Guest editorial: Scott Ratner on The Myth of Detective Fiction: “Fair Play”

This is the first time I’ve offered space to a fellow Golden Age of Detection enthusiast to express his views, but I couldn’t resist this opportunity to bring this interesting material to a wider audience. Scott Ratner and I have gotten to know each other through a Facebook group devoted to Golden Age Detection (GAD) as fellow aficionados who share an interest and have gone deeply into it; our views are generally similar, but occasionally quite different. As it should be. Over time, I’ve come to respect his knowledge and analysis.

I’ve known for a long time that Scott has a well-developed argument about the words “fair play” in the mystery context, and I’ve  read short comments that interested me in hearing the full argument.  Recently, in the course of a wide-ranging discussion on various GAD topics, Scott mentioned that he wanted to lay out this argument, but didn’t have anywhere to publish the result; I offered him the space below.

To the best of my knowledge, there are no actual spoilers in the material below but it’s possible that you will learn more than you wish to about the plot and construction of various Golden Age mysteries by a number of authors. I’ll approve on Scott’s behalf any comment that seems relevant to the discussion (I draw the line at advertisements disguised as general praise) as fast as I can manage.  Scott’s opinions are his own; I’ll comment or not as I see fit, and I didn’t edit his work (although I’m sure I reflexively corrected a typo or two; I can’t help it, it’s a disease).

Thanks to Scott for his contribution — I hope you enjoy it and find it thought-provoking!

The Myth of Detective Fiction: “Fair Play”

by Scott Ratner

Time to ruffle some feathers. I’ve already upset and inadvertently insulted someone I admire with my views on this subject, but I know that that’s no good reason to deny my own convictions. And please note this disclaimer: if the arguments I present do not all seem to hold, please consider that it may be not that the ideas themselves are unsound, but rather that my ability to convey them is weak. At any rate, here goes:

“Fair play” is one of the key and most oft-cited principles of Golden Age and Puzzle Plot Detective Fiction. However, what is rarely examined is what that term really means, how it can be measured, and whether it even really exists in relation to the genre.

First, it should be noted that “fairness” (and by this term, of course, I mean its definition relating to equitability, not lightness of hue, or or attractiveness) is always treated as an objective concept, and always considered in reference to a presumed exact and objective standard.  Our language reflects this: we speak of “fairness” in binary, “lightswitch” terms– things are either “fair” or “unfair.”  Moreover, the very fact that questions of fairness are disputed is evidence of its perceived objective status; subjective concepts cannot logically be disputed– one may argue the merits of a work, but a sincere subjective statement such as “I don’t like it” is inherently and inarguably true– the maker of the statement is the sole arbiter… he doesn’t like it!

As with the concept of justice, we may not agree upon where the standard of fairness lies, but recognize that, if it indeed exists, it exists independent of our personal judgment. A phrase such as “that’s more than fair” further demonstrates a recognition of the exactitude of that standard, suggesting a level of generosity beyond it. Even such subjective statements as “that strikes me as unfair” or “it seems fair to me” do not imply a subjective standard, but rather indicate a subjective understanding of an objective standard; that is, they assert “the line of fairness exists, and I believe this is where it lies.”

This is an intuitively understood notion, and its value is realized even by the small child. The child cries, “It’s unfair!”, and while he may be feeling merely that wants more of something or that he is unhappy with the treatment he is receiving, he appeals to this presumed objective standard, a threshold above which he is being treated fairly, and below which he is not (in many cases with children– and even with adults– this is equated with equal treatment: “you let Tommy do it!”). He realizes, even at this early age, that reference to this standard carries more persuasive weight than a mere expression of his desire;  even if all the grownup  responds with is “no, it’s not,” in disagreeing where the standard lies he is confirming the concept of the standard, and that it is a valid basis for decision. For many children, this is perhaps their earliest attempt to get their way via reason; realizing that while they can only express a desire, they can argue a point of fact (fair or unfair).

The concept of “British Fair Play,” which is most probably the direct source of its use in detective fiction, may seem more casual and inexact, based on a personal, subjective sense of “gentlemanly” conduct– indeed, one might think I’m taking the whole matter too literally. But this use of the term is also integrally related to the others, and just as solidly tied to the concept of an objective standard. It is a reference to the very rigid and explicit rules of British sports (“it’s not Cricket!”) and military regulations, which are in turn presumably based on the “real,” objective standard of fairness. Thus, while our personal decision of what constitutes giving an enemy or opponent a “fair” or “sportsman’s” chance may be entirely intuitive, that intuition is presumably based on what is truly fair, independent of our belief.

The point of all this is not that there is necessarily an exact, objective standard of fairness (I don’t really know if there is), but rather that the concept is always treated as such, and that every use of the term “fair,” “fairness,” or “fair play” implies and references such a standard, regardless of its actual existence.

So, how does this apply to the detective fiction genre? Well, in citing fair play, the reader of such a work is holding it up to an subjectively felt, though recognized-as-objective standard. And because he recognizes the standard as objective, if he feel the work falls short of it he does not complain that “this is not satisfying to me!” but rather that “this is unfair!” However, unlike with the child, it is not sufficient for the author to reply “No, it’s not!”– not sufficient, that is, for either his sales or his pride. It is important to him that the reader believes that the standard has been met. And that’s where the “rules” of the genre fit in. They are cited to define the standard of fairness, to arbitrate whether a work is fair or unfair.  But can they really achieve this?

