I’ve almost finished Pat Barker’s Regeneration (a wonderful book, and I’m looking forward to the two sequels), and I’ve just run across an interesting conundrum in punctuation. A sentence on page 202 reads “Sassoon, Rivers left till last, and found him lying on the bed in his new room, wrapped in his British warm coat.” I was taken aback by the first comma, which seemed to me wrong (there’s not normally a pause after a preposed object—cf. “Hegel I’ve never been able to read”), until I mentally rewrote it without the comma and had “Sassoon Rivers left till last,” which temporarily threw off the sense of syntax and perhaps suggested a phantom character named “Sassoon Rivers.” So I turn to you, my picky and keen-eyed readers; comma or no comma? (No fair suggesting a rewrite of the sentence; it’s perfectly good English, you’d say it without a second thought, and what the mouth can say, the pen—or pixel—should be able to reproduce.)

Addendum. I’ve just (Sept. 6) run across “British warm” (see comments for explanation) in In Parenthesis, on page 97: “A young man in a British warm, his fleecy muffler cosy to his ears, enquired if anyone had seen the Liaison Officer from Corps, as one who asks of the Tube-lift man at Westminster the whereabouts of the Third Sea Lord.”


  1. I’m sorry, but I really must insist that the sentence should be rewritten; for me, at least, it’s outright ungrammatical, because it involves an extraction from one of two conjoined verb phrases. It would be okay if you deleted the him, making the extraction across the board:
    Sassoon, Rivers left till last, and found lying on the bed in his new room, wrapped in his British warm coat.
    I would keep the disambiguating comma after Sassoon. But while we’re at it, I’d also like to reverse the order of British and warm.

  2. What’s wrong with recasting it? Written prose should not be a mirror of speech. If this were dialogue in the mouth of a character you might want to mirror speech patterns, but the syntax of written exposition should conform to a different standard.
    To me, the “, and” is the problem here. It is the main cause of the confusion. Make the latter clause unambigously dependent or independent; don’t make the reader have to consciously parse the sentence.
    “Sassoon Rivers left till last, finding him…”
    “Sasson Rivers left till last. Rivers found him…”
    And for some reason “British warm coat” seems out of order. I would write it “warm, British coat.” I don’t know why, but the latter flows more naturally–at least to my ear.

  3. I agree that the sentence should be rewritten. As written, I had to stop and figure out what the author was trying to say, but taking out the comma leads to the confusion you noted. If you have to choose one or the other, I would go with no comma, but only if the previous context makes it clear that there are two separate characters named “Sassoon” and “Rivers.”
    “British warm coat” makes it sound like this fellow has a bunch of warm coats, but it’s the British one he’s using.
    I vaguely remember seeing something somewhere on rules for adjective order though I don’t know where.

  4. Although at first I did what you did, remove the comma, but it did sound like there might have been something called “Sassoon Rivers.” When I read it over in the original way, the only thing thing that bothered me a whole deal was “British warm coat,” unless he was wearing something called a “warmcoat,” I prefer, “warm British coat.” I can’t tell you why, but there must be a reason, at least thats what Linguistics has told us so far 😛

  5. A ‘British Warm’ is another name for a greatcoat. I read it as a British warmcoat.

  6. How about dropping a semicolon in after Sassoon – the preposed object feels like a topic to me.

  7. Q. Pheevr and Dave Wilton: Great minds think alike! You’re both right, the “extraction from one of two conjoined verb phrases” is iffy and probably made the sentence niggle at me even though I wasn’t consciously aware of it. Given that, I agree that rewriting is probably a good idea.
    But as Jaarizz points out, British warm is a fixed expression.

  8. I hadn’t known about “British warm.” Now I do. I often learn new things reading language hat.
    I think I like Dave Wilton’s suggested rewritings better than the one I originally proposed–I was trying to change as little as possible, which is often a mistake.

