A sad case

« previous post | next post »

A few days ago, Ben Goldacre, or someone pretending to be him on twitter, tweeted

dear everyone, when i read your passive sentence constructions i sort of have to convert them into active ones in my head because i'm thick.

As Geoff Pullum recently observed

I despair when I see this kind of drivel. What on earth comes over people when they write about language? It’s not just their ability to use dictionaries that disappears, it’s their acumen, their numeracy, their common sense.

The despair in this case is especially intense, because Dr. Goldacre is unusually well supplied with acumen, numeracy, and common sense, and is an accomplished practitioner of "The noble and ancient tradition of moron-baiting". So it's with a heavy heart that I turn to the first paragraphs of his most recent three Bad Science posts, noting the "passive sentence constructions":

While the authorities are distracted by mass disorder, we can do some statistics. You’ll have seen plenty of news stories telling you that one part of the brain is bigger, or smaller, in people with a particular mental health problem, or even a specific job. These are generally based on real, published scientific research. But how reliable are the studies?


Now, there are dozens of different ways to quantify the jobs market, and I’m not going to summarise them all here. The claimant count and the labour force survey are commonly used, and number of hours worked is informative too: you can fight among yourselves for which is best, and get distracted by party politics to your heart’s content. But in claiming that this figure for the number of people out of work has risen, the BBC is simply wrong.


I made a documentary about prospective cohort studies in epidemiology, they’re the tool we use to find out if one thing is associated with another, where trials are impossible. It’s really good. Instead of reading about it, listen here:

Or listen live when it’s repeated tonight on Radio 4 at 9pm

I make that six "passive sentence constructions" in 19 or 20 tensed clauses, depending on how you count, for a rate of 30% or so.  No doubt in a larger sample there would be some regression to the mean, but this is an unusually high percentage. Quoting Geoff Pullum again ("The passive in English", 1/24/2011):

As mentioned on Language Log here and elsewhere, the people who criticize the passive the most tend to use it more than the rest of us. George Orwell warns against the passive in his overblown and dishonest essay "Politics and the English language". E. B. White does likewise in the obnoxiously ignorant little book he coauthored with Strunk, The Elements of Style. Both of these authors have a remarkably high frequency of passives in their work: around 20 percent of their clauses with transitive verbs are cast in the passive, a distinctly higher frequency than you find in most of the prose written by normal people who don't spend their time pontificating hypocritically about the alleged evil of the passive.

In this case, I didn't limit my calculation to the proportion of passive in "clauses with transitive verbs", but also included copular sentences like "the BBC is simply wrong", and intransitive verbs like "fight among yourselves for which is best". There are actually only three transitive tensed verbs in this sample that are NOT passive — so in six of the nine cases where a passive tensed verb construction was available, Goldacre went for it. A couple of these are arguably "adjectival passives", but still…

I should note here that Ben Goldacre is mistaken here, not just hypocritical. In my opinion, none of these examples of "passive sentence construction" would be improved by being recast in the active voice. Nor do I feel the need to "sort of convert them into active ones in my head" in order to understand them — and I don't believe that he does either.

So why is he carrying on about his difficulties in understanding passive-voice sentences?

Well, his twitter account may have been hacked by operatives of the fish-oil cartel, or by minions of Baroness Greenfield, aiming to undermine his credibility as a critic of pseudo-scientific nonsense.

Or perhaps hypoglycemia was to blame.  Subsequent tweets in the same series veer off into dystopian science fiction and irrelevant echoes of Mark Twain on German syntax, before concluding with a pathetic search for scraps of food:

actiually, someone could build a website that converts passive sentence constructions into active ones. i'll shut up now.

but it's like reading german and waiting for the verb. okay, i've really shut up now.

what was that Swift line? "if i the architect of the german language was, i the verb in a slightly less irritating position would put"

something like that anyway. i'm so hungry i can barely think. there must be some stale ryvita somewhere.

I swear I'm not making this up.

But the most likely diagnosis, I'm afraid, is temporary idiocy brought on by reckless indulgence in unprotected linguistic analysis. The patient's system may have been weakened by  prior exposure to Orwell.

[Tip of the hat to andrewjshields]


  1. Randy Hudson said,

    September 15, 2011 @ 7:14 am

    OK, Mr Goldacre, here's your web site:
    Real, published scientific research generally bases them.
    That sounds so much better…

    [(myl) If we take "These are based on real scientific research" to be a genuine passive here (rather than an essentially adjectival form), then the corresponding active would be "<someone> bases these on real scientific research", where <someone> is the journalists involved, or perhaps the news publication as a collective entity. In fact it's not really relevant in this case exactly who is responsible for the basing — the salient relationship is between the news stories and the corresponding research. So Goldacre's use of this agentless construction was entirely appropriate, and a translation into an "active sentence construction" — in his head, or by means of a "website", or whatever — would make the text harder to understand, not easier.]

