Cryptic thoughts
by Hot
24 years ago, or so, I wrote a manifesto of sorts, “Cryptic Thoughts”. I addressed what I felt were two problems in the US cryptic universe: (1) A shortage of difficult cryptics, and (2) narrow rules that prevent the emergence of a variety of voices among constructors. The response from my fellow puzzlers was largely negative. I was told by most that American cryptic “rules” represented the only possible square dealing rules, because, well, that’s how it is. There were some exceptions — a few expressed agreement with my points, a few agreed with some points but not others, and a very few disagreed but made clear they realized that those conventions were just that: conventions they happened to like.
I largely gave up on acting upon my views, and in 1995, when Sibyl asked me to work with Trazom as a cryptic editor for The Enigma, I joined the cryptic mainstream in enforcing the dominant US consensus on our contributors. With Trazom’s help, I wrote strict guidelines for Enigma cryptics. I am proud of the work we did as editors, which culminated in the National Puzzlers’ League Cryptic Crosswords book.
Nevertheless, I held on to my misgivings, which I expressed at an NPL con panel about cryptics in 2007, and to which I return today.
In many ways, things have improved since my first cryptic manifesto. One big change is that if one is looking for challenging cryptics that are made in the USA, one has to look no further than The Enigma, or Ucaoimhu’s website. I’m not as strong a solver as some other NPLers, so I am quite content with the current range of difficulty in US cryptics.
However, we still do not have either the range of difficulty or the range of styles available in the UK. When it comes to a greater variety of voices, if you look very carefully, you may see a slight relaxation of the “rules” in the US cryptic scene. But in a blind test, very few solvers, if any, could tell one US cryptic constructor from another solely by looking at the clues. Thus, we lose a pleasure we have in other puzzle forms (such as flats and US-style crosswords): that of getting to know a constructor, and appreciating their style and their quirks. Also, I imagine that constructors with different personalities would appeal to solvers with different personalities. This uniformity in style is largely due to the continuing broad consensus among editors, constructors, and some solvers that US cryptic conventions deserve to be rules.
The result of that consensus can be seen most clearly in American black square cryptics. Perhaps you enjoy solving those. If so, don’t let me rain on your parade, but as for me, like just about every experienced solver I know, I don’t find those very interesting. Even though some great clues can be found in the best of those puzzles, they have a limited audience. On the one hand, they are too difficult for people who have not been initiated into this arcane hobby. But on the other hand, experienced solvers don’t stick with them, and prefer to solve variety cryptics, and British cryptics. I blame this state of affairs on the continuing narrowness of vision in the US cryptic scene.
We could learn a lot by looking at what the Brits do right. I know some people are impatient, occasionally even contemptuous, of the UK scene. But look at what they’ve got! Five daily cryptics, a wide range of styles and difficulty, multiple blogs about cryptics, and huge numbers of solvers. How do they do it? Part of the answer is that they do not obsess over rules.
The Brits have plenty of healthy debate about cryptic clues and themes, a lot more than we do in the US, but not so much about adherence to supposed rules. In fact, one British expat told me her father was a cryptic solver on the other side of the pond, and that he was amazed when he found out that in the US, there were rules to this game.
As far as I’m concerned, the only rules worth having are these two: the cryptic reading of the clue must lead to the answer, and the definition must be correct. Constructors should feel free to use any and all sorts of wordplay, and not be limited once and forever to a short list of “legal” clues. Here are some of the constraints on diagrams and clues that I believe we should dispense with:
[See Appendix 1 for one consequence of these constraints.]
- no three-letter words
- no etymological overlap between wordplay and definition
- no breaking up compound words into their parts
- no splitting entries between different locations in the grid
- no leakage between wordplay and definition
- no wordplay that does not fall strictly under the permitted types
- avoidance of uncommon words
There are also some aesthetic criteria that have been elevated to near-rule status, such as a preference for simple elegance over riotous excess, and thus for short clues and spare themes.
And here are some wordplay opportunities that are either unused or underused in US cryptics:
- puns
- themes in black-square puzzles
- compound anagrams
- clues with more than two parts
- rebus clues
- letter banks
- heteronymy in clues
Trazom and I have used all of these in our puzzles, and while we know they have irked some of our friends, they have almost always been highly prized by most of our test solvers. [See Appendix 2 for examples.] I have no problem with constructors sticking to US conventions if they want, and some do it brilliantly. But as a solver, I have enjoyed solving puzzles that break some or all of these conventions, and I resent the fact that there is no venue where such puzzles can be published in the US. (The only exceptions I know of are unedited puzzles such as Richard Maltby’s in Harper’s, and the puzzles Trazom and I construct as Out of Left Field.)
