If Then.

As regular readers know, I am a staunch descriptivist; that is, I hold to a scientific, fact-based description of language: whatever native speakers of a language say is ipso facto correct (unless, of course, they make a slip of the tongue). It’s absurd, for example, to claim that the way the vast majority of English-speakers use “beg the question” is incorrect and the only acceptable usage is an obscure philosophical one that makes no sense in English (since it’s a literal translation of the Latin petitio principii; see this LH post). I deplore prescriptivism, the idea that there is some Platonic ideal of language that is eternally “correct” and that all deviations from it must be chastised and if possible forbidden, and I mock the peevery it inspires.

And yet I myself am human, all too human, and I have my inner peever who occasionally demands to be taken out for a walk. So just as back in 2003 I complained about contrary-to-fact past may have (“if he’d run faster he may have caught the ball”) — which still bothers the hell out of me — I now feel compelled to gripe in public about another bête noire, the use of the if … then construction in something other than a conditional statement (“If you get good grades then you will get into a good college”). This is prompted by having read a particularly grating example of what I cannot help but call misuse in Annette Gordon-Reed’s 2017 NYRB review of Geoffrey R. Stone’s Sex and the Constitution: Sex, Religion, and Law from America’s Origins to the Twenty-First Century, currently available online here; she writes:

For Luther it was celibacy that was devilish, while sex was as “necessary to the nature of man as eating and drinking.” If the early Christians had especially dire views on these matters, then Protestantism, which has been the dominant religious tradition in America from the beginning, rejected those views.

I couldn’t understand the second sentence until I went back and reread it, so alien is that usage to me. It’s bad enough to use “if” in a vague non-conditional way, as a sort of equivalent to “while,” but I’m more or less used to that; the problem is that people are thereby tempted to follow it with “then,” which to me directly and ineluctably implies a conditional statement. Stop it, all of you! Use English in a way that makes sense to me!

Whew. OK, having gotten that off my chest, I now return LH to its regular schedule of calm, scientific discussion of language and related matters.

Addendum. J.W. Brewer, in the comments, mentions Mark Liberman’s 2004 Log post on concessive (bleached? conditional?) “if,” with examples back to 1965; he adds:

Significantly, however, the examples given of that usage of “if” tend not to be followed by a “then,” which might mean that using the “then” with an other-than-truly-conditional “if” really is a more recent innovation calculated to aggravate those who weren’t acclimatized to it in their formative years. Indeed, Prof. says regarding the specific example that had prompted the post: “In fact Wilgoren & Justice don’t use ‘then’ — it would have been weird if they had.”

Comments

  1. Stu Clayton says:

    I strongly sympathize, and would encourage you to take up the cudgels again as soon as possible. One time that I encountered that kind of “if…then”, I thought “this was invented by an American journalist, and is now taught in journalism schools”. Don’t know why I think that, nor do I have any evidence. It just looks like borrowed plumage. Synthetic flamingo for the prom.

  2. Stephen C. Carlson says:

    I think there’s another of this ‘if’ elsewhere in the article (albeit without the ‘then’): If Stone holds Augustine responsible for promoting the idea of sex as an evil force, he presents Saint Thomas Aquinas as “the man most responsible for the hardening of the Church’s attitude toward same-sex sex.”
    What an unusual usage!

  3. Yes, that’s the then-less form I said I was more or less used to; thanks for plucking it out as an example. I don’t like it, but it doesn’t make me froth the way the form with “then” does.

  4. Stu Clayton says:

    Now I see part of what’s happening here. “If…then..” is being used to create the impression that a train of thought is underway. In fact the cars are not connected, and are still standing separately in the marshalling yard. Fake choo-choos !

  5. I don’t know if this helps any, but I always think of these as a way of hedging. The author is saying, in effect, “Proposition p is kind of controversial, but IF you will join me for a moment in assuming p, THEN it is also true or relevant that q”. In other words, I read this as saying, “Did early Christians have especially dire views about sex? I guess people could have arguments about that, but it doesn’t seem unreasonable to say so, and if we do say so, then here’s another relevant fact: early Protestants rejected those views.”

  6. Nice try, but I’m pretty sure it doesn’t work. There is no implication that p is in any way controversial; it’s purely a manner of speech, and a piss-poor one.

  7. AJP Crown says:

    Also,
    Protestantism, which has been the dominant religious tradition in America from the beginning
    As well as being a peculiar premise, dominant? “Protestantism” is a dreadful word too. And it’s not a tradition, she just couldn’t think of anything better.

    Furthermore… I object to “scientific” when it’s not applied to the natural sciences. And yes, I know it comes from Hegel and social science and what you mean by it but scientific as opposed to what? Emotional, artistic, random, scatterbrained? Nothing good, that’s for sure.

  8. Owlmirror says:

    Hm. I see it as being a variant on a comparative conditional.

    X may be [something], but Y is [more/less something].

    So:
    If Maybe the early Christians had especially dire views on these matters, then but Protestantism, which has been the dominant religious tradition in America from the beginning, rejected those views.

    Or better: “The early Christians may have had”, etc

  9. As well as being a peculiar premise, dominant? “Protestantism” is a dreadful word too. And it’s not a tradition, she just couldn’t think of anything better.

    I find your objections very peculiar. How, exactly, would you rephrase the sentence? Protestantism not only strikes me as a very ordinary word, I can’t think of any other word that succinctly conveys the idea of “various Protestant sects.” And why not “dominant”? Protestants were a substantial religious majority when the United States was founded and remain so today.

