“20 Rules For Writing Detective Stories” By S.S. Van Dine, One of T.S. Eliot’s Favorite Genre Authors (1928)

Every gen­er­a­tion, it seems, has its pre­ferred best­selling genre fic­tion. We’ve had fan­ta­sy and, at least in very recent his­to­ry, vam­pire romance keep­ing us read­ing. The fifties and six­ties had their west­erns and sci-fi. And in the for­ties, it won’t sur­prise you to hear, detec­tive fic­tion was all the rage. So much so that—like many an irri­ta­ble con­trar­i­an crit­ic today—esteemed lit­er­ary tastemak­er Edmund Wil­son penned a cranky New York­er piece in 1944 declaim­ing its pop­u­lar­i­ty, writ­ing “at the age of twelve… I was out­grow­ing that form of lit­er­a­ture”; the form, that is, per­fect­ed by Edgar Allan Poe, Arthur Conan Doyle, and Wilkie Collins, and imi­tat­ed by a host of pulp writ­ers in Wilson’s day. Detec­tive sto­ries, in fact, were in vogue for the first few decades of the 20th century—since the appear­ance of Sher­lock Holmes and a deriv­a­tive 1907 char­ac­ter called “the Think­ing Machine,” respon­si­ble, it seems, for Wilson’s loss of inter­est.

Thus, when Wil­son learned that “of all peo­ple,”Paul Grim­stad writes, T.S. Eliot “was a devot­ed fan of the genre,” he must have been par­tic­u­lar­ly dis­mayed, as he con­sid­ered Eliot “an unim­peach­able author­i­ty in mat­ters of lit­er­ary judg­ment.” Eliot’s tastes were much more ecu­meni­cal than most crit­ics sup­posed, his “atti­tude toward pop­u­lar art forms… more capa­cious and ambiva­lent than he’s often giv­en cred­it for.” The rhythms of rag­time per­vade his ear­ly poet­ry, and “in his lat­er years he want­ed noth­ing more than to have a hit on Broad­way.” (He suc­ceed­ed, six­teen years after his death.) Eliot pep­pered his con­ver­sa­tion and poet­ry with quo­ta­tions from Arthur Conan Doyle and wrote sev­er­al glow­ing reviews of detec­tive nov­els by writ­ers like Dorothy Say­ers and Agatha Christie dur­ing the genre’s “Gold­en Age,” pub­lish­ing them anony­mous­ly in his lit­er­ary jour­nal The Cri­te­ri­on in 1927.

One nov­el that impressed him above all oth­ers is titled The Ben­son Mur­der Case by an Amer­i­can writer named S.S. Van Dine, pen name of an art crit­ic and edi­tor named Willard Hunt­ing­ton Wright. Refer­ring to an emi­nent art his­to­ri­an—whose tastes guid­ed those of the wealthy indus­tri­al class—Eliot wrote that Van Dine used “meth­ods sim­i­lar to those which Bernard Beren­son applies to paint­ings.” He had good rea­son to ascribe to Van Dine a cura­to­r­i­al sen­si­bil­i­ty. After a ner­vous break­down, the writer “spent two years in bed read­ing more than two thou­sand detec­tive sto­ries, dur­ing with time he method­i­cal­ly dis­tilled the genre’s for­mu­las and began writ­ing nov­els.” The year after Eliot’s appre­cia­tive review, Van Dine pub­lished his own set of cri­te­ria for detec­tive fic­tion in a 1928 issue of The Amer­i­can Mag­a­zine. You can read his “Twen­ty Rules for Writ­ing Detec­tive Sto­ries” below. They include such pro­scrip­tions as “There must be no love inter­est” and “The detec­tive him­self, or one of the offi­cial inves­ti­ga­tors, should nev­er turn out to be the cul­prit.”

Rules, of course, are made to be bro­ken (just ask G.K. Chester­ton), pro­vid­ed one is clever and expe­ri­enced enough to cir­cum­vent or dis­re­gard them. But the novice detec­tive or mys­tery writer could cer­tain­ly do worse than take the advice below from one of T.S. Eliot’s favorite detec­tive writ­ers. We’d also urge you to see Ray­mond Chan­dler’s 10 Com­mand­ments for Writ­ing Detec­tive Fic­tion.

