I haven’t read the whole thing yet, but Charles Cooke has a long essay in the latest issue of National Review:

The law of defamation is useful for awarding civil damages against those who peddle outright lies — that is, against those who do real damage to a person’s reputation by abusing plain facts that can be easily verified and adjudicated in court. In such cases as it is claimed that Jones beats his wife or Smith is a drug addict, the relevant facts fall easily within the competence of a civil tribunal, and litigation does not threaten to impose a chill on the public discourse. But when a plaintiff files a libel suit involving a matter of political or scientific controversy, the calculus is quite different indeed. When the merits of a libel claim implicate contested questions of science and statistical methodology, judges and juries are so ill suited to pronounce a verdict that allowing the public authority to have the final say is inconsistent with the very concept of free inquiry. The whole point of the scientific enterprise is to resolve controversies through open debate, not through the final decree of government officials.

Even where no verdict of guilt is ultimately pronounced, allowing litigation over criticisms of the validity of scientific research has a deleterious effect on the public discourse. It prompts critics to trim their sails in order to avoid the cost and headache of a lawsuit, thus establishing a climate of fear and quiet rather than of boisterous agitation and open discussion. Hanging the prospect of punishment above the heads of participants in scientific disputes serves not to yield greater accuracy but to invite censorship, the toning down of rhetoric, and the avoidance of hyperbole — of anything, indeed, that could invite a libel complaint. Which is to say that it shuts up the dissenters.

Which is, of course, the entire point.