Recent academic study has apparently been able to reduce Agatha Christie’s plots down to an formula which enables the reader to predict, within statistically tolerable boundaries, who has done the murder. That the formula is complex is inevitable – there are so many variables for Agatha to play with. That it relies on the reader spotting the relevant factors and is unreliable where these are missed is an equally inevitable constraint on its successful application. That is completely misses the point of reading Christie is surely also abundantly clear. For me, at least, its use is akin to tackling yesterday’s Times crossword with the solution open beside you. Reading Christie is about solving the puzzle she sets, differentiating the clues from the red herrings and reaching a conclusion based on the evidence presented – assuming you have not missed a vital clue through her masterly misdirection.
So it is with some trepidation that I offer the following analysis.
Its purpose is to explore features and patterns in Christie. And in this at least it shares some common ground with the research described above. But it does so not for the purpose of helping the would-be solver of the mystery – for that he or she must read the tales themselves. Its scope is broader than merely examining the murderer, but gives equal consideration to the victims and to the nature, location and method of the crimes. In this it shares some common ground with the studies of Kathryn Harkup, who has examined in minute detail the uses of poison in Christie’s work, in her book A is for Arsenic.
However, it considers solely the Poirot cases and, in order that it is not disproportionately concerned with murder – which is invariably the case with the novels – it does so only in respect of the 51 short stories which feature Poirot. For the record the stories include only the longer versions where two different versions exist (eg The Mystery of The Spanish Chest is used rather than the shorter The Mystery of the Baghdad Chest) and the Poirot version of stories which are also available in different form with another detective such as Parker Pyne.
Of the 51 short stories, three feature no actual crime. To avoid spoilers, in this, as in every other feature discussed, I shall not identify the specific stories (since there being no crime may, but may not, be the twist in the tale for a crime story). So we are reduced to 48 stories, featuring murders, thefts, blackmail and so on – some stories, of course, featuring more than one type of crime or, indeed, more than one instance of the same type of crime (though serial killers are a little thin on the ground, not that the term had been coined during the Golden Age). The actual breakdown is:
It may also be interesting to note in passing that fully half of the thefts are of jewels or jewellery worth fabulous sums. I suppose that just lends more colour to the story than anything so sordid as stealing bags of notes.
But who are the victims? What are they like?
In total there are 55 victims. Of these, the majority are men – 36, compared to 18 women. That ratio holds true for the murder victims – 25 male compared to 12 female.
The age of the victims is more difficult to ascertain. Rarely does Christie give a specific age of a person in her short stories although often a person will be described, for example, as “a few years over thirty”. The reader therefore has to infer from other information, such as having adult children who are working but do not themselves have children that they are perhaps in their forties or fifties rather than in their thirties. On the basis of such information, the age profile of the victims is as follows (in necessarily broad age bands).
I have separated out the murder victims from the victims of other crimes (theft, blackmail etc) because there is a distinct difference in the profiles. The majority (64%) of Christie’s murder victims are under 50 whereas less than half (44%) of the victims of other types of crime are under 50. Indeed, the difference is most marked in the 20s and 40s age bands where there are significantly fewer victims of crimes other than murder. Interestingly, there are very similar proportions of murder victims and victims of other types of crime who are in their 30s, which was the age-band Agatha had reached at the time most of the Poirot short stories were written. Perhaps she simply felt comfortable writing about people of a similar age to herself.
It is a widely held view that Christie wrote mostly about the upper- and upper-middle classes. England at the time she was writing these stories, between the two world wars, was indeed a very class-conscious society and there were nuances of class which are blurred or lost today. I have, therefore, attempted to reflect these differentiations in the analysis.
The data confirms Christie’s bias towards the victim being drawn from the upper- or middle-class. The interesting difference between the murder victims and the victims of other crimes is that socialites (celebrities??) are only ever victims of murder but not other crimes, whereas the rich businessman is more likely to be the victim of a crime other than murder.
So can we draw up a profile of the typical victim in a Poirot short story? We might suggest that the murder victim was most frequently male, middle class and in his 40s. The typical victim of other crimes might most frequently be male, slightly older, perhaps in his 50s and more likely to be a rich businessman.
