How Many Is a Couple?

Anne Curzan at Lingua Franca discusses an interesting phenomenon. To her, as to me, “a couple of” basically means two, but she had a revelation:

While discussing language peeves in my introductory English linguistics course, one student, Katelyn Carroll, volunteered that it drove her nuts when people used the phrase a couple (of) to refer to more than two things. I heartily concurred, along with a few additional students, but a good number of other students in the course felt we were being persnickety — and, perhaps, were just flat-out wrong.

Katelyn ended up doing research on a couple (of) as a quantifier for the class usage guide, and it has changed my copy-editing practices. In a very small survey of undergraduates at the University of Michigan, she discovered that only one third of them believed a couple (of) could refer to only two items, and some of them believed that it always had to be more than two (i.e., equivalent to several, which is typically seen as more than two). Almost all the respondents agreed — as do I — that a couple (of) is informal, whatever it means.

Dictionaries indicate that I am not on solid footing with my restrictive definition. Merriam-Webster online provides the definition “an indefinite small number: FEW” for the word couple, with the example “a couple of days ago.” The online Webster’s New World concurs. The definition in the American Heritage Dictionary of the English Language online supplies a very similar definition: “Informal A few; several: a couple of days.” American Heritage defines several as more than two or three (but less than many); few is simply a small number. In any case, both dictionaries reflect much more flexibility than I was exhibiting in how many things are encompassed by a couple of things.

This was as much a revelation to me as to her, so I ask the assembled multitudes: what does “a couple” mean to you?

Comments

  1. Between 2 and 4, inclusive.

  2. For me (a speaker of Australian English), it means something like “two, or perhaps three, or perhaps even four”. I think that the extent to which the referent is individuated influences the interpretation: I would expect “a couple of beers” or “a couple of hours” to be pretty vague (because people often count these things inexactly), whereas “a couple of friends” would be more likely to refer to exactly two people.

  3. To me “a couple”, as in “I’ll send it to you in a couple of days”, could stretch as far as four. “Several” starts around 4 or 5. Being told “I’ll give you several of them” and getting just 3 would leave me feeling a bit hard done by. I’d be happy with 5 though.

  4. January First-of-May says

    Obligatory XKCD.
    (To be honest, I’m not quite sure how did the article’s author forget it.)

    I personally would probably say that all four of the terms listed there (including, yes, “a couple”) are more about the 3-7 or even 3-8 range than the 2-5 range. I would be a little surprised to see either of them used for 2 (though of course “couple” can still refer to a matched pair consisting of one specimen of either gender).

    Incidentally, Russian пара (cognate to “pair”), and especially the diminutive парочка, had also ended up as similar quantifiers (though “2, but sometimes up to 5” is probably still the correct interpretation of those words – they haven’t became “between 3 and 7, or in some specific cases 2” yet).

  5. Based on practical experience, “couple of days” means one, two or more days…

  6. I think I wouldn’t notice if “a couple” turned out to be anything up to four. After that it would be weird.

  7. Lucy Kemnitzer says

    A couple is “more than one, probably two, definitely no more than five.” I find myself saying “a couple-few” which I think means “less than 8, unless the whole conversation is about large numbers of things in which case it’s less than 8 hundreds or thousands or whatever.” “A few” is a very relative word, and can only be defined in the context of the conversation–it’s kind of like a statistical word, as is “many”–but “several” doesn’t scale up, I think.

  8. A couple of is indefinite for me, up to perhaps five. A couple, of course, is exactly two.

  9. As another Australian speaker, my instinct is to say that “a couple” means “two,” but on reflection Matt M. is onto something there re the referent. “Grab a couple of beers from the fridge” = two beers, no more or less. “We went out for a couple of beers” = could easily be three or four beers (each!).

  10. A couple is a pair. My wife and I are a couple. By the usual process of extension, couple can mean any pair of things, and then, any two of something even if that thing doesn’t come in pairs, and then, by a process of vagueness and imprecision, a number that is more than one of something and is about two. But the fundamental meaning which has never been lost is a a pair.

  11. As a speaker of American English (North Carolina, Virginia, DC) born in the ’60s, I agree with most of the comments above. “A couple of” does not have to be 2. It’s not equivalent to “several”, as Curzan says; that’s too many.

  12. A couple is a pair, but if you’re speaking informally and use it to mean a small number then I won’t so much as blink.

    Several is more than just “greater than two”, it has to be, like, more than five or something. If you have somewhere in between a couple of items and several items then you have a few items. I’m aware that other people do not have this mental definition. I don’t tell them it bugs me.

  13. Does anyone have a different order of the sequence couple – few – several – many?

  14. A couple for me usually means more than two (perhaps three or four), although it could also refer to just ‘two’.

    “Having a couple of friends over” could mean 3 or 4, maybe even 5. Restricting it to just two friends would be extremely literal minded. This is different, of course, from ‘having a couple over’.

    Incidentally, for me it’s always ‘a couple of xxx’, although I’ve noticed many Americans write ‘a couple xxx’, which sounds, well, wrong to me.

    I must say, a number of comments to the article are quite prescriptive in the usual way. For example:

    If “a couple” doesn’t imply two, and you’re “in a couple”, how many are involved in the relationship?

    A married couple might not want to be referred to as a married few…unless something else changes!

    And what about a coupler, or coupling–think of train cars or a device for bringing two things together. Couple definitely implies two.

    In my book, a romantic relationship between two people is always “a couple” — add another person to the picture and you’ve got “a threesome,” not a loosely-defined couple.

    Which is why I like GeoffD’s comment:

    We have here the makings of some good fake prescriptive poppycock.

    For example: rule: ‘several’, reflecting its consanguinity with ‘sever’, should be used only where the items are being divided or at least moving apart, rather than being joined or brought together:

    Wrong: ‘Several new members came to the meeting.’
    Right: ‘The Black Knight lost several limbs.’

    By extension, it should only be used for items that are heterogenous in space, time, or character:

    Wrong: The Shire of Bland has several attractions for visitors.
    Right: There were several attempts on the poles before Peary and Amundsen.

    Wrong: ‘Dan Brown wrote several successful novels…’
    Right: ‘…and used his earnings to buy several country houses.’

  15. I notice that Mollymooly participated in the blog’s contest in 2012 to make up a new prescriptive rule:

    The pronouns “somebody” and “someone” are illogical, and one should use “a person” instead.

    Reasoning: “Some” means either “a number of” (with plural nouns) or “an amount of” (with singular nouns). Since “body” and “one” are singular, “somebody” and “someone” mean “an amount of person” and is thus only appropriate for cannibals: “Would you like somebody? I’ve just taken a juicy missionary out of the oven.”

    See https://www.chronicle.com/blogs/linguafranca/2012/10/12/one-for-the-usage-books

  16. For me a couple signifies an uncertain number that is most likely two. I wouldn’t say “a couple days ago” if I thought it was three or four. Only if I wasn’t entirely sure it was two.

    In Chicago, at least in 20th century Chicago ethnic dialect, (where ethnic has its own specific meaning – white but not yuppie) one can’t use “a couple of beers” to mean 3 or 4, because there is a phrase for that – “a couple two-three beers,” with the “two-three” specifically moving the estimate away from the idea that it would most likely end up being only a couple.

  17. Slightly off topic, but…

    I (a non-native speaker) have always thought of “several” as meaning “a few, but more than expected”, but I’ve been noticing that dictionaries do not mention this nuance at all, and also I hear it used by native speakers in situation where this interpretation wouldn’t make sense, like “only several”. Yet, online I do find support for my interpretation

    https://english.stackexchange.com/questions/21418/is-it-bad-to-combine-only-and-several

    Am I right that at least some speakers interpret “several” in this way?

  18. Eli Nelson says

    @dainichi:

    I wouldn’t say “several” particularly implies to me “more than expected”, but I agree that “only several” sounds at least a bit odd. I would say “only a few” instead. When trying to think of how to define “several”, the phrase that came to my mind was “more than one“. “Only more than one” obviously seems very strange. The people above seem to me to have oddly high definitions of where “several” starts (to me, saying “several times” seems fine when something happened only three times, for example; I definitely wouldn’t assume it meant “five or more times”), but I guess we can generalize that it means “more than [some undefined but small lower bound]”.

    The definition of “few” that first comes to my mind is “not many”.

    I guess “a few” seems to me to mean “some”, more or less, although probably there are contexts where they can’t be interchanged. “Only some” sounds better to me I think than “only several”, but worse than “only a few”. This is consistent with the order of their frequencies according to the Google Ngram Viewer.

  19. I’ve long assumed that all native speakers take “a couple of” to mean “basically two but possibly more, but not too many” when referring to non-binary sets. This usage is colloquial and considered incorrect by some, but I never thought there existed L1 speakers for whom “a couple” meant strictly “two” in sentences such as “I need to get away for a couple of days” or “I have a couple of tricks up my sleeve.” But I admit there must be Russian speakers who interpret “меня тут пару дней не будет” strictly as “I won’t be around for two days” and take offense when the speaker is still away on Day 3. After all, if you think you might need three, rather than two, days off, why don’t you say “пару-тройку дней” instead of just “пару”?

  20. A couple of days could indeed stretch to three or four, but with each day over two, I get more and more irritated.

  21. To me, as a non-native speaker, ‘a couple’ is ‘two or perhaps a little more, as long as the difference doesn’t really matter’, just like Polish parę [+ GEN], which is etymologically the accusative of para ‘pair’.

    Not unlike the middle member of the Pirahã quantifying system.

    In the experiment conducted by Frank et al., spools of thread were provided in increasing number until ten were present, and participants were asked how many spools were present after each increase. The same elicitation was done in decreasing order. The results revealed that only three terms in Pirahã were used to express each of the quantities of spools, and those terms were hói, hoí, and baágiso.

    In the increasing elicitation:

    Hói — used to describe one object
    Hoí — used to describe two or more objects
    Baágiso — used to describe quantities of three or more.

    In the decreasing elicitation:

    Hói — used to describe quantities up to six
    Hoí — quantities between four and ten
    Baágiso — quantities between seven and ten.

