THE U.S. GOVERNMENT seems to be changing its mind about the possible dangers of low-level radiation. After several decades of prevarication, evasion and positive thinking, the government is now prepared to acknowledge that some of the key questions are not only unanswered but, perhaps, unanswerable. In the past, most officials have felt that public discussion of these risks could only inhibit weapons research and, more important in recent years, the development of nuclear power. Our own view is precisely the opposite -- that the past failure to deal openly with radiation issues has contributed greatly to the present public suspicion of nuclear reactors.

This turn in federal attitudes began, apparently, with President Carter's decision last spring to ask for another study. This time, significantly, the responsibility was lodged with Hoseph A. Califano, the secretary of health, education and welfare, rather than with any of the agencies running weapons or energy prodrams. Most of the task-force committees have now made their reports, and HEW has published a summary that is notable for its balance and clarity.

The heart of the matter is the relationship between radiation, in small doses, and cancer. Earlier in the nuclear age, most specialists believed that there was a certain threshold of exposure below which radiation was harmless. After all, half of the radiation to which the average American is exposed comes from natural sources that are as old as the universe. (Most of the other half, incidentally, comes from medical uses.) Since people have been living with this background radiation from the beginning, it seemed reasonable to think that the race had adapted to it. But now, HEW reports, some specialists have begun to belive that the danger of cancer is disproporitionately greater at low exposures. This whole subject is highly uncertain and controversial. But several studies suggest that natural background radiation "may account for between 50 and 70 percent of all cancer, instead of the one percent predicted by current assumptions."

That possibility has several diffeent ranges of consequences. It raises a need for new caution and restraint in even the most routine medical and dental X-ray procedures. It also suggests that the country has a heavy moral obligation to the people -- some of them soldiers, some of them workers in the nuclear industries -- who were exposed to substantial amounts of radiation in the days of the early, inadequate standards of protection. The United States was testing bombs in the open atmosphere until 1962 and sometimes soldiers were posted close to these explosions. Now, a quarter of a century later, cancer is turning up with dismaying frequency among those men. There is no way to distinguish between the cancer caused by radiation and the cancer caused by any other influence. That makes it very difficult to prove legally that any individual veteran's illness is the direct result of a bomb test. But the statistical evidence suggests a large national responsibility to the people who worked close to radioactivity in the early days of the threshold theory.

For the future, there is the question of which agency and which interest should control the larger federal studies of radiation's effects that will follow this preliminary survey. It should absolutely not be any agency with an interest in nuclear power -- like, for example, the Energy Department. There is nothing in the Hew/ report that rules out power reactors, although it certainly underscores the importance of safety precautions. Even at the highest current estimates, the radiation risks of the nuclear reactors are substantially less than the health risks from air pollution of conventional coal-burning generators. But we regard that question of relative risks as one that needs to be left open for continual review. If the results are to be persuasive, Mr. Carter had best leave control of this research in the hands of HEW and its cancer epidemiologists.