  10. AJP Crown says:

    How, exactly, would you rephrase the sentence?

    I wouldn’t try. It’s a dumb premise.

    I can’t think of any other word that succinctly conveys the idea of “various Protestant sects.

    Then say “Protestant sects”. If you can’t see that at 5 syllables P’ism is an ungainly word, that’s not my prob. (it’s subjective.)

    Dominant doesn’t mean ‘in a majority’. It means among other things ‘having power over others’ and it carries all sorts of baggage not consistent with a separation of church and state (the great American contribution to Earth’s affairs, as far as I’m concerned). If she meant Wasps, her meaning would be clearer if she’d said so, but there are plenty of Protestants who aren’t Wasps (WASPs, if you prefer; I don’t).

  11. I wouldn’t try. It’s a dumb premise

    That’s a cop-out. You’d already made it clear that you disagreed with the premise. That wasn’t my question. Even ideas that we disagree with can be conveyed in language.

    As for the rest, five syllables does not strike me as a particularly large number, and Protestantism is no more objectionable to me than any other philosophical noun ending in -ism.

    And “separation of church and state” is a laudable abstract ideal, of course. However, lets not pretend that Protestants haven’t played a dominant role in American history. Indeed, for the first 150 years of the republic, there were virtually no non-Protestants in positions of national power. If that’s not “dominance,” I don’t know what is.

  12. David Eddyshaw says:

    The writer was clearly thinking in Classical Greek, as we properly educated folk tend to do: this is evidently simply an attempt to compensate for the fact that barbarian languages like English lack clear equivalents for μέν and δέ.

  13. David Eddyshaw says:

    Which gives me all the excuse I need to link

    http://faculty.georgetown.edu/jod/texts/housman.html

    (referring to the Chorus’s advice on being Good or Bad.)

  14. I don’t like using if that way, but I don’t see how adding then makes it any worse (or better). I wouldn’t have particularly noticed the sentence if you hadn’t pointed it out.

  15. AJP Crown says:

    However, lets not pretend that Protestants haven’t played a dominant role in American history. Indeed, for the first 150 years of the republic, there were virtually no non-Protestants in positions of national power. If that’s not “dominance,” I don’t know what is.

    Yeah, yeah, this is all beside the point. Forget American history & national power; what she’s talking about is:

    Protestantism, which has been the dominant religious tradition in America from the beginning

    That’s a cop-out.
    No amount of taunting will provoke me into rewriting rubbish; only money. I would do it for $1500, payable in advance. Besides, my comment was meant only as support for Language’s post. It’s no biggie. I just got carried away.

  16. David Eddyshaw says:

    Come to think of it, our wrongstoryshortener Tacitus uses ut … ita … as a sort of calque of μέν … δέ. But then he was a Roman, so he naturally spoke Greek, like all educated people.

  17. I’ve seen this construction a few times and found it alien to English, too. Who do we blame? The Germans? The French? Please let it be the French!

  18. Jen in Edinburgh says:

    I have no particular problem with the construction, but I don’t think I could use it in the original example – it has to be something in the shape of ‘if A was bad, (then) B was worse’.

    Stephen C. Carlson’s example is fine – Augustine was strict, Thomas Aquinas was stricter.

  19. J.W. Brewer says:

    Norvin’s gloss seems plausible to me, but I can’t tell if hat’s rejection of it is focused on this particular usage or the broader pattern that has inspired his peevishness. The general usage strikes me as weird and inelegant but I haven’t noticed it enough (maybe now I’ll start because I’ve been primed?) to have a good sense of how it’s used in practice. In this particular instance, however, the possibility that “if” is being used as a hedge indicating that p is not uncontroversial seems plausible insofar as the review goes out of its way to note that “The problem is that it is very difficult to speak of the ‘early Christians’ or the ‘ancient Hebrews’ as if these terms describe fixed, monolithic groups.”

  20. Concessive “if” is fine; merely contrastive “if” is a stretch.

    I think some authors use “if” in place of “while” or “whereas” in obeisance to Orwell’s Second Rule. If the two contrasting clauses are long and have internal commas, the author puts in a “then” to separate them more clearly. I don’t mind that; I guess I’ve already drunk the Kool-Aid on non-conditional “then”.

  21. In this particular instance, however, the possibility that “if” is being used as a hedge indicating that p is not uncontroversial seems plausible insofar as the review goes out of its way to note that “The problem is that it is very difficult to speak of the ‘early Christians’ or the ‘ancient Hebrews’ as if these terms describe fixed, monolithic groups.”

    It doesn’t go out of its way to note that, it just notes it. And the point is not to maintain that early Christians were not sex-negative; she goes on to say:

    The early Christians did not think up religiously inspired sexual continence all by themselves. […] What the early Christians did have was a set of highly effective salesmen for their beliefs, beginning with Saul of Tarsus, later Saint Paul the Apostle, who, a number of theologians believe, brought to Christianity many of the beliefs about women and sex that he had learned from the Judaism he left behind. But it was Saint Augustine, Stone argues, “who crystallized the early Christian understanding of sex and who, in so doing, ultimately helped shape traditional American views of sexuality more than a millennium later.”

