How to write a good cryptic clue - an American perspective
Rodders
part.
Rodders
eGroup home: http://www.eGroups.com/list/sob_softly_small_spasm
I guess it's time for me to weigh in. I'm a professional cryptic constructor
(Games, Tough Cryptics, The New Yorker) with a
relatively-brief-but-sufficient eight years of experience in the American
cryptic industry. Here's what I can tell you about writing a good clue,
insofar as you take "writing a good clue" to mean "writing a clue that an
American cryptic editor will find particularly praiseworthy":
1. It MUST have a good surface sense. The stronger, the better. Ideally
this is done by linking two parts of the wordplay, or the wordplay and
definition halves in some way, so they look more like a single conceptual
whole. RWANDAN could be clued as "African ran around stick (7)," but it's a
lot nicer when you clue it as "African competed in a relay, holding baton
(7)."
2. It should be creative, from a solver's standpoint. If you're doing "Trim
a tree (6)" for SPRUCE or "Heavy metal star (4)" for LEAD, you should try
harder, because most dedicated solvers will have seen them before.
(Exception: if you're trying to make a clue easy for newcomers, this can be
acceptable.) This also applies to the word choices and wordplay: Cluing
TEMPERA as TEMPER + A has been seen a zillion times. A REP MET, backwards,
is still uncommon enough to be considered nicer. A hidden word (or a pun on
TEMPURA) might get bonus points for creativity, since most clue-writers look
at TEMPERA and can't see beyond the charade.
Also, a rarer word (GASKET) or a phrase (TWO OF HEARTS, which is "woof"
inside "the arts," by the way) is preferable to words that everyone has seen
(like TEMPERA) when you can do it. Just be sure to get good wordplay out of
your new recruits.
3. In America, at least, the clue should be efficient: any charade that's
broken into five parts has been sloppily thought out. (Or, more likely,
wasn't thought out at all. There's not much you can do with a word like
SPHINX that doesn't involve breaking it into tiny little parts. What's fun
about that?) My personal rule is to try to minimize the number of "parts"
every wordplay has (each part being a charade section, a reversal, an
anagram, etc.), doing "one-property" words whenever possible (perfect
reversals, single-word-containing-single-word containers, elegant whole
anagrams rather than partial ones) and NEVER doing more than four "parts" in
any clue. (And never more than three in words with fewer than seven
letters.)
4. Finally, play by the rules. Puzzle editors are trying to sell magazines,
and people are most happy with puzzles they can understand and solve.
Challenge, yes, but do it comprehensibly and on common ground, on the theory
(quite sensible in America) that most cryptic solvers are relatively new to
the form and still find the standard games stimulating enough. Most
"rule-breaking" clues are, upon closer analysis, simply lazy clues that
could
have been rethought, or bad clues that need to be brought up to code. (The
example in the original post is, to an American editor's mind, just plain
illegal: every clue has to have a definition distinct from its wordplay, or
it'll inspire angry letters.)
The chief exception is clues that use loose wording in the service of a
surface sense. I once clued DAIQUIRI as "Randy Quaid seen with one Rhode
Island drink (8)." "Randy" is not a particularly good anagram word--it
suggests loose morals, not physical looseness or insanity--but the editor
let
it stay because he felt that the ideas of looseness and corruption were
clear
enough by association, and the surface sense was worth preserving. Surface
sense can trump a general rule as long as the rule is made "fuzzy" rather
than completely destroyed; the basic mechanics of cluing, however, cannot be
waived.
FINAL NOTE: Almost all of these guidelines are predicated on the idea that
the puzzle is a way to introduce an informed solver to the interesting
wordplay properties of certain words, en route to a complete solution. So
this presupposes that a., all the words will actually HAVE some interesting
property, and b., the solver will eventually get it. American cryptics are
designed to be solver-friendly, at least in part because it's still a
relatively new puzzle form, and easier puzzles are a happier way to obtain
new recruits and (we hope) addicts.
This is why American cryptics have two rules that don't seem to be followed
outside the colonies:
A. Wordplay should not be etymologically related to the word it is
cluing--otherwise it makes a clue less inherently interesting. It's not
uncommon to see a British puzzle that clues SIDEWALK as a charade of SIDE
plus WALK. This would never wash in America, where you'd be expected to go
another direction--maybe AWED backwards inside SILK. After all, noticing
that SIDEWALK can be broken into its two obvious parts isn't noticing
anything particularly neat.
This means, by the way, that the two halves of double definition clues can
never be etymologically related to each other in American cryptics. I've
seen a British cryptic where the word HAND was clued as "Give applause (4)".
It caused me quite a bit of culture shock.
This is not an optional felicity in America. This is a RULE. Never submit a
clue that breaks it. (A common mistake: cluing GAZEBO as GAZE + BO. As it
happens, "gazebo" is etymologically related to "gaze," so the wordplay's not
legal.)
B. If you've solved the clue you should know what word is being clued
without any additional information. This means that homophone indicators
should always be placed to avoid ambiguity. "Antelope reporting information
(4)" could be either GNUS or NEWS, and the only way to find out is to solve
for the crossing words. This isn't unthinkable abroad, but in America, the
word "reporting" would have to be moved to one end of the clue or the other,
to make absolutely clear which half contains the wordplay.
