« previous post | next post »

Helen DeWitt writes with a question about "afterward(s)" and "backward(s)":

I've had comments back from my editor on a book that is to come out in late October.  He mentioned that when he started going through the document he changed "afterwards" to "afterward" and "backwards" to "backward" but later stopped, so the words could be left as they stood. It might, he thought, be better to be consistent (the text was not entirely consistent).

Much of the book is written in the voice of an American salesman.  I ran a couple of searches on Google (since it was not clear why the change would have been suggested in the first place), and found various people claiming that, while both were acceptable, "afterwards" and "backwards" were more characteristic of British usage. I have no idea whether this is true; if so they would be wrong for the character.  My slight sense was that, in America, "afterwards" and "backwards" would be more characteristic of an oral, colloquial style (a sense that is slightly strengthened by the fact that most of the first 100 hits for "ass backward" bring up "ass backwards" or "ass-backwards" – though, on the other hand, a search for "ass backwards" gets 638,000 hits, but a search for "ass backward" gets 1,840,000, and while many of the instances without final s are presumably attributive I'm not sure how to find out how many are the s-less adverbial form). If this intuition was right the form with final s would in fact be right for the character. Even if right for "backwards" this might not be true of "afterwards" (or, for that matter, "towards," which also came up).  I wondered whether you might be willing to ask readers of LL what they think.

Before going any further, please do yourself a favor and buy a copy of Helen's novel The Last Samurai.  Trust me, if you're the sort of person who enjoys reading Language Log, you'll enjoy reading The Last Samurai even more.

Now that you've called your local bookstore, or placed your order on line, we can learn a fair amount about the whole X-ward(s) business with a few quick searches in available corpora. What we learn is complex, helping to explain why Helen and her editor are unsure about things.

  • The choice between X-ward and X-wards is subject to variation in all regions and registers.
  • In the case of back- and after-, American English uses a higher proportion of -ward forms than British English does (and the proportion of afterward forms in British English is very low).
  • In the case of backward(s), in both American and British English, conversation shows a lower proportion of -ward forms (and thus a higher proportion of -wards forms) than academic prose does.
  • In the case of afterward(s), American English also shows more -wards forms in conversation than in academic prose, while British English seems to go the opposite way, though there are too few -wards forms overall to be confident about the relationship.
  • At least in American English, these patterns seem to be stable over time.

Some more quantitative results follow. ["BNC" is the British National Corpus (100 million words); "COCA" is the Corpus of Contemporary American English at BYU (400 million words); "LDC CTS" is the LDC's collection of conversational telephone speech (25 million words)].

Summary for backward(s):

backward backwards % backward
BNC overall 632 1739 27%
BNC spoken 32 211 13%
BNC academic 141 139 50%
COCA overall 6213 3127 67%
COCA spoken 243 998 24%
COCA academic 781 310 76%
LDC CTS 37 170 18%

Summary for afterward(s):

afterward afterwards % afterward
BNC overall 38 4446 0.8%
BNC spoken 6 468 1.2%
BNC academic 1 292 0.3%
COCA overall 7161 4211 63%
COCA spoken 312 1649 16%
COCA academic 551 525 51%
LDC CTS 69 624 10%

There are a number of other -ward(s) words to look at; and no doubt there are effects of the house style at different publications; part of speech (adjective vs. adverb) no doubt makes a difference; and as Helen notes, particular expressions (like "ass backward(s)") no doubt also have their influence.

One more piece of quantitative evidence — in COHA, the U.S. percentage of backward usage in decades from 1840-1850 to 2000-2008 has remained quite stable:

81 75 79 77 83 82 82 86 87 81 76 76 70 75 78 72 78

The OED explains that

In English the history of -wards as an advb. suffix is identical with that of -ward; beside every adv. in -ward there has always existed (at least potentially) a parallel formation in -wards, and vice versa. The two forms are so nearly synonymous (the general sense of the advs. being ‘in the direction indicated by the first element of the compound’) that the choice between them is mostly determined by some notion of euphony in the particular context; some persons, apparently, have a fixed preference for the one or the other form. Sometimes, however, the difference in the form of the suffix corresponds to a difference in the shade of meaning conveyed, though it would not be possible to give any general rule that would be universally accepted. Where the meaning to be expressed includes the notion of manner as well as direction of movement, -wards is required, as in ‘to walk backwards’, ‘to write backwards’. In other instances the distinction seems to be that -wards is used when the adv. is meant to express a definite direction in contrast with other directions: thus we say ‘it is moving forwards if it is moving at all’, but ‘to come forward’, not ‘forwards’ (see further the note on forward adv.); so ‘to travel eastward’ expresses generally the notion of travelling in the direction of an eastern goal, ‘to travel eastwards’ implies that the direction is thought of as contrasted with other possible directions. Hence –wards seems to have an air of precision which has caused it to be avoided in poetical use.