    However, the details of this case are more or less irrelevant, since I’m responding to the broader pattern that has inspired my peevishness. I’ve seen lots and lots of examples (and have removed them when they occurred in material I was editing), and I assure you that any relation to Norvin’s hedging hypothesis (“IF you will join me for a moment in assuming p”) is purely coincidental. People are not using it that way, they are simply using it as a vague handwave to join two thoughts.

  22. I think some authors use “if” in place of “while” or “whereas” in obeisance to Orwell’s Second Rule. If the two contrasting clauses are long and have internal commas, the author puts in a “then” to separate them more clearly.

    That seems a reasonable hypothesis. Now, lay off that Kool-Aid!

  23. J.W. Brewer says:

    If the individual example isn’t the point, then the accuracy or inaccuracy of various hypotheses about what the pattern might mean are not going to be illuminated by it (I hope that was an ok use of the normal if/then construction?), and as I said I haven’t noticed enough other instances to have an independent sense of what’s going on.

  24. Stu Clayton says:

    Crown tacks down the principle – readers should be paid in advance to read such paltry prose. Absent klingende Münze, I refuse to read anything that’s harder to understand than what it purports to explain. In this vale of tears, who needs eye-wateringly confused concessives ?

    David Eddyshaw then sews it up with the link to Housman: “Mud’s sister, not himself, adorns my shoes”.

  25. If the individual example isn’t the point

    I didn’t say it wasn’t the point, but it’s just one example; it works for me, and I thought I’d given enough evidence to show why, but if it doesn’t work for you, that doesn’t change the facts. I hope you will start noticing it and see that it is not a logical construction in the Norvinian sense.

  26. David Marjanović says:

    In German this seems to require the weird verb-initial construction instead of any word for “if”: Hatten die frühen Christen noch […], so lehnte der Protestantismus […] diese Ansichten ab. Note I smuggled “still” in there to warn people that the next clause is going to be about a later time; omitting it seems downright ungrammatical.

  27. Stu Clayton says:

    Hatten die frühen Christen harsche Meinungen zu solchen Themen, sollte der Protestantismus diese Ansichten ablehnen.

    That’s one elevated-register technique to signal “later”.

  28. Hm, doesn’t bother me that much, because I think I parse it as shorthand for something like:

    “IF we all accept my description of X as a valid way to describe X, THEN you will surely also agree 1) with the way I shall now characterize Y, and 2) that it’s valid for me to now make the following interesting comparison between X and Y.”

    It’s a way of asking for the reader’s implied agreement with the first part of the whole premise and then leading them from that into agreeing with the second. For example (to use the original example), someone who didn’t agree with the first IF idea, that “the early Christians had dire views on these matters”, would THEN not agree that Protestants rejected them, but the author is hoping that everyone’s on board.

    Basically (missed it on first reading): what Norvin said.

  29. My feeling* is that this usage may have originated with the construction that Norvin describes (and in that form, it seems basically unobjectionable). However, even if that is where it started, the pattern has expanded quite a bit further, leading to examples like the one cited in the original post (which, I agree, feels like very bad style).

    * Personal feelings and impressions can, of course, be extremely unreliable guides for identifying the origins of particular expressions and idioms, so I may be totally off here. (As an example of this kind of unreliability, I remember once when John McWhorter was a guest on a public radio call-in program, he was talking about how new words and expressions arise in contemporary English. One man called into the show to suggest that McWhorter should pay attention to African-American culture as an important source of new words and meanings. That the caller was evidently unaware that John McWhorter is himself black was understandable, given the medium, but the weird part was that the caller rattled off, by way of examples, three or four neologisms that he was sure had originated in the African-American community—except that none of them actually had.)

  30. To take things one meta-level up, this passage reminds me of something that’s been on my mind for a while:

    I am a staunch descriptivist; that is, I hold to a scientific, fact-based description of language: whatever native speakers of a language say is ipso facto correct… I deplore prescriptivism, the idea that there is some Platonic ideal of language that is eternally “correct” and that all deviations from it must be chastised and if possible forbidden, and I mock the peevery it inspires.

    This seems like a false dilemma. Are those really the only two options here? There must be a middle way between those extremes, where we can look at past or proposed language changes from an engineering perspective and say, this one was (or would be) a design improvement and that one was/would be a design deterioration, without assuming any sort of metaphysical capital-p Perfection as a reference point.

    Ultimately language is a tool for communication, and tools can be more or less well-suited to their use depending on their design. It’s certainly true of computer languages, to take one example I know from personal experience; I don’t have to believe in any Platonic perfection to know that I’d rather write Python than COBOL. Why should natural languages be any different?

  31. Because they are. Natural languages have almost nothing in common with computer languages, and it is a fatal error to assume otherwise. You are, of course, free to look at language changes from an engineering perspective and decide on design improvements and deteriorations (from your point of view; you can’t expect others to agree), but it seems a pointless way to spend one’s time and mental energy.

  32. Let me put it another way, then: is there any other activity or behavior, any single one in all of human experience, where you would say “the way people do this thing changes over time, but the changes are never for the better or worse; in fact they cannot possibly be better or worse, and even thinking about it in those terms is an error”?

