Intermittent indicators tell the solver to select letters at regular intervals from a word or series of words within the clue. So in the clue “Notice odd bits of swede (3)” the odd-numbered letters from the word SWEDE must be selected, producing SEE, and in the clue “Canny golfer regularly cut dogleg (5)” alternate letters in CANNY GOLFER must be deleted to produce ANGLE.
With indicators such as ‘regularly’, the solver must determine both the first letter to be selected and the sampling frequency – typically every second letter is to be extracted, but every third or even fourth letter could legitimately be required. Many setters take the view that at least three letters in total must be selected when an indicator such as ‘regularly’ is used, as just about any combination of two letters could be considered ‘regular’.
The indicators can dictate either selection or deletion of letters. With a ‘regular’ selection indicator, the effect for the solver is unaffected, so ‘regularly’ and ‘regularly cut’ come to the same thing. With ‘even’ and ‘odd’ indicators, the effect of deletion is the opposite of selection, so ‘odd pieces from author’ gives ATO but ‘author ignoring odds’ gives UHR (ie the even-numbered letters).
There are very few intermittent selection indicators which stand up to close scrutiny, but some are certainly better than others. In this list, indicators are rated A or B for validity. Those in first group I would consider relatively sound, while those in the second group are ones which I have seen used in puzzles but would consider highly questionable. Indicators which involved the words ‘evens’ and ‘odds’ have been downgraded to ‘B’ status because although ‘evens’ and ‘odds’ are nouns, they have nothing to do with even or odd numbers.
The list below can be sorted alphabetically on Indicator (default), Selection or Validity, in either ascending (default) or descending sequence. The Search box allows full and partial searching of the first two columns in the list.
Date last modified: 10/09/25
Recent changes: (10/09/25) added prime selection from/prime bits of, downgraded expressions containing evens and odds to ‘B’ status.
| Indicator | Alternative form(s) | Selection | Validity |
|---|---|---|---|
| alternate bits of | alternate parts of | alternate letters | A |
| alternately | alternate letters | A | |
| at intervals | regular letters | A | |
| dropping every second | wanting seconds, seconds out | odd letters | A |
| even bits of | even letters from, even pieces of | even letters | A |
| evenly | even letters | B | |
| evens out | odd letters | B | |
| every other piece | leaving every other piece | alternate letters | A |
| every second | even letters | A | |
| every so often | regular letters | A | |
| from time to time | regular letters | B | |
| ignoring the odds | even letters | B | |
| intermittently | regular letters | A | |
| now and again | every now and again | regular letters | B |
| now and then | every now and then | regular letters | B |
| occasionally | regular letters | B | |
| odd bits of | odd letters from, odd pieces of | odd letters | A |
| oddly | oddly chosen | odd letters | B |
| odds of | odd letters | B | |
| odds cut | even letters | B | |
| on and off | off and on | regular letters | A |
| periodically | regular letters | A | |
| prime selection from | prime bits of | letters 2,3,5,7 etc | A |
| regular letters from | regular letters | A | |
| regularly | regularly ignored | regular letters | A |
| regulars in | alternate letters | B |
How about “Dad turns mean and yells, occasionally in costume (7)” for APPAREL (i.e. AP (<pa) + PAR + EL)? Grateful for feedback please
Hi Neil
‘Occasionally’ isn’t one of my favourite intermittent selection indicators because it doesn’t suggest regularity, but it often appears in published puzzles (as do the other ‘B grade’ indicators), so ‘yells, occasionally’ is fine for EL (or YLS).
However, the clue as it stands does have a problem – ‘Dad turns’ on its own can produce AP, but this can’t be part of a continuous series of elements. It’s a point that many setters overlook, but each element of a charade clue like this must yield a noun expression. In this regard, ‘mean’ and ‘yells, occasionally’ are fine, but ‘Dad turns’ needs to be ‘Dad turning’. When elements appear one after another with no indicator between them, imagining an ‘and’ between each one will make any problem with the cryptic grammar clearer. So ‘Dad turns on former partner’ is ok as a wordplay for APEX because of the juxtaposition indicator ‘on’, but ‘Dad backs former partner’ (=’Dad backs and former partner’) is no good (it needs to be ‘Dad backing’). I hope that makes sense, but let me know if not!