  9. Sassoon he left until last, and found him lying on the bed in his new room wrapped in a warm, British coat.

    This requires a previously offered reference for the initial ‘he’, but it reads much better than consecutive names. Other fixes:
    1. I hate ’till’.
    2. As per previous commentors, ‘British warm’ is odd and distracting in my dialect.
    3. Too many masculine pronouns (especially with my ‘he’ fix), so I dumped ‘his’ coat for ‘a’ coat. We can assume it’s his. He’s wearing it.
    Of course I’ve totally ignored your restrictions. While it’s arguably grammatical, I would neither say it nor write it and would look horrified at someone who did either, so unless the author is deliberately adopting a perverse idiom, I’d recommend a rewrite and not feel bad about it. But I’d say the comma is the preferred choice in the false dichotomy.

  10. Whoops. I somehow missed the British warmcoat explanation before commenting. Mea culpa. I’d definitely kill the space if that’s an accepted spelling.

  11. Call me a heretic, but I like the sentance as written. That comma needs to be there, and I can mentally “hear” this sentance and it makes sense to me.

  12. For me, it depends how much stress I’m putting on the topic; to take a simpler example, “him I don’t like at all” and “him, I don’t like at all” both seem correct to me, but the latter implies much more emphasis on the word “him.”
    In this case, the version with the comma posed me almost no trouble, but I think I’d have been quite confused had it been lacking.
    That said, the whole pronoun thing is a serious problem here, because as Q. Pheevr has pointed out above, the first “him” really should be deleted; but then once that “him” is gone, the “his”-es later in the sentence look like they’re referring to Rivers rather than to Sassoon. My first instinct, on reading Q. Pheevr’s revised version, would be that Rivers found Sassoon on Rivers’s new bed, wrapped in Rivers’s coat; and my second instinct would be to wonder why Sassoon had no respect for other people’s privacy.

  13. I am bemused by how many people totally ignored your restrictions. Wilful folks, these copyeditors.

  14. Plenty of sentences work in speech but not in writing, and vice versa. This one needs to be rewritten. If it must be preserved, it shouldn’t have the comma, which doesn’t belong there at all. With or without the comma, the sentence will cause at least momentary puzzlement for readers, which is why it should be rewritten.

  15. If it were me, I would put it exactly the same. It’s almost like extra information. The only other way I can see to make sense of it myself is to add whom, but I would still keep the comma.

  16. Comma. The purpose of punctuation is to help the reader, not to satisfy rulemeisters.
    Plug: Comments welcome on my 2006 update of Strunk’s original, The Elements of Style Revised.

  17. Huh. An interesting sentence, one which it would never have occurred to me to say or to write — not idiomatic in my idiolect 🙂 I see the sense of the criticisms of it, but I also like its economy. And its meaning seems clear to me.
    I agree with most that if the sentence stands so must the comma.
    But the British warm coat brings me to a question. Why does “big red house” sound right to me while “red big house” sounds wrong? (Unless we’re using the color to differentiate between a number of big houses.) What’s wrong with a British warm coat, and why is it made right by “British warm” being a specific kind of coat?

  18. A terrible sentence. Too many things that can go wrong with it. That said, I agree with Ran’s notion: the preposed topic can be integrated or supplementary (marked by no comma, or comma); both are grammatical, though integrated is more usual. But to avoid the written appearance of a single two-word proper name, the comma is required.
    Both occurrences of ‘his’ are ambiguous, but that’s just because it’s a pronoun; it’s not a fault of this particular sentence. Presumably previous context would guide us, or failing that the default pragmatics that each person is in their own coat and their own room.
    I’d put ? rather than * on the extraction. “Chocolate I like, and eat (?it) all the time” – no, it’s bad, since the fully grammatical alternatives “and eat” and “and I eat it” are crying out to be read in its place, but I think it’s degraded rather than impossible.
    Bad, messy sentence.