THE DETECTIVE sto­ry is a kind of intel­lec­tu­al game. It is more — it is a sport­ing event. And for the writ­ing of detec­tive sto­ries there are very def­i­nite laws — unwrit­ten, per­haps, but none the less bind­ing; and every respectable and self-respect­ing con­coc­ter of lit­er­ary mys­ter­ies lives up to them. Here­with, then, is a sort Cre­do, based part­ly on the prac­tice of all the great writ­ers of detec­tive sto­ries, and part­ly on the prompt­ings of the hon­est author’s inner con­science. To wit:

1. The read­er must have equal oppor­tu­ni­ty with the detec­tive for solv­ing the mys­tery. All clues must be plain­ly stat­ed and described.

2. No will­ful tricks or decep­tions may be placed on the read­er oth­er than those played legit­i­mate­ly by the crim­i­nal on the detec­tive him­self.

3. There must be no love inter­est. The busi­ness in hand is to bring a crim­i­nal to the bar of jus­tice, not to bring a lovelorn cou­ple to the hyme­neal altar.

4. The detec­tive him­self, or one of the offi­cial inves­ti­ga­tors, should nev­er turn out to be the cul­prit. This is bald trick­ery, on a par with offer­ing some one a bright pen­ny for a five-dol­lar gold piece. It’s false pre­tens­es.

5. The cul­prit must be deter­mined by log­i­cal deduc­tions — not by acci­dent or coin­ci­dence or unmo­ti­vat­ed con­fes­sion. To solve a crim­i­nal prob­lem in this lat­ter fash­ion is like send­ing the read­er on a delib­er­ate wild-goose chase, and then telling him, after he has failed, that you had the object of his search up your sleeve all the time. Such an author is no bet­ter than a prac­ti­cal jok­er.

6. The detec­tive nov­el must have a detec­tive in it; and a detec­tive is not a detec­tive unless he detects. His func­tion is to gath­er clues that will even­tu­al­ly lead to the per­son who did the dirty work in the first chap­ter; and if the detec­tive does not reach his con­clu­sions through an analy­sis of those clues, he has no more solved his prob­lem than the school­boy who gets his answer out of the back of the arith­metic.

7. There sim­ply must be a corpse in a detec­tive nov­el, and the dead­er the corpse the bet­ter. No less­er crime than mur­der will suf­fice. Three hun­dred pages is far too much pother for a crime oth­er than mur­der. After all, the read­er’s trou­ble and expen­di­ture of ener­gy must be reward­ed.

8. The prob­lem of the crime must he solved by strict­ly nat­u­ral­is­tic means. Such meth­ods for learn­ing the truth as slate-writ­ing, oui­ja-boards, mind-read­ing, spir­i­tu­al­is­tic se’ances, crys­tal-gaz­ing, and the like, are taboo. A read­er has a chance when match­ing his wits with a ratio­nal­is­tic detec­tive, but if he must com­pete with the world of spir­its and go chas­ing about the fourth dimen­sion of meta­physics, he is defeat­ed ab ini­tio.

9. There must be but one detec­tive — that is, but one pro­tag­o­nist of deduc­tion — one deus ex machi­na. To bring the minds of three or four, or some­times a gang of detec­tives to bear on a prob­lem, is not only to dis­perse the inter­est and break the direct thread of log­ic, but to take an unfair advan­tage of the read­er. If there is more than one detec­tive the read­er does­n’t know who his cod­e­duc­tor is. It’s like mak­ing the read­er run a race with a relay team.

10. The cul­prit must turn out to be a per­son who has played a more or less promi­nent part in the sto­ry — that is, a per­son with whom the read­er is famil­iar and in whom he takes an inter­est.

11. A ser­vant must not be cho­sen by the author as the cul­prit. This is beg­ging a noble ques­tion. It is a too easy solu­tion. The cul­prit must be a decid­ed­ly worth-while per­son — one that would­n’t ordi­nar­i­ly come under sus­pi­cion.

12. There must be but one cul­prit, no mat­ter how many mur­ders are com­mit­ted. The cul­prit may, of course, have a minor helper or co-plot­ter; but the entire onus must rest on one pair of shoul­ders: the entire indig­na­tion of the read­er must be per­mit­ted to con­cen­trate on a sin­gle black nature.

13. Secret soci­eties, camor­ras, mafias, et al., have no place in a detec­tive sto­ry. A fas­ci­nat­ing and tru­ly beau­ti­ful mur­der is irre­me­di­a­bly spoiled by any such whole­sale cul­pa­bil­i­ty. To be sure, the mur­der­er in a detec­tive nov­el should be giv­en a sport­ing chance; but it is going too far to grant him a secret soci­ety to fall back on. No high-class, self-respect­ing mur­der­er would want such odds.

