Puzzles regularly require knowledge from outside the stated dictionary. This is particularly true of barred thematic puzzles, where the theme will be unfamiliar to many solvers, who will need to do some research to corroborate details. (Let’s not, for now, get into the tricky issue of ensuring solvers unfamiliar with a theme are pointed in the right direction for research purposes.)
This may be as simple as noting that a key quotation is in the Oxford Dictionary of Quotations (ODQ), or that the thematic entries can be verified in Brewer’s Dictionary of Phrase and Fable (Brewer’s). One long-running series is especially fond of Brewer’s and understandably – it is an absolute treasure-house of lists and cross-references.
A key purpose is also consistency. While it is unlikely that anyone will fail to quote something like ‘Do not go gentle into that good night’ correctly, there are plenty of conflicting references to other quotes out there. One of the justifications for reference books is to act as a common source. It may be that a quotation has been reported multiple times in slightly different ways (see, e.g. Quote Investigator for a site that searches through variant versions) – the reference book’s is the one used in that particular puzzle.
[Aside: I once owned an edition of Bartlett’s Familiar Quotations which fell open at Conan Doyle one day, and there it was: “Elementary, my dear Watson”. I did toy with using that as a justification for including it in a puzzle – but I decided relatively few people would also own the 1910 edition.]
The question in these internet days is how to ensure that consistency. I have ODQ 4th edition, brought over from the UK fifteen years ago. Even then it was out-of-date, and I think the current edition is the 8th. My ODQ4 certainly did not have a quote used in a recent puzzle. Now, I have never seen the ODQ on sale in a New Zealand bookshop, and the recent work to strengthen the central Wellington library against earthquakes has seen its stock distributed around five or six locations. ODQ8 might have been in the central library; it might even be in the outpost set up near my place of work, but it requires a bit more work to verify that. Google is easier.
With Brewer’s the situation is similar – the latest edition seems to be the 19th, from 2013. I have the Millennium edition (no, published in 1999, actually) so I’m clearly going to be missing some things (I augment it with Brewer’s Twentieth Century Phrase and Fable). It was missing one of the entries in a recent Spectator puzzle (though I knew the answer, and the Twentieth Century volume came up trumps), but I expect the 2013 edition had it.
It is certainly possible to subscribe online, but the thing about these stalwart references is that once you have ODQ4 you have something like 80% of ODQ8, and so on. Does the additional material justify the subscription? With Google and Bing in the wings, it’s not so clear. On the other hand, my electronic edition of Chambers is updated as soon as a new version is available, because it’s used every day.
As Terry Pratchett’s introduction to the Millennium edition enthuses, the joy of Brewer’s is to open it to check something and to emerge some hours later having found 23 other things and forgotten what you went in for. That doesn’t seem to happen online, which is another count against the subscription. My paper copy has nearly everything and I can fossick. (Incidentally, isn’t it obvious from the Discworld books that Pratchett has swallowed at least one edition of Brewer’s and perhaps a Reader’s Handbook as well? I suspect J K Rowling is a devotee as well.)
The subscription question develops an additional dimension when people edge further afield. A recent puzzle referenced the Oxford Dictionary of Proverbs. I’ll see your Dictionary of Proverbs and raise you the Oxford Dictionary of Beliefs and Religions, which is sitting almost within reach. How about Chambers’ Dictionary of World History? Or Kobbé’s Complete Opera Book? One of these appearing as a source almost guarantees the use of Wikipedia. It’s certainly the case that these days when I’m developing a theme for a puzzle, I will repeatedly search for it online throughout the setting process, just to ensure it is accessible via multiple routes, and the sources thrown up by those searches remain consistent over a period of time. This is even if it has a clear reference book source – I no longer think it safe to assume people will have that.
Perhaps the time has come to find some way to indicate reliable web search locations. We have had a few puzzles over the years where a web address has been the ‘hidden message’, thereby ensuring solvers found a common resource. Maybe that needs to sneak into preambles.