Friday 24 January 2014

Pharrell-timed discussion

I think there is this idea that a seminar is a place for lively discussion and informed debate. In practice, it's not like that. More often, in my experience, it's the seminar leader who does most of the work, and the similarity of the endeavour to drawing blood from a stone varied depending on a multitude of factors: group size, group ability, group demographics, work set, how much other work has been set that week, what time the seminar is, how hot the room is...

Today I had the first proper seminar for my spring morphology module. It started off badly, with only 50% attendance. After getting just 75% to the lecture, this is slightly worrying, so I hope it picks up. Anyway, the group itself is a good one - both seminar groups are filled with bright, keen students. Today I wanted them to discuss a chapter I'd asked them to read. It's an interesting and important discussion of the kind of data we use by the ever insightful Maggie Tallerman, who taught me everything I know about morphology.

I wanted this discussion to be interesting for the students, more so than just doing exercises. I put some discussion questions up on the projector to get them going. I even started them out with an easy exercise to get them in the right frame of mind. Then I split them into groups of 5, not so big that they would fight to be heard, but big enough to generate discussion. Trouble is, with only ten students, that's only two groups, and the room was suddenly silent. Each group could hear the other, and it was too intimidating, and no one said anything.

First, I tried to stimulate conversation by joining each group. That was successful for about three seconds. Then, fantastically, I realised the problem was feeling self-conscious, and put some music on. I had to ask the students the best way to do this, of course, but having been told to YouTube something, I just picked the first song on the 'music' channel, which was Pharrell Williams' 'Happy'. Straight away, discussion was easier because they weren't aware of the other group and me being able to hear everything they said. OK, they were still shy and quiet, but it really did make a difference. They talked till the end of the song (well, with some lulls) and then we discussed the questions in the full group.

In future, all seminar discussions will be accompanied by songs.

Thursday 9 January 2014

Passive voice

Yeah, I'm going there.

UPDATE: As I was writing this post, Geoff Pullum blogged on Language Log, noting that he has completed a paper describing his excellent position on the matter of passives, and you can read his forthcoming paper there.

So anyway, what I was going to talk about was just a couple of interesting times I've noticed it.

First: someone tweeted Alexander Armstrong to say that he wished he would 'stop being referred to as Zander':

As well as being a bizarre thing to say, this is a bizarre syntactic construction. I guess he has phrased it this way to avoid using a vague subject like 'people' or 'contestants and Richard Osman'. After all, one of the main reasons for using the passive is when you don't know or don't want to make a big thing about the subject. Unfortunately, using stop often implies some sense of agency, when it's predicated of a person. Not always: He's stopped burning would not be any more agentive than the fire's stopped burning, of course. I think it must be because stop occurs with another verb in the present participle form, and when that verb is agentive the whole construction is agentive. And refer is agentive. This gives the weird idea that Xander has to somehow stop a process from happening to him that's entirely beyond his control.

Second: passives betraying speaker attitudes. I was listening to the radio and someone told a very funny story about Stevie Wonder, in which he apparently used to freak out his house guests by picking them up at the gate and driving them half a mile or so up the drive to his house (this being terrifying on account of his being blind). The way the person phrased it (sorry I can't give more info or be specific - I had the details but they were lost in a technical snafu) was this:
Stevie Wonder was taught to drive from his gate to his house.
Was taught is passive, indicating that someone is doing something to Stevie Wonder. Passive reverses the roles in the sentence, so giving it in the active would mean the opposite: Stevie Wonder taught someone to drive. However, we have a nifty little thing whereby we have an active verb that means more or less the same as the passive form of teach: learn. (It doesn't mean exactly the same, but in some contexts, including this one, it's near as dammit.) So we now have a choice with no meaning difference, between these two sentences:
Stevie Wonder was taught to drive.
Stevie Wonder learnt to drive.
If there is no meaning difference, we make the choice on (among other things) stylistic grounds. Some would say don't use the passive - I'm not one of those people. Go ahead and use it if you like. But in this case, I think there is another reason not to use it, and that's again related to agency, although this time not just on grammatical grounds.