    As evidenced by the difference in quantities expressed by the same term in the increasing and decreasing elicitations, each term is shown to represent only a relative idea of quantity, and is not defmed by fixed cardinal boundaries, or “exact quantity.”

    https://etd.ohiolink.edu/rws_etd/document/get/oberlin1316100344/inline

    Let me add that even hói does not always mean precisely ‘one’, but rather ‘a very small number or amount’. A set of two may be hói or hoí to Pirahã speakers depending on various additional factors (a general impression of ‘as few as two’ versus ‘as many as two’).

  22. “A couple” for me would mean (outside specialised cases where it obviously means exactly two, such as married people and hounds) “I think there are two of them but I’m not sure”. I would definitely not use “couple” in a situation where I knew there were at least three of something.

    “Several” is fewer than “many” and more than “a few”, but what number each refers to is flexible. A few peas on a plate could be anything up to thirty. A few cars in a car park could be ten. But if I were counting elephants, I think I would get to “several” and “many” much more quickly. Even if my back garden contained only six or seven elephants, I would still be prepared to use “several” or “many”.

  23. David Marjanović says

    In German, ein Paar “a pair” and ein paar “a few” are not even perceived as related; while I haven’t read a lot of German prescriptivism, I’m not aware of any prescriptivists even noticing.

    Interestingly, Paar does not take its plural ending (-e) when it follows a numeral: zwei Paar Schuhe. Rather than a noun in the really strict sense, it’s a measure-word, like a Chinese classifier.

    Back to English: several seems to mean “more than one, contrary to your expectations”, in practice also “more than two” because if you mean “two” you’d just say “two”… or so I thought before I found out that, as Matt above, some people seem to use a couple of as the unstressed form of “two”.

  24. J.W. Brewer says

    I like David M’s Paar/paar distinction and would say that in AmEng “couple” and “coupla” are perhaps similarly distinct. “A coupla X’s” (meaning an indefinite number probably <5) is related to "a couple" (meaning a matched pair, with different stress pattern) only etymologically.

    The classic nostalgic look at teenagers trying to start a rock band circa 1965 from the super-jaded music-business-pro POV of 1979:

    "Down in Joe's garage
    We didn't have no dope or LSD
    But a coupla quarts uh beer
    Would fix it so the intonation
    Did not offend yer ear."

    Who would think that meant exactly two quarts, no more and no less (whether per member or for the band as a whole)?

  25. In my own speech, couple means exactly two. For my wife, it is an unspecified number. So, to avoid misunderstanding, I don’t assume it means two. And to avoid being misunderstood, I mostly restrict using it to inherent pairs, of the “one for each of us” sort.

  26. “Several” is fewer than “many” and more than “a few”, but what number each refers to is flexible. A few peas on a plate could be anything up to thirty. A few cars in a car park could be ten. But if I were counting elephants, I think I would get to “several” and “many” much more quickly. Even if my back garden contained only six or seven elephants, I would still be prepared to use “several” or “many”.

    That’s how the Pirahã system works. The size of objects being quantified certainly plays a role. Three little fishes ma be hói (≈ 1), while one satisfactorily big fish may be hoí (≈ 2).

    I wonder if the widespread singular/dual/plural distinction evolved as a formalisation of the more primitive fuzzy categories of “very little/few” vs. “a little/a few” vs. “much/many”.

  27. J.W. Brewer says

    To maybe try to be a bit more formal, is there a conflict between a parse that’s “[a [couple [of X’s]]]” and one that’s “[a [[couple of] X’s]]],” with the former being presumptively exactly two and the latter not? MMcM, for example, says what “couple” means to him and I’m obviously not going to say he’s wrong about his own idiolect, but I’m thinking for many speakers that’s a different question than what idiomatic “couple of” means and thus not necessarily the relevant question to be asking and answering.

  28. Michael Eochaidh says

    Ryan caught the “couple two three” usage I remember. I lived in Wisconsin for twenty years, three hours from Chicago, and have spent a fair amount of time in Chiicago. It might have been used in Milwaukee as well.

  29. I think beer is the canonical example. ‘A few beers’ is more than two, ‘several beers’ is more than a few, perhaps surprisingly more than a few. ‘A couple of beers’ is at least two, but possibly a few more.

  30. I agree with Matt M for Australia usage. However, the Macquarie Dictionary only recognises “couple” as referring to 2. This may be one of the rare instances where the MD is normative (prescriptive) rather than descriptive.

    In my native Croatian, “par” (literally “pair”, but more readily translated by “couple”) means – according to the Croatian Encyclopaedic Dictionary:
    1. a. Two similar objects or persons that normally go together, work together, have the one function or form the one whole, eg. boots, gloves, shoes; dancing couple; married couple. …
    3. (indeclinable) colloquial. a small number of something. compare “nekoliko” (meaning “a few”).

    So in Croatian, the “small number” meaning is an indeclinable noun, differentiating itself from the “2” meaning which is a declinable noun.

    It seems to me that a similar thing happens in English. Namely, “couple” by itself means “2” as in a “married couple”. Whereas a “small number” is denoted by what can be caled an indefinite possessive: “a couple of”, as in “a couple of biscuits”, “a couple of beers”

  31. The Pirahã study was interesting. As I recall there was a suggestion that numbers up to 4 are hard wired into our brain. Higher numbers need to be learned.

    The implication was that it was very difficult to teach numbers and mathematics to adults in pre-numerate societies.

  32. Interesting stuff here! The whole issue would be a good thing to throw at people who think language is logical and exact.

    I (a non-native speaker) have always thought of “several” as meaning “a few, but more than expected”, but I’ve been noticing that dictionaries do not mention this nuance at all

    I don’t have the “more than expected” nuance at all; it’s just “maybe more than a few.”

    “A couple” for me would mean (outside specialised cases where it obviously means exactly two, such as married people and hounds) “I think there are two of them but I’m not sure”. I would definitely not use “couple” in a situation where I knew there were at least three of something.

    I think my usage is similar to ajay’s (although with beers, of course, it’s fuzzier).

  33. Hmm, a question for the “several > few” people: can you actually use them as constrastive terms? To me, sentences like “He didn’t just make a few mistakes, he made several” or “This didn’t just happen a few times, it happened several times” sound weird. Do they sound natural to anyone?

  34. It does sound weird, but (I think) only because both terms are inherently vague, and trying to use them as if they were definite sounds weird. I wouldn’t contrast them like that, I just think I would be more likely to use “several” for a larger number than “few.”

  35. @Eli Nelson

    They seem OK to me. I think that just because the sentences are contrastive, one thinks “Well, there must be a contrast here.”

  36. @languagehat, MattF: Thanks for the replies! Interesting. I found an example of a sentence like this from a Google Books search: “All things! Not just a few things, not even several things, but all things” (NKJV, The New Spirit-Filled Life Bible for Women, by Thomas Nelson). But I think I would prefer the wording “Not just a few things, not even many things, but all things.”

  37. Interestingly, Paar does not take its plural ending (-e) when it follows a numeral: zwei Paar Schuhe. Rather than a noun in the really strict sense, it’s a measure-word, like a Chinese classifier.

    Or like “brace of duck”, “stand of arms”. Or indeed “couple of hounds”. (“Stand of arms” is weird because it’s a measure-word meaning “one”. 15,000 stand of arms = 15,000 weapons.)

  38. The few/several contrastive sentences sound entirely natural to me. What doesn’t is contrasting “several” with “many”…

  39. Thanks, so it’s not only me who thinks “several” means something like “more than one, contrary to expectation”. For example, I found this on https://www.democracynow.org/2012/4/24/death_on_the_border_shocking_video

    > He said that one of the officers, while in detention, had kicked his ankle, where he had surgery several years prior.

    This, to me, is weird. It would only make sense if the sentence was trying to downplay the harm done by the kick, which is obviously not the case (if you read the surrounding text and/or know democracynow). “A few”, “some” or even “many” would not have that effect.

  40. Huh. Sounds perfectly unexceptionable to me.

  41. January First-of-May says

    To maybe try to be a bit more formal, is there a conflict between a parse that’s “[a [couple [of X’s]]]” and one that’s “[a [[couple of] X’s]]],” with the former being presumptively exactly two and the latter not?

    It’s an interesting idea, and would certainly explain the “2, but sometimes up to 5”.

    The same thing happens with Russian пара, though, even though the Russian equivalent to “of” is the genitive case (though I guess the underlying parse tree could still be similar).

    Incidentally, in case anyone missed it, a comment I posted yesterday has finally came out of moderation (it is currently the fourth comment on this post).
    Other commenters had brought up the Russian word (and mentioned similar examples in German, Polish and Croatian… which makes me wonder how common it is in non-European languages – aside from Pirahã, obviously), but the link apparently hadn’t been posted by anyone else or discussed yet (which is weird, because it’s very related).

  42. Without reading any of the posts, and understanding I am an American:

    – What is ‘a couple’? (Two people who are partnered in some activity not necessarily sexual in nature.)
    – OK, smartass, what do you mean by ‘a couple of…’? (Formally, two. Informally, two, maybe three, probably not four. Possibly ‘more than one but less than four-ish’ if I can’t remember an exact number. I have been known to use the compound “a couple-three”.)
    – Isn’t that really the same as ‘few’? (No, ‘few’ means three to five, OR it means a relative quantity. For example: ‘The poll shows that about one out of fifteen Americans is an atheist; oddly, a few of them say they believe in God.’ This could mean thousands of people but far fewer than the total.)
    – How about ‘several’? (I might be weird but in my mind this is heavily influenced by ‘seven’.)
    – Anything else? (Pleeeese tell the Irish, especially my husband, that ‘going to the pub for an hour’ means sixty or so minutes and not to get mad because I’m not trying to be awkward when I ask, ‘that’s one hour, right, not until the beer runs out?’.)

  43. @Hat:
    Is “only several” natural for you, then?

    @David:
    >Back to English: several seems to mean “more than one, contrary to your expectations”,

    Yes!

    >in practice also “more than two” because if you mean “two” you’d just say “two”…

    Unless 1. you’re not sure how many, except that it’s more than one, 2 you want to stress the plurality, and “several” sounds more impressive than “two”.

