PRESIDENT BUSH has done his bit to reinforce one of our hoariest myths about the Supreme Court: that you really can't pack the Court because there's no way to predict how a nominee will vote, once granted life tenure and confronted with the awesome responsibilities the job entails.

Like all myths, this one has some truth in it. It especially fits the case of Judge David Hackett Souter, the Bush nominee whose Senate confirmation is expected soon after the Court begins its new term tomorrow.

In his testimony to the Judiciary Committee this month, Souter came across as smart, fair-minded and moderate-spirited -- neither liberal crusader nor conservative revanchist. But his delphic responses to specific questions, like his prior record, provide remarkably little basis for discerning his leanings on abortion, affirmative action, church-state relations and other great issues that divide the court and the nation.

If Souter votes to reaffirm broad judicial protection of abortion rights, Bush can give a shrug, disavow (again) the anti-abortion "litmus test" for judges in each GOP platform since 1980, and say something like, "Just goes to show, you never can tell what these Supreme Court guys are gonna do."

But history teaches that you usually can tell what a nominee will do on the issues that loom large at the time of appointment -- that is, if predictability is what you're looking for. The trick is to find someone whose philosophy is well thought out and whose opinions are known.

That's what Franklin D. Roosevelt did when he appointed eight supporters of his New Deal programs to the Court between 1937 and 1941. That's what presidents including Washington, Jackson and Lincoln did earlier to promote their own respective visions of the Constitution.

And that's what President Reagan did when he moved the court sharply to the right in the 1980s, by elevating Justice William H. Rehnquist to chief justice and adding Justices Sandra Day O'Connor, Antonin Scalia and Anthony M. Kennedy.

Scalia, for example, has consistently called for overruling Roe v. Wade, crusaded against racially preferential affirmative action, opposed strict separation of church and state and championed sweeping presidential powers. That's just what Reagan wanted him to do, and what those familiar with his record predicted when he was chosen in 1986.

That's also what Robert H. Bork would no doubt have done but for the newly Democratic Senate's 58-42 rejection of his nomination three years ago.

The ideological screening was less rigorous in the case of O'Connor, whom Reagan chose more for her sex and extrajudicial accomplishments than for her philosophy, and Kennedy, whom Reagan chose for ease of confirmation after Bork was rejected. But it was predictable that O'Connor and Kennedy would bolster the Court's conservative wing, as they have done. Reagan succeeded better than any president since Roosevelt at using his appointment power to change the court's direction. In part because of his success, however, blank-slate nominees like Souter may become the norm in the coming years -- and "the myth of the surprised president," as Harvard law professor Laurence H. Tribe calls it, may become the new reality.

This is not because of anything inherently unpredictable about Supreme Court nominees. It is because, in an era of divided government and ideological polarization over the court, unpredictability has become a sine qua non for avoiding a bloody confirmation battle. Bush might have pushed an avowed critic of Roe through the Senate, but only after a bloody battle that would have cost him dearly in popularity and political capital, commodities that Bush has proven most reluctant to spend. Instead, Bush chose a well-credentialed nominee whose views on abortion and just about everything else of current controversy were unknown.

This choice of a "stealth nominee," with the enthusiastic backing of White House Chief of Staff John Sununu, the conservative former governor who had put Souter on the New Hampshire Supreme Court, spawned suggestions that Bush or Sununu must have had some secret assurance that Souter would vote to overturn Roe.

The suggestions were both too cynical and not cynical enough: Too cynical because it is far-fetched to assume the president, his aides and Souter himself all conspired to lie to the Senate and the public; not cynical enough because it is naive to assume that Bush cares deeply about getting Roe overruled.

To the contrary, the record, beginning with Bush's flip-flop from his early 1980 position that Roe v. Wade "was right," suggests his position on abortion is shaped not by moral conviction but by political calculation. While Bush's anti-abortion posture is still essential to sustain his support among conservatives, his political fortunes might well be hurt if the court overruled Roe and further energized pro-choice voters who have already sent Republican candidates scrambling for cover.

