Please use the comments form attached to this page for any thoughts, ideas, questions etc about cryptic crosswords. The feedback page should be used for comments on the site itself.
Recent Posts
Recent Comments
- Doctor Clue on Notes for Azed 2,773
- Andrew Wardrop on Notes for Azed 2,773
- Matthew H on Notes for Azed 2,773
- Jay on Notes for Azed 2,772
- Luciano Ward on Notes for Azed 2,772
- Doctor Clue on Notes for Azed 2,772
- Andrew Shields on Notes for Azed 2,772
- Luciano Ward on Notes for Azed 2,772
- Mike Thomas on Notes for Gemelo 17
- Andrew Wardrop on Notes for Azed 2,772
Hi again Doctor Clue,
I gather you tend to avoid “links” between wordplay and definition in your own clues, but I wondered if I could ask your thoughts about their use in double definitions.
I’m in a slightly difficult situation in that both definitions (and the solution) are plural nouns, but using “are”, would be cryptically incorrect (as I understand it). “Becoming” is the most obvious solution, but I’m wondering about “can be”, as a smoother alternative in this particular case. My thinking is that cryptically this would indicate “word meaning X can be [the same as] word meaning Y”. What do you think?
Hi Anon
Your question happens to have arrived just as I am putting together a new list for the site…of link words and phrases! You’re right that I try (hard) to avoid them in clues for my barred puzzles, but I’ve got nothing against them when used fairly, and they can be particularly useful in blocked puzzles to help solvers work out where the definition is.
Two points that I try to address in the introductory text for the new list are relevant here. The first is the cryptic grammar: ‘Pointed stick is hazard’ works for STAKE but ‘Pointed sticks are hazards’ (as you understand correctly) does not, since the cryptic reading still requires ‘is’. The second is that in a double definition clue either definition is treated as the wordplay (the setter can decide which one) – this might seem a bit strange, but it made more sense in the days when the wordplay was called the ‘subsidiary indication of the answer’, and although the two words are spelt the same, only one of them can be the answer (in effect a double definition clue is a ‘definition of answer plus definition of homograph’).
So pretty much any link word or phrase that doesn’t require a plural subject would be ok – ‘can be’, ‘could be’, ‘also’, ‘and’, ‘yielding’, ‘making’, ‘for’, ‘from’ etc. The only sort I would rule out would be those which explicitly indicate that multiple elements are involved, eg ‘combining in’.
I hope that makes some sense!
That’s really helpful – thank you. It’s excellent news that you’re compiling a page dedicated to this topic for the site!
It’s now published. Any suggestions for corrections, additions or improvements will be welcomed!
It’s brilliant – extremely useful. I wonder if “in” should go the other way – having read this recently on Alberich’s site?
“…as a link word, “in” implies “found in” so clues of this type should read [DEFINITION] in [WORDPLAY].”
https://www.alberich-crosswords.com/more-articles/link-words
Thanks, Anon – I’m glad that you feel it’s a useful addition to the site.
That’s an interesting point about ‘in’. I’m not sure why ‘in’ should imply specifically ‘found in’, but in any event if we say “Milk and salt are [found] in cheese”, the components are on the left and the result on the right. I accept that one meaning of ‘in’ given by Chambers is ‘consisting of’, but it is only seen with this sense in phrases like ‘a play in four acts’ or ‘a story in words and pictures’, when no further details of individual components are given; by contrast, there are, for instance, “one capital E, one small E and one L in ‘Eel'”.
I know one very fine clue writer who is quite relaxed about the ‘directions’ of certain link words/phrases, and I suspect that he would accept ‘in’ as being bidirectional, but I still need a bit more convincing!
I find this completely fascinating!
Intuitively, I’m closer to Alberich on this one, in that I think what he means is along the lines of
‘defined word is found within correct parsing of wordplay’. Looking at Chambers I suppose that might be something like definition ‘by way of’ wordplay.