  44. I grew up in Delaware and went to Princeton for my undergraduate. The usage of “a couple of” to mean any number between 2 and 5-ish seemed completely normal anywhere in the greater Philadelphia area.

    I then went to graduate school in Utah, where “a couple of” always meant strictly 2 and no more.

    I’ve lived about twenty years in the D.C. area now, and while I suspect the former usage is the natural one around here, I myself hew to the more restrictive usage, just because there’s no way to know which background any given interlocutor will have.

  45. About the several > few thing: No, it doesn’t sound weird, but I’ll tell you what does sound weird… “I only made a few proofreading mistakes in this book.” “No, you made several.” Why? Because I know how many proofreading mistakes are made in an average-length book, and “several” is less than that.

    By the way I’m another vote for “only several” being natural to my ear and having a natural meaning. “Only several of the girls in the ninth grade voted against trousers in the school uniform.” It’s an absolute number of between seven and ten, regardless of the total number of girls in class, unlike “few”, which just means “a vastly smaller quantity” when it does not mean “three-ish to five-ish”.

    If “few” doesn’t mean more than five, and “several” doesn’t mean less than seven, what word do I use for six? (Silly. “Six”.)

    I have no idea in the world what the “contrary to expectations” part of “several” even means.

    If I ask you for “a couple of those” and you always hand me a pair of them, I will always be perfectly satisfied with that even if I don’t actually mean exactly two.

  46. I was corrected recently by a grammar cop for using “several” in a context where it was clear I meant exactly two. I countered that I had always used “several” to mean more than one (as several other people above have mentioned). Nope, I was told, “several” means at least three. I’ve always used “a couple of” interchangeably with that sense of “several.” In general, I would use “a couple of” in spoken language, and “several” in written language.

  47. @Hat:
    Is “only several” natural for you, then?

    Nope, sounds odd to me.

  48. Probably two, but could very well be three or four.
    (I’m from California)

  49. For me (American native speaker of English), “a couple of x” means two or three, maybe four. I would agree that it is the equivalent of “a few.” I see it as more indefinite than definitely two.

  50. I’m with Hat in the reacting to dainichi. The “contrary to expectations” connotation is alien to me, I don’t really understand how it applies to the example given, and I don’t see the connection to “only several”, which sounds odd to me too.

  51. Except possibly for “several,” I basically concur with the definitions given by xkcd: Words for Small Sets.

    However, I have run into enough people who were picky about “a couple” meaning exactly two that I tend to use “several” to distinguish cases that involve more than two. Once, a friend pointed out that I was saying “several,” even when I really meant only two; so I sarcastically corrected myself to “a brace.”

  52. For my usage, “a couple of” is two things, “a few” is three to seven, while “several” is seven or more (note the overlap between few and several). And “a handful” is always five, because that’s how many fingers are usually on a hand.

    But it doesn’t faze me a bit when other people use “a couple” how I would use “a few”.

    I do find it odd that so many people responding here think “several” can mean as few as three (or even two). To me: several >= few > couple > one

  53. “Only several” sounds utterly bizarre to me. It doesn’t collocate. I’d only ever use “a few” or “a little” to indicate quantity after only, and I think the same would be true of any negating adverb. But, to me, several has no nuance whatsoever of “more than expected.” Several simply means “more than two but less than six.” But it usually isn’t used after any sort of negation. (Californian English).

    Like most other respondents, I’m sure I use “a couple of” fairly loosely to mean both “exactly two” and “two, or so,” depending on context.

  54. J.W. Brewer says

    Like Craig, I grew up in Delaware, so our childhood exposures to the meaning of the phrase match up well. Unlike Craig, I have never attended grad school in Utah. I’ve certainly never been corrected on my usage of “couple of” by people I’ve interacted with whom I knew to be Utahns. But maybe outside their home environment they felt too intimidated to do so? But what I do know about Utah in this regard is that COCA lives on servers at BYU. And I scanned the first page of COCA’s hits (there are tens of thousands) for “a couple of” without any examples leaping out at me where “two and only two” seemed the most obvious or natural interpretation — largely because even in an informal register if you want to communicate “two but definitely not three” there are more natural ways to do it. If for example you say “a couple of days ago” or “a couple of weeks ago” rather than “day before yesterday” or “week before last” (which are both precise without sounding unduly stuffy or formal in register), it probably means either that you’re not 100% sure that it was two days/weeks ago rather than three or that the difference between it being two and being three is just immaterial to the point you’re trying to make.

  55. J.W. Brewer says

    Of course, sometimes for rhetorical reasons people shift back and forth between a vaguer quantification and a more exact one for the same referent within the same discourse. So e.g. there’s a hit that starts with the sentence “It’s the first official weekend of the fall movie season and there’s a couple of movies on tap that look to lift the box office out of its summer slump.” It then proceeds to list exactly three movies. And the beauty part is that the headline (which obviously might have been written after the text) is the more precise (as well as more concise from a character-count perspective) “Trio of new films kick off first weekend of fall movie season.”

  56. “Beer” is a special case. People don’t want to be too specific about how much they’re drinking, so they use a term that allows them to under-estimate their consumption without quite fibbing about it. This is often done with an implied wink. “A coupla beers” is socializing; “a few beers” is getting a load on.

  57. “Stand of arms” is weird because it’s a measure-word meaning “one”. 15,000 stand of arms = 15,000 weapons.

    This reminds me of the use of troops to mean ‘soldiers’. As a child reading about the Vietnam War in the papers, I was always confused by reading that 10,000 troops were going to be sent to such-and-such a place. To me, troop meant a fair number of soldiers (looking it up now, the size of a U.S. Army troop varies from 16 to 50), so I would always have to figure out how many soldiers that might be, and wondering in annoyance why the papers couldn’t just say “200,000 soldiers” or whatever.

    And “a handful” is always five, because that’s how many fingers are usually on a hand.

    Really? For me a handful always means literally as much as you can hold in your hand: “a handful of sand” is a huge number of grains, “a handful of strawberries” maybe ten.

  58. Really? For me a handful always means literally as much as you can hold in your hand: “a handful of sand” is a huge number of grains, “a handful of strawberries” maybe ten.

    Surely, not “always.” “I’ve told a handful of people about my plans” does not refer to the number of people you can hold in the palm of your hand.

  59. J.W. Brewer says

    The “spaghetti western” movie titled in English “A Fistful of Dollars” (“Per un pugno di dollari”) was followed by the sequel “For a Few Dollars More” (“Per qualche dollaro in più”). Someone could tote up exactly how much cash is seen onscreen at various key plot points in both films and explore how that illuminates the meaning of these somewhat vague quantifiers, either in English or in Italian.

  60. Laowai, my hands are VERY large.

  61. Ксёнѕ Фаўст says

    The interesting thing about Polish parę (which is for me synonymous with kilka except for register difference) is that it hasn’t remained uninflected (like some other numeral-like words, e.g. dużo ‘many, much’) but has acquired Frankenstein case forms with numeral endings (similar to the numeral ‘2’ and the word kilka ‘a few’): paru (genitive/dative/locative/human masculine nominative), paroma (instrumental). The noun para ‘pair, couple’ inflects regularly and differently from parę.

  62. In languages with paucal number on nouns, paucal seems to be typically 3-4 or 2-4 (depending on whether there’s a dual too).

    Russian uses the genitive ‘singular’ for nouns counting up to four. My question is, is that limit strict? Has anyone used the plural for four of something?

  63. J.W. Brewer says

    I expect that unlike John Cowan, the XKCD fellow (linked above in Brett’s comment) must have been thinking of “handful” solely in its metaphorical extended sense and not at all in its literal and physical sense because I can’t imagine him seriously thinking that “a handful of M&M’s” would mean “two to five M&M’s.” Note also that the metaphorical sense has a strong connotation of “fewer than you might expect and/or want” whereas the literal doesn’t — a recipe calling for “a handful of” garlic or cilantro or what not may (depending on how large a volume of other ingredients it’s being combined with) lead to more of the flavor in question than many might find optimal.

  64. As E said, “a couple of” refers to a number that is *probably* two. You aren’t lying if you say it will take a couple days and it takes one, or three, or even four. Five… well, you estimated the probabilities really badly. It’s a probability curve, not a static range.

    Some one asked me how much “a few hundred” was, the other day, and after some pondering I said, “probably between 200 and 600…?” same thing, though. You’re saying the curve peaks somewhere around, I dunno, 400.

  65. I think my usage is similar to ajay’s (although with beers, of course, it’s fuzzier).
    Everything’s fuzzier with beers. 😉

  66. Just the other day I started panicking when I was warned I had only par minuta left to meet a deadline, only for the person to update me after 3 or 4 minutes that there were now only 5 or 6 minutes remaining. Very anticlimactic. But anyway, looks like par at least, can go all the way up to ten for some people.

    When I was young I used par to mean something like “at least two, but less than five”, until someone pointed out to me that that’s actually kind of strange when you think about it, and that it would make infinitely more sense if people just used it for “two(ish)”. That all sounded so very logical to me at the time, so I decided to do just that (with a couple of, as well, since I’ve always just thought of them as one and the same). Trouble is, people just don’t think two-ish when they hear the expression. If anything, it seems most people here agree with the students who wouldn’t use it for just two of something. Much more likely to hear it used for higher quantities (though to get close to 10 really is prety extreme). Turns out training your brain to think of a common expression in a way that’s totally counterintuitive to everyone around you for no reason at all but silly pedantry can make things kinda confusing occasionally. Hey, it made sense at the time.

    Several has no connotation of “more than expected” to me. I think I do think of it as more than a a few, but I couldn’t say why.

  67. I would never say “I”m having several friends over”. In that context, “a couple” means anywhere from 2-6.

  68. I would never say “I”m having several friends over”

    I don’t think I would either, which brings up the point that these words may be specialized in terms of context. I’d say “several times,” “several days,” but not “several friends” or “several cats.”

  69. But I would certainly say “I’m having a few friends over.”

  70. I also would never say “I’m having several friends over” or “they have several cats” (where I would say “I’m having a few friends over” or “they have a few cats”), but “several friends told me they might not be going” or “several cats have been hanging around in the back yard” sound fine to me. But I can’t really tell what distinction I’m making there.

