Ask Language Log: "differ to"?

« previous post | next post »

Trevor Butterworth, "Top Science Journal Rebukes Harvard's Top Nutritionist", Forbes 5/27/2013:

In an extraordinary editorial and feature article, Nature, one of the world’s pre-eminent scientific journals, has effectively admonished the chair of the Harvard School of Public Health’s nutrition department, Walter Willett, for promoting over-simplification of scientific results in the name of public health and engaging in unseemly behavior towards those who venture conclusions that differ to his.

Barry R. asks:

"… that differ to his?"  Is this a common usage? Or is it as wrong as it sounds to me?

Stan Carey has a long and insightful discussion of a related question, preposition-choice with different ("Different from, different than, different to", Sentence First, 4/20/2011):

If you see nothing immediately wrong with the phrases different from, different to, and different than, you might be surprised by all the ink spilt, keys poked and eyebrows furrowed over their respective permissibilities. Not only is different than often mistakenly called a mistake, it has been described as flagrant, eyebrow-raising, revolting, abominable, and ridiculous. More on that later. First, an introduction to the use and distribution of the expressions.

Different from is by far the most widely used and accepted form, different to is common in British English, and different than is spoken regularly in different varieties of English, including US English and BrE. All have their uses. The predominance of different from, particularly in written English, is shown by these figures from the Collins Cobuild Bank of English, which I found at alt.usage.english:

Here are the raw string-frequencies in the BNC and COCA corpus:

different to 474 405
different from 3,246 13,439
different than 50 3,846
% to 12.6% 2.3%

The string frequencies are only an approximation — many of the 405 instances of "different to" in COCA, for example, are things like

And I'm wondering whether it feels different to you.
Well, we all bring something different to the table.
So, he actually said something different to his wife than he did in his press conference.
She had something different to live for.
It turns us on when you wear something different to bed.

Still, this proxy measure confirms that "different to" is more common in British English. What about prepositions in the complement of the verb differ?

differ to 8 11
differ from 457 1819
differ than 0 0
% to 1.7% 0.6%

Again, from is the most common; and than is genuinely shunned, even without any fussing from peevers.

And in this case, only one of the 11 COCA examples is genuine — and that one is in the transcript of an interview with a British officer:

Quite clearly, the requirements for peacekeeping differ to those of becoming a rapid reaction force.

The others are all things like

Financial planners differ to some extent about how such cash should be used.
How does the approach of the journalist and the academic differ to this sort of classified information?
The systemic azoles all differ to some degree in drug-drug interactions, adverse reactions, and clinical application.

Similarly, only one the 8 "differ to" examples in the BNC is genuine:

When acid secretion was expressed corrected for lean body mass and fat free body weight these findings did not differ to those using acid output expressed as mmol/h.

All the rest involve things like "… differ to the extent that…", "… differ to a greater or less extent from each other…", "… differ to some extent over the precise reasons…"  There are 5 more examples of differed|differs|differing to, of which two are genuine to-complements:

If their opinion differs to mine, I believe that they know best.
Peter Harris on how the British approach of designing and buying a home differs to our European neighbours.

But if we weigh these 3 instances of [differ] to against the 1,099 (almost all genuine) examples of [differ] from, we see that BNC writers and speakers are choosing [differ] to only a few tenths of a percent of the time.

Nevertheless, looking at the current Google News index turns up dozens of relevant examples, all from British or ex-British Empire sources. For example:

Woodford’s separate £10.3bn Income portfolio, for instance, is not at risk of sector ejection because its own year-end date of 31 March differs to that of the High Income fund. [Investment Week, a .uk site]
Additional details on Tekken Revolution, including the content featured in the base game, information on micro-transactions, and how the game differs to previous Tekken titles have yet to be revealed. [Videogamer,  by a writer based in London]
But it is more precise to talk about a plurality of racisms than a single phenomenon; racism in Australia certainly differs to racism in say France, which differs again to racism in Norway, and so on. [The Guardian]
How does it compare and differ to your past albums? [Lancashire Telegraph]
Animals Australia executive director Glenys Oogjes said although her group's views differed to those of to rural lobby groups on live export, it was “profoundly disappointing to see this attack on Coles for supporting an unrelated campaign that encourages consumers to support higher welfare methods of production”. [The Land (Australia)]
How will this tour differ to all your other tours? []
None of the leaders, however, stood up to give an opinion that differed to Mbela's nor move to speak and tone down the inciting message. [The Standard (Kenya)]
Claiming that “no one ever left their desktop in the back of a taxi cab”, Morris’ opinions differ to that of Apple CEO Tim Cook, who believes we are now living in a “post-PC era.” [Technday (New Zealand)]

