Rhetorical testosterone and analytical hallucinations

« previous post | next post »

In her most recent column ("Obama: Our first female president", 7/1/2010), Kathleen Parker argues that Barack Obama writes like a girl:

If Bill Clinton was our first black President, as Toni Morrison once proclaimed, then Barack Obama may be our first woman President. […]

No, I'm not calling Obama a girlie President. But … he may be suffering a rhetorical-testosterone deficit when it comes to dealing with crises […]

What's her evidence for this lack of "rhetorical-testosterone"? Along with a lot of vague stuff about how Obama is "a chatterbox" who shares with "Ronald Reagan and Bill Clinton" (!) the ability to "assume feminine communication styles effectively", the column includes exactly one relevant fact:

Obama's [oil spill] speech featured 13 percent passive-voice constructions, the highest level measured in any major presidential address this century, according to the Global Language Monitor, which tracks and analyzes language.

If you're not a regular reader, please take a few minutes to scan our last discussion of linguistic "analysis" from Paul Payack's Global Language Monitor ("Language guru runs with the journalistic pack", 6/17/2010). According to Mr. Payack, president Obama's address on the gulf oil spill was excessively "professorial" because its average sentence length was 19.8 words. I checked on president George W. Bush' post-Katrina speech, and found that its average sentence length was 23.5 words, suggesting either that Bush was even more "professorial" than Obama, or that Mr. Payack was full of it.

So what about those passives?

The first thing to say is that there isn't the slightest evidence that passive-voice constructions are "feminine".  Women don't use the passive voice more than men, and among male writers, number of passive-voice constructions doesn't appear to have any relationship at all to real or perceived manliness. The "passive is girly" prejudice seems to be purely due to the connotations of (other senses of) the term passive, misinterpreted by people who in any case mostly wouldn't recognize the grammatical passive voice if it bit them on the leg. (See e.g.  "When men were men, and verbs were passive", 8/4/2006; "How to defend yourself from bad advice about writing", 11/1/2006; "'Passive Voice' — 1397-2009 — R.I.P.", 3/12/2009.)

[And the idea that women are "passive" in a non-grammatical sense is an equally silly stereotype.]

Still, there's a point of fact here. Did president Obama's speech really have more passive-voice constructions than "any major presidential address this century"?

Getting a meaningful number for "percent passive-voice constructions" requires some definitions. What are we taking a percentage of: sentences? tensed clauses? all clauses? all constructions where voice could be varied? And what counts as passive? do get-passives, adjectival passives, passive-participle modifiers, etc., add to the total or not?

I'm not sure what definitions Paul Payack used. For some evidence that he's among those who don't have a clue what a passive-voice construction actually is, in the traditional grammatical sense, see Ben Zimmer's post "There will be passives", 11/7/2008.  I don't have time this morning to try to figure out how Mr. Payack derived his passive percentages, if any information about this is available — I'll have more to say when I've looked into this further.

But I did just make a quick analysis of president George W. Bush's post-Katrina address to the nation. I count 142 sentences, 25 of which contained one or more passive-voice tensed verb constructions. That's 17.6%.

Doing the same thing with Barack Obama's post-oil-spill address, I count 135 sentences, 15 of which contain one or more passive-voice tensed verb constructions. That's 11.1%.

If instead I calculated the percentage of tensed verbs that are in the passive voice, or the percentage of voice-relevant constructions that are passive, I'd get somewhat different numbers. But I don't believe that any of these alternative calculations would rescue Mr. Payack's assertion, or Ms. Parker's little exercise in empirically vacuous meme-replication. She wrote:

We've come a long way gender-wise. Not so long ago, women would be censured for speaking or writing in public. But cultural expectations are stickier and sludgier than oil. Our enlightened human selves may want to eliminate gender norms, but our lizard brains have a different agenda.

Parker's lizard brain, I'm sorry to say, seems to have the agenda of promoting — with less than no evidence at all — one of the currently fashionable journalistic tropes about Obama.

