Penguin no. 58: The Poisoned Chocolates Case by Anthony Berkeley

"How often don't you people at the Yard stumble on some vital piece of evidence out of pure chance? How often isn't it that you're led to the right solution by what seems a series of mere coincidences? I'm not belittling your detective work; but just think how often a piece of brilliant detective-work which has led you most of the way but not the last few vital inches, meets with some remarkable stroke of sheer luck (thoroughly well-deserved luck, no doubt, but luck), which just makes the case complete for you. I can think of scores of instances...Don't you see what I mean? Is it chance ever time, or is Providence avenging the victim?"

The police have exhausted their leads and all but abandoned the search for the murderer of Joan Bendix, a woman who met her death by sampling from a box of poisoned chocolates, apparently the inadvertent victim of a murder plot which misfired. The chocolates had originally been received in the mail by Sir Eustace Pennefather, a predatory and indiscreet womaniser, and a man certain to have many enemies amongst the discarded mistresses and cuckolded husbands he leaves in his wake. But in a fit of pique, he passes the tainted sweets onto Bendix, who takes them home and shares them with his wife; later that afternoon Bendix is taken ill, and his wife dies. The police now speculate that the murderer may have been a pious though deranged individual intent on ridding the world of the menace that is Sir Pennefather. It is a convenient hypothesis, and one that explains why they have been unable to solve the crime.

But this puzzle which baffles the police offers the members of the Crime Circle an intellectual challenge, and an opportunity to test their wits against both the Yard and each other. The Crime Circle functions as an exclusive club facilitating discussion between individuals with an intense interest in criminology. To date only six people have passed the rather formidable entrance requirements: four writers, one barrister, and the rather timid and unassuming Ambrose Chitterwick. Through this organisation Anthony Berkeley parodies the real life Detection Club, an enterprise he established in partnership with Dorothy L. Sayers, and which included amongst its members Agatha Christie, G.K. Chesterton and John Dickson Carr.

After being provided with all the information the police have on the murder, each member of the Crime Circle is charged with the task of independently investigating or reflecting on the crime for one week, and then offering a solution. They plan to present their findings on consecutive nights, in an order specified in advance, and determined randomly. It follows that each 'criminologist' must be prepared to argue their case in front of their peers, and these are people who are competitive, each certain their solution is the correct one, and each equally determined to destroy the opposing cases. This competitive element is the great strength of the novel, for it mocks the tradition of the single detective, alone capable of making a correct inference from the evidence, and the only one with sufficient insight to identify the criminal. Over the successive nights it becomes clear that despite seeming to convince their audience when given the stand, each criminologist is only partially right, and each is partially wrong, but collectively their different perspectives map out a path which ultimately uncovers the truth.

Perhaps more importantly, it also becomes clear that each of the sleuths is to some extent biased. Sometimes this is because they form their hypothesis of the murderer's identity first, and then selectively search the facts for the ones which support their view, and sometimes it is because of how they view the world, of the importance they place on psychological or historical factors. In all cases there is a tendency to infer from any information available the interpretation which best suits the case they are developing. And they are selective in applying their critical faculties, accepting without question the aspects which support their individual cases, while minimising or dismissing those which don't.

And when it comes to the presentation, each criminologist behaves almost as an illusionist, using rhetoric and drama to command the audience's attention, while carefully smuggling past them the defects in the case. And interestingly, even at this level of expertise, and with this level of scrutiny, the reasoning is not faultless. Morton Harrogate Bradley, for example, gives almost an almost perfect example of proof by Prosecutor's Fallacy, and though his logic is exposed by a conclusion which is clearly absurd, and although his critics question his assumptions, the flaws in his reasoning are never actually highlighted. And yet this is a real problem, and not only in detective fiction, for confusing unlikeliness with a probability of innocence has led to mistaken convictions.

Anthony Berkeley developed this novel from his short story The Avenging Chance, and in the process created something original, clever and subtly critical of the genre's conventions. He gives six capable and confident sleuths the same set of clues and gets six different, though apparently provable answers. Detection is not quite as straightforward, even for the most talented of criminologists, as it is often made to seem.

Professor Peter Donnelly: How juries are fooled by statistics (on youtube)

By the same author:
Penguin no. 590: Trial and Error