  71. I would say that a couple is the second number on the inexact counting scale: one, couple, few, several, many. With values of
    one = 1 ± 0
    couple = 2 ± .5
    few = 3.5 ± 2
    several = 7 ± 4
    many = 13 ± 5 or more (though, probably not more than a hundred, which would warrant more elaborate language)

    Of course, all these numbers are inexact, and I wouldn’t necessarily expect someone else to agree at the cutoff points, so they are, necessarily, approximate.

    Also, for continuous or typically inexact things (inches, beers, measurements of time), these ranges could grow. For instance, a couple hours could easily be anywhere between half an hour and 4 hours.

    Additionally, for especially small things the values are likely larger: a couple peas is probably at least 3; a couple seconds is probably at least 5, more likely 10 and maybe even 30. (This does not happen with large things, a couple buildings is most definitely 2 and no more buildings.)

    Another model for understanding: if you had just a glance at an image, but did not have time to count, and were asked how many marbles were in it, and there were three, you might have only seen two of them, and therefore answer “a couple”. You might have seen three, and therefore answer “a few”. Or, you might have even thought you saw 5, and therefore answer “several”. In either case, it would be a reasonable answer. But, you almost certainly wouldn’t answer “many” because it was only three.

    As for having several friends over, it would have to be “for a party”, in which case, it probably means 7 to 15.

  72. The “person who wants to follow etymology always” in me wants to say that a couple must refer to two, but the way I actually use it means… a small number (less than 5?) of things. I can say “I have a couple of cats” and if I have three, that would be fine. If I say “I have a couple of cats” and I have nine, that’s quite something else.

    My pragmatics teacher polled the class on the numerical limits of a few, a couple, several, and other such quantifiers. I don’t remember if anyone had this particular restriction, but I wouldn’t be surprised if they did.

  73. David Marjanović says

    The few/several contrastive sentences sound entirely natural to me. What doesn’t is contrasting “several” with “many”…

    and

    I have no idea in the world what the “contrary to expectations” part of “several” even means.

    Methinks you’re both outside the Standard Average European Sprachbund, and Danish is in… the usage I described matches German mehrere. “Contrary to expectations” means emphasis: “you probably thought there’s just one (or two), but no, there are several!” And yes, that completely rules out only several as self-contradictory.

    German eh, while I’m at it, means “in accordance with your or my hopes or sarcastic/cynical expectations”…

    Russian uses the genitive ‘singular’ for nouns counting up to four. My question is, is that limit strict?

    I don’t think that’s comparable. The genitive singular is used specifically with the numerals 2–4 (and 22–24, 32–34 and so on), never without them. Measure-words like “pair” go with the genitive plural no different from what happens in English (if we count “of” as a genitive).

  74. Sure, I wasn’t too precise; but still, it’s related to the semantic concept of paucity, and I wonder how specific it is or has been.

  75. How do you see “some” fitting into all this? eg. i had a couple of friends over, vs a few friends, vs several friends, vs some friends.

    To me this appears largely synonymous.

  76. I think “I”m having several friends over” is ok if you stress the ‘several’. It would only work in specific contexts, though.

    “We really would prefer it if you refrained from having any more than one or two people at at time coming into the communal space.”

    “Well, in fact, I’m having several friends over for a party tonight. Is that a problem?”

    But that only highlights your point. ‘Several’ isn’t so likely to be used if you’re just speaking casually.

    Another situation: you have a problem and are looking for ways to deal with it. In each case the quantity word is slightly stressed:

    There are several ways of dealing with it. One is to …

    There are a few ways of dealing with it. One is to …

    There are a couple of ways of dealing with it. One is to …

    There are some ways of dealing with it. One is to …

    Of these, ‘some’ seems to be the odd one out. It could mean that, contrary to expectations, there are actually ways of dealing with it. “Don’t be so negative! There are some ways [admittedly not many] of getting a child to eat his vegetables. For instance…”

    Or it suggests that there might be drawbacks and risks. “There are some ways of dealing with it. But all of them involve the use of virulent poisons.”

    ‘A few’, ‘several’, and ‘a couple’ are all used in a positive sense without these implications.

    Nothing like introducing stress to highlight the semantics of a sentence!

  77. The constitution of the USA speaks of “the several states”, which can jar a modern reader, and not just because fifty seems like more than several. I’m sure that in those days “several” was not so much a vague number word as a word meaning something like “individual” or “separate”.

  78. I think of “a couple of” as strictly or literally two, casually or figuratively two or not much more than two.

  79. “I’ve told a handful of people about my plans”

    I’ve seen that in writing often enough, but I wouldn’t say it myself.

    several cats

    I wouldn’t use a paucal of my friends’ or mine own cats, because it seems weirdly underspecific; I would say “I have three cats.” (When TSE calls Mungojerrie and Rumpelteazer “a notorious couple of cats”, I think he is using the noun couple, not the paucal.) However, as an observation, where I don’t have exact knowledge, a paucal is fine: “I saw a few/several/some cats crossing the field today.” If I ran an animal shelter, I would certainly say “We have many cats” as a replacement for “316 cats”, or whatever.

    several states

    The oldest meaning of several in the OED is ‘separate’, and I think that’s the sense here. In the law of debt, a group of co-debtors may be bound to a creditor (a) jointly, (b) severally, or (c) jointly and severally. Respectively: (a) the creditor may claim the full amount of the debt from one debtor, some debtors, or all the debtors, but when the debt is paid in full by any subset, the rest are relieved; (b) each debtor only owes his individual part of the debt; (c) the debtors are jointly bound to the creditor but severally bound to each other to even up any surplus paid by some debtors.

  80. David Marjanović says

    I’m sure that in those days “several” was not so much a vague number word as a word meaning something like “individual” or “separate”.

    Or it just meant “explicitly plural”.

  81. I spent my high school years in Japan and largely missed out on American car culture, so I had a lot to learn when I came back after high school and spent the summer working in my uncle’s gas station in rural Virginia. I quickly learned that I shouldn’t take “a coupla dollars worth” of gas to mean more than two, which in the late 1960s in the U.S. bought you six gallons, more or less (c. 33 cents/gallon).

    The other problem I had was finding the damned gas tank: often behind the rear license plate, but on some cars behind a taillight.

  82. Ø, yep, it’s related to “separate” and “several”.

    May I only remark there were not fifty states at the use of that word.

    On the number terms, i polled my first-grade children and they tell me “a couple” is more than “a few”. They tell me they’re about five and two respectively. I hope they’re just severely unlicked cubs on this, not that this is the usage of their generation, because I will never internalize that.

  83. >I then went to graduate school in Utah, where “a couple of” always meant strictly 2 and no more.

    Ironic, in the only place in the country that ever allowed a more expansive interpretation of the married couple.

    The radicalism of the convert.

  84. >Except possibly for “several,” I basically concur with the definitions given by xkcd: Words for Small Sets.

    XKCD clearly is not a parent, or he’d understand that one can often be a handful.

  85. The constitution of the USA speaks of “the several states”, which can jar a modern reader, and not just because fifty seems like more than several. I’m sure that in those days “several” was not so much a vague number word as a word meaning something like “individual” or “separate”.

    This secondary definition of several is by no means obsolete, just rare; you’ll still find it in any dictionary. “They have broken the clan and gone their several ways. We who are here this morning have remained true to our fathers, but our brothers have deserted us and joined a stranger to soil their fatherland.” (Chinua Achebe Things Fall Apart)

  86. David Marjanović says

    The other problem I had was finding the damned gas tank: often behind the rear license plate, but on some cars behind a taillight.

    what

  87. I had an argument with my six-year-old daughter about “couple” the other day. I told her that “a couple” strictly meant two, but could also be used more loosely to refer to a small group (equivalent to “a few” or “several”). She insisted that “a couple” definitely had to be more than two, and thought that my use of it to refer to a group of two of something was incomprehensible.
    I don’t know where she got this idea from. Being six years old, she refuses to believe that her ideas about language come from anywhere. When we differ, it is simply that she is right and I am wrong.

  88. The other problem I had was finding the damned gas tank: often behind the rear license plate, but on some cars behind a taillight.

    what

    A common feature of American-made cars before the 1980s. A Google Image search will show you examples.

  89. Re: the difference between “a couple” [2] and “a couple of” [2-c.4] Compare:
    [A]
    * “He has few friends” [negative]
    * “He has a few friends” [positive (though not so much as to qualify for “contrary to expectations”)]
    * “He has only a few friends” [negative]
    [B]
    * “I have one or two questions” [I have at least two, probably more]

    Someone with LexisNexis access must be able to find courtcases where the plaintiff asserts “couple” means “exactly 2” and the respondent insists they only meant “a few”.

    I remember learning the word “several” aged about four and assuming it was related to “seven” and therefore must mean “seven or so”. “Several” is at the formal end of my common register. I would not say “I had several friends over” unless I was under oath or writing a novel set in Surrey. Also:
    * Burt Ward : Gee, Batman, is there anything you don’t know?
    * Adam West : Yes, Robin; several things.

    “Pair” is definitely different from “couple”. A romantic couple is about the only couple that could equally be described as a pair. Apparently “pair” can also sometimes denote more than two: MW sv “pair” has “3 chiefly dialectal : a set or series of small objects (such as beads)”

  90. To me (Mid-Atlantic AmEn), a couple is two or about-two-but-it-doesn’t-really-matter. Several is three, four, or five—six would be pushing it—and I had no idea so many people would interpret it as more than that, which makes me suddenly feel very anxious about everything I’ve ever written. A few can be the same as several or it can be more, depending on whether you’re talking peas or elephants, as ajay said. “A number of” is more than several or a few. And no matter what, additional beers are always going to be less problematic in my book than additional days of waiting for something.

  91. Barin Sergatchsky says

    In most Irish I have heard and read, cúpla means “a few” (most often 2 or 3, maybe 4) as in English, not usually exactly “a pair”. I wonder how far this goes back in Irish.

  92. I would certainly say “We have many cats”

    Huh. For me, positive “many” is part of written register; in speech, I use it only in the negative (“I don’t have many books on Albanian”).