There appears to be no appreciable difference in meaning between the prepositions toward adj. and adv. and towards prep. and adv.; the latter is now, at least in British use, more common colloquially.

The OED's rather tentative effort to create a sometimes-categorical difference in meaning does not seem to stand up to trans-Atlantic test. Thus BNC has 1 instance of "walk backward" versus 15 of "walk backwards" — but COCA has 25 instances of "walk backward" versus only 17 of "walk backwards".


  1. Nick said,

    February 21, 2011 @ 9:37 am

    I always hear this apply to toward/towards. As a native AmE speaker though I definitely have both and I find myself producing both regularly. I think I agree with most of what the OED is saying about -ward(s) in general.

  2. Amy Stoller said,

    February 21, 2011 @ 9:58 am

    AmE NS here: In conversation (and in writing, too, I think) I do both in afterward/s, backward/s, but not in toward and forward, where I am completely s-less (QED). Huh. Wonder what that means.

    In beside(s) I have -s exclusively except when describing actual relationships: beside the bed but besides that, … .

  3. Spell Me Jeff said,

    February 21, 2011 @ 10:22 am

    As a fledgling novelist in college (1980s) I subscribed to the less is more philosophy, and began the habit of dropping the s from these words. I still like efficient prose (which is not to say spare) so I still write the s-less forms. I have no idea which I choose when I speak. I suppose I'm inconsistent.

    The other extreme. As a teacher of college freshmen and sophomores, I regularly distribute writing samples for discussion. I do not grade on grammar and style, nor do I teach it, because doing so is hopeless. I emphasize organization and content. Yet, invariably, as students struggle to find things to say, a handful fall back on what they think are the correct rules of usage. "Toward" is a frequent target. (As if the success of an essay turned on a single letter.)

    "Shouldn't there be an 's'?"
    "It's optional."
    "It also has nothing to do with our discussion of transitions."
    "But . . ."

  4. Coby Lubliner said,

    February 21, 2011 @ 10:22 am

    How about anyway(s)?

  5. John Lawler said,

    February 21, 2011 @ 10:29 am

    It's fixed at [Singular] -ward in the phrase "X____ Ho!"
    And I agree that individual and therefore unconventional semantic distinctions are rife in situations where everybody knows both forms but has form preferences. For instance, to me afterwards has always referred to a more serious thing than afterward, which is just a less awkward variant of after it.

    Afterwards is much more expansive (I always took it to be plural, I think), and refers to an event, a story, an interaction, an epilogue, something definite,
      … if there is an afterwards
      Afterwards there'll be dancing.
    and this is abetted by the fact that afterwards gets mixed up with After Words because I pronounce them identically.
    But I wouldn't expect anybody else to understand that that's what I mean when I say afterwards; it's not a convention, just a preference.

  6. The Ridger said,

    February 21, 2011 @ 10:35 am

    Just a note: the novel has nothing to do with the Tom Cruise movie of the same name…

  7. Pflaumbaum said,

    February 21, 2011 @ 10:37 am

    @ Cory Lubliner

    I always thought anyways was purely AmE, but I've been hearing it in England recently (or seemingly recently anyway). Would I be right in suspecting that it's much more common as an adjunct at the beginning of a sentence with a pause after it than at the end?

    Also a ways. I'm not sure I've ever heard it in speech but Thomas Pynchon uses it, and I can't completely parse it. Does it mean a little way or away or something else?

    Speaking of Pynchon, there's a great stoned discussion of the phrase ass backwards in Gravity's Rainbow (pp683-4 in the Jonathan Cape first edition).

  8. GeorgeW said,

    February 21, 2011 @ 11:22 am

    Pflaumbaum: Then, there is euphemistic 'bass ackwards' which I think I have heard more than 'ass backwards.'

  9. Brett said,

    February 21, 2011 @ 11:22 am

    I once had a copy editor complain that my writing had both "toward" and "towards" in roughly equal numbers, arguing that I should pick one or the other. I told him that, to my mind, the two are different (albeit closely related and synonymous) words. The OED quote above sums up how I select one or the other: "[T]he choice between them is mostly determined by some notion of euphony in the particular context…."

  10. language hat said,

    February 21, 2011 @ 11:25 am

    I'm glad for this post, because I've been wondering about this myself (I've been enforcing a style guide that requires -ward on the MSS I edit, but I wasn't sure what the facts of usage were — including my own).

    Also, I heartily second the recommendation to buy and read Helen's first book; if you want more details, I babbled enthusiastically about it here. I'm very much looking forward to her next.