However, reading your view (‘components found in results’) persuades me (for one) that both ways are in fact valid.
Interestingly, on that same webpage Alberich advocates ‘to’ as a occasional possibility, which I don’t especially like, but I now see that Chambers has ‘with the object or result of’ among its definitions. I suppose that a phrase such as ‘rags to riches’ could be understood to indicate X becoming Y.
This is one of those situations, as with many things cruciverbal, which comes down to shades of opinion. The decision typically hinges on what is likely to be understood/accepted by the solver, and I am softening to the idea of ‘in’ being bidirectional. My particular objection was to the expectation that the solver should infer ‘found in’ from ‘in’ – why not the more obvious ‘are/is in’, or the equally reasonable ‘resulting in’? But one could say that “There is joy in eating éclairs’, and TS Eliot wrote “I will show you fear in a handful of dust’, so it’s going to get a thumbs-up.
Several prepositions (eg ‘by’) have ‘headline’ senses which suggest that they might be valid as links, but although the cryptic reading of a clue does not have to make sense as a piece of English prose, it must accurately tell the solver how to reach the answer. It’s possible, I think, to construct a clue where ‘by’ is involved in a link from definition to wordplay, but it requires additional words which do not directly contribute to the wordplay or definition, eg “Get Z by putting X around Y’. I would say that ‘to’ falls into the same category, but ‘X holding Y to Z’ doesn’t strike me as fair.
Hello Dr Clue. Wonderful site – thank you. Re: matching parts of speech – what’s your view (/the accepted view) on using a present participle to define an infinitive, and vice versa? To give a simple example, “die” as a definition for “expiring” seems plausable to me, as would the reverse.
(I realise a present participle can of course act as a noun (i.e. gerund), adjective, etc depending on context, too.)
Hi Matt, and thank you for your kind words.
I would generally apply the ‘substitution test’ as an initial sense check in this sort of situation. If we take the infinitive ‘to eat’ in the sentence “I like to eat éclairs”, it could be replaced by ‘eating’ without changing the meaning. ‘Eating’ in this example is a gerund, which suggests that it might be reasonable to use an infinitive to define a gerund, in your example ‘to die’ for ‘expiring’ (‘To see is to believe’ / ‘Seeing is believing’). The infinitive in English includes the word ‘to’, so in the reverse direction ‘expiring’ would only be valid as a definition of ‘to die’ (not ‘die’ on its own), and ‘to die’ would not be a valid answer in a crossword.
However, crossword solvers generally do not like definitions where the parts of speech appear not to match, and they are likely to write such definitions off as being in error, resulting in dissatisfaction. For this reason, I would suggest that regardless of the justification it would be unwise to use an infinitive in a clue to indicate a gerund answer, and I would expect an editor to reject such a thing.
The present participle is at the other end of a one-way arrow when combined with an inflection of the auxiliary verb ‘be’, so ‘are expiring’ is a valid definition of ‘die’ (and ‘is expiring’ of ‘dies’), with this equivalence being sufficiently familiar to solvers to be acceptable; the reverse is no good, since ‘are expiring’ would not be a valid grid entry.
I hope that makes sense.
Very insightful – thanks!
Assume a clue contains an insertion and an anagram. Would it be ok to have insertion indicator in past participle (say, squeezed by) and anagram Ind in imperative or intransitive verb form (doctor/struggles, for example)
My response below is based on the established principles of sound clueing – a regular correspondent (and clue writing expert) observed recently that he is seeing a lot of ‘plonkers’, these being wordplays where two or more elements have been ‘plonked’ together with no regard to the overall grammatical soundness.