  93. I wonder how far this goes back in Irish.

    Here’s the eDIL entry:

    n m. (Engl. or Romance loanword). Us. with dental infl. in pl. couple, pair : lucht an bheatha ché an c.¤ (a mother and child), DDána 37.22 ( IGT Decl. ex. 908 ). na sé ch.¤ (of apostles), 34.12 . trí ch.¤ dho chaiplibh, IGT Decl. ex. 39 . cuingir . . . cupla, Eg. Gl. 148 . cūpla . . . sārc[h]loinne / rucc si twins, ZCP xiii 18.15 . croinn ‘n-a gcúplaibh in pairs, DDána 111.22 . In sense of a few, a couple (of): c.¤ uair gach lá, Mac Aingil 1778 . c.¤ bráthar, Beatha S.F. 2615 . cūpla do chaillechoip, Fl. Earls 144.21 .

    In other senses of Engl. couple : sreath gcúpladh gcorcra couplings (of roof of castle), DDána 119.33 . a fhosdadha ┐ a chupladha, Celtica ii 125.174 . ó da ṡín c.¤ ar [c]comuind the bonds of our affection, Ériu iv 220 § 32 . Of versification: c.¤ do na ceithre féarsa fichead ‘couplet’, O’Gr. Cat. 296.14 .

    So, 17th century (Mac Aingil, Archbishop of Armagh, 1571-1626).

  94. Barin Sergatchsky says

    Go raibh maith agat, a Hata!

  95. Pleeeese tell the Irish, especially my husband, that ‘going to the pub for an hour’ means sixty or so minutes

    I am sure that he would point out (if the Irish are anything like the Scots) that he rarely even spends as much as an hour going to the pub. He may, once he has gone to the pub (however long that took), spend an hour in the pub, and then anything from twenty minutes to five hours leaving the pub, but going to the pub will be an act accomplished rapidly, purposefully and in the minimum time practical.

  96. On elephants, I think I may have been unconsciously influenced by M’Bu, a apprentice dealer in exotic animals in Terry Pratchett’s “Moving Pictures”:
    “How many elephants we got?”
    “Three, boss,” said M’Bu firmly.
    “Are you sure?”
    “Yes, boss. It’s easy to be sure, with elephants.”

  97. Huh. For me, positive “many” is part of written register; in speech, I use it only in the negative (“I don’t have many books on Albanian”

    Oh? Even, say, with the word times (e.g. “I’ve been to Paris many times.”)?

    For me, there’s nothing even the tiniest bit unusual about using many in a positive sentence in spoken English. While with a quotidian noun like “cat” my default would probably be lots of/a lot of, I’m pretty sure that with more abstract nouns (difficulties, obstacles, challenges, concerns, etc.) many would be my default in either spoken or written English.

  98. To me, several retains some of its distributive sense: several visitors = visitors arriving independently, not as a group, as opposed to a few visitors. I would judge I have waited several years (year by year, in a sequence) to be completely natural but I paid several pounds, outright ungrammatical.

  99. Marja Erwin says

    I don’t usually mix a couple-several with a few-many, because to me “a couple” and “several” are the same in all contexts– 4 can be several and 5 is always several– while “a few” and “many” depend on the context.

    An exception is a few hundred-several hundred, a few thousand-several thousand.

  100. Oh? Even, say, with the word times (e.g. “I’ve been to Paris many times.”)?

    Yup. I might say it, but it would be with the conscious adoption of a formal register, and I would sound stilted to myself.

  101. Positive many is not as formal-sounding as positive much (in the sense of a lot of), but I can’t think of many contexts where I’d use it in speech. The same is true for positive few (as opposed to a few) and even more for little (in the sense of not a lot of).

  102. Yeah, those are my judgments too.

  103. To me, several retains some of its distributive sense

    Yes, I agree, & this, I think, is the reason that “several friends told me they might not be going” and “several cats have been hanging around in the back yard” sound fine to me but not “I’m having several friends over” or “they have several cats”.

  104. “Being six years old, she refuses to believe that her ideas about language come from anywhere. When we differ, it is simply that she is right and I am wrong.”

    This is (sadly) completely normal for people of all ages.

  105. When I saw Cats, Mungojerrie and Rumpelteazer were explicitly a romantic “couple.”

  106. In Irish cúpla can mean “a few”, like cúpla duine “a few people”. However when used of people Is cúpla iad doesn’t mean “They are a couple” as in English, it means “They are twins”. The constellation Gemini is An Cúpla.

  107. Athel Cornish-Bowden says

    When I was a child I was sure that “a couple” was a smallish number, but not necessarily two. I had a fierce argument about it when I was about ten, and was quite cross to learn that I was wrong.

    “Several” has always meant “about seven” to me, probably on account of the similar sound.

    In English “pair” always means exactly two. In Spanish, however, they talk about un par de semanas, meaning a couple of weeks, but not necessarily exactly two (at least, they do in Chile).

  108. Positive many is not as formal-sounding as positive much (in the sense of a lot of), but I can’t think of many contexts where I’d use it in speech. The same is true for positive few (as opposed to a few) and even more for little (in the sense of not a lot of).

    Interesting. Much + noun in an affirmative sentence seems “formal register” to me, but none of the other words do. Perhaps I habitually speak more formally than most.

  109. Laowai. the most unusual of those for me is little, which I’m not sure I’d use in any register. Would you really say something like “I have little money”?

  110. Not I. But I would say “We know little of this.”

  111. Laowai. the most unusual of those for me is little, which I’m not sure I’d use in any register. Would you really say something like “I have little money”?

    Not without further elaboration (i.e explaining the reason why I have little money at the moment). However, I’d certainly say, for example, “I have little patience for fools.” In fact, I said that very sentence during a meeting last week.

  112. Eli Nelson says

    Geoff Pullum just made a Language Log post about the polarity of “few” in “Lydia knows few things”: Accentuate the negative. It seems likely to me that the disagreement between different speakers is at least partly a result of the rarity of the construction.

  113. “Several” has always meant “about seven” to me, probably on account of the similar sound.

    Huh, cool. I don’t have that association in my mind, but it kinda makes sense.

  114. An addendum: I have a number of widely-used British and American ESL textbooks in my library, so I pulled down a handful to see what they have to say about much/many/little/few in affirmative sentences.

    All of them note “Much is usually not used in affirmative sentences, especially in spoken English,” or words to that effect. None of them make any such suggestion about many, few or little. On the contrary, they all provide numerous examples of such affirmative usages, all of which sound perfectly natural to me (e.g. “We need to hurry. We have very little time left.” “Few people speak more than three languages.”)

  115. David Marjanović says

    I think very little is more common than little, though, and likewise with few, much, many.

  116. I would say “We know little of this.”

    I would say it only in a moment of whimsy, deliberately adopting the manner of a white-coated scientist in a B movie.

    I’d certainly say, for example, “I have little patience for fools.”

    That doesn’t sound quite as absurdly formal to me, but I still wouldn’t say it unless I was for whatever reason adopting a formal manner. Not part of my everyday speech pattern.

    On the contrary, they all provide numerous examples of such affirmative usages, all of which sound perfectly natural to me (e.g. “We need to hurry. We have very little time left.” “Few people speak more than three languages.”)

    Interesting; if I encountered such a textbook, I would shake my head and think “They should do a better job of distinguishing spoken from written registers — their students are going to sound like books.” But clearly there are English-speakers who speak that way.

  117. if I encountered such a textbook, I would shake my head and think “They should do a better job of distinguishing spoken from written registers

    If there’s an ESL textbook that does a good job of teaching colloquial English, I’ve never encountered it in the course of a long teaching career flitting in and out of ESL.

    However, as I’ve implied, for me this isn’t a question of register, as the expressions aren’t “written register” to me at all. Of course, we all have peculiarities of dialect related to region, social class, education, etc., but it doesn’t seem like this could be anything idiosyncratic to my native English. Which is to say, a textbook should be extremely wary of proscribing something as “written English” when there is a group of native English-speakers who routinely use it in spoken English. When confronted in a textbook with a usage that is not my own (typically an obscure Briticism) I long ago learned to say “Well, I wouldn’t say it that way, but some do.”

  118. Which is to say, a textbook should be extremely wary of proscribing something as “written English” when there is a group of native English-speakers who routinely use it in spoken English.

    I couldn’t agree more. I just hadn’t been aware until this thread that there was such a group.

  119. vrai.cabecou says

    How much is “a number of”? Is it the same as “several”? More? I always want to change it to save a little space.

  120. Yes, that’s an odd phrase, isn’t it? Everybody uses it, but I doubt anyone could define it.

  121. @David:
    > Methinks you’re both outside the Standard Average European Sprachbund, and Danish is in

    Interesting! Is there a reason why this feature would be likely to be affected by the Sprachbund?

    And yes, I’m painfully aware that I’m probably mapping it mentally to Danish “flere”. However, assuming based on the judgment of several native speakers that I was completely wrong, that still doesn’t explain why “only several” is strange, since “only 5 to 10” (or whatever you map “several” to) is not. If it’s because of its distributive sense, as @Piotr has suggested, then I would say that the sentence

    > He said that one of the officers, while in detention, had kicked his ankle, where he had surgery several years prior.

    should still be strange, since we’re not talking about a “year by year, in a sequence” scenario, but just a chunk of time.

  122. The constitution of the USA speaks of “the several states”, which can jar a modern reader, and not just because fifty seems like more than several.

    When the US constitution was drafted, there were only thirteen states in the USA.

    Would thirteen count as several today?

  123. David Marjanović says

    Is there a reason why this feature would be likely to be affected by the Sprachbund?

    Why not? One of the defining features of a Sprachbund is that constructions with similar meanings in different languages end up having exactly the same meaning.

  124. Would thirteen count as several today?

    As pointed out somewhere above, several here is very clearly the original adjectival meaning of the word (“separate; individual”) not an indefinite determiner.

    It surprises me that anyone thinks of several as more than five or six, at the most, but judging from this thread some people do. Personally, I only ever use the word to refer to a number that is more than two, but very small. For me, it’s exactly synonymous with a few.

