Let's assume that Israel has a top spy in Tehran, the son-in-law of the Supreme Leader, Ali Khamenei, or his predecessor, Ayatollah Khomeini. And let's assume the quality material the spy is relaying to Mossad is being pumped into the veins of the most senior ranks of decision makers, waiting to hear what he has to say on the most important question of them all: has Iran crossed the line between developing a nuclear infrastructure and building a nuclear weapon?

An affirmative answer will put an end to the disagreement between Israel and its friends in the west, on deciding whether it is still possible to wait before embarking on a military operation against Iran's nuclear facilities. Such a spy could turn out to be an asset who actually works against his masters' interests.

The Iranian spy is just a fable, an imaginary parallel to the Egyptian Ashraf Marwan, Gamal Abdel Nasser's son-in-law and the man who gave Israel Anwar Sadat's secrets. (Marwan died in mysterious circumstances in June 2007, falling from the fifth floor of his London apartment. )

What are your thoughts on this issue? Follow Haaretz.com on Facebook and share your views.

Last week saw the start of yet another round in the debate between Mossad and Military Intelligence on Marwan: was he a loyal agent who saved Israel with his last-minute warning about 1973's Yom Kippur War, or did he mislead and make Israel drop its guard? During earlier rounds, Zvi Zamir - the head of Mossad from 1968-74 - clashed with Eli Zeira, the head of Military Intelligence during the year which preceded the Yom Kippur War. This time, it was Zeira's successor at Military Intelligence, Shlomo Gazit, who criticized the way Marwan was handled by Zamir.

In an article written in a periodical of the intelligence community, Gazit rejected the status granted to Marwan as a super source "who volunteered and conscripted himself." He was considered "more valuable than gold" and also cost a great deal of "gold," criticizes Gazit.

"The head of Mossad also contributed to this unusual situation, and he took an exceptional step, which runs contrary to the principles of intelligence work," accuses Gazit. Zamir "decided to disseminate the original reports received from Marwan to a limited number among the senior political-military leadership" - then Prime Minister Golda Meir, Defense Minister Moshe Dayan, minister Israel Galili, Chief of Staff David Elazar, a handful of generals - instead of first passing them through the research filter of Military Intelligence and then cross-checking them with other materials for a more considered assessment.

"Thus was created a new and strange reality, where the decision makers had to choose between the drily-written analysis documents - prepared in the careful intelligence jargon - and the exciting reports of Marwan," becoming hooked to the point of dependency.

From mid-September 1973, "troubling reports accumulated, first on the Syrian front and then also on the Egyptian front, which under normal circumstances would have seen Israel responding to them with seriousness and concern, declaring a state of readiness at the Israel Defense Forces and perhaps even calling in reserves." However, "so long as Marwan's reports kept coming in and the reports were not evaluated, dealing with the troubling information was delayed."

Gazit considers the most serious possibility to be that Marwan was not actually a double agent, "because that would mean we fell into a hole we ourselves had dug, with the exalted status given to the agent by his handlers, those evaluating his reports and especially those at the top, who eagerly read his exciting reports."

The source whose reports go directly to the top is rare indeed (leaders also directly receive information from their counterparts, including Golda from King Hussein ). The general problem is that senior decision makers with a defense background may believe that their experience and skills free them from the methodical process of intelligence analysis - especially when their mind is already made up. Must Ehud Barak, who was head of Military Intelligence when current chief Aviv Kochavi was a 19-year-old corporal, rely on the ant work of the professionals?

Of course, it is not the method that is troubling those who wish to bomb Iran, but the result. Were Military Intelligence and Mossad to lie to themselves and provide data fully supporting an operation, Barak and Prime Minister Benjamin Netanyahu would use this to justify their fervor. Before a country goes to war, it must check its sources well - and even more itself. 1973 is a clear example - not the first one and not the last one - of fatal contempt. It may happen again.

Read this article in Hebrew