In regard to one aspect of detective fiction, I believe they can. That is the realm of what might be described as “narrative fairness” (not a particularly satisfactory term, but I’ve not been able to come up with a better one). By “narrative fairness” I am simply referring to the question of which techniques the author is or is not allowed to employ in the “telling” of the tale. People may argue about what be the rules should be, but at least regarding this aspect it is possible to establish and cite clear-cut rules.  I myself subscribe to Dorothy L. Sayers’ notion in that there is only one thing an author may not do in this respect, and that is to make a false statement “on his own authority.” In other words, a third-person narrator cannot lie. This does not prohibit the author from employing deception– deception by omission, deception by misleading inference, or falsehoods by first person narrators, who, as Sayers reminds us, are “not necessarily the author.”  Thus, the Christie’s Murder of Roger Ackroyd is exonerated on several counts (it’s rather stunning how “clean” this once-controversial book is in this regard), while a rarely-questioned work such as Death on the Nile turns out to actually be unfair, based on an extremely minor technicality. A book such as Carr’s Seeing is Believing is admittedly difficult to judge, but that doesn’t affect the rule– the question of whether it plays fair depends upon how one interprets the tricky ambiguities of the English language. Similarly, the narrative fairness of Christie’s A Murder in Announced must  based on how we answer the question of whether that which we call ourselves is our true name. Whether these works follow the rule is in question, but the rule itself remains constant. Now, others my argue that narrative fairness consists of more or less than my (or Sayers’) single rule, and I’m not insisting that I’m right about it. I’m just pointing out that that it is possible to define clear-cut criteria for this question, and judge works according to it.

But what about the issue of clue sufficiency? Here’s where it all blows up. Let’s look at some of the offered “rules” regarding this question. The first category would be those rules that state “the reader may not be denied any clues granted the detective” or “the reader must have equal opportunity with the detective for solving the mystery” (there are several other versions of this which say the same thing). And that’s fine as far as it goes– I’m sure that most would agree that fairness dictates that the reader is provided with all the clues granted the detective. The problem is, it’s a rule with no minimum standard. For, if that’s all there were to it, a story in which the detective arrives at the solution based on little or no evidence must be deemed fairly-clued, just as long as the reader has been provided with the same sparse or nonexistent evidence. As you can see, that rule really gets us nowhere.

Nearly all other clue sufficiency rules consist of variations of the idea that “the reader must be provided with all the clues necessary to solve the case.” This initially appears to be much more useful, until one faces the task of defining or measuring its terms. What is really meant by “all the clues necessary”? Indeed, what does it even mean to “solve the case”? (I can’t help thinking of Robert Benchley’s hilarious “Does the average man get enough sleep? What is ‘enough sleep’? What is ‘the average man’? What is ‘does’?). Seriously, though, what does qualifiy as “solving” a mystery? If a reader has arrives at the solution of mystery thru sheer guesswork or an arbitrary hunch, can he be said to have solved it? If not, does the fact that a reader has employed indications (clues) provided by the author to arrive at the correct solution mean that the he has “solved” the mystery?

Suppose that I arrive at the solution that Phillip Latterby was killed by his nephew Nigel based on the fact that Nigel owned the crossbow employed in the commission of the crime, and that Phillip had stated that he planned to disinherit him. Can I be said to have solved the mystery if that turns out to be the correct solution? If so, then what about another reader who decides that the culprit was Phillip’s wife Adeline, who may have stolen the crossbow from Nigel’s’s room (it had been earlier established that she had once been arrested for shoplifting), and whose disagreement with Phillip’s political beliefs was well known? Is this reader less correct than I am, or is he justified in claiming that the author was not “fair,” that he had not provided the reader with “all the clues necessary to solve the mystery”? Again, we are referencing some invisible but objective standard.

The question, then, is clearly:  how many indications qualify as “enough”? How many constitute “all the clues”? One? Five? 50? Outside of the standard of “some” clueing (which means at least one clue– and I doubt that many would agree that the inclusion of a single clue guarantees that a work is sufficient to be called fairly-clued), there is only one standard of clue sufficiency that can be clearly defined and universally agreed upon as sufficient, and that is the standard of total deductive provability.

Now, total deductive provability is a great, solid standard, against which no cries of “unfair” could ever be raised, but unfortunately it entails certain problems in relation to detective fiction, not the least of which is that no works of detective fiction have ever met it! A bold statement, I realize, and one that I certainly can’t back up from personal knowledge– I haven’t read (nearly) all works of detective fiction. There is certainly the possibility that I am wrong about this point. But I have read a great deal of the most lauded works of the genre (all of Christie, most of Carr, Queen, Berkeley, Brand, and several others), and none of what I’ve read (or heard about) suggests that there are any works that qualify.

Admittedly, there are occasional works that prove that “x and only x could have committed the crime” (though even these are rarer than it would seem, as the “logic” that “proves” this point is more often than not flawed). However, even those works that do arrive at this point by unassailable deductive logic do not meet the standard, as the solution to the mystery in these books never (in my experience) consists solely of this single point.

Rather, the solutions to detective stories (presumably) all consist of a scenario of contentions, some of which may be arrived at deductively, but which are all linked together by abductive reason (inference to the best explanation). This abductive link itself can not be proven, and very often the details it connects (and which subjectively strengthen the credibility of the solution) cannot be deductively proven either.

