Twenty selected Coalface errors

« previous post | next post »

Those who have read about the great Queensland grammar scandal about the "Coalface" teachers' guide and the ensuing coverup and counterattack may have wondered just what the crucial errors of grammatical analysis were, because the press coverage mentioned only a scant half-dozen. I thought Language Log readers might like to see fuller details in browsable form (Huddleston's full presentation in PDF format is available here). Below I give a terse listing of twenty sample errors in Lenore Ferguson's first two articles in the "Grammar at the Coalface" series (the listing is not exhaustive).

How important these are depends on how seriously you take grammar as a subject. It's true that linguistics is not like earthquake engineering — if someone misclassifies a word or botches a definition, nobody dies; but on the other hand getting factual claims right is part of the job description for scholars and teachers. From the point of view of the public controversy, however, the relevant question is just how many of these blunders could conceivably be dismissed in the way Lenore Ferguson and Gary Collins have tried to dismiss them: as (1) minor errors of typing or formatting, or (2) mere "matters of opinion", or (3) simple terminological differences, or (4) substantive differences between one theory and another. There is not a single one. Which means the blustering ETAQ responses are entirely dishonest. Where anything to do with "functional grammar" is relevant at all, Ferguson has generally either contradicted its tenets or contradicted herself. Here is the select list of 20 particularly clear errors:

  1. Won't in The small boy won't eat his lunch called an adverb. (It's an auxiliary verb.)

  2. Capable of in The small boy is capable of eating his lunch called an adverb. (Not a grammatical unit; it's an adjective followed by a preposition.)

  3. A pair called an adjective. (Not a grammatical unit; it's a determiner followed by a noun.)

  4. Set of in a set of bowls called an adjective. (Not a grammatical unit; it's a noun followed by a preposition.)

  5. More and most (in more/most swollen) called adjectives. (They're adverbs.)

  6. Your in your folder called a determiner and a pronoun. (Those are presented as distinct categories; under Ferguson's assumptions no instance of a word can be both.)

  7. Sam's in Sam's folder called a possessive pronoun. (It's the genitive form of a proper noun.)

  8. What in They saw what lay before them and who in the explorer who saw the carnage called conjunctions. (They're pronouns.)

  9. Something called a pronoun but everything called a noun, though pronoun and noun have been presented as distinct categories.

  10. Have a peep and wants to help called verbs, though verb is a word class. (They are syntactic constructions, the first being a verb followed by an object noun phrase and the second being a verb followed by an infinitival clause.)

  11. "Subject" is glossed as "doer of the action". (Clauses like I don't know don't express actions, so not every subject is a doer; and Ferguson says elsewhere that in passive clauses like The race was won by the boy the doer is "relegated" to the by-phrase, so not every doer is a subject.)

  12. Non-finite clauses are claimed not to permit a subject. (Cannot be right, because of the subjects (underlined here) in clauses like the bracketed part of [For you to give up now] would be a great pity, or in We objected to [her children being given special privileges].)

  13. Complement function presented as contrasting with object, as in traditional grammar, but then defined in the glossary as including object, as in Functional Grammar.

  14. Glossary defines complement as an element with "the potential to be the subject of the verb" (the lunch in a clause like Sam gave the lunch to Jodi can be made subject by changing to passive voice yielding The lunch was given to Jodi by Sam). (Cannot be right because of complements like ill in Sam was ill: notice that *Ill is been by Sam is ungrammatical.)

  15. Participants are said to be realised by noun groups, but an example on the same page has a participant with the form of a preposition phrase, so not all participants are noun groups. Later participants are actually said to be the same thing as noun groups, but in fact noun groups often do not function as participants in the functional grammar sense (e.g. the underlined one in He watched TV all morning), so not all noun groups are participants.

  16. "Clause complex" is said to be the highest unit in the "rank scale", but five lines later the "clause" is claimed to be the highest unit. (These are incompatible claims.)

  17. The account of adverbs says they don't inflect. (Not true; note soon, sooner, soonest, and other such cases involving adverbs that don't end in -ly.)

  18. Initially "finite verbs" are contrasted with auxiliary verbs and lexical verbs as in functional grammar, but the glossary definition is based on traditional grammar, with finite verbs defined in terms of inflection for tense, etc. (On the glossary's account, went is a finite verb, but on the functional grammar account it is not, because it is a lexical verb.)