14. The method of mur­der, and the means of detect­ing it, must be be ratio­nal and sci­en­tif­ic. That is to say, pseu­do-sci­ence and pure­ly imag­i­na­tive and spec­u­la­tive devices are not to be tol­er­at­ed in the roman polici­er. Once an author soars into the realm of fan­ta­sy, in the Jules Verne man­ner, he is out­side the bounds of detec­tive fic­tion, cavort­ing in the unchart­ed reach­es of adven­ture.

15. The truth of the prob­lem must at all times be appar­ent — pro­vid­ed the read­er is shrewd enough to see it. By this I mean that if the read­er, after learn­ing the expla­na­tion for the crime, should reread the book, he would see that the solu­tion had, in a sense, been star­ing him in the face-that all the clues real­ly point­ed to the cul­prit — and that, if he had been as clever as the detec­tive, he could have solved the mys­tery him­self with­out going on to the final chap­ter. That the clever read­er does often thus solve the prob­lem goes with­out say­ing.

16. A detec­tive nov­el should con­tain no long descrip­tive pas­sages, no lit­er­ary dal­ly­ing with side-issues, no sub­tly worked-out char­ac­ter analy­ses, no “atmos­pher­ic” pre­oc­cu­pa­tions. such mat­ters have no vital place in a record of crime and deduc­tion. They hold up the action and intro­duce issues irrel­e­vant to the main pur­pose, which is to state a prob­lem, ana­lyze it, and bring it to a suc­cess­ful con­clu­sion. To be sure, there must be a suf­fi­cient descrip­tive­ness and char­ac­ter delin­eation to give the nov­el verisimil­i­tude.

17. A pro­fes­sion­al crim­i­nal must nev­er be shoul­dered with the guilt of a crime in a detec­tive sto­ry. Crimes by house­break­ers and ban­dits are the province of the police depart­ments — not of authors and bril­liant ama­teur detec­tives. A real­ly fas­ci­nat­ing crime is one com­mit­ted by a pil­lar of a church, or a spin­ster not­ed for her char­i­ties.

18. A crime in a detec­tive sto­ry must nev­er turn out to be an acci­dent or a sui­cide. To end an odyssey of sleuthing with such an anti-cli­max is to hood­wink the trust­ing and kind-heart­ed read­er.

19. The motives for all crimes in detec­tive sto­ries should be per­son­al. Inter­na­tion­al plot­tings and war pol­i­tics belong in a dif­fer­ent cat­e­go­ry of fic­tion — in secret-ser­vice tales, for instance. But a mur­der sto­ry must be kept gemütlich, so to speak. It must reflect the read­er’s every­day expe­ri­ences, and give him a cer­tain out­let for his own repressed desires and emo­tions.

20. And (to give my Cre­do an even score of items) I here­with list a few of the devices which no self-respect­ing detec­tive sto­ry writer will now avail him­self of. They have been employed too often, and are famil­iar to all true lovers of lit­er­ary crime. To use them is a con­fes­sion of the author’s inep­ti­tude and lack of orig­i­nal­i­ty. (a) Deter­min­ing the iden­ti­ty of the cul­prit by com­par­ing the butt of a cig­a­rette left at the scene of the crime with the brand smoked by a sus­pect. (b) The bogus spir­i­tu­al­is­tic se’ance to fright­en the cul­prit into giv­ing him­self away. © Forged fin­ger­prints. (d) The dum­my-fig­ure ali­bi. (e) The dog that does not bark and there­by reveals the fact that the intrud­er is famil­iar. (f)The final pin­ning of the crime on a twin, or a rel­a­tive who looks exact­ly like the sus­pect­ed, but inno­cent, per­son. (g) The hypo­der­mic syringe and the knock­out drops. (h) The com­mis­sion of the mur­der in a locked room after the police have actu­al­ly bro­ken in. (i) The word asso­ci­a­tion test for guilt. (j) The cipher, or code let­ter, which is even­tu­al­ly unrav­eled by the sleuth.

You can find S.S. Van Dine’s detec­tive nov­els on Ama­zon.

Relat­ed Con­tent:

Ray­mond Chandler’s Ten Com­mand­ments for Writ­ing a Detec­tive Nov­el

H.P. Love­craft Gives Five Tips for Writ­ing a Hor­ror Sto­ry, or Any Piece of “Weird Fic­tion”

Stephen King’s Top 20 Rules for Writ­ers

Josh Jones is a writer and musi­cian based in Durham, NC. Fol­low him at @jdmagness