A large category of such details are behavioral discrepancies, clues which very often (in my personal opinion) offer the most fascinating, satisfying and convincing of evidence, and yet which can never be deductively proven. Examples of such behavioral discrepancies are the suddenly heightened volume of Simon Doyle’s voice in Death on the Nile, Avory Hume’s abrupt apparent change in attitude toward Jimmy Amswell in The Judas Window, and the uncharacteristic comportment of the two Generals in Chesterton’s The Sign Of The Broken Sword. The solutions of these stories not only explain these discrepancies, but are made more interesting and convincing by them. The explanations fit in with everything else in the solution, and reinforce the solution’s sense of inevitability. Yet none of them can be deductively proven, as there are countless other possible explanations for these behavioral discrepancies. For instance, Simon Doyle’s sudden vocal volume increase might have been due to the fact that at that moment he felt a sudden surge in pain from his injured leg. Or, he may have suddenly gone deaf in one ear and was attempting to compensate. That such explanations have no clues to support them and do not otherwise bolster the solution is of no importance; the point is that they are no less provable than the more satisfying explanations ultimately given, and in fact no less logically possible.  Furthermore, not only are the explanations to behavioral discrepancies unprovable, they in turn prove nothing.

Are then works that consist largely or solely of such clues—works that are richly and satisfyingly clued (IMO) and include many of the most lauded works of the genre—“bad” detective stories? Or are they not even detective stories at all? Carr, Van Dine and others call the genre a game, but if these works cannot “play fair” (which, as we’ve seen, is nearly impossible to do), do they not qualify as of the genre? Certainly Christie’s Five Little Pigs and Chesterton’s Father Brown stories (again, for me and others, beautifully and satisfyingly-clued) never strive for anything even remotely approaching total deductive provability (heck, not even partial deductive provability)—are they not legitimately detective stories?

Note, moreover, that any (possible) detective story of total deductive provability would also have to exclude motive as part of its solution. After all, due to the impenetrability of the human mind, motive can never be deductively proven. Sure, we might be able to prove that Uncle Phillip threatened to disinherit his nephew Nigel, that Nigel threatened Uncle Phillip (“I’ll kill you before I let you change you will!”)… even that he DID kill him, and shouted afterward “I killed him because he was going to change his will!” But we still cannot prove that that was the reason he killed him. All that we can prove is that he had a strong possible motive. (Though people often refer to a strong possible motive as a motive, only the actual desire to commit a crime [or other action] constitutes an actual motive. Otherwise, any person with a weak possible motive [“I’ve never cared for Australians”] must be deemed to have a motive for, as with the matter of clue sufficiency, there is no way to objectively define the threshold between weak and strong possible motives).

And, as I mentioned before, even if we were able to deductively prove all the individual points of a detective story solution (which would be an incredibly tedious and lengthy process), we would still not be able to prove the abductive chain that links them (the cause and effect relationships  between them). So then, am I suggesting that the greatest works of the Golden Age masters are all failures? Well, set against the either uselessly vague or virtually unattainable standards of “fair play” I’d say… yes, they are.

Now, before anyone brings out the tar and feathers and starts referring to me as the “21st-century Edmund Wilson,” let me state emphatically that I love Golden Age Detective Fiction! It is my favorite genre, and John Dickson Carr is my favorite author (with Agatha Christie running a close second). Further, I consider their greatest works (along with those of Queen, Berkeley, Brand and several others) as masterpieces of their art. But I consider them brilliant examples of what they are, not of what they’re not, just as I consider Twelve Angry Men a triumph of drama and a failure as a musical comedy. The fault then, dear Brutus, lies not in these works but our model. And that faulty model– that model that does not fit the genre– is that of the “game.”

Now, there’s no doubt that games and games-playing were extremely important to the world of Golden Age Detective Fiction. The people who both wrote and read GA fiction were by and large games-playing people, the type that Anthony Shaffer memorialized with the character of Andrew Wyke in his play Sleuth (though most of them were presumably more likable and kind-hearted than Wyke, of course). Games were indeed all the rage in that era, and it is quite natural that a type of fiction bearing resemblances to games would be appealing to those people who reveled in playing them.  Games-playing and GA Detective Fiction undoubtedly fed and fed off each other. But resemblance is not the same thing as identity, and just as singing at a karaoke bar does not constitute a concert, I maintain that a work of detective fiction is fundamentally distinct from a game.

Of course, much depends on how one defines the concept of a “game.” There are many definitions out there, some of them admittedly broad enough to include detective fiction, but those definitions are also broad enough to be of no use in resolving the question. For instance, the first definition of a “game” on is “an amusement or pastime.” Well, yes, by that definition, a detective story clearly is a game, but then so is watching The Sound Of Music. That really doesn’t help us, I’d say. One might enjoy or not enjoy The Sound Of Music, but the mere watching of it does not constitute playing a game, and even those who do not like the film wouldn’t claim that is unfair in not giving the viewer sufficient opportunity to “win” (whatever that would mean in this case).

Another “game” definition (same source) is “a competitive activity involving skill, chance, or endurance on the part of two or more persons who play according to a set of rules, usually for their own amusement or for that of spectators.” This is clearly closer to the definition we seek, but it matches rather badly with the genre, as we’ll see below. But let’s first take a look at the descriptions provided by the people who were actually insisting on the connection in the first place. First, here’s the way S. S. Van Dine put it:  “The detective story is a kind of intellectual game. It is more – – it is a sporting event.”

Similarly, John Dickson Carr wrote: “It is a hoodwinking contest, a duel between author and reader.”

So, what both are suggesting (and also corresponding to that second, more specific dictionary definition) is that, more than just a pastime, detective fiction is specifically a competitive match, a battle of wits between the author and the reader. But is it? I’d call attention to several points that illustrate the distinction between this pastime and all other competitive games. I’ll start with my weakest assertion.