  125. David Marjanović says

    Conversely, I’m surprised by all these precise number ranges so many people have given. Like ajay and the Pirahã, I’d adapt these ranges to what is being counted.

  126. Human eye perceives number of items less than four immediately (without need to count).

    So “few” or “several” must actually mean something more than four, there is no need to use the word otherwise.

    “Few Men in a Boat (to say nothing of the dog)” sounds weird, doesn’t it?

  127. “Few Men in a Boat (to say nothing of the dog)” sounds weird, doesn’t it?

    Only because it’s incorrect English. Since you’re expressing a positive idea, the indefinite article is mandatory: “A Few Men in a Boat”

    And the entire point of the words several and few can be summed up as “the number is small, but the exact number is not important.” If the number were in any way significant (as it is in the Jerome story) then, naturally, an exact number would be used.

    I reiterate: for this native speaker a few/several is invariably six or less, most often three to five. If I had seven or more of something, I’d choose a different expression (some, many, a number of, lots) Likewise, if someone told me they’d been to London “several times” and it subsequently turned out to be 10, I’d wonder why they’d been dissembling.

  128. So “few” or “several” must actually mean something more than four, there is no need to use the word otherwise.

    Your “must” contradicts the facts.

  129. Original hominid counting system:

    One, two, a couple, several/few, many

  130. I’ve been trying to resist adding Danish data, but seeing that Dainichi mentioned flere:

    Danish doesn’t have the ‘couple’/’pair’ distinction, par applies to shoes as well as lovers. But to escape the precision of et par we say et par stykker (‘a couple of pieces’ literally). Also adskillige shares (calques?) the original sense of separability and the current sense of “not a lot but not as few as you might expect either” that I get for ‘several’.

    For flere, note that Danish has a strange inversion in the pragmatics of scalar flexion, if that’s the right phrase — the comparative is used where the positive is too large! En større by is not as large as en stor by, though it cannot be very small, and I will become en ældre mand before I’m gammel. Similarly, flere is more than a few, but not many. (This doesn’t apply when comparing two sets or objects, though, if I have flere penge than you I do have the most money).

  131. How much is “a number of”? Is it the same as “several”? More?

    I think the phrase contains a silent “significant”. Which is to say: if I say that there is a number of X, what I mean is “there are enough X that you should pay attention to them”. “There’s a number of reasons why not”. “A number of people disagree with you”. (“A few people disagree with you” implies “but you can probably ignore them”.)

    Hence, I suppose, “a number of elephants in my back garden” would be one or more, as there is no number of elephants so small that I would not feel compelled to pay attention to them. But “a number of British voters in favour of unilateral disarmament” would have to be a few million.

  132. I think the phrase contains a silent “significant”. Which is to say: if I say that there is a number of X, what I mean is “there are enough X that you should pay attention to them”.

    Makes sense to me.

  133. “Few Men in a Boat (to say nothing of the dog)” sounds weird, doesn’t it?

    Which implies some other literary classics:

    Some Feathers
    The Lots Of Nights and One Night
    A Tale of a Couple of Cities
    Fahrenheit Many
    Slaughterhouse-Handful
    A Number of Musketeers
    Catch-Lots
    One of a Few Men

  134. And Russian classic “Several chairs”

  135. I will become en ældre mand before I’m gammel.

    English has exactly the same absolute use of the comparative as a lesser degree than the positive: an older man is not as old as an old man. Note that suppletion is not required.

  136. Suppletion just happened in the examples I thought of. Does it extend to more vs many, though?

  137. for this native speaker a few/several is invariably six or less, most often three to five.

    This doesn’t work for me, at least not in all circumstances. Say you ask a girlfriend or wife coming back from party “Were there many there?”. If she said “Oh, there were a few people there but I only knew a couple of them.” “A few people” could easily mean 15-20, maybe a lot more depending on the size of the venue, and “a couple” could be 3-4.

  138. Yes, I suspect this is all much more context-dependent than would seem apparent at first glance.

  139. J.W. Brewer says

    For me at least there’s a marked divergence between few and several in the specific situation of a number that is pretty small in percentage terms in the relevant context but not small (don’t know about <7 as the absolute max but that's in the right neighborhood) in absolute terms. Hundreds of X's could plausibly still be characterized as "a few X's" in a context involving ten thousand or more X's in total, but could not idiomatically be characterized as "several X's." At least not for my particular native-speaker ears.

    Note, by the way, that one of the ways in which the usage of "several" in the U.S. constitution is a different and archaic sense is that even when there were only 13 states, "the several states could mean 13 out of 13," so the fairly strong implicature of <many, and thus necessarily <most and <all, that the more modern sense of several shares with few is totally missing.

  140. I should again point out that the number indicated by “there were a few people there” could vary wildly with intonation and delivery.

  141. J.W. Brewer says

    I don’t disagree with Bathrobe but surely most if not all English sentences can be given wildly differing meanings with sufficient variation in intonation and delivery?

  142. laowai: “if someone told me they’d been to London “several times” and it subsequently turned out to be 10, I’d wonder why they’d been dissembling.”

    Ten is probably a decent estimate of how many times I’ve been to London, and I wouldn’t feel at all shifty in telling you I’d been there “several times”. In fact I can’t think of a better way of saying it without exact numbers: “many times” would feel like an exaggeration, and “a few times” would be downplaying it. But context and expectations are definitely involved, for me: ten seems to me a respectable if unremarkable tally for someone living across the Irish Sea, so “several” is about right. For an Australian, I imagine ten would merit “many times”. For someone from Slough, I’d expect ten to translate into something like “only a few times”.

  143. Which implies some other literary classics:

    There are also films in this category:

    The Magnificent Several
    A Number of Angry Men
    Quite a Few and a Half

  144. Yes, I suspect this is all much more context-dependent than would seem apparent at first glance.

    Words that should be written up over every linguistics classroom and professor’s office.

  145. The Bunch of Commandments.

  146. marie-lucie says

    The meanings of French plusieurs

    In France this word is approximately equivalent to English several. The TLFI traces it to a late Latin form plusiores, formally a comparative form of plus ‘more’, so literally ‘the more mores’, meaning ‘the larger part of (a number)’, hence the common phrase plusieurs d’entre (a number, a group) ‘several of/among …s’.

    In francophone Canada the meaning of the word seems to have considerably expanded.

    I once discussed the topic with a colleague from Québec as I was not quite sure of the range of plusieurs. As an example I used the situation of going to a place where a number of random people are expected, such as a waiting room, entrance hall, theatre, or similar places where one might encounter mostly strangers. If I said je croyais être très en advance, mais il y avait déjà plusieurs personnes ‘I thought I was very early, but there were already several persons’, how many people do you think there were? Personally, my own answer would have meant ‘between five and eight, certainly no more than ten’, but he thought there must have been around twenty-five people.

  147. Say you ask a girlfriend or wife coming back from party “Were there many there?”. If she said “Oh, there were a few people there but I only knew a couple of them.” “A few people” could easily mean 15-20, maybe a lot more depending on the size of the venue, and “a couple” could be 3-4..

    Well, not my partner. Completely impossible. But, then, we extremely seldom attend social gatherings with more than a dozen participants. For us, twenty party-goers would amount to a “huge” party, at which there were “a bunch” (lots, loads) of people.

    Breffni’s hypothetical about distance from London works far better for me in suggesting the contextual slipperiness of the words, though I don’t think it applies to my own usage of the words. Which is to say, when I say “a few,” I mean literally three to six of X, and nothing else, no mater what the context. I do understand, however, that others use the terms more loosely and/or metaphorically to refer to “a small number relative to expectation.”

  148. bgibson135 says

    My girlfriend and I are a couple. Are you going to ask me who the third or fourth person is in our relationship?

  149. David Marjanović says

    That’s its meaning as an ordinary noun, not as a measure word.

  150. I meant to comment on something several years ago, but I forgot. I don’t think this is quite the thread I meant to comment in either, but as I cannot seem to locate the correct one, I am leaving this here.

    In the last section* of Stormbringer,** Elric, upon his return to the dreaming city (his own capital, which he destroyed in the first story Moorcock wrote about him) dreams of the city in its heyday:

    And across the azure sky lazy dragons flapped their way toward the caves where thousands of the beasts were stabled, unlike the present where scarcely a hundred remained.

    To me, “scarcely a hundred,” “barely a hundred,” or similar constructions indicate that while there are at least a hundred, it is probably not much over that and, in any case, less than might be expected. The part about being less than expected clearly applies in this instance, but I was surprised when, a few pages later, Moorcock mentioned that the precise number of dragons was actually only ninety-eight.

    * The title is “Doomed Lord’s Passing,” the least evocative of the four possessives Moorcock used as titles for the four novellas. The others are: “Dead God’s Homecoming” (the best of the bunch), “Black Sword’s Brothers,” and “Sad Giant’s Shield.” I’m not sure whether Moorcock may have changed titles around shortly before publication though. A teaser for the fourth novella in a previous issue of Science Fantasy (which published all four) refers to the climactic story as “Doomed Lord’s Prayer.” This might indicate a change by Moorcock, or it might just have been a production error. (With interference from “the Lord’s Prayer”? Perhaps the compositor at Science Fantasy was a devout high church Anglican?) Moreover, in the table of contents of the issue where the finale actually appeared (but not on the cover or the first page of the story), the title was “Dead Lord’s Passing,” perhaps confusing Elric with Darnizhaan from the first of the four novellas.

    ** In which Elric’s story comes to its conclusion, and which, quite frankly, should equally have been the last thing Moorcock wrote about the character. Even in the two original fix-up books, there are already sections that are practically unreadable (e.g., “Kings in Darkness”), and things generally get worse from there.

  151. David Marjanović says

    The TLFI traces it to a late Latin form plusiores, formally a comparative form of plus ‘more’, so literally ‘the more mores’

    …and obviously this plus-ior-es is what mehr-er-e is calqued from. I had long wondered why it looked like a comparative of a comparative!

    The part about being less than expected clearly applies in this instance, but I was surprised when, a few pages later, Moorcock mentioned that the precise number of dragons was actually only ninety-eight.

    I interpret such wordings as “at most 100 or very slightly above, probably a bit less”, so I wouldn’t have been surprised.

