It's long been an article of faith that the airwaves are a scarce resource. On this notion rides the existence of the Federal Communications Commission, which regulates the airwaves, not to mention the ownership of great swaths of the spectrum by a variety of public and private interests.

What if the scarcity turns out to be an artifact of history and outmoded technology? That's not a new thought, but it's back on the table for discussion in tech and policy circles. If scarcity can be overcome, the implications are both exciting and disruptive -- a cornucopia of communications that foreshadows woes for some of our biggest telecommunications companies. Late last month, David P. Reed gave a provocative talk to the Federal Communications Commission's Technological Advisory Council. He told the group of experts, in effect, that the FCC's fundamental mission is flawed, maybe obsolete.

Reed is no newcomer to the tech scene. He holds a Ph.D. from the Massachusetts Institute of Technology, where he taught computer science and headed the Laboratory for Computer Science's Computer Systems Structure Group. He was chief scientist at Lotus Development and Software Arts, two of the pioneering software companies, and worked at the now-closed Interval Research, the Paul Allen-funded think tank in Palo Alto. Lately he's been a consultant, entrepreneur and researcher.

He's been involved in Internet technical details for several decades, and even has a ``law'' named after him. ``Reed's Law'' isn't as famous as Moore's Law, but it's a big one. The importance of the Internet, under Reed's Law, is at least as much about the formation of groups that communicate and collaborate as about person-to-person contact.

In a panel discussion and interview last week at the O'Reilly Emerging Technology Conference in Santa Clara, Reed put in plain English some of the concepts he discussed at the FCC and which he has put online at his Web site (www.reed.com/dpr.html). Simply put, he said, we have to start looking at spectrum as an almost limitless commodity, not a scarce one.

The current regulatory regime that allocates spectrum ``is a legal metaphor that does not correspond to physical reality,'' he said.

Why not? First, he said, the notion of interference has more to do with the equipment we use to send and receive signals than with the physics of radio waves.

``Radio waves pass through each other,'' Reed said. ``They do not damage each other.''

In the early days of radio, the gear could easily be confused by overlapping signals. But we can now make devices that can sort out the traffic.

The second way that reality defies the old logic is what happens when you add wireless devices to networks. I won't go into the details of Reed's argument, which you can find on his site, but he contends that you end up with more capacity -- the ability to move bits of data around -- than when you started.

``In principle, the capacity of a certain bandwidth in a certain physical space increases with the number of transceivers in a given space,'' he said. Yet the FCC regulates the airwaves as if the capacity was a fixed amount.

Yes, he said, this is counter-intuitive. And, to be sure, there are experts who disagree with him.

But if he and others in his camp are right, we have a lot of work ahead to fix a hopelessly broken regulatory system. And if that happens, the sky is literally the limit for future communications -- but the consequences for some of the most powerful companies in our economy may be grim.

Reed wants the FCC to open up some spectrum for these more open wireless networks, giving entrepreneurs a new public space in which to innovate and create value for the rest of us. He's not sure who'll make money in this space, but surely equipment manufacturers and other companies, especially software companies, will be in the middle of a wave of innovation.

Software is a key, perhaps the key, to the future Reed envisions. Most radio-like devices using today's spectrum -- radios, televisions, mobile phones and the like -- are based on the old way of doing things, constrained by hardware to receive and transmit signals in specific ways and in specific places of the airwaves.

To get the capacity multiplier effect, he said, we need devices with fairly generic but powerful hardware components. ``Software defined radios'' will be vastly more adaptable, and useful, than their old-fashioned cousins, according to Reed and others who are promoting the concept. The military has been using these devices, also called ``agile radio,'' for some time; civilian availability is getting closer as costs come down.

Who stands to lose? Apart from regulators whose jobs might be largely unnecessary, consider the potential plight of the phone companies. Their business model is based on economics that Reed's notions, should they become reality in the marketplace, would shred.

Getting from here to there is a huge, perhaps insurmountable task given the business interests that would object to changes in the rules. Some regulation would still be necessary in at least some areas, no doubt.

Imagining this new world has another attraction. It conjures a boost for a civil liberty we take for granted in America but which has been dampened under the current regulatory scheme.

I'm talking about free speech. Regulation of the airwaves has specifically included curbs on speech, such as the FCC's commands to the nation's TV and radio broadcasters about what may or may not be said on the air.

Restrictions on speech have been justified under the idea that the spectrum is a public and limited resource. If that is not true, there's no reason to regulate speech in this way. Maybe, someday, the First Amendment will mean something when people broadcast their views, not just when they put them on paper or on the Internet.

The worst direction for the FCC to move right now, Reed said, is to keep giving or auctioning spectrum to ``monopoly owners'' that won't use it efficiently. A new kind of open space is all about the public good, he said, and there's a fine analogy in recent history.

``We need to do for spectrum,'' he said, ``what the Internet did for the network.''