As I mentioned, this is the weakest of my assertions, and I don’t expect everyone to agree with it. But when I read the works of Golden Age Masters– who clearly never knew of me or my level of intellect– am I really competing against them? Doesn’t their lack of opportunity to rebut or parry against my moves disqualify it as a competition? And if I’m able to arrive at both the identity of the culprit and the motive of a mystery prior to the author revealing it (as I did with Christie’s The Body In The Library) did I actually outwit them? It would be nice to think of myself of as the man who outwitted Agatha Christie (clearly I’m much more clever than she), but I don’t honestly consider it an valid claim.

And even if we do accept the idea that someone totally unaware of us (and who in certain cases has died before we were born) can be competing with us, it certainly gives detective fiction a unique status among games. Admittedly, in such activities as crossword puzzles, the puzzle has been designed without knowledge of us or our intellectual capacities (and the crossword puzzle deviser might too have died before we were born). But no one refers to a crossword puzzle as a competition or battle of wits between the person trying to fill in the answers and the puzzle deviser. And there is also another important distinction between a crossword puzzle and detective fiction…. :


I suspect that many who buy into the detective-story-as-game scenario think this one is covered. What about, they may say, the lists of rules set forth by Van Dine, Knox, Gorell, Milne, even Carr? To which I call attention to one monumental point they’re overlooking… the matter of just who these rules are written for! Van Dine’s rules are titled “Twenty Rules For Writing Detective Stories” and, similarly, the rules propose by Knox, Gorell, etc… are all placed upon the writers of the stories. If the detective story is, as proposed, a competitive match between the author and the reader, where are the rules that the reader must follow?

To my knowledge, none have ever been suggested, let alone laid down as law. I can only think of one possible rule placed upon the reader, and that is the tacit rule that he mustn’t peek at the end of the book. But whereas the reader may call “foul” at the writer not following the “rules” (whosever’s rules they choose to appeal to), no one is insisting upon (or even mentioning) that the reader must heed the “no-peek” rule– it is both unspoken and self-enforced. What other competitive game lays down rules for players on one side and not on the other? None which I can think, which brings up the next distinction.


This axiom applies to all competitive games, from thumb wrestling to baseball to hopscotch to championship chess.  When the players themselves call the decision (as in, say, a card game) it is in reference to a specific set of rules, calling upon such rules to provide an objective arbiter of victory. Other competitions do admittedly have more subjective rulings (e.g. a beauty pageant, a dog show, or a singing competition), but these too are presumably following specific set guidelines and, more importantly, in such cases the judges are not the players themselves but external arbiters. The outcome of the detective fiction “game” is neither decided by mutual assent of the players (J.D. Carr is not there to agree that I outwitted him), nor is there an external judge deciding the outcome (“No, Scott, you did not properly solve this one before Ellery Queen revealed it. I’ll be back next Thursday, and have the check postdated”). No, the outcome of a detective fiction match is decided by a judge solitary, subjective and “of the players”… the reader himself. And what if that reader arrives at a solution he deems superior to the solution subsequently revealed in the book? Was he wrong? Did he “win” or did he “lose”? Who is to make the call? Not only is the reader himself not an objective arbiter, but he has no standards to appeal to other than varying, unstandardized sets of “rules” (we play cards according to Hoyle, but are we playing the detective fiction game according to Van Dine? Knox? Carr?). Further, the most frequent grounds for crying “unfair”– insufficient clueing– has, as we’ve seen above, either no objective standard to appeal to, or else an objective standard that is never met. In essence, only the gut of the reader can decide whether he is victorious, and certainly no other competitive game is decided by the subjective belief of one of the players.

So far, I’ve noted that in at least three important ways, detective fiction is unique from competitive games of the type suggested by those who promote the “whodunit-as-game” theory: it has players often unaware of each other’s existence, it has no rules set forth for players on one side, and it offers no objective (or external subjective) arbiters of success. I think these points alone are enough to raise serious doubts that detective fiction falls into the category of games. But I believe the fourth distinction puts it beyond doubt:


No doubt, one can enjoy playing a game even if one loses it. And there also unrelated reasons for desiring to lose a game (“If I let her win, she’ll sleep with me, give me the promotion, etc…”). But I can think of no game which many people play actually hoping– for no other ulterior reason– to lose. Yet, there are many, many people (myself included) who would a actually prefer to “lose” the detective fiction “game.” For, if detective fiction were indeed a game, “winning” (for the reader) would consist of correctly arriving at the solution to the mystery prior to it being revealed by the author, and “losing” would mean not anticipating it (or arriving at an incorrect solution). And a substantial portion of the mystery reading public would actually rather be proven wrong, to “lose” under this definition. Why? Because, if the author is able to successfully conceal the truth from them until the moment at which he chooses to reveal it, the reader may experience– in the dramatic way the author intended– a pleasing sense of “sudden retrospective illumination” (or paradigm shift, or epiphany, or in Aristotelian terms, anagnorisis)– that is, the sudden simultaneous sense of surprise and inevitability.

If you are not among the people who prefer this sensation to correctly anticipating the answer, I invite you take a survey of fellow mystery readers. I’m not suggesting the that our way of enjoying detective fiction is superior to the other, only that we constitute a substantial portion (perhaps even majority?) of the mystery readership.

