In the opening of his essay on the analytic-synthetic dichotomy, Peikoff provides the following anecdote about a discussion he had with a professor which illustrates how Objectivists use definitions and logic to evade facts while assuming the very point at issue:
Some years ago, I was defending capitalism in a discussion with a prominent professor of philosophy. In answer to his charge that capitalism leads to coercive monopolies, I explained that such monopolies are caused by government intervention in the economy and logically impossible under capitalism.... The professor was singularly unmoved by my argument, replying:
"Logically impossible? Of course -- granted your definitions. You're merely saying that, no matter what proportion of the market it controls, you won't call a business a 'coercive monopoly' if it occurs in a system you call 'capitalism.' Your view is true by arbitrary fiat, it's a matter of semantics, it's logically true but not factually true. Leave logic aside now; be serious and consider the actual empirical facts on this matter."
Doubts arise, of course, as to whether Peikoff has accurately related the professor's argument. But even if this professor said what Peikoff claims he said, the professor nonetheless has a point. Objectivists do in fact tend to resort to definitional arguments. Such arguments suffer from the fallacy of begging the question. Grant someone's definitions, and the rest follows, logically. But since definitions merely establish what one means by the words one uses, this is not enough.
If Peikoff's professor had really said, "Leave logic aside now," he was making a poor argument for his case. He would have been better served by saying, "Leave definitions aside now; let's start with the actual facts of the matter."
While Peikoff is entirely free to define capitalism any way he pleases, he's not free to assume that his definition accords to anything we find in reality. That's empirical question dependent on the actual facts relevant to the case at hand. Peikoff's proof that "coercive" monopolies cannot arise under capitalism is entirely circular. It's contained in his definitions of capitalism and coercive monopoly. Arguing from definitions is not how one ascertains truth. If you're serious about acquiring true knowledge of reality, you have to go beyond your definitions of words and actually test your actual beliefs (as opposed to the meanings of your terms) against reality. And that's what Peikoff does not want to do, because it's much more difficult to establish his views on the basis of empirical testing rather than on speculation based on arbitrary definitions. To begin with, capitalism as he defines it has never existed, nor is it plausible that such a system ever will exist. If we seek out systems that approximate Peikoff's definition, we find that his assertions about "coercive" monopoly a tad bit exaggerated. What we find when we look at the "lightly" regulated capitalism of the nineteenth century is a very strong desire on the part of many firms to discover and attain monopolistic advantages, which sometimes led to monopolies which even Peikoff would admit are "coercive." It is often through such monopoly advantages that profits are secured against the uncertainties of market competition; and profits are the lifeblood of every business.
It matters little if Peikoff responds by insisting that his (and Rand's) definitions are "true." Definitions define the meaning of words. How one goes about defining one's terms is, at least initially, entirely "arbitrary." Once one's terms are defined, one needs to stick to the initial definitions, or risk equivocation. But the meaning chosen is neither true or false. As I have repeatedly elucidated when discussing the Objectivist view of definitions, words by themselves don't tell us anything about reality. It's only when something about reality is asserted using these terms that the issue of truth is broached.
The issue of monopolies under "capitalism" is very complex. Any sophisticated understanding of political systems strongly suggests that the sort of "laissez-faire" or "unregulated" capitalism imagined by Rand is a fantasy. Human beings tend to respond to incentives; and built-in to any social system featuring widespread division of labor are incentives, spread across various factions, against the vague, poorly defined "unregulated" capitalism advocated by Objectivists. Politicians don't want laissez-faire; lawyers don't want it; speculators and investors don't want it; most businessman don't want; the rich don't want it; the poor don't want it. It turns out that, when the matter is examined closely, almost everyone wants at least something from the government: security for their investments, security for their employment, security for the old age, security against natural disaster, security against the vicissitudes of competitive life. Real life (as opposed to the type of existence imagined by Rand) remains threatened from all sides by uncertainties. Individuals seek hedges against this uncertainty; and the government, as the most powerful institution in society, remains for many the most attractive hedge of all. Given the constitution of human nature, it's extremely unlikely that very many individuals can be persuaded by a mere ideology not to regard government as an attractive hedge against uncertainty.
So when we come to investigating the monopolistic side of "capitalism," it simply will not do to dream up how things would work in an imaginary, political unfeasible version of the system. Empiricism requires that we stick to plausible versions of that system, not merely imaginary ones. Once we do so, we are confronted by a far more ambiguous set of facts, none of which yield the simple, easy answers propagated by political ideologies.