Setting Things Right: The Manifezto Pt.7

18 min read · Wed 22 Jul 2026

Hi all, after a bit of a hiatus 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!

Part 7: Plonkers

Well, after a few lengthy blogs on 'definition and wordplay' clues, you might think the subject exhausted. Far from it! In this instalment, I was hoping to wrap up a few loose ends, including thoughts on punctuation and capitalisation, as well as looking at &lit and reverse-engineered clues. But it turns out that the first issue I wanted to address provides more than enough material for a single blog post ... so I'm delaying those 'loose ends' til next time, and instead will focus here on plonking.

What's plonking?

We're all always learning. It's only relatively recently that I've come across the concept of plonking, and it's one that has certainly influenced how I think about setting. In previous posts, I've mentioned the idea of 'active' vs 'passive' wordplay, and how, to me at least, mixing those different 'modes' often feels jarring and so, at best, sub-optimal. The 'plonking' concept firms up that instinctive feeling, and on reflection provides a stronger argument that mixing modes is perhaps not just jarringly sub-optimal, but incorrect – grammatically inaccurate.

So, what's plonking all about? It's perhaps best understood by considering a simple charade. In a charade, the various elements are just presented – plonked – side by side, so the 'wordplay' doesn't depend on any grammatical construction. It's not explicitly ungrammatical, but can maybe be thought of as non-grammatical.

Grammatically, though, a series of plonked bits of wordplay can be read as 'Element 1' and 'Element 2' and 'Element 3', and so on. And it's in this implied grammar where problems might arise.

The following example from a reader at the Clue Clinic illustrates the point:

Dad turns mean and yells, occasionally in costume (7)

Here we have three bits of wordplay plonked together (albeit with an explicit "and" involved): a reversal of PA ("dad turns"), PAR ("mean" ... perhaps a bit of a stretch), and EL ("yElLs, occasionally"), with that wordplay "in" the definition APPAREL, "costume".

The query here was actually to do with that use of "occasionally" (an issue looked at in this blog way back in Part 1), but the excellent Dr. Clue of that parish highlighted another possible problem:

"'Dad turns’ on its own can produce AP, but this can’t be part of a continuous series of elements [...] each element of a charade clue like this must yield a noun expression. [...] ‘Dad turns’ needs to be ‘Dad turning’."

That is, imagining a full complement of ands, you have "Dad turns and mean and yells occasionally". And that clearly doesn't make sense grammatically, because the "dad turns" is a verbal phrase, not a 'thing' that can appear in a list of 'things'. "Dad turning" is a noun expression, a 'thing', so that can appear in a list of 'things'.

Dr. Clue, thorough as ever, provides further examples, here considering possible wordplay for APEX:

"Dad turns on former partner" is fine; although 'Dad turns' is an active verbal phrase, it is used with the juxtaposition indicator ‘on’ (rather than just being plonked in a list) so the expression makes sense as a whole, showing in what way Dad turns. It forms a single, coherent (though complex) 'active' wordplay unit.
"Dad backs former partner", however, gives one 'active' and one 'passive' wordplay unit simply plonked together. Reading it as "Dad backs and former partner" doesn't make sense. That first wordplay unit should be "Dad backing" so that it is a 'thing', a passive noun expression, that can be part of the charade.

How about something a tad more complex than a simple charade? Dr. Clue doesn't disappoint there either, with a further example:

Chap accepts independent European state (5)

Here we have MAN around I ("chap accepts independent") plonked next to E ("European"), to give MAINE. Put in the implied 'and', and you have "Chap accepts independent and European" ... and if read as a list (A and B) that first bit (A = "chap accepts independent") isn't a 'thing'. For the grammar to work, the verb "accepts" has to work on the whole plonked list "independent and European" as its object (not just "independent"), so really the clue leads to MA(IE)N. That is, "A accepts B and C" can only mean "A accepts (B and C)", not "(A accepts B) and C".