Why then, one might ask, do we “hopeful losers” still try to solve the mystery while reading it? Well, I certainly can’t answer for everyone here, but I can explain my own reasons. I try– earnestly and intently– to solve the mystery, all the while hoping in my heart to be proven wrong because, if the author can surprise me with a richly clues-solution I had not foreseen despite my best (and frankly, “seasoned”) efforts to anticipate it, my regard for his skill will be all the greater, and my pleasurable experience of “sudden retrospective illumination” all the more intense and powerful. Thus, I’m employing my own “puzzle solving” prowess as a measure by which I judge the quality of the work. And this I would characterize far more as an act of “art appreciation” than of “games playing.”

Moreover, there are many readers who claim to read a detective story without trying to solve the mystery at all– they’re just there for the ride. How does that fit in with the games concept? Quite simply it doesn’t. Which brings us to another point about games:


Of course, many people do try to solve the mystery they are reading, and would rather arrive at the correct solution prior to being given it by the author. It is quite fair to say that these readers are treating the  detective story as a game– they are “playing” it as such (serving as their own rule makers and arbiters of success). But there’s a fundamental distinction: a detective story exists as an entertainment independent of its employment as a game– one can actively participate in its function as designed (i.e. one can read it and enjoy it) without anyone treating it as a game. This same is not so of entities designed solely or even primarily as games. Yes, one can enjoy baseball or chess as a spectator, but someone must be playing it as a game in order for anyone at all to enjoy it. Not so of detective fiction.

Also note that ultimately any entertainment– not just detective fiction– can be treated by an individual (or even a group) as a game. Even the aforementioned activity of “watching The Sound Of Music” can easily be turned into a drinking game (take a shot every time Gretl cries “Fräulein Maria!”). But this doesn’t mean that The Sound Of Music or the act of watching it is inherently a game. Admittedly, the puzzle provided by a detective story more readily invites its treatment by individuals as a game– that is, they make a game of it for themselves. But as with The Sound Of Music, The ABC Murders can be enjoyed as an entertainment without the reader choosing to treat it as a game. Thus, if we say call detective fiction a game– merely because it can be treated as such– it follows that we must say the same for all types of fiction, and indeed for all types of entertainment.

Speaking of comparison to to other entertainments, let’s make a comparison of the activity of reading a whodunit (say, Death on the Nile) with playing an actual game (we’ll use baseball, though the comparison would work with chess, backgammon, croquet, or any other real game) and with watching the film Citizen Kane:

FullSizeRenderI believe that side-by-side comparison makes it easy to recognize what type of activity detective fiction more closely resembles.

One further point (and it is indeed an important one): that element of “sudden retrospective illumination”– a key element of the detective fiction genre, and described by Carr and other genre experts as a euphoric, almost religious experience– is  found nowhere in games. One might be surprised by the outcome of a game, but games are not specifically designed to provide an ending that both surprises and seems retrospectively inevitable. It is however, found elsewhere in art, not only in detective fiction, but in other genres as well (e.g. the 1945 romance film Brief Encounter— anything but a murder mystery– concludes with a revisit to the first scene, with a new, more intense audience understanding of the meaning of the events).

And so, one further comparison:


An interesting case is that of Cluedo (or Clue, as it is known here in the States) which, much as the character in Chesterton’s The Man With Two Beards is described as the reverse of a ghost (“not the antic of the soul freed from the body. It was the antic of the body freed from the soul“), is in several respects the exact opposite of detective fiction: whereas a detective story is a fiction that in some respects resembles a game, CLUEDO is a game that resembles detective fiction. For, despite involving many of the stylistic trappings of the classic Golden Age Detective Story (the Victorian British setting, the stock character types, the genre-common instruments of death), it is indeed a true game which is played by employing strict deductive logic. Moreover its solution offers no sudden retrospective illumination. One might be surprised that Colonel Mustard committed the murder in the conservatory with a lead pipe, but there’s nothing in the game designed to make that scenario seem any less likely than any of the others. Conversely, there’s nothing (in the way of clueing) provided to make one feel, “Of course! I should’ve known! It was there before my eyes all the time!” At the same time, it does provide the true “fair play” which detective fiction cannot.

Finally, what is my point in “attacking” the idea that detective fiction is a game and the notion of detective fiction “fair play”? I assure it is not to upset the apple cart, nor is it to spoil the fun. And it is certainly not for the purpose of criticizing or belittling the genre. On the contrary my purpose is rather to glorify the genre… I come not to bury GA Detective Fiction, but to praise it. However, to call the detective story a game merely because some readers think of it as such is actually to do it a disservice. For, while one may admittedly use a shoe to drive a nail into a wall (indeed, I have), to then call a shoe a “hammer”–merely because it can be employed as such– is to call attention to all the ways in which it is inferior to those objects (real hammers) that were designed expressly for that purpose. Similarly, to call the detective story a game both highlights the many ways that detective stories fall short as games, yet overlooks the wonderful pleasures they offer that games cannot.

The Greek Coffin Mystery, by Ellery Queen (1932)

The Greek Coffin Mystery, by Ellery Queen (1932)


Ellery Queen is a fictional detective in the books by Ellery Queen … who is  a fictional writer.  The fictional writer whose name is on a set of novels from 1929 to 1971 was actually two people, cousins generally known as Frederic Dannay and Manfred Lee, whose joint Wikipedia entry is found here. As Wikipedia makes clear here, quite a few books ascribed to Ellery Queen were actually written by other authors; this one, however, is certainly the product of Dannay and Lee. Dannay also managed the affairs of Ellery Queen’s Mystery Magazine (the original version of this post erroneously suggested that it was Dannay and Lee in tandem), and the Ellery Queen name appears on the cover of many books of anthologized short stories reprinted from the magazine. Complicated, isn’t it? There’s also an old-time radio program, a series of vintage movies, a television series, comic books, a game or two, and even reference books about the character and the authors.

