Setting Things Right: The Manifezto Pt.3

Hi all, and welcome back to Setting Things Right, aka The Manifezto. The usual caveat: these are just my ideas and opinions – please do add your own thoughts and disagreements in the comments! Last time I introduced the idea of a clue as an illusion, like the famous Duck-Rabbit, with the Duck as its surface reading and the Rabbit as the cryptic reading. Now we’ll get into the detail of actually drawing that Rabbit.

In case you missed 'em:

• Part 1: Say What You Mean
• Part 2: The Duck-Rabbit

Part 3: Types of Rabbit

Give us a clue

There are lots of different ways to construct cryptic clues. Many ‘how-to’ guides are structured around different clue types. Even Ximenes provided a list of basic types, these being: (1) Two or more meanings; (2) Reversals; (3) Charades; (4) Container and contents; (5) Puns; (6) Anagrams; and, (7) Hidden.

I’m not convinced that these lists – or at least, any that I’ve come across – are quite right. Ximenes’ first example, “Two or more meanings”, now usually referred to as multiple definitions, certainly describes a particular type of clue. But others from his list – “reversals”, say, or “anagrams” don’t really describe a type of clue, just a type of wordplay that might be used within a clue. An anagram might appear within a container, perhaps even followed by a reversal ... or a substitution? And what of purely cryptic definitions?

So, my approach will be slightly different. I propose there are really just four main clue types:

1. Cryptic definitions
2. Multiple definitions
3. Hidden answers
4. Definition and wordplay

Whilst all are commonly employed, perhaps the most common is definition and wordplay, and this particular clue type is the one that then requires some further analysis to fully describe it – there are all sorts of possible elements that might be employed, and various ways to put those elements together. In fact, there’s so much involved that I’ll save this clue type for later posts. Instead, we’ll start with cryptic definitions: in my view, the ‘purest’ form of cryptic clue.

Cryptic definitions

A cryptic definition, to me, is the perfect encapsulation of the Duck-Rabbit: it appears to be one thing, but looked at in a different way it is clearly something else at the same time – with no further trickery getting in the way of that illusion. As such, I feel it deserves to be considered first – even though it’s tempting to leave it until later, as it’s often the trickiest to get right and feels quite ‘advanced’ when compared to a hidden word or an anagram, say. One reason a cryptic definition is tricky to get right is that it can be difficult to ensure you draw a Duck-Rabbit rather than a Duck-Goose – you are, after all, presenting the definition in plain sight, and it has to be recognisable. Coming up with a suitably deceptive form of words, that appears at first sight to refer to something completely different, is difficult!

• Maybe John Lennon’s Revolver? (8)

Here, the surface appears to reference the Beatles record, but if you know that Liverpool’s airport is called John Lennon, then a “revolver” belonging there could be the baggage CAROUSEL. I think that works as both a Duck and a Rabbit (though I’m not going to make any claim to be particularly good at cryptic definitions – they are difficult, which is why a really good one is so satisfying!)

• In which you might enjoy satisfying penny-drops among the lights? (9,6)

This is, I think, weaker as a cryptic definition – but still does the job? The enumeration – and the context – may make you think of CROSSWORD PUZZLE, but a few crossing letters will soon disabuse you of that notion. The answer is, instead, AMUSEMENT ARCADE. Now you could easily think of the intended answer first – and if you had a few crossers in place, you might not even think of the deceptive CROSSWORD PUZZLE at all – so the clue might well fall a little flat ... where’s the Duck? And it doesn’t stand alone in its own right – either answer would fit the clue, so you need crossing letters to decide which one is right, which some solvers find unfair. But as I’ve mentioned, it’s a big ask for the setter to produce 30 or so perfect illusions, and I think this does present two distinct images – which is what’s necessary. In the same puzzle, I had:

• Perhaps the last place you’d visit on holiday abroad? (7,8)