  33. Esli… to… is sometimes used in much the same way in Russian. Contrastively rather than conditionally.

  34. Jim, art is an obvious example.

  35. Stu Clayton says:

    More such activities or behaviors:

    1) expressing your opinions.
    2) reading science fiction novels
    3) espousing a religion
    4) favoring one baseball team over another
    5) preferring not to, with reference to any of the above

  36. @Stu, asserting that of expressing your opinions is to argue, about free speech, a position more extreme than I would accept. There are many sorts of opinions where it is reasonable to contend that their becoming less commonly asserted is for the better, or a recent trend to assert them is for the worse: sexism, racism, xenophobia, homophobia, transphobia etc.

    I agree with you about some activities which affect only the person doing them. I would even classify language here, when used to express private thoughts. Language is another matter though, when anyone uses it to communicate ideas to others. Here I side with Jim: it is a tool for communication. If the listener doesn’t understand, someone is wrong, and it might be the speaker.

    I myself don’t bother trying to decide on design improvements and deteriorations, but I might from time to time think about some language usage I’ve just noticed. If I then go on to consider its advantages and drawbacks, is that such a pointless way to spend my time and mental energy?

  37. dainichi says:

    Now that Jim has opened the can of worms…

    > whatever native speakers of a language say is ipso facto correct

    Sure, but it’s not necessary standard. What I fear is lost in the descriptivism-vs-prescriptivism feud is a healthy discussion about whether there should be a standard, what should be in it, who should design it, in what situations its use should be encouraged etc…

    Also, I fear that the lack of a regulated standard means that the de facto standard is modeled after the language of the powerful. But I’m not sure to what extent this is a problem for languages without regulatory bodies, or if the situation is any better for languages with them.

  38. Owlmirror says:

    You are, of course, free to look at language changes from an engineering perspective and decide on design improvements and deteriorations (from your point of view; you can’t expect others to agree), but it seems a pointless way to spend one’s time and mental energy.

    This looks like it could be a dig at those who work on engineered languages like Lojban.

    If it’s not, what do you think would be the difference between what you write and what those projects are doing or have done? Did you only have in mind trying to make engineered changes to non-engineered languages (and thus, languages that are too widely-diffused among probably recalcitrant populations to be easily engineered)?

  39. minus273 says:

    I feel this to be a high-register Gallicism, imitating the contrastive si.

  40. David Marjanović says:

    Also, I fear that the lack of a regulated standard means that the de facto standard is modeled after the language of the powerful. But I’m not sure to what extent this is a problem for languages without regulatory bodies, or if the situation is any better for languages with them.

    In German, it’s rather the powerful who have taken up the standard than the other way around. The standard started as the written language of bureaucrats, followed by Luther’s desire to make his Bible translation as widely understood as possible. There is no regulatory body, except on orthography since 1901.

  41. whatever native speakers of a language say is ipso facto correct

    I can’t agree with this, because it ignores the existence of dialects. A usage within a particular dialect may be correct, but could be incorrect when brought into a different dialect. For example, my father regularly referred to his doctor as “the quack”, something that would be totally normal in Australia, but when he moved to other countries it seriously shocked people.

    Further I have noted some examples of words that people pick up by ear and don’t quite get right. One is “specific” and “Pacific”. If someone says “I can give you a Pacific example”, I would call that incorrect. Also “for all intensive purposes”, and I think there are many other examples that have been discussed here.

    There are also spelling-only examples like “free reign”, but perhaps that is a different category. (“Things that get printed in the newspaper because Microsoft Grammar Check doesn’t catch them.”?)

    In regard to the if-then question, I thought I would take a look in Fowler (I have the 3rd edition, 1996), and it does not address the matter. Hmm. It’s something that I mentally characterize as an “NPR usage”, things that I hear from people who get interviewed on NPR. Like the “biscuit conditional’, was that what we called it? (“If you’re hungry, there are biscuits in the cupboard.”) Because of Terry Gross, “If you’re just joining us, we’re talking to Joseph-Michel Montgolfier, the distinguished balloonist.”)

  42. Stu Clayton says:

    I feel this to be a high-register Gallicism, imitating the contrastive si.

    What the triangular heck is a “contrastive si” ? Tip: example suffices.

    I am grateful to maidhc for the notion of “biscuit conditional”.

  43. AJP Crown says:

    I’m grateful to maidhc for the whole comment. But should we criticise the biscuit conditional after we’ve noticed and remarked on it? We understand Terry Gross; ‘if you’re just joining us’ clearly isn’t part of a philosophical question. It’s a phrase, and I feel better off with it than without it, unlike the reviewer’s if-then construction which I can easily do without.

  44. I generally read these sort of “if…then… “ constructions as containing an implicit “if we accept/assume”. E.g. If {we accept that} the early Christians had especially dire views on these matters, then {we must concede that} Protestantism, which has been the dominant religious tradition in America from the beginning, rejected those views.” But maybe this is my brain trying to impose order on chaos.

    Using “if” without “then” to simply mean “while” is not English for me.

  45. Stu Clayton says:

    Using “if” without “then” to simply mean “while”

    While you have ham, you can have ham and eggs, while you have eggs.

    If you have have ham, you can have ham and eggs, if you have eggs.

  46. Stu Clayton says:

    But should we criticise the biscuit conditional after we’ve noticed and remarked on it?

    Of course not. I see a whole new grammatical category to match the Latin ablative. We can just call it the English conditional. So far we have

    – the conditional of biscuit
    – the conditional of contrast
    – the conditional of tuning in
    – the conditional of clickbait

  47. Owlmirror says:

    Maybe it’s really a clickbait conditional.