2633Publication Data:

This volume is the fourth Ellery Queen novel to be published by the cousins. The first nine books in the series each have a number of common features; there is a nationality in the title, here “Greek”; there is an introduction written by someone known only as “J.J. McC.”, now not considered canonical, and the famous “Challenge to the Reader”.  This challenge stops the action of the book and speaks directly to the reader, asserting that every piece of information necessary to solve the mystery is now in the reader’s hands. This is, in fact, the case; this volume is a strict-form puzzle mystery as I have elsewhere defined this term. One interesting conceit of this particular book is that each chapter has a single-word title; examination of the table of contents reveals that the initial letters of the chapter titles, considered acrostically, spell out “The Greek Coffin Mystery By Ellery Queen”.

The book was first published in 1932 by Frederick A. Stokes in the U.S. and a little later by Gollancz in the UK.  The first paperback edition is Pocket #179, seen at the head of this post. Many paperback editions exist; this book has only sporadically been out of print since its publication. It is now available in multiple e-book formats.

The Greek Coffin Mystery, 1960 - illus James Meese-1Although I have a VG copy of the first paper edition shown above, I actually used an e-book from an unknown source as my reference copy for this review (I found it in my files and have no idea where it came from, possibly as part of a gift of a bundle of e-books from a colleague); pagination is impossible to guarantee and I have chosen to not give page citations.

About this book:

Standard spoiler warning: What you are about to read may discuss in explicit terms the events of this murder mystery in GREAT detail. Please read no further if you wish to preserve your ignorance. You will also probably find here discussions of the content of other murder mysteries, perhaps by other authors, and a similar warning should apply.

IF YOU HAVE NOT READ THIS BOOK, STOP HERE AND GO READ IT BEFORE YOU RETURN. YOU WILL THANK ME. I can’t be any clearer — your first reading of this book should be unsullied by any knowledge of its contents, and the less you know in advance, the happier you will be. 

index-3_1The story begins with the death of wealthy Greek-American art dealer and connoisseur Gregor Khalkis; for once in a murder mystery, there’s nothing suspicious about the death. He’s been suffering from heart troubles for years that have left him blind and under the full-time care of a physician. It’s the disappearance of Khalkis’s will that is baffling everyone; five minutes before the funeral it was there, after the funeral it’s vanished. The house is searched, to no avail, and Mr. Woodruff, the family lawyer, calls in District Attorney Pepper. More searching, and no results. No secret passages or hidden compartments in the furniture or walls; no evidence that it was destroyed. Apparently the disappearance of the will is connected with its provisions, and someone’s desire to return to an earlier testamentary disposition of the Khalkis estate … but no one can figure out what happened. Finally Pepper calls in Ellery Queen, who deduces that the only possible location is inside the only object that’s left the house unsearched — Mr. Khalkis’s coffin. He convinces the authorities of the validity of his logic and they obtain permission to dig up the coffin. Unfortunately the coffin doesn’t contain the will. What it does contain is the strangled body of an ex-convict, a convicted forger named Grimshaw, jammed in on top of the late Mr. Khalkis. 

We soon meet the household and learn that Grimshaw had been admitted to a private interview with Khalkis shortly before their deaths. Khalkis has household staff (including the beautiful British secretary, Miss Brett, who might be romantically involved with Khalkis’s handsome young nephew Alan), relatives (including his mentally handicapped cousin Demmy, who acts as a kind of valet for the blind Mr. Khalkis) and the various employees of his art gallery and other business operations.

Ellery directs the activities of his father, Inspector Queen of the New York Police, with the assistance of DA Pepper, and a large group of officers immediately begin to learn everyone’s every movement. As is common in such fictional situations, it soon becomes apparent that most of the people in Khalkis’s life had recent acrimonious interactions with him, and many of them may well have had interactions with the deceased forger. Promptly upon the start of investigations, multi-millionaire Wall Street baron James Knox, friend of both the President and the late Mr. Khalkis, insists upon being briefed upon progress; Ellery announces that the case is solved. <gasp>

index-5_1A few chapters previously, the people around Ellery were baffled by his insistence on performing a number of experiments with the contents of a tea-urn in Khalkis’s office, and the surrounding used teacups, lemon, et cetera. He boils water, pours it out, measures amounts — no one understands what’s going on, and they think he’s losing his grip. As well, Ellery seems curiously interested in Mr. Khalkis’s neckties; he’d had some new ones delivered for the use of his handicapped cousin in executing his valeting duties. Ellery doesn’t explain until this point, when he reveals that, first of all, the details surrounding the neckties reveal that Mr. Khalkis has spontaneously regained his vision, and second, that two mysterious people who visited Khalkis in his study the night before his death were not actually two people, and that Khalkis had gone through an incredible rigamarole to make it seem as though two other people had been there. This idea, Ellery reveals, is the result of his analysis of tea-cups and tea water. And therefore — Khalkis murdered Grimshaw.

Immediately upon this revelation — about halfway through the book — two things happen. One is that Miss Brett reveals that, oopsie, she forgot to mention that the used teacups were differently arranged than when they were found by Ellery, and Knox reveals that there was indeed a third man in that meeting with Khalkis and Grimshaw.  How does he know?  Knox was the third man.

