Anyone who is interested in crosswords will know of Ximenes
(1902-1971). He wrote crosswords for the Observer from 1939 until his death, when the present incumbent, Azed,
took over. Ximenes is widely regarded as the finest compiler who has ever been. He wrote clues which showed an ingenuity
second to none; witty, taxing and above all, fair. I have heard him described as the Mozart of crosswords (which I suppose
makes me a drummer in a pub band somewhere!). That is not to say there have not been great compilers since – Azed,
Dimitry (“Listener” and Sunday Telegraph “Enigmatic Variations”) and the
Guardian’s inimitable Araucaria are three great compilers who spring readily to mind. And
don’t forget those clever Times compilers, who remain anonymous.
Ximenes’s name lives on not just for his puzzles, but for the work he
did to ensure that a reasonable standard of crossword writing was achieved in his day and thereafter. The phrase “Ximenean
standards” is (or should be) a phrase familiar to every aspiring crossword writer. What exactly does this mean? And
are they still relevant now?
There is not enough room here to go into great detail about Ximenes’s
thoughts on crossword writing. For an in-depth treatment I recommend his own book – Ximenes on the Art
of the Crossword by D.S. Macnutt (Swallowtail Books). A briefer but useful guide is to be found in the Chambers
Crossword Manual by Don Manley (Chambers). In a sense, it is possible to sum up Ximenes’s ideas in one
sentence: CLUES MUST BE FAIR. That’s it. Essentially, all he strove to do was ensure that the
most important contributor to any crossword – the solver – got a fair deal.
Here is an example of a clue of which Ximenes would not have approved:
Beast of burden bothered the players (9)
The “bothered” suggests an anagram, but an anagram of what? “The
players” provides too many letters, and “beast of burden” doesn’t fit at all. Now, let us suppose
you have got all the letters towards it:
O _ C _ E _ T _ A
Well, the only obvious word that fits is ORCHESTRA, and after some thought
it’s possible to realise that ORCHESTRA is an anagram of CARTHORSE (yes, that old chestnut!) and that a carthorse is
a beast of burden...
The clue is unfair. It asks the solver to work out an anagram of an unknown
word – an almost impossible task. The only way to solve this clue is to get the letters towards it and guess –
then work backwards. There is not enough information in the clue itself to lead to an answer – hence its lack of fairness.
This was the kind of thing Ximenes was trying to eradicate from crossword puzzles intended for publication in respectable
Here’s another example.
One who sleeps late could be ideal (3-4)
The answer is an anagram of “be ideal” to give LIE-ABED –
but the problem with this clue is that there’s a word missing. It should really read
One who sleeps late could be be ideal (3-4)
Obviously this makes no sense (not that the original makes much either) but
at least it is a fair clue. In the first case, the word “be” is doing double duty – being part of the anagram indicator and the anagram itself. How can the solver reasonably be expected to
Here’s another example of a clue which would have Ximenes turning in
Leading nationalist with Roman organisation finds wisdom (6)
The idea is that N + anagram of ROMAN gives NORMAN (Wisdom)
There are three serious errors here. “Leading nationalist” does
not, in any grammatical sense, convey the letter N. Far better would be “nationalists’ leader”, “leader
of nationalists” etc. Nor does “Roman organisation” really suggest that the letters in Roman actually move
– nouns do not modify other words in the way adverbs, adjectives or some verbs do. Better here to have “Roman
uprising” or “Roman militant” or something similar. And the last error, perhaps worst of all, is the attempt
to clue a person’s name whilst omitting the capital letter all proper nouns should have. This is an easily solved problem –
just put it at the beginning of the sentence!
So we could have
Wisdom found by leader of nationalists meeting Roman militant (6)
It’s still not a brilliant clue but at least it gives the solver a fair chance.
Very often, it is the “easier” crosswords – those that appear
in magazines for example – which provide the least fair clues. I can think of several times when, puffed up with pride
at having finished the Times in under ten minutes, I have picked up the Anytown Quick TV Guide expecting to
complete the crossword while the kettle boils – and taken longer than I did on the Times! Sometimes
clues in these publications are so woolly and vague that more guesswork than skill is required. I have spoken to one crossword
editor (not for any of the nationals, I hasten to add) who blithely told me that “we don’t bother with that Ximenean
stuff here”. I have seen some of the clues printed under his imprint and wondered how anyone actually managed to solve
Having said all this, it’s important that aspiring setters don’t
get the idea that writing clues is a sterile process that has to follow lots of complicated rules. Provided you have a reasonable
grasp of grammar – and you wouldn’t be able to do crosswords if you didn’t – it soon becomes second
nature to check that your clues give a fair indication of what the solver is required to do.