    If [you thought that] the early Christians had especially dire views on these matters, then [check out this one weird theological trick where] Protestantism, which has been the dominant religious tradition in America from the beginning, rejected those views.

  48. Stu Clayton says:

    # If you think this is weird, you ain’t seen nothin’ yet.#

  49. David Marjanović says:

    Maybe it’s really a clickbait conditional.

    I think you’re on to something.

  50. John Cowan says:

    You are, of course, free to look at language changes from an engineering perspective

    That, indeed, is the perspective not only of computer language designers but of (added: some) conlangers, and to a lesser degree of language standardizers. But what standardizers do can become the living language of actual people. This is most obvious in Hebrew, but also in Greek, where the contemporary standard language that most Greeks speak is a messy and unsystematic compromise between the artificial Puristic (itself a compromise with Ancient Greek) and the naturally evolved Demotic. The post-junta constitution even specifies the national language to be Demotic “without extremism”, meaning not hyper-Demotic.

    Like the man said: “[People] make their own [language], but they do not make it just as they please; they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past. The tradition of all dead generations weighs like a nightmare on the brains of the living.”

    If you have have ham, you can have ham and eggs, if you have eggs.

    Nice example there of an actual language error (in the sense “a construction that the speaker/writer would disown”).

    If the dog would catch a rabbit, we’d have rabbit stew for dinner — if we had a dog.

  51. AJP Crown says:

    In England (I’m not sure about elsewhere) everyone now writes “whether or not”; whether the “or not” is needed or not (as it is here).

  52. Language is another matter though, when anyone uses it to communicate ideas to others. Here I side with Jim: it is a tool for communication. If the listener doesn’t understand, someone is wrong, and it might be the speaker.

    Of course something is wrong; things go wrong all the time. But the problem is self-correcting, both in any given case (“Huh? I didn’t get that” followed by rephrasing) and in general (if language change leads to genuine ambiguity that causes genuine problems, speakers will route around it and develop new ways of saying what needs to be said). The basic confusion here is the prevalent but false idea that language can be “fixed” in both senses: changed for the better and for good. The most important fact to know about language is that it changes, constantly and unstoppably. Once you get that into your head, you realize the futility of all ideas for “improving” it. Language is fine and will go on being fine.

    This looks like it could be a dig at those who work on engineered languages like Lojban. If it’s not, what do you think would be the difference between what you write and what those projects are doing or have done? Did you only have in mind trying to make engineered changes to non-engineered languages (and thus, languages that are too widely-diffused among probably recalcitrant populations to be easily engineered)?

    I confess that to me constructing engineered languages seems a pointless way to spend one’s time and mental energy, but all that means is that it’s not my thing — to someone else, my reading obscure Russian authors and following the Mets would seem a pointless way to spend my time and mental energy. I applaud people doing whatever makes them happy as long as it doesn’t frighten the horses. But yes, I had in mind trying to make engineered changes to non-engineered languages.

    non-engineered languages (and thus, languages that are too widely-diffused among probably recalcitrant populations to be easily engineered)

    I’m not sure how much irony is intended here, but I will take it seriously as written to make an important point: It has nothing to do with wideness of diffusion. Natural languages are, like art and religion, an inherent element of humanity that are not amenable to engineering, and the very idea of engineering them gets my hackles up because it reminds me of Stalin’s “engineers of human souls.” Leave humanity alone!

  53. Stu Clayton says:

    weighs like a nightmare on the brains of the living

    To get the same effect today, Charlie would need to write “the tradition of all dead generations makes me uncomfortable”.

  54. Also, invented languages are historically tied to the idea that natural languages are inferior (“illogical” and what have you) and humanity would be improved by using the inventor’s brilliant idea for a replacement. This I reject with loathing as both megalomaniacal and anti-human; the whole idea that people should be as “logical” as computers is loathsome and anti-human. To the extent that modern constructors of languages are just enjoying themselves in a similar way to those who construct crossword or chess puzzles, I have no problem with them, but to the extent they are carrying on in the tradition of attempting to “improve” humanity I deprecate them. Ah has spoken!

  55. J.W. Brewer says:

    I should like to introduce hat’s “It’s bad enough” in the original post to this vintage LL post and the OED definition of one not-particularly-conditional sense of “if” that it references, with examples back to 1965. http://itre.cis.upenn.edu/~myl/languagelog/archives/000408.html

    Significantly, however, the examples given of that usage of “if” tend not to be followed by a “then,” which might mean that using the “then” with an other-than-truly-conditional “if” really is a more recent innovation calculated to aggravate those who weren’t acclimatized to it in their formative years. Indeed, Prof. Liberman says regarding the specific example that had prompted the post: “In fact Wilgoren & Justice don’t use ‘then’ — it would have been weird if they had.”

  56. Interesting, thanks — I’ll add it to the post.

  57. AJP Crown says:

    the prevalent but false idea that language can be “fixed” in both senses: changed for the better and for good. The most important fact to know about language is that it changes

    I don’t think many people are really trying to fix English. It’s more about identifying the English that you yourself would like to use; getting that straight in your own mind. Except for newspaper commenters, nobody over the age of 30 is expecting to change the world.*

    *(allows for grad school plus 2 yrs in a job)

  58. I don’t think many people are really trying to fix English.

    Oh, I think you’re vastly underestimating the number; they’re a small subsection of the number of people trying to fix the world in general, but still quite a few. People have a deep desire to make everybody else conform to their own ideas of what is right and proper.

