#42: Defining ‘Fair Play’ in detective fiction

book stack

Partly because I’m feeling brave, and partly because it’s something that has niggled at the back of my mind given some recent reading experiences, I wanted to have a bit of a discussion – or, well, monologue – about what constitutes a fair play detective novel.  I shall attempt not to go the Full Knox, and I’m not imagining that anything I say will be remotely original, but it’s on my mind and it’s my understanding that this is how blogging works.  So…

I think I’m relatively safe in saying that for many people the appeal of the detective story is the opportunity to have a go at fitting the puzzle together before the author explains all at the end (differentiating here from the crime novel or the thriller, which for brevity’s sake we’ll simply say have different intentions).  It’s also quite secure ground to say that finding out you never had a chance of solving the mystery for whatever reason – insufficient clueing, frank lies on behalf of the author, etc – adversely affects one’s enjoyment of such books.  Having all the information there for you to colligate and deduce is what motivates a lot of said reading, but the precise nature of what constitutes a ‘fair play’ novel is somewhat hard to rigorously define.  I’ll suggest up front that my own conclusions probably won’t tally with your own, but that’s all part of the fun, what?

It becomes quite murky quite quickly, so I apologise if this gets a touch cluttered.  Arguably, the purpose of a detective story – if you don’t mind, we’ll assume I mean fair play ones from now on to save me typing it repeatedly – is to provide a crime, the clues necessary to solve it, and then a chain of reasoning that leads to a guilty party.  My first contention is that it doesn’t necessarily have to lead definitely to a single party; part of the joy of such stories is that there must be some misdirection, some suspicion sprinkled about and arrow-shaped smelly fish pointing in potentially two directions – to both the actual guilty party and the patsy intended by the author to bear the brunt of our readerly speculation.  Inevitably it comes down to interpretation.  We the reader are intended to interpret a clue one way while the detective interprets it in another; enough interpretations are collated to provide a weight of evidence, and that series of connections the interpretations combine to make it seems sensible and largely unarguable that the proffered guilty party at the end of the story is indeed the murder/blackmailer/highwayman.  I can name at least three novels off the top of my head where the suggested guilty party is still protesting their innocence come the close of the book, suggesting an element of interpretation still available within the narrative, but each of these is arguably a fairly-played novel of detection.

Equally, there must be sufficient declared clues and no deliberate narrative chicanery on behalf of the author in withholding something without appearing not to.  Countless classic-era detective novels have their detective protagonist making a phone call or receiving a letter the contents of which are not revealed to the reader until the fatal summing-up.  Immediately these are discounted from consideration.  However, there are books which appear to be abiding by the spirit of the game but are in fact playing rather close to the edges of such cheating.  As an example – and hopefully you’ll excuse my repeated deliberate vagueness, the peril of this kind of undertaking is trying to be specific while also not spoiling anything for a potential audience whose reading to date I am not privy to – I have read three books in the last couple of years where a crime scene map is given as a way of bringing the reader up to parity, but some element of that map, a crucial requirement given the commission of the crime, is either excluded or obfuscated.  In one case the presence of a window makes all the difference, but the windows on the map are indicated only by the line used to draw in the walls thinning out by about three microns (note: actual thickness may be slightly greater) which is not going to be apparent until you really go back and know you’re looking for it.  Technically it has been declared, but in reality I’d argue this isn’t fair.  To be fully fair, the window should be called a window, should be obviously a window at first glance, and if this means your crime is solvable on page 43 then you need a better crime.

Similarly, burying the key information in a higgledy-piggledy mess of deliberately confusing cross-talk does not, to my mind, make it fairly declared.  We might as well stop beating around the bush here and call someone out: this is in part motivated by the 1950 novel Smallbone Deceased by Michael Gilbert which I had intended to review this coming Wednesday.  Look it up on any book review site you like and it is frequently called a masterpiece, a classic, a wonderful piece of detective fiction, and, crucially, scrupulously fair in its clues.  I am aware what I’m taking on here, pretty much the entire swathe of public opinion is against me (and do bear in mind that I didn’t finish it).  The opening 50-ish pages – which is about as far as I got – is deliberately awash with reams of this kind of writing: relaying conversations of several people in a room at once without always saying who is speaking, referring to characters first by a given name, then by a nickname, then by a titled name, switching suddenly from light-hearted banter to dry drudgery…there’s so much going on and it’s such tremendous hard work that I honestly couldn’t even begin to keep it all straight (or, indeed, care about any of it).   A friend who has read the book – and loved it, incidentally – tells me that there are several key clues dropped in during this part of the book, and I’m sorry but it doesn’t count as fair play to just mix in one or two spare comments amidst pages of white noise (see also, Not to be Taken by Anthony Berkeley).  Don’t care, I’m not having it.