    (I remain continually surprised that there are people who have de facto numerals for ranges like “3 to 6” out there.)

  152. The Piraha counting is discussed at length by a palaeo-anthropologist
    https://johnhawks.net/weblog/numbers-are-a-cognitive-technology/

  153. That article has two AI-generated illustrations. Eccch.

  154. I only noticed that AI-generated images are a thing some two weeks ago:(

  155. On the original subject, I’ve always thought of “a couple (of)” as meaning two. I am aware there are others that use it more loosely, but didn’t realize how common that is.

    Looking at examples in Wiktionary, “a couple (of) miles” or “a couple (of) minutes” as not meaning precisely two don’t bother me…I don’t expect that to be exact. But “a couple (of) friends” for anything other than two friends seems wrong to me. Or at least imprecise.

    I’ve never thought about the distinction between “a couple” and “a couple of”. I find it sounds odd to me with the “of”. Thus it’s interesting that Wiktionary claims that “a couple of” means two, but can also, informally, mean a small number, but “a couple” (without of) only means “a few”.

    Merriam-Webster, in it’s noun definitions, has it meaning a pair (one definition) or “an indefinite small number” (few) (a separate definition). Which, suggests it means two when the two are a pair (such as one of the other definitions). The adjective (that is, without “of) it has as primarily meaning “two”, but also “a few”.

  156. I used AI-generated illustrations for two out of 3 most recent entries on my translation blog.
    My gut feeling is that poetry needs no illustrations, but it’s a customary thing to add at least one image everywhere (except LH!! which is great IMVHO). And it’s probably more appropriate to play with turning lines of verses into AI-generated illustrations than to try finding someone else’s image loosely fitting the topic & hope that the rightful owners don’t mind…

    What do you think?
    http://letras-de-tango-en-ruso.blogspot.com/2022/12/ojos-muertes.html
    http://letras-de-tango-en-ruso.blogspot.com/2022/12/ninguna.html

  157. Don’t like them. They look like artificially-flavored drinks taste.

  158. Stu Clayton says

    I like them well enough. The temperature of the images matches that of the lyrics.

  159. I didn’t like them, but don’t worry, there is no taste in me (as a missionary told a cannibal). The problem is that the poems (in their Russian version) have a character and verve and images have only temperature. Stu, did you make a pun on stable diffusion or am I imagining things?

  160. Sometime in the late 1960s or the 1970s I visited a friend in San Francisco. One day, she pulled into a filling station and asked the attendant for “a couple of dollars’ worth.” Under my breath I said to her, “Tell him exactly how much you want.” Her response was, “I have.”

    It turned out that in San Francisco the word still had its original meaning, at least when prides were concerned.

  161. “a pun on stable diffusion ”

    That’s how I learned that AI-generated images are a thing (the news that Kickstarter kicked Unstable Diffusion out)

  162. Athel Cornish-Bowden says

    When I was about ten or younger I was convinced that a couple was just a small number, and was quite surprised when someone insisted that it meant two. Since then I’ve adopted the “two” meaning.

    I’ve always been a little surprised that in Spanish “a couple of weeks” normally comes out as “un par de semanas”, i.e. it doesn’t restrict “par” in the way English does with “pair”, to things like socks and shoes that constitute a natural pair.

  163. That’s what pareja (and related terms) are for, innit? (Granted, the way Spanish splits the semantic space between that and par isn’t the same as English pair vs couple, but all the relevant distinctions can be expressed nonetheless.)

  164. Dmitry Pruss says

    The problem is that the poems (in their Russian version) have a character and verve and images have only temperature.
    Really, you think that an image accompanying a verse must transmit all aspects of the poem’s emotional, mental, rhythmic and alliterative twists 🙂 ? Ditto book cover for a book or an album jacket? It’s just an eye catching little vignette IMO. If it isn’t a distraction and doesn’t lead you astray while catching your attention, then it works as intended

    (And thanks for the nice words about my translation, it’s appreciated 🙂 )

  165. Ok, then. It doesn’t distract me because if I didn’t look for them especially, I wouldn’t notice 😉

  166. David Marjanović says

    They look like artificially-flavored drinks taste.

    “Green juice, extra green!”

    “It tastes… green.”

    I’ve always been a little surprised that in Spanish “a couple of weeks” normally comes out as “un par de semanas”, i.e. it doesn’t restrict “par” in the way English does with “pair”, to things like socks and shoes that constitute a natural pair.

    See above for ein Paar “a pair; a married or similar couple” vs. ein paar “a few; definitely not exactly two except by rare accident”.

  167. Unfortunately, English pair is not only socks, and Englsih couple is not just a number:(
    Twisted pair, galvanic couple, this loving couple makes a great pair.
    Force couple, kinematic pair.

    P.S. no, “makes a great pair” does not sound great but “great pair” and “loving couple” both exist.

    P.P.S. but of course, to native speakers who want to be able to distinguish several items and a pair of socks, a pair of socks and a loving couple, a loving couple and a great twisted pair, English provides the necessary means:)

  168. Athel Cornish-Bowden says

    That’s what pareja (and related terms) are for, innit? (Granted, the way Spanish splits the semantic space between that and par isn’t the same as English pair vs couple, but all the relevant distinctions can be expressed nonetheless.)

    I know how to use pareja, but I didn’t say that I didn’t understand, but only that I was “a little surprised”?

    A more confusing question of dividing up the semantic space occurs between French and Spanish, where the distinction between pour and par in one is not exactly the same as the distinction between para and por in the other.

  169. I didn’t mean to imply any slight, just checking that my intuition actually applied to varieties beyond the ones I’m fluent in. (I would not be surprised at all if the par/pareja distinction had gone the way of the dodo in some dialects.)

  170. John Cowan says

    Wikt defines pair of stairs as ‘flight of stairs’ and labels it British and dated; I’ve seen it in literature.

  171. If you speak about a “couple” that is getting married or in love, you don’t assume there are 2, 3 or maybe 4. You know it means 2. A couple means two, not a few or several. IT,S TWO!

  172. When I speak of a couple that’s getting married, it’s 2. When they say they plan to have a couple of kids, they mean somewhere between 2 and 4. When they say it’ll take them a couple of years to get established before they have the first one, I assume they mean anywhere up to 5.

  173. Yes, it’s all a matter of context.

  174. David Marjanović says

    If you speak about a “couple” that is getting married or in love, you don’t assume there are 2, 3 or maybe 4. You know it means 2. A couple means two, not a few or several. IT,S TWO!

    For a lot of people, as explained above, the similarity of a couple and a couple of is accidental – they’re two separate expressions that mean two separate things: “a coupled pair” and “a small number”.

  175. @David Marjanović: I can’t imagine anyone thinking the similarity there is truly accidental. The two terms are obviously related. Rather, insisting on restricting a couple to two on that basis would be an etymological fallacy.

  176. David Marjanović says

    Fair enough, but it actually surprised me how obvious the etymology seems to be to everyone. With Paar and ein paar, that is not the case.

  177. J.W. Brewer says

    FWIW I just came upon the random fact that (according to a secondary source which I am assuming is reliable) C.P.E. Bach stated in his 1773 autobiography that he had composed “ein Paar Duzend Sinfonien” (“a coupla dozen symphonies”). However, only 18 of his symphonies, including 4 composed after 1773, are known/extant today, so this suggests that some significant number have been lost. How many were lost? Well, that depends on what range of numbers you think “ein Paar Duzend” might have encompassed as and when Bach used the phrase.

  178. I pretty much take for granted that a lot of music composed in the 18th century has been lost — there wasn’t the institutional/commercial structure that ensured preservation from the 19th century on.

  179. Stu Clayton says

    a coupla dozen symphonies

    I like his nose-to-tail attitude towards creativity. It is also mine:

    Weil ich meine meisten Arbeiten für gewisse Personen und fürs Publikum habe machen müssen, so bin ich dadurch allezeit mehr gebunden gewesen, als bey den wenigen Stücken, welche ich bloß für mich verfertigt habe. Ich habe sogar bisweilen lächerlichen Vorschriften folgen müssen; indessen kann es seyn, daß dergleichen nicht eben angenehme Umstände mein Genie zu gewissen Erfindungen aufgefodert haben, worauf ich vielleicht ausserdem nicht würde gefallen seyn.

    “The practice of nose-to-tail cooking or eating means ensuring that every edible part of an animal is prepared and consumed, as opposed to using only the high-value cuts.”

  180. Stu Clayton says

    The sound of a harpsichord – two skeletons copulating on a tin roof in a thunderstorm.
    (Sir Thomas Beecham)

    [Source: Klaviergeschichte]

  181. I pretty much take for granted that a lot of music composed in the 18th century has been lost

    How did Bach’s umptillion compositions get preserved so completely?

  182. Stu Clayton says

    What risk was there of their not being so preserved ? Umptillion compositions are no hindrance to preservation. Just print them and pile them on a shelf. CPE was more popular than his father, says the Bach Archiv, so perhaps there was more demand for dead tree versions in every discerning home. Not just preservation, but redundancy as well. The père rose to glory only long after his death.

    Als hoch geachteter und verehrter Komponist, dessen Berühmtheit die seines Vaters bei weitem in den Schatten stellte, verstarb Carl Philipp Emanuel Bach am 14. Dezember 1788 in Hamburg.

  183. @Y: They didn’t. I have seen estimates that as many as half of J. S. Bach’s works have been lost. However, that O(10³) scores did survive probably has a lot to do with the fact that two of his sons were also prominent classical musicians and composers. Although their father’s music was not performed extensively in the decades immediately after his death, both C. P. E. Bach and J. C. Bach were committed to his legacy and worked to promote his compositions. Hundreds of original manuscripts remained in the hands of Bach family members, until interest in the Baroque masters picked up again in the nineteenth century.

  184. Stu Clayton says

    A victory over the initial Boring Baroque Reponse.

  185. How did Bach’s umptillion compositions get preserved so completely?

    As Brett says, they didn’t. I didn’t say everything got lost, did I? But a lot did.