  19. Not supplementary. Ignore ‘supplementary’. Was using it as synonym for “set off by intonational boundary”, which it isn’t.

  20. “Rivers left Sassoon till last.” Obviously.

  21. Sure, the sentence might with profit be recast. But I’m with those who think that, being perfectly sayable as idiomatic English, it ought to have the regimentation of its punctuation considered as a separate and interesting issue. How, after all, would you write it down in reporting direct speech?
    When I address this issue, I find no compelling case against the comma after Sassoon. It is sometimes acceptable to use a comma to clarify the structure of a sentence, for example where in speech the structure is made clear by emphasis (Sassoon Rivers left till last,…), a device not always elegantly available in the written text.
    The comma is overworked in English, and it has long shown the sad effects of this overwork. Sometimes there is no perfect solution in recording spoken English. But the comma can still be made to cover many sins, if we who wield it do so with great care, and with consideration for its current frailty.

  22. Yeah, it’s grammatical, and it works out loud, but in writing you don’t have any equivalent of the clue of auditory emphasis that lets you know in speech that “Sassoon” is the topic of the sentence.

  23. I think the sentence as it stands causes too much confusion.
    The comma causes the reading pause that is not natural for the sentence.
    The juxtaposition of proper nouns makes it too ambiguous, and it can’t be fixed with punctuation the way, aurally, oral intonation makes the sense understood.
    If, however, it had been full names, it would work a little better because it’s easier to figure out where proper name starts and another proper name ends:
    Sassoon Smith Brad Kennedy left till last
    And as an aside, quite often when listening to a conversation in a language I am not fluent in, whenever I hear a proper noun that I am not familiar with, I think that it’s a missing piece of vocabulary and I assume the word to be a normal noun or verb and completely misunderstand the sentence.

  24. How, after all, would you write it down in reporting direct speech?

    I’d require my informant to restate it, on pain of torture if necessary. No, I’m mostly kidding.
    However, this does highlight an assertion by the hat that’s been heretofore ignored and which I think needs to be addressed. viz,

    what the mouth can say, the pen—or pixel—should be able to reproduce

    I don’t believe this is true, both from experience and from the simple mathematical proof that the amount of information conveyed acoustically in speech (via pitch, wave-shape, dynamics, tempo, etc) absolutely dwarfs the amount of information that can be conveyed by a stream of glyphs selected from a population of a few tens. Honestly, it’s orders of magnitude. So rather than assuming it should be possible to translate back and forth between the two universally, I’m inclined to think it’s a complete miracle that you can even come close in some situations.

  25. Maybe it’s my dialect, but I take issue with preposed objects. “Ulysses I attempt to read every year”, albeit true, is awful for me.
    Also, an ‘and’ after a comma always suggests to me that the following is a conjunct of whatever occurred before the embedded part, which is closed off by the comma. So in this sentence, my immediate reading of the ‘and’ is that it conjoins ‘Sassoon’ and ‘found him…’ but that is of course ridiculous.
    Finally, the ‘British warm coat’, well i’d take ‘British’ to be that other sort of adjective, I don’t know its name, but it operates like ‘red’ in ‘red wine’. One speaks of a ‘good red wine’, not a ‘good, red wine’ and certainly never a ‘red good wine’.
    I therefore vote for:
    “Sassoon, Rivers left until last and found him lying on the bed in his new room, wrapped in his warm British coat.”
    But the preposed object still upsets me.

  26. It’s true that without the comma “Sassoon Rivers” looks like a single name, but with the comma I have no idea what to make of the sentence — it seems to be addressed to Sassoon and talking about when Rivers left, but then it peters out into nonsense. It’s a “garden path” sentence that cries out for rewriting.
    As with “There are three ways to spell /tu/”, there’s really no way to present the spoken sentence reasonably without resorting to linguistic notation of some sort.

  27. It’s the second comma that’s the mistake. “Sassoon, Rivers left till last and found …”
    “British warm” is the standard soldier’s greatcoat.