1. Imperatives. If the result of an imperative construction is followed by a separate wordplay element, there normally needs to be a preposition in between. So ‘Doctor has humanity’ (a ‘plonker’) is not valid for SHAMAN but ‘Doctor seen by socialist’ is ok for SNEERED. When there is a second cryptic manipulation (the situation you describe), things are rather different. ‘Doctor tried catching cold’ is fine for DIRECT because the C can be ‘caught’ prior to the rearrangement. Similarly ‘Doctor seen squeezed by doctor’ is valid for DENSER, the rationale being the same – there is nothing in the clue that says in which order the manipulations are to be carried out, so SEEN can already be in the grip of DR when it is rearranged.
2. Indicatives of intransitive verbs. The situation is very similar when there are separate wordplay elements, so ‘Ace struggles hard’ is not valid for EACH while ‘Ace struggles near Hungary’ is fine. When it comes to multiple manipulations, though, it’s rather different – ‘More fragments squeezed by French’ is not valid, since the order of the words means that there is only one way to evaluate the wordplay, which leaves the main verb ‘fragments’ out on a limb. There needs to be a conjunction between ‘fragments’ and ‘squeezed’, so ‘More fragments when squeezed by French’ is ok. ‘More fragments, squeezed by French’ comes to much the same thing. If the intransitive verb comes at the end of the wordplay, though, all is well. ‘Boil squeezed by yours truly bursts’ is ok for MOBILE, because the BOIL can already be ‘squeezed by’ ME when it ‘bursts’.
In situations like these, the setter has to ask themselves whether the wordplay can be read in such a way that it either (i) consists of elements which can stand alone, or (ii) forms a coherent whole which can legitimately be interpreted in the intended way. I hope that makes sense.
The invisicomma, again? While utterly unsound, “[d]octor has, humanity” conveys the separation of elements where the absence of punctuation fails? I continue to rack up offences and risk becoming defensive.
I’m afraid the comma, visible or invisible, doesn’t help here. If the wordplay starts with an imperative verb, the remainder must continue the instruction. ‘Rearrange X Y’ or ‘Rearrange X, Y’ tell the solver to rearrange X and Y; it is expecting too much to ask them to infer ‘Rearrange X [and add] Y’. Examples of valid wordplays for SHAMAN would be ‘Doctor has, add humanity’ or ‘Doctor has ahead of humanity’ – and ‘Has doctored humanity’, which is a charade of two elements, ‘doctored has’ and ‘humanity’, each of which delivers a result.
I concede the point–not that you needed my approval. I do not routinely employ the imperative form and have one fewer sin to confess than originally believed.
Hi Dr Clue. Seeking another bit of your sage advice… So, the latest OED (and I think Collins) list “starkers” as meaning “2. completely irrational or out of touch with reality” However, Chambers only lists the more common usage: i.e. stark naked. I feel the second OED definition makes the word fair game as an anagram indicator… What do you think? (It would offer some lovely surface misdirection…!) Or is this just inviting accusations of foul/unfair play…?
Hi Anon
Where a primary reference is explicitly given for a particular puzzle, eg Chambers for most barred crosswords, I would say that the ‘bonkers’ sense of ‘starkers’ is a non-starter. For most blocked puzzles, no primary reference is shown, but the Collins English Dictionary (14th ed, 2023) is often used by editors as the basis of their guidance to setters, largely because it contains a good selection of better-known proper nouns such as place names as well as a selection of biographical entries. If a required meaning such as ‘starkers’ = ‘bonkers’ is in there, then it would potentially be fair game – the question that you would need to ask yourself is “Is my target audience likely to know the sense required by the cryptic reading or have ready access to the relevant dictionary?” Many solvers of blocked puzzles see Chambers as the ultimate arbiter, so I would personally view as high risk the use of any unlisted sense of an existing headword in Chambers as an indicator.
Thanks for that. It’s for a private blocked puzzle that will be a gift. However, I took the step of buying the Collins in paperback today (rather than the online version) and it’s not in there except in the sense of naked, so I’m going to give it a miss. Ah well! Perhaps in 10 years time, when Chambers and Collins have caught up with Oxford, I’ll be able to make the most of it!
