Let us be optimistic for a moment and assume that the president follows through on his State of the Union pledge to propose a comprehensive national energy strategy. In the much-needed debate that will follow, the hottest and most perplexing charges and counter-charges will be aimed at nuclear power.

One side will say that efficiency and conservation can only go so far and that the only sure answer to pollution, greenhouse warming and oil import dependence is nuclear energy. The other side will assert that nuclear is the highest cost option among many and that safety, proliferation, waste disposal and other considerations make nuclear power a last resort at best. One thing is certain, the vehemence of the arguments will bear no relation to the judgment the marketplace has rendered.

The last time a nuclear reactor was ordered that was not later canceled, Spiro Agnew was vice president, the Vietnam War was still being fought and Three Mile Island was six years in the future. It was 1973. Yet no subject so reliably brings discussions of energy policy to a screeching halt.

As hard-core proponents see it, the nuclear industry has been the victim of a conspiracy they variously attribute to the president, to Congress, to a biased elite that has misled an otherwise pro-nuclear public or to a strain of "irrationality" that infects public opinion. The other side sees a uniquely favored technology that was allowed to sidestep political checks and balances through secrecy and a special regulatory arrangement and grew into a sloppy industry incapable of ever managing a demanding technology.

A few battle-scarred souls believe there is a middle ground. In their view, nuclear power may be an important element of the nation's long-term energy mix if it can regain public trust and lower its costs. Changes in federal policy, such as streamlining the licensing process and requiring a standardized reactor design, are worthwhile but will contribute less than the industry hopes to either goal. Regaining trust will be achingly slow. Attempts to force the pace, for example by excluding public intervenors or rushing to choose a new reactor design, are guaranteed to backfire. There must also be a solution to waste disposal. Most sensibly, that would entail dropping the ludicrously inflated official goal of assuring the wastes' safety for 10,000 years in favor of an achievable target, but doing so may be politically impossible.

The key to a saner future life with nuclear energy is understanding what went wrong. From the beginning, neither government nor industry took this technology seriously enough. Some in the industry still don't. They put too much effort into buying poll after phony poll showing overwhelming public support for nuclear energy and too little into the demanding task of technical and managerial reform. Despite efforts to tighten up after Three Mile Island, the industry is still plagued by utilities that simply shouldn't be in the nuclear business. Not long ago regulators discovered that a plant in Pennsylvania was being run by operators who slept and played video games while on duty, apparently with management's knowledge. Until such incidents are only dim memories, the industry need look no further for the source of its woes. As Peter Bradford, chairman of New York's Public Service Commission, bluntly puts it, "The lesson that Wall Street learned from Three Mile Island was that a group of federally licensed operators -- not appreciably better or worse than any other crew -- could convert a $2 billion asset into a $1 billion cleanup job in about 90 minutes."

Government bears a heavy responsibility for the industry's poor record. Victor Gilinsky, a former federal regulator, points out that the first medium-sized commercial reactor was licensed in 1964. Before it had operated even a single day, 38 larger reactors had been licensed. Most of today's 115 reactors were licensed and built at the same time, leaving no opportunity to learn from others' mistakes. Worse, the great majority of nuclear utilities operate just one or two plants, which means they also have had no opportunity to learn from their own experience. It is no coincidence, says Gilinsky, that the most notorious problem plants -- Diablo Canyon, Shoreham, Seabrook, Zimmer and others -- were each their respective utility's first commercial nuclear project.

The solution is to consolidate the more than 50 nuclear utilities, leaving fewer and better management teams each in charge of a sizable number of plants. That process could also ease safety concerns by removing some of the chronically poor performers, which would in turn lighten the regulatory burden for all.

Even after all this has been achieved -- a perfect operating record, improved management, growing public confidence, more efficient regulation, a functioning waste disposal system and an improved reactor design -- a question mark will remain. Will nuclear power be more or less costly than other means of supplying electricity, including efficiency improvements? No one knows. Today it is one of the most expensive options. Reforms should lower costs, and pollution surcharges for fossil-fired plants will help nuclear compete, but the ultimate outcome is impossible to predict. It will be about a decade before reactor construction resumes, if it does. There is time to turn down the heat. If the industry can end its fruitless hunt for villains and its long habit of promising more than it can deliver, and if nuclear opponents can relax long enough to allow regulatory changes that will improve performance, a way might yet be found to come to terms with this promising, troubled technology. The writer, vice president of World Resources Institute, writes this column independently for The Post.