Term Paper Features Crossword Puzzle

There’s an experience familiar to anyone for whom one puzzle a day is not enough. We particularly enjoy it at Crossword Blog: the Cluing Coincidence.

Recently, for example, Raich in the Independent had this clue ...

12ac Self-disciplined – sure, ate differently (7)
[Anagram of “sure ate”]

... on the same day that the Times cryptic had this one ...

4d Severe features relaxed — not feminine (7)
[Anagram of “features” without abbreviation for “feminine”][Anagram of EATURES]

... both of them asking for AUSTERE. As sober sorts, we rarely suspect the kind of skullduggery uncovered by Will Travers in the aborted TV thriller Rubicon, where keywords to a seditious plot appear simultaneously in the US’s posh papers.

In that fictional world, Will’s boss is circumspect:

I’ve seen this before. Crossword editors, quirky sense of humour. Probably some kind of inside joke.

In other words, a harmless conspiracy. But, for both of them, the coincidence is evidence of agency (as in intended action, not necessarily on the part of the CIA), whether in the name of pranking or murderous treason.

What are the chances, though, of such concurrences being merely happenstance?

The Times’s archive editor and feedback nabob Rose Wild has tackled this head on, in response to a reader who solves the Times2 quick crossword and that of another, unidentified, paper that he gets free at the gym. Not cryptics in this case, but the same thing is happening: “Over the course of a week or so, both puzzles include a number of the same answers.”

Wild asked Times2 crossword editor John Grimshaw (also feared by solvers of the Listener, the Enigmatic Variations and the Magpie as the puckish Dimitry) for an explanation, and it’s compelling:

Each puzzle selects about 25 words from, say, about 80,000 suitable for such a puzzle, leading to a probability of a particular word in one puzzle appearing in another of 25/80,000.

But, Grimshaw goes on, the odds of a coincidence are much higher in practice, because, as we saw here in some wonderful data journalism, some words appear much more often in crosswords than they do in what non-solvers might call “real life”.

There are lots of Is and Os knocking about in words, whether in puzzles or the written word, but puzzles need words to intersect with these letters. So, crossword English has far more words that end in I or O than does the normal version of the language. Grimshaw expands:

There are similar effects that probably mean the selection is very much biased towards some types of words cropping up more often, especially where there are few possibilities to match crossing letters in longer words that can’t be changed as easily.

For example, he says, a setter will sooner or later get to a point in a grid where he or she needs a word of the form __T__H, and so ITCH and ETCH come into play – or indeed UTAH, which alarmed MI5 quite so much when it appeared in the Telegraph puzzle in 1944.

This considerably changes the probability of coincidences, although it’s still low — but not so low that one wouldn’t expect them to occur on occasion.

And, of course, if a word is topical (as AUSTERE surely is), the chances are higher still.

To my mind, this is a feature of crosswords, rather than a bug. In his Chambers Crossword Manual, Don Manley (meet him here in our Q&A with Pasquale) gives further examples of “crosswordy” words. We should “learn to regard them”, he says, “as old friends”.

And, for the beginner, coincidences can be instructive. Looking at the same word clued in different ways is as good a way as any of getting to grips with the styles of different setters. The many shades of meaning in the English language, and the limitless possibilities of the cryptic toolkit, mean we can take many routes to the same destination. Here’s Hectence in a recent quiptic, the Guardian’s stepping-stone puzzle for those wanting to leap from quick to cryptic:

8ac Jane’s brief religious education is puritanical (7)
[Truncated surname of a “Jane” plus abbreviation for “religious education”]

So, as well as being topical and having great letters, AUSTERE is also a word with a range of meanings (self-disciplined, severe, puritanical) – and that’s before the world of politics adds its own nuance and spin.

By Stephen Battersby

TACKLING a crossword can crowd the tip of your tongue. You know that you know the answers to 3 down and 5 across, but the words just won’t come out. Then, when you’ve given up and moved on to another clue, comes blessed relief. The elusive answer suddenly occurs to you, crystal clear.

The processes leading to that flash of insight can illuminate many of the human mind’s curious characteristics. Crosswords can reflect the nature of intuition, hint at the way we retrieve words from our memory, and reveal a surprising connection between puzzle solving and our ability to recognise a human face.

“What’s fascinating about a crossword is that it involves many aspects of cognition that we normally study piecemeal, such as memory search and problem solving, all rolled into one ball,” says Raymond Nickerson, a psychologist at Tufts University in Medford, Massachusetts. In a paper published earlier this year, he brought profession and hobby together by analysing the mental processes of crossword solving (Psychonomic Bulletin and Review, vol 18, p 217).

1 across: “You stinker!” – audible cry that allegedly marked displacement activity (6)


Most of our mental machinations take place pre-consciously, with the results dropping into our conscious minds only after they have been decided elsewhere in the brain. Intuition plays a big role in solving a crossword, Nickerson observes. Indeed, sometimes your pre-conscious mind may be so quick that it produces the goods instantly.