Noun expressions

Note that the objection highlighted by Dr. Clue is that the plonked wordplay units have to be noun expressions, or what I've previously referred to as 'passive' bits of wordplay. But pinning down what's meant by a 'noun expression' perhaps needs a little more thought. Consider a couple of further examples for MAINE:

Important European state (5)
Mother in European state (5)

A couple more simple charades, MAIN+E and MA+IN+E. But in the first case, there's an adjective ("important") included, and a preposition ("in") in the second. Does "Important and European" make grammatical sense as a list of 'things'? "Mother and in and European"?

Well, the difference here is that both IMPORTANT and IN are nouns! In the cryptic reading, "important" isn't an adjective, but a token representing the set of letters MAIN: it's a noun expression, "[the set of letters from] something meaning important". And "in", in the cryptic reading, isn't a preposition – it's a bit of fodder, again a noun expression: "the set of letters IN". (And, of course, that "European" has all along been a noun rather than an adjective too, in the cryptic reading – simply a token representing E.) It's only in the cryptic reading that those elements have to be noun expressions, and individual tokens and fodder are always nounal in the cryptic reading.

In "Chap accepts independent", it's the "chap" and "independent" that are tokens (resolving to MAN and I), but the whole wordplay unit is not a noun expression – the "accepts" makes it verbal; it describes the action of what those tokens are doing. In contrast, "Chap accepting independent", although using a verb, can be read as a noun expression because it describes the result of the action. Compare "This is a chap accepting an award" with "This is a chap accepts an award" and hopefully the distinction becomes clear.

But is plonking really an issue?

Now, let's take a step back. There's clearly a compelling case to carefully consider grammar when plonking. But right at the start, I mentioned that a series of plonked elements is essentially non-grammatical, rather than ungrammatical. So is it fair to then impose grammatical restraints, insisting on accuracy in accordance with what is only implied grammar? As a setter you have an obligation to 'say what you mean', but in a plonked list you're not quite saying "Element 1 and Element 2 and Element 3..." ... otherwise, surely you should have put in those commas and ands?

Think about that original clue for MAINE. It gives a sort of visual representation of the word, split into different elements: the sequence "Chap accepts independent" / "European" corresponds exactly to what you see, "MA(I)N" / "E". It doesn't rely on a grammatical construction.

Until recently, my minor objection here wasn't so much that non-'things' shouldn't appear in a charade, but simply that the first part ("Chap accepts independent") is 'active', whilst the second ("European") is 'passive' – and it's that clash of 'modes' that is slightly jarring. I guess that amounts to pretty much the same objection? The nuance is that I viewed the clue as sub-optimal due to the clashing modes, rather than incorrect due to poor cryptic grammar. Not the most elegant construction, but the sort of compromise that I'd have found within the bounds of reason.

And I wasn't simply disregarding my repeated insistence on respecting cryptic grammar – I felt there was a coherent justification. As explored above, that MAINE clue certainly isn't, as a whole, grammatically correct – but then again, it isn't pretending to be. The setter hasn't included any explicit juxtapositions or otherwise tried to make that first element do anything other than show what it says; it's not being moved around or reversed or put inside anything, for example. So is the construction really inaccurate or unfair?

Well, perhaps unfortunately (as it likely invalidates many clues I've happily written in the past!), after a lot of thought I've concluded that it really is inaccurate. I no longer accept that 'visual representation' argument, because it splits the wordplay without any regard for grammar, essentially presenting as wordplay not a recognisable Duck but a series of randomly plonked animals. That spoils the illusion – you want to 'see' two distinct different things at once, not a random collection: a Duck-Rabbit, not a DuckMooseTiger-Rabbit.

So, plonked elements in wordplay do indeed need to respect cryptic grammar. But maybe that then leads to a further possible issue with plonking. It involves what I'm going to call the ultimate plonker.

The ultimate plonker?