[For some background that may help you to evaluate the credibility of Paul J.J. Payack as a linguistic analyst, see "There will be passives", 11/7/2008; "The million word hoax rolls along", 1/3/2009; "Forbes on neologisms, and the return of the million-word bait-and-switch", 4/23/2009; "The millionth word in English could be 'sucker'", 5/12/2009; "End times at hand", 6/6/2009.]

[Update, July 2 — On June 16, "admin" at Paul J.J. Payack's "Global Language Monitor" site posted an item "Obama Oil Spill Speech Echoes Elite, Aloof Ethos", which includes this:

The President’s first Oval Office address came in at a surprising high tenth-grade reading level, with some 13% passive constructions, the highest level measured in any major presidential address in this century.  In political speaking, the passive voice is generally used to either deflect responsibility, or to have no particular ‘doer’ of an action.

The post seems to have been updated several times since June 16 — for instance, it includes a link to Kathleen Parker's column, using the curious link-text "Kathleen Parker’s ‘Empiracally Vacuous Meme-replication’".

But I could find no discussion of the methodology used to determine the "13% passive constructions", nor any information about what else counted as a "major presidential address this century", nor any data on what passive percentages Mr. Payack assigned to these other speeches.

A couple of years ago ("There will be passives", 11/7/2008), Ben Zimmer examined in detail Payack's assertion (picked up by CNN and others) that in the vice-presidential debate

Passive voice can be used to deflect responsibility; Biden used active voice when referring to Cheney and Bush; Palin countered with passive deflections.

Ben observes (following Gabe Doyle) that this was simply false: Gov. Palin didn't use a single passive-voice sentence in any relevant region of the transcript.

Ben also notes this passage from a CNN article about Obama's victory speech, where Mr. Payack cites a single example of "passive voice" — which isn't actually in the passive voice:

Though most of Obama's verbs were in the active voice, 11 percent of the sentences were in the passive voice, a dependable method of deflecting responsibility, Payack said. He cited Obama's "There will be setbacks and false starts" as an example.

"He's spreading the responsibility around," Payack said. "He didn't say, 'I will have setbacks. I will be wrong. I will make mistakes.' He used the passive voice for those types of constructions."

Ben's conclusion:

Lay discussions of the "passive" can tell us a lot about folk-linguistic beliefs (see, for instance, the treatment of "folk grammaticality" in Nancy Niedzielski and Dennis R. Preston's 1999 monograph Folk Linguistics, pp. 270ff.). But in the CNN article we have a self-styled "linguist" throwing around words like "passive" in a spectacularly uninformed fashion. We've set the bar pretty low if we're listening to an "expert" whose knowledge in his professed field of expertise wouldn't get him very far on "Are You Smarter Than a 5th Grader?".

I could find nothing in the GLM's analysis of Obama's oil-spill speech, or indeed anywhere else on the GLM site, to suggest that Mr. Payack has now mastered grade-school grammar.]

[For more on the relation between passivity and passive voice, see "Mark Steyn uses the passive to avoid passivity", and "More on the stupidity of Kathleen Parker".]


  1. Mark P said,

    July 1, 2010 @ 9:19 am

    It's no use. We'll just have to wait this out until some reporter somewhere thinks up something new, and everyone rushes off to start mining that vein.

  2. Outis said,

    July 1, 2010 @ 9:31 am

    In the spirit of descriptivism, maybe it's time we redefine "passive voice" as anything that sounds "professorial" or "feminin", which in turn could be redefined as anything Fox News and their contributors say it is.

    Like it or not, Fox News & co. are possibly the single-most cohesive force in the ongoing American English evolution.

    [(myl) Let's not indulge in stereotypes here. Kathleen Parker is a columnist for the Washington Post, who was also recently established as co-host (with Eliot Spitzer!) of a new CNN show due to start in the fall.

    And neither Paul J.J. Payack nor his "Global Language Monitor" has anything at all to do with Fox News, as far as I'm aware, nor has Fox been nearly as prominent as various other outlets in pushing Payack's various nonsenses.]