  11. KevinM said,

    February 21, 2011 @ 1:13 pm

    Re: the possible AmE/BrE divergence. Just for fun, I googled "arse-backward" and "arse-backwards" The vast majority of the hits were for the latter. Caveat: a lot of "asses" came up, too, leading me to believe that the google heuristic captures both when one is sought.

    My personal preference is "backwards," because it preserves a useful distinction — I reserve "backward" for the sense of "retarded in development." To my ear, "backward" also fits better as a description of the direction of motion, as opposed to orientation.

  12. KevinM said,

    February 21, 2011 @ 1:16 pm

    Sorry, should have mentioned: My ear is a mid-Atlantic, American one (and so is my other ear).

  13. GeorgeW said,

    February 21, 2011 @ 1:18 pm

    What would the be origin of the /-s/ in 'backwards?' Did it arise through analogy? With what? Presumably it would be a plural.

  14. John said,

    February 21, 2011 @ 1:49 pm

    Amy: Are you sure you're completely s-less for forward? I thought I was, too, but then I thought of it in alternation with backwards:

    Are we going backwards or forwards/*forward?

  15. keri said,

    February 21, 2011 @ 1:53 pm

    I was going to comment that it feels like there's a slight difference between the 's' and no-'s' forms of the words, but I couldn't quite place it. Then KevinM commented, and I think he pinpointed at least one of those differences. I don't know if I actually use the forms differently, but it feels as though I might, because of the slight difference in connotation that I perceive.

    As for the American vs. British usage – I am constantly typing "afterwards" into my browser, only for Firefox to give it the squiggly red underline, so I change it. I tend to think that without the 's', it should be spelled "afterword", which is a completely different thing. I'm an American speaker from N.E. Florida, by the way, with family from Chicago and Pittsburgh, and rural Florida/Georgia.

  16. Will said,

    February 21, 2011 @ 1:58 pm

    Re: "a ways" or "a way"

    I use this occasionally in casual AmE speech… I don't think the added "s" particular changes the meaning, except to make it less formal, and both forms are acceptable. The phrase itself just means some unspecified, but not insignificant, distance.

    Can be modified into other more descriptive phrases, such as "a little ways" and "a long ways."

  17. mollymooly said,

    February 21, 2011 @ 2:46 pm

    The OED's note at "-wards" dates from the first edition and is merely the personal opinion of the relevant editor (Henry Bradley?). By 1926 Fowler was already calling the putative distinction spurious or obsolete.

    The Atlas of English Dialects' maps of "forward(s)" and "backward(s)" show the -s form is more widespread for the latter than the former.

  18. Michael Johnson said,

    February 21, 2011 @ 3:31 pm


    If I recall my Old English studies correctly, it's not a plural analogy, but a genitive one.

    In OE one could turn nouns into adverbials by putting them in the genitive case, e.g. 'nihtes' (at night). The analogy spread so that other adverbial expressions (anyway, backward) also got the "genitive" s.

  19. Michael Johnson said,

    February 21, 2011 @ 3:32 pm

    Just to add to that, 'since' also got the "genitive" s. It went from 'syththen' to 'syththens' to 'since'.

  20. Dan Lufkin said,

    February 21, 2011 @ 4:09 pm

    @Michael Johnson
    We still do that in Modern English: I like a dram of whisky of an evening, but mornings give me a mug of tea. Swedish has a similar construction: Nu är det dags att sticka. "Now it's [the time] of day to leave.

  21. Craig said,

    February 21, 2011 @ 6:49 pm

    Another mid-Atlantic AmE speaker here, I usually lean toward the s-forms, but, as I just showed, it's a question of euphony and the desire to get rid of the Firefox spell-checking squiggle.

    @Michael Johnson, I'll also throw in that German and Dutch also have -s genitives. The German -s genitive is a regular part of the language for masculine and neuter nouns. To the best of my knowledge, the adverbial time constructions "on X(s)/in the X" are derived from this case, e.g., the days of the week (montags, dienstags, usw.) and the times of the day (morgens, mittags, abends, nachts).

    Although Dutch has the same forms ('s maandags, 's dinsdags, ens.), the language underwent an official reform in 1946/1947 where it lost its case endings, so these forms are sort of a frozen artifact.

  22. Helen DeWitt said,

    February 21, 2011 @ 8:20 pm

    Thank you – that's tremendously helpful. I should emphasize that my editor, Jeffrey Yang at New Directions, has not insisted on these or any other changes to the text, I just wanted to know what would be right for the usage of the character (I think in my own usage the s-less form predominates, but this felt vaguely wrong for the character and I now see why). Though irrelevant to the book, it's extremely cheering to learn of the extension of the OE genitive and the official abolition of case endings in Dutch in 46/47.