Remember the full MAINE clue reads:

Chap accepts independent European state (5)

There's another element in the clue: the definition, "state". And that is just plonked there, too. In the revised, 'pure' version (with "accepting" rather than "accepts"), you could insert the implied ands and have a grammatically correct clue: "MAN accepting I and E and the definition". But a slightly different clue, where "accepts" would perhaps be acceptable, might be:

Chap accepts independent chief (4)

I.e., MAN "accepts" I / definition: "chief" = MAIN. That first (and only) wordplay unit is now an active verbal phrase, "Chap accepts independent", not a 'thing', a noun expression. Yet it's simply plonked next to the definition as if in a list. If we're insisting on complete grammatical accuracy, surely there needs to be some sort of link to avoid that inaccurate plonking of the definition? "Chap accepts independent to become chief"?

When it comes to plonking, maybe the most frequent case is the one we've just seen: where a definition is provided without any link words, instead just being plonked next to the wordplay. It's a point raised by Ximenes in relation to 'say what you mean', the key principle put forward by his mentor Afrit (and sometimes known as Afrit's injunction). Here he refers to the clue "Rending granite all to pieces (7)" (TEARING; defined as "rending", clued as an anagram):

"If I am really saying what I mean, oughtn’t I to put a dash or a colon between “rending” and “granite” to indicate the two parts, as a careful writer would? Now here I’m in a weak position, because I believe Afrit would have said “Yes, you ought.” It spoils the misleading sense, and the clue would have to be reworded; but I think he, as a purist, would have insisted" (Ximenes on the Art of the Crossword, in Chapter 5, "Cluemanship")

Insisting on "a dash or a colon" (or, a grammatically acceptable link word) to separate definition and wordplay certainly feels too restrictive. I'm with Ximenes, and I think the vast majority of setters, on this.

The difference in Ximenes' example is that his "granite all to pieces" can be read as a noun expression, so the expanded clue "[A word meaning 'rending'] and [the set of letters GRANITE, anagrammed]" still makes grammatical sense. With "Chap accepts independent chief" we have "[The set of letters from a word meaning 'chap'] accepts [a letter that represents 'independent']" which, as a verbal expression, can't then be associated with the definition "[a word meaning 'chief']" by means of simply inserting an "and".

So in "Chap accepts independent chief (4)", does the clue work as it is, or does it need something to avoid the definition being ungrammatically plonked next to the verbal wordplay? And if the definition can be casually plonked alongside something more 'active', why not other parts of the clue ... providing they are doing nothing other than giving a 'what you see is what you get' representation? On the one hand it's argued that 'plonked' elements should follow grammatical rules, but on the other it seems a definition can be 'plonked' at the start or end of a clue with impunity. Is that inconsistent? As a whole, you're essentially providing a 'visual representation' ... and if that's good enough for the definition, it should be justifiable for other bits of the clue.

Or so I thought.

But the key here is that the wordplay and the definition are two very different things, doing separate jobs within the clue. It's not a case of random exceptionalism to treat the definition differently, because the definition really is a genuine 'special case'.

And this is implicitly covered in that quote from Ximenes above. He notes that the definition strictly ought to be separated from the wordplay by "a dash or a colon". He doesn't say that it needs some implied grammatically correct 'link' such as an "and". Instead, the definition serves as a stand-alone 'caption'. "Chap accepts independent: chief". The wordplay remains as a grammatically correct unit; the definition provides the caption that clarifies what you're looking at. The elided colon works to separate the two – and it wouldn't work within the wordplay, because we don't want to separate out the wordplay elements as "Duck : Moose : Tiger" – that Duck needs to be a single coherent Duck.

So, the 'ultimate plonker' really isn't a plonker after all. (Rodney Trotter is vindicated!)

Captioning

Just to finish off this discussion of apparently 'plonked' definitions (and really just as an excuse to highlight a great clue!), consider the following recent beauty from Fed in The Guardian:

David Beckham makes one eat dessert first (10)

Tough? Let's add in the 'elided colon' so that the clue's definition, its 'caption', is clear. And also add in an 'elided comma' to clarify the wordplay:

David Beckham makes one: eat, dessert first (10)

Now the wordplay gives us HAVE ("eat") with AFTERS ("dessert") appearing before it ("first"). Which gives us AFTERSHAVE ... a product ol' Golden Balls does indeed make.