Yes, I think with words like that you have to wait for the dictionaries to catch up – and although it’s reached the OED, it hasn’t made it into the Oxford Dictionary of English yet.
Two small/short points, Doc:
1. ‘Small’ is widely used by setters to indicate ‘s’. S=Small is in your list of abbreviations but it isn’t in Chambers. I think Azed has accepted it in clues submitted, but because of the omission from Chambers I was dubious about using it and I’m fairly sure I never did. Was I over-cautious?
2. Chambers has SW=short wave (as do you). Could that justify using ‘short’ to indicate ‘s’? I wouldn’t be happy.
😀
1. How interesting! I don’t know how it got in there, and it will have to go. The intention was that the list would accurately reflect the abbreviations in the latest edition of Chambers, since that’s the closest to ‘definitive’ that I can get. I can’t immediately see that Azed has ever allowed S for ‘small’ in a competition clue (he might have used it himself, but he’s also used C for ‘cadet’ at least once). The abbreviations in Chambers are, to put it generously, idiosyncratic, and there are other obvious omissions such as ‘Live/Neutral’ (L/N), ‘Large’ (L), and ‘Won/Lost/Drawn’ (W/L/D), along with some mysterious inclusions such as D = ‘deserted’. The ‘back pagers’ have their own lists of single-letter abbreviations, often including S = ‘small’ and L = ‘large’ but excluding many of the ones found in Chambers. I think there could be merit in adding a ‘blocked puzzle single-letter abbreviations’ list, albeit it could never be precisely applicable to all blocked puzzles.
2. No, I couldn’t accept ‘short’ for S, and I think there are a number of single-letter abbreviations in Chambers which are distinctly questionable since they are only ever seen as part of longer terms – ‘Academy’ and ‘Athletic’ spring to mind, although there are others. Leaving aside the fact that Chambers is right even when it’s hopelessly wrong, I can’t see that any single letter abbreviation which cannot stand alone in real life can stand alone in a clue.
Perhaps only on this side of the Atlantic is ‘S’ used to indicate a “small” size for clothing (along with M, L, XL, and such)? As such, ‘S’ should a perfectly serviceable abbreviation for “small”.
In the UK, S/M/L are ubiquitous (and universally understood) for small/medium/large as clothing sizes, likewise L/E/N for live/earth/neutral on electrical equipment and W/D/L for won/drawn/lost on sporting league tables. The problem is that when it comes to barred puzzles Chambers is the definitive reference for abbreviations, and of the foregoing it lists only ‘m’ = medium, ‘E’ = earth and W = won; the last of these relates to the Korean currency, while the context of the other two is unspecified. So S (small) and L (large) were inconsistent entries in the existing abbreviation lists on the site, and have been removed. However, I am applying the finishing touches to a new table, which will bring together all the single letter indicators (not just abbreviations) likely to be found in blocked puzzles – this will include S/M/L but (sadly) not L = ‘live’ and N = ‘neutral’ as I don’t believe that these are allowed in any UK ‘back pagers’.
Hi Dr Clue,
Quick question relating to a point below about deletion indications using formulations like “X when Y is removed” i.e. passive or “X when Y departs” (intransitive)… Can one also fairly use a participle-based formulation “X, when losing Y,”?
Cheers!
Hi Anon
Quick answer: yes!
There are many possible constructions, but as long as (i) they are grammatically sound, and (ii) they suggest that the (X-Y) state exists in the present (or will do in the immediate future), they are likely to be ok. The examples that you give are all valid, as would be (say) ‘X with Y departing’ and ‘X once Y has been removed’.
Examples of invalid constructions would be ‘X Y departs’ (‘depart’ is intransitive) and ‘X had Y removed’ / ‘X that Y left’ (these describe a past state with no indication that it persists).
I hope that’s helpful.
Wonderfully helpful – thank you!
Is ‘s (apostrophe s) a valid telescopic ind especially when there are multiple words in the fodder (considering ‘s attached only to the last word)? I understand some setters are ok with it but felt it’s a bit weak.