  59. Kate Bunting says:

    ‘Contrary-to-fact past may have’ has long been a pet peeve of mine. I’ve often seen statements such as “If X had happened he may not have died” when we know that the unfortunate victim did die.

  60. Exactly!

  61. John Cowan says:

    Hat: Since standardization does affect language change (see Nick the Greek’s post that I linked), how do you differentiate between standardization and engineering?

  62. Rodger C says:

    What’s happening there is that “may” no longer has a past tense “might.” It annoys the hell out of me too.

  63. how do you differentiate between standardization and engineering?

    I don’t like either one.

  64. AJP Crown says:

    you’re vastly underestimating the number

    Perhaps in the world but not here at Language Hat, surely. And yet we too have our pet, Peeves. So what are we trying to do? Should we kick the habit of analysing other people’s mistakes and our own dialectal idiosyncrasies?

  65. Oh, no, I meant in the world, of course. Here at Language Hat we are all sensible folk. We should definitely carry on analyzing other people’s mistakes and our own dialectal idiosyncrasies.

  66. David Eddyshaw says:

    At some stage I internalised the may/might thing so thoroughly that sentences like “If he hadn’t done that, he may not have died” quite genuinely cause doubletakes for me, irregardless of any peevery. I have to mentally translate them into Properspeak.

    But then I also say “I shall” where normal people say “I will.” I should probably just give up and join the Tory Party.

  67. ktschwarz says:

    whatever native speakers of a language say is ipso facto correct

    More precisely, whatever a speech community accepts as normal is ipso facto correct, and you can empirically investigate what’s accepted. There are people with brain damage who produce word salad and don’t notice anything wrong with it; that doesn’t make their language correct. And who decides which people have brain damage? A speech community.

    The phrase “community accepts” suggests questions to ask when talking about language correctness: What exactly is a community? How unanimous does it have to be? And what if people go around using and hearing something routinely without ever noticing it, until somebody points to it and says “That’s against the rule!” — and they all nod and say oh yes, mustn’t violate the rule, and go on just as they were?

  68. David Eddyshaw says:

    And what if people go around using and hearing something routinely without ever noticing it, until somebody points to it and says “That’s against the rule!” — and they all nod and say oh yes, mustn’t violate the rule, and go on just as they were?

    This actually happens a lot (especially with syntax), and by no means only in languages with a tradition of writing and/or pseudogrammatical peevery. People’s insights into how they use language are often inconsistent with their own unreflective usage: like the man said, you can’t trust your Sprachgefühl, even in your mother tongue.

    In Kusaal, for example, people often use “animate” gender pronouns for inanimates; if you draw their attention to the phenomenon, they substitute inanimate pronouns, right up until they are concentrating on something else. (I suppose that in fact this just shows that “animate gender ” has actually become “unmarked gender” in informal speech.)
    You don’t see this in texts at all, but you hear it in speech all the time if you eavesdrop.

    Both sets of facts are interesting and worth describing, I’d say; but all it shows is that people’s insights into the structure of their own languages are often not very solidly based in reality. Which is hardly surprising a priori, and why there is an actual ecological niche for linguists.

    Most of what prescriptivists want to say which has any connection with linguistic reality (and there’s a fair bit of it) properly belongs to the perfectly scientific realm of sociolinguistics.

  69. J.W. Brewer says:

    Just re the dialect-variation-versus-“standard” point, things get even more complicated in a pluricentric language like English, where spellings, pronunciations, syntactic features, word usages, etc. that are so deviant from the standard/prestige variety of AmEng as to be Just Plain Wrong from a scientific sociolinguistics perspective are sometimes in fact part of the standard/prestige variety of English in some other Anglophone part(s) of the globe. I was reminded of this just now via free-association from the “crackpot filter” discussion on another thread because I’m pretty sure I first became aware back in my teens that British people tend to pronounce “garage” in a Totally Weird and Wrong Way via the Clash song with the opening line “Back in the garage with my bullshit detector.”

  70. Etienne says:

    Minus273’s comment today (4:46) echoes my own impression, that this use of “if” in English is a Gallicism.

    Alex K. (1:10): I suspect the Russian ESLI…TO structure is also a Gallicism (Could some Hatters familiar with the history of the Russian language confirm or deny this?)

    Stu Clayton: In answer to your question (Today, 6:18), a “contrastive si” in French is a “si” which can best be translated into English as “whereas”: thus,

    “Si moi j’aime les chiens, ma soeur, elle, aime les chats”, which we may render into English as “Whereas I like dogs, my sister likes cats”.

    Importantly, this is perfectly good written French and quite common in more colloquial registers: to my ear it seems more natural if both propositions have a disjunctive subject pronoun (as in the example above: “Si j’aime les chiens, ma soeur aime les chats” strikes me as grammatically correct standard French, but as slightly artificial/unnatural as an ordinary spoken French sentence).

  71. David Eddyshaw says:

    Bullshit detectors are very useful. We could club together to get one on Amazon for … anyone who might need one.

    They are indeed best kept in garridges. If you keep them in the house they can impair your social life significantly.

  72. J.W. Brewer says:

    Of course, “whereas” tends to be used in English near-exclusively in less colloquial registers, making the proffered translation of the French example seem oddly formal!