  23. m.m. said,

    February 22, 2011 @ 2:08 am

    Coby Lubliner said,
    How about anyway(s)?

    Pflaumbaum said,

    @ Cory Lubliner
    I always thought anyways was purely AmE, but I've been hearing it in England recently (or seemingly recently anyway). Would I be right in suspecting that it's much more common as an adjunct at the beginning of a sentence with a pause after it than at the end?

    A quick ngram has "anyways" AmE usage currently more than doubled compared to BrE.

    Oddly, I've used "anyways" as long as I can remember, but have only been called on its use a few times, always on text. Never orally.

  24. John Walden said,

    February 22, 2011 @ 2:31 am

    Isn't there a difference between "a backward(s) glance" where the s is optional and "a backward child" where "s" would make the child face the wrong way? In fact you could even call it "a child backward"

    To show how little I know about English, I'm not even clear about what the word modifying a verb-turned-into-noun is:

    glance look step leap etc

    Is "a backward(s) glance" the same as "a glance backward(s)"?

  25. rootlesscosmo said,

    February 23, 2011 @ 2:34 pm

    What would the be origin of the /-s/ in 'backwards?'

    Compare German "vorwärts." (A teacher at my New York high school ca. 1956 criticized a student for pronouncing German as though it were Yiddish: "You sound as if you're reading the Forverts backverts!")

  26. Gwillim Law said,

    February 24, 2011 @ 9:55 am

    Would it be off-topic to discuss the distinction between among and amongst, amid and amidst, while and whilst, unbeknown and unbeknownst? My understanding is that the -st forms are very rare in AmE. In the UK, Australia, and Canada, there seem to be some people who think the -st forms are more formal, some who think they're merely pretentious, and some who choose on the basis of euphony.

    Checking up on that in the BNC and the COHA, I find that the -st forms are in fact less common in both AmE and BrE. The frequency of the -st forms in AmE has tended to decline over time. Unbeknownst is an exception. It has the highest relative frequency in AmE historically (66%), and even more so in the recent past; but its absolute frequency is low, so it has only a weak influence on the totals. Amidst is the most popular of the variants in AmE for the period 2000-2009 (21%). Combining the four pairs, the -st forms account for 12% of the total frequency in the BNC, and 2.7% in the COHA. Considering only the 2000s, it's 1.2% in the COHA.

  27. maidhc said,

    February 25, 2011 @ 3:22 am

    There's an entertaining discussion of "ass backwards" in Gravity's Rainbow.

    I would prefer to hear
    "You've got that ass backwards"
    "He entered the room ass backward"

    That's just by ear, I haven't attempted to craft a general rule.

  28. dglp said,

    February 26, 2011 @ 5:52 pm

    'Then, there is euphemistic 'bass ackwards' which I think I have heard more than 'ass backwards.''

    Agreed. But it doesn't come across strongly as having or not having the s. I'm as likely to say bass ackward as bass ackwards. And I think Google should take account of the Spoonerific spelling in their results.

    As an AmE speaker in Br (E), I note the addition of -s to various things, and wonder if there's a colloquialism that isn't based on philosophy. For example, why does Safeway become Safeways? Is it a case of greengrocer's apostrophe in reverse? Or do people just habitually add an -s to certain words? Could that be the case with -wards?

  29. samboha said,

    May 6, 2011 @ 7:15 pm

    Regarding dglp's question about "Safeway" > "Safeways" : I've noticed this occurring a lot lately among people I know, almost exclusively with place names and retail locations – Kroger(s), Foxfield(s), even Mellow Mushroom(s). I mentally process it, and perhaps the speakers do too, as a possessive morpheme ("Mellow Mushroom's") in analogy to other establishments actually named after their owners. I suspect this has provided the tragic basis for a personal bias against the "afterwards," "towards," and "backwards" forms.

    For the record, I'm an AmE speaker at college in central Virginia.

  30. S-Series II: Backward(s) « Motivated Grammar said,

    June 14, 2011 @ 9:23 am

    […] we have for the back story for backward(s), what's their current status? Mark Liberman had a well-researched profile of these words on Language Log a few months ago, and his two main conclusions were […]

  31. Carolyn said,

    June 14, 2011 @ 2:48 pm

    I came over from the Motivated Grammar post to see if anybody had said what KevinM said about the separate color of 'backward' as an adjective, as applied to, say, a region where indoor plumbing was not yet common. I agree with him there, and also on the movement versus location distinction. (I might pull a car backward into a space, unless I noticed that only one other car was in backwards.)
    Also happened to notice that the percentage given for 'COCA spoken', in the first table, should be 19 or 20 per cent, not 24.

RSS feed for comments on this post