Since the 1970s, Israel has reversed her instructions to soldiers for talking to their captors. In line with Soviet military doctrine, IDF originally banned its soldiers from divulging any information to the enemy except their name, rank, and serial number, which are required for Red Cross identification. The policy worked: by far most Jewish captives in the Yom Kippur war refused to divulge any information to the Egyptians. True, in terms of intelligence-gathering it matters little that, say, 80% of the POWs refused to cooperate; the rest were more than enough for the enemy to obtain any information they wanted.

Two factors led to the policy reversal: the widespread torture of non-cooperative POWs and substantial information leakage by the weak. The torture, however, was not directly related to intelligence gathering: the Egyptians brutalized and murdered Israeli POWs simply for revenge. This, by the way, makes abominable the silence of the Israeli government when a few years ago Egypt accused us of killing their POWs; we did, sure, but on a scale not even approaching the Egyptian murder of our soldiers. (I’m speaking of relative figures; since Egyptian POWs greatly outnumbered captured Israelis, we probably killed more of their POWs than vice versa.) Back to the point, talking to the enemy can save a few high-ticket captives from torture. But when every captive talks, their information is of little value, and they will be tortured anyway.

This attitude reflects a typical Israeli way of institutionalizing failures. Yes, there will always be weak people, willing to divulge our military secrets to the enemy. That is not a proper reason for everyone to descend to their level. If their information is valuable, they must be encouraged to commit suicide rather than talking to their captors―or soldiers can be instructed to kill suspected collaborators among their fellow POWs.

Which brings us to the question, what information is valuable? During the Korean War, the Chinese ran extremely successful ‘conversion’ programs in POW camps. North Koreans generally did not succeed in breaking American POWs with beatings, but the Chinese achieved whopping 80–90% cooperation rates by subtle psychological means. They started by eliciting from POWs just marginal criticism of the American system (“there is some poverty in the United States”) and equally marginal praise of communism (“communism seeks equality of people”), and then kept prodding the captives to logically expand on those statements. In result, the returnees from Chinese camps were heavily indoctrinated in communism and anti-Americanism. Interrogators employ a similar approach. They ask the POW trivial questions, establish a rapport, and slowly move to more sensitive questions, which he is psychologically unable to refuse, having been ensnared in the trap of cooperation. Unless the captive is especially smart and well-versed in psychological manipulation, he won’t be able to dodge his interrogators with trivial answers for long.

In the age of mass media that scrutinizes every aspect of national economic and military affairs, most soldiers really possess no information important to their interrogators or critical to national security. That would seem to leave them free to talk to enemy interrogators. Such an approach is mistaken. Though our enemies are nowhere near the acute psychologists the Chinese were, and cannot stage mass conversions such as were seen in the Korean War camps, cooperation with the enemy leaves a psychological scar no less than torture leaves physical scars. Much more important is the problem that free-talking, know-nothing POWs create for the few soldiers and officers who do possess security information. Their uncooperative behavior stands out, putting them at risk of being tortured severely and forced to reveal their secrets. With modern means of torture, including chemical agents, very few people can keep themselves from divulging their secrets, and it might not always be possible for them to commit suicide.

If uncooperative behavior were universal among POWs, the enemy would be forced to sift through hundreds, perhaps thousands of captives, in order to find out a few intelligence assets. By the time they succeeded, their tactical information would become outdated. There is a good chance that the enemy would find it impossible to comb through scores of uncooperative POWs, and the assets would preserve even strategic information. Tellingly, Germans during WWII tried to identify high-value POWs immediately upon capture, and abstained from large-scale efforts at sifting through large groups of POWs.

It is absurd to suggest, as the army’s doctrine does, that soldiers cannot tell the enemy anything important because the battlefield changes too fast. No war is our last war, and secrets about battlefields are not the only secrets we have. Tens of thousands of IDF soldiers went through special training camps and bases which the enemy would die to know about. The massively lopsided Tannenbaum exchange shows how highly we value some of those secrets.

To suggest that the POW’s life is all-important is the same nonsense which leads to disproportionate prisoner exchanges. Every life, and certainly a soldier’s life, has a price tag, however regrettable that is. It is far preferable for society to equip soldiers with poison and instruct those who know major secrets to commit suicide if there is no other chance to preserve the information.

Generals, as the truism goes, fight the last war. The current “tell-them-everything” rule might have worked with the primitive Syrian and Egyptian interrogators of the 1973 war. They knew next to nothing about Israel and could easily have been inundated by a flow of irrelevant information. In the current situation, the army’s willingness to allow its soldiers to talk to interrogators almost freely may well cause them to divulge our every secret.