More than a decade ago, my Post colleague Joel Garreau gave people a new way to think about urban development when he coined the term "edge cities" -- those suburban clumps of high-rise office buildings, shopping malls and clustered housing that sprouted out of the suburbs in the 1980s. Prime examples were Tysons Corner and the area around White Flint Mall at I-270 and the Beltway.

Edge cities challenged the notion that the benefits of density could be achieved only in traditional downtowns. Sure, many of them developed into soulless cultural wastelands plagued by endemic traffic jams. But Garreau's point was that edge cities were the wave of the future, a natural evolution of the urban idea in response to changing economics, technologies and lifestyles.

Now comes another local, Robert Lang, director of the Metropolitan Institute at Virginia Tech's Alexandria campus, arguing that edge cities may have been nothing more than a passing phenomenon. In a new book titled "Edgeless Cities," published by the Brookings Institution, Lang argues that most suburban economic growth over the past 30 years has occurred outside any identifiable cluster, in low-slung office buildings, strip shopping centers and townhouse developments scattered along secondary highways.

"Edge cities have lost their edge," declares Lang, who predicts that the future of the American metropolis is one of continuous sprawl and random development.

Recent census figures seem to confirm Lang's prognosis. As reported in yesterday's Post, two-thirds of the Washington region's population growth in the past year was in the outmost counties such as Virginia's Loudoun and Stafford counties. Many of the residents there are commuting from bedroom communities to jobs in Washington or closer-in edge cities. But once these communities achieve a critical mass, history shows it is only a matter of time before malls, restaurants and office buildings will follow. As Lang sees it, these activities are unlikely to cluster themselves naturally into nodes that resemble anything like a traditional city.

"Edge cities are perhaps the last stop on the road away from traditional urban forms," writes Lang. "One day, edge cities may be seen as a transitional urban form -- an attempt to build auto-based, low-density downtowns before developers realized that for the most part, cars made such places unnecessary."

Nonsense, replies Garreau, who is writing a book on the search for modern-day utopias. He remains convinced that the advantages of density, geographic proximity and face-to-face contact are profound and enduring and can be realized only in urban environments.

At the intellectual level, this debate tends to be waged over semantics, competing data sets and historical metaphor. The stakes are significant, however, not just for real estate developers trying to decide where to make the next investment but also for citizens and local officials trying to decide how to manage growth.

There seems no doubt that the stronger economic forces these days push in the direction of sprawl, even as political forces are marshaling to push back the other way. The result is likely to be a metropolis organized spatially, more like Los Angeles and suburban New Jersey than Chicago or Philadelphia.

For fast-growing places like Loudoun and Stafford counties, I suspect that means that the current campaign to use zoning and other restrictions to stifle growth is likely to be as successful as shoveling sand against the tide.

But those communities still have the opportunity to use a variety of tools -- including, in particular, major investment in public transit and sewer and water systems -- not to stop growth, but to channel it into well-planned edge cities that avoid some of the mistakes of the past. Think of the recent successes of Ballston or the main streets of Reston and Bethesda.

As Garreau pointed out yesterday, just as Rome wasn't finished in a day, we shouldn't expect Tysons Corner to be any different.

The edge city may simply be a work in progress.

E-mail Steven Pearlstein at pearlsteins@washpost.com.