  2. Stan said,

    September 15, 2011 @ 7:23 am

    Goldacre's tweets were disappointing but not surprising, since the passive voice is such a common object of ire even among people who find good use for it. I referred him (via Twitter) to The Greenbelt's helpful and sensible Virtues of the Passive Voice post, but have no reason to suppose he read it.

  3. Claire said,

    September 15, 2011 @ 8:10 am

    That's really quite bizarre. His Twitter feed appears to be a hitting the gin on a number of occasions. There is just no excuse for bad spelling.

    [(myl) "Appears to be a hitting the gin"? Those who live in glass bottles should be wary of casting the first stone, given Murphry's Law and all.]

  4. Stan said,

    September 15, 2011 @ 8:23 am

    By "people who find good use for [the passive voice]", I suppose I mean everybody — unless there are extremists who guard against its every possible occurrence in their speech and prose. And who know what it is. Which seems improbable.

  5. ENKI-][ said,

    September 15, 2011 @ 8:44 am

    I suppose people who actually use the passive voice may be more familiar with the concept — connecting it to "that thing that Orwell and Strunk and White don't like" rather than "that thing that my English teacher mentioned once in sixth grade". (Maybe it's shades of the 'eschew obfuscation' rule — wherein those who understand the rule don't necessarily benefit from it?)

    [(myl) "People who actually use the passive voice"? That would be people, period. Writing a significant amount of English prose without ever using the passive voice would be a difficult and artificial endeavor, roughly like writing without ever using the letter 'a'.]

  6. Josh said,

    September 15, 2011 @ 9:10 am

    Ah, but remember that Orwell's bombast is tempered at the end by the instruction to "break any of these rules" to avoid writing an absolute monstrosity.

    [Josh: you may have missed this highly relevant golden oldie on Orwell, from the Language Log archives: http://languagelog.ldc.upenn.edu/nll/?p=992.]

  7. Ben Hemmens said,

    September 15, 2011 @ 9:14 am

    1: The better someone can write, the less they are able to explain how to write well (privately I think of that as Pullum's Law; I'm sure GKP would rather reserve the title for something more dignified, but it was his LL posts that made me aware of it).

    2: Perversity of the Stimulus
    People learn to speak languages, and develop writing skills, while being fed lots of wrong ideas. See http://dr-ben.at/slowcom/1375/questions-about-teaching/ for a quick sketch.

  8. Moira Law said,

    September 15, 2011 @ 9:22 am

    I'm curious. Did you tweet back at Mr. Goldacre?

  9. Jerry Friedman said,

    September 15, 2011 @ 9:25 am

    Writing a significant amount of English prose without ever using the passive voice would be a difficult and artificial endeavor, roughly like writing without ever using the letter 'a'.

    A challenge! What amount is significant?

    And how noticeable would it be? I was recently involved in a quiz question involving a bit over 100 words with every letter except 'l'. A number of quite astute people failed to notice.

    Once, for a stunt, I wrote a couple pages without a figure of speech. That was hard.

  10. Ran Ari-Gur said,

    September 15, 2011 @ 9:35 am

    "A couple of these are arguably 'adjectival passives'" seems like understatement: at least three of the six (both "distracted"-s, and the "associated") seem to be clear-cut adjectival passives. And I'd argue that any case that could reasonably be an adjectival passive shouldn't be counted, even if a true verbal passive seems more likely, since we should err on the side of assuming non-hypocrisy. No?

    [(myl) Both of the "distracted" cases come with by-phrases and therefore have clear active counterparts with subjects actually named in the text ("while the authorities are distracted by mass disorder" ~ "while mass disorder distracts the authorities"). So those look pretty much like "passive sentence constructions" to me.

    But there are still only three active-voice transitives in the sample, so if we remove the examples you cite from the discussion, that's still three passives out of six tensed transitive verbs.]

  11. Seonachan said,

    September 15, 2011 @ 10:00 am

    For an actual example of "constrained writing", see Enerst Vincent Wright's 1939 novel Gadsby: A Story of Over 50,000 Words Without Using the Letter "E"

  12. GeorgeW said,

    September 15, 2011 @ 10:06 am

    Can anyone point to a substantial writing that lacks any passives and doesn't sound contrived and artificial?

  13. Ran Ari-Gur said,

    September 15, 2011 @ 10:15 am

    Well, according to CGEL, a by-phrase doesn't mean it's not an adjectival passive. ("Very distracted by ___" and "become distracted by ___" are both well attested, for example.) But I guess I have no idea what Goldacre thinks he means by "passive sentence constructions". :-P

    [Language Log has provided a detailed description of what people ought to mean by "passive" if they are following anything like the usual conception of English grammar: http://languagelog.ldc.upenn.edu/nll/?p=2922.]