  3. Dierk said,

    July 1, 2010 @ 9:50 am

    Simple test for those coming up with grand claims based upon President Obama's perceived wimpiness or femininity: Let them deduce the sex of authors they only read actual texts without knowing who wrote it. Like, say, George Eliot's novels*.

    *I know, it's an obvious one, but I've lived through numerous occasions in which English majors made her a him.

  4. Cecilieaux Bois de Murier said,

    July 1, 2010 @ 10:01 am

    You stole my letter to the editor. Thank you, thank you, thank you. She'll at least notice you.

  5. David L said,

    July 1, 2010 @ 10:06 am

    I admire your diligence in debunking these absurd "quantitative" arguments. But as long as the debunking remains in LL, it's reaching a limited audience. Have you thought about writing an oped piece for the NYT or WaPo on these topics? It's the kind of thing I could see appearing in the Sunday Review/Outlook sections. No guarantee they would take it, of course, but it would reach a larger audience and provide a handy reference for bloggers and others who want to argue against George Will and his incurious buddies.

    After all, the reason Payack gets so much ink is not simply because of the credulity of too many reporters, but because he works to publicize his stuff.

  6. Spell Me Jeff said,

    July 1, 2010 @ 10:20 am


    Please note that I couched my "counter-program" in all manner of conditionals and qualifiers. I don't expect any thinking person to buy it. Merely to show that if one wanted to engage in such shallow gender-mapping, another route leads to a different conclusion.

  7. Yuval said,

    July 1, 2010 @ 10:27 am

    Seriously though: isn't there some way respectable linguists such as yourselves can issue a fraud-warning against Payack in a way at least the major news outlets can see it and avoid quoting him anymore? I'm assuming (or hoping) some natural sciences have such mechanisms.
    The post-hoc warnings and analyses in this blog are clearly not doing the job.

    [(myl) If we consider the record in areas such as the vaccine-causes-autism canard, HIV-AIDS denialism, fish oil for schoolchildren, homeopathy, etc., the prospects for innoculation of mass media against bogosity don't look very good — though I agree that we should keep trying, and perhaps look for a higher platform or a louder megaphone.

    In our own case, at least, our failure has been a non-partisan one. Despite years of trying, we never succeeded in slowing down the Bushisms industry.]

  8. Spell Me Jeff said,

    July 1, 2010 @ 9:44 am

    What a bunch of hooey. As writers from Orwell to Joe Williams have pointed out, the passive voice is often (but certainly not always) the voice of choice for anonymous bureaucrats who wish to obscure the agency of nefarious behavior: eg, "Bombs were dropped," etc.

    [(myl) The earliest example of such a claim that I know of is William Cobbett, writing as Peter Porcupine in 1797 ("All verbal assassins speak in the passive voice"). Despite its antiquity, however, I don't think that any actual evidence for this assertion has ever been put forward. And Orwell undermines his own credibility (as he generally does in the matter of linguistic commentary) by using the passive voice more often than average (see "Passive aggression", 7/18/2006).]

    I suppose it might be argued that such obscurantist language is characteristically male, since we are more inclined to associate it with chicanery and the worst sort of war-mongering. I'm not sure I wish to make that case, but if the Parkers or Tannens of the world wish to make it, I think the fodder is there.

    But to make such a case in Obama's speech (notice I'm hypothetically arguing for a position contrary to Parker's) counting passives would be insufficient. One would have to count the number of times that agency was obscured, and furthermore the number of times when the obscurity was significant. Eg, were there sentences along the lines of, "Mistakes were made"?

    [(myl) I think you'll find it very hard to get agreement about how to count instances of "significant obscuring of agency". For example, when you write "to make such a case in Obama's speech", you don't specify the agent who would make the hypothetical case. Was agency therefore "obscured"? If so, was the obscurity significant? What about your "one would have to count"? Does the generic pronoun "one" obscure agency? And so on.

