What is the most common image of crossword solvers among non-addicts? The odds are good that most people would see the stereotypical solver as mild-mannered, middle-aged, slightly bookish – maybe a member of the clergy or an academic – who likes to retreat from life’s little annoyances for half an hour every day into a world where flowers are rivers, there are more doctors than the NHS could ever hope for, the Queen turns up a lot too and the 1950s concept of U and non-U is alive and flourishing. Yet take a look at some crossword discussion boards, and from time to time you will see certain aspects of clueing being debated with more ferocity than one might expect when the subject is basically a pleasantly distracting word game. Many of those who air strong opinions on this are well-regarded or promising setters themselves. Not that there’s anything wrong with people showing passion for a hobby – thank goodness the anti-intellectual “laddism” of the 90s which branded almost any interest other than sex, lager and football “nerdy” is truly gone, perhaps in part due to the popularity of Sudoku puzzles. But what is it that people get so steamed up about?

Essentially, many setters and serious solvers fall neatly into one of two camps. These are the “Ximenean” and “non-Ximenean” factions. The Ximeneans believe that all cryptic clues should follow the guidelines stipulated in Ximenes’s book Ximenes on the Art of the Crossword (D.S. Macnutt, Swallowtail Books) and other pronouncements on crosswords by this great compiler. For a brief guide to principles of Ximenean clueing click here. The non-Ximeneans argue that since Ximenes died in 1971 it is surely time to move on and explore new possibilities for pleasantly baffling the solver. Sometimes this camp is labelled “Araucarian” since the Guardian’s most popular compiler by far, Araucaria (John Graham, who received an MBE for his work in 2005) is perhaps the most widely known compiler whose work deviates from Ximenean principles. I think this title is rather misleading since Araucaria is not the only opponent of strict rules and he has never actually set down an alternative set of rules or principles himself. Of course many setters and solvers have their feet in both camps. As a solver I do too, though as a setter I’ve tended more and more towards Ximenes over the years, so that only my little toe is left in the non-Ximenean or “libertarian” camp.

When Ximenes wrote Ximenes on the Art of the Crossword in 1966 there was a very great need for something like this. Look back at old Times crosswords, for example, from the 50s and many of the clues in them don’t pass muster by almost any standard today. There are anagrams without definitions, plain quotations, and the literary knowledge required is far too specialist to attract new solvers today – although that could reasonably be blamed on our so-called progressive education system which shrinks from teaching almost anything about British heritage and culture. If there hadn’t been some sort of guidelines laid down to ensure that these puzzles gave the solver a fair chance, crosswords could well have become an activity for an ever-decreasing exclusive group of scholars, rather than the popular activity they undoubtedly are today.

Yet Ximenes himself was by all accounts an unassuming and modest man, like many geniuses, and it is inconceivable that he would have intended that crossword writing should stick rigidly to his principles and his alone. He more likely intended to provide a framework within which the setter could use his or her inventiveness and entertain solvers for years to come. It does seem, though, that some of today’s post-Ximenean setters cling so rigidly to what they see as Ximenean rules that they provide clues which are scrupulously fair but a little...well, dull.

Perhaps I can indulge myself with one of my many musical analogies here. The two composers Mozart and Salieri were contemporary rivals. Salieri’s music was scrupulously “correct”, following to the letter the established rules of harmony and composition of the time. Mozart, on the other hand, followed the rules too, but allowed himself a few liberties in order to give his music greater expression. How much of Salieri’s music do we hear today?

That is not to suggest that Ximenes was dull – his crosswords were incredibly inventive – or that all of his adherents today are either. Mr Magoo and Dimitry, who both set for the Listener, and Pasquale, the only staunch Ximenean among the Guardian’s team of compilers, always provide fresh and entertaining puzzles. However I can’t help feeling that the Listener and similar puzzles’ (probably necessary) insistence on strictly Ximenean clueing does sometimes frighten some setters into playing safe to avoid rejection or excess editing, and this is apparent in some of their clues. I would cite my own first Listener puzzle, Europe’s Ports by Alberich, as an example of this. I met a well-regarded Listener setter at the annual Listener dinner who told me that “it would be nice if we could throw away the straitjacket and write some fun clues sometimes” or words to that effect.