At this halfway point in the novel, Ellery’s house of logical cards collapses and he sinks into depression; this event actually affects the remainder of his career and all subsequent books that feature him. He determines that because he has revealed the results of his analysis and been disproven, he will never again speak about his investigations until he is absolutely, completely certain of the identity of the murderer (rather like Saul’s conversion on the road to Tarsus). Although it’s not referred to specifically in later volumes, his detective career is forever changed by this event; it also changes the way in which his work is presented. When you think about it, it’s not sensible for a detective to hide the progress of his investigations from the police; this situation was apparently set up by the authors to create a structure for future novels that would delay the solution until the end of the book.

Knox now starts the second half of the plot in motion.  He had been dickering with Khalkis for the right to purchase a Da Vinci painting that had previously been thought to have been destroyed. But Grimshaw had become involved by going to Knox, announcing that he had stolen the Da Vinci some years ago for Khalkis, and Khalkis had apparently been unable to pay him for his labours. Finally Khalkis had agreed to make out his will in favour of Grimshaw and in the interim gave him a promissory note. Khalkis, Grimshaw and Knox had all met and drunk tea on that fateful evening, and then some unknown person had tampered with the physical evidence in order to lead Ellery away from the truth. Ellery soon determines that that unknown person must logically have been in partnership with Grimshaw.

Knox refuses to hand over the Da Vinci and announces that he’ll deny having it in his possession — and that it’s a copy anyway. Ellery then realizes that his deduction of Khalkis having recovered his sight was also incorrect; instead, handicapped Demmy is revealed to be colour-blind. Ellery grimly acknowledges his mistakes and gets back to work on solving the case.

Events now progress more rapidly.  The investigation receives an anonymous tip that the manager of Khalkis’s art gallery, Gilbert Sloane, is actually Grimshaw’s brother. The police discover that an empty house in Khalkis’s neighbourhood was the temporary resting place of Grimshaw’s corpse (until the murderer had the bright idea of disposing of it in the coffin) and they discover a shred of the burned will in a furnace in the empty house, confirming that the missing will indeed left the huge Khalkis estate to Grimshaw. This means that Sloane will actually inherit through his brother; they find a key to the empty house concealed in the Sloane home. Everyone rushes to the Khalkis Gallery to arrest Sloane — and he’s been shot. Superficially it looks like suicide, but Ellery makes a deduction that proves it to be murder. And everything grinds to a halt, because Ellery cannot find a thread of the tapestry upon which to pull in order to make progress with the case.

index-221_1Miss Brent reveals herself to have been an agent of the British Museum, employed to track down the Da Vinci; she’s hired by Knox to help him with his executor’s duties on the Khalkis estate. And the British Museum is about to pull the lid off the case unless Ellery solves it in a hurry.  Soon, the missing promissory note shows up — half of it is used as the paper upon which a blackmail note is typed. The actual typing of this note is of interest; there’s a tiny typographical error that is shown to the reader but not further explained.

At about this point, the above-mentioned “Challenge to the Reader” breaks the flow of the action; you now have in your possession enough information to solve the mystery and identify Grimshaw’s partner and the murderer.  I will from this point on be reticent about what happens; I haven’t yet told you anything that would make any difference to your ability to solve the murder, since if you read the book everything will be available to you.  But henceforth, I will cut back drastically on my comments for fear of spoiling things for you.

It is safe to say, though, that there is a common theme in nearly all Ellery Queen stories that is repeated here; the false solution, then the true. At this point, Ellery makes an announcement about who is guilty of precisely what; this leads to a series of events that brings us to the final solution. Ellery has set a trap for the real killer, and I wager that you will be very, very surprised by the answer, which is revealed dramatically with Ellery being shot in the shoulder and the murderer dying in a hail of gunfire at the end of Chapter 33. Chapter 34 consists of Ellery recuperating from his wound and explaining everything, in great detail, to an assembly of suspects and investigators.

04b_GreekWhy is this book worth your time?

The year of publication of this book is 1932.  In 1932, Agatha Christie had published a mere dozen novels, but including one of the most difficult mysteries ever written (The Murder of Roger Ackroyd). Ngaio Marsh was two years away from her first book; Margery Allingham was at the beginning of her career; John Dickson Carr had not yet published a Gideon Fell or a Henry Merrivale novel; Anthony Berkeley had published a number of excellent books including 1929’s The Poisoned Chocolates Case — and the “Golden Age” mystery was in its early stages. It was not completely newborn; perhaps adolescent; still finding its way, outlining the ideas that define the form, the boundaries of the genre, its passions, its likes and dislikes, its enthusiasms and hatreds. S.S. Van Dine and Ronald Knox had both published sets of rules as to what detective stories should and should not be; clever writers like “Ellery Queen” were casting off the old strictures and extending the boundaries of the form.

This particular story has to be one of the most difficult strict-form puzzle mysteries ever written and, frankly, they don’t make ’em like this any more. This book has more sheer logic and detection in it by the halfway point than in the entire oeuvre of your average cozy author; and by the end of the novel, more difficult chains of logic than the entire oeuvre of ten cozy writers. This book was written at a time when readers did not cavil at being faced with an extremely difficult puzzle and it has, over the years, maintained its place as one of the finest examples of such a puzzle. I haven’t worked out the ramifications of this in great detail, but I’ll suggest that this is one of Queen’s top two books — the other being The Chinese Orange Mystery — and one of the top 25 puzzle mysteries ever written. Just don’t make me name the other 23, please!