  19. Declarative clauses are said to be defined by the order subject + finite, but in many declarative clauses the subject does not precede the finite verb (e.g. Never had this been seen before).

  20. Imperative clauses are said to be defined by the absence of subject + finite. (Wrong because imperatives can have subjects: Nobody move!, You be quiet, Don't you speak to me like that, etc.).

None of these mistakes could plausibly be given a defense based on claiming that they are not errors, or that they reflect a theoretical difference, or that it is a matter of opinion. There simply aren't any grammarians who believe the false claims about grammatical analysis in the above list. Whether they favor functional grammar or traditional grammar or more recent theories of grammar makes not one whit of difference. In particular, despite what Ferguson and Collins appear to have been suggesting in some of their comments to the press or on the radio, not a single one of the errors above amounts to a disagreement about an analysis unique to the Cambridge Grammar: if that work had never been published, all of the above would still be transparently obvious errors. And it is quite clear that none of them could be plausibly portrayed as typos or formatting slips. It is hard to believe Ferguson and Collins could seriously suggest it. But that is what they have done.


  1. Mark Liberman said,

    June 22, 2008 @ 5:50 am

    I think it's important to recognize the damage that is done by trying to teach children incoherent and inconsistent material. When homework and test answers are arbitrary and impossible to predict on any rational grounds, kids lose respect for teachers and for school in general, as well as for the subject being taught.

    I have a very distinct memory of having this problem with diagramming sentences, at the age of nine or so. I was frustrated by being completely unable to determine, from the instruction we were given, what the "right" answer was supposed to be. We were never given any useful arguments about why a diagram should be done one way rather than another — the teacher acted as if this was obvious to the initiated, whose ranks we would join if we only listened and did the exercises. At one point I gave an answer that seemed to me to be the one specified by the examples we'd been given to study, but still, I was told it was wrong. I complained and showed the teacher the study sheet; she gave me a completely unconvincing explanation about why the examples were different; and I had the sudden insight that the teacher didn't know what she was talking about, and was making stuff up to hide her ignorance and the incoherence of the material she had presented.

    I had previously felt that school was sometimes boring, but never that it was nonsense.

    I don't know whether the fault in this case was the Reed-Kellogg system of sentence diagramming, or my teacher's understanding and presentation of it — or perhapsmy own inadequately subtle reasoning powers. But I do know that it's a bad thing for everyone if teachers present students with a purported well-defined analytic framework that makes no sense and is inconsistently applied.

  2. Lester Piggot said,

    June 22, 2008 @ 6:17 am

    Mark Liberman wrote, it's a bad thing for everyone if teachers present students with a purported well-defined analytic framework that makes no sense

    Yes and no. You learn to not take the teacher's word for everything, which isn't a bad lesson.. When I was six, I had a similar experience to yours. Our teacher wrote a heading on the blackboard: Similar Means The Same. We were supposed to make a list of words with the same meaning, something like that. Or was it similar meanings? I was confused, similar doesn't mean 'the same'. It means, in fact, not quite the same. That was forty-nine years ago, and I remember it as if it were yesterday. After some time, I figured out where the problem was. The lessons you learn at school aren't all in the curriculum.

  3. Andy J said,

    June 22, 2008 @ 7:53 am

    This whole ETAQ debate (if that is the appropirate descrpition of what is going on here) only serves to underline my disquiet about those who describe linguistics as a science. Certainly lingusitics uses the same analytical approach as, say, the physical sciences, and an aim of linguists may be to find a set of universal definitions, similar to, say, the laws of physics. But following a particular process towards an intended outcome does not make it a science, otherwise we could say that cooking was a science. To my mind there are too many contested and competing methodologies (cf theories and hypotheses in the 'traditional' sciences) in linguistics for the true implementation of the kind of peer review which characterises the traditional scientific process of validation.
    Perhaps we should take a lead from the International Astronomical Union on the subject of classifying Pluto, and call lingusitics a dwarf science.

  4. Mihai Pomarlan said,

    June 22, 2008 @ 8:56 am

    IRT. Andy J.:

    I am no linguist, but that claim still looks off.