  14. ben goldacre said,

    September 15, 2011 @ 10:21 am

    haha that's awesome, and i'll totally take it on the chin. i think it's possible that some readers might have understood my brief comment to be railing against a certain kind of clunking sentence construction, which i could probably have tried to define in 140 characters as my blood sugar collapsed below 3.0, but youre absolutely right.

    i might, though, say one thing on all your "i'm totally serious" "someone pretending to be him" banter, about these tweets that i wrote as i was frantically trying to finish work so i could finally eat. (less generously summarised by a commenter above as "on the gin"). it's good humoured but touches on something i've noticed more broadly, around the demands that some people make of your tweets once you've got a few more followers than average.

    i'm a normal person, sitting normally, using twitter in my home, like anyone else, mixing "here's an interesting PDF from a journal" alongside "here's an everyday human observation about a trivial aspect of my life".

    i think that's desirable, and it's what i like about other peoples' use of twitter. but more than that, i think it's *acceptable*, that anyone should be able to use twitter normally like i do, and feel free to write frantic tweets about the recurring human struggle to find food before mental collapse sets in, even if they have a few followers. people don't act or communicate in a formal mode all the time, and we shouldn't demand that they do.

    so your commentary on my humanity – and the appropriateness of its display in a forum where people are human – feels a bit overdone, but also kind of inappropriate for the context of twitter. i know it's colour, and i'm responding more to the phenomenon directed at me and at others rather than your instantiation of it, but i think i don't like it because i want everyone to keep on being human on twitter. just a thought.

    your critique of my misguided clunk about the passive construction, on the other hand, is extremely welcome, and very funny.

  15. Mark Etherton said,

    September 15, 2011 @ 10:23 am


    Or Georges Perec’s La Disparition, also without the letter e, or his Les Revenentes, with no vowels other than e, or almost anything produced by Oulipo.

  16. Sili said,

    September 15, 2011 @ 10:26 am

    Pompous windbaggery is unfortunately as common among skeptics, as it is elsewhere. The Skeptics Guide to the Universe goes off the rails when they touch upon language (I've suggested they interview a LanguageLogger, but they must be getting too many suggestions), and Jerry Coyne is run of the mill prescriptivist.

  17. Russell said,

    September 15, 2011 @ 10:26 am

    Clearly by "sort of have to convert them into active ones" is a slight confusion. He just means he takes longer than most people to compute the underlying kernel sentence that gave rise to the surface passive clause. Maybe he has to add a little oil between his be and en and everything will work out.

    On a more serious note, I think the presence of a "by" phrase does not automatically mean you have a real passive. Since the verb can describe a change into a state but the adjective cannot, one way to cut up the pie is to say that if the candidate passive clause describes a transition, it's a real passive; if not, it's been adjectivalized. Major disadvantage to this approach is not being able to tell what you've got on any given occasion.

  18. Theodore said,

    September 15, 2011 @ 10:28 am

    OK, maybe I am being thick myself, and maybe myl and all commentators are in on the joke. I don't follow Goldacre on Twitter to have a good feeling for his usual style, but this tweet:

    dear everyone, when i read your passive sentence constructions i sort of have to convert them into active ones in my head because i'm thick.

    Seems more than anything to be poking fun at passive-voice-peevers than anything, in the manner of "look at me, I'm too stupid to comprehend a common construction in my native language."

  19. Jonathan Mayhew said,

    September 15, 2011 @ 10:36 am

    Josh @9:10 a.m. That implies that breaking a "rule" would be a rare event undertaken only to avoid a worse alternative, a "monstrosity." But if the so-called rule is broken routinely, as in Goldacre's or Orwell's prose, then why call it a "rule" in the first place? Do you need to break this so-called rule in 2/3rds of the sentences you write?

  20. Neil Coffey said,

    September 15, 2011 @ 10:45 am

    It is curious that the passive seems to be such a preoccupation of linguistic folklore.

    Now, Ben Goldacre– despite his occasional rantings about the state of his fridge– is an intelligent man and devotes a good deal of his time to persuading people of the importance, for example, of referring to empirical scientific studies and not relying on the opinions and subjective interpretations of journalists.

    So I wonder if part of the problem is simply that people underestimate the extent to which language is also susceptible to scientific study? Perhaps it simply doesn't occur to people that questions such as "Do people find passive sentences more difficult to interpret than their active counterparts?" are not simply matters of Orwell and others' opinion, but are testable hypotheses and may be (and have been) the subject of scientific study?

  21. Dougal Stanton said,

    September 15, 2011 @ 10:51 am

    Like Theodore, I assumed the message was a joke. Was there follow-up discussion to suggest he was serious?

  22. Neil Coffey said,

    September 15, 2011 @ 11:12 am

    Dougal — I think the "I'll shut up now" is sort of Ben-speak for "this is an actual rant, not just a joke".

  23. J. W. Brewer said,

    September 15, 2011 @ 11:17 am

    I'm not sure if this makes the "is associated" example an "adjectival passive," but it seems really quite hard to come up with an active-voice equivalent (what one comment called the "underlying kernal sentence," although that seems to me to be unnecessary and perhaps unhelpful chomskyan phlogiston) for that one. You can't even really stick "someone" in as the dummy subject for the active-voice tensed verb, because the point of the usage in context presupposes that the "association" between A and B exists or doesn't exist prior to any human researcher inquiring into the subject and potentially "associating" A with B.