When I’m analyzing a puzzle mystery, there’s a process I go through that is crucial to determining its level of quality. Simply put, once I know whodunnit, I go through the novel again from the murderer’s point of view and see if everything makes sense. And I think you would be surprised at how often things just do not make sense when I do that. For instance, I recently looked at a poorly-written mystery by Frances Crane, The Applegreen Cat. (My analysis is here.) Among other problems, the plot consisted of a mystery that was difficult from the point of view of the reader — but ridiculous from the point of view of the murderer, who apparently deliberately waited until the country house was filled with house guests before embarking upon a killing spree among the servants. Another example is an early novel of Harlan Coben’s whose name slips my mind along with most of the details. Three-quarters of the way through the book, the protagonist discovers that the murderer has a cabin  in the woods filled with evidence, and this provides everything needed to bring the book to a close. The problem is, as I realized even before reaching the end of the novel, no murderer in his right mind would have left all that tasty evidence in place, sitting in an empty cabin for anyone who happened by. It’s rather like one of those plots where the murderer has the detective at his mercy, but stops to deliver a complete detailed confession before disposing of his nemesis. It helps out the book a lot, but lowers the murderer’s IQ by 50 points in an instant.

If you go through the process of analyzing things from the murderer’s point of view, everything in this book continues to make perfect sense. The murderer’s motives are clear; they make sense and continue to make sense once you know what they are. The only thing that trips up the killer is a trap set by the detectives that is also based on something that the murderer needs to see happen. The tiny clues left by the murderer are tiny accidents; they aren’t taunts left by the killer, or foolish oversights, but something small and careless like closing a door when it shouldn’t have been closed, or not predicting that a character may confess something that is not in his best interests in order to cooperate with the police. And there are not many puzzle mysteries about which this can be said. Nothing depends on coincidence, chance, acts of God or ridiculous motivation. Just about the only logical flaw in the entire novel is the size of the fragment of the will that is found in the furnace of the empty house, and the fact that it contains precisely the information that is needed to move forward; this is a bit of a stretch, but, you know, it could happen. All the clues you need are fairly there, and the Challenge to the Reader is accurate.

The other part of this book that is beautifully crafted is the false trail that the reader is meant to follow. I read this book as a teenager and I remember the sense of exultation with which I came to the conclusion that the authors wished me to reach; I’d spotted the tiny clues, I’d noticed the snippets of dialogue, and I’d realized what they meant. I felt smart. By golly, this mystery business wasn’t so hard after all, I thought. And then I realized that I’d been well and truly fooled, and that was what the authors had meant to happen. Up until that point, I’d merely failed to solve the mystery, or I’d guessed sort of randomly at a possible solution. This time I’d tried to solve the mystery, and I’d been fooled. And it may well be this book that started me on a lifetime of challenging my wits against those of the author.

In short — this is one of the finest strict-form puzzle mysteries that you will ever have the pleasure of failing to solve. In the past, for the benefit of a friend who hasn’t yet had the pleasure of encountering this mystery, I’ve taken a cheap paperback and torn it in half at the point at which the Challenge to the Reader appears, in order to give my friend the chance to give this mystery the attention it deserves without the opportunity to spoil it by peeking. There are not many mysteries worth doing that with. If you enjoy the experience, and you see a cheap paperback copy go by, pay it forward for a friend.

Notes for the Collector:

As of this writing, AbeBooks has on offer a Good copy of the first edition, inscribed by Frederick Dannay to his sister-in-law, for $500, and two unsigned copies of the first for $236 and $250. The second edition will set you back $175, and a copy of the first UK from Gollancz is listed for about $60. I am aware of an interesting edition from International Readers League in 1933, with a street map and floor plan of the Khalkis house (like the ones reproduced here, which are also in the first paper edition), and Abe has a copy for $75.

Some crazy person on ViaLibri wants $500 for the Bestseller Mystery/Mercury edition of 1941, and I can only think that it has about $490 in cash tucked between the pages. Amereon reprinted this title in 2001 and I can’t think why this particular book is bringing prices in the $75 range for an undistinguished hardcover with no jacket.

In paper, the 1942 first paper edition from Pocket is quite collectible because it’s a low-numbered book in that pioneering series, collected by many, even though, as you can see from the illustration at the top of this post, the cover art is downright unattractive — muddy, unexciting and dull. (When you look at the gaudy but exciting cover of The French Powder Mystery from the same company at about the same time, you wonder if the publishers were trying to make the Greek Coffin look boring!) Mine is a relatively nice copy and what appears to be a similar one on Abe is listed for $20; I’ve seen many copies of this book and many of them appear to have vertical creases in the cover, rolling, etc. There is a Penguin greenback available, of which there are many collectors, and many other editions.

1808330There’s a Cardinal edition that has a great piece of “girlie leg art” on the cover and, for once, it actually depicts a scene from the book. One quirky favourite edition of mine has always been a uniform set of Signet paperbacks from the early 70s with a tightly-kerned Helvetica title and cover art of a pretty model posed within a box, holding an oversized prop that has something to do with the plot.  Possibly this has something to do with the fact that in many cases this was the first edition that passed through my hands; at this remove, they look quite camp. Your mileage may vary. The point is that, depending on what your budget and collector’s instincts might be, there’s something for you. My own recommendation would be the signed first, which is quite scarce with any signature, and for smaller budgets the best copy you can afford of the Pocket edition, unless you like “girlie leg art” in which case the Cardinal edition may suit you best.

2014 Vintage Mystery Bingo:

This 1932 volume qualifies as a Golden Age mystery; third under “D”, “Read a book already read by another challenger.” This volume was reviewed on February 17, 2014 at a blog called “Classic Mysteries”; the review is found here. For a chart outlining my progress, see below.

Vintage Golden Card 001