Here, the intention is to make you think of somewhere like North Korea, maybe, or the industrial suburb of a resort rather than the beach – the last place you’d want to visit. Of course, the answer is actually AIRPORT TERMINAL, literally the final place you might go to on your holiday. But just how ‘cryptic’ is that, really? There isn’t any particular distraction that might make solvers think there’s some wordplay involved (nothing suggesting an anagram or letter selection, say), the question mark emphasises that it's probably something a little ‘tricky’ like a cryptic definition, and the leap from “least likely” to “final” isn’t huge – so the solver’s experience may well be that it’s pretty much a straight definition. However, again I do think there’s still an alternative reading that presents a sufficiently ‘different’ scenario – so whilst it may be a particularly easy example of the clue type, I think it remains valid.

So, I guess there’s nothing really to add with regards to cryptic definitions beyond the description of what they are – and an acknowledgement that they are very, very hard to get right! They simply rely on the setter’s imagination and ingenuity, and they are often going to be a little Marmite as some solvers will see the Rabbit immediately and maybe miss the Duck altogether.

Multiple definitions

Another clue type without wordplay, a multiple definition simply presents two or more definitions of the same word. For example, this classic double definition from Spurius:

• Mad, passionate lovers (7)

BONKERS can be defined as both “mad” and “passionate lovers”, and those two bits go together nicely to make a coherent and deceptive surface – fabulous!

An issue that’s sometimes seen in multiple definitions is ‘same-sideyness’, where the two definitions are too close, like our earlier “Match replica (6)” for DOUBLE. Often, this is unfair on the solver because two words that mean the same thing are likely to have a number of possible further synonyms, and it’s not clear which of those is required. But even if there’s only one feasible synonym that fits the enumeration, the clue still fails because essentially, you’ve drawn a Duck-Goose – there’s no surprise you can make a word appear to look like others with pretty much the same meaning. A good way to check for same-sideyness is via the dictionary (Chambers of course being the recommended resource here) – if the two definitions appear under the same ‘headword’, with the same etymology, they’re likely to be same-sidey. Some nuances may have drifted far enough apart in everyday meaning to be acceptable, even though they share the same root, but more often than not a shared root is a red flag.

Link words in multi-definitions

Another common issue with multiple definitions is the use of link words. In a later post I’ll say much more about link words generally, but for now it’s worth noting that any ‘directional’ link should go from the wordplay to the definition. Why? If you think of the wordplay as a coherent picture or story, showing or telling you how you arrive at the definition – that is, it’s a proper Rabbit – then any link has to go the right way in order for the clue to make sense as a whole. Otherwise, you’re implying the answer, effectively the story’s end, is already known and instead telling the solver how to arrive at some weird way of presenting it – not a very satisfactory story! (Well, I guess that’s the sort of story the likes of Christopher Nolan could work with, but hopefully you see the point.)

But in a multiple definition, we don’t have any wordplay (you could say it’s simply a picture, not a story) – so there really shouldn’t be a directional link between the definitions. If you want a link word, it needs to suggest that the two (or more) things just happen to be the same, not that one somehow leads to the other. So non-directional words like “and” or “or” can do the job, but not “from”, “makes” or “for”, say. There are other links that are bi-directional – such as “is”, "in" (arguably) and “with” (even more arguably!) – that could be used either way in a wordplay & definition clue: “wordplay is definition” and “definition is wordplay” both make sense. But here, there is still that implication that one thing ‘provides you with’ the other – such words aren’t quite neutral, so I’d argue even these are usually best avoided for a satisfying multiple definition.

More than two definitions?

Most often, a multiple definition clue will contain just two definitions. Despite the vagaries of the English language, there aren’t that many words that lend themselves to more definitions without succumbing to same-sideyness. But triple, quadruple and even more definitions do crop up. Personally, I’m not usually a huge fan. Remember, the clue ought to provide exactly what you need and nothing more ... so bunging in an extra definition feels, to me, a less ‘pure’ way of constructing a clue. Perhaps even more so if link words are involved – whilst “Def1, Def2 and Def3” feels fine as a construction, “Def1 or Def2, Def3”, say, doesn’t have that same natural feel to it. If you can string multiple definitions together in a pleasingly coherent surface, then it certainly can be effective and is, without doubt, a clever thing to do – but unless that surface really zings, I find it’s a construction that’s more satisfying for the setter than the solver.

