Trial By Osmosis: Amanda Knox, Raffaele Sollecito and the Nightmare of Italian Justice

By Andrew GumbelFebruary 11, 2014

Trial By Osmosis: Amanda Knox, Raffaele Sollecito and the Nightmare of Italian Justice

ANYONE FOLLOWING THE BYZANTINE TRIALS of Amanda Knox, the American exchange student accused with her onetime boyfriend Raffaele Sollecito of murdering her roommate in Perugia, will have noticed that criminal justice in Italy doesn’t work the way it does in other countries. First they were guilty, then they weren’t, now they are again. In the United States, this is commonly referred to as double jeopardy and is barred under the Constitution. In Italy, it’s pretty much business as usual.

When the pair was first arrested, more than six years ago, they were left to rot in jail and for months — in Sollecito’s case in solitary confinement — before charges were brought. They didn’t qualify for bail because bail does not exist in Italy. The prosecution regularly leaked information to the media but did not formally share its investigative findings with the defendants or their lawyers until the summer of 2008, by which time the public was broadly convinced they were no ordinary college students, but rather, depraved sex addicts who had forced the victim, 21-year-old Meredith Kercher, into a satanic orgy before brutally stabbing her to death.

To this day, it remains doubtful whether evidence ever existed to substantiate such a scenario.

The case went to trial with the prosecution refusing, despite repeated requests, to hand over the raw data on which it based its forensic analysis. The data was crucial because the prosecution claimed it had found traces of Kercher’s DNA on the tip of a kitchen knife believed to be the murder weapon (Knox’s DNA was on the handle), and traces of Sollecito’s DNA on a torn bra strap recovered from the crime scene. Neither claim would survive independent scrutiny.

That scrutiny, though, did not come until after Knox and Sollecito had already been convicted. The first trial judge, Giancarlo Massei, decided he didn’t need to ask the prosecution to hand over its full data, as would be a matter of course under US rules of evidence. He figured he could sort out the competing DNA claims without it. Massei also saw no problem with admitting evidence from the rough all-night interrogations Knox and Sollecito had endured in the run-up to their arrests, even though the country’s highest court, the Corte di Cassazione, had deemed it inadmissible. The reason Massei could do this was because the criminal case ran concurrently with a brace of civil suits brought by parties seeking financial compensation — something else that commonly occurs in Italy —- and the high court ruling applied only to the criminal case.

The latest developments in the case are, if anything, more perplexing still. The principal forensic evidence against Knox and Sollecito was thrown out on appeal, along with much of the eyewitness testimony, after it was shown to be false, unreliable and, quite possibly, spun out of whole cloth. That’s why the two defendants were exonerated, and why Knox was able to fly home to the United States at the end of 2011. But it was not the end of the story. Under Italian law, no verdict is considered “definitive” until it has been reviewed and approved by the high court. And the high court elected last March, as it often does, to send the case back for retrial at the appellate level. By this point the evidence against Knox and Sollecito was in tatters, and nothing new had emerged to incriminate them further. Yet the new appeals court in Florence decided that they were guilty anyway and, under the same bright media spotlight that has shone since day one, recommended sentences of 28-1/2 years for Knox and 25 years for Sollecito.

Much ink has been spilled on the injustice of the case — how Knox, in particular, became an international media sensation the Italian authorities could not resist, even after the man  whose DNA was indisputably all over the crime scene, Rudy Guede, was arrested, tried and convicted ahead of her. (The lingering contention is that they all committed the murder together.)

Critics, at least on this side of the Atlantic, have split into two broad camps. First, there are those who have treated the case as an anomaly, a miscarriage of justice fueled by tabloid hysteria whose equivalent one could, without too much difficulty, imagine unfolding over here. Sloppy police work, an over-zealous prosecution team, misreading of the crime scene, trial by media, a refusal to admit mistakes: these are things we can comprehend as universal human failings and have seen in one form or another at home. Then there are those who have leapt on the case as an opportunity to lord it over the Italians and express relief that nothing similar could possibly happen here. (The West Memphis Three might beg to disagree.)