There's no problem that the definition appears to be a verbal phrase plonked next to some passive wordplay, because this is a case of captioning rather than plonking. Plonking is only a wordplay phenomenon.

Before signing off, there's still a bit more detail related to plonking. Beyond presenting it on its own as a simple charade, what can you fairly 'do' with a plonked series of wordplay units?

Indicator 'scope'

So far, we've mostly looked at plonked charades in isolation. But often you'll see an indicator of some sort applied to multiple elements that are plonked together.

We have actually seen this in the discussion above, with the wordplay "Chap accepts independent European" being interpreted as "[MAN] accepts [I and E]" to get MAIEN – the plonked list is being used as fodder, the 'thing' that's accepted by MAN. (Note that if you're going to be doing something with the plonked list, it really does need to be a noun expression – and so it can only be made up as a series of noun expressions. Even that dodgy 'visual representation' excuse can't be wheeled out here!) So, when can indicators apply to multiple plonked elements?

For example, in "Primarily, you are weary, naturally showing signs of tiredness (5)" the indicator "primarily" tells you to take the first letters of all of "You Are Weary Naturally Showing" to get YAWNS, "signs of tiredness".

But that wouldn't work as, say, "Beginning to yield and wane, naturally showing signs of tiredness (5)", because the "beginning to" can only really apply to the first word in the plonked list. The scope of the two indicators is different.

Often (as in the YAWNS examples above, I hope) the indicator's scope is naturally obvious, but that's not always the case. Consider:

Refutable avowal turned out to be true (4)

That gives us RefutablE and AvowaL plonked together and both "turned out" (i.e., emptied), "to be" the answer REAL ("true"). But does "turned out" naturally apply to both words, or is it only really applicable to the immediately adjacent "avowal"?

Or how about this one from PostMark:

Characters central to Jaguar poached racing designer (5)

We're looking for the characters that are central to jaGUar, poaChed and raCIng, to get the designer GUCCI. But does "characters central to" naturally apply to that whole plonked list?

In both cases, I think the intention is clear, and that goes a long way. But mere intention isn't enough: "first light" for L is arguably clear but, I hope you'll agree, also incorrect. These examples are a little more subtle. Both would be clearer if the fodder list wasn't plonked together, but explicitly presented as a list: "Refutable and avowal turned out ...", say, helps to clarify the scope. Even just a comma, "Refutable, avowal", makes things a tad clearer. And of course in crosswords, solvers can expect a little 'telegramese' economy, elided commas and the like, so perhaps then omitting that explicit list indicator should be OK, just as in a simple plonked charade? Inserting the implied ands (and separating out the definition 'caption') gives us:

Characters central to Jaguar and poached and racing: designer (5)

Which I think clearly works to indicate all of those central characters.

The clue was submitted for a MyCrossword 'Featured' puzzle, and a 'hidden motive' of that series is to show that an insistence on accuracy isn't a barrier to fun! (Sometimes, less-than-accurate clueing is 'justified' by saying it allows more freedom and more fun; I strongly disagree! If that Rabbit isn't really accurate, I think it's inherently less fun.) Now in hindsight, and having had the opportunity (through writing this blog) to think a little more deeply about it, I'd say the clue works nicely. But at the time I wasn't really sure, and as editor I did – some might say, rather curmudgeonly! – ask for reconsideration. The update from PostMark:

Characters central to Jaguar, racer and racing designer? (5)

By making the fodder explicitly a list, using punctuation and "and", I think the scope of "characters central to" is now clearer. And surface-wise, I don't think the clue is any weaker than the original, at least not enough to prefer the 'plonked' version over the re-write. The original was, of course, already good; the re-write I think could be argued to just surpass it because it more clearly says what it means. Where the fodder words are simply plonked together and the solver is left to interpret the indicator as applying to all of them, maybe it's not quite as fair on the solver? Even the casual-sounding term 'plonking' maybe suggests something inherently sub-optimal.