    ALL sciences had a period of groping around for methodologies, hence the "competing" and "contested" parts were a fact of life for all of them.

    It still IS, really. Sure, the basic outline of the scientific method is the same but we are discovering new things that need new ways of describing: see all the math jazz in theoretical physics, some of which is not just mental masturbation and is actually applied in high-end electronics, or spintronics; neo-Darwinism is filled with debates about what works and how in the living world like punctuated equilibrium vs (more or less) smooth and steady etc.

    Science is one big debate, not some collection of facts set in stone. We have a few of those, but not that many.

    And the cooking analogy is wrong because a) cooking does not attempt to produce research papers (but food) and b) the owner of the famous restaurant the Fat Duck has a show on Discovery called The Science of Cooking. So there.

  5. Tim Silverman said,

    June 22, 2008 @ 10:00 am

    I don't see that the ETAQ kerfuffle has anything to do with the status of linguistics as a science. People talk nonsense about physics, too. Does that undermine the status of physics as a science? People forget which side of the road the post office is on. Does that mean we as a species possess no reliable knowledge about the concept of spatial location?

    The debates within linguistics about the best classification of words and so on are generally quite subtle, affecting small groups of words in particular languages. It's not like like people are proposing to abandon the concept of a preposition.

    As far as teachers presenting false or incoherent material is concerned—I don't think I ever, at any age, believed teachers were infallible; but it still lessened my respect for them when they made obvious errors, and still more when they tried to cover them up with bluster (which happened, I hasten to add, very rarely). It's fine to learn that things aren't perfect, but that's not a reason not to try to improve them!

  6. Andy J said,

    June 22, 2008 @ 10:42 am

    Can I just preface these remarks by saying that a. I am no linguist (or scientist) and b. my earlier post had a certain amount of Sunday afternoon ennui and light-heartedness behind it! Perhaps I should have used a few emoticons.
    @Mihai Pomarian. I fully agree that many sciences and the development of scientific theories have debate at their heart. The classic that comes to my mind at present is that of 'Steady State' vs 'Big Bang' in astrophysics. But that was, as if often the case, about how certain observable 'facts' are interpreted. My point with linguistics is that while there are many areas in which there is a large amount of agreement, when we have relatively qualified people arguing about whether a particular word is a verb or abverb, or whether another word is a noun or pronoun, it seems to be that we lack the widely accepted fundamentals (the 'facts') on which to base the classification of linguistics as a fully grown-up science. I acknowledge that the particular ETAQ spat is rather like someone from the Flat Earth Society engaging in debate with a geophysicist. And perhaps my closing analogy with Pluto was inappropriate because however hard it works at it, Pluto can't become a planet one day (unless the IAU changes the rules again LOL) whereas linguistics surely can overcome my current objections. And finally, on the subject of cooking, you mention the work of Heston Blumenthal. Surely his angle is that he harnesses science in a more specific way in his cooking, in contrast to most chefs who would style themselves artists, cooking by intuition rather than by laboratory methods.

  7. Andy J said,

    June 22, 2008 @ 10:55 am


    By my browser's rendering of my previous posting, in line 8, it should read "as is often…" and in line 12, "it seems to me that…"

  8. Bonechar said,

    June 22, 2008 @ 11:02 am

    Some of those errors are astounding to me. The misclassification of something like "capable of" as an adverb seems like it cannot even be a failure of understanding – it must certainly be the result of not having thought about it at all, as I cannot imagine any set of definitions under which that would make sense. They must have just stuck "grammar" words on a dart board.

    Saying the ETAQ "debate" shows linguistics is not a science is like saying the Intelligent Design "debate" shows biology isn't a science.

    It speaks to the point of the Mark Liberman's comment, though. There is, at least in pretty much every (pre-University) school I've had the opportunity to observe, a worrying tendency to try and teach the kids what to think rather than how – it's much easier, of course, to standardize and test, and it means you can teach material you don't understand yourself since it need only be repeated verbatim. The greatest danger in this is not that children will discover all adults are liars and fools, which is inevitable, but rather that they come to assume this reflects how things actually are; that is, that grammar, mathematics, etc, are totally arbitrary and that the soundest argument is the argument from authority. This is precisely why highly educated people can still believe in Creationism – they have no capacity to evaluate evidence because they've never been given evidence for anything.

