempt to determine matters of fact on the basis of logical deductions from broad abstractions. Why should Peikoff reject arguments based on broad abstractions when used by Leibniz, but not when used by Rand?
Objectivists circumvent this difficulty by simply declaring that their abstractions, however broad, are "true" of reality, while Leibniz's abstractions "float" and are "detached". In other words, broad abstractions are fine when used by Rand to justify her own predetermined conclusions; they are "floating abstractions," cut off from reality, when used to justify anyone else's conclusions, predetermined or otherwise.
The real problem with broad abstractions is that they don't convey enough information. Although their meanings are precise enough, their reference, because it applies to so many different things, is not exact enough. Objectivism, with its unit economy, should have understood this. But Rand's penchant for rationalist arguments seems to have led her to deny the referential vagueness of overly general terms:
A widespread error ... holds that the wider the concept, the less its cognitive content -- on the grounds that its distinguishing characteristic is more generalized than the distinguishing characteristics of its constiuent concepts. The error lies in assuming that a concept is nothing but its distinguishing characteristic. But the fact is that in the process of abstracting from abstractions, one cannot know what is a distinguishing characteristic unless one has observed other characteristics of the units involved and of the existents from which they are differentiated....
To grasp a concept is to grasp, and, in part, to retrace the process by which it was formed. To retrace that process is to grasp at least some of the units which it subsumes [26-27]
Rand contends that, as long as concepts are formed "properly" (i.e., in the Objectivist way), broad abstractions do not in fact contain less "cognitive content." But this misses the point. Indeed, Rand's argument here is based on an error. Concepts do not contain information. A concept is a meaning, an item of description. By itself, it contains no information. To be sure, when used in propositions, concepts can help convey information. Even then, the information conveyed is not necessarily "cognitive" (in the sense of conveying knowledge about matters of fact). As I have repeatedly noted throughout this series on the Objectivist Epistemology, concepts can be used to describe truth and falsehood, fact or fiction, hypothesis or dogma. The truth or falsity of a proposition does not arise from the truth or falsity of the concepts used in that proposition.
Essentially, Rand argues that vagueness in terms is caused by lack of precision in definitions, which in turn is caused by improper concept formation. This is completely wrong. Vagueness in terms is caused by insufficient specificity: when the term used is too broad for the purposes at hand, vagueness results. Precision of meaning has nothing to do with the issue, nor will providing a more "precise" meaning render a broad abstraction less vague. Consider the following two propositions:
- I went to the zoo and saw the animals.
- I went to the zoo and saw red pandas, flamingos, crested screamers, jaks, rheas, and a golden pheasant.
Which statement is less vague, more presise? Obviously, the second statement is more precise. It contains more information; it describes more precisely what is meant by "animal" in the first statement. Where did the precision come from? From more precise or "truer" definitions? No, the suggestion is absurd. You do not increase precision by increasing the precision of your definitions (or making them more "correct" or "true"); you increase it by using more specific, less general terms.
Rand admits that knowledge involves condensation. However, she refuses to admit that this condensation has trade-offs. She suggests that you can condense without losing anything. But this just isn't true. With greater condensation comes greater vagueness. More general terms, when used as items of description, are more vague. Precision is increased, not through more "precise" definitions and/or "proper" concept formation, but by using more specific terms.