New York City has taken down much of its elevated and unsightly West Side Highway along the Hudson River and replaced it with a waterfront boulevard at a cost of nearly $1 billion. Boston is about to spend more than $2.5 billion putting its elevated Central Artery, the "green monster," underground. San Francisco's overhead Embarcadero Freeway, thanks to an earthquake, is about to descend into history.

So why won't Washington, at much less cost, tear down the Whitehurst Freeway, a deteriorating fragment of a never-finished inner city freeway network conceived in the 1940s that continues to exist as an urbanistic misfit along Georgetown's waterfront?

While other cities have recognized at long last the inappropriateness of elevated highways slicing through historic downtown neighborhoods or along scenic, economically valuable waterfronts, the District's Department of Public Works has decided to preserve the Whitehurst Freeway -- not really a freeway but a four-lane, limited access arterial street several blocks long -- and upgrade it at a cost of nearly $70 million.

It's the wrong thing to do.

To do the right thing, the city should follow the suggestions of the Committee of 100 on the Federal City and its transportation subcommittee member Joseph Passonneau, an engineer and architect specializing in transportation planning and urban design.

Passonneau and the committee argue convincingly that the Whitehurst should be torn down and replaced by a six- to eight-lane, tree-lined avenue, a landscaped, sun-filled K Street NW. According to Passonneau, a properly designed avenue would cost little more than refurbishing the existing Whitehurst structure and would carry rush-hour traffic more effectively than the Whitehurst.

Equally important, Passonneau's proposal suggests that the avenue alternative would help alleviate traffic congestion along M Street in Georgetown and actually would improve access and movement conditions at the ends of the freeway, the complex interchanges around 27th Street on the east and Key Bridge, M Street and Canal Road on the west.

These interchanges, not roadway width, are what restrict rush-hour traffic flow along the Whitehurst-K Street-M Street corridors. Indeed, widening the Whitehurst without radical changes at its endpoints will only make the freeway a larger parking lot during peak traffic periods.

Passonneau's boulevard proposal envisions four through-lanes tunneling under the Wisconsin Avenue intersection, plus direct surface connections via controlled-movement intersections to Georgetown's north-south streets between 29th and Wisconsin. A new ramp would connect Key Bridge directly to the boulevard.

This strategy would improve vehicular access to and from portions of Georgetown south of M Street where millions of square feet of new commercial, residential and parking space have been developed during the last 15 years. Further, it would eliminate the visual and psychological barrier between upper Georgetown and the Potomac River.

Passonneau bolsters his proposal with sharp criticism of the Public Works analyses and findings. He challenges the credibility and reliability of their construction-cost data, their computer modeling of traffic flow and performance, and their road design standards.

Bureaucratic momentum and vested interest in "getting on with it," in protecting previously made assumptions and policy positions, rather than objective assessment and good urban design, are driving the process, Passonneau asserts. And time for reconsideration is running out, since Public Works is forging ahead with drawings for reconstructing the freeway.

I'm no expert on traffic engineering, but even if replacing the Whitehurst Freeway with a well-designed boulevard costs more than improving the existing structure, or if traffic flow via a waterfront boulevard were less than perfect, the city still should take down the Whitehurst Freeway for aesthetic, humanistic and ultimately economic reasons.

The history of inner-city freeways, whether elevated on stilts or depressed in ditches, no matter how well designed, has taught two lessons.

First, urban freeways eventually become congested with traffic to the point of performing no better than a network of conventional city streets with timed traffic signals. Second, they are divisive, noisy, polluting eyesores that tear apart a city's fabric and compromise the amenity of neighboring properties and adjacent streetscape.

Even Hollywood knows this. Urban, elevated transportation structures routinely are used to convey the imagery of sinister, seedy, perilous environments best suited for staging car chases. It's a movie cliche born of reality.

When inner-city freeway systems were being contemplated a little less than a half-century ago, planning considerations rarely went beyond creating efficient traffic patterns, facilitating rapid movement of automobiles, and acquiring rights-of-way at least cost. Other urban design issues -- environmental impact, neighborhood preservation and revitalization, nurturing of pedestrian activity, cityscape visual character -- hardly were on the list.

Generations earlier, the same kind of narrow, pragmatic thinking produced many urban waterfronts dedicated purely to industrial and transportation uses. Along American rivers and lakes were some of America's most profitable but ugliest manifestations of human commerce, the inevitable result of treating bodies of water as nothing more than sources of energy and means of shipping.

With such traditions dating back centuries, it's no surprise that planners continued thinking in this century that derelict waterfronts constituted logical sites for elevated highways. After all, their only neighbors would be warehouses, cheap bars, storage depots and polluted harbors.

Georgetown's waterfront was no exception, once having been the site of a busy port alongside industrial buildings and railroad yards.

But land use and cultural values changed in recent decades. Today, an elevated Whitehurst Freeway fragment would never be proposed, much less seriously considered, if it didn't already exist.

It would be a great tragedy if the city compounds the egregious transportation mistake it made earlier -- not providing Metro subway service to Georgetown -- by investing tens of millions of dollars in remodeling an anachronistic transportation structure that should be scrapped and that, after remodeling, would be just as congested as it is now.

How wonderful if Georgetown finally could acquire a waterfront boulevard coupled with waterfront parkland and buildings. Like M Street with its hundreds of shops and restaurants, the new K Street would offer an alternative east-west artery along which people could drive or promenade, enjoying the river and cityscape simultaneously.

I hope the powers that be -- Mayor Marion Barry and the D.C. Council, the Commission of Fine Arts, the National Capital Planning Commission and any other agency with relevant authority -- will heed Passonneau's plea, along with those of many Georgetown residents, tenants and property owners, and see fit to reconsider plans for the Whitehurst Freeway before Public Works lets contracts or starts reconstruction. There's still time. Roger K. Lewis is a practicing architect and a professor of architecture at the University of Maryland.