Nice one, thanks. Very new to this so that’s extremely helpful
Revealing myself to be slow of wit: is the reason why “Dad backing” yields the necessary noun expression and “Dad backs” does not? I am not as much the ‘plonker’ as I fear, but still grapple with what seem to me to be the finest of lines between acceptable and unacceptable indicators.
Hi Dr Daniel
I think that the issue perhaps becomes clearer in a clue like “Chap accepts independent European state”, intended to give the answer MAINE ((MAN around I) + E). The main verb ‘accepts’, though, must take as its object ‘independent European’, yielding MAIEN or the like. Or “Better referee holds Northern Ireland back”, meant to result in FINER – only the NI can be reversed here.
My question was garbled, but you nevertheless interpreted it correctly. I remain puzzled, however: are you stating that “accepting” or “holding” apply only to the next word, while “accepts” and “hokds” would not?
What I’m saying is that in crosswordese ‘A holding B C’ can mean ‘(A around B) + C’ or ‘A around (B + C)’, but ‘A holds B C’ can only mean ‘A around (B + C)’.
THANK you. You have invoked “crosswordese” as the reason, and “crosswordese” is a legitimate reason. It is a relief to know that it I am not misunderstanding the rules of ordinary grammar.
Stagger is intriguing. Presumably one could use well chosen link words to get something from “staggering” and “stagger”…?
e.g.
Staggering kitchen produces gelato = ICE
Stagger guardsman exposing body of stars = URSA
Both a little clunky but you get the idea!
I do, and I think a case could be made for both those clues being valid.
Great – thanks! This led me to wonder about similar constructions with the verb “alternate”, and I did just find an example on fifteensquared. (There may well be others, but searching for “alternate” brings up countless “regularly” clues…)
Touching and crude, alternating when required = ON CUE (Stamp, Independent 12,037)
I suppose that “alternate”, used cryptically as an imperative verb, disguised as an adjective in a clue’s surface, could be quite effective
I’m a little surprised to not see “frequently” given even a B here (despite the word “frequency” appearing on this page!). It seems about as justifiable as “periodically” to me, unless I am missing something silly…?
Hi Supware, and thanks for the question
The problem with ‘frequently’ is that – unlike ‘periodically’ – it carries no suggestion of regularity; ‘periodically’ means ‘at regularly recurring or definite intervals’, but ‘frequently’ means ‘at frequent or brief intervals’, with no indication that those intervals are equal. Without such indication, there is no reason why the solver should not select (say) the first, third, fourth and sixth letters of the operand (that seems to satisfy the requirement of ‘frequently’). I think one could argue that it was no worse than the B-rated ‘occasionally’ (meaning ‘now and then’, itself another ‘B’). However, George Ho’s database tells us that ‘occasionally’ is an occasional, albeit not frequent, visitor to blocked puzzles, whilst the database contains only one example of ‘frequently’ being used in a newspaper puzzle (the FT in March 2020). On that basis I’m not inclined to add it (the likely implication of its absence being that most editors will not allow it), but I could be persuaded to add it as a ‘B’ if it were to receive the support of other readers.
‘Primes’ to indicate positions 2,3,5,7 etc
Hi Anon (or may I call you Al?)
Thanks for the suggestion. My first thought was that I wouldn’t be keen on ‘primes of’ but that I would be happy with ‘prime selection from’ to indicate the second, third, fifth etc letters of a word.
However, that prompted me to consider indicators such as ‘odds of’ and ‘evens out’. When ‘last of’ (for instance) is used as a letter selection indicator, it is a pronoun, and ‘lasts of’ to select multiple final letters makes no sense. ‘Odds’ and ‘evens’ cannot be adjectives, nor can they be pronouns, so they must be nouns. Would we accept ‘odd of me’ for M? If not, then we shouldn’t allow ‘odds of success’ for SCES or ‘luck evens out’ for LC. The existence of the nouns ‘odds’ and ‘evens’ might seem to make these constructions feel ‘natural’, but these nouns have very specific meanings – as letter selection indicators, they are actually no different from ‘earliests’ or ‘ultimates’.
If ‘odds of’ is ok, then ‘primes of’ is most certainly acceptable. I feel an update to that list coming on, but I’m not certain what form it’s going to take.
…I have added ‘prime selection of’ and ‘prime bits of’ as ‘A listers’, and downgraded indicators containing ‘evens’ and ‘odds’ to ‘B’ status.
