After yesterday's report about four-year-old girls taking off their sandals in the airport security line, one reader writes:

Several years ago, a TSA agent at the Islip/Macarthur NY airport made us remove a pacifier from the mouth of our toddler daughter before going through the metal detector. After she started crying, he said, rather sarcastically, "If that's the worst thing that happens to her all day, it's a pretty good day."

I haven't punched anyone in the face since I was fourteen years old, but I kind of regret not slugging the jerk. (Though I'd probably be in federal prison right now if I had...)

But another demurs:

It certainly sounds stupid to make 4-year old girls remove their sandals. But here's what I think is going on: The TSA doesn't want their employees working on the security lines to exercise independent judgment about what constitutes a potential threat and what doesn't. These men and women have a limited amount of training and it certainly isn't enough to be able to spot a potential terrorist.

I think this approach is correct. Yes, it does lead to some absurdities, like making 4-year olds remove their sandals. But this sort of rigid application of the rules doesn't terribly inconvenience anybody and it doesn't appreciably add to the length of time it takes to get through security. Better that everyone be subject to the same rules than that some 26-year old TSA employee, with a 2-week training course under his or her belt, be charged with the discretion to decide who looks like a possible threat and who doesn't.

Could they revise the protocol so that travelers don't need to remove sandals? Maybe that would make sense, but bear in mind that the more rules there are and exceptions to rules, the less reliable the system will be. If there are too many rules or exceptions to rules, more mistakes will be made by the TSA personnel. That's not in anyone's best interest.

It's a fair point that rules are rules, and that as soon as you allow or require each TSA agent to make judgment calls, you're asking for new complications. Lines would probably be longer and, if anything, more confrontational, since each individual agent's judgment, rather than "the rules," would be the source of intrusions we didn't like.

Still, the fundamental problem with "security theater" is that it elevates the appearance of greater security, plus the machinery and process of seeming safety-concerned, over the reality. In my view, the no-exceptions, no-common-sense-allowed application of rules undermines the long-term faith that the security authorities know what they are doing. We're really making the pilots of the plane (along with four-year-olds) give up their bottled water at the checkpoint? What do we think they're going to do with it? If a pilot is a secret agent bent on suicide terrorism, confiscating his water isn't going to make any difference. And in all other circumstances, since we are after all trusting him to fly the damned plane, why won't we trust him with his water?

In a larger sense this is why it's a shame that the TSAs's "intelligence-based" program to identify probable threats, as opposed to its "apply the same rules to everybody" approach at the airport checkpoints, has so far proven disappointing.  (Background here.) And someone with long involvement in this field wrote to me recently, more and better intelligence-based security is the only sensible way ahead for the TSA:

You can't have it both ways -- thumping on TSA for security theater in its traditional checkpoint screening and then, when TSA introduces an intelligence-driven, non-intrusive layer of security, whack 'em again.... TSA needs encouragement to do more things that are intelligence-driven. They are very risk-averse from the public affairs point of view and if, even when they venture out into smart security, people like you beat on them, it will cause them to stick with the old stuff.

Again a fair point. So, for the record, I henceforth resolve to be supportive and constructive -- including with constructive criticism! -- in urging TSA to develop the intelligence-based systems that will mean less hassle for four-year-olds, teething infants, and uniformed flight crews, and more on more-probable malefactors. (And, yes, I realize that the next stage is a debate about "profiling" -- that is for another time. "Profiling" is merely "intelligence" done clumsily.)

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.