There was an innocent time when almost any ''civil rights bill'' could be debated with candor and clarity. That usually meant that the legislation could safely be supported without fear that it would establish one injustice to cure another. But that is far from true of the more controversial provisions of the proposed Civil Rights Act of 1990.

Why? A clue appeared in a recent review in The Post's Book World. David Garrow, reviewing Hugh Davis Graham's ''The Civil Rights Era,'' was struck by ''the extent to which the EEOC {the federal agency that monitors employment discrimination}, with a helping hand from Warren Burger's Supreme Court in the 1971 case of Griggs v. Duke Power, was able to transform to the point of outright reversal some of the statutory language of the 1964 {civil rights} act.''

In that startling but candid nutshell lies a brief history of the recent trend in job-bias litigation and enforcement. A nation that had repudiated racial quotas at every legislative opportunity has cheerfully accepted them in practice. Whether the discrepancy between promise and performance was wholesome or harmless (or to what degree a bit of both) is debatable. That is why the current argument over the 1990 bill's ''correction'' of recent Rehnquist Court decisions has been so murky.

What is scarcely debatable is that the congressional compact that accompanied the passage of Title VII of the 1964 act, banning employment discrimination, has been greatly modified if not shattered by actual enforcement practices.

That is demonstrable historical fact. The high principles of racial impartiality enunciated by Congress in past civil rights acts tend, in judicial and bureaucratic practice, to slide into mechanical guidelines -- to which, in turn, the adoption of de facto quotas (usually euphemized as ''goals'') is often the line of least resistance for business, industry and educational institutions.

Implicit, if unavowed, quotas are often the least costly and abrasive way to apportion jobs and entitlements to the historically disadvantaged. And employers tend to adjust their compliance practices to government policy in ways that cause them the least pain and inconvenience. This is the way the world works; or at least the way it worked until last spring, when the Rehnquist Court decided Ward's Cove v. Atonio.

But the obvious is still denied. For instance, William T. Coleman, a distinguished lawyer who chairs the NAACP Legal Defense Fund, has argued as follows in defense of the 1990 civil rights bill: ''Many of the civil rights acts of our own era were reviled as covert attempts to guarantee minorities quotas in employment, education, public office, or other aspects of American life. Happily, Congress and the country paid little heed to these accusations.''

Bill Coleman is right on the first point but far astray on the second. Foes of civil rights legislation habitually charge that ''reverse discrimination'' or ''favoritism'' to blacks and others is the true aim. But Congress has never ignored these charges, and certainly did not ''pay little heed'' to them when they were widely heard in 1964. It is fair to say that this was the chief issue regarding Title VII.

As Senate floor manager, Hubert Humphrey gave fervent and explicit assurances that the aim was not to create hiring quotas. In those more candid days, to treat reverse-discrimination charges cavalierly would probably have doomed Title VII to failure.

Last spring's Ward's Cove decision shifted the burden of proof from employers to plaintiffs, and toughened the standards for proving job bias on the basis of figures. In other words, it moved the law back toward the ''colorblindness'' ideal of 1964, sacrificing effect to principle. The 1990 ''correction'' bill would shift the law back toward the practice prevailing before last year's decision, sacrificing principle to effect. It would almost certainly encourage the adoption of de facto quotas in many situations.

If the social or legal wisdom of one course or the other were starkly obvious, Congress would not be worrying over this bill. And judges wouldn't have been quarreling with one another over the issue for the past quarter century. It is no service to public interest to pretend, as some defenders of the bill do, that its effect on the law would be trivial.

For years we rather assumed that affirmative action could serve, and would be congruent with, the goal of colorblindness. Experience in our claimant society has shown that to be a pleasant delusion. That is what all the shouting is about.