  28. It’s the second comma that’s the mistake. “Sassoon, Rivers left till last and found …”
    “British warm” is the standard soldier’s greatcoat.

  29. michael farris says

    As for your original question:
    Yes, I’d put a comma there. The only way the senence can Make Sense (for me) in spoken English is for there to be a intonation break between Sassoon and Rivers (I have high-rise-fall on Sassoon and low-level on Rivers) and a comma is as good a means for indicating that as any barring tone marks. Sassóòn Rìvèrs left… if it were a single name it would be Sasoon Rívèrs
    As for rewriting (it is a little awkward) while changing as little as possible:
    For me, all that would be needed is adding back ‘he’ as follows.
    “Sassoon, Rivers left till last, and he found him lying on the bed in his new room, wrapped in his British warm coat.”
    That makes it perectly legible though maybe more haphazardly informal than is customary in formal written prose.

  30. michael farris says

    Among the other variables going on with people’s reactions, I’m wondering how much sensory preference is playing a role?
    My preferred senses are muscle memory followed closely by hearing with sight a not very close third. I almost always subvocalise to a lesser or greater extent (muscle memory) and ‘hear’ writing in my mind’s ear (I was astonished to find out there are people who claim they don’t ‘hear’ printed text).
    I wonder if those who insist in recasting are more visually oriented?
    I’ll also mention here that my taste in second/foreign languages runs toward those with a relatively close match between the written and spoken forms. I pretty much avoid Chinese, Japanese, French, Arabic and the like, not because I dislike the languages but the visual information is just too different from the audio for me.

  31. There seem to be many people sidetracked by the ‘Britsh Warm’. Let me describe it: a double breasted, camel-coloured, fine wool coat with shoulder straps. Its context: commonly worn by British officers in WWI, usually by majors and above but never with badges of rank on the shoulder straps. Worn later, but rarely seen of late. Sometimes regarded as the mark of a bounder (like a blazer with a Guards regiment buttons where the wearer never served in the Guards). Very very comfortable – and warm.

  32. Roger Depledge says

    As a Brit connected to a web full of all sorts of English (“no fair”!), I am struck by the vehemence that Americans bring to these issues. It reminds me of Allen Walker Read’s comment, diplomatically hidden away in a Britannica article: “Rather than observe the language around them, as Englishmen commonly do, Americans give up their autonomy and fly to a dictionary to settle questions on language.” Calm down, chaps.

  33. If this sentence is read in the context of the story, presumably readers would know that Sassoon and Rivers are two different people, especially considering that Sassoon is a relatively uncommon first name, and Rivers’s surname had no doubt been given beforehand. If the comma had been left out, would it have made it onto a blog?

  34. David Marjanović says

    – I’d drop the comma, but then German orthography outright forbids that comma… I still haven’t completely understood the English comma rules (being one of the very few people who have apparently understood the German ones!)… and German grammar is a lot kinder to preposed objects than the English one (despite the lack of a case marker on names).
    – If “British Warm” — note the capital W! — weren’t a proper name, the adjective order would be flat-out wrong. I have been explicitely taught the English rules on adjective order (because they aren’t quite identical to the German ones — they are in this case, however –, or at least they are generally stricter than the German ones).

  35. Roger: I’m not quite sure what you’re on about, unless you’ve just been dying for a chance to use the quote and decided you couldn’t wait any longer. Nobody’s flying to a dictionary; we’re discussing, we speakers and users of English, whether the sentence works better with or without a comma. Don’t tell me Brits don’t care about commas, because I know for a fact many of them do. Calm down yourself.
    For my part, I’ve been convinced by this discussion that the sentience does work better with a comma, which doesn’t surprise me, because Barker is a damn good writer.
    Saif: Thanks very much for that excellent description of the British Warm! Barker could have saved readers unfamiliar with the term some trouble by using a capital W.