I am not suggesting that anything at all goes. [See Appendix 3.] Constructors and editors need to be committed to fairness: clues should be solvable, and trade-offs between entertainment and correctness should be weighed conscientiously. But the preoccupation with narrow rules has not been successful in bringing cryptics to the masses, and some experimentation should not be considered off limits.
In particular, the etymological taboo needs to be retired. It makes too many puns unacceptable, even though, to normal people, puns are the most common form of wordplay. The etymological taboo is not respected by Richard Maltby, and was not by the late Henry Hook, who once brilliantly clued I STAND CORRECTED with a reference to orthopedic shoes. Even Cox and Rathvon have been known to deviate from it (e.g. cluing FENCES with “Fights barriers (6)”).
It’s time to lighten up. Cryptic crosswords are for fun, and the overriding concern of their creators should be entertainment, not correctness, according to immovable rules. Editors, constructors, and solvers should be as open to innovation, variability, and creativity as they are when it comes to any other word puzzles.
When Trazom and I took over the Nation puzzle in 2011, we decided to incorporate some of what we liked in our predecessor Frank Lewis’s style, as well as some NPL forms of wordplay, and to only stick to the parts of US conventions that we agree with. I hope that others will join us with their own parameters, and that the cryptic scene will continue to diversify.
[Many thanks to Trazom for his feedback on earlier drafts of this essay.]
Appendix 1
One problem in US cryptics is that long entries are almost always clued by anagrams. I personally don’t enjoy long, non-apposite anagrams: while they’re easy to construct, I don’t think they’re fun to solve. You may like them, but still, we’d all be better served as solvers if there were more variety in the cluing of long entries. Opening the door to some of the above options would definitely help, and may even increase the number of long entries in the puzzles we solve.
Appendix 2
Here are some convention-flouting clues. They may not be the greatest clues ever written, but really, are they so bad that they should be barred from publication? All of them are from past or future Nation puzzles, so you can sic the cryptic police on Trazom and me. (More drastic violations are of course possible, but we have not moved all that far from the US mainstream.)
HERACLES He clears stable? No and yes (8)
[“He clears stable? No” is the wordplay (an anagram of ‘he clears’) and “He clears stable? Yes” is the definition, a reference to Heracles’ labor in cleaning the Augean stable.]
NEON LIGHT Darkness engulfs Libya’s capital after many years–this could vanquish it? (4,5)
[The definition (“this could vanquish it”) doesn’t stand on its own, because “it” refers back to the “Darkness” mentioned in the wordplay part of the clue. We acknowledge the fact that readers do not forget one part of the clue as they read another.]
BEETHOVEN’S THIRD E is for “Eroica” (10,5)
[Rebus clue]
TEA FOR TWO AND TWO FOR TEA Song excerp2 from 19T5? (3,3,3,3,3,3,3)
[Phonetic rebus]
PERSPIRE React to the heat with steeplechasing agent in retreat (8)
[Heteronym in the clue. Insert a space between “steeple” and “chasing”.]
ENID Eat uptown in Oklahoma? (4)
[Reversal clue for a Down word. We suppressed the space between definition and wordplay.]
HAIKU Recited aloud, an exalted, brilliant stroke — this clue, for instance (5)
[Phonetic charade: high coup]
SHREWD Cache rudimentary audio holding — that’s clever (6)
[Phonetic hidden word]
TROPICAL Hot car and pilot in smash-up (8)
[Compound anagram of “car” and “pilot”. US convention is that all the letters in an anagram need to be contiguous, with the absurd consequence that some constructors use an ampersand in clues such as this one.]
CATCH-AS-CATCH-CAN Snatch, using every element as needed in whatever way possible (5–2‑5–3)
[Letter bank: snatch]
TO BE OR NOT TO BE Existential question latent in Brontë’s letters (2,2,2,3,2,2)
[Letter bank: Brontë]
RING Call band “Surround Sound” (4)
[Quadruple definition]
GANDER Goose ranged freely for a look (6)
[Wordplay between two defs]
LOCK, STOCK AND BARREL Totally overheard: conversation concerning smoked fish and mineral (4,5,3,6)
[Homophone of “lox talk and beryl,” breaking up the phrase almost as in the original]
SPEAKER OF THE HOUSE Part of home stereo system for politician (7,2,3,5)
[Pun]
ORGAN RECITAL What you might hear in church: “liver, kidneys, pancreas…” (5,7)
[Pun]
Appendix 3
Lest you think I’m arguing for total chaos, here are some constraints I approve of:
- No missing or incorrect indicators
- No extra words that ruin the cryptic reading (I don’t object to harmless extra words)
- No indirect anagrams (and, a fortiori, no indirect letter banks)
- No parts of speech inconsistency between def and sol
[This article originally appeared in the Guide to the Enigma, and an earlier edition can be found in the print version of the Guide. Hot has provided a recent (2026) update to this article, and it has been formatted for the web]