Friday, March 27, 2009

The self-righteous fury of the blogs

If you don't read Carl Zimmer's blog, The Loom, you should. He has a magnificent post about the cl/rash between blogs and the mainstream media (Zimmer also writes for Discover, The New York Times, and other more traditional news sources) in which he writes,
"Take, for example, a recent righteous rant from Ben Goldacre, a British doctor who writes the blog Bad Science. Goldacre is doing fabulous work on his blog, taking on irresponsible sensationalism and misinformation about vaccines, quack medicine, and other offenses. His guns are blazing particularly hot in a recent post about some awful reporting on recent studies on prostate cancer. He concludes with this sweeping condemnation of our trade:

Journalists insist that we need professionals to mediate and explain science. From today’s story, their self belief seems truly laughable.

Goldacre is big on evidence and down on anecdote, but as Zimmer notes here, he's guilty when it comes to condemning the MSM based on cherry-picked stories.

When I am critical, as I often am, I do worry about this, and I hope that I have taken a fair line with who I blame. If I haven't I trust you to let me know.

Monday, March 23, 2009

Coordinating and stacking modifiers

One of the teachers here came by Friday with a question about commas in pre-head modifiers in NPs. In other words, she was asking about the difference between things like:
  • a serious political problem
  • a serious, unsolvable problem
where these are both noun phrases (NPs) headed by the noun problem and modified by two words coming before the noun. Notice that one has a comma and one doesn't.

The textbook that had led her into this said that there are two types of adjectives: cumulative and coordinate. With coordinate adjectives, you put a comma, but with cumulative ones, you don't.

It's hard to know if this is simply a confusing description of a real situation or whether it is wrong. It suggests that adjectives belong to these two types, which is wrong, but if it really means that adjectives can be listed in two ways, then that's true.

The important point is that it has nothing to do with any particular type of adjective as you can have exactly the same two adjectives with a comma or without. Rather it is a question of coordinated modifiers versus stacked (or what they call cumulative) modifiers. Coordination looks like this: (m , m) h, but stacking looks like this: (m (m h)), where m is a modifier and h is a head noun.

1a) a powerful, new tool (coord)
1b) a powerful new tool (stacked)

In 1a, we have a tool that is both new and powerful. In 1b, we have a new tool that is powerful. This is probably not a distinction worth drawing, but look at this:

2a) a powerful, political force
2b) a powerful political force

In 2a, we have a force that is powerful but also political (as opposed to, say, moral). In 2b) we have a political force that is powerful (as opposed to weak).

3a) a small, happy family
3b) a small happy family

In 3a, the family happens to be both small and happy, but in 3b, the happy family is small (in comparison to most happy families, which we may expect to be big).

Where a modifier is a noun, it is usually indicative of a type and so lends itself to stacking. This also indicates why nouns will come at or towards the end of a string of modifiers.

4a) a big, ugly brick house

But it can also be coordinated:

4b) a big, ugly, brick house

4a is a brick house that is big and ugly. 4b is a house that is big, ugly, and made of brick.

Friday, March 20, 2009

Grammatically Speaking incoherent

Richard Firsten continues to provide questionable answers to grammar question in his "Grammatically speaking" column for TESOL . I must say, I'm disheartened that nobody in the editorial department there seems to have cottoned on to it yet. This time around there are enough problems that we're going to have to break things out into multiple posts.

We'll begin with the following question: is the -ing word in this sentence a present participle or a gerund?
He was famous for consistently picking stocks that gave high returns.
Firsten answers,
"What gives this away as a present participle and not a gerund is the word consistently. A gerund is like the noun form of a verb, so if it's preceded by nothing at all or by an adjective, we can see that it's a gerund, working as a noun."
Amid the confusion here, there are two useful hints. Adjectives do not modify verbs and adverbs do not modify nouns. So if your choice is between a word being a noun and a verb, and it's being modified by an adjective, you can bet the house it's not a verb. Equally, if it's being modified by an adverb, it's not a noun.

The rest of this is nonsense. "A gerund is like the noun form of a verb." And this is "a gerund, working as a noun." So, putting those together, we get: picking described as being like the noun form of a verb, working as a noun. So, is it a noun or a verb? Firsten doesn't seem to know, or if he knows, he's not telling. Maybe he's been reading Cowan's Teacher's Grammar of English.

