A wit with a flair for euphemism, and an unrealized vocation for politics, once defined a fib as an artistic molding of the unshapely clay of truth. Following the president on television Tuesday night, Rep. Dan Rostenkowski (D- Ill.), chairman of the House Ways and Means Committee, was forced by circumstances to do some artistic molding.

Rostenkowski is built along the lines of a cement mixer, and like most cement mixers he is more methodical than spectacular. He is a graduate of the Richard Daley School of Rhetoric, which stresses succinctness, as in: "Here's yours. Where's mine?"

Tuesday night, Rostenkowski had a tough act to follow and did it artistically, saying: Tax reform is our -- the Democrats', damn it -- idea. Look here, Truman praised it in 1948.

The rather-more-than-hairline fissure in the foundation of his argument is this: since Truman spoke, Democrats have controlled the House -- in which the Constitution vests the power to originate revenue bills -- all but two years. The senator most important in shaping the tax code has been a Democrat, Louisiana's Russell Long, who until 1981 was chairman of the Finance Committee.

The pedigree of the idea of tax reform is less important than the fact that the president has got everyone's attention. But can he keep it? In 1861, Tolstoy visited Turgenev, who handed Tolstoy a manuscript of "Fathers and Sons." Tolstoy read a few pages -- then fell asleep. The tax-reform proposal may be, like "Fathers and Sons," a masterpiece. But public attention is hard to hold with anything short of a war or a hot "yes or no" like, say, the Panama Canal treaties.

It is doubly difficult to hold attention with an issue that is a salad of smaller issues, most of them as strange as Mongolia to most Americans. And when public attention wanes, the Law of Intensity-Over-Numbers takes hold. That law is: Democracy is superficially about numbers, but essentially about intensity. For example, there are 235,873,426 (well, approximately) cheese eaters in America, but a lot fewer cheesemakers. Yet there are cheese-import quotas, because cheesemakers are organized and intense about cheese.

Another problem for tax reform is that the game of legislation is more like football than baseball: there are too many huddles, and the game is a slave to a clock. When Congress reconvenes on June 3, there will be 78 legislative days remaining on this year's calendar. The two tax-writing committees could spend all of those days sifting evidence about the possible consequences of any one of a dozen important proposed changes of the tax code, a code that is woven into the expectations of various industries.

Congress dare not delay, yet cannot responsibly act with dispatch. Delay will subject whole industries -- real estate, for example -- to paralyzing uncertainties as investors hesitate, awaiting clarification of tax treatment of investments.

The president's proposal for less-generous depreciation schedules probably would require many owners of buildings to raise rents to recoup losses. The variety and severity of the secondary consequences of this are unpredictable.

To the close observer of government, or the conscientious participant in it, nothing is more consistently dismaying than the thinness of the information on the basis of which officials must act. But the legislative timetable and the need to minimize uncertainty impose strict limits on how long things can be prudently studied.

Delay spreads economic paralysis, but dispatch amounts to leaping into the dark. The restaurant industry is one of the largest employers of minorities and unskilled labor. No one can know the extent of the damage restaurants will receive from restrictions on the deductibility of business spending. Perhaps we ought to go ahead and find out. But some people are going to get hurt in the learning process and, as usual, the allocation of pain among the social classes will be regressive.

It is in the nature of things -- of complex industrial societies, of the limited predictive tools of social analysis -- that no amount of study will produce a well-founded consensus about the consequences of the reform plan. Besides, even if study did produce such consensus, there would be no consensus about how to assess the equity of the consequences. So just keep saying, "You can't make an omelet without breaking eggs," and keep hoping you are not an egg.