Australian Crosswords

11-across in this puzzle – CORPORATION OF OLD, POSSIBLY? (4,3) – is a cryptic definition, not a clue type I use a lot. I’ve talked about these before: it’s rare for such a clue to be solved without some or all of the cross letters because a solver has no idea this isn’t a “standard” two-part cryptic clue until all normal solving options have been exhausted. They do provide the most “aha” moments though, which is why some people really like them. This one is particularly difficult because I’ve used two irregular definitions in the clue. CORPORATION, meaning a pot belly, and OLD, referring to a style of beer.
So how do I judge how far to go, how obscure, how oblique, and what expectations do I have regarding the extent to which solvers will go to solve my puzzles? First I make a conscious effort to keep cryptic definitions to a minimum – you won’t see one in every puzzle. Next I limit my cryptic definitions to my own realm of knowledge, that is, I wouldn’t construct one that I couldn’t solve without checking references. (BTW, I’m no word-geek with a huge vocab, in fact, English was my worst subject at school – explains a lot, doesn’t it?). Of course, that doesn’t mean everyone will be able to do the same, and that leads me to my last point: solvers should expect to need to check at least a few things in major references, as I believe learning the odd thing from a cryptic crossword is worthwhile, and it’s not possible for solvers to remember everything about a word or term that they have encountered in their lifetime.
I often get asked about the use of references when solving cryptics. I use them when compiling, so it’s only fair solvers use them when solving.

I’ve been asked to write something about the use of linking words in clues, that is, words that either link sections of wordplay or, more typically, words that link wordplay and definition. I won’t be able to cover what could be a very technical and lengthy essay in just a couple of paragraphs, but I can provide a summary of my view.
My focus is always on the solver and I especially want to provide consistency to allow new solvers a firm platform to build on. As a result, I’ll only use Wordplay<->Definition linking words if I have to, and, when I do, I use them in a consistent way so that solvers can rely on my cluing. For example, I don’t subscribe to the concept that wordplay comes FROM a definition: it makes more sense from a solver’s point of view that the answer, defined by the definition, is determined FROM the wordplay. It’s true the setter wrote the wordplay from the answer, but how that’s relevant to the solver I don’t know.
If you solve my crosswords regularly you’ll know I prefer to construct clues in two distinct parts and I use definitional clues rather than instructional or descriptive types of clues. As a result, my Wordplay<->Definition link words tend to be “is”, “may be”, “in” etc, reinforcing the separate two-part nature of good cryptic clues. Having said that, I understand why setters want and exercise some licence when it comes to linking words. The more word ammunition available, the easier it is to write clues and more wordplay variety is available. It takes a lot of time to massage a clue idea into the best clue it can be, and sometimes that time isn’t available.
As for linking words in the wordplay, I’m sure my philosophy is different from most setters. That’s a story for another time.