  24. Philip said,

    September 15, 2011 @ 11:28 am

    Don't past participles used as adjectives come from passive voice sentences?

    Someone lost the papers—>the papers were lost by someone—>the papers were lost—->the lost papers.

  25. Ben Hemmens said,

    September 15, 2011 @ 11:54 am

    Disparaging the passive, without understanding it, is just something scientists do:
    does it, but so does the President of the Royal Society: http://www.tsn.org.uk/writing-styles.htm

    And I suppose, under extreme hunger, latent grumps just come to the surface and we should be glad it was something this mild.

    Moral of the story, don't tweet b4 u eat.

    I think Neil Coffey's right; I doubt many scientists ever give a thought to the possibility that language experts might have anything empirical and workable to say. "Here be Humanities", they think to themselves as they pass by certain buildings on campus, and behind the windows they imagine a folk as strange as hobbits, babbling postmodernly.

  26. Anne said,

    September 15, 2011 @ 12:09 pm

    Does Herr Goldacre hate German, with all its strange verb placement and capitalization of common nouns, so much that he refuses to capitalize … well, anything?

  27. Rodger C said,

    September 15, 2011 @ 12:30 pm

    Maybe Claire really meant that Goldacre had been a-hitting the gin.

  28. Bobbie said,

    September 15, 2011 @ 1:20 pm

    I have problems with low blood sugar. When my blood sugar gets low, I tend to get surly and cranky. At those times, minor irritations become "super-clear" to me (but not to anyone else.) So I can understand Ben Goldacre's peevishness (even if his rant does not make a lot of sense.)
    Too many donuts for breakfast, Mr. Goldacre?

  29. Chandra said,

    September 15, 2011 @ 1:27 pm

    @Dougal Stanton – I was with you and Theodore at first, but the follow-up tweets, and then Ben Goldacre's reply to this very post, show that it was not in fact meant as a joke.

  30. Jerry Friedman said,

    September 15, 2011 @ 1:55 pm

    @Ben Hemmens: That's a good point. The closest I can come up with is "to find out if we may legitimately associate one thing with another", but even that's not the same, because the is associated formulation suggests that the association is an objective property of the things and has nothing to do with us.

    [(myl) Forms of "associate(d) with" can have eventive as well as stative interpretations. Independent of that semantic question, the use of "associated with" to mean "has a statistically significant co-occurrence with" has an active-voice counterpart, as in these example from Google Scholar:

    Our results indicate that a suite of polymorphisms associate with differences in flowering time …
    Event-related activation in the human amygdala associates with later memory for individual emotional experience
    Cardiac conduction defects associate with mutations in SCN5A
    Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse

    So even if this usage is stative rather than eventive — i.e. is more like the traditional adjectival passive "the door is closed" than like the eventive "the door was closed by an unseen hand" — it still occurs in corresponding active-voice/passive voice pairs.]

  31. Szwagier said,

    September 15, 2011 @ 2:01 pm

    Can we please stop calling him Mister Goldacre? He's a real doctor, you know.

    [(myl) Sorry — fixed now, at least the one instance in my post.]

  32. Rubrick said,

    September 15, 2011 @ 2:32 pm

    I, too, am with Neil Coffey on his "it just doesn't occur to people that language could be, and in fact is, studied empirically" hypothesis.

    And I'm curious: What work has been done, by linguists or others, on the question of why some passives seem objectionable, while others are so natural that even the misguided anti-passivists don't notice them?

    Take the first "hoist by his own petard" example sentence above:

    While the authorities are distracted by mass disorder, we can do some statistics.

    The choice of the passive for the first part of the sentence seems right and proper. But what about:

    While the authorities are distracted by mass disorder, some statistics can be done by us.

    I think most of use would agree this is pretty dreadful. It's this sort of thing which people who rail against passives think they're railing against. But why is it dreadful? Can we do any better than the "I know it when I see it" test? Is there science to be done here, or are vague explanations about "vigor" and "crispness" all that we've got?

    [(myl) There are several much larger questions here, for example about how the placement of noun phrases and pronouns affects textual coherence — see for example the work on "Centering Theory". In this case there's some parallelism between what "the authorities" and "we", which would be lost if "statistics" became the subject of the second clause. And there are also some more-or-less discourse-independent reasons to dis-prefer "statistics can be done by us". We can start by observing that verbal complexes of the form "can do NOUN" often resist passivization, perhaps because they're somewhat idiom-like — COCA has 75 instances of "can do business" vs. 3 of "business can be done". And by-phrases whose object is just a pronoun are somewhat rare, unless the pronoun is focused. Compare "by me" with 1465 hits to "by my" with 5137; "by us" with 1049 compared to "by our" with 3367. The pattern {[n*] can be done by [pp*]} gets no hits at all, though {[n*] can be done by} gets 37.]

