Now presenting… Muphry's Law

« previous post | next post »

Success has many fathers, the old saying has it, and the same goes for a well-turned maxim. We've noted a number of different originators for what Jed Hartman called the Law of Prescriptivist Retaliation: corrections of linguistic error are themselves inevitably prone to error. Around 1999 this truism was hit upon by no less than three independent sources: Hartman, Erin McKean, and alt.usage.english contributor Skitt. And 90 years before that, Ambrose Bierce expressed much the same sentiment. Now it appears that the law has yet another eponymous author: the mythical Mr. Muphry.

"Muphry's Law" came to my attention in the comments on Stephen J. Dubner's "Freakonomics" blog on the New York Times website. As I wrote in the post "Of pasties and pastries," Dubner faulted an article in The Economist that referred to "Cornish pasties" for sale outside of Mexico City. He thought pasties was an error for pastries, since he was unfamiliar with the British culinary term pasty (instead construing pasties as the plural of the American term pastie, the stripper's ornamentation). Of the commenters who corrected Dubner's correction, one wrote: "You’ve just encountered Muphry's Law (no, not Murphy’s)." Dubner picked up on the comment in a subsequent post (in which he describes receiving a Cornish pasty in the mail, sent from the wags at The Economist to spare him further confusion).

Muphry's Law, it turns out, was so named by John Bangsund of the Victorian Society of Editors in Australia. In a 1992 article in the Society of Editors Newsletter, Bangsund set out a four-part principle:

(a) if you write anything criticizing editing or proofreading, there will be a fault of some kind in what you have written;

(b) if an author thanks you in a book for your editing or proofreading, there will be mistakes in the book;

(c) the stronger the sentiment expressed in (a) and (b), the greater the fault;

(d) any book devoted to editing or style will be internally inconsistent.

Dubner's goof might not be the best example of Muphry's Law, however. The canonical case of prescriptivist retaliation requires an actual error to exist in the first place, with one or more errors occurring in the correction to the original error. As Jeremy Cherfas pointed out in a comment to my "pasties" post, "Dubner was pointing out an error that was not an error." So Dubner falls more into line with the examples discussed by Mark Liberman in his post, "Why are so many linguistic corrections incorrect?" Mark borrows the term "incorrection" from William Safire ("a correction that is itself incorrect"), but Coby Lubliner has argued that "miscorrection" is a better term for this phenomenon (which has come up so often in our posts on the Cupertino effect).

Note the last link there, for the Wikipedia page on the Cupertino effect. As it happens, there's also a brand-new Wikipedia page on Muphry's Law, with a panoply of variations from Bierce onwards. Both articles show the dissemination of Language Log scholarship beyond the friendly confines of LL Plaza. Nice to have some online company for the eggcorns and the snowclones.


  1. John Cowan said,

    July 21, 2008 @ 7:55 pm

    I added an explicit defense of inconsistency (Bangsund's type d) to my revision of Strunk's "little book":

    I have attempted to remain within the scope of the original. This book, therefore, is intended as a compendium of helpful advice to novice writers in freshman composition classes, not a code of general laws of writing for all works by all writers in all circumstances. Violations of the rules can be found within the book itself — this is neither inconsistent nor hypocritical, as The Elements of Style Revised is not a paper written for a composition class.

  2. Quicksand said,

    July 21, 2008 @ 8:17 pm

    Would it, perhaps, induce me to generate HTML errors of my own if I point out that the Wikipedia link for eggcorns, in the final sentence of the original post, is malformed (extra "http//")?

  3. Adrian Morgan said,

    July 21, 2008 @ 8:40 pm

    There are places on the Internet (particularly Usenet) where the received wisdom is not so much that a pedantic correction will contain an accidental error (by causality) as that it must contain a deliberate error (by obligation)…

  4. Benjamin Zimmer said,

    July 21, 2008 @ 9:05 pm

    Quicksand: Well, there couldn't be a post about Muphry's Law without at least one error. Now fixed (I hope).

    Adrian: Don't you mean a *pendantic* correction? (That's one of many ingroup shibboleths on the newsgroup alt.folklore.urban.)

  5. language hat said,

    July 22, 2008 @ 8:41 am

    John: Might it not be a good idea to indicate (by color or typeface) what you have added to or changed in Strunk's text? This would be easy to do in a Web-based text, and would comfort readers who care about attribution and authorship.

  6. Martyn Cornell said,

    July 22, 2008 @ 4:19 pm

    Is there a difference between an incorrection ("a correction that is itself incorrect"), and a hypercorrection (a "correction" of something that was not incorrect in the first place), which is what I would argue Stephen J. Dubner's "pastries" was?

  7. Benjamin Zimmer said,

    July 22, 2008 @ 4:24 pm

    Martyn: I suppose you could say a hypercorrection is a type of self-{in/mis}-correction, one in which the user over-extends a particular pattern (usually a phonological one) beyond the intended target. I don't quite see how Dubner's error was a hypercorrection, though — at least in the way that linguists tend to use the term.

  8. John Cowan said,

    August 1, 2008 @ 5:24 pm

    Hat: There is a diff at , which I link to using the words "revisions to the original" in my introduction.

RSS feed for comments on this post