Without doing a quantitative comparison, it seems to me that there are more of these Out There than we would expect if the true rate of [differ] to usage among British (in the broad sense) writers were only a few tenths of a percent. Perhaps things have changed since the early 1990s, when the BNC was collected, or perhaps journalists are special in this respect. It's relevant that the author of the Forbes article, Trevor Butterworth, is "an Irish journalist and writer who has lived in the United States since 1993 and presently resides in Brooklyn".

In any case, the web-search evidence seems to confirm that [differ] to is essentially never used by Yanks. The only Google News example that I found from an American source is in a transcript quoting my old friend George Doddington, where it's a mis-transcription of "defer to".  And if British (ex-) Empire usage is largely limited to journalists, that would help explain Barry R.'s reaction.


  1. Different from, different than, different to | Sentence first said,

    June 10, 2013 @ 8:01 am

    […] at Language Log received a query about the phrase differ to, and presents some data and analysis of prepositions used with the verb differ, quoting from this post by way of […]

  2. And what about said,

    June 10, 2013 @ 9:29 am

    —I beg to differ with you. ?to, from? than?

  3. Joe said,

    June 10, 2013 @ 9:40 am

    I think I linked to this before, but the BYU Corpus of Global Web-Based English is a good place for comparisons with COCA and the BNC.

  4. Lazar said,

    June 10, 2013 @ 9:53 am

    I'm surprised to learn that "different from" is so common in both countries. I've had the impression that it's characteristic of careful speech, and that in informal speech it's been largely replaced with "different than" here in the US and "different to" in Britain.

  5. Victor Mair said,

    June 10, 2013 @ 10:02 am

    My recollection is that I've been saying and writing "different from" for the last seven decades, while "different than" sounds awkward to my ears, and "different to" strikes me as positively bizarre. (I'm from northeastern Ohio, but have spent my adult life on the East Coast and at various far-flung locations around the world).

  6. John Lawler said,

    June 10, 2013 @ 10:05 am

    Perhaps it's a confusion with defer to?

  7. J.W. Brewer said,

    June 10, 2013 @ 10:13 am

    It's of course hard to tell these days how much this sort of web content is edited in the traditional sense (with Forbes copy-editing presumptively done by AmEng speakers who would not share Butterworth's own possible non-AmEng intuitions about the acceptability of this usage). The odd thing is that it wasn't an available non-AmEng option I was aware of. If Butterworth had e.g. used "whilst" for "while" I would have reacted "Oh, he must be from someplace foreign (you know, the sort of place where "Trevor" is a more common name than it is in the U.S.) and it's interesting that Forbes didn't impose an AmEng house style on his natural voice." But here it just Sounds Wrong.

  8. Robert Coren said,

    June 10, 2013 @ 11:50 am

    I'm accustomed to seeing "different to" in British-English writing. I can't recall ever seeing "differ to".

    I was brought up in the belief that "different than" was wrong, wrong, wrong (I've recovered). I remember my father challenging my brother's use of it by asking "Do you differ than us?" (He then rather whimsically segued into "Are you differ than us?" It didn't occur to me until years later that, if he was really being a stickler, he should have said "we" rather than "us".)

  9. Alan Gunn said,

    June 10, 2013 @ 1:31 pm

    In "Line by Line," Claire Kehrwald Cook says (p. 175 in my edition) that "in educated American usage, one thing is different from another, not different than another." She adds, though, that "than" is preferred when a clause, rather than just a single noun, follows "different," as in "We use a slightly different method than they do." Sounds normal to me.

  10. dw said,

    June 10, 2013 @ 2:31 pm

    The British phrasing "different to" is, I have always thought, by analogy with "similar to", or "identical to".

    Is the US "different than" by analogy with "greater (or less) than"?.

    In that case, the US usage can be seen as analogous to the use of "" to represent inequality (as in the programming languages BASIC or Pascal), while the British usage is analogous to the use of "!=" for inequality (as in C, Java and many other other programming languages).

    "Differ" and "defer" (meaning "delay") are etymological doublets, both from Old French differer.