Should all clues in every crossword be Ximenean? Opinion is divided on this.
Ideally, the answer is yes, since all being Ximenean means is being scrupulously fair to the solver. Yet occasionally, I believe,
there is nothing wrong with a few minor infringements here and there if the clueing is imaginative. Any decent crossword compiler
would be ashamed to produce any of the examples given above, but some compilers do bend the rules of grammar slightly
if the clue is clever and amusing enough to justify it and the result is still relatively fair.
For example, I used to buy the Guardian almost solely for the pleasure
of seeing “Set by Araucaria” above the crossword; his puzzles
are probably the most enjoyable and imaginative I have found. He – and his excellent colleagues Bunthorne and Enigmatist,
to mention but two – has been criticised in some quarters for minor infringements of Ximenean standards. Of course none of
them would ever produce anything awful like the examples above, but sometimes they do stretch things a bit. Yet
given the imagination shown in their clues, does it really matter? Personally, I think not. After all, crossword writing is
a creative art, and if creative artists didn't strain at the leash occasionally, there would have been no Picasso, no Wagner,
and of course no Shakespeare! All these artists have one thing in common, however; they may have bent the rules of their art,
but they leave their public in no doubt as to what they intended to say. This, too, is what the compiler should strive for.
For my own part, I aim to produce clues which are all, or almost all, of Ximenean
standard. A trawl through my puzzles on this site will no doubt unearth a few that aren’t – either by mistake
or because I felt the clue justified the infringement. For example, I don’t really have a problem with superfluous link words
as long as they aren’t misleading. Compare these two clues
Animal returns to grass (4)
Saturnalia upsetting to national (10)
In both cases, “to” is a link word. It is also largely superfluous,
although it can be argued that “to” means “leading to” i.e. the wordplay is “leading to”
the answer. Ximenes would have no truck with the first example, and neither would I. DEER returns to give REED (the answer)
– fair enough at first sight, but the solver could legitimately claim s/he was looking for a synonym of the verb “to
grass” – to inform on. I don’t think staunch Ximeneans would approve of the second either, but here I would
argue that there is no verb “to national” and that it is obvious enough that an anagram of SATURNALIA leads to
a national of a country (AUSTRALIAN).
I would say the same for one of the most oft quoted of non-Ximenean clue
types, such as
Man changing line around Gateshead (5)
We’re looking at an anagram of LINE around G – to give NIGEL – and some would (technically rightly) argue that as there is no apostrophe in Gateshead, it doesn’t signify the head
of gate, G. Yet Gateshead, Maidenhead, egghead have long been used as indicators for the letters G, M and E and most solvers
know what is expected by such a clue. And it’s far better than “leading gate” or any such nonsense.
If you are looking to set puzzles for the Listener or Enigmatic Variations,
then you will find that you will need to stick very closely to the type of clueing expected by Ximenean standards. It would
be as well to read around the subject in some detail, particularly the two references I have given above. These are complex
thematics which involve obscure vocabulary, and there is no room for ambiguity. The Times and Guardian crosswords
aren’t strictly Ximenean, but as said above, the quality of the clueing usually justifies this. That doesn’t mean
that anything goes, of course – you'll get an angry letter from their crossword editors if you start giving woolly
definitions or miss out anagram indicators. Other publications, from the tabloids downwards, are very variable on this. This
is where most people start their crossword writing career, and it is important to remember that the readers of these publications
regard their crosswords as a brief distraction, rather than an obsession. In other words, the clues have to be easy.
Obviously fair clues are easier than unfair ones, so here too it is best to stick to Ximenean practice as much as possible
– if you deviate too much from it, your clues will be too hard and you will get several complaints from your readers!
To finish, I would say that for anyone who intends to become a serious crossword
writer, it is best to study and follow the Ximenean way of clueing. Only
depart from it if you are confident that the clue is still reasonably fair and solvable, and is good enough to justify
bending the rules of grammar. In any case, if you are writing fair clues, you are most likely following Ximenean practice
If you have any questions about whether a clue you have written is Ximenean
or not, click here.