On CNN, she was a round blue blob, haloed by dark hair, hovering over a strand of pearls, expressionlessly emitting rape accusations, memory lapses, whimpers, and sobs.

On Court TV, she was a flickering checkerboard mosaic, a high-tech mask.

She had a human face and a name only for the six jurors and the handful of spectators squeezed into the tiny Palm Beach courtroom. The rest of us could catch fragments of her identity only when the TV people mistimed the annoying bleeps they used to censor out each utterance of the name of history’s most famous rape victim.

Or is she just a famous perjurer?

We couldn’t see whether her gaze was steady or shifty, whether those sobs looked as genuine as they sounded, whether the woman herself was more believable (as some who saw her say), or less, than the disembodied voice behind the blob.

Was all this really necessary, while she was face to face with her alleged rapist in the courtroom, trying to send him to prison and destroy his reputation forever? Is it still necessary, after the jury has found William Kennedy Smith not guilty in just 77 minutes’?

We all knew his name. We all saw his face. And while he was legally presumed innocent unless and until proven guilty, no one was showing much solicitude for his privacy. Wherever he goes, people will know him as the alleged Palm Beach rapist.

So why should she be shielded from public view while testifying at a public trial? Why, indeed, so many months after everyone in Palm Beach knew who she was, and after two national news organizations first used her name?

Protecting a rape complainant’s anonymity (if she wants it protected) is, in fact, very good journalistic policy, up to a point. But in my view (not my employer’s), that point was reached in this case the moment she appeared on the witness stand-or, if not then, at least by the time that the jury had found Smith innocent.

That’s because those who still protect her anonymity now thereby signify that they still view her as presumptively a victim-and him as presumptively a rapist.

Must a man accused of date rape be journalistically presumed guilty for the rest of his life? Even after, as here, he goes to trial, asserts his innocence, gives an account more plausible than his accuser’s, and is acquitted?

Even after the defense not only creates a more-than reasonable doubt, but also makes a fairly persuasive case that the "victim" is really a perjurer?

Of course as in most "she said, he said" cases there is no way for any objective observer ever to be sure whether she or he, if cither, is telling the truth. We may never know whether the Palm Beach complainant is the victim of a monstrous wrong or the perpetrator of a monstrous lie. Or whether both are lying, and the facts are somewhere between what he and she said.

The dilemma for journalists is that the presumption of innocence we owe the accused, especially after an acquittal, collides ineluctably with the presumption of truthfulness that, in decency, we owe the accuser.

There’s no avoiding this collision. We can only try to strike a fair balance between conflicting ethical imperatives.

For starters, we can avoid such politically correct slanting as Newsweek’s indignant complaint last week (in a piece by Eloise Salholz, with five others) that "the Smith proceedings have confirmed what many women suspected all along: that when a man is accused of a sex offense, the woman goes on trial."

Well, of course she does. When it’s her word against his, they both do. That’s as it should be. Unless defense lawyers, judges, and journalists pursue aggressively any doubts as to her credibility, we effectively presume his guilt. (The Newsweek piece comes close to suggesting that we do just that: "Research suggests that the notion that women invent rape charges is statistically unfounded and psychologically implausible.")

Still, in my view, the working presumption of truthfulness to which rape complainants should be entitled tips the balance in favor of respecting their anonymity (if they want it) in the pretrial stages.

That’s because for many women, involuntary public identification as the complainant in a case like this can cause immense harm. Aside from whatever stigma may still attach (though none should) to women who cry rape, publicity about the details of such cases strips away a precious residue of their privacy. And this prospect may well deter rape victims from seeking justice.

This argues for using blobs and bleeps instead of faces and names when the only thing on the other side of the scales is the public’s substantial but less-than-compelling interest in knowing everything about the complainant.

It is not enough to say, as David Kaplan did In another piece in last week’s Newsweek (dissenting from the company policy of protecting ( anonymity), that "fairness dictates consistency" for as long as we do not "seriously consider the legitimate privacy interests of those other unfortunates we cover-the hostage, the plague carrier, the family of the plane-crash victim." We should respect everybody’s legitimate privacy interests, except when we have good reason to sacrifice those interests.

Nor is the fact that some competitor has already blown the complainant’s cover a good enough reason to assume she has no privacy left to protect. So The New York Times reasoned in April when, in the wake of an NBC News broadcast exposing the Palm Beach complainant’s name, it did the same. After a firestorm of criticism, the Times-stopped using her name.

But the argument for exposing the complainant’s name and face becomes far stronger at the point at which fairness to the defendant argues for doing so. This will typically be true in any acquaintance-rape case that goes to trial, because the complainant’s demeanor and credibility will inevitably be crucial issues.

Herein lies the critical difference between the Palm Beach case and New York’s Central Park jogger case. There was no doubt that the josser had been brutally assaulted and raped. The case against the defendants did not depend on the credibility of her (non-existent) recollections.

Under those circumstances, the defendants were not prejudiced in the slightest, either in court or in public opinion, by the anonymity that most news outlets accorded the jogger.

In the Palm Beach case, on the other hand. Smith was potentially harmed from day one by the complainant’s anonymity.

Even in the pretrial stages, it diminished, at least marginally, the prospect that people in Palm Beach or elsewhere might come forward with any evidence casting doubt on her credibility.

More important, when she testified at trial. TV censors blobbed out the kind of demeanor evidence our legal system treats as critical to juries’ and judges’ assessments of witnesses’ credibility.

Most fundamentally, a decision to protect a date-rape complainant’s anonymity contains within it an implicit statement that she is presumed truthful. After all. journalists would properly feel that they should expose, not shield, the identity of a complainant whose story appears fabricated.

If she is presumed truthful, it follows, as the night the day, that he is presumed guilty.

So unless we name the date-rape complainant in every case, or in none-or in every case when she appears in open court-we must make case-by-case, day-by-day determinations whether to treat her as a presumptive victim or a presumptive liar.

Journalists aren’t comfortable making such determinations. We would generally rather defer to the conclusions reached by the legal system; thus, the journalistic convention after a conviction has long been to drop the allegedly’s and treat the defendant as indisputably guilty, even when the evidence still leaves us in doubt.

Acquittals are far more ambiguous than convictions, grounded as they can be in the presence of a mere reasonable doubt. An acquittal in a date-rape case does not by itself justify treating the complainant as a perjurer.

But fairness to the acquitted defendant suggests that at this stage (if not before) we stop treating his accuser as a victim, at least when there is more than a reasonable doubt, and the evidence suggests she is probably lying.

Can we fairly conclude that the Palm Beach accuser is probably lying? I think she is, about some things, if not everything. Others think not.

The difficulty of drawing lines based on subjective assessments of the evidence drives me toward the view that we should protect a date-rape complainant’s anonymity up to the point when she appears in open court, and should use her name and face from then on. That’s an easy line to draw, in every case.

I might strike a different balance were I a publisher, properly worried about the costs of offending readers by treating rape complainants as forfeiting all claim to anonymity once they go to trial. But is it fair to exempt rape complainants alone from the publicity with which all defendants, and all other accusers, must live?