Men running for office live and die by the polls, but the truth is that polls cannot measure the future; they are at best a measurement of what has already happened and, at worst, can be misinformation about the cause and effect process they attempt to test. Take a photograph of a train in motion and try to guess how fast it is going; bet the odds at the race track all the time and see how much money you make; examine a colony of ants under a microscope and you may see what they are doing, but you'll never discover why they are doing it.

To say that polls are limited, however, is to tell only half the story. What they have become is important weapons in the perception war that takes place in political campaigns. In an era in which at least seven national polls are discussed periodically in the newspapers and on the television networks, the perceptions that come from people's readings of the polls take on a life of their own -- one that may have little to do with the facts contained in the polls themselves.

An example can be seen in the mythology surrounding the political history of the Vietnam War. It is accepted fact today that Eugene McCarthy raised the issue of the war in the New Hampshire primary in 1968 and that the substantial vote he received in that primary was an indictment of our participation in the war. Well, the Nixon campaign did some post-primary polling in New Hampshire and found that 60 percent of the people who voted for McCarthy did so in spite of his position on the war and another 12 percent did not know that McCarthy and Johnson differed on the war. That left only 28 percent of the McCarthy voters who might have voted for him because he was against the war.

McCarthy got about 40 percent of the total Democratic votes in New Hampshire. Twenty-eight percent of that is 11.2 percent, a figure that was not out of line with what national polls were showing as the true opposition to the Vietnam War. Yet because McCarthy campaigned against the war, the assumption was made that a vote for McCarthy was a vote against the war. The truth was that Johnson was a despised man, and voting for McCarthy was a way for voters to show their displeasure without having to worry that McCarthy might become president.

Today, it would be impossible to separate fact from fiction about the political history of the Vietnam War. Besides, it doesn't really matter. The perception that the war might be unpopular grew until that perception replaced the facts. In politics, unchallenged perception becomes fact and, once perceptions are established, it is unprofitable to try to go behind them.

In the present race, all polls agree that Carter is in trouble because of his handling of the economy. But the real question is whether Reagan can establish and so ingrain the perception that people will vote for him because of it. So far, he has not done this. His early mistakes, plus the time spent in pursuing Carter on the Stealth bomber leaks, have detracted attention from the economy as an issue.

Similarly, Carter's poor performance on energy cannot be exploited unless Reagan's own credentials are established and, regardless of how strongly the polls show this country's prestige to have suffered from Carter's stewardship of foreign policy, Reagan will not win votes from him unless he is perceived as better able to handle the Russians, ready to provide the proper blend of strength and restraint and capable of foreseeing upcoming dangers.

I once caused quite a commotion at a gathering of political scientists by maintaining that the Founding Fathers first decided to revolt and then approved the Declaration of Independence. I was attacked for suggesting that the result had preceded the reason and that the statements of principle contained in the Declaration were in some respects an afterthought. Well, one good instinct of Americans is that they sense what the right thing to do is well before they can prove it. As much as the science of polling would have you believe that voters make up their minds by a rational process, it is, in fact, an emotional process.

Right now, a majority of the voters are fighting the idea that they must tolerate another four years of Carter as their president, but Reagan is not giving them enough help. There is still time for him to do so, but with only six weeks left in the campaign he has a lot to do.

Campaigns are won by a mysterious thing called momentum. It is a confidence that is transmitted by a candidate to his supporters and radiates out from there to engulf the undecideds and switch an opponent's soft supporters. During the last phases of a campaign, you can feel it, if you have it, every time you walk into a room full of people. The electricity is there and the excitement of anticipated victory takes hold.

Reagan must seize the momentum by the middle of October or it will swing to Carter because of his incumbency. If Reagan doesn't grab it, people will stop fighting the idea of another Carter term and give in to it.

You see, the confidence that sets momentum in motion is actually the self-expression of a candidate's will to win; it is an extra reserve that wells up from within. Pollsters cannot measure momentum, they can't predict it, they don't understnd it. They can only try to explain it after it happens. It is always humorous to hear pollsters trying to rationalize momentum. You can't rationalize a hurricane, you can't capture the wind and you can't beat a man who won't be beaten.