If you compare the two, with learnt, it was Stevie Wonder's volition to drive. With was taught, agency is given to some other, unnamed person. While I'm quite sure it was still Stevie Wonder's decision to do so, the sentence may be interpreted as reflecting a subconscious belief that people with disabilities can't or don't play jokes, make decisions or control their own lives to the extent that non-disabled people do.

Language doesn't influence thought (oversimplification - sorry), but it can reflect beliefs and reinforce attitudes.

Monday 6 January 2014

Another Sweary Blog Post

A fellow linguist tweeted (ages ago):


To me, this is ungrammatical. For me, the pattern goes like this, where the asterisk next to (4) means it's ungrammatical:

  1. Answer some emails
  2. Respond to some emails
  3. Answer the fuck out of some emails
  4. *Respond to the fuck out of some emails
Answer and respond are a pair of verbs that have nearly identical meanings but behave differently in terms of their syntax. Answer needs a noun phrase complement like some emails, while respond also has to be followed by the preposition to. The structure is nearly the same in both cases, where the verb has some kind of phrase as its complement (that's what the brackets mean):
[answer [some emails]]
[respond [to [some emails]]]
If we add in the emphatic the fuck out of, then the fuck becomes the object of the verb, and we get something more like this:
[answer [the fuck] [out of [some emails]]]
It should be understood as meaning that the fuck, whatever it is, is the thing that is removed from the emails due to their being answered so comprehensively.

So if, like for the tweeter, it's grammatical with respond as well, then we must have an identical construction: the fuck is the complement of the preposition to, and we've got this:
[[respond [to the fuck]] out of some emails]
But for me, respond to the fuck doesn't work so well. It doesn't even have the idiomatic meaning of 'doing something so well it removes all the fuck from it', which we saw above. out of some emails has to be an adverbial phrase, I think, being something the location of the whole action. You'd have to respond the fuck out of some emails, and in fact that is much more acceptable to me.

Saturday 4 January 2014

Because survey.

I made a survey, because the construction that I called 'because+noun' back in July 2012 has suddenly become word of the year and we still don't know how it works! If you could take it I would be pleased. Many thanks. If you can't see it below, it's here

Thursday 2 January 2014

What are the chances?

I've been catching up on iplayer episodes of the Christmas 'University Challenge', and in episode 7 Rory Bremner gave the same wrong answer I did: King Cnut (Canute? Not sure of how we spell that these days).

But the thing is, I was guessing any random historical figure, as the question was to say which historical figure some poem was about and I'd never heard of the poem or the poet, or indeed the correct answer, so I didn't even know what century to go for. (History is not my quiz forte.)

So why did we both say Cnut (Canute)? Maybe simple chance, but it's possible we were primed. That's when you hear or see a word and it makes you more likely to say a word that's similar phonetically or semantically. Paxo ended his question with the word 'brute', so maybe we subconsciously picked a word that rhymed.



Ten hundred words of confusing syntax

There's an interesting blog called 'Ten hundred words of science', in which people try to explain their research using only the thousand most common words of English (go to the page to find out why and how these words were determined). 

Here's a couple of screenshots: 

You'll notice that using simple words can have the unfortunate consequence of requiring very complicated syntax. The first paragraph is very hard to parse, especially that first sentence which has a very long subject [Big human like men animals that people go to see at parks they pay to get into]. It also has a comma following the subject which is strictly incorrect, but which would only add to the difficulty if it wasn't there. Similarly, the last sentence uses incorrect commas to try to clarify a very awkward construction.

The second entry, on the right here, has an almost incomprehensible sentence in it, the first sentence in the second paragraph. It does get better, and of course I picked two of the worst ones in this respect, but it just shows: simpler words does not necessarily mean simpler writing.