Anyhow, let's look at another sentence: He was famous for his tan. It would be very unhelpful to say the NP his tan is "working as a noun." Rather, we would say it's functioning as the object of the preposition for. So, being the consistent folk we are, we'd say the clause consistently picking stocks that gave high returns is functioning as the object of for in the sentence in question.

What about picking itself? It's a verb. It's modified by an adverb, as Firsten points out. So, despite some confusion, he gets us to the right answer. But rather than stopping there, he bumbles on:
"If we tweak the sentence and have He was famous for picking stocks that gave high returns, we have a gerund."
No! No! No! This is to suggests that picking is a gerund by default, and that it is only by adding the adverb consistently that we turn it into a verb. That's backwards. Rather, it is because picking is a verb that it becomes possible to modify it by an adverb. But whether there is indeed an adverb there or not is irrelevant.

Then he gets the next part right:
"We also see that picking has a direct object, stocks, so that's another way we know that in this case, picking is a present participle"
Of course, this directly contradicts what he said earlier about it being a gerund, but don't let that bother you. It doesn't seem to bother the editors at TESOL.

Next time, what is backshift?

Thursday, March 19, 2009

From blogosphere to hereosphere

Goofy from Bradshaw of the future came by today and we had a nice chat. If there's anyone else reading in the Toronto area, look me up.

pre- and post- as preposition-forming prefixes

[please see the reconsideration of this post here]
This morning coming in to work, I heard Andy Barrie on CBC say something like "How is the water post-Walkerton?" I've forgotten the actual sentence, but it ended "post-Walkerton." (See here for an explanation of the Walkerton Tragedy.)

It struck me that post-Walkerton sounded, prosodically, like one word rather than two, and that, if it were one word, it would be a preposition (albeit an intransitive one, explained briefly here.) The Cambridge Grammar of the English Language claims that pre- and post- can form adjectives, nouns, and verbs, but doesn't allow for the possibility of prepositions. Clearly though, the word post-Walkerton as used above is neither adjective, verb, nor noun.

The other way to look at this, I suppose, is that post is itself a preposition rather than a prefix.

Tuesday, March 17, 2009

TOEFL TV

Just in case you needed to know, the good folks at ETS have found another way to market their juggernaut Test of English as a Foreign Language: their very own YouTube chanel. Teachers and students share their TOEFL tips in some of the most compelling TV you've seen. Actually, it's probably no worse than most of what's on YouTube.

Monday, March 16, 2009

Japanese as a distant language?

We all make mistakes, big and small. I know I do. So I usually do not criticise others for trivial mistakes. Not that I am particularly magnanimous; it is simply that, hey, life is short, let's look for good rather than bad.

So I am not lamenting nor ranting when I say I often come across strange things about the Japanese language written by, of all people, professional linguists. They baffle me. I will give you just a few examples.

While discussing Japanese classifiers on page 70 of Concise encyclopedia of grammatical categories, Craig gives the following:

empitsu ni-hon 'two (long) pencils'
inu no-hiki 'two (animal) dogs'
no' chiyo 'the (animal) chicken'
ixim wah 'the (corn) tortilla'

All these are labelled 'Japanese', which happens to be my mother tongue; yet I have difficulties. The first one is fine. The second contains a simple mistake: 'no-hiki' should read 'ni-hiki'. But I simply cannot make out what the third and fourth examples are. Maybe they are some esoteric expressions that I know nothing about - or something has gone awfully wrong (perhaps these two are from some other language).

An introduction to language and linguistics, edited by Fasold & Conner-Linton (2006), is a fine textbook, but apparently the editors did not have enough time to check certain simple facts. On page 29, you have the following warning:

In Tokyo, you want to be careful to order [bi:ru] 'a beer' rather than [biru] 'a building,' and to ask directions to a certain [tori] 'street,' rather than a [to:ri] 'gate' or [to:ri:] 'bird.'