  73. Minus273’s comment today (4:46) echoes my own impression, that this use of “if” in English is a Gallicism.

    Your example certainly suggests the likelihood of that idea.

  74. “Si je préfère les chats aux chiens, c’est parce qu’il n’y a pas de chat policier.” (By Cocteau or Prévert, according to the internet. I saw it in a book by Siné.)

  75. …chats policiers, in some versions.

  76. David Eddyshaw says:

    Police cats would be terrifying, at least if they could be bothered to be.
    I imagine something like Porfiry in Crime and Punishment.

  77. Etienne says:

    J.W. Brewer: Hmm, what about “while”: “While I like dogs, my sister likes cats”: that seems less formal in English, and thus might be said to be more faithful to the French original in terms of register.

  78. J.W. Brewer says:

    “While” would be colloquial enough for my own sense of colloquial register, although I’m not sure I can speak for the whole range of variation in AmEng. But I think the more idiomatic way to express the point in AmEng would be just “I like dogs, but my sister likes cats.” For most purposes I wouldn’t think of that as an excessively loose translation of the French, because preserving word order should not be very high on the list of desiderata for even fairly “literal” translations. Unless perhaps the French expression is itself more “marked” compared to some standard/default idiomaticway of making the same point in French, which I don’t have any insight into due to my near-complete lack of fluency in French. And of course the woodenly-literal word-order-preserving translation may be useful in a textbook-type setting where you’re trying to get Anglophone students of French to understand and internalize the French construction — in that context unidiomatic English is to be expected and is indeed arguably a feature rather than a bug of the specific genre.

    EDITED TO ADD: I should perhaps have thought of this earlier, but what I gave as the “normal” idiomatic way to make the point in English is sort of in a vacuum, i.e. considered as a freestanding sentence devoid of context. One of the OED examples in the old LL post that hat added to his post once I mentioned it is “If my father’s people were mill-workers.., my mother’s people were agricultural workers.” Even if you came up with a different word to swap in for the “if,” that’s a bit oddly-phrased in a vacuum, but my inference is that something about the previous discourse created a context in which that approach worked better than “my father’s people were X but my mother’s people were Y” would have. “While I like dogs . . .” would be a more plausible approach if dog-liking were part of the topic already on the table.

  79. ktschwarz says:

    David Eddyshaw: by no means only in languages with a tradition of writing and/or pseudogrammatical peevery … hardly surprising a priori

    Hardly surprising to you as a seasoned professional, but to me those Kusaal facts are amazing and awesome. Thank you!

  80. Yes, I love DE’s Kusaal facts!

  81. For more on this subject, Kaaden’s A Semantic Map Approach to Adverbial Clauses is just the thing. I haven’t had time to more than skim it.

    From the abstract: “The manuscript applies the method of building and analyzing a semantic map to the grammatical category of adverbial clauses, more specifically to adverbial clauses of concession. The category of concession and its markers are described in detail and the different semantic-pragmatic functions of concession that are inherent to the category are developed based on the descriptions. These functions are then arranged in a network, the conceptual space of adverbial concession. In a last step, the concessive conjunctions of four sample languages will be distributed on the map. The sample of languages is restricted to English, German, French and Latin.”

  82. David Eddyshaw says:

    I’m always happy to rabbit on about Kusaal given the least encouragement. And, indeed, without any encouragement at all, or even in the face of polite discouragement.

    I am, however, no seasoned professional, but only a wizened amateur.

  83. David Marjanović says:

    Yes, I love DE’s Kusaal facts!

    Who doesn’t 🙂

  84. ktschwarz says:

    Scooped by J.W. Brewer! I swear I was about to post that link to Language Log last night! I think that Pullum’s example,

    If Eskimos have dozens of words for snow, Germans have as many for bureaucracy.

    is especially valuable because we don’t even need to interpret it, we can compare the many variant expressions of the Ur-Snowclone, such as:

    Just as Eskimos distinguish between a dozen different sorts of snow, so too do the Swiss appreciate a wide range of Gruyeres.

    You know how Eskimos have like 100 words or so for snow? Why don’t people in Washington have 100 different words to rain?

    Presumably all these forms mean the same thing, and people are independently re-inventing the different syntaxes.

    MWCDEU judges “if in the sense of ‘though'” to be standard, but their examples are only simple pairs of adjectives, like “a routine, if exhaustive, search”; they don’t consider the question of this kind of “if” with a clause. They do point to some who don’t like it, such as Garner in Oxford Dictionary of American Usage and Style: Garner says it’s “oddly precious” and “carries a tone of affectedness”, though his example is just a pair of adjectives as well.

    I don’t feel like it’s precious — it is highfalutin, I’ll grant that — and I haven’t noticed any obtrusive “then”s, so I searched the most serious nonfiction that I have in electronic form: Alan Turing: The Enigma by Andrew Hodges. Paydirt! Here are a few:

    For if Natural Wonders had introduced him to the post-Darwinian world, Einstein took him into the twentieth-century revolution of physics.

    But if he was twenty, and preparing to confront the work of European mathematicians, he was still a boy away from home, away from Sherborne.

    If at Sherborne his sexuality was described in terms of ‘filth’ and ‘scandal’, he now also had to come to terms with that other kind of labelling that the world found so important: that of the pansy, an affront and traitor to masculine supremacy.