  33. Aristotle Pagaltzis said,

    September 15, 2011 @ 2:45 pm

    FWIW, in proper German that would be “Wäre ich der Architekt der deutschen Sprache, legte ich das Verb an einen etwas weniger lästigen Platz”. Translated to English while keeping the exact same word order, that is “Were I the architect of the German language, put-would I the verb in a slightly less irritating place”, where “put-would” is my attempt to emulate Konjunktiv II for the verb “put”, a form of subjunctive that does not exist in English.

    Admittedly in less formal German few people use the proper K2 forms of verbs, and instead express the K2 by combining their regular forms with “würde” (equivalent to the added “would” above), yielding “würde ich das Verb an einen etwas weniger lästigen Platz legen” — “would I the verb in a slightly less irritating place put”. The construction of the first clause with the verb at the end, however, is just plain silly; no one speaks or writes German like that.

  34. Ben Hemmens said,

    September 15, 2011 @ 3:23 pm

    I once had a German teacher who introduced his course by saying what a good language German was, because it has fewer words and leaves the least important word, the verb, till the end ;-)

  35. Ben Hemmens said,

    September 15, 2011 @ 3:35 pm

    While the authorities are distracted by mass disorder, we can do some statistics.
    While the authorities are distracted by mass disorder, some statistics can be done by us.

    Well, that one's easy. It allows us to set up the sentence to emphasize the contrast between what the authorities are doing and what we are doing. The active-active version

    We can do some statistics while mass disorder is distracting the authorities.

    doesn't work for that purpose any better than the passive-passive one. Thanks to myl for the link on centering theory, which looks promising.

  36. J. W. Brewer said,

    September 15, 2011 @ 5:02 pm

    myl's instances (in his addition to Jerry Friedman's comment) of active-voice "associates with" in scientific writing where the subject is a phenomenon (e.g. "a suite of polymorphisms") rather than a scientist or other conscious being are interesting, not least because they are totally ungrammatical in my idiolect. Not merely inelegant and clunky or unusual, but actively wrong in a reflective-native-speaker-intuition kind of way. Does anyone else have this reaction? Maybe it's just a specialized usage, such as can occur in any profession's jargon (I'm sure my own profession's internal prose style has stock phrases that could strike an outsider as ungrammatical rather than merely awkward), but otoh maybe it's a weird by-product of anti-passive-voice paranoia?

    I think the problem may be that active-voice "associate" in this sense is in my idiolect sort of mandatorily ditransitive? "Professor X's research associates phenomenon Y with phenomenon Z" would be a clunky but grammatical active-voice construction, but it specifies an actor capable of the mental act of association. That is fully consistent with the less clunky "Y is associated with Z (according to Professor X's research)" but not "Y associates with Z," which makes it sound to my ear as if the phenomena are hanging out in the same questionable social circles. A probation or parole condition forbidding the offender to e.g. associate with known drug users or gang members seems to my ear to be using "associate with" in a significantly different sense than the statistical-correlation one.

    Note that the thesis of Sir Robert May of the Royal Academy (contained in one of the links upthread) was that the passive-voice "it was observed that" style of writing up research results was bad because it supposedly reflected "the antique delusion that work was more scientific if performed by the impersonal forces of history rather than by real people." This "phenomenon X associates with phenomenon Y" usage is interesting because it avoids an arguably passive construction while still concealing the human scientist behind the curtain, which was May's objection. It thus remains exactly as "vague about agency" as the "X is associated with Y" construction would be.

  37. Jerry Friedman said,

    September 15, 2011 @ 5:06 pm

    @myl: Isn't A associates with B and A is associated with B a very anomalous active-passive pair? Normally you'd expect A associates with B to form what the CGEL calls a prepositional passive: B is associated with by A. Is there any other active-passive pair like this one?

    [(myl) What I had in mind, and didn't explain clearly, was that the intransitive "X associates with Y" has a causative form "Z associates X with Y", which in turn has a passive "X is associated with Y [by Z]". This would be analogous to (say) intransitive "The chapter ends with a discussion of X", causative "The author ends the chapter with a discussion of X", passive "The chapter is ended with a discussion of X". Is this the right way to think about "X is associated with Y"? Well, there are a lot of verbs (associate with, connect to, relate to, etc.) that have similar patterns, and are commonly used in a passive-like form, less commonly in an intransitive form, and rarely in the causative form that might be the source of a passive.]

    I looked at Google Books. The earliest visible example of "associates with changes" in this sense is from 1989, and all the examples are scholarly. On the other hand, there are over 200 examples of "is associated with changes" from the nineteenth century, many of them medical. I strongly suspect this sense of "associate with" was mostly invented within the last few decades to satisfy prescriptions, and maybe even journals' rules, against the passive voice.

    [(myl) Yes, but the OED tells us that transitive forms of associate, both active and passive, go back to the 14th century, e.g.

    a1513 R. Fabyan New Cronycles Eng. & Fraunce (1516) I. cxxvii. f. lxiiiv, He‥associate vnto hym certeyn wanton persones.
    1561 T. Norton tr. J. Calvin Inst. Christian Relig. Table Quot., She was associated unto him in marriage.