On this subject see The Times today (2025-09-22), Cryptic No. 29341:
24d: Select prime locations in Bearsden for facility (4)
The answer is EASE, revealed by taking the 2nd, 3rd, 5th, and 7th letters, i.e. the ‘prime locations’, not alternate letters.
Thanks, Matthew
I came across a very similar clue the other day in a puzzle which hasn’t yet been published. It seems to be something that setters have started picking up on as a novelty, and I fear that overuse may quickly see that novelty wear off, as it did with (say) ‘bit of a kip’ for AT. I’m not unhappy with ‘prime locations’ or ‘prime positions’, particularly when combined with a verb, as in your example.
I wonder how long it will be before we see something like “Isolated squares in Limoges and Poitiers” for LONE. Or “Rock Fibonacci’s taken from olid teledu” for OOLITE…
Or how about “Cool cubic places found in insipid correspondence courses (3)” for ICE? 🙂
Now that’s just silly 😉
As far as I know, the first use of the ‘prime’ device was in the clue “Prime parts of iconoclasm are fashionable” (COOL), namely 6ac in Times 23105, which appeared on Oct 8th 2005 (i.e. 20 years ago next week). The editor at the time said (to me, because the puzzle was mine 😉) he’d never before seen the device, which elicited comments in online platforms. I’ve used (but not overused) it a few times since then, but it does seem to have caught on.
Thanks, Monk
How interesting – and of course your original clue was, as we would say here at the Clinic, ultrasound. It’s sometimes possible to identify the first appearance of a particular type of puzzle (eg Eightsome Reels), but being able to pin down the debut of a clueing device is something else! Not that I solve a lot of puzzles, but I don’t think I’d seen the device at all until a few months ago, since when I have encountered it several times. I happen to know that it will be making another appearance this Sunday, just three days before it turns 20 🎂. I’m still not sure about ‘primes of’, but I don’t know how one would use it effectively in a clue anyway.
Thanks Dr Clue, not least for your sterling efforts in updating this site so meticulously. FWIW, I [too] doubt that “odds of” or “primes of” respectively imply the intended “odd [bit,part,etc]s of” and “prime [bit,part,etc]s of”. Faced with a potential ambiguity or lack of precision (generally invited by focussing on the SR at the expense of the CR), it surely behoves the setter to prioritise [necessary] accuracy in the CR and, if that damages the [desired] SR, to back out and consider another tack.
Thank you!
I agree entirely. I must admit that I had got it into my head that (say) ‘ignoring the odds’ was reasonably sound, but it’s really no better than something like ‘missing preliminaries’ to indicate the loss of multiple first letters.
Incidentally, the word ‘meticulous’ always puts me in mind of a job application I received many years ago which contained the assertion “I am maticulate in my attention to detail”.
😆 But can you please be a bit more pacific about what the applicant was inferring? [A panellist made the second error on QT last night, and another said that a word was “synopsis” with another 😯 . What is the whirled coming two?
What about staggers/staggered/staggering? One definition is to set, arrange, or incline alternately, for example, “The tiles were staggered for a pleasing effect.” I’ve used this myself, but never seen it elsewhere.
Hi Darren, and thanks for the suggestion.
One could argue that very few of the intermittent selection indicators stand up to even moderate scrutiny, so there’s certainly some leeway available, particularly since unlike, say, anagram indicators, there aren’t many of them available. Chambers gives a transitive sense of ‘stagger’ as ‘to dispose alternately’, which sounds quite promising; there is no similar intransitive sense, so ‘staggers’ and ‘staggering’ have less potential.
The problem I see is that while tiles certainly can be staggered, this means each individual tile going one way or the other. If we extend that to words, it would suggest that the words, rather than the letters therein, should be affected. It seems a slight stretch to infer from ‘Portia staggered’ that alternate letters of PORTIA should be selected, but that’s a problem with most of the indicators in this group, and one could argue that it’s at least as good as ‘Portia now and then’, which doesn’t even suggest alternation. I’d certainly be happy with “Of the finest quality, Portia’s letters staggered me’ for PRIME. ‘Pre-eminent Portia staggered me’ for PRIME? Yes, I think I’d view that as pleasingly innovative, albeit mildly questionable, and I’d much rather see it than ‘occasionally’ or the like.