Pierre Bayard, Who Killed Roger Ackroyd?

Editions de Minuit 1998 / The New Press 2000, pp. 59-65

Part II, Chapter 3, Improbabilities

Apart from the general structure of the book, which is a machine for producing the undecidable, Poirot's solution, when we examine it more closely, reveals a fair number of improb­abilities. Pronounced too quickly in a state of misleading eu­phoria, it hardly resolves the many problems raised by the investigation.

Provided the novel is read carefully, it contains such a great number of improbabilities that, for the clarity of exposition, we will group them into three categories: those that concern the hours preceding the murder; those that involve the act of murder itself; and those that concern the period following the murder.

To investigate the period before the murder is to investigate the motive. Moreover, Sheppard clearly poses the question to Poirot:            "What on earth had I to gain by murdering Ackroyd?", to which Poirot replies in these words:

"Safety. It was you who blackmailed Mrs. Ferrars. Who could have had a better knowledge of what killed Mr. Ferrars than the doctor who was attending him? . . . If Ackroyd had learnt the truth he would have had no mercy on you - you were ruined for ever.

Let us consider, for a moment, that Sheppard may indeed be the blackmailer. Just how would his safety have been threat­ened, and, following the hypothesis that it were, how would the murder dispel this threat? If he is guilty, Sheppard is not taking much of a risk. A blackmailer does not use checks, and transfers of funds cannot be traced.

Moreover; the epistolary confidences of a murderess - Mrs. Ferrars, in her letter of accusation, owns up to her husband's murder - who is dead besides, may be easily discredited. In the face of a letter of accusation, delivered posthumously, the most reasonable attitude is denial, attributing such accusations to amorous spite or any other motive, and putting the burden of proof on the investigators, who will find it very difficult in the absence of the victim.

But this is the case only if the murder solves the problem. There is no guarantee that Mrs. Ferrars confined herself to a single letter and did not send a second - to the police, for example. If a blackmailer's victim is determined to harm him, she may do more than simply denounce him to her lover - especially when she eliminates further risk by committing suicide. How could Sheppard have known that only one letter accusing him existed, and that Ackroyd's murder therefore made sense?

Which leads us to another problem aside from motive, namely, the psychology of the alleged murderer. There is a fundamental contradiction between the suggested portrait of a cold and calculating mysterious killer and that of a Sheppard going mad at the thought of a danger that is nonexistent (or at least that should be confirmed), at risk of imperiling his safety - this time for real - by committing a murder. These two images do not coincide.

More broadly, Sheppard, presented as a weak-willed fellow, hardly fits the occupational profile. These character traits are confirmed several times by his sister, but also by other characters, notably Poirot. Yet this weakness does not square with the decisive frame of mind the murderer must have had to kill Ackroyd only a few hours after the announcement of Mrs. Ferrars' death.

Improbabilities connected to the act of the murder itself are still greater, and rest for the most part on the fact that Shep­pard has very little time at his disposal between the moment he learns of Mrs. Ferrars' death that morning and the time he is supposed to have killed Ackroyd that evening.

Concerned as he is with establishing to the minute how Sheppard occupies his time, Hercule Poirot does not seem to pay attention to a simple problem. Let us carefully review the doctor's activities on the day of the murder. His morning is devoted to medical visits, then to seeing pa­tients at his office. He lunches with his sister, and afterward goes into his garden to do some weeding. This is when he makes the acquaintance of his neighbor, Hercule Poirot, with whom he has a long conversation. At the end of this talk, he goes back into the house, where he has a new discussion with Caroline. After a certain time, on the pretext that he has a client to see, he goes off to the Three Boars Inn, where Ralph Paton is staying. The book communicates to us only the be­ginning of the conversation between the two men, and, after a gap in the account, we find Sheppard in front of the house at Fernly Park, a little before 7:30 P.M.

The text does not give us Sheppard's precise schedule, and we have to entertain several conjectures. Suppose that his lunch with his sister took place at around 1:00 P.M. and lasted an hour. The long conversation with Poirot might then have taken place between 2:00 and 3:00, the shorter talk with Caro­line between 3:00 and 3:30 P.M., the meeting with Paton be­tween 4:00 and 4:30 or 5:00. We would also have to add travel time between the Three Boars and Sheppard's house, and be­tween the doctor's and Fernly Park.

Why are these precise details of schedule so important? Be­cause the killing of Ackroyd, at least in Poirot's version, not only requires the possession of a weapon, it also implies the construction of a sophisticated device, namely a kind of pro­grammed timer that Sheppard must invent and construct from scratch out of a dictaphone. The question that arises, then, is When? The time for the conception and construction of this object is very brief since Sheppard learns only that morning that he might be in danger, and until then had nothing to worry about. But the time for manufacturing such a complex preci­sion instrument is even briefer, since Sheppard has a maximum of two hours in a house where Caroline leaves him no respite, and where he has little hope of isolating himself to fabricate (this is 1926) what must have been the first clock-radio in technological history.

Ackroyd's murder does not depend only on the lightning-quick fabrication of a timing mechanism. It also depends on a pair of shoes, which Sheppard is supposed to have filched from Paton in order to leave misleading footprints.