    The transitive sense glossed as "To join, combine in action, unite (things together, or one thing with another)" is noted to be "Mostly refl. or pass." — that is, "Z associates X with Y" occurs mostly as "Y is associated with X".

    1578 J. Banister Hist. Man v. f. 70, The thyrd veyne of the ventricle is very small, not associated with any Arterie.
    1660 R. Boyle New Exper. Physico-mechanicall Digress. 352 The inspired Air‥does there associate it self with the Exhalations of the circulating Blood.
    1751 Johnson Rambler No. 158. ⁋7 Faults are endured without disgust when they are associated with transcendent merit.

    Whether these are "real" passives is also open to argument, I guess, since it's not clear whether there's any agent. But the existence of the reflexive forms makes the analysis as passives more plausible.

    (There's also a (rare) intransitive sense "To combine for a common purpose, to join or form an association".)

    The use of any form of this word to indicate an observed statistical connection couldn't possibly be more than a century old or so, just because the concept hasn't been around much longer than that.]

  38. Jerry Friedman said,

    September 15, 2011 @ 5:14 pm

    Well, maybe I don't suspect it that strongly. It could just be the kind of shortening that happens in professional jargon.

  39. Nathan said,

    September 15, 2011 @ 9:51 pm

    Behold the power of the focusing voice!

  40. Tom S. Fox said,

    September 15, 2011 @ 10:58 pm

    Aristotle Pagaltzis:

    FWIW, in proper German that would be “Wäre ich der Architekt der deutschen Sprache, legte ich das Verb an einen etwas weniger lästigen Platz”.

    Or, in even more proper German: “Wäre ich der Architekt der deutschen Sprache, setzte ich das Verb an eine nicht ganz so lästige Stelle.”

    The construction of the first clause with the verb at the end, however, is just plain silly; no one speaks or writes German like that.

    Not true. Translated literally, it would become the completely correct “Wenn ich der Architekt der deutschen Sprache wäre …”

  41. J. Goard said,

    September 16, 2011 @ 12:33 am

    @Ben Hemmens:

    Almost a great post there…but IMHO you muck it all up with a ridiculously inaccurate invocation of the notion "poverty of the stimulus", which you refer to in the title. Nobody of any theoretical significance would think for a second of equating relevant stimulus with explicit instruction: Skinner didn't, James didn't, psychologists studying learning in 1957 or 1977 or 2011 don't, applied psychologists (like in sports) don't, and usage-based linguists sure don't.

    The basic argument against linguistic nativism is that general learning processes plus the nature of the learning environment are sufficient to explain acquisition. Often, especially in the last decade, a third factor is often added — cultural evolution of language selected for by general features of (usually child) brains. But the "stimulus" which nativists after NC argue to be inadequate (or logically must, for their argument to go through) is the socially-situated learning environment in its totality. This is what made NC's proposal shocking and radical rather than trivial, but it's not the subject of your post — in fact, it's precisely what you yourself find to be effective.

    By invoking the concept the way you do, you're giving NC credit for a very different claim, one that's obvious to the point of silliness. True, language isn't learned by reading a treatise on it, but neither are ice skating, playing tennis, or playing a trumpet, skill sets for which almost all thinkers would never consider proposing an innate, domain-specific acquisition device. I'll bet that people who start early in life at these endeavors, putting in tens of thousands of hours on task with both intrinsic and extrinsic motivation and rich social feedback, can similarly survive a dozen bad advisors along the way.

  42. maidhc said,

    September 16, 2011 @ 4:02 am

    I'm a little confused by the references to scientists, because books on technical writing generally recommend the use of passive structures in scientific papers.

    This is not to make things harder or easier to understand, but because there is certain information that should not appear in a scientific paper because it is irrelevant and distracting.

    You should not say "My research assistant Sally heated the sample to a temperature of 300 degrees C". You should say "The sample was heated …"

    Active voice is fine too, if it is something like "The tests showed that the sample contained…"

    Based on the fact that scientists are specifically taught to write this way, I would expect the use of passive constructions to be higher in scientific papers than probably any other field of literature.

    I note that in the TSN reference it says that most scientists believe that the passive structure should be used by scientists (as opposed to children). I don't know about Nature and Science, but most other professional journals appear to prefer the absence of information not directly connected to the topic of the paper.

    One can imagine another world, without the divide between science and the arts that so plagues our own, where a scientific paper might contain something like

    It was a warm spring day in Cambridge, and a soft breeze through the open window of our lab ruffled the thick red hair of Sally, my gorgeous research assistant (too bad she has a boyfriend!), as she heated the sample to 300 degrees C. It was only while discussing the test results later over delicious shu mai and a bottle of Chilean sauvignon blanc at a delightful little bistro just a block from campus, that we realized that the ore sample had a significant amount of contamination…

    but the page charges would add up pretty fast.