  36. Hmm, did not know British warm was an expression. Sweet.

  37. I think Andrew has a good point. LH, when you were reading the passage, the names Sassoon and Rivers were presumably well established, so you wouldn’t have thought “Sassoon Rivers” was a single name (any more than we’d think “Frodo Gollum” was one). Would you even have noticed the sentence if the comma hadn’t been there?

  38. Maybe not, although the “and found him” might have jarred me a bit just because of the odd pronoun extraction.

  39. I like the original better than any of the suggested alternatives. The rhythm’s nicer.

    “what the mouth can say, the pen—or pixel—should be able to reproduce”
    I don’t believe this is true, both from experience and from the simple mathematical proof [there is not room for the proof in this margin]….

    Here’s a different way to put what I suspect is a similar sentiment: Some of the best writers in English sometimes feel compelled to write a particular thing the mouth can say or has said, and therefore feel that it should be possible for them to do so. Foolhardy or not, that compulsion has been responsible for some of the best writing in English, and so, although I don’t wish to dispute sidereal’s argument, I can hope it doesn’t spread.
    In cultures with a sharper distinction between written and spoken language, I suppose the question doesn’t come up much.

  40. “For my part, I’ve been convinced by this discussion that the sentience does work better with a comma”
    Nice slip!

  41. I haven’t a clue what the sentence means, though maybe I’d have to read the context to find out more. As it stands, this group of words makes no sense and therefore needs to be rewritten.

  42. Nice slip!
    Yep, sentience sure does do wonders for a coma.

  43. I’ve ofter heard the noun ‘British Warm’ but wouldn’t have written it with the qualifying ‘coat’ because a British Warm is a coat.

  44. I hope I am not too late to weigh in here. I just found this blog and I’m qute satisfied that I did. I love language and to find a blog dedicated to it is quite a find.
    After reading the excerpt, this is what I understood it to mean. Please let me know if I am correct. I have not read the book.
    The character Rivers was doing some kind of search and he left the character, Sassoon, for last in this search. When Rivers found Sasson, he found him lying on the bed in his new room, wrapped in his British Warm coat.
    If this characterization of the excerpt is correct, then so is the use of the comma after Sassoon.

  45. It is (except that he was making his rounds as a doctor, not doing a search), and I agree. Thanks for the kind words!

  46. Wikipedia has a brief section about English order of adjectives in the Adjective article.

  47. a tort from a macadamien;
    “Clinton, Bush left till last, and found him lying on the bed in his new room, wrapped in his Manish warm Jacket.”

  48. I like the comma. To me it indicates a pause that forces the required emphasis on Sassoon mentioned by Noetica.
    Though it’s a different construction altogether, this reminds me of the opening of Joyce’s Ulysses: “Stately, plump Buck Mulligan came from the stairhead, bearing a bowl of lather on which a mirror and a razor lay crossed.” Is stately here an adverb (“majestically, loftily”) describing how he came from the stairhead, or another adjective (“impressive, grand”) describing Mulligan? The comma, added by Joyce on the galley proofs, I believe, creates ambiguity.

  49. tim atherton says

    A note to further emphasise the British warm coat issue. It should indeed be British Warm. It would not normally be qualified with the addition of “coat”. That would would be rather like saying “he was wearing an anorak coat or a Macintosh coat”
    (As noted the British Warm was a woolen overcoat generally worn by British Officers in the First World War and later. In fact I recall them being worn in the 1980’s by Army Officers.)

  50. So, Tim Atherton, is British Warm the standard expression for that kind of coat? Being unfamiliar with the term, I parsed the phrase as “British warmcoat” — something akin to “tropical greenhouse” in contrast to “tropical green house.”
    A better analogy would be “Rhode Island Red,” which certainly wouldn’t normally be written as “red Rhode Island hen” nor even “Rhode Island red hen” (with “red” uncapitalized).

Speak Your Mind