When a court-appointed special master last year rejected the claim of an Alabama couple that their daughter had suffered seizures after a vaccination, she explained her decision in part by referring to material from articles in Wikipedia, the collaborative online encyclopedia.

The reaction from the court above her, the United States Court of Federal Claims, was direct: the materials “culled from the Internet do not — at least on their face — meet” standards of reliability. The court reversed her decision.

Oddly, to cite the “pervasive, and for our purposes, disturbing series of disclaimers” concerning the site’s accuracy, the same Court of Federal Claims relied on an article called “Researching With Wikipedia” found — where else? — on Wikipedia. (The family has reached a settlement, their lawyer said.)

A simple search of published court decisions shows that Wikipedia is frequently cited by judges around the country, involving serious issues and the bizarre — such as a 2005 tax case before the Tennessee Court of Appeals concerning the definition of “beverage” that involved hundreds of thousands of dollars, and, just this week, a case in Federal District Court in Florida that involved the term “booty music” as played during a wet T-shirt contest.

More than 100 judicial rulings have relied on Wikipedia, beginning in 2004, including 13 from circuit courts of appeal, one step below the Supreme Court. (The Supreme Court thus far has never cited Wikipedia.)

“Wikipedia is a terrific resource,” said Judge Richard A. Posner of the United States Court of Appeals for the Seventh Circuit, in Chicago. “Partly because it so convenient, it often has been updated recently and is very accurate.” But, he added: “It wouldn’t be right to use it in a critical issue. If the safety of a product is at issue, you wouldn’t look it up in Wikipedia.”

Judge Posner recently cited a Wikipedia article on Andrew Golota, whom he called the “world’s most colorful boxer,” about a drug case involving the fighter’s former trainer, a tangent with no connection to the issues before his court. He did so despite his own experience with Wikipedia, which included an erroneous mention of Ann Coulter, a conservative lightning rod, as being a former clerk of his.

“I have never met Ann Coulter,” he said, but added that he was heartened that the friend who spotted the error could fix it then and there.

That friend was Cass R. Sunstein, currently a visiting professor at Harvard Law School. “I love Wikipedia, but I don’t think it is yet time to cite it in judicial decisions,” he said, adding that “it doesn’t have quality control” He said he feared that “if judges use Wikipedia you might introduce opportunistic editing” to create articles that could influence the outcome of cases.

He added, however, that he could not fault a use like Judge Posner’s, which “seems too innocuous for a basis of criticism.”

Many citations by judges, often in footnotes, are like Judge Posner’s, beside the main judicial point, appear intended to show how hip and contemporary the judge is, reflecting Professor Sunstein’s suspicion, “that law clerks are using Wikipedia a great deal.”

The Supreme Court of Iowa cites Wikipedia to explain that “jungle juice” is “the name given to a mix of liquor that is usually served for the sole purpose of becoming intoxicated.” In the Florida case, the court noted that booty music has “a slightly higher dance tempo and occasional sexually explicit lyrical content.”

As opposed to these tangential references, Wikipedia has also been used for more significant facts.

Such cases include a Brooklyn surrogate court’s definition of the Jewish marriage ceremony and the Iowa Court of Appeals’ declaration that French is the official language of the Republic of Guinea. In 2004, the Court of Appeals for the 11th Circuit, in Georgia, referred to a Wikipedia entry of the Department of Homeland Security’s threat levels in a ruling concerning magnetometer searches of antiwar protesters.

In a recent letter to The New York Law Journal, Kenneth H. Ryesky, a tax lawyer who teaches at Queens College and Yeshiva University, took exception to the practice, writing that “citation of an inherently unstable source such as Wikipedia can undermine the foundation not only of the judicial opinion in which Wikipedia is cited, but of the future briefs and judicial opinions which in turn use that judicial opinion as authority.”

Recognizing that concern, Lawrence Lessig, a professor at Stanford Law School who frequently writes about technology, said that he favored a system that captures in time online sources like Wikipedia, so that a reader sees the same material that the writer saw.

He said he used www.webcitation.org for the online citations in his amicus brief to the Supreme Court in Metro-Goldwyn-Mayer Studios v. Grokster Ltd., which “makes the particular reference a stable reference, and something someone can evaluate.”

Wikipedia is increasingly becoming the default reference for the curious. According to comScore Media Metrix, there were more than 38 million unique visitors to Wikipedia sites in December in the United States, making it the 13th most popular destination.

Stephen Gillers, a professor at New York University Law School, saw this as crucial: “The most critical fact is public acceptance, including the litigants,” he said. “A judge should not use Wikipedia when the public is not prepared to accept it as authority.”

For now, Professor Gillers said, Wikipedia is best used for “soft facts” that are not central to the reasoning of a decision. All of which leads to the question, if a fact isn’t central to a judge’s ruling, why include it?

“Because you want your opinion to be readable,” said Professor Gillers. “You want to apply context. Judges will try to set the stage. There are background facts. You don’t have to include them. They are not determinitive. But they help the reader appreciate the context.”

He added, “The higher the court the more you want to do it. Why do judges cite Shakespeare or Kafka?”

...a new computer program, which can determine who alters Wikipedia information, traced the culprit behind the addition of the words "jerk, jerk, jerk" to President Bush’s Wikipedia profile. The source? A New York Times computer. There was also no discussion of a similar incident involving the insertion of the word "wanker" to Bush’s entry from a BBC computer.

