Invent a new communications technology recently? If so, beware: the U.S. government may require you to build it in a way that will enable federal agents to eavesdrop by court order. Otherwise, the New York Times’s Charlie Savage reports, you’ll end up paying a court-ordered fine.

If the requirement sounds reasonable to you -- maybe you think you really “didn’t build that,” after all -- then think again. Would the mail system, which was once private, have been invented if its creators had been required to facilitate censorship on every delivery route across Europe? What about writing itself? Would you have invented it if you had been forced to teach some Sumerian censor to read, and make every clay tablet accessible to his (court-ordered) glance?

Of course, the U.S. has a long tradition of government wiretapping of phone conversations. But it’s important to note that this practice came about only after the telephone had been invented. It is a very different thing for the government to regulate existing technology than for it to dictate before the fact what technologies may be permitted to exist.

Although the new proposal being vetted by the White House doesn’t prohibit outright any technology that can’t be tapped by the government -- an earlier proposal did -- the substitution of fines for a ban doesn’t help much. If the fines are large enough to be effective, they will have the same effect as a ban: to create an incentive structure in which innovators cannot put into practice any technology that wouldn’t be susceptible to government surveillance.

To see why this proposal would stifle innovation, consider that centralized technologies and surveillance come out of the same intellectual and technological milieu. The phone line must run through an interchange, and that is where a classic phone tap was set up. The bottleneck was the locus of government control in exactly the same way that the bottleneck was the key to the functioning of early telecommunications technology: The switchboard was all.

Today, decentralization is the name of the technological game, from the famously decentered Internet itself to newer communications techniques that run peer to peer, often accompanied by decentralized encryption.

Central Control

Yet the modern government isn’t decentralized in the same way -- and it never will be. Federalism, the decentralizing impulse of the late 18th century, didn’t really withstand the Age of Henry Ford. A government that was going to regulate industry and labor, and fight wars abroad, had to be more centrally powerful. The Federal Bureau of Investigation eventually followed along with it. While local law enforcement still does the vast bulk of ordinary policing, it typically lacks the capacity to do national, let alone international, investigations of the kind necessitated by networked crime.

A centralized security apparatus cannot fully monitor decentralized communications. So the government’s impulse will be to order the innovators to do its will -- by reasserting some option of centralized surveillance. The cost to innovation can be significant, since a whole new direction must be made to conform to old norms.

The government’s best argument for potentially standing in the way of innovation is, as always, necessity: It’s a dangerous world, and someone has to police it. As a matter of cost-benefit, then, maybe the loss of technological and commercial progress is outweighed by the gains to security.

Don’t bet on it -- or at least, don’t assume this argument is any truer in this situation than the many others in which it is made by the state. Criminals and terrorists operated perfectly well before the Internet, and they will continue to do so even if their communications aren’t totally secure. Indeed, the smarter the bad guys get about realizing there is no true privacy on the Web, the more they will resort to personal, face-to-face communications that will make their networks harder to break. Osama bin Laden managed to hide for as long as he did after the Sept. 11 attacks partly because he was from a generation that was able to maintain telecommunications discipline.

The government’s impulse here is driven less by a particular security need than by the powerful state interest in total surveillance, even where privacy rights demand a warrant. It is the idea, not the reality, that is doing the work here -- and the idea is that someone somewhere might be able to communicate without the government being able to listen.

It is sometimes suggested that the Web represents its own civilization with its own anarchic structure -- Google Inc.’s Eric Schmidt and Jared Cohen make the argument in a much-hyped new book, “The New Digital Age.” But old-fashioned state power stands in the way of such an independent existence.

The state is not going to wither away, and that’s a good thing. Yet we should be able to make sure it has good reasons for restricting liberty and innovation -- or else that state will cease to be our technology of governance, and become something else entirely.

(Noah Feldman, a law professor at Harvard University and the author of the forthcoming “Cool War: The Future of Global Competition,” is a Bloomberg View columnist. Follow him on Twitter. The opinions expressed are his own.)

To contact the writer of this article: Noah Feldman at noah_feldman@harvard.edu

To contact the editor responsible for this article: Tobin Harshaw at tharshaw@bloomberg.net