  11. dw said,

    June 10, 2013 @ 2:33 pm

    Last comment with correct HTML encoding:

    The US usage can be seen as analogous to the use of "<>" to represent inequality (as in the programming languages BASIC or Pascal), while the British usage is analogous to the use of "!=" for inequality (as in C, Java and many other other programming languages).

  12. Pete R. said,

    June 10, 2013 @ 3:01 pm

    The first time I saw 'to' instead of 'from' in this context, I thought it was an ESL-ism. In my own experience (East coast US), I've almost always heard 'from' and occasionally 'than', and always read 'from'.

  13. Eric P Smith said,

    June 10, 2013 @ 3:08 pm

    Like Victor Mair and Robert Coren, I was brought up to say “different from”. I can see however the usefulness of We use a different method than they do. Since *We use a different method from they do is impossible, sticklers for “different from” have to say something more roundabout, and arguably more clumsy, such as “We use a method that differs from the one that they use”.

    Over the years I have got used to “different to” and “different than”. Nevertheless, “differ to” sounds odd to me.

  14. dw said,

    June 10, 2013 @ 4:13 pm

    @Eric P Smith

    "We use a method different from theirs".

  15. Rubrick said,

    June 10, 2013 @ 4:14 pm

    What I find interesting is just how wrong whichever versions one is least familiar with feel — when I hear "iOS is different to Android in many ways" my gut reaction is that it's as absurd as "I swam my new airplane to Tulsa".

    And yet on reflection, it's clear that "than", "from", and "to" are virtually content-free in this context. "On", or "with", for example, would serve just as well so far as meaning is concerned. The choice is purely a matter of convention and habit.

    I think "on reflection" is what's missing from most peevers' analysis.

  16. Faldone said,

    June 10, 2013 @ 4:35 pm

    Back in the day, when they still allowed riffraff on the Vocabula Review message board, I was involved in an ongoing argument with a member of the different-from-only club. He maintained that than could only be used with comparatives, e.g., "X is newer than Y. Unfortunately they limited participation on the board to subscribers to VR before I managed to ask him if he would then say that he unconditionally recommended use of different from rather than different than. But then, perhaps rather is the comparative of rath: rath, rather, rathest. I never could rationalize paying $40/year to read the stuff on the Journal of Prescriptivist Poppycock.

  17. J.W. Brewer said,

    June 10, 2013 @ 4:36 pm

    I would say slightly contrary to Rubrick that it is the fact that preposition usage often tends to be so arbitrary/contextual/content-free that makes it useful to have conventions (not prescriptivist zombie rules, but actual conventions identifiable by descriptivist methods that are in practice tacitly but consistently followed by native speakers). That said, there certainly are contexts in AmEng where more than one preposition is in reasonably common use, sometimes with a semantic nuance; sometimes without. MWDEU has a bunch of entries on situations like that, and I'm not in favor of stomping those existing variations out on One Right Way grounds. But I am less interested in refereeing between different from and different than than in the evidence confirming my own intuition that "different to" is just not part of AmEng. With all due respect to the journalists of the Commonwealth, it doesn't seem like we're missing much.

  18. Jerry Friedman said,

    June 10, 2013 @ 5:31 pm

    "And what about said,": Beg to differ with you (in all the relevant hits at COCA).

    Speaking of which, where did "difference with" come from? COCA examples:

    "Ray, that's another huge difference with Syria." [That is, between Bahrain and Syria.]

    "I think, you know, the other difference with 1980 is that there seems to be a meanness of spirit now in the country, certainly in the tone of the debate."

    "Explain to us, Dr. Hendi, what's the difference between these UVA and UVB rays? There was already protection, supposedly, against UVB. But what's the difference with those and the ultraviolet-A rays?"

    These are only a minority of uses of "difference with", so I didn't look for any historical trends. However, I blame Wikipedia, which used to show you "Difference with previous version" (I think) when you clicked on "diff". It now shows you "Difference between revisions".

  19. Alex Leibowitz said,

    June 10, 2013 @ 9:59 pm

    When we say things are different, I think we must consider them to be separate and distant, insofar as they are different, and "from" seems like a good way of indicating their separation, whereas "to" might imply the opposite sense, that they are, in their difference, somehow moving towards each other. I like to think there is some logic in our choice of prepositions, however faint and analogical.

  20. zeenoster said,

    June 11, 2013 @ 4:38 am

    I am completely different against you. (?)