I stress that being wary of being too hidebound by perceived “rules” is not the same as saying “anything goes.” I certainly have no time for sloppiness, and there is a huge difference between a well-considered new idea and what is essentially poor clueing. I hesitate to use the term “unsound” as it makes it appear there is only one right way to do things. Yet if I’m sent a puzzle that has “leading politician” as an indication for the letter P in the first clue I solve, I don’t approach the rest of the puzzle with much confidence. I can see no justification for “leading” as a first letter indicator, other than it is an ellipsis for “that which is leading” and like a joke, if you have to go round the houses to explain it, it isn’t much good. Other bugbears are missing indicators for hidden words or anagrams, redundant words inserted just to make good surface reading, “A inside B” when B is really inside A or missing capitals. To exemplify the last two:

Fashionable attempt inside goal (6) = TRENDY: END in TRY

This is just plain wrong, although it crops up with depressing regularity, and not just among novice setters. The wordplay is the wrong way round, rather amusingly described as “arfacese” on the now defunct Guardian talkboards (go figure).

Create king in west? (4) = MAKE: K in MAE

Should be West as we’re referring to a proper name here, and even taking into consideration the spread of lower case advertising logos and of course text speak, a puzzle designed to celebrate the English language should not abuse it in this way. It is a grammatical error, pure and simple.

(While we’re on this clue, I’d say that surely Mae = West is long due for retirement – how many people under 40 actually know who Mae West was? It’s not as if she was Prime Minister or anything, splendid lady though she was. Same goes for “girl” = Ada or Maud and "yob" =  Ted – nobody talks about teddy boys any more. And although I’d rather stick needles in my eyes rather than write “u r there? Gr8” in any sort of message, I’d take you = U any time over the absurdly outdated U = posh/acceptable.)

Let’s move on from the merely sloppy to the more inventive but decidedly non-Ximenean type of clueing. Two examples spring to mind here:

Beaten, an achievement indeed (8) = DEFEATED: FEAT in DEED

Walter now is to change (5) = ALTER: WALTER minus W

The first of these is an example of what may be the most often debated type of clue. Does indeed equal in deed? The Ximenean argument against it is that it is an abuse of spacing and therefore it doesn’t, and this a very valid point. However others argue equally cogently that almost all solvers read clues aloud in their heads and so can easily interchange “indeed” and “in deed” (or “infer” and “in fer”, or whatever) and thus the clue is fair, especially if there is a question mark or an “as it were” at the end of the clue. I remember this device being done to death in the Telegraph, I think it was, in the 70s – there was always one clue that contained “indeed” and you knew to write DE and ED at the start and end of the answer before even making any attempt to work it out. I don’t feel cheated by this device, and have seen some novel uses of it, notably the use of “incontinent” in this way to put some letters inside ASIA. I have not used this trick for a very long time, but I did once provide a tongue-in-cheek clue for a thread on the old Guardian talkboards, which was:

He probably wouldn’t have approved of this clue, ultimately mine: “Playing INXS” (7)

No prizes for guessing the answer to that one!

As for the second clue, this is a very different matter. The idea that “now” = no W” is one step too far for almost all compilers, certainly including me. Reading aloud, “now” bears no resemblance to “no double-u” and although on the surface this may seem like a clever idea, it is too unfair to be considered a good clue, even when the name Walter is given rather than the more vague “man”. It’s what I’d call a “yeah, right” clue – guess the answer, figure out how it works, then mutter rather dissatisfiedly “yeah, right.”

After receiving a communication issued by a petty civil servant, objecting to the ending of a sentence with a preposition in official documents, Winston Churchill wrote in the margin: “This is the sort of pedantry up with which I will not put.” Here I hope to show where Ximenean rigour can be taken to extremes. It is considered fair by almost all compilers, Ximenean or not, that while “colour” adequately defines “red”, “red” is only an example of a colour rather than a definition of colour, hence this needs to be indicated. “Perhaps”, “maybe”, “possibly” or a question mark are usually used to do this. Now let’s look at a possible clue for the word CARPET. It’s not an easy word to define, as “floor covering” is hard to fit into a sentence and is in any case a bit of a giveaway. So suppose we decide to define by example – using Axminster, one of the most famous types of carpet. Looking at the wordplay, one obvious way to break the word down is into CAR and PET. Rover is a type of car and also a popular dog’s name, and of course dogs are common pets. So a possible clue would be