    Even if you could find a way to make such annotation intersubjectively valid, I'd be somewhat surprised to find reliable associations with gender.]

  9. Jerry Friedman said,

    July 1, 2010 @ 11:15 am

    It's been clear since one of the earlier discussions of Strunk and White that it would be great if "agentless" constructions could be defined. Then there could be studies of correlation with gender, author's political position when known, clarity as determined by comprehension tests, etc.

    I've thought since one of the earlier discussions of Strunk and White that I'd be glad if linguists could define "agentless" constructions. Then they could look for correlations with gender, author's political position when they know it, clarity as determined by comprehension tests, etc.

    (I don't see the reason for the count to be limited to tensed verbs; your count was only of passives, not of agentless constructions.)

    [(myl) One reason to limit the count to tensed verbs is that there's a sort of slippery slope here. There are passive infinitive clauses after active main verbs ("We believe this passage to have been corrupted" — corrupted by whom?); post-nominal passive modifier phrases ("We attribute this text to a man generally called Master Zhuang" — called by whom?); and pre-nominal modification by passive participles ("a well-known forger" — known by whom?). But the same sorts of evasions, if you want to think of them that way, can be expressed with adjectives ("a notorious forger" — notorious to whom?) or nouns ("the forgery of this document" — forged by whom?).

    You can plausibly argue for various alternative ways to define the denominator of the proportion. But if you open it up to any mention of any action or process that might in some alternative mode of expression have had a subject — or worse, any of the semantic roles that subjects commonly express — things are going to get very messy.

    Looking instead for proportions of cases where agency is specified is not going to be any clearer. Again, what's the criterion for identifying cases in which an agent might have been specified? What about cases where knowledge of agency, or even the existence of agency, is explicitly denied? What about hypothetical or generic actions, where it may not make sense to specify an agent? What about cases where the agent is contextually obvious, but not locally repeated?

    You could make up a set of guidelines for inter-subjectively reliable annotation of agency, but it would be require the usual complex "common law" sort of process of specifying how to deal with the long tails of various family-resemblance chains of circumstance.]

  10. Geoffrey K. Pullum said,

    July 1, 2010 @ 11:39 am

    Mark says:

    The "passive is girly" prejudice seems to be purely due to the connotations of (other senses of) the term passive

    Quite so. But this is not only a vague and confused equation; even in non-linguistic terms it crumbles into dust if you spend ten seconds thinking about it. What is the sense in which women are supposed to be passive? Even the worst stereotypes have them as more active than men in various domains (nagging; shopping; gossiping). What are they supposed to be passive at? It isn't remembering to do the laundry, at least in many households, that's for damn sure. So is there a hint here that what we're talking about is sexual practices? Is the "passive = womanlike" thing supposed to be about taking a passive role in sexual acts? Girls lie down on their backs and stay still while guys get on top of them and bounce up and down? It looks to me like the people who suggest such things must have a very narrow range of sexual experience if they haven't noticed that the variation in sexual aggressiveness within a given sex is greater than that between the averages of the two sexes. It really is utterly fantastic that someone can think there is a point about sex characteristics to be made here at all. And I say that even before I get to work on replicating Mark's work to see whether Mr Payack, who we already know can't tell actives from passives, is even roughly accurate in his figures linking it to language. My God, there is a lot of garbage about language passed around in the media. Kathleen Parker, you should hang your head in shame.

  11. Jerry Friedman said,

    July 1, 2010 @ 11:50 am

    Thanks to a Language Logger for trying to fix the sentence I'd messed up, as I mentioned in my next post. However, it didn't come out quite right. The last sentence in my post above should read, "I don't see the reason for the count to be limited to tensed verbs, or for you to limit your count to tensed verbs."

  12. Amelia Eve said,

    July 1, 2010 @ 11:53 am

    This article also raises a different question for me: girlie vs. girly. In my understanding, girlie is a noun and girly would be the adjectival form. We do not say things that smell bad are stinkie or that computer programs with numerous errors are buggie.

