A year ago, the notorious Cosa Nostra killer Giovanni Brusca was arrested in Palermo and everybody cheered. This was the man, after all, who pressed the button that set off the bomb that blew up the heroic anti-Mafia magistrate Giovanni Falcone in May 1992. This was the man, too, who had punished a Mafia informer by dissolving the body of his 11-year-old son in an acid bath.
Then last summer Brusca decided he too would start collaborating with the justice authorities, and the cheering suddenly went quiet. Did the state really want to start offering protection, not to mention a salary and the promise of judicial leniency, to a monster of a man nicknamed U Verru, The Pig? Could his confessions really be trusted, let alone merit reward?
Thus began the most prominent debate surrounding the whole issue of informers, orpentiti, "repentant" mafiosi, as they are informally known. In the first three months, much of what Brusca said turned out to be either unverifiable or false, and a growing chorus of politicians called for a tightening of the whole collaboration system.
Then last month Brusca suggested that some of his fellow informers had committed murders while under the state's witness protection scheme, and the same crowd of politicians wondered whether a pentito could ever be trusted again.
The public concerns are understandable, but unfortunately they are based on a fundamental ignorance of the way that Mafia collaboration works. Moreover, the debate has been skewed by a section of the political class with a direct interest in discrediting all evidence based on the testimony of wrong-doers, not just members of the Mafia.
The risk now is that the whole system of collaboration, an invaluable resource for prosecutors which is responsible for more than half of all investigations into the Mafia now underway, will be undermined by a new law about to go before parliament.
A published draft of the law rightly recognises that too many informers and their families - more than 8,000 people, all told - are being protected regardless of the quality of the testimony they have to offer, and it attempts to sort them into different categories. But it also obliges would-be collaborators to tell everything they know, and within a very limited time scale, before they can find out if they qualify for a witness protection scheme.
Since it can take months for informers to start giving really valuable information, and several years before they attack the most sensitive subjects such as the Mafia's relationship with politics, the quality of the evidence is almost sure to go down. Not only that, but most mafiosi might feel that revealing all, with no guarantee of a quid pro quo, is reason enough to keep their mouths firmly shut.
"We don't think witnesses will tell everything they know in the allotted time period, only as much as they think they need to," said the Palermo prosecutor, Antonio Ingroia. "And after the six months are up they will be required to attest that they have nothing more to say, thus barring them from ever testifying again, even if they want to."
The first Mafia informer, Tommaso Buscetta, still has not stopped talking after 13 years.
Since the confidence of Mafia informers depends crucially on the commitment of the state to combat organised crime and offer true protection to those who disassociate themselves from it, evidence inevitably comes in waves over a long period of time.
The public concerns about occasional disinformation are misplaced, first because everything is rigorously checked against other witnesses and material evidence, and secondly because lies are part of the long process of evolution that mafiosi go through when they decide to turn state's evidence. Brusca, in fact, has not yet proved reliable enough to be considered a full informer - he is described as a dichiarante, or talking witness.
"There is a problem with the word pentito because it implies a moral transformation, whereas what we are doing is bargaining with these people for information," said Pier Luigi Vigna, Italy's top anti-Mafia prosecutor and a man with long experience of judicial interrogation of witnesses.
"These people might even go back and commit crimes, but that doesn't change the fact that what they previously said may be true."
Why is the most successful prosecution tool to have emerged in Italy in recent years being messed with? Who stands to gain?
One obvious beneficiary is Giulio Andreotti, the veteran Christian Democrat politician whose trial on charges of Mafia collusion is based almost exclusively on pentito evidence. As a man still commanding widespread public respect, he has been quite successful in convincing his old electorate of the existence of Mafia plots to discredit him.
But the problem goes deeper than just one man. The whole political class has an interest in curbing the powers of the magistrature to ensure that the vast judicial investigations into official corruption of the early 1990s run no risk of being repeated.
Not just the Mafia, but also the old political system, were both dismantled through the evidence of pentiti who negotiated a better deal for themselves by spilling the beans on someone more important.
This prosecution method is questionable in the case of political corruption because nobody in the end goes to jail at all - only the big fish get hefty sentences and they either defer them in the appeal courts or avoid them by going abroad. In the case of the Mafia, though, the informer system works brilliantly because what matters is not so much the guilt of an individual as the structure of the system as a whole.
"In a closed world like a terrorist group, if you capture the top five or ten leaders the whole things falls apart. But the Mafia is not like that," Mr Vigna said. "The Mafia is not closed to the world, indeed in certain places it is the world."
For the past decade, the door on that world has been opened for the first time. However, it may be about to be slammed shut again.
Join our new commenting forum
Join thought-provoking conversations, follow other Independent readers and see their replies