    It's also how people can think one group of people's failure to understand what they're talking about throws the whole area into doubt. If the argument from authority is the only one you know then when authorities disagree you have no recourse but to throw up your hands, and that is the damage done by teaching inconsistent, incoherent material.

  9. Leicester Pig said,

    June 22, 2008 @ 11:15 am

    @ andy J:

    Perhaps this was part of your Sunday afternoon point, but you know that telling linguists that linguistics isn't really a science just irritates the hell out of them. They (some of them, at least) find it useful to see things scientifically, and as far as I have seen there is nothing to be gained telling them otherwise. And if you look for example at the bits of Liberman's work that he shows on Language Log it pretty clearly is science, he just happens to be equally knowledgeable about the arts and humanities side of language.

  10. Eli Morris-Heft said,

    June 22, 2008 @ 11:36 am

    @Andy J, who said, "My point with linguistics is that while there are many areas in which there is a large amount of agreement, when we have relatively qualified people arguing about whether a particular word is a verb or abverb, or whether another word is a noun or pronoun, it seems to be that we lack the widely accepted fundamentals (the 'facts') on which to base the classification of linguistics as a fully grown-up science."

    I doubt you'll find many linguists debating whether a particular word is a verb or an adverb, or whether a word is a noun or a pronoun: those sorts of things tend to be relatively well-resolved within the space of linguistics. What you will find is debate over how exactly verbs do their job, or what the differences between a noun and pronoun are exactly, a distinction that can be rather subtle if the evidence is examined.

    If your model is physics, then you will always be disappointed by linguistics, for the simple reason that physics would be the same if there weren't people to examine it: the universal constants would all be set at the exact same precise values, light would still exhibit duality, the standard model would still apply. However, linguistics exists because humans exist – our facts are taken from observation of those humans. Also, language changes: something that was true a few hundred years ago may no longer be true now (especially in the what-category-is-this-word department). It is indeed difficult to make some of these empirical calls when different speakers of (supposedly) the same language give different answers to the same questions. How would you feel if any two randomly-selected electrons were found to have different masses? Or if the mass of an electron depended on its placement in spacetime, but not in a predictable way?

    Despite that, we do have universal truths and facts to base our theories off of. Observation of the evidence has yielded linguistic universals, and guided us to create theories that make predictions that we can test and prove. This sounds mighty familiar to me.

    Let us also not forget that modern astrophysics has been around for (depending on your view) about three to eight times as long as most parts of modern linguistics. Linguistics is a young science, but it is a science: a discipline devoted to discovery of facts about a phenomenon by way of empirical observation, logical reasoning and theorizing from those observations, and testing of those theories by more observation.

  11. Mark Liberman said,

    June 22, 2008 @ 1:57 pm

    I don't think it's interesting or useful to argue about who is scientific and who isn't. Let's leave that argument to the physicists, who seem to be having a fine old time of it.

    Often, I would argue, anxiety about whether or not a field is adequately "scientific" leads to bad results, such as pointless over-formalization, proliferation of jargon, excessive prestige given to complex physical and mathematical apparatus, etc. (…as opposed to appropriate and useful levels of all of those things, of course…)

    It seems more useful to me to worry about whether we're asking the right questions, and whether we're using the rational evaluation of relevant evidence to make progress in answering them, or at least in asking them more clearly.

    In the end, these standards aren't really any different in the humanities than in the social or natural sciences, in my opinion. Or at least they shouldn't be.

    I do recognize that this has been a minority viewpoint for the past century or so.

    But Charles Darwin, whose 200th birthday we'll be celebrating in February, is an excellent example of someone who asked the right questions, and used the rational evaluation of relevant evidence to make progress in answering them — using methods essentially indistinguishable from those used by many humanistic scholars of his day.

  12. Jangari said,

    June 22, 2008 @ 6:50 pm

    Re: Bonchar, while it seems the errors, especially capable of to which you refer, are completely inexplicable, I showed (over here) that they are the predictable result of the theoretical framework that ETAQ used, Systemic Functional Linguistics.

    The main idea is, SFL, like just about all modern theories on language broadly, identifies a form-function division, and, unlike other models, focuses on the function, that is, the interpersonal and ideational interaction in larger bodies of text. ETAQ took this framework and simplified it too far, such that form versus function is no longer recognised, and such that function takes precedence. The result? Formal categories (noun, verb) are instead labelled with the functions (participants, processes, etc.) that often correlate with them.