A quick check reveals, almost inevitably, that there is in fact no example of a middle class, male murder victim in his 40s (though there 8 of the 37 murder victims – more than 20% – display two out of those three characteristics).
There is in fact one example of a victim of other crimes who fits the profile of a rich, male businessman in his 50s and there are five others who satisfy two of the three criteria (so that’s fully one-third of the 18 victims, which is starting to look more impressive).
But what of the murderer? Or the perpetrator of other crimes?
There is a marked difference in the gender of the murderers compared with the gender of perpetrators of other crimes. As can be seen below, murderers are overwhelmingly more likely to be male (87% if you add in the males who are halves of murderous couples or gang members), whereas there is an even split for perpetrators of other crimes. I have separated out murders carried out by couples as this is a distinctive feature of murder stories – there are no other crimes committed by husband and wife “teams”.
On the face of it this might be attributable to the theory that men have greater physical strength and so are more capable of carrying out a murder, but that pre-supposes that the method used requires an element of brute force. Given that Christie is supposed to have a preference for poisons – which make no such physical demands on the perpetrator – it will be interesting to examine the methods used in the Poirot stories to see if this has any bearing on the gender bias in Christie’s murderers.
However, before we look at that, we must consider the age of the perpetrators to build up our criminal profile. As with the victims, so too are the perpetrators’ ages rarely specifically mentioned so once again, it is necessary to infer the likely age band from the ancillary information about the characters – a father of teenagers is unlikely to be younger than his mid-thirties and may be somewhat older still, for example.
There are, it transpires, subtle differences in the age profiles of Christie’s murderers and other criminals. The murderers tend to be slightly older – we see them going on into their 60s in a way that doesn’t seem to arise in other areas of crime (maybe criminals retire like the rest of us?). And fewer murderers are in their 20s than commit other types of crime. Though it has to be said that age seems to be no bar to any type of criminal activity in Christie’s book – they are at it at all ages in very similar numbers.
Which brings us back to that archetypal English question of class. Does Christie have a better class of murderer?
To which the answer would appear to be yes. The vast majority of crimes are committed by upper- and middle classes. In the case of murders this is 77%; and 72% of other crimes. There are some notable peculiarities. Members of the medical professions (doctors and nurses) are disproportionately more likely to be involved with murders. Maybe it is the easy access to poisons? But there is an amusing discrepancy between the penchant for murder – an exclusively naval urge – and the willingness to commit other crimes where members of both Army and Navy are equally likely to be involved. Intriguingly show-business types appear as both murderers and other criminals. Did Agatha have something against the stage? Maybe that accounts for her reluctance to attend performances of The Mousetrap?
And what of the nature of the relationship between the criminal and his (and as we have seen in the case of murder it is more likely to be “his”) victim?
Here there is a very marked difference between the murders and other crimes. There is almost invariably a long-standing, and frequently close relationship between the murderer and victim. In the case of other crimes, however, there is almost always none – or at best acquaintance only. Although listed below as friends for the purposes of comparison, it frequently emerges that these “friendships” have often been struck up by the criminal expressly for the purpose of committing the crime.
It can be seen that the most likely culprit in the case of a murder is the spouse. Whereas for other crimes, in 61% of cases, there is no relationship beyond an apparent friendship or acquaintance. (“Criminal” in the chart above indicates there is no relationship other than that of criminal and victim of that crime.)
So our prime suspect in a murder story is a middle-class male, in his 30s or 40s, married to the victim. (Let’s forget that the typical victim is male – the Golden Age pre-dates same sex marriage by several decades!) Miraculously there is indeed one such example satisfying all four criteria. There are however, three other murderers who are middle class male in their 30s or 40s and a further three if we allow members of the medical profession and officers in the armed forces (both of which categories would be middle class in the period). Thus our profile (ignoring the relationship) fits nearly one in five cases.
In the case of other crimes, there is no clear profile as the the criminal is equally likely to be either gender and any age. They are more likely to be middle class but then the social setting for the majority of the stories is middle class. So that doesn’t help much. Perhaps the only guidance that can be taken from the data is to beware of military men and those in show-business with whom you have a slight acquaintance. And vet your servants carefully!
Another aspect to consider is the criminal’s motive.