When you refer to a ‘telescopic’ indicator, I believe this is what I would think of as a ‘hidden indicator – but correct me if I’m wrong!
I don’t think that the ‘attachment’ of the apostrophe-s is an issue per se, with “red hat is destroyed” and “red hat’s destroyed” being equally valid wordplays for THREAD – conventionally, a series of words can be treated as a single string for the purposes of cryptic manipulation. However, if apostrophe-s is being used in a ‘hidden’ clue, it must either be (i) a shortened form of ‘has’ in the sense of ‘bears’, or (ii) a possessive.
Considering (i), in normal usage it is always the auxiliary verb ‘has’ which is shortened, as in “He’s just written a clue”; this avoids any potential ambiguity where an object is involved – “He’s a wonderful friend” would invariably mean that the person we are talking about is a wonderful friend, while if we were referring to two people we would say “He has a wonderful friend”. While one might conceivably say something like “He’s no time for fools”, this is surely non-standard.
Considering (ii), could “Chancellor’s instrument” be valid for CELLO? The possessive can on occasion mean ‘belonging to’ in the sense of being a component part, so I think there’s an argument for it being ok, even if it doesn’t ‘feel’ quite right. I would suggest that “Donald Rumsfeld’s instrument” for DRUM is no better and no worse. I wouldn’t use the construction myself, but I would be hard pushed to say that it was unsound.
Thanks. Very helpful examples
While “He’s no time for fools” is perhaps a non-standard construction, one would not be troubled by “I’ve no time for fools.” Is the extension to the third person too much of a stretch?
With “I’ve” there is no ambiguity, since it can only be a contraction of ‘I have’; that said, my Irish granny would have been quite likely to say “He’s a way about him” or the like. The deciding factor for me is that at least one UK ‘back-pager’ explicitly outlaws this usage in clues, which for me puts it into the ‘discretion advised’ category – I doubt that many solvers would be put out by the device, but some editors will undoubtedly reject it.
I noticed “drunk” is listed as a Standard insertion Ind. I can’t think of a logic where A drunk B suggests A in B. Also, Extra is shown as a container Ind (though Advanced). Extra can be outside the scope when used as a prefix but does that suggest containment ?
Similar question on “chaps” too. To chap means to cause to crack rather than to crack. Is that enough to suggest insertion?
It’s one that has been in the list for as long as I can remember, though I’ve never used it myself. The Chambers def of ‘to cause to crack or divide’ suggests that it might be allowable, but based on the examples in OED I think it’s decidedly marginal – the one that offers most support is “The extremely cold winds…chap the timber, and kill the cattle”, which perhaps does suggest the winds actually getting into the timber. That list will be getting a thorough review soon, and ‘chaps’ will certainly be subject to further scrutiny!
Thank you. This is helpful
Thanks for that. The appearance of the past participle ‘drunk’ without a preposition is inconsistent. For the sake of both consistency and clarity it should (and now does!) read ‘drunk by’. So ‘A drunk by B’ would be the typical construction; while “B A’s drunk” (or ‘B A has drunk’) would also be valid (just as would ‘B A drinks’), that is equally true of all the past participle indicators, since they form their perfect tense with ‘has’. I will add a note to that effect in the introductory text.
Incidentally, the Deletion list contains a number of indicators (mainly for ‘departure’) which similarly lack prepositions. I am already in the process of improving that list.
Regarding ‘extra’, it is as you say shown as an Advanced indicator, and that is based on the Chambers entry for extra[2], meaning ‘outside’. This would not be allowed, say, in UK ‘back page’ crosswords, but would be acceptable in barred puzzles such as The Listener (where I have used it myself).
A question on adjectival anagrinds – Is it generally ok to use adjectival anagrinds after the fodder? I have seen setters using them after the fodder but we normally use adjectives before the noun in English so thought there could be some grammatical explanation.