    The <capable of example is completely foreseeable; in the sentence he is capable of eating his lunch, in a purely function-based analysis, you might say that it functions as a modifier of the most functionally prominent verb in the sentence, 'eat'. Since you don't have any formal categories with which to label parts of speech independently of their functions, it gets called an adverb.

    I would also add that SFL is not at all to blame for this; it is a framework of textual criticism and was never designed to be used as a structural analytical tool. Furthermore, ETAQ have corrupted it so heavily that it can only correctly make predictions as often as a broken watch tells the right time, i.e., by coincidence.

    Its use here by ETAQ was completely inappropriate to begin with, but this corrupted version has no educational benefit whatsoever, and that is the indefensible part. ETAQ should find someone who is even cursorily qualified in these areas to inform their English teachers how to teach an accurate model of the structure of language.

  13. Andy J said,

    June 22, 2008 @ 7:11 pm

    Thanks to everyone for the good-humored responses to my inflamatory comments! As Leicester Pig surmised, I did rather expect to ruffle one or two feathers. And apologies to Geoff Pullum for going off-topic. I guess I must have been so amazed that the ETAQ supporters should have held their corner so long in the face of such a thorough analysis of the errors in their original work and the subsequent defense of their position, that my previously private skepticism about other matters spilled out. I rather think the ETAQ saga will continue awhile.

  14. Mabon said,

    June 22, 2008 @ 10:23 pm

    I very much appreciate Bonechar's cogent commentary on the nature of authority-based assessment of fact. The example of educated people clinging to a belief in Creationism is very much on-point. Because some people haven't developed a mechanism for evaluating evidence, they fall back on believing, as Bonechar puts it, that scientific arguments "are totally arbitrary and that the soundest argument is the argument from authority".
    Well stated!

  15. möngke said,

    June 23, 2008 @ 4:06 am

    I must say that I'm frankly astonished by the errors that have been pointed out in the teachers' guide. In any case, I really don't believe that a majority of schoolteachers would take these recommendations seriously, at least if education of English teachers in Australia is on at least a broadly respectable level. I might be wrong, but I seriously doubt that deference to authority would lead teachers to actively forget the knowledge they have been taught and wholeheartedly embrace the 'New Recommendations'.

    Of course, it might be that I'm skewed by my own primary- and secondary-school experience of teachers of Slovene (my native language) – everybody knew what they were talking about, for the most part. That said, they were mostly baffled by 'more complex' linguistic concepts (such as IPA notation, LEXICAL tone, or the subtleties of numeral-qualified noun phrases). Needless to say that I was unable to explain the phonetic differences between 'h-sounds' in Arabic to my secondary school Slovene teacher.

    The debate about word classes and usefulness of sentence diagramming (actually an enormous part of the Slovene language curriculum) reminds me of a traditional 'problem' in Slovene grammar, whereby elements of verb phrases that cannot be put into any of the existent word classes according to their strict definitions are given a word class all of their own, despite the absence of any overt similarity – and the fact that, if an analysis that would actually take spoken language use into account was attempted, there would be no categorisation problems.

  16. Ellen K. said,

    June 23, 2008 @ 10:41 am

    I wonder if I'm the only one who sees a certain logic to these errors. Jangari calls the "capable of" error inexplicable. Me, I don't find it inexplicable. Inexcusable, yes, inexplicable, no. Eating = action word. Thus, it might (incorrectly) be called a verb. The idea "capable of" modifies the idea of eating. Thus incorrectly called an adverb. Still an inexcusable error. But an explicable error.

  17. MikeA said,

    June 23, 2008 @ 11:24 am

    möngke seriously doubts that teachers would "check their minds at the door", so to speak, and teach officially approved nonsense. I have no knowledge of Australian or Slovenian education, but in the U.S.A, students are graded by "standardized tests", schools are funded (or punished) on a basis of their students' test results, and teachers who dare to stray from "teaching to the test" are soon looking for other employment. This would be marginally acceptable if the standardized tests were not a highly politicized process, but that would be some other U.S.A.