That's why Bush's look-Ma-no-litmus-test approach makes perfect political sense. The surest way to insulate himself from blame for whatever Souter ends up doing is to make a virtue of ignorance.

The court's current precarious balance, combined with the constraints on a Republican president seeking to get a nominee through a Democratic Senate, make Bush's deliberate choice of such an unknown unusual. But Bush is one of a long line of presidents who have had more immediate priorities in choosing nominees than influencing the court's ideological direction. Into this category fall most -- though not all -- of the presidents who have been chagrined by actions of people they put on the court.

President Eisenhower, for example, complained of the liberalism of Chief Justice Earl Warren and Justice William J. Brennan Jr. But Eisenhower did not choose Warren and Brennan for their philosophical leanings. His appointment of Warren in 1953 was driven by politics, including the need to pay a debt to Warren, a popular governor of California who had helped deliver the Republican nomination to Eisenhower in 1952.

Eisenhower's choice in 1956 of Brennan, a highly regarded member of the New Jersey Supreme Court who was also a Catholic, working-class, liberal-leaning Democrat, was designed to broaden the president's appeal in an election year.

It was President Truman who said that "packing the Supreme Court simply can't be done . . . . I've tried it and it won't work . . . . Whenever you put a man on the Supreme Court he ceases to be your friend."

Despite his complaint, the four undistinguished personal and political cronies he put on the court generally supported him on domestic issues. His disappointment came on an issue nobody was thinking about when Truman made his appointments: Two of his appointees joined in the landmark 1952 decision to strike down Truman's seizure of the nation's steel mills during a Korean War labor dispute.

President Theodore Roosevelt, irked by the vote of his appointee Justice Oliver Wendell Holmes Jr. in an antitrust case, complained, "I could carve out of a banana a judge with more backbone than that." But that was just one case; over the years, Holmes compiled a record of brilliant progressivism of which Roosevelt should have been proud. It has been more rare for a justice to change a position that a president had specifically relied upon in choosing him. It has happened, however.

In 1864, when the death of Roger B. Taney opened up the position of chief justice, Lincoln was anxious to find a man who would uphold the "greenback" legislation that had financed the Civil War.

"We cannot ask a man what he will do," Lincoln told a confidant, "and if we should, and he should answer us, we should despise him for it. Therefore, we must take a man whose opinions are known."

Lincoln took Treasury Secretary Salmon P. Chase, a prime mover of the greenback legislation. But Chief Justice Chase saw the issue differently than had Secretary Chase. He wrote the decision striking down the greenback law in 1870.

More typical was the experience of President Nixon. He chose his four appointees primarily with a view to cutting back on Warren court decisions that had vastly expanded the rights of criminal defendants. Nixon's appointees did just that.

But quite apart from their role in the 8-0 decision ordering Nixon to turn over the Watergate tapes, three of them (all but Rehnquist) also proved surprisingly activist in other areas that Nixon was not thinking of when he chose them. Most dramatically, in 1973 Chief Justice Warren E. Burger, Justice Lewis F. Powell Jr. (both retired now) and Justice Harry A. Blackmun all voted to legalize abortion in Roe.

Blackmun was a predictable conservative when he first joined the court, but has moved across the spectrum and is now considered a liberal. But this was in part because public debate has come to focus on abortion, affirmative action and church-state issues that were in the background when Blackmun was appointed. And even he has rarely if ever abandoned strong convictions on issues that were apparent when he was appointed.

So history suggests that President Bush could have found a predictable vote to overrule Roe (or to uphold it, for that matter) if he had really wanted to. But it would be unwise to demand that kind of predictability on an issue that, given pro-choice successes at the polls and development of an abortion pill, may soon cease to dominate debate about the court.

The nation will be better served by a justice who will give both sides the kind of sympathetic, open-minded hearing that Souter promised than by one equipped with a bundle of preconceived views on the issues of the moment, frozen in the certitude that guarantees predictability.

Learned Hand, one of our greatest judges, said it best: "The spirit of liberty is the spirit which is not too sure that it is right."

Stuart Taylor Jr. is a senior writer for American Lawyer Media.