    I’m not sure whether I would defend all these or not. The first “if” could be replaced with “just as”, but perhaps it’s meant to suggest that “post-Darwinian” isn’t the most obvious adjective. The second “if” could be replaced with “though”; perhaps it suggests something like “If you’re thinking of him as an adult now, remember that he was just barely adult.” I’m not sure what the third one means, maybe “whereas”.

    Aha, here’s one with THEN!

    For if Cambridge embodied class, then Princeton spoke wealth.

    This could have used “whereas”, but I think Hodges must have been going for an emphatic parallel structure with exactly the same number of words on each side, hence the “then” to match the “if”.

    I’m not ready to ban all of these yet, even though they all have alternatives. Words are allowed to be polysemous. More study needed, and I should note, there were other books where I didn’t find any.

  85. Well found — excellent examples all!

  86. @ktschwarz: I did not remember this kind of construction as part of Hodges’ particular style in Alan Turing: The Enigma, so those instances must have seemed unremarkable to me as I was reading.

    It is probably not significant that Hodges is, by profession, a mathematician. That fact does, however, make me wonder whether any of his examples could actually be biconditional—e.g. But he was twenty, and preparing to confront the work of European mathematicians, if and only if he was still a boy away from home, away from Sherborne.

  87. Ellen K. says:

    There’s also “Although I like dogs, my sister likes cats”.

  88. Bathrobe says:

    Sorry, didn’t have time to read the whole thread.

    The construction reminded me of this:

    “If Trump is the Hitler of the new orthodoxy, Bannon is its Goering”. (Totally made-up example)

    I think Owlmirror hinted at something like this with his “clickbait conditional”.

    The sentence that Hat is complaining about is a pretty piss-weak example, but it seems to me to fit into this mould (at least to some extent).

  89. Bathrobe says:

    Also, as pointed out:

    Just as Eskimos distinguish between a dozen different sorts of snow, so too do the Swiss appreciate a wide range of Gruyeres.

  90. Stu Clayton says:

    @Etienne: many thanks for the example of constrastive si in French ! The clarifying role of the “disjunctive subject pronoun” makes perfect sense. I have internally modified it for future use of “lui/moi/…” in my speech. By “modify” I mean erasing the technical term “disjunctive subject pronoun” from my memory. I find that explicit terminology too easily becomes an obstacle to proper practice. It’s like the trainer wheel on a child’s bicycle.

    In English, the contrast can be enhanced (if desired) by the use of “myself/…self” or, in speech, by stressing the pronouns: “[whereas] *I* like dogs, my SISter likes cats”.

  91. If I peeved about every word I hear an American pronounce, as J.W. Brewer put it, “in a Totally Weird and Wrong Way”, I’d be peeving more about that than about everything else combined. Even with the bar set higher, things deserving being peeved at still arise.

  92. David Marjanović says:

    I’m not sure what the third one means, maybe “whereas”.

    That’s a clickbait conditional. “If you thought being described in terms of ‘filth’ and ‘scandal’ were bad enough, just wait till you get to the mindset that underlies all this.”

    But he was twenty, and preparing to confront the work of European mathematicians, if and only if he was still a boy away from home, away from Sherborne.

    What sense would that make? How does If he was twenty become “But he was twenty”?

  93. AJP Crown says:

    Pullum’s example,

    If Eskimos have dozens of words for snow, Germans have as many for… buns named after their cities (hamburgers, berliners etc.)

    The last time I read about the Eskimos in Language Log, many years ago, this was supposed to be a myth. Eskimos have the same number (three or four) of words for snow as others in cold climates. So I think Geoff Pullum really did mean IF in this case. “Just as Eskimos…” wouldn’t be as good.

  94. That double if thing: Hvis du vil købe tobak, skal du vise billed-ID, hvis personalet beder om det. Official Danish verbiage to the effect that “if you want to buy tobacco, you have to show picture ID, if the staff requests it.”

    It’s not part of my syntax for Danish, let’s just leave it at that.

  95. David Marjanović says:

    Pullum is the author of a paper titled The Great Eskimo Vocabulary Hoax. He was being sarcastic about his pet peeve.

    It’s not part of my syntax for Danish, let’s just leave it at that.

    Unremarkable in spoken German, but I’d definitely avoid it in writing.

  96. “Si je préfère les chats aux chiens, c’est parce qu’il n’y a pas de chat policier.” (By Cocteau or Prévert, according to the internet. I saw it in a book by Siné.)

    I on the other hand saw it in a Fabulous Furry Freak Brothers comic. “Dogs are fascists. Ever seen a police cat?”

  97. Unremarkable — I’m pretty sure that even when speaking I’d make a repair by putting the second condition in a separate coordinated (verb and subject-less) sentence: …, men kun hvis … = ‘… but only if …’. The type of repair would depend on the logical relation of the two conditions, but in this case ‘but only’ is clearly what is meant.

    More colloquially, you (I) could start a new sentence intonation and start Hvis altså … = “That is, if …”.

    Also Fritz couldn’t be a dog, now could he?

  98. John Cowan says:

    Unremarkable in spoken German, but I’d definitely avoid it in writing.

    Toadally cromulent in written English, at least mine.

  99. ktschwarz says:

    I was too concise: Pullum didn’t write the sentence about German bureaucracy, he quoted it from the Economist.

    Surely this “if” goes back further than 1965? Mark Liberman didn’t quote all of the OED’s citations. I’ll try looking in some older books.

Speak Your Mind

*