Alas, they will still be lost in Tokyo, ordering some fried gates or grilled streets in a restaurant. The correct combinations, if you care, are: [to:ri] - street, [tori:] - gate, and [tori] - bird. To my knowledge there is no such word as [to:ri:] in Japanese, except perhaps for phonetic translation for 'Tory'.

Incidentally, the same textbook (on page 410) gives a Japanese sentence meaning 'The boy who hit the dog is my brother' - but wait. The Japanese sentence says 'The boy who hit against [i.e. whose (entire) body hit] the dog is (my) brother'. Small difference on paper, big difference in the real physical world.

Even Pinker sometimes nods. In his wonderfully entertaining The stuff of thought, he gives bakatara as a Japanese word meaning 'stupid' (on p.336 of the Penguin edition); the correct word is bakatare. Yes, it's just a single small letter - a tiny letter that distinguishes, say, fad and fed ('The dog has just been fad'? hmm ...)

I will readily admit that language professionals are usually careful, indeed skilled, in dealing with foreign languages (in fact, they are the ones who get my Japanese name right every time, while banks, shops and other institutions get it wrong). George Lakoff discusses the Japanese classifiers in his Women, fire and dangerous things with admirable accuracy; yes, he checked with a Japanese speaking person (whom he thanks in the Acknowledgments). All it takes is just a little time and care.

But then, that's the hardest thing to do - especially when what you are dealing with is something as distant as the Japanese language - or, in my case, something as distant as the English language. That's why it takes me so long to translate - oops, time to get back to work. I have some translation to do, a distant trip. I'll get all trippy.

Sunday, March 15, 2009

Pronunciational self delusion

I was having dinner earlier this week with a number of colleagues when the topic got around to pronunciation. After an initial discussion of being laughed at by Americans for our pronunciation of out, the conversation went something like this: A and B are Canadian, and C is British and knows a lot more about phonology than I do.
  • A: Canadian English is the clearest. Even the main US networks are hiring Canadians as anchors or their anchors sound Canadian.
  • B: Clearer than British English you think?
  • C: Oh, yeah.
  • me: Which British English?
  • C: If you could hear the way my brother talks...
  • me: But it just depends what you're used to, doesn't it?
  • C: But Canadians pronounce all the sounds clearly. It must be easier for students learning English to study here.
  • A & B: Sure.
  • me: Come on, how do you say Toronto?
  • A: toe Ron toe
  • me: Really? You don't say Trawna?
  • B: Maybe the younger people do, but not our generation.
Far be it for (OK, from) me to argue, but I just wanna point out that Patrick Corrigan's editorial cartoon in today's Toronto Star has mayor David Miller's bicycle being eaten by a pothole. A street sign in the background says, "Trawna".

Saturday, March 14, 2009

The decline of the subjunctive

There's no doubt that the subjunctive in English has lost much of its historical scope. There used to be present and past subjunctive for all verbs, but the only vestige of the past subjunctive is the irrealis use of were that I referred to yesterday. When you've only got one form, it's hardly worth calling it a system. The simple past tense has taken over much of the work in sentences like If you saw it, you'd know.

Many claim that the subjunctive continues its fade today, but there are a brave few, such as Charles Finney of the University of Tennessee, who actually argue that it's making a comeback. I'm very pleased to see that Finney is at least somewhat interested in the data. He keeps a list of examples that he runs across to illustrate his point. The problem is that you can't show a revival by looking at a single point in time.

So to get a time series, I went to the TIME Corpus and the COCA and used that [pronoun] be as my stand-in for the whole of the subjunctive. I'd say it's a reasonably representative sample, though clearly far from perfect. If you're playing along at home, the query would look like this:
The result from the TIME corpus shows a clear drop off over the last century with only the 1930s and 1990s going slightly against trend. The most recent decade has, however, been very hard on the subjunctive:
The drop off in the last two decades is confirmed by the COCA, though it doesn't look quite as desperate:
The thing I found most interesting is that spoken English appears to be the biggest user of the subjunctive, though academic English isn't far behind.
So it appears that the subjunctive really is continuing its decline.

Friday, March 13, 2009

Whether Grayling know the subjunctive

In making a point about the (in)accuracy of internet sources, A. C. Grayling remarks in passing on the grammar of the following:

Look thy last on all things lovely, Every hour. Let no night Seal thy sense in deathly slumber Till to delight Thou have paid thy utmost blessing.

