Over at crookedtimber, Kieran Healy notes that the NSA's program of monitoring every phone call is actually quite familiar to a sociologist. That is, what the NSA is doing is very similar to the sociologists pipe-dream of having an infinte dataset, unconstrained by any practical considerations. The sociologists just lack the freedom from ethical prerogatives and monetary constraints that the NSA enjoys.
To me, this suggests an interesting argument: we don't let scientists collect ideal datasets in the way the NSA has done out of ethical considerations-we're happy to set up Institutional Review Boards which constrain the ways in which you can treat people you're researching. So why should the NSA be any different? There's a certain tug once you think about the topic. A little voice says "it's for national security*, we can do more to protect ourselves from attack than we can for the sake of science." Upon reflection, that's just wrong, though. National security is a fleeting and ephemeral thing which only benefits one nation, often at the expense of another. Science, on the other hand, is a progressive body of knowledge which is the shared property of humanity. Whatever case you can make for a measure being justified in terms of national security, you can make for it being justified for the sake of science.
I admit that I have some trouble seeing this in the case of sociology, probably because I don't know the field. But when I think about epidemiology, for instance, it just feels obvious that invading people's privacy to get good data would be far more useful than it would be for the NSA. Yet we still gladly live with these restrictions on epidemiological research. I think it's clear that in this case, it's the calmer reasoning employed in the normal case of scientific research that is worth heeding. Given the status quo, any inclination that people have to turn tail and accept the NSAs actions, or amend the law to make it legal in the future, is hysteria speaking.