CONGRESS just passed, and President Bush hurriedly signed, a law that amends the legal framework for the electronic interception of various kinds of communication with foreign sources. Almost immediately, commentators concluded that the law was unnecessary, that it authorized a lawless and unprecedented expansion of presidential authority, and that Democrats in Congress cravenly accepted this White House initiative only for the basest political reasons. None of these widely broadcast conclusions are likely to be true.

All sides agree that some legislative fix is required because of changes in telecommunications technology. Where once it made sense to require warrants when one party to a foreign conversation was in America, this ceased to be the case when American routers became the transit points for foreign conversations that might or might not involve a person in the United States.

Once linear, analog, point-to-point communication has been replaced by the disaggregated packets of the Internet, two people talking to each other in Europe could find their conversations going through American switches. It also became difficult to determine the true origin of any communication that was routed through the United States. If a terrorism suspect in Pakistan is having conversations with someone on a computer with a New York Internet protocol address via a chat room run by an Internet service provider in London, where exactly is the intelligence being collected? If the answer is the United States simply because the servers are here, of what possible relevance could that be to the protection of the rights of Americans?

Amending the statute to focus on protecting American people rather than an American address would not have dealt with a larger and more profound problem. The change in the global communications infrastructure is both a driver and a consequence of a change in the nature of conflict. The end of the cold war was brought about in part because of technologies that empowered the individual and whetted people’s appetites for more control over their lives. These same developments also empower networks of terrorists, and the war they will soon be capable of waging has little in common with the industrial warfare of the 20th century. Accordingly, foreign intelligence tasks will also change.

It made sense to require that the person whose communications were intercepted be a spy when the whole point of the interception was to gather evidence to prosecute espionage. This makes much less sense when the purpose of the interception is to determine whether the person is in fact an agent at all. This sort of communications intercept tries to build from a known element in a terror network — a person, a telephone number, a photograph, a safe house, an electronic dead-drop — to some picture of the network itself. By crosshatching vast amounts of information, based on relatively few confirmed elements, it is possible to detect patterns that can expose the network through its benign operations and then focus on its more malignant schemes.

For this purpose, warrants are utterly beside the point. As Judge Richard Posner has put it, “once you grant the legitimacy of surveillance aimed at detection rather than at gathering evidence of guilt, requiring a warrant to conduct it would be like requiring a warrant to ask people questions or to install surveillance cameras on city streets.” Warrants, which originate in the criminal justice paradigm, provide a useful standard for surveillance designed to prove guilt, not to learn the identity of people who may be planning atrocities.

A statutory fix that simply waived the warrant requirement when both parties to a conversation were foreign would scarcely address this problem. Technology is changing the nature of the threat, not merely the mechanics of collection. The statutory change is unnecessary, I suppose, if you believe that there is in fact no real threat, that it’s all hype by the White House to expand its powers — presumably to some other end — and that all we have to fear is fear itself. Doubtless, some people do believe this. If the editorialists and columnists in the news media make this assumption, they should frankly say so (and hold their breath until the next attack).

Furthermore, there is an unstated assumption that warrantless surveillance is lawless surveillance. There is, however, judicial precedent for warrantless searches, even if you can’t tell this from the public debate. The president of the American Bar Association objected to the new statute by sarcastically observing, “The last time I checked, the Fourth Amendment is still in the Bill of Rights,” which he doubtless believed to be a withering salvo.

In fact, there are many instances in which warrantless surveillance has been held to be permissible under the Fourth Amendment. Searches in public schools require neither warrants nor a showing of probable cause. Government offices can be searched for evidence of work-related misconduct without warrants. So can searches conducted at the border, or searches undertaken as a condition of parole. Searches have been upheld in the absence of a warrant where there is no legitimate expectation of privacy. The Clinton administration conducted a warrantless search — lawfully — when it was trying to determine what the spy Aldrich Ames was up to. The day after Pearl Harbor, President Roosevelt authorized the interception of all communications traffic into and out of the United States.

Then there is the widespread charge that Democrats supinely accepted all this on political grounds. There probably were Democrats who adapted their long-held views on the Foreign Intelligence Surveillance Act to political necessity. But these are most obviously to be found on the other side of the vote. Senators Hillary Clinton, Christopher Dodd and Joseph Biden — all of whom are running for president — voted against this legislation, when their records are otherwise quite forceful where national security issues are concerned. With respect to those voting in favor of the statute, I find it hard to believe that Senator Daniel K. Inouye of Hawaii and Senator Jim Webb of Virginia are concerned about appearing insufficiently sensitive to security threats to the country.

Why would we be troubled in any case when a politician in a democracy votes the way he thinks the people want? Polls show that the American public is not as anti-security-minded as the American Civil Liberties Union. That’s why we need an A.C.L.U., I imagine.

One good reason not to want popular politics to guide such decisions arises when the public is not well-informed. Partly this can be laid at the door of the incumbent president, the Great Miscommunicator. But mainly it lies with those people who don’t bother giving reasons, don’t explain or give arguments, who prefer to traduce the people with whom they disagree by attacks on their characters, which are presumed to be insufficiently stalwart.

In Robert M. Gates, the defense secretary, Mike McConnell, the director of national intelligence, and Gen. Michael V. Hayden, the director of central intelligence, we have about as good a team as it is possible to imagine. Most people in Congress know that. Why not assume they are proposing a solution to a real problem? Developments in technology are forcing a long-overdue statutory change — and those developments will be with us long after the politics of the moment have passed.

Philip Bobbitt, a professor of law and the director of the Center for National Security at Columbia University, was a National Security Council senior director from 1998 to 1999.

Warrants, which originate in the criminal justice paradigm, provide a useful standard for surveillance designed to prove guilt, not to learn the identity of people who may be planning atrocities.