Several improbabilities are connected to this enterprise. Paton, though merely passing through King's Abbot, has to have brought along several pairs of shoes, or he would have soon noticed the theft. How could Sheppard have been sure of this in advance? We later learn that he returned to the inn when Paton was gone in order to steal the shoes. This makes some sense, since clearly, it would not be so easy to steal a pair of shoes from a room at an inn in their owner’s presence. Shep­pard is then reputed to have brought them in his medical bag. This, too, is not completely impossible, even if we must imag­ine him armed with an enormous sack, since he is already carrying a weapon and a dictaphone in it besides! Finally, in Poirot's logic - for otherwise Sheppard would incriminate himself as the person who stole them - Sheppard has to return the shoes to the inn, where he again runs the risk of being noticed.

All this is extremely difficult to execute but not impossible. On the other hand, another point is entirely improbable. When Poirot, in a turn of phrase, casually mentions the shoes Shep­pard must have stolen from Paton to fabricate false clues, he tends to underestimate the difficulty of imitat­ing someone's footprints. Indeed, we learn during the investi­gation that the weather had been dry, and it was thanks to a chance bit of damp gravel that the tracks had been identified. Now, when Sheppard was plotting his murder, that is something he could not have foreseen.

Even more surprising than the dictaphone and shoe escapades is the telephone call, on which Poirot's solution hinges almost entirely. Let us recall that on his return from Roger Ackroyd's house on the evening of the murder, Sheppard receives a tele­phone call attributed to Parker, the butler, supposedly inform­ing him of his master's death.

We concur with Poirot and see this as the key to the whole affair, but by interpreting it in a radically different way. As Poirot sees it, Sheppard wants to be sure he is at the scene of the crime when the body is discovered, and has asked one of his patients, a ship's steward leaving for America, to call him from the train station. He then tells his sister he has received this call from Parker and must go immediately to Fernly Park.

Far from incriminating Sheppard, the story of the telephone call tends to suggest his innocence if we apply a modicum of common sense. Even assuming that he needs to be at the scene of the crime at that moment, we may well wonder why he resorts to such a complicated and risky procedure, which is so dangerous and so useless as well. The complication is obvious. It is risky since Sheppard's entire strategy rests on the hope that the steward will not forget to call. The danger is amply dem­onstrated by what happens, namely, that the police easily man­age to trace the call.

But most of all, the telephone call is perfectly useless, and this is why it suggests Sheppard's innocence. If he had wanted to be on the spot, he could easily have done what everyone (including characters in other Agatha Christie novels) does who wants to return to a place they have just left: pretend to have forgotten something, such as his valise or some other medical object needed for his visits the following day. This is a simple, foolproof method that involves no external partici­pation or any lie that the police might exploit.

This ruse would also have the advantage of resolving a problem that the solution imagined by Poirot leaves un­touched: For Sheppard's presence at the crime scene to have any meaning, he needs to be left alone in the room to retrieve the dictaphone and slip it into his bag. He manages this by sending Parker, the only one awake, to inform the police. But the risk is great, in Poirot's version, that the other inhabitants of the house will wake up, informed of Ackroyd's death by the doctor's arrival.

Sheppard's attitude after the murder is not very clear. First of all, it is difficult to understand why he devotes so much energy to putting the police on the right track. After Ackroyd's de­mise, there is no longer anyone, by definition, who might know about the blackmail of Mrs. Ferrars. Yet it is Sheppard himself who informs the police of this blackmail and tells them about the blue envelope.

Another cause for surprise is Sheppard's attitude toward Paton. As we know, the doctor spirits the young man away from the investigation by hiding him in a psychiatric hospital. This initiative can be explained if Sheppard is protecting the young man. It becomes more enigmatic if he is the murderer, and, curiously, Poirot does not raise this point during his ac­cusation. Just how does concealing Ralph Paton help to dis­guise the truth? (We can see how effective it is when Poirot forces Paton from his refuge.) Save by killing him, too, Shep­pard cannot hope to keep Paton sequestered forever. Once the young man is free, Sheppard's part in his concealment will necessarily come out.

Sheppard's attitude after Poirot's accusation is also quite odd. This man who is reputed to have killed in cold blood to protect himself does not even dream of denying Poirot's recon­struction, which nonetheless rests on extremely shaky ground and contains no real proof. The small discrepancy in Shep­pard's schedule, like the displacement of the highbacked chair, hardly merits much attention. More seriously, the telephone call may have a different explanation from Poirot's, as we shall see. Sheppard, however, offers no protest, thanks Poirot for the evening, and runs off to commit suicide. Strange behavior, requiring a few explanations.

This series of improbabilities does not lead us, we see, to proof positive of Sheppard's innocence or a declaration that he could not possibly participate in the murder. But Poirot's accusation leaves a number of gray areas. It cannot be excluded that Sheppard is the killer, but it is also possible that he is not. In any case, a jury would be unlikely to condemn him on such flimsy evidence, with so many elements unexplained and no attempt to clarify the remaining murky facts. All indicators make it desirable, then, to reopen the investigation.