-Walter de la Mare

Grayling claims that Thou have is an error, and one committed by the poet. It should, he says, be Thou hast. But Alan Larman, of Congleton, Cheshire, UK, writes to NewScientist that this is the subjunctive mood and no error at all.

In modern English, the use of the subjunctive is highly restricted and occurs mainly in subordinate clauses introduced by that (e.g., It's important that he be on time; I take were in if I were you to be an isolated irrealis form and not part of the subjunctive, which uses the plain form of the verb). My problem is this: I don't know enough about the grammar of early modern English (the style, if not the time, in which la Mare was writing) to know whether the subjunctive is possible here. The poet was writing in the early part of the twentieth century, at which time thou had been out of use for perhaps 250 years. So it is indeed possible that, in trying to be poetic, he goofed.

There is, however, a still-used frozen form that displays exactly the pattern in question: until death do us part, which suggests strongly that the error is Grayling's.

If anybody can settle this definitively, I'd appreciate hearing from you.

Pearson Test of English

On Wednesday night, we were discussing how to deal with graduates of our college who had great credentials and ability, but who couldn't land a job because of their spoken English. We agreed that one issue is having some kind of objective benchmark of speaking ability, and so I brought up (again) the Versant tests.

The next day the associate dean sent me an e-mail about the Pearson Test of English (PTE), which has been in development for a number of years and will be available this October. Pearson is trying to get colleges and universities to say they will accept the PTE for admissions purposes, and somebody there had contacted our registrar about it. She'd forwarded it to the associate dean and he forwarded it to me with the question, "Is the oral test you spoke about yesterday a part of this test?"

This wikipedia article claims that that indeed appears to be the case, though there's nothing to back the claim up.

The makers of the Versant tests have claimed for a long time that they can produce valid reliable scores automatically in a fraction of the time it takes a behmouth like the TOEFL. The problems is that, even if it's true, nobody believes it. The PTE is about 3 hours as compared to Versant's 10-15 minutes, so it will be interesting to see how the statistical properties of each line up.

Tuesday, March 10, 2009

No word for X

A standard (journalistic) trope for exoticising a language is to claim that it has no word for x. According to an article in the Native Times, Doris Jean Lamar McLemore is the last person alive who can fluently speak "the language of the Wichita Indians," a language so complex that
"it takes 60 to 90 minutes to transcribe even a small portion of it.

There are no words for “hello” or “good-bye,” for example. The closest to a “hello” you can give someone is a phrase for “How are you?” – “e : si : raci : ci.”
I can think of another really exotic language with no words for hello: French. The closest to a "hello" you can give someone is to wish them a good day or good health. I can't say I know much about transcribing, but it seems to me that the speed would have to do much more with familiarity than with complexity.

Finally, it's interesting that nowhere in the article is the language actually named. It's simply called the language of the Wichita Indians.

Saturday, March 07, 2009

Only a couple letters off

A few months ago, I wrote about a couple + plural noun (e.g., a couple things vs. a couple of things). Today it's in the news, as pointed out by Geoff Pullum over at Language Log. When there was a typo on a gift from Hillary Clinton to her Russian counterpart Sergei Lavrov, the Clinton adviser Philippe Reines is reported to have protested that the word they printed is only a couple letters off.

Geoff puts only a couple letters off in quotes, but I'm not sure whether he means to attribute that to the newspaper story or to Reines. I can't actually find any stories with those words in quotes, and indeed a Google news search only turns up two stories with that wording, one from Fox and one from Asian News International.

Friday, March 06, 2009

A love letter

Walrus Magazine has just announced the winners of their love letter contest. The following took second prize:

Dearest;

I still remember the day we met; my grade-eight grammar teacher recommended you with forget-me-not, fire-engine-red ink in place of a capriciously employed em-dash.

Only the bravest and worthiest of young grammarians were permitted to adopt you in their prose; hence, from that day forth I've worn you like a badge of honour.

I still recall the moment my juvenile narrative first profited from your mature wisdom. A reluctant convert, I snuck you into a three-hundred-word handwritten exposition on "Casey at the Bat"; I even used you outside of the quotation marks.