Multiple definitions combined with wordplay?

Occasionally you might find multiple definitions together with some elements of wordplay. Much like triple-or-more definitions, I find this unsatisfying as the clue no longer just gives you “exactly what you need and nothing more”. For me, if there’s wordplay involved, there really should then be just the one definition to hone in on – providing that extra definition might feel helpful to the setter (after all, it’s another element that provides further confirmation of the answer) but in my experience it’s more likely to be frustratingly distracting for the solver.

And that’s why, really, I think the multiple definition clue deserves to be treated as its own separate clue type. One or more of the definitions might well itself be cryptic, so these two clue types can be combined, but there’s no wordplay involved, which makes it – for me – a fundamentally different thing to a regular ‘wordplay & definition’ clue.

Hidden answers

Perhaps it’s surprising to treat hiddens, or lurkers as they’re sometimes known, as an individual clue type, rather than just as a wordplay element like an anagram (and they are, of course, very similar to clues using ‘container-and-content’ wordplay). Seeing as there’s no rules, it would be possible to construct clues using a lurker as just one element of the wordplay: a hidden anagram, say? But I find it difficult to imagine a situation where such a clue would remain both fair to the solver and stylish in a way that it doesn’t feel like a fudged convenience for the setter. And so, I view the hidden answer as fundamentally different, its own individual clue type.

One particular distinction I see in a hidden answer, as opposed to a ‘wordplay & definition’, is that the definition can often be given as an intrinsic part of the clue as a whole.

• Scrap radio talk shows (4)

Here, and in similar cases, you’re being asked to find the definition that ‘belongs to’ or ‘comes from’ the fodder. The definition is included as part of the clue as a whole; that is, the indication “radio talk shows” does not stand alone: radio talk shows ... what? Instead you need to read the whole clue as the description of what’s intended: we're looking for the particular "scrap" that "radIO TAlk" displays. (This sort of construction can work in other types of clue, but needs to be carefully and precisely worded to ensure it remains accurate, fair and stylish.) That doesn’t have to be the case:

• Scrap that radio talkshow's hosts (4)

Now, the definition, “scrap”, and the indication, “that radio talkshow's hosts”, can be read as distinct parts - the latter perhaps a little cheekily grammar-wise, with an elided comma after the "that" and the possessive "'s" just part of the string of fodder. Both constructions work fine, so it’s up to the setter to decide which provides the better surface.

Extra words and ends on show

An important point regarding lurkers is that the fodder words hiding the answer should all play their part. Adding any further words is simply ‘surface padding’, which is poor stylistically, unfair on the solver by providing red herrings, and inaccurate as the word is only hidden in those parts that contribute – CAT is not hidden in “respected publiC ATtorney”, only in “publiC ATtorney”.

Another important point here is that the answer should be completely hidden amongst the fodder – it shouldn’t appear at the start or end. When you have an indicator that suggests the answer is inside the fodder, it really must – as a point of accuracy – be fully within that set of letters. But sometimes the indicator isn’t quite so specific – for example, it’s not inaccurate to say that CAT is a “part of” CATCH. But the general idea remains the same – you’re looking for something hidden, and if some of it is openly on display, it’s less fun to find. In a game of hide-and-seek, why would you hide in a cupboard with the door open, or behind the sofa with your feet sticking out? So even if the indicator isn’t specific about hiding rather than just showing, I’d say the clue fails in term of style.

Reverse hiddens

Whilst I view lurkers as a particular clue type, they can still incorporate a little wordplay in the form of a reversal. As hiddens are generally relatively easy – with everything in plain sight, all together in the right order – they’re usually only seen once or twice in any puzzle. But combined with a reversal, you might be able to stretch to a couple more without unbalancing the puzzle overall. We’ll look at the detail of reversals when we consider that device amongst the many possible elements of wordplay next time.