What few people have focused on is the light that the case casts on the real idiosyncracies, eccentricities and shortcomings of the Italian justice system — in other words, how it operates day in and day out. The country has a long history of high-profile cases, many with wider-reaching ramifications than the Kercher murder, that have fallen prey to similar confusions, incompetence, ill will, abrupt about-turns and lack of resolution. More profitable than a flag-waving, my-country’s-justice-system-is-better-than-yours attitude may be to inquire why this is so.


A century ago, the British expatriate writer Norman Douglas toured the Italian south and was appalled by what he saw as the “fundamental perversity” and “elaborate brutality” of the justice system. In his gloriously dyspeptic travelogue Old Calabria (1915), he denounced harsh punishments handed down on the flimsiest of pretexts, “paleolithic” rules of evidence and courtroom procedure worthy of a Gilbert and Sullivan operetta. The courts were not the province of justice, he wrote, but of “elegant partisan non-truth”:

Rhetoric, and rhetoric alone, sways the decisions of the courts […] The rest is mere facts; and your ‘penalista’ has a constitutional horror of a bald fact, because there it is, and there is nothing to be done with it. It is too crude a thing for cultured men to handle. If a local barrister were forced to state in court a plain fact, without varnish, he would die of cerebral congestion; the judge of boredom.

What Douglas was witnessing, above all, was the legacy of a system developed by Italy’s old colonial masters, most notably the Bourbons, the Bonapartist French and the Catholic Church. That legacy, gussied up with pomp and classical learning and great outpourings of public emotion, resulted in a system where the presumption of innocence was fragile at best, defendants’ rights were next to non-existent and witnesses were regarded as playthings to be bent to the will of the investigating magistrate who almost always doubled as the prosecutor.

Italy has been struggling with this legacy ever since. The Fascists blew off some dust with a penal reform known as the Codice Rocco, but they also made the system more explicitly authoritarian and more political. With the founding of the Italian Republic after World War Two, the judiciary insisted on — and obtained — a large degree of independence but did almost nothing to alter the Fascist code, which remains substantially in place to this day and includes, for example, laws making it a crime to disrespect state institutions or insult public officials. The code provides for criminal suspects to be held for up to a year without charge while prosecutors assemble their case. It makes no professional distinction between prosecutors and judges (they are all fellow magistrates) and it allows trial judges to decide cases substantially on their own. Nominally, the judges are assisted by a panel of giudici popolari, “popular judges” who are in fact self-selecting members of the public. They are the closest thing Italy has to a jury — except they do not get to deliberate by themselves, are not vetted for their advance knowledge or opinions and are free to seek out information beyond the courtroom. In practice, they act as a rubber stamp for whatever the judges have already decided.

The biggest check on the system is its unusual three-tiered structure. Almost all criminal cases get heard in their entirety twice, first in the Court of Assizes and then in an appeals court. An initial guilty verdict can be overturned, but so too can a “not guilty” verdict. In the end it is up to the high court to sign off on the appeals court’s decision, or modify it, or to send the case back for retrial. This multi-layered approach can help break the coziness that may exist between local prosecutors and lower-court trial judges. It is extremely common, in fact, for criminal defendants to receive an initial guilty verdict on their way to a full exoneration. But the system is also wildly inefficient. Cases drag on for years, landing the innocent and the guilty alike with astronomical legal fees for which they are solely responsible and raising serious questions about fundamental fairness. If, on the way to a final conviction, a defendant has been exonerated, how does that exoneration not in and of itself constitute reasonable doubt invalidating the charges?