    I've seen this -ie spelling of the adjectival suffix for girl elsewhere, so it's not unique to the WaPo. MS Word passes both spellings, so I don't think its a Cupertino. I can only conclude that the copy editor was a dummie.

  13. Writes Like A Girl « Shitty First Drafts said,

    July 1, 2010 @ 12:00 pm

    […] is so screwy that it's hard to figure out exactly what she is saying), though as the Language Log helpfully reminds us: The first thing to say is that there isn't the slightest evidence that […]

  14. Will said,

    July 1, 2010 @ 1:30 pm

    @Amelia, what about a horse-and-buggy?

    But of course your point wasn't about words ending in -y, but about words ending it -ie. It's true this is a suffix (marking diminutiveness or cuteness) that turns a noun into another noun, and never into an adjective.

    But I think girlie is a special case where it can simply be part of the lexicon as an adjective in it's own right, separate from girly. And in fact I think it does have a somewhat different meaning than girly.

    I think the meaning has comes as a generalization from the meaning in the idiomatic phrase "girlie boy", which is frequently spelled like that (for a crude basis of comparison, "girlie boy" has 53k ghits and "girly boy" has 50k ghits). Even "girlie man" (which is not really as idiomatic like "girlie boy") has a third as many ghits as "girly man".

    So I think a "girlie President" has different connotations than "girly President", and at least in my vocabulary the former sounded like it had the better usage in context.

  15. Taylor Selseth said,

    July 1, 2010 @ 3:19 pm

    Is obviously seems like to me to be a fusion of said negative associations of the term "passive" together with popular prescriptivist shibboleths that rant against using the Passive Voice as "weak writing" or such nonsense that was driven into my skull in high school English.

    But That's probably over-thinking what is essential an anti-Obama hit piece using any stupid reason they can find to bash him.

  16. Jerry Friedman said,

    July 1, 2010 @ 3:45 pm

    @myl: If I can post on the topic of syntax and out of my depth…

    I think there are two separate issues. One is your count of passive and active verbs. I don't see why you didn't count infinitives. Is it sometimes ambiguous whether an infinitive is passive or active? I can't think of any, and I don't see that you mentioned any in your reply. I can see ambiguity about whether a past or present participle is more like a verb or an adjective, but linguists have probably solved this one (several ways), and if you can state that it's a verb, wouldn't it clearly be either active or passive?

    The other issue is my call for a definition of agentless constructions. You brought up a number of problems I hadn't thought of. Yet another one is a construction that has a syntactic agent but gives no information about it: "Many people consider…" No doubt (there's another one) this has many variations.

    On the subject of when the agent could have been specified, I think the question relevant to style (Obama's, or advice from Strunk and White) is whether it should have been. The problem is lack of explicitness, and there's no problem where the speaker or writer has been explicit about an unknown agent or the lack of one. (There may be some problem where the speaker or writer has grudgingly mentioned the agent and later avoids mentioning it in a disingenuous attempt to de-emphasize it.) Here the problem would be deciding when identifying the agent helps the reader and when it doesn't, which could be a matter of taste.

    You could make up a set of guidelines for inter-subjectively reliable annotation of agency, but it would be require the usual complex "common law" sort of process of specifying how to deal with the long tails of various family-resemblance chains of circumstance.

    You linguists' ability to solve problems like that is why you're the wealthy and glamorous figures you are.

    Seriously, I think the interest in "passive" constructions is really an interest in agentless constructions, except when it's political bunk, so some people might be interested in a linguistically valid approach to the problem. (Here I expect a linguist to point out that someone has solved it, with a reference, and the solution hasn't made the Washington Post.)

  17. Bill Walderman said,

    July 1, 2010 @ 3:47 pm

    What's with this green type? It's hard enough to read against a white background, but impossible against a grey one.

  18. Nina said,

    July 1, 2010 @ 3:47 pm

    Kathleen Parker gets weirder and more nonsensical every year.