Note, by the way, that the fact that I've written all of this does not mean
that I think all these rules are optimal. (I personally would LOVE to see
American cryptics start using stranger words, for example. Then I could
finally write that double definition clue on "flageolet"!) But I think most
of them represent commonsense balances of fairness, complexity, and
aesthetic
appeal. And I know--from sometimes painful experience--that American cryptic
editors would all agree with what I've written. That's about as official as
something like this gets.
--David Ellis Dickerson
------------------------------------------------------------------------
Peter T. Daniels
> Saw the below in a cryptic crossword group which is quite interesting in
> part.
>
> Rodders
>
> eGroup home:http://www.eGroups.com/list/sob_softly_small_spasm
>
> I guess it's time for me to weigh in. I'm a professional cryptic constructor
> (Games, Tough Cryptics, The New Yorker) with a
Is that really a commendable credential? I don't recall anything
specific about David Ellis Dickerson's work, though I certainly know
the name, but The New Yorker and Games Magazine World of Puzzles
carried shockingly easy puzzles (the New Yorker ones essentially half-
sized). Every puzzle in GMWoP carries a difficulty rating from 1 to 3
stars (including half-stars), and every cryptic, no matter how simple,
is rated 3.
(It's a bimonthly magazine, and until the middle of last year the only
challenging item was the contest, which had you solve a usually
several-step puzzle and send a solution on a postcard, with a first
prize of $500, randomly selected from all correct entries. They then
discontinued that, and I stopped buying the magazine after they
published the solution of the last one, two issues after it appeared.)
There were usually two straight cryptics in each issue and one
"variety cryptic," i.e. with a gimmick like the Cox & Rathvon Atlantic
ones, but almost never at their level of challenge; for one or two
issues early in 2009 the level was suddenly higher, and they may have
gotten many complaints, because by the time I stopped buying it, they
were below what had been their normal level.
Peter Biddlecombe
Looking at the content of the advice, I'd agree with points 1 and 2 as
stated, for anyone writing cryptic clues. Likewise point 4 subject to
the thorny problem of what the required rules should be. Point 3
shouldn't really need stating explicitly - it's very difficult to chop
a word into five or more parts, give fair instructions about using
them to make the answer, and still follow rule 1 or write something
conspicuously epic - so it will rarely be broken anyway. _If_ a
setter can find a way of expressing 5+ parts succinctly, why not?
American cryptics seem stuck between two problems, resulting in a
Catch-22 problem: gaining an audience from people used to other xwds,
and retaining that audience by challenging its experienced members at
least some of the time. One of these requires easy puzzles all or
most of the time, the other requires difficult puzzles some of the
time, and frequent enough puzzles for new solvers to start bridging
the easy/hard gap before they forget what they learned from the last
easy puzzle. To justify the frequent puzzles, you need the audience,
but to maintain that audience, you need the frequent puzzles so that
you can have some tough puzzles. We should probably remember here that
British cryptics as seen in daily papers came about by a rather
different route - over a few decades they evolved from the plain-
definition puzzles imported from the US in the twenties, and there are
still a few changes taking place gradually.
One day the English-speaking world might have a global "newspaper",
accessed by technology we can barely imagine, and my guess is that it
will have two kinds of crossword - US-style non-cryptics and British-
style cryptics, each varying in difficulty. A few people will tackle
and enjoy both.
Peter T. Daniels
We have lots of monthly crossword magazines, but even more magazines
of the less challenging puzzles like Word Search. We have no magazines
of cryptics, and only a few books of cryptics a year, and even fewer
of "variety cryptics." The general puzzle magazines include crosswords
of both kinds, but never at a NYT Friday or Saturday level.
Peter Biddlecombe
> On Jan 25, 7:18 am, Peter Biddlecombe
>
>
> > One day the English-speaking world might have a global "newspaper",
> > accessed by technology we can barely imagine, and my guess is that it
> > will have two kinds of crossword - US-style non-cryptics and British-
> > style cryptics, each varying in difficulty. A few people will tackle
> > and enjoy both.-
>
> We have lots of monthly crossword magazines, but even more magazines
> of the less challenging puzzles like Word Search. We have no magazines
> of cryptics, and only a few books of cryptics a year, and even fewer
> of "variety cryptics." The general puzzle magazines include crosswords
> of both kinds, but never at a NYT Friday or Saturday level.
Not very different from here. I'm not aware of any cryptics magazine
that's available in a newsagents - they exist, but all by private
subscription. Of the "broadsheet" cryptics, only the Times and
Telegraph can be relied on to produce books of cryptics annually -
there was a recent FT book but that was the first I can remember in 20-
odd years. No Guardian book for about 2 years, and no Independent one
for 5 or more. Barred-grid or "variety" cryptics: we probably get
something about every other year - the last one I remember is the
Listener collection from 2008.
dinizi...@gmail.com
I looked around- but couldn't find any contact info so trying here...