The answer to that question has to do with the Italian legal establishment’s attitude to the very idea of reasonable doubt. In short, they don’t like it, don’t trust it and, despite the explicit introduction of a “reasonable doubt” standard in a legal reform introduced in 2006, don’t generally base the way they prosecute cases on it. Giuliano Mignini, the first public prosecutor in the Meredith Kercher case, sounded almost alarmed by the concept when he made his closing statement in the first trial:

Yes it’s true you need to find the defendants guilty beyond a reasonable doubt, as the high court has said, but the high court was merely codifying a principle that already existed in our jurisprudence […] It doesn’t mean you need to find the absolute truth, which is the province of God alone […] You need only be certain enough for the purposes of a trial. What does that mean? It means two things, essentially: that the reconstruction of the facts is based on logic, and that its elements are not in contradiction with each other.

Tellingly, Mignini did not direct the court to look at the evidence per se, but to look at the logic of his reconstruction of the murder. And the high court endorsed his view when it sent the case back to trial last March. It was a mistake, the high court said, to focus on the shortcomings of the evidence piece by piece; the new trial judge needed to absorb the facts of the case “by osmosis” to appreciate the story in its totality.

The high court ruling was bizarre, even by Italian standards, because it all but directed the new appeals court to return a guilty verdict. (“This evaluation by osmosis will be decisive […] in demonstrating the presence of the two accused at the scene of the crime.”) But its philosophy of what constitutes evidence — also bizarre to Anglo-Saxon ears — was consistent with the way many Italian prosecutors and judges think and act. What they are interested in are less the facts themselves than the narrative they can build from their interpretation of those facts. They see themselves akin to poets, aspiring to a greater truth beyond the accumulation of evidence and dusty piles of official documentation. Their principal tools are rhetorical persuasion, instinct and artistic creativity. When they talk about logic, as Mignini and many of the other lawyers in the Kercher case have, they mean constructing a story that fits the predetermined view that the defendants are guilty.

Sometimes that story changes, even in the middle of a trial, and in the Kercher case it changed a lot. Mignini started out asserting  that the murder was the result of a sex game gone wrong; he then dropped that theory at the end of the first trial and asserted instead that Knox resented Kercher for being straight-laced, for having too many English friends, and for complaining when Knox forgot to flush a toilet. Judge Massei discarded both theories in his own explanation of the case and ascribed the crime to a spontaneous act of “pure evil.” Last March, the high court went back to the unsupported sex-game-gone-wrong theory. And the new prosecutor, Alessandro Crini, rejected that and went back to unflushed toilet theory.

What all of them were seeking, to varying degrees of narrative plausibility, was to fill in the gaps of what they did not know. Instead of acknowledging there was no physical evidence placing Knox and Sollecito at the scene and stopping the case right there, they each constructed a narrative — different in each case, some versions more unintentionally humorous than others — imagining what could have happened if the pair had been there. Put another way, they weren’t weighing evidence; they were writing fiction.

This is exactly what Norman Douglas railed against a century ago. The problem has not gone away and perhaps it never will. In the Anglo-Saxon mindset, informed by that old Protestant work ethic, truth is something we generally believe we can attain as long as we strive hard enough. Our court system is constructed accordingly. In the Italian, Catholic-inflected mindset, truth is a divine mystery — the province of God, as Mignini put it — and to presume to know it is to commit the most serious of the cardinal sins, the sin of pride. Reasonable doubt is hardwired into the human condition, and seeking to overcome that doubt is an exercise in futility. All we mortals have, in the end, are competing stories; we can only hope they make sense on their own terms.


Nothing about the Italian mindset prevents effective justice. On the contrary, Italy has often produced outstanding jurists whose subtle investigative minds, alert to the foibles of the human condition and forever conscious of their own doubts and limitations, can cut through the seemingly impenetrable complications of, say, a Mafia racketeering case or vast networks of political patronage and corruption. There is no overstating the extraordinary achievements of, say, Giovanni Falcone and Paolo Borsellino, who decapitated the top leadership of the Sicilian Mafia in the late 1980s and early 1990s and ended up paying with their lives.