At other times, you might need to take a more methodical approach and consider possible solutions one by one, perhaps listing synonyms of a word in the clue.

Even if your list doesn’t seem to make much sense, it might reflect the way your pre-conscious mind is homing in on the solution. Nickerson points to work in the 1990s by Peter Farvolden at the University of Toronto in Canada, who gave his subjects four-letter fragments of seven-letter target words (as may happen in some crossword layouts, especially in the US, where many words overlap). While his volunteers attempted to work out the target, they were asked to give any other word that occurred to them in the meantime. The words tended to be associated in meaning with the eventual answer, hinting that the pre-conscious mind solves a problem in steps.

Should your powers of deduction fail you, it may help to let your mind chew over the clue while your conscious attention is elsewhere. Studies back up our everyday experience that a period of incubation can lead you to the eventual “aha” moment. Don’t switch off entirely, though. For verbal problems, a break from the clue seems to be more fruitful if you occupy yourself with another task, such as drawing a picture or reading (Psychological Bulletin, vol 135, p 94).

So if 1 across has you flummoxed, you could leave it and take a nice bath, or better still read a novel. Or just move on to the next clue.

1 down: Sounds like… sounds like Umberto’s (6)

Pre-conscious processing is hidden from us, so it is not clear how the mind sifts through our mental lexicon to answer a clue. As written language is only a recent reflection of the long-evolved spoken word, Nickerson suspects that sounds are important. He illustrates this with a simple puzzle: quickly think of four-letter words ending in -any, -iny, -ony, -uny and -eny. When you’ve done it, read on.

You probably had little trouble with the first four, but may have struggled with the last one. Nickerson thinks that is because the only common word ending in -eny has a different pattern of stress from the natural way of reading the three-letter fragment. Research supports this idea, showing that a three-letter syllable forms a more effective clue than three other consecutive letters. So our mental dictionary is not just alphabetical, but also phonological. In which case, it may help to say the clue or your guesses out loud.

2 down: Nearly, nearly all, at the tips of many of solvers’ tongues (6)

When solving these puzzles, you might initially have a strong feeling about whether you know the answer or not – and these intimations are likely to be right. Given a mixture of solvable and unsolvable word association tests, subjects tend to guess correctly which ones they will and won’t be able to answer. In crosswords, says Nickerson, this “feeling of knowing” can be useful. If you are pretty sure you know the answer, you sensibly spend more time trying to get it; if you are certain that you don’t, you move on and try to get intersecting words instead.

Psychologists make a fine distinction between this feeling of knowing and a sense of something being “on the tip of the tongue”. The latter, more irritating state is the feeling that an answer will come soon, rather than that it will come eventually. It is often false, as the phantom of revelation fades away. One theory is that a wrong word retrieved from our memory blocks the way for the right word – a state that Nickerson recognises in crossword solving when an initial wrong guess makes it more difficult to find the true solution.

3 across: Choose from among various electronics (6)

Be careful with the more difficult puzzles – cryptic crosswords can warp the mind in surprising ways. Michael Lewis of Cardiff University in the UK came to this conclusion while investigating why results from police line-ups are so unreliable. He was following up research showing that face recognition can become temporarily impaired after a task known as the local Navon stimulus. The subject is presented with a large alphabetical letter made up from repetitions of a smaller letter, and is asked to read out the smaller letters while ignoring the larger one. This seemingly innocuous preparation made them much worse at a face recognition test.

“Be careful with the more difficult puzzles – cryptic crosswords can warp the mind in surprising ways”

Nobody is likely to perform this obscure task before they are called to pick out someone in a line-up, so Lewis decided to look at more common waiting room activities: sudoku puzzles, reading a book, literal crosswords and cryptic crosswords. Lewis thought the sudoku puzzles would have the biggest effect; the crosswords were only there as controls. But the subjects tackling the first three tasks all achieved roughly the same results in face recognition tests, whereas those wrestling with cryptic clues performed far worse (Perception, vol 35, p 1433).

Lewis speculates that some form of suppression may play a role. In the Navon task you must suppress the global picture, and in cryptic crosswords it helps to suppress larger linguistic units and break up phrases to look for hidden wordplay and definitions. As a side effect, that seems to suppress our ability to see a face as a whole unit. The phenomenon goes beyond visual and verbal realms – Navon stimuli also affect wine-tasting ability, says Lewis. “It suggests there is some overlap in processing between all these tasks.”

Crosswords naturally probe connections between ideas and words, and Nickerson suggests that psychologists could make more use of these puzzles when studying cognition. The human mind is itself a fiendish puzzle, so perhaps it’s not surprising that they cast light on its workings. Even if that light turns out to be oblique; aslant; indirect; elliptical…

Answers: 1A Eureka, 1D Echoes, 2D Almost, 3A Select

More on these topics:


Leave a Reply

Your email address will not be published. Required fields are marked *