Since that day I've been your faithful admirer. My longing for your is even evident when I type en français; my heart commands that I not separate you with a space from the word that precedes you.

In my vintage 1982 edition of The Chicago Manual of Style yours is the best thumbed of all pages–dog-eared number one-forty-seven–where into six little vignettes from 5.68 to 5.73 those know-it-all Chicagoans squeezed your every known use in the English language; e.g. when you appear before the expression e.g. if the break in continuity is greater than that signaled by a comma. (For comparative examples see 5.54.)

But I know you're worth much more than these grammarians perceive; surely the greatest poets and philologists toiling in tandem could never describe the fullness of your beauty.

Oh, how I've sinned in the name of loving you: pride, in objectifying you when no other mark can satisfy me punctually; lust, for the time when I may again see at the tip of my pen your seductive point-and-bend; envy, when those like Vonnegut and Hemingway treat you better than I; wrath, my rage when I see a perfidious comma in what was previously your domain; sloth, when thoughtlessly I start a new sentence rather than employ a coordinate clause; greed, when I wish to hoard you all to myself and not share your muted grandeur with others; and gluttony, my sin when I use you six times in a horrendously long sentence.

Mine is a love unrequited, I know, for you are just a fleeting fleck between a colon and a comma.

But dearest, I would never misuse you, as so many do, next to a callous conjunction that connects two independent clauses, and my love for you is greater than my desire to exploit you as do your many Internet sycophants.

Should you never return my love still I shall remain forever your humble and faithful applicant; accordingly, never will I remit watchfulness of your applicability as semi-colon, the connector of two coordinate clauses; semi-colon, the clarifying attribute dividing a series of two or more objects in a post-colonal sequence with internal punctuation; and semi-colon, the super comma, where an ordinary comma would just be too damn confusing!

I love you always ;
R.A. Johnson

Thursday, March 05, 2009

Simplifying rules

After the talk I had with my son the other day, I suggested that he ask his teacher the same questions he asked me. I was disappointed to learn that the response was: "sometimes authors (who begin sentences with and) forget the rules." This was not what I was hoping for, so I sent a gentle letter to school (my wife insisted on editing it to ensure gentleness). I received the following response:
Thank you for sharing your expertise! You certainly know the rules thoroughly!
For grade 2, to simplify things, we generalize that we use 'and' once in a sentence, as a joining word or part of a list. Also, to reinforce the idea of joining words, not to start a sentence with 'and'. Certainly as they enter older grades, refinements as to the use of 'and' and 'but' will be made.
Thanks again for sharing!
I'm glad to learn that she views this as a simplification rather than a real rule. As I said last time, I don't believe it's a needed simplification because I feel confident that a seven-year-old can understand that you can use and at the beginning of a sentence as long as you don't do it too much. I'm also less sanguine about the chance of refinements being made in future grades. Most of the students (pre-service teachers) in my current TESL class came in believing in a no-sentence-initial-coordinator rule, and in some cases I haven't been able to overcome a lifetime's faith in it.

Wednesday, March 04, 2009

National Grammar Day

In the US, today is National Grammar Day. Bradshaw of the future has an apropos explanation of what grammar is.

Monday, March 02, 2009

It has begun

I think today is the first day that my son asked me about a prescriptive grammar rule. He's in grade two, and he asked me if it was true that you can't start a sentence with and. I asked him why he was curious, and he said that he'd seen it in books but his teacher had said that it was against the rules.

I asked him why his teacher might do that, but he couldn't imagine a reason, so we talked about how kids often tell stories with and between every "sentence" and I asked him how it sounds if you use the same word(s) too many times. In the end, we agreed that you can start a sentence with and but that you just shouldn't do it too much.

Which seemed to satisfy him, until he added, "but you can't have two ands in the same sentence, right?" So we got to look at the difference between coordinating clauses and phrases, and he had no trouble seeing the difference once it was brought to his attention.

If a bright seven-year-old can notice a discrepancy between what he's being told and what he sees, and if he can understand the facts of grammar with a little Socratic questioning, why do we teach all these fake oversimplifications?