More complex lurkers

And it’s not just reversals that can make your lurkers that little bit more tricky. Consider this classic clue from John Henderson (Io, Elgar, Enigmatist, Nimrod):

• The real reason for the merger meeting of Volkswagen and Daimler? (6,6)

Here, the fodder is given but not directly, instead using a juxtaposition – “the merger meeting of Volkswagen and Daimler” gives you the actual fodder VOLKSWAGENDAIMLER.

And what indication do we have that the answer is hidden in that? Well, in fact the answer isn’t entirely there, and instead we need to rely on a bit of reverse engineering. Within that fodder, we have a “hidden” word: VOLKSWAGENDAIMLER. The full answer spells this out: it’s a HIDDEN AGENDA.

And the definition? Well, the whole clue – with it’s concluding question mark – provides a possible example of a hidden agenda. It's an &lit ("... and literally ...") with the clue itself serving as definition. Genius! But very tough ... which isn’t perhaps what you’d expect from what is, essentially, a lurker.

Using juxtapositions in your hidden answer fodder – separating them with an “and”, say, or changing their order with an “after” – is a fun way to make lurkers a little trickier, but be cautious: many solvers do frown on this practice! It could be argued that the setter is introducing unnecessary layers of complication ... and thus moving closer to providing a riddle rather than a clue. As usual, the answer is not to adhere to any arbitrary rule, but to trust your judgement (and that of your test solvers too) of whether the clue remains accurate, fair and stylish.

(As for ‘reverse engineering’ and '&lits' ... well, more of those a little later. Or maybe quite a bit later, there’s the regular engineering and definitions to consider first!)

No rules?

By now, you might be thinking my original mantra of “no rules!” is pure poppycock. “Don’t use directional link words in multiple definitions.” “Don’t use multiple definitions with the same etymological root.” “Don’t use any words in the fodder for a hidden if they don’t contribute to the answer.” “Don’t leave either end of a ‘hidden’ word out in the open.” And so on. But I still contend that these aren’t rules as such – they are just examples of the consequences of following the core principles: accuracy, fairness and style. Directional link words in multiple definitions don’t really ‘say what you mean’ – they’re inaccurate. Multiple words with the same root are usually unfair on the solver, as they could resolve to numerous alternative synonyms, and lack style as there’s no surprise that one can be linked with the other. Extraneous words in lurkers are unfair, as they introduce red herrings, and inaccurate as you’re saying rather more than what you mean, whilst leaving a hidden word’s end on show lacks style because there’s no real hiding going on. If you like the clarity and comfort of a rules-based approach, by all means add these pointers to your (ever-expanding) list. But if you can find ways to break these supposed ‘rules’ that do genuinely adhere to the principles of accuracy, fairness and style then ... go for it!

Definition and wordplay

OK, so that leaves us with the final, and most common, of the main clue types: definition and wordplay. There’s a huge variety involved, so I’m saving the details for the next post, but as a brief overview ...

A definition and wordplay clue has various elements. One of these, of course, is the definition. The others could be words or letters in ‘plain sight’, synonyms, abbreviations, letter selections, letter mixtures, aural wordplay (homophones or Spoonerisms), often with further descriptions or instructions for how to manipulate them to construct the answer you’re after – with all sorts of indicators telling you when and how they are put together, moved around, reversed, deleted, put inside one another, anagrammed or substituted, say.

It might seem odd to lump all of these into just one clue type, but I think it does make more sense that way. An anagram is not itself a type of clue, just a type of wordplay within a clue. And there’s little point thinking of an anagram and a reversal and a homophone each as a separate clue type, because then you’d have to also have an almost endless list of clue types such as “charade of anagram with reversal”. So instead we have just the single clue type, within which there are many different ways to play with your words. And next time we'll start to get into the detail of all of those many wordplay elements, not just thinking about types of Rabbits, but closely examining their anatomy.

Til then, cheers!

Fez