    [(myl) Glynnis MacNichol attributes the column under discussion to the aftereffects of a "MoDo pill". Another theory might be that Ms. Parker is just preparing herself for the characteristic intellectual level of cable news.]

  19. Scott said,

    July 1, 2010 @ 4:17 pm

    Yeah, that's the first thing I look for in a woman: a good set of passive verbs…

  20. Daniel Ezra Johnson said,

    July 1, 2010 @ 5:00 pm

    Why did the comments turn green?

  21. Bloix said,

    July 1, 2010 @ 5:12 pm

    I have only an ordinary high-school understanding of passive voice, but in my review of the speech (admittedly a once-over), I found only eight passive-voice constructions and dozens of active voice constructiions. The passive voice constructions do not hide agency:

    Seventeen others were injured.

    Millions of gallons of oil have already been removed from the water …

    Over five and a half million feet of boom has been laid across the water…

    And this fund will not be controlled by BP.

    [T]he account must and will be administered by an independent, third party.

    The plan will be designed by states, local communities, tribes, fishermen, businesses, conservationists, and other Gulf residents.

    At this agency, industry insiders were put in charge of industry oversight.

    Oil companies … were essentially allowed to conduct their own safety inspections and write their own regulations.

    [I also counted the seven sentences containing these sequences:

    …in areas where beaches are not yet affected.
    … whatever resources are required to compensate the workers and business owners who have been harmed…
    … the necessary precautions would be taken.
    … were essentially allowed to conduct their own safety inspections …
    … the path forward has been blocked.
    … the same thing was said about our ability …
    … a tradition that was brought to America long ago by fishing immigrants from Europe.

    for a total of 8+7 = 15.

    Nice to see that I didn't miss any obvious ones.]

  22. Eric said,

    July 1, 2010 @ 5:30 pm

    "people who in any case mostly wouldn't recognize the grammatical passive voice if it bit them on the leg"

    Don't you mean:

    "people who in any case mostly wouldn't recognize the grammatical passive voice if they were bitten on the leg by it"?

  23. Andrew said,

    July 1, 2010 @ 6:03 pm

    I've read complaints about Obama saying "I" too much, and "we" too much, and now the passive voice. What's left? "Barak Obama is emulating Bob Dole?"

  24. Kylopod said,

    July 1, 2010 @ 7:34 pm

    @Jerry Friedman

    Yet another one is a construction that has a syntactic agent but gives no information about it: "Many people consider…" No doubt (there's another one) this has many variations.

    Isn't that a form of what's called weasel words? It includes attempts to give your assertion a sense of authoritative backing by vaguely referring to some unnamed consensus. Sentences which begin "It has been said that…" are examples of this practice, and whether or not they use passive voice is incidental.

  25. How Wagner Called Hollywood’s Tune - Idea of the Day Blog - NYTimes.com said,

    July 2, 2010 @ 6:41 am

    […] Female President' in the Way He Communicates – Kathleen Parker, The Washington Post (A Linguist Dissents — Mark Liberman, Language […]

  26. the other Mark P said,

    July 2, 2010 @ 6:42 am

    We've come a long way gender-wise. Not so long ago, women would be censured for speaking or writing in public.

    Quite apart from the ridiculous "passive" is feminine thing, how could anyone write this?

    Really, when was the last time it was considered reasonable that a woman be censured for speaking or writing in public just because she was a woman?

    I'm not suggesting that women have had a fair shake of the stick, but they have been active members in public discussion for rather longer than "not so long ago". I would not consider Margaret Chase Smith to be recent history, and women have had some voice (if not an equal one) for far longer than that.

    But, I suppose facts shouldn't get in the way of a good story.

  27. Leo said,

    July 2, 2010 @ 10:37 am

    Dierk: "Simple test for those coming up with grand claims based upon President Obama's perceived wimpiness or femininity: Let them deduce the sex of authors they only read actual texts without knowing who wrote it."