Hi
That’s a very fair question. By convention, adjectival anagrinds can be placed either before or after the fodder on which they act. As you say, this is contrary to typical English usage, but it is not unknown, as in (for instance) ‘attorney general’ or ‘amphibians proper’, or instances of anastrophe (‘in the forest dark and deep’). It’s important to make the distinction between what is commonly encountered in ‘real world’ English and what could legitimately arise, so I think the foregoing is sufficient justification, but since adverbial anagrinds are generally accepted although they cannot legitimately act on a noun expression, there is in any event a requirement for solvers on occasion to make certain inferences. Thus in ‘lean badly’ for LANE the solver must infer something like ‘arranged badly’, and therefore with ‘spoilt rotten’ for PISTOL they could similarly be expected to interpret this as ‘in rotten state‘.
The same applies to letter selection indicators, eg ‘sunny hollow’ for SY.
New homonym (to me anyway) in today’s Times Quick Cryptic No 3126.
9a: Podcaster’s way of running portal (4)
GATE, a homonym of gait.
Hope this helps and thanks for a great site.
Hi Matthew
Thanks for this, and for your kind words.
Seems pretty sound to me, and I like people to think that I’m moving with the times (no pun intended). It will be added at the next update.
Each Azed/Gemelo crossword is accompanied by the tagline “The Chambers Dictionary (yyyy) is recommended”.
In the last century I won a copy of Chambers from The Independent and have been using it ever since for all my dictionary needs, as they say. I often wonder how much I’m missing by not using the latest printed edition.
So I just downloaded the Chambers app and think it’s not something I’ll be using too often, unless it’s got stuff I really need. You see I like the serendipity of opening the wrong page and finding something new, or opening at the word I think I need but finding I’m wrong and the word i actually need is several entries away from it.
All that is a long-winded (I’m Irish, we use English like it’s going out of style) way of asking if you prefer the printed edition or the app?
I actually won twice and for the second chose the thumbnail-index version, which I’m saving as a gift for me new grandchild.
I would suggest that you are missing very little – there are a few new words (and meanings) in each edition, but not many. The 12th edition included a number of ‘enriching’ words, which were highlighted as being of special interest; when the editors said that they wanted to omit these asterisked words from the 13th edition (2014), somebody obviously took them at their word and left them out altogether (whoops!) – they were reinstated in the 2016 printing, which might yet end up being the last new edition of the dictionary. I would say that all barred puzzle setters should have access to the latest (2016) version in either paper or electronic form.
I must admit that while in years gone by my well-thumbed copy of the big red book was one of the first things to be packed when we went on holiday, I now use almost exclusively the electronic versions, for three main reasons. One is that certain words can be hard to find in the paper version due to cross-referencing failures – an example would be BLADDER SENNA in the recent Azed, which appears only under the entry for SENNA. The second is that, as a setter, I am able to take advantage of the search facilities that exist in the electronic versions. And the third is that I can travel a little lighter when going on my hols 😉.
As someone fairly new to Azed clue writing competitions, I’d very much like some feedback on my submission for the last competition. The clue word was ECBLASTESIS and here is the clue…
Budding A-list celebs distraught when left abandoned by society (11)
(A list celebs -L)* + S (society)
On reflection I think the definition “budding” is a little weak, though I note that there are two clues in the VHC section of the slip which use this definition.
Appreciate any thoughts.
It was a very difficult word to define, and Azed clearly gave competitors a degree of latitude – ‘budding’ seems pretty sound to me, given that the Greek word ἐκβλάστησις means ‘shooting or budding forth’.
Your clue seems to me to flow nicely from beginning to end (in the surface reading, nothing is obviously ‘tacked on’). The problem that I see is the word ‘when’ connecting the initial anagram fodder with the element to be removed. The ‘abandoned’ in the cryptic reading is an adjective (a participial one), and therefore the clue requires something like ‘with’ between ‘distraught’ and ‘left abandoned’; a comma would also be valid, and would preserve the intended surface reading, albeit the flow would be broken somewhat.