The problem arises when jurists — good, bad or indifferent — run into a system that combines latent authoritarianism with maddening inefficiency. And the problems are only compounded when the case takes on any kind of public prominence. Throughout the Cold War, Italy was beset by a string of “mysteries”: kidnappings, bombings and assassinations that the government reflexively blamed on the militant left, and the militant left believed had been orchestrated by neofascists and government intelligence agents backed by the United States to discredit the Community Party. Almost invariably, trials arising from these misteri were reduced to incoherence, endless retrials, conflicting verdicts and controversy that refused to end.

Prosecutors and their overactive imaginations were often to blame. Frederic Spotts, a former US envoy to Italy, observed how such trials tended to “sink into mystification, conspiracy theories and wild speculation.” Italians call this dietrologia, literally the habit of looking “behind” surface realities for signs of lurking conspiracies and deception. When Ali Agca, the Turkish national who shot and almost killed Pope John Paul II in St Peter’s Square, was tried in the early 1980s, the judiciary all but willed him to be part of a broader plot. They fell hook, line and sinker for a CIA disinformation campaign — later exposed and discredited — that Agca was part of a “Bulgarian connection” ultimately leading to the Kremlin. “In some secret place,” prosecutor Antonio Albano wrote in one court brief, “where every secret is wrapped in its turn by another secret, a politician of great power […] made a decision, in accordance with the higher interests of the Soviet bloc, that it was necessary to kill Karol Wojtyla.” He had no concrete evidence of this, but he knew he had a receptive audience.

Italy has botched big mafia trials — most notoriously because of the repeated 11th-hour intervention of a Sicilian high court justice who earned the nickname ammazzasentenze, or sentence-killer. It botched the war crimes trial of Erich Priebke, a former SS captain responsible for one of the worst civilian massacres of World War Two, who was brought back from hiding in Argentina in the 1990s and found guilty, only to be excused prison time because of the “mitigating circumstance” that he had been obeying orders. The high court eventually reversed this finding, agreeing under considerable public pressure that it might be a good idea for the Italian justice system, 50 years after the fact, to absorb the lessons of Nuremberg. By the time Priebke was reconvicted he was too old to spend another day behind bars.

Often, the system fails to ensure that well-connected criminals pay for their misdeeds. In the early 1990s, the judiciary launched a huge investigation into a spider’s web of bribes and kickbacks underpinning the entire political system — an operation known as Mani Pulite, or Clean Hands. Magistrates made more than 2,500 arrests, destroyed the country’s two biggest political parties and essentially forced the country to reinvent the way it was governed. Fewer than 10 people, however, received prison sentences.

At other times, defendants in high-profile cases get nailed seemingly in spite of the evidence. In 1990, a left-wing activist and writer named Adriano Sofri was belatedly convicted of ordering one of the most controversial murders of the postwar period: the shooting of a Milanese police captain who in 1969 had allowed an anarchist he was interrogating to fall to his death from a fourth-floor police station window. (That event inspired Dario Fo’s play Accidental Death of an Anarchist.) Sofri had been on the suspect list for years because his organization had written a notorious article all but inciting the police captain’s murder, but the case against him rested almost entirely on the testimony of a low-life bank robber turned police informant of little credibility. The informant claimed to have driven the getaway car but got crucial details wrong, including the color of the car and the route it took away from the crime scene. The informant’s uncorroborated account of meeting Sofri and receiving instructions from him was similarly riddled with inconsistencies. Crucial evidence, including the car and the fatal bullet, was unaccountably destroyed or discarded before trial. At the very least, Sofri’s prosecution and 22-year prison sentence did nothing to heal the social and political divisions that the case had created. Sofri’s friends and supporters felt it was an act of revenge, pure and simple.