The murderers’ motives vary. The most common is to speed up an inheritance. This is also a complicating factor in two of the murders “for love” – in an almost Trollopian regard for the need to finance whatever subsequent love-nest is desired. No doubt insurers will be relieved to note that murder for the insurance money is a rare occurrence (even in respect of spouses to be bumped off so that the way is clear for another).
Other crimes are, sadly, almost invariably motivated by money. Theft, blackmail, kidnapping: the end object is pretty much always the same. So nothing much in the way of interesting analysis of motives emerges there. The only exceptions are two later stories which feature attempts to frame someone as a drug dealer. One is left wondering what provoked Christie to use this plot device not once but twice in a relatively short space of time.
Now it has already been mentioned that Christie is said to favour poison as a method for her murderers to use. This, of course, draws on her own training in the hospital pharmacy when serving as a nurse for wounded soldiers during the First World War. Nevertheless, it is instructive to look at the means employed in her Poirot stories to see that this expectation is borne out.
It is perhaps worth considering that the poisons used included three uses of cyanide and two of arsenic at a time when such chemicals were more freely available to purchase than they are now. Pest control between the wars left a much scope for abuse! As did the prevalence of firearms, especially as a result of bringing home “trophy” weapons from the War. It is hard to imagine such conditions in the UK now with our much stricter gun controls. In her choice of methods, Christie simply allowed her murderers to use what was readily to hand in many upper- and middle-class households.
The “other” methods include being pushed from a train, being pushed downstairs and two unspecified methods.
Having proved that Christie did indeed prefer poison, did she also prefer that archetypal Golden Age setting for her crimes: the country house? On the face of it, the answer is no.
The location Christie used most often was somewhere in London, usually central London, though also sometimes in the suburbs. And for the record, where she has specified a suburb at all – or where it may be inferred from other information such as references to the nearest county outside the city (such as Essex or Surrey) – she shows no preference for either north or south of the river!
That said, lumping all London locations in together does sidestep rather than face head on the fact that the commonest type of location Christie uses is the country manor house. So perhaps there is quite a bit of truth in the stereotype.
She does, however, for a quintessentially English form, use foreign settings surprisingly often.
She is also fond of transport for her settings. As well as the two murders on trains, one of the murders and one of the other crimes “abroad” take place on a cruise ship and a train respectively while a further crime takes place in a moving car in a UK country (village) location.
Which brings me to a potentially controversial concluding section on Poirot’s own involvement in the stories.
You see it has always struck me – perhaps sacrilegiously – that being around Poirot is a pretty damn dangerous, nay even fatal perhaps, place to be. He may, like the Mounties, always get his man (or woman), but he doesn’t always succeed before another murder has been committed. So I thought it would be interesting to look at when he is called in relative to when the crime takes place. In short, how many times does someone request his help only to get themselves murdered before he has got himself around to investigating what is going on.
What I found, somewhat shockingly, was that out of the 37 murders he investigated, he was involved before they took place on 10 occasions. And he only foiled 3 of the possible murders. Okay, in fairness, some of the time he was merely a bystander before the murder took place and then investigated after the event but still… If I saw Hercule coming I’d run.
Of course, this being an analysis of crimes – albeit fictional crimes. It is impossible to end without considering the solving of those crimes.
Here, this being Poirot, his clean up rate is 100%. He always solves the case but, interestingly, he is concerned with justice in its broadest sense. So the criminal doesn’t always face the justice of the legal system. Two killers are allowed to go free – one because it was an accidental killing of a blackmailer and the other because the killing was done to prevent the “victim”, who was himself a murderer, from killing again. And 5 of the 18 criminals, other than murderers, were also allowed to walk away, either on the grounds that they had learned their lesson or their crimes were not too serious.
This is more important than it might appear. At the time these stories were written, the penalty for murder was death. So Poirot’s success in identifying the murderer was, in effect, a death sentence. Christie was of a generation that had no problems with this (though interestingly Dorothy L. Sayers’ amateur detective Lord Peter Wimsey, perhaps as a result of his experiences in the War, appeared to suffer agonies over this fact – not that it stopped him solving crimes at the time, you understand, but afterwards he was often seen to be deeply conflicted by the results of his actions). Yet she allows Poirot occasionally the latitude to extend his mercy to those for whom he (and therefore Christie, we may speculate) felt that the full rigours of the law would be an injustice.