With ‘when’ followed by the thing to be lost, a passive construction is needed, ie ‘when left is abandoned’ (or “when left’s abandoned”). If the verb has an intransitive form, that could also be used, eg ‘when left goes’.
I hope that is helpful.
That’s very helpful, thank you.
In the US (where I began setting cryptic crosswords), barred grids are exotic, often reserved for especially-challenging puzzles: much as you describe in your second paragraph. The same appears to be true in Canada, where I now reside. In my view, blocked grids are visually attractive before and after filling; barred puzzles less so. Even so, I would prefer to set barred puzzles, but my insistence that every puzzle have a theme constrains my options (as mentioned already).
I recall some discussion of blocked vs barred cryptics. Perhaps because I am accustomed to (North) American styles and structures, but more because the themes of my grids are seldom possible to achieve in a barred puzzle, the majority of my cryptics (>150 published to this point) are blocked. Some patrons have requested more barred grids because of the larger number of checked letters; while I would wish to accede to their request, a blocked grid allows me to include many more themed entries. [Quality is of course more important than quantity, yet having ten theme words rather than four makes for an easy choice.] Note, though, that more recent grids include some bars between words, resulting in hybridized structures that are quite satisfying.
My contribution (above) to the “Psychology Ward” was in response to something I dimly recall: perceived difficulty in setting blocked vs barred cryptics. Is my recollection inaccurate? Perhaps Doctor Clue was describing challenges associated with having blocked puzzles published; perhaps there was nothing there at all.
Going back to the mid-1930s, many blocked puzzles were already recognizable as ‘normal’ crosswords, where roughly half the letters in the entries were checked, the words used were the sort that would be familiar to most solvers, and there would be no theme of any kind. Barred puzzles were almost another world – created and developed by academics such as Torquemada and the early Listener setters, they were thematic, demanded a wide general knowledge (especially of the classics, although Listener no. 3 apparently required a familiarity with Hindustani), drew on a broad vocabulary, and bordered on the arcane. Their only concession to solvability was the higher proportion of unchecked letters in each entry. Afrit was something of an ‘outlier’, in that his puzzles were blocked but often included pairs of adjacent cells which produced two-letter entries that were simply ignored when it came to the filling of the grid; effectively his grids therefore combined blocks and bars.
By the middle of the last century, the division was very clear – blocked cryptics were light entertainment, ranging from the easy up to the Times crossword and the like, while their barred counterparts were battlegrounds, where the setter was as likely to come out on top as the solver, not always by fair means. Then in 1966 Ximenes set out the ‘rules’ for cryptic crosswords in his book ‘On the Art of the Crossword’, and it was not long before the majority of barred puzzles were being produced to Ximenean standards, while blocked puzzles continued to feature clues that Ximenes would have considered unsound.
More recently, though, the divisions have blurred. The clues in a typical Guardian Genius blocked crossword could equally well belong to a barred puzzle, and blocked puzzles regularly include explicit themes, ghost themes or hidden ‘extras’ such as Ninas or puns. At the same time, barred puzzles such as Azed are often plain, have no theme, and typically present less challenges than a high-end blocked cryptic.
In the UK, if you’re not a setter with one of the newspapers and you want to see a puzzle in print, then barred puzzles (Listener, Inquisitor, Enigmatic Variations) are potentially a good option since they offer a route for all comers, with no contracted setters. There are routes for blocked puzzle setters to reach the papers, but they are poorly defined.
From a setting point of view, the difference depends more on the series for which one is setting (and thus the expectations of the solvers) rather than the grid construction. I’d agree that the amount of specific thematic words that can be accommodated is typically higher in a blocked grid, at least without recourse to the dusty corners of Chambers, but the barred puzzle is better suited to creating more complex effects, including changes to the initially filled grid and the production of pictures.