  43. ‘A sad case’ – ‘Language Log’ Blog - English Teaching Daily said,

    September 16, 2011 @ 5:10 am

    […] A sad case […]

  44. Mark F. said,

    September 16, 2011 @ 8:12 am

    maidc — I have read papers where the authors scrupulously avoided the first person, and relied heavily on the passive voice in the methods section. I think this is the kind of writing that scientists who complain about the passive have in mind. It is OK to say "We measured the density after 5, 10, and 15 seconds," and depending on the flow of the paper, that may or may not be better than "The density was measured after 5, 10, and 15 seconds." "My research assistant sally measured…" is just a straw man, although it is amusing.

  45. Sarra said,

    September 16, 2011 @ 8:25 am

    Oh, dear.

    Goldacre doesn't fit into your bugbear-framework. He really doesn't. This is such a misguided post.

    He's a scientist, and British. His Britishness is important, because we don't have this pervasive pedant's pickle with the passive over here – nobody really gives a toss. (Which is why, incidentally, I really don't think the idea that the comment is meant as a peeve back at the peevers.)

    His profession is important because scientific publications are PAINFULLY passive. This isn't about 'proper English'; scientists don't believe that the passive voice should be used religiously anywhere else, just in reports, papers etc. We're taught this aged 11 when we do our first experiment write-ups. It's a practice that is intentionally specific to scientific writing.

    It's just a bugger to read. Hence Goldacre's comment. There's something tiring about reading an entire report where passive after passive after passive appears specifically to cut out any 'I's or 'we's. I'm not talking about any other type of passive construction, by the way (viz. very interesting comments on 'is associated with', for example, above – I proofread public health papers and I don't feel that sort is a problem), just these ones that erase the subject in a way that, after a while, comes to be so tedious you just want to shout at the paper just say you did it!

  46. Sarra said,

    September 16, 2011 @ 8:29 am

    NB low blood sugar here too. Sorry!

  47. anthony alcock said,

    September 16, 2011 @ 9:45 am

    I think Nick Williams might disagree with Geoff Pullum about how dead Cornish is.
    Personal message to GP:

    The opening riff of Money for Nothing: you should try playing it on an acoustic guitar, mate.

  48. jan said,

    September 16, 2011 @ 9:57 am

    Here's a story about modern science and scientific research, as developed by the Native Americans.


    It includes:

    How the Lizard Giants Became the Birds

    a vision received by

    Too Curious Otter

    The same author has a lot of other stories and essays.

  49. LDavidH said,

    September 16, 2011 @ 1:17 pm

    Best ever use of the passive voice: "A good time was had by all". It really mystified me the first time I heard it – not because I didn't understand it (I did), but because it seemed such an awkward way of saying "Everybody had a good time".

  50. Ben Hemmens said,

    September 16, 2011 @ 1:59 pm

    >We're taught this aged 11 when we do our first experiment write-ups.

    This is true, and it's perfectly all right to want to correct that tendency. But scientists giving instruction to other scientists on how to write tend to make up what they teach all by themselves without noticing that there are people who actually work on this stuff and can say something about
    a) what is or is not correct English (empirically- now that ought to appeal to scientists),
    b) how to make a text clear, coherent, concise and readable.
    All the advice on language I can remember getting during my scientific career (every rule anyone ever told me!) belongs to the set of things the LL people have shown to be completely wrong. Surely we could do better than that.
    And secondly, most of the science in the world is being written in English as a second or foreign language, by authors for whom the content of what they are explicitly taught should be even more important than for native speakers. But as far as I can see, they are being told the same guff, and worse.

  51. Ben Hemmens said,

    September 16, 2011 @ 2:46 pm

    @J Goard,

    Thanks for your concern, but I think you're taking me a bit more seriously than I aimed to be taken.

    As for the PoS, I freely admit to not having a crystal-clear idea of what it's about. But in my defence, I will say that in the only paper I have read on the subject, if I remember correctly, the authors (Scholz and Pullum) spent about half of the space trying to work out a definition of PoS, based on the writings of people who use the term. And if *they're* not sure, I'm sure as hell not sure.

    OK, in return everyone gets a free pass to make jokes (at my expense) using the names of theories from biochemistry, which is the subject I have some qualifications in.

    Slightly more seriously, I'm not sure that successful learning in the presence of wrong instruction is always because the learners compensated for it by picking up correct content somewhere else. I suspect that learners (in language, football, singing or whatever) may produce "right" responses to inputs which are objectively wrong (or demonstrably not properly understood in terms of the bare content).

    And as for "skill sets for which almost all thinkers would never consider proposing an innate, domain-specific acquisition device", I don't see why we should be more inhibited about this for other activities than for language. Certainly a lot of other skills – at least in certain individuals – seem to be driven by a powerful innate motor as much as, or more, than they are picked up in "thousands of hours on task". A certain young boy by the name of Johannes Chrysostomus Wolfgangus Theophilus might have had a pushy father, but it wouldn't have come to much if he hadn't been the particular genius he was.