Could something similar have happened in the Kercher case? Could prosecutors and judges have taken advantage of the fluid approach that Italian jurisprudence takes to trial evidence and made a conscious decision to frame Amanda Knox and Raffaele Sollecito? On the surface, it’s hard to see how a young American exchange student — 20 years old at the time of the murder, 26 now — and the Italian boyfriend she had known only for a week could have inspired such institutional bloody-mindedness. Yet one person who appears to believe such a thing is at least possible is Claudio Pratillo Hellmann, the judge (now retired) who presided over the first appeal and found not only that Knox and Sollecito were not guilty under the law but that they affirmatively “did not commit the deed.” “I could tell in advance that the high court would quash my decision and send the case back,” he said in a newspaper interview last year. “The prosecutors’ lobby is very strong within the judicial system.”

Hellmann did not elaborate. But we do know, from trial records, that Mignini and his fellow prosecutors were deeply aggrieved by the criticism their work received in the American media and aggrieved, too, by the intervention of Maria Cantwell, a Senator from Knox’s home state of Washington, who issued a statement criticizing the initial guilty verdict and the Italian criminal justice system as a whole. We also know that the case’s saturation media coverage was due to Knox being a good-looking young American and the murder victim a good-looking young British woman. If they’d all been Italians the case would barely have made the papers outside Perugia.

It doesn’t take a conspiracy theorist to look at the pattern of events and conclude that the judiciary’s reaction to the criticism — to the flood of books and newspaper articles and TV specials and movie projects — was to double down and say, in effect, we’re not done with you yet.


One of the dangers of the Italian system’s chronic indecisiveness is that it diminishes the responsibility of any one judge issuing a guilty verdict. The head judge in the Florence appeal, Alessandro Nencini, gave a newspaper interview the day after his decision acknowledging how difficult it had been to hand out long sentences and saying he realized that debate over the case would continue. He was immediately reprimanded for speaking out before issuing his formal explanation of the verdict, due within 90 days. Lawyers for the defendants wondered if his words didn’t indicate a distancing, however slight, from his own decision. It’s not impossible: judges have been known to issue a guilty verdict to please their colleagues on the prosecutor’s bench and then fill their sentencing reports with enough inconsistencies and leaps in logic to provide the next court with ammunition to overturn their decision.

One of the best insights into the mindset of the Italian judiciary was published 80 years ago by Piero Calamandrei, a savvy jurist and terrific raconteur who explained in his book L’elogio dei giudici (In Praise of Judges) how judges can issue guilty verdicts knowing  they may be mistaken and still live with themselves:

It may be that half of the sentences handed down are unjust […] and therefore half of those in prison are innocent; but by the same reasoning half of those acquitted and set free are in fact guilty and should be in prison. Instead of worrying about individual cases, it's important to look at the bigger picture and understand that every error is compensated by another in the opposite direction. So the scales of justice are in balance and we judges can sleep easy at night.

That “bigger picture” again — a catch-all to deflect all moral and legal responsibility. While the judges sleep soundly, Amanda Knox and Raffaele Sollecito’s nightmare continues.


Andrew Gumbel, a regular contributor, spent five years as a foreign correspondent in Italy and co-wrote Raffaele Sollecito's account of the trials, Honor Bound: My Journey to Hell and Back With Amanda Knox (Simon & Schuster, 2012).

LARB Contributor

Andrew Gumbel is an award-winning journalist and author with a long track record as an investigative reporter, political columnist, magazine writer and foreign correspondent. He contributes regularly to The Guardian, among other publications, and has written extensively on subjects ranging from politics, law enforcement and counterterrorism, to food and popular culture. His books include Oklahoma City: What The Investigation Missed — And Why It Still Matters (HarperCollins, 2012), Down for the Count: Dirty Elections and the Rotten History of Democracy in America (The New Press, 2016), and, most recently, Won't Lose This Dream: How an Upstart Urban University Rewrote the Rules of a Broken System (The New Press, 2020).


LARB Staff Recommendations

Did you know LARB is a reader-supported nonprofit?

LARB publishes daily without a paywall as part of our mission to make rigorous, incisive, and engaging writing on every aspect of literature, culture, and the arts freely accessible to the public. Help us continue this work with your tax-deductible donation today!