    There is a website which attempts to do this, algorithmically. I won't link it here because I don't want to endorse it, and it's quite old now anyway, but it's quite easy to find using your favourite search engine, if you're interested – strictly for a laugh of course.

  28. Michael Koplow said,

    July 2, 2010 @ 10:48 am

    Parker is quite explicit about linking the grammatical passive with real-life pasivity: "When he finally addressed the nation on day 56 (!) of the crisis, Obama's speech featured 13 percent passive-voice constructions, the highest level measured in any major presidential address this century, according to the Global Language Monitor, which tracks and analyzes language.

    "Granted, the century is young — and it shouldn't surprise anyone that Obama's rhetoric would simmer next to George W. Bush's boil. ****But passivity in a leader is not a reassuring posture****" (emphasis added).

    By the way, I copied the comments into word and changed the font to black.

  29. Michael Koplow said,

    July 2, 2010 @ 11:53 am

    Hoops. In the last sentence of the previous comment, I meant "Word" instead of "word."

  30. Army1987 said,

    July 2, 2010 @ 1:53 pm

    If I had been forced to associate a gender to the passive voice, I would have said it's masculine. (The ‘reason’ is that the passive voice is typically associated to technical language, which in turn is typically mostly done by men.)

  31. Oh, good, this again « Motivated Grammar said,

    July 2, 2010 @ 3:07 pm

    […] in any major presidential address this century". And that's something, except for Mark Liberman's discovery that it's not nearly true. Bush's similar post-Katrina address had 17% […]

  32. Diane said,

    July 2, 2010 @ 3:21 pm


    So, what *do* you think the difference is between a "girlie President" and a "girly President"?

  33. Joe said,

    July 2, 2010 @ 4:38 pm

    Probably off point, but I love the headline, "Obama Oil Spill Speech Echoes Elite, Aloof Ethos." It's easy to parse, but still. . .

    [(myl) Just try to say it five times fast!]

  34. Jerry Friedman said,

    July 4, 2010 @ 10:57 am

    @Kylopod: The American Heritage Dictionary defines weasel word as "An equivocal word used to deprive a statement of its force or to evade a direct commitment." I suppose an example would be "Save up to 35%!" The "up to" evades any commitment. (I always thought the etymology was that such words allowed you to "weasel out" of the commitment, but AHD says it's "From the weasel's habit of sucking the contents out of an egg without breaking the shell." Without breaking it in two, I take it.)

    At Wikipedia, and maybe elsewhere for all I know, weasel word is used in the way you mention, as in my example of "Many people consider". I'm guessing the connection between these two senses is that "many people consider" can be an attempt to evade the speaker's responsibility for the statement, but Wikipedia doesn't like it because it gives the statement a specious air of authority, as you say.

  35. Kylopod said,

    July 4, 2010 @ 9:16 pm

    Wikipedia adopted the term "weasel words" for its own policy because it has a problem of questionable contributions worded this way (as in, "Observers have noted that George W. Bush has a close physical resemblance to that of a chimpanzee"). But you're right that this was previously never the main sense in which the phrase was used.

    It would be nice if we could just have a general term for the practice of using language to obfuscate one's intentions or deflect responsibility. "Orwellian" comes pretty close, but is usually applied only to the most visible examples. Barring that, it's easy to see how the passive voice has become the fall guy among the less grammatically literate.

  36. Kylopod said,

    July 4, 2010 @ 11:07 pm

    Let me amend my previous comment about the term "Orwellian." Although Orwell probably did more than any other writer to call the public's attention to the ways in which language is used to obfuscate, the adjective named after him is usually used in a very narrow sense: for euphemistic phrases that brazenly try to pass themselves off as the opposite of what they really are. More mundane examples, like O.J.'s "We had abusive situations," are rarely termed Orwellian.

  37. Obama, First Woman President « Feminist Philosophers said,

    July 7, 2010 @ 11:54 am

    […] Philosoraptor has pointed us to this awesome debunking of the linguistic "facts" in which the article was based.   Comments […]

RSS feed for comments on this post