Rover, a dog in Axminster (6)

Now, although that reads well, there is no indication that both elements of the wordplay and the definition also are all examples. So applying Ximenean principles to the letter, we get

Rover, perhaps, maybe a dog in Axminster possibly (6)

The result is technically OK, but even if you replace the last “possibly” with a question mark it still sucks, as our transatlantic friends would say. Let’s change it to:

Could be Rover, a dog in Axminster? (6)

“Could be” is a perfectly acceptable replacement for “perhaps” and phrased like this, as a question, the clue reads a million times better than the previous example. There may be some debate as to whether the “could be” can cover both Rover and the dog, but the clue is quite solvable and I leave you to decide which you prefer. Incidentally, I saw a clue similar to the second example, stuffed with “perhapses” and “maybes” in a national newspaper once and remember thinking that it looked so out of place in an otherwise decent crossword that it must have been an oversight.

This clue raises another point which divides serious cruciverbalists. This is the use here of the word “in”. Some would argue that as a link word “in” is directional, e.g. that a clue can only be presented in the form [answer] in [wordplay]. It is true that the answer is to be found in the wordplay rather than the other way round, and some setters (including me, these days) try to adhere to this concept. Yet several other setters – some of them devoted Ximeneans – treat “in” as a bidirectional link word.

I will finish by mentioning a couple of old chestnuts which Ximeneans have criticised but which have taken on an almost mythical status over the years.

The first is

HIJKLMNO (5)

for WATER, i.e. H to (two) O. Is this a brilliant piece of innovation or an absolutely unfair, rotten clue? I think it’s both! It has no definition. There’s no homophone indicator, and anyway “to” and “two” aren’t really homophones. Nor does the clue actually mean anything. Yet people remember it many years on, and it has to be admitted there is a kind of beautiful simplicity about it. No compiler would dare use it again, though!

Heggs? (11)

is another example that displays all the faults of the previous clue. The answer is EXASPERATED, i.e. “eggs aspirated”. I have a fond memory of this clue as it is the first ever cryptic clue I got on my own.

However, I can muster up no enthusiasm for the clue

E? (13)

which seems to be quite popular in some circles. I would never have got it myself and for a long time I didn’t understand the answer even when I saw it. Eventually I saw an explanation that the answer, SENSELESSNESS, works on the basis that if you take the letters of NESS from SENSE you’re left with E. Apart from the obvious objection that the letters to be removed appear in the wrong order and there is no indication of this, I do not think there is enough information here for the solver to have a chance to solve it. It’s a bit like setting “6” as a maths problem and expecting the student to come up with 13 – 7. Very much a “yeah, right” clue. Perhaps this clue could be rescued to some extent by supplying a definition along the lines of

E? Absolute stupidity (13)

I still think the wordplay is unfair but the definition does give the solver more of a chance, and the clue makes a true statement about the foolishness of taking dangerous drugs.

After all this I guess I need to draw a conclusion – and will do so with yet another analogy (not musical this time). I once saw an old lady making her slow and painful way up to the counter in a post office just before one o’clock one day. The petty, jobsworth clerk stood and watched as she did this, then as she reached the counter at exactly one o’clock he pulled down the “Closed for Lunch” sign. I hasten to add that I am not for a moment suggesting that Ximenes or any of his present-day followers are in any way like that horrible clerk – I have met several of them and I am sure that they would have gladly waited for the old lady and served her with courtesy. I merely illustrate that following rules slavishly to the letter rather than the spirit can sometimes have the opposite effect of what the rules were actually intended for, and we need to beware of being so dogmatic that we end up following a set of rules just because they are there. Ximenes’s guidelines today still give an excellent indication of what generally works and what doesn’t in a clue and it is good that there are plenty of people around who feel passionately about upholding the standards he laid down. Yet in essence crosswords are there for entertainment and fun, not precise exercises in some scientific discipline, and we should welcome original thought when it is entertaining and doesn’t cheat the solver, even if we have to relax the rules a little from time to time. Much of the time this happens anyway – many people tackle Araucaria with the same enjoyment that they set about the staunchly Ximenean Azed puzzles – and long may this continue.