Re(2): Cryptic Crossword Research - new publication IP: 22.214.171.124 Posted on May 6, 2016 at 09:49:23 PM by Kathryn Friedlander
Thanks for this, Don.
Yes, vocabulary and general knowledge certainly look important for American-style crosswords, and we did acknowledge (Discussion, page 17) that lexical ability was no doubt valuable for UK cryptics, too – but that it “does not appear to be a critical discriminator of high expertise among elite solvers”. We’re basing this on the very strong association between cryptic crossword solving and mathematics/computer programming in our survey – and this clearly increases significantly with expertise. Conversely, a background or interest in words or English language was least commonly found among our elite solvers.
As the article points out (page 4) “even if a word referenced by a cryptic clue is not known to the solver, it can often be deduced from the wordplay, and there are potentially two quite distinct avenues to the clue’s solution (Coffey, 1998): the crystallized route, tapping general knowledge and vocabulary to intuit the response, perhaps using cross-checking letters; and fluid intelligence which taps the ability to ‘derive logical solutions to novel problems’ … using clue components.”
Tricks of the trade, such as ‘tar=sailor’ and ‘leg=on’ are pieces of crystallised knowledge that solvers pick up as they gain experience with the crosswords. There are a great number of ‘Teach-Yourself’ guides to crosswordese which list words of this nature, and a solver doesn’t necessarily have to come to cryptic crosswords with a large store of these already at his finger-tips; but we would argue that they do need good problem-solving ability right from the start.
That said, we did study lexical breadth and word retrieval in some of our later research (see article, page 17), but unlike the Nottingham trials these were not our primary focus. We’ll be publishing these findings further along the line, so watch this space!