  52. Jerry Friedman said,

    September 16, 2011 @ 2:54 pm

    Nature says, "Nature journals prefer authors to write in the active voice ('we performed the experiment…') as experience has shown that readers find concepts and results to be conveyed more clearly if written directly."

    (Nice "to be conveyed" and "written".)

    @MYL: Thanks for the detailed explanation. I do indeed have trouble seeing these constructions without agents as passives. However, you've given me more of an idea of how it's possible to see them that way.

    The use for statistically observably connections is indisputably as recent as you say, but it strikes me as an extension of the use for "frequently occurs with", which is quite a bit older: "It must, however, be granted, that generally an inflammation of the conjunctiva, in a greater or less degree, is associated with it [inflammation of the iris]", John Cunningham Saunders, A Treatise on Some Practical Points Relating to the Diseases of the Eye, 1811.

    (A lot of instances of "associated with" at that time are from geology—one mineral is associated with others—but in that case I wonder whether the agent is God.

  53. Mark Mandel said,

    September 16, 2011 @ 2:57 pm

    Would this post have been better titled "A sad voice"?

  54. Mark F. said,

    September 16, 2011 @ 4:40 pm

    I am still influenced by the "avoid passive" meme, but I never interpreted "avoid" to mean "scrupulously avoid". I took it to mean "Sometimes people overuse the passive voice in formal writing, even though sentences phrased in the active voice tend to be easier to understand. Watch out for that." It's debatable whether the part about active sentences being clearer is actually true, but that's how I took the advice at the time. (In methods sections of scientific papers, I still buy the clarity claim.)

    It's pretty clear that the strong form of the advice is just silly. But to empirically test the weak form of the advice, it's not enough to show that good writing uses the passive voice moderately frequently. In fact, good writing is completely the wrong thing to study. What you need to know is whether more bad writing would be improved by using the passive voice less than would be improved by using it more. And I don't know the answer to that.

    If it's really true that the people who rail against the passive the most are those who use it the most, perhaps that's because it's more salient to them since they're always having their temptation to use it even more.

  55. ShadowFox said,

    September 16, 2011 @ 7:49 pm

    Why not challenge every blogger/twitter/teacher who complains about passives to go 48 hours without writing one case of passive? (and give some minimum number of words/sentences that must be included in that bet to make it completely fair–so that they cannot claim lack of access to internet or some other silly excuse to turn that into a sucker bet)

  56. Ken Brown said,

    September 17, 2011 @ 1:29 pm

    What Sarra said. I studied science in school and universities in England and the only advice about writing style that I remember ever being given at all was to put lab reports into an anonymous passive. There was no explict prescriptive instruction about writing style in any other context. It wasn't part of our academic culture.

  57. Dan said,

    September 17, 2011 @ 3:32 pm

    Actually, the opposite effect is found by me. Whenever active sentences are read by me, they sort of have to be converted by me into passive ones. It is guessed by me that I'm ridiculous.

  58. Janice Byer said,

    September 17, 2011 @ 8:15 pm

    Dan, the ridiculousness surmised by you is contradicted by your reasonableness being well established here on LL, and further by these five passive clauses written to illustrate the threshold for ridiculousness has been raised.

  59. Mike Maxwell said,

    September 17, 2011 @ 8:22 pm

    Slightly related to the topic of the passive voice being avoided, or scribing phrases while avoiding any 't' or 'e', I've always wanted to talk about Douglas Hofsteder's short story on Albert One Stone's work without any Greek or Latinate words:

  60. Sili said,

    September 18, 2011 @ 11:33 am

    I have to withdraw some of my criticism of the Skeptics Guide to the Universe.

    They came out very strongly in favour of descriptivism yesterday.

    Sadly, they still didn't bother doing any research. It's rather surprising that nerds do not not know the etymology of "irregar'less".

  61. Writing Progress: Week Ending September 17, 2011 « The Undiscovered Author said,

    September 18, 2011 @ 8:13 pm

    […] of how passive voice weakens the story.  But then I read some interesting little squib like this, and my brain goes tic-toc.  (Shame on you, Strunk & White.)  Or you can find style advice […]

  62. Tracy said,

    September 19, 2011 @ 7:19 am

    Language Log's fulminations on passive vs active have confused me so much that I would never dare to use them as descriptors any more. I follow Dr Goldacre on Twitter, saw that tweet and thought "Uh, oh, Language Log will tar and feather him for that!" and here we are. I adore Language Log and highly respect Dr Goldacre, long may they both live.

    (and yes, Skeptics Guide to the Universe can get utterly overbearing on topics they know little about – my pet peeve is their knowledge of education.)

  63. Jerry Friedman said,

    September 19, 2011 @ 12:37 pm

    @Mike Maxwell: No doubt you know the essay that inspired Hofstadter, Poul Anderson's "Uncleftish Beholding". There must have been things like this in the 19th century, but not about atomic theory.

    And I'm still interested in what avoiding the passive voice would look like and how hard it would be.

RSS feed for comments on this post