  186. I guess when you start with two umptillion…

  187. J.W. Brewer says

    Mass-published sheet-music became a big business in the 19th century, but there were certainly styles and genres of music it tended not to capture so lots of stuff went undocumented and unpreserved until the rise of recording technology, with it not being until after 1920 that (in the U.S., things were probably slower elsewhere) some doughty capitalist hustler was trying to capture absolutely any sort of music that might potentially sell, even if to a niche audience. One great example is the enigmatic Buddy Bolden, whose public performing career ended abruptly in 1907 when he was relegated to the Louisiana State Insane Asylum (perhaps better than being sent to prison, which was the other potential realistic option). He may have done as much as any other single individual to invent jazz, but we know nothing about what he sounded like other than decades-later testimony from interviews with the first generation of New Orleans jazzmen who actually recorded in the 1920’s, who pretty uniformly talk about how mind-blowing and transformative it had been the first time they heard him play. (Over the decades there have been intermittent rumors that someone had unearthed an Edison cylinder or something with a faint/scratchy trace of Buddy playing. Those rumors always turn out to be false, but people want them to be true.)

  188. @Brett I have seen estimates that as many as half of J. S. Bach’s works have been lost.

    I think that claim is out of date. (Although I wish I could cite where I saw it rejected.)

    His publisher was producing a ‘collected edition’ of keyboard works at Bach’s death. That’s where the tantalising unfinished last contrapunctus of the Art of Fugue would be printed. Bach supervised the collection closely, so if any pieces were omitted that’s more likely because he thought them not worthy. wp has

    Bach collected his cantatas in annual cycles. Five are mentioned in obituaries, three are extant.[42] Of the more than 300 cantatas which Bach composed in Leipzig, over 100 have been lost to posterity.

    Firstly that’s not “as many as half”.

    But later research has concluded Bach recycled his works (as did most composers of the time[**]), such that most ‘lost’ cantatas were in fact earlier versions of the ones we have. wp also has

    Bach composed and transcribed concertos for one to four harpsichords. Many of the harpsichord concertos were not original works but arrangements of his concertos for other instruments, now lost.[173] A number of violin, oboe, and flute concertos have been reconstructed from these.

    YMMV as to what “lost” means.

    [**] For example many of the P`s&F’s of the second book of the 48 are earlier ad-hoc compositions, in some cases transcribed to a different key. There’s one where the P is registered in flats but the F in sharps (they’re in the same key because Well-tempered), and because the P was an earlier composition. cite: a Youtube video I watched a month or so ago.

  189. @AntC: Indeed, half is the highest estimate I have ever seen for how much of J. S. Bach’s output was discarded. Moreover, it does seem very likely that a lot of the discarding was done by the composer himself. As you say, he rewrote and recycled a lot of his works; sometimes he kept the both sets of scores, and sometimes he tossed the the older ones. Some compositions he discarded without even trying to rework them, presumably because he was fundamentally unsatisfied with them. The result is that the works that he personally thought deserved to be preserved have probably almost all survived.

  190. David Marjanović says

    “A Russian BMP was hit, with infantry on top. I suspect that out of 9, only several made it out alive.”

    halfway down this post (not counting the comments); none of the currently 9 comments is about this; judging from the video that follows, “several” cannot mean “about 7” here, but has to be much less, more like 2

  191. That sentence is weird/non-idiomatic. As if from a non-native speaker. I’d say “… only a few made it out alive.”

    Without re-litigating the discussion here above, for me ‘several’ means more than 2; but yes less than 7. OTOH it’s vague (fog of war), so I wouldn’t be greatly surprised at either figure. Also possible the sentence got drafted with the opposite polarity, then was swizzled into nonsense.

  192. David Marjanović says

    Continued adventures with numerals: “in about two handful places“, apparently using “handful” the same way as “dozen” in ongoing resistance to decimalization.

  193. to me, “a couple [of]” has always meant two, three, or maybe four of something. it’s indefinite. i think “a couple [of]” is distinct from the usage of the word “couple” as in, “the happy couple”. that usage has a distinct meaning equivalent to “a pair”.

    when you ask someone if you can have a couple of their french fries, are you *really* going to eat exactly two, and not a single fry more? didn’t think so.

  194. Continued adventures with numerals …

    I’m not seeing the point of continuing. Ragnar Bjartur Gudmundsson is either a non-native speaker, or writing in some sort of military shorthand. I don’t think they’re aware enough or deliberate enough with the subtleties to impute motives like “resistance to decimalization”.

    Apart from that plain number crunching exercise done we may well expect …

    Also is “barbells” a typo for ‘barrels’? If not, what is the Russian Army using barbells for on the battlefield?

  195. when you ask someone if you can have a couple of their french fries, are you *really* going to eat exactly two, and not a single fry more? didn’t think so.

    If couple means ‘two’, then it is not the word you would use under those circumstances. (Except under special more specific ones, such as if you wanted one each for two dogs you were walking.)

  196. Trond Engen says

    David M. (quoting): I suspect that out of 9, only several made it out alive.

    Misguided elegant variation from ‘some’?

  197. David Marjanović says

    Ragnar Bjartur Gudmundsson is either a non-native speaker, or writing in some sort of military shorthand.

    Whoever Torska is, they aren’t Bjartur and aren’t quoting him in that passage, just citing him as a source for the estimates.

    Besides, it would still remarkable enough if coming from a nonnative speaker; eine Handvoll in German, for example, will certainly not be understood as “five”, and *zwei Handvoll is nonsensical. (Like… I’d have to come up with a very strange recipe saying you’re supposed to scoop salt/sugar/flour with your hand twice.)

    I don’t mean “resistance to decimalization” as a conscious motive; I mean decimalization of the way people think is slower to catch on than I thought.

    (…Entering “zwei handv into the Google search field does bring up a lot of suggestions that start with zwei Handvoll, but they all look like song or film titles where poetic license applies.)

    Also is “barbells” a typo for ‘barrels’?

    I’m sure it’s autocorrupt for a typo of barrels. In fact, I just typed barrells; I’ve switched the autocorrupt off, so I don’t know if the one in my browser would turn that into barbells, but it wouldn’t surprise me.

    Misguided elegant variation from ‘some’?

    I wouldn’t expect that in this context; there’s no some, or similar, pretty far up and down from that line (in a sequence of one-liners separated by embedded tweets).

  198. J.W. Brewer says

    I think a bunch of these words may work differently (in terms of what specific numerical range of X’s they can correspond to) when the total number of X’s in the entire group is made explicit or at least has a clear likely upper bound. “Several children in the class” did such-and-such probably has a higher upper numerical bound than “several of my children” does, at least if one has certain default expectations about the typical range of total children in a class versus a single family.

    I myself have five children. I think “several” of them could be three but couldn’t be four. Five is probably too small a total number for “several” to cover a different range than “a couple of.” In the combat example quoted above, “only several” creates a further complication. I could see four plausibly being “several” out of a total number of nine, but not “only several,” which probably maxes out at three. It may also depend on what predicate is being associated with the “several.” If it’s a less dramatic predicate I think as noted above three out five could be several but if it’s as dramatic as surviving some sort of potentially-fatal catastrophe “several” probably can’t be a majority of the total, even a bare majority.

  199. “Several” has the connotation of a subjectively larger number, though low on that hierarchy, below e.g. “quite a few” and “many”; thus “I’ve told you several times to pick up your socks” would not work with “a few” substituted. I think that’s why “only several” doesn’t work.

  200. J.W. Brewer says

    It is certainly true that “only several” is much rarer (per the google books ngram viewer) than “only a few,” but one can find instances of the former (in this usage – there are also false positives for other constructions with the same bigram sequence) in apparently edited books that are not obviously ESLish glitches. E.g.: from a discussion of why the more fervently anti-clericalist figures in Third Republic France opposed women’s suffrage: “[Leon] Richer argued that women needed extensive education to be weaned from the ‘black peril’ of the clergy. ‘Among nine million women who have attained their majority, only several thousand would vote freely; the rest would take their orders from the confessional.'” That’s obviously translated from something Richer said in French, but I don’t think the translation is unacceptably unidiomatic.

    Or in speech from some 1948 Congressional hearings on “Communist Espionage”

    Q. Alger Hiss? [meaning in context how many times did you meet with him]
    A. I have said only several times.

    [The witness’ testimony about Hiss a few pages earlier was “I have never met him except to bump into him in some other company I should say two, three, not more than four or five times.” ETA: the witness was this guy: https://en.wikipedia.org/wiki/Frank_Coe_(government_official) ]

  201. PlasticPaddy says

    Part of the difficulty with only is that only + [AUGMENT] works only(!) in certain cases, i.e., he was “only a little (bit) tired” and “only very tired” are not really interchangeable, e.g., after “After running up the hill” (I suppose some English could start with “Only he was very tired” but the “he was only very tired” works for me only! after something like, “he was not being rude on purpose.. “)

  202. J.W. Brewer says

    The function of the “only” in the “only several X’s” examples is make it signal “not enough X’s to really amount to a significant number in context.” Whereas by contrast plain old “several X’s” could well be enough X’s to be significant. “Only a few X’s” would send the same signal of non-significance. Maybe “only several X’s” could mean “arguably more than ‘a few’ X’s but still not enough X’s to be significant in context so let’s not waste time on that irrelevant argument”?

  203. only several thousand

    I have said only several times

    I don’t think either of those count. “Only several thousand” is much less unidiomatic than “only several” tout court, and in the latter I think the “only” modifies the whole phrase: only [several times], not [only several] times (where “several” corresponds to “two, three, not more than four or five”).

  204. J.W. Brewer says

    I don’t necessarily disagree w/ hat’s bracketing, but I’m not sure where he’s going with it. Wouldn’t e.g. “only a few times” likewise be [only [a few times]] not [[only a few] times]? Put another way: don’t count toward what?

    In the clunky-sounding example upthread, surely it means the same thing as “only several of them made it out alive.” Does it sound less unidiomatic if the “of them” isn’t elided?

  205. Hmm. Now I’m confused.

  206. David Marjanović says

    That’s obviously translated from something Richer said in French

    …and I bet the original has quelques, which could be translated about equally well as “a few” (at least without more context), rather than plusieurs.

Speak Your Mind

*