I tried to look into the niceties of the grammar here but came to something of a dead end. As a general rule of thumb, adjectives and noun modifiers (like "start of") tend to be 'sticky', with local scope only applying to the immediately adjacent object. And adverbs (like "initially") tend to be 'floating', with global scope, framing everything to describe the overall status of the statement. But although in everyday language adjectives tend to be 'sticky', in crosswords we take advantage of the possible ambiguity and use adjectives all the time to refer to multiple elements:

Oddball Alice might like Stonehenge (10)

Here, the adjective "oddball" applies not just to the first element ALICE, but also to the plonked MIGHT (giving MEGALITHIC, "like Stonehenge"). So, with adjectives (and I'm sure other examples too), the scope is often uncertain. "Red cars, vans and lorries" might mean "Red [cars and vans and lorries]". Or it might mean "Red cars, and [any colour] vans and [any colour] lorries". That latter interpretation is perhaps just a tad more natural, reflecting the adjective's tendency to stickiness. But in the 'plonked' version, "Red cars vans lorries", it could be argued that omitting the commas and ands actually helps to clarify the intended scope. That "cars vans lorries" is perhaps more easily interpreted as a single unit, with the "red" applying to the whole group, than "cars, vans and lorries", where the comma and "and" more clearly separate them out as individual things rather than a unified group? So it can work both ways: on the one hand a 'plonked' construction might help to clarify the intention, on the other it could be that mentally inserting the ands might make things clearer. (I'll admit that here, I'm back to relying on what feels right to me ... I'm sure a genuine grammar expert could provide a better line of reasoning. But everything I know about grammar has been learned from crosswords!)

So maybe the most reliable option here is, as might be expected, to follow the esteemed Dr. Clue's advice: simply mentally insert those ands into the plonked list (remembering that each token and bit of fodder is a noun expression), and then check that your intended meaning is naturally expressed. (And for completeness, then take them back out and judge whether that 'plonked' version is justified – perhaps even helping to clarify the intended scope – or whether you're actually being fairer to solvers by 'unplonking' the list.)

Just as I, perhaps, should've done with PostMark's original GUCCI – mea culpa!

Summary

Plonking puts wordplay units together side-by-side, with an implication that they're joined together (by a series of ands, say)
As this produces a list of 'things', each plonked element should be a noun expression for the cryptic grammar to work
Plonking is an issue in wordplay; where a definition appears to be 'plonked' alongside the wordplay, that's fair captioning
So we have, for example:
- Chap accepts independent European state (5) ❌ (to remain accurate, you can't plonk the verbal phrase "MAN accepts I" next to the noun expression "European")
- Chap accepts independent chief (4) ✔️ (you can fairly provide your definition as a 'caption' alongside the wordplay, even if that wordplay isn't a noun expression)
Where a set of plonked elements is playing a role beyond just a simple charade – where some sort of indicator is involved – you can check that the indicator's scope encompasses the whole plonked list by mentally inserting those implied ands

It's only through writing the blog that I've really been forced into thinking more deeply about 'plonking'. I've always maintained that mixing 'active' and 'passive' elements can be jarring, and thus 'sub-optimal'. But thinking about things through the lens of 'plonking', and about the grammar surrounding 'verbal phrases' vs. 'noun expressions', that instinctive idea has been given a firmer foundation.

As mentioned, the rounding up of loose ends – including punctuation and capitalisation, and &lit and reverse-engineered clues – will be the focus of the next instalment. But even then, I won't pretend that I've been able to cover everything ... as has been re-inforced for me through writing the blog, there's always more to learn!

Last time, I also mentioned that "if time permits", we'd re-visit the 'Ximenean vs Libertarian' debate, thinking about how 'libertarian' you can be without straying into 'inaccurate' or 'unfair' territory ... and how 'Ximenean' you can be without stifling fun and creativity. Well, clearly time doesn't permit, and it probably won't next time either! So instead I'll make that the focus of the next next instalment!

Til next time, cheers!

Fez

0 comments

Be the first to comment!