Additionally then, specialist knowledge.  Ulf Durling’s Hard Cheese which I reviewed last week is a key example and another motivator for this rant post; while you do arguably have the information needed to solve the crime, you also need a fair amount of specialist esoteric knowledge of at least two subjects to be able to do so.  Now, the novel may just provide enough of the information required for you to go and do the research, but the fact that you need to do research in my mind stops it qualifying (even though, technically, its fair playness could be argued on the grounds of the information being there in the first place).  This is the exact some reason I never understood the popularity of the TV show House – most of the audience never stood even the vaguest chance of working it out! – but that’s another issue and another post for another time (not by me though, I’m sure it’s already been done).  Relying on general ignorance or urban myths is fine, so long as they’re corrected and not used as a lazy get-out – a body found outside the Empire State Building having been killed by a coin dropped off the top (a scientific impossibility) would not be okay in any way, shape or form – but ignorance of the facts given a prevailing false assumption in general perception is perfectly acceptable.  Indeed, it is to be encouraged!

A pedantic but necessary corollary of the above paragraph should also include nonsense invention – no poisons previously unknown to science, not-of-this-world unexplainable influences, or just plain old invention for the sake of surprise.  John Dickson Carr – well, I had to mention him at some point – could be accused of a lack of likelihood in many of his scheme, but they’d still be possible if certain circumstances were to collude (even though one of my favourites of his pushed this rule a bit far, it’s not like he has a cat trained as a ninja or is reliant on time travel or anything in that sense of the impossible).  We may also exclude communing with ghosts, reading of tea leaves, etc, etc.  Including these as part of your minatory milieu is naturally fine, but having them or their implied influences provide a key part of the solution or workings makes your plot no longer that in the detective subgroup (grrr, closed-minded simpleton that I am, etc, etc).

So, where does that leave us?  Put simply, if I get to the reveal of a detective novel and can see how each crumb along the path of reasoning to the solution was given to me to pick up and examine at my leisure then it’s a fair play puzzle.  Said crumbs should combine to make only one loaf of bread (er, I might drop this analogy now) – which is to say, there should be no major contentions with the final solution reached and it should be as tightly-constructed as possible (I maintain that my solution to Agatha Christie’s Sparkling Cyanide is far more elegant than her own; it even removes one character from the novel and places suspicion in a variety of different places…sigh, I really must write that book one day).  Some will do this better than others, of course – Edmund Crispin and Freeman Wills Crofts have both gone as far as to provide page numbers of the relevant clues and information, as I’m sure have others –  but it’s possible to still have some doubts in your final workings and the book be fair play.  However, the information must be there.  No sudden intuition, no undeclared postcards or telephone calls, no esoterica not discussed within your pages.  It sounds easy, but an alarming number of novels fail to do these things (and at least 80% of you reading this already disagree with something I’ve said, anyway).  I like to imagine there’s some detective fiction Xanadu where all books do this and everyone is happy, but I think we’re more likely to continue to be disappointed when authors cheat, and so start a blog in which we implicitly insist that we’d do it better if only we didn’t have those damn forms to fill in or that lightbulb to change.  And – ah! – my back!

Sincere thanks to those of you who successfully negotiated these 1800 words of not very much.  Your thoughts, as ever, are encouraged and appreciated…

9 thoughts on “#42: Defining ‘Fair Play’ in detective fiction

  1. At the Kyoto University Mystery Club (and some other clubs), there’s been a decades-old tradition of “whodunnit” stories; short stories featuring a challenge to the reader. First half of the session, participants are given the first part of the story (up to the challenge). When they think they know who the culprit is, then can go to the writer of the story and explain their deductions, based on the clues in the story. In the second half, everyone is given the answer-half of the story. These short stories (which often use the Queen-like elimination method of determining the culprit) are essentially intellectual puzzles, so they are often a bit lacking in ‘literary value’, but are incredibly fun.

