Daniel Patrick Moynihan came back to Harvard last week in search of common ground. The New York senator, the former Harvard professor, delivered three lectures on the American family, or to be more precise, on American families.

It was no coincidence that Moynihan's words came on the 20th anniversary of his famous -- or infamous -- analysis of the Negro family. In that work he described the shattering relationship between family disintegration and poverty.

This time, Moynihan reappeared as a social policy-maker in an era of limits. The problems he described have been engorged by social change. Today, the feminization of poverty, the child-ization of poverty are endemic to the whole society. A national family policy never did evolve. The rate of poverty among the very young in the United States is nearly six times as great as among the very old. In 1983, 23 percent of the preschool children lived in poverty.

But this neo-Moynihan didn't offer any massive program for change, nor did he support any social-science solution to family troubles. There is no single way for government to turn around the social forces that are affecting families. Instead, said Moynihan, "Social policy must flow from social values and not from social science."

In a flight of academic humility, the professor-cum-politician said, "We do not know the processes of social change well enough, so as to be able confidently to predict them."

What the senator tried to do then was to cull out of the morass of conflicting values and information a number of things we can, as a nation, agree upon. "What is necessary," he believes, for solving the deepest problems of poor families, "is the willingness and ability to act in some coherent manner in accordance with some coherent objective." Even in this inhospitable era, "We can act if we can agree . . . ."

If we can agree, for example, that government should not tax people into poverty, then, Moynihan said, we can enlarge tax exemptions so that once again they are large enough to push families out of poverty. If we can agree that certain government programs do work, saving lives and saving money, then we can improve on the model of Head Start or job training for women.

If we can agree that the needs of poor children aren't being met, then we can tie their benefits, like those of the elderly, the disabled and the veterans, to inflation. If we can agree that drugs are destroying families, then we can make a wider commitment to law enforcement.

This piecemeal approach to problems does not quite make a national family policy, and the senator was the first to admit that. He talked in the most general philosophical terms -- "We value self-sufficiency. We are offended by poverty" -- and in the most specific about money and programs. There was a tentative, pared-down quality to his words. No panaceas, no promises, just pieces of progress. No, we do not know everything about making families strong, he implied, but "there are places to begin."

What I liked about the speeches was that he sounded less like a politician than like an archaeologist trying to salvage pieces of common ground. In the last two decades Americans have learned what we can't do to eliminate poverty or to empower families, but in the process many have lost the belief that we can do anything at all.

The philosophers in power today maintain that government itself is the problem. But the man who coined the misunderstood phrase "benign neglect" says, " . . . no government . . . can avoid having policies that profoundly influence family relationships."

The current inertia of social policy malignantly neglects the young. Moynihan's disparate approach is not new, or neat, but it fits the pragmatism of the times; we have to focus our vision on what we do know and what does work. If we are obsessed by major arguments over "the family," then it's time to find pockets of consensus.

If we can agree, we can act, said the senator. We can help -- at least help -- the lowest and youngest quarter of our population. The problem is that first we have to agree that we want to act.