  21. GeorgeW said,

    June 11, 2013 @ 5:42 am

    @J.W. Brewer: I don't think that preposition meaning is quite so arbitrary. The abstract meaning is often related to the concrete meaning. The concrete meaning of 'from' involves distance or separation. So, it is, I think, likely that speakers would generally select a pronoun with a similar meaning for an abstract context.

    But, anyone who has learned a foreign language realizes that this is not completely predictable. As an example, in Arabic, one is married 'from' (min) their spouse.

  22. J.W. Brewer said,

    June 11, 2013 @ 9:16 am

    Ex ante the spatial metaphors where the different-from-each-other things are facing toward each other (emphasizing consciousness of the distinction, could perhaps result in "to") and where the distinct things have their backs turned to each other (emphasizing separation and perhaps movement further away, could perhaps result in "from") and even where the distinct things are being viewed by some third-party observer as occupying different points along some sort of scale or continuum (could perhaps result in "than") all seem entirely plausible. Maybe I shouldn't have said arbitrary without further qualification – but as with "in line" versus "on line" there may often be multiple candidate prepositions where an extended meaning would be plausible, with the process that controls which becomes (or become, if more than one option is in common use) the fixed idiom for the particular context seeming essentially arbitrary.

  23. GeorgeW said,

    June 11, 2013 @ 9:39 am

    @J.W. Brewer: I think the point is that prepositions, in abstract ideas, are not perfectly predictable nor are they completely arbitrary. There is often a relationship between the concrete and the abstract meaning.

  24. Greg Malivuk said,

    June 11, 2013 @ 3:08 pm

    Interestingly, GloWbE seems to have 19 distinct true examples of "[differ] than" in its database. All of them sound intensely ungrammatical to me, but evidently less so to others.

  25. Dan Hemmens said,

    June 11, 2013 @ 3:17 pm

    I can see however the usefulness of We use a different method than they do. Since *We use a different method from they do is impossible, sticklers for “different from” have to say something more roundabout, and arguably more clumsy, such as “We use a method that differs from the one that they use”.

    While I would never underestimate the irrationality of the peever, I think even people who insist that "different than" is "wrong wrong wrong" would be okay with "we use a different method than they do" – if nothing else, peevers tend to get hung up on very *specific* rules and I suspect that therefore "different [word] than" won't trigger the same endorphins that straight "different than" would.

    For what it's worth, I'd probably say something like "we use a method that is different from theirs" but then again, I might say "he went to a different university than I did".

  26. Ken Brown said,

    June 11, 2013 @ 4:59 pm

    In my English I can say "he went to a different university than I
    I did" but I think I'm more likely to say "he went to a different university to me". "From me" and "than me" work as well. But not *" from I did".

  27. Geoff Nunberg said,

    June 12, 2013 @ 12:42 am

    Dwight Bolinger observed (somewhere) that there was a contrast between "How different this seems from Paris" and "How different this seems than Paris," where the latter suggests a comparison of events rather than places.

  28. mollymooly said,

    June 12, 2013 @ 5:55 am

    I don't recall ever seeing or hearing "differ to" in Ireland (or anywhere else). Perhaps Butterworth was trying and failing to adapt his native idiom to that of his readers; though I suppose it should be the copyeditor who makes such decisions, or at least fixes them if they go wrong.

  29. Peter CS said,

    June 12, 2013 @ 12:47 pm

    In 1950s Scotland I was taught to use 'different from' when comparing two largely dissimilar things, but 'different to' when the objects were more similar. So the climate of Iceland is different from that of the Congo, but a Bactrian camel is different to a Dromedary in having two humps instead of one.

  30. blahedo said,

    June 12, 2013 @ 3:43 pm

    Interestingly, one justification I've heard from different-from-only peevers is that "different" should use "from" precisely because "differ" uses "from", and that we don't say "different to" because we don't say "differ to". The correct response to that argument is that language isn't always (or even usually) very logical, of course.

    But it does make me wonder if the increasingly-accepted usage of "different to" is influencing "differ" and triggering usage of "differ to".

  31. The Negative Canon: Different Ways of Being Different | Caxton said,

    July 21, 2013 @ 2:56 am

    […] Others have trodden this path before me, and I need do little more than refer readers to Stan Carey’s comprehensive post on different from, different than and different to, and to Mark Liberman’s additional analysis of the corpus evidence, and his examination of the verb differ to, on Language Log. […]

RSS feed for comments on this post