  18. Bonechar said,

    June 23, 2008 @ 1:12 pm

    Ellen K.:

    I should just mention that I was the one who called the "capable of" error inexplicable – Jangari was the one who explained it.

  19. Brad Encore said,

    June 23, 2008 @ 4:33 pm


    "None of these mistakes could plausibly be given a defense based on claiming that they are not errors, or that they reflect a theoretical difference, or that it is a matter of opinion."

    Shouldn't it be:

    "None of these mistakes could plausibly be given a defense based on claiming that *it is* not *an* error*, or that *it* reflects* a theoretical difference, or that it is a matter of opinion."

    "None" is tricky.

  20. Ellen said,

    June 23, 2008 @ 7:08 pm

    Ah, thanks, Bonechar. I couldn't (can't) follow what Jangari wrote, so I didn't realize he/she was referring to what you said there.

  21. Jangari said,

    June 23, 2008 @ 7:11 pm

    Ah, I was just about to reply to that. Thanks Bonechar.

    I believe I said "while they may seem inexplicable, they're in fact predictable errors…"

  22. Bryn LaFollette said,

    June 23, 2008 @ 9:33 pm

    @Brad Encore: I am assuming that your comment was made in the service of Irony, for to make such corrections of Mr. Pullum's prose is a dangerous path to tread! I recommend that anyone who takes such things seriously should consider this post on the subject.

  23. Michael said,

    June 23, 2008 @ 11:43 pm

    *Ill is been by Sam is clearly ungrammatical because a better transformation would have to be Ill has been by Sam.


  24. Yoda said,

    June 24, 2008 @ 4:39 am

    Like, I do, "Ill was Sam" Certain ring, has it.

  25. Mary Laughren said,

    June 25, 2008 @ 2:53 am

    Education Queensland is a grand bastion of ignorance as far as understanding the nature of human language or particular languages is concerned. It displays no interest in acquiring mainstream knowledge about English grammar in particular. It refuses to recognize a Major in linguistics within a Bachelor degree as part of approved education pathway for teachering in Queensland Schools. It refuses to invite recognized grammarians such as Prof. Huddleston or the members of the Linguistics Programs at The University of Queensland or Griffith University (also in Queensland) to run professional skills upgrading for English language teachers; it insists on running in-house courses where material of the sort critiqued by Profs. Huddleston and Pullum are presented. Education Queensland simply refuses to engage with either traditional grammar or with current mainstream grammatical analyses of English (or of any other language). They have concocted their own weird blend!

    Many recent graduates of Queensland High Schools arrive at University with no idea about the relationship between form and function or form and meaning in language. They have no idea of word class or that sentences are not just random strings of words; some students have come into my room in floods of tears because they can't distinguish a verb from a noun and have never heard terms like preposition or adjective before, while they see that some fellow students seem to know this stuff (usually middle-aged students, or lucky ones from schools where teachers wisely ignore the edicts of Education Queensland and teach traditional word classes and basic grammar). For many high achieving students, this is the first time they have been shown up as not knowing something. They are usually quite angry about being let down by Education Queensland, and rightly so. Of course our challenge in teaching linguistics is to introduce our students to meaningful ways of analysing language which comply with the exigencies of the scientific method, and which will be useful to them in their own writing, in learning other languages and so forth. Unfortunately few of our students go into teaching, so Fortress Education Queensland is unlikely to be whiteanted from within, just yet.

  26. Ellen said,

    June 25, 2008 @ 10:20 am

    P.S. Jangari, after your second post I was able to go back and read your first post and understand it. Interesting to know there is a system that looks at language that way.

  27. woliver said,

    June 25, 2008 @ 2:45 pm

    Do you suppose that even an Annie Hall moment, with Michael Halliday taking over the cameo role of Marshall McLuhan, would make ETAQ back off?

    WOODY ALLEN: Tell him.

    MARSHALL McLUHAN: — I heard, I heard what you were saying. You, you know nothing of my work. How you ever got to teach a course in anything is totally amazing.

    WOODY ALLEN: Boy, if life were only like this.

  28. Zvi Agam said,

    April 18, 2009 @ 9:52 am

    You say somewhere, 'I just received an email from…'. Do you mean by that that you only received an email from…'? Or did you mean to say, 'I've just received an email from…'?

RSS feed for comments on this post