    Because the stories are deliberately designed to be solved, in a group, simultaneously with other people, the most important aspect of any whodunnit story is that it’s solvable, based on the clues, in a realistic manner. Anyone can write an unsolvable story; it’s writing a solvable story, that is also entertaining for the reader, that is difficult. There are of course some unwritten rules, like there’s always one culprit, there exists nothing outside the story-world, all characters act intellectually and nobody except for the culprit lies, etc. At the end of each session, there’s always a discussion about the story itself, what was good, what was bad, was it fair, etc. It’s this process that has cultivated many Japanese mystery writers, as they developed within this culture of emphasizing fair-play, while exploring the limits of the (unwritten) rules. Note that Van Dine/Knox’s rules are a lot more limiting.

    Because of that, I’ve actually become very fond of mystery stories that do dare to use ‘nonsense inventions’, ghosts and such. The concept of a ghost is not unfair on its own. It’s how you use it. During my year at the Kyoto University Mystery Club, I remember we did a whodunnit session which featured a fantastic story with only ghosts with various powers appearing (and the narrator was dead and a ghost too) . Yet it was completely fair.

    Liked by 2 people

    • I read about the KUMC in your afterword to The Decagon House Murders, Ho-Ling, and it sounds absolutelt brilliant.

      There is absolutely scope for ghosts, etc., if the rules are established appropriately, you’re quite right. I suppose I was being a bit sweeping in regarding ‘typical’ detective fiction – Christie, Carr, Sayers, etc. relying on such a thing would leap two-footed significantly beyond the scope of the fair play universe. But properly done and in the correct context – and I’m hugely curious now as to how this story did it – it could work.

      This reminds me of an Isaac Asimov story in which the narrator is turned into a demon and has to escape from somewhere using his demonic powers…technically it is fair-play, but the inclusion of this story in collections alongside more traditional authors has raised more than a few eyebrows over the years…


  2. I have always enjoyed Gilbert’s books but it has been too long since I read SMALLBONE to comment either way, though I suspect that the lack of potential fair play would bother me too much. I know what you mean JJ< but it is a question f degree for me – In the past I have definitely finished some books and wished that I had had fewer clues which I interpreted correctly before the ending because actually I would have preferred the surprise!

    Liked by 1 person

    • Oh, completely agreed; the degree of fairness is another factor entirely, not least because of how much each individual reader spots and connects. I’m tempted to write something on the subject that completely spoils a particular book (nothing in mind, but would just pick one to illustrate the points), under the understanding that people must definitely have read it before reading that post. Hmmm, will mull it over and see what occurs to me.


  3. Pingback: A Mexican Delight in Todd Downing’s Murder on Tour (1933) | crossexaminingcrime

  4. Ok, will try to reconstruct my earlier post that got swallowed up…

    Just to make sure that I’ve heard you rightly:

    1. The puzzle should be made up of pieces given to the reader for reconstruction.
    2. There should be no ‘narrative chicanery’: vital information withheld from the reader.
    3. Vital information should not be obscured through an excess of pointless detail or convoluted narrative.
    4. No specialist information please.
    5. And no fantastical inventions that circumvent the reader’s assumption of reality.

    I think I agree with all these principles. I suspect I may be less stringent when it comes to (3), insofar as a detective sometimes has to deal with a messy crime scene, and handle a barrage of sensory experiences/ observations as part of the detecting process. But I would have a problem if clues are hidden from the reader simply by way of a challenging narrative, rather than by way of a clever puzzle.

    Narrative chicanery is an interesting criterion. Perhaps different readers might have different levels of threshold for what is acceptable? I recall one or two Ellery Queen novels where Ellery receives a telegraph or a verbal confirmation that the reader is not privy to (Dutch Shoe Mystery?) – but I believe the information merely affirms a prior deduction made on the basis of shared information with the reader. Regarding ‘Decagon House Murders’ I thought the twist was devious, but not in fair. One of Ayatsuju Yukito’s other novels – ‘Maze House Murders’ – contained a twist that I thought was a chicanery beyond the reader to reasonably discern.

    Liked by 1 person

    • I really appreciate the effort rewriting this must have taken, so thank-you for putting the time in, and you raise a very interesting challenge to point 3. I suppose the confusing nature of a fictional crime investigation will bring up all manner of confusing aspects, that’s the fun of reading these stories, but you’re correct in what I’m referring to is a deliberately confusing narrative.

      Joel Townsley Rogers’ The Red Right Hand is a perfect terrible example of this: the story is told is a hopelessly disjointed manner purely to confuse you further. When you finally get to the end and take the time to put it together in a linear way, it’s a stupidly simple puzzle that (in my opinion) is told in that way to be deliberately impossible to follow and so provide surprises. The information is arguably given to you to solve it, but in such a way as to be preposterously beyond most people to spot in anyway. That’s the kind of thing I have in mind.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.