5 Different collocations or units of meaning = different networks of words
We saw in the second post in this series that coming up with clear-cut definitions of single words is difficult because individual words impact on each other and change their ‘meanings’ slightly. For example, Hanks argues that the typical usage of raging fire is different to blazing fire in that it can be seen to be a both different kind of fire (an accidental one as opposed to one that’s deliberate) and a different kind of ‘big’ (big and out of control – as opposed to big and warming). There is a further knock-on effect of collocation though. That is that each collocation will, in turn, have different collocations and networks of words connected to it.
So if we take raging fire, an associated network might be turn into (a raging fire) / call the fire brigade / out of control / rescue / die (= people, things) / fight to control and put out. With blazing fire, however, you might expect to use or hear words such as build / sit round / die (= the fire) / camping / the night sky / keep us warm, and so on.
This idea can apply to ‘simpler’ collocations where we might consider the individual words to be more obviously ‘separate’. Take new car / old car. Obviously, both collocations will share some collocates, such as ‘drive’, but in the main, the collocates will be different – think how you’d expect these phrases to be completed: I got rid of my …. ; I can’t afford a(n) …, I’m picking up my …; my …. died. Beyond these collocations of the collocations, we are also going to have other word associations that we might expect to come up in a ‘text’ about old cars. Again, these will mostly be different to those associated with a new car. For example, rust / fail its MOT (British English) / [part] fell off / the [part] is broken / [number] miles on the clock – and so on.
Also, of course, we might add that these networks will be totally different from the networks around things like new / old house; new / old wine; new / old friends.
This is speculation, but perhaps this is another benefit of language organising itself around collocation or what Wray call ‘units of meaning’. If you work round units of meaning, you reduce the number of likely collocates and associations when compared to trawling through all the collocates of, say, new and all the collocates of car. This narrower focus may well enable our students to speed up recall and construct text more easily in real time.
Applications of this principle
In a way, the above shows the limitations of teaching words in collocation – and the importance of getting students to engage with vocabulary beyond this level. You can do this through:
– asking questions about words and collocations connected to usage – see the Exploiting exercises section.
– getting students to ‘text’ words by reading lots and putting words into stories or conversation (some may also call this ‘free practice’ or doing tasks)
– teaching and learning groups of words round a collocate or connected to a collocate by a text – rather than learning words in hyponomous sets (car, motorbike, train, etc.).
Very interesting examples in this post. You do make words “work” that way, they get meaning in real environments and this way more words are taught. It is the contexts and thus collocations that will help us retrieve those words & collocations when we need them. I think this fits very well with my approach to vocabulary lists as suggested in the post “In so many words” on Lexical Lab.
Thanks for taking time to read and comment Bruno.
Yes, think there are plenty of overlaps with your post.
Have you read the Patrick Hanks book, LEXICAL ANALYSIS?
He explores the problems lexicographers have of nailing down one explicit, fixed meaning to words in light of the fact that meanings shade one way or another depending on context. Highly recommended reading.
I did not know that book, Hugh. Thanks for the tip! I will check it out.
James Thomas has an interesting approach partly influenced by Patrick Hanks work. You can get an idea of this from the paper Stealing a march on collocation – Deriving extended collocations from full text for student analysis and synthesis [https://books.google.fr/books?id=1xbyCAAAQBAJ&pg=PA85&lpg=PA85&dq=Deriving+extended+collocations+from+full+text+for+student+analysis+and+synthesis%3F&source=bl&ots=sEucVAepHH&sig=FzwsF1tvQmtqkRULy_v6Aa4FG-c&hl=en&sa=X&ei=OwRnVYr-EMyoNsOjgPAC&redir_esc=y#v=onepage&q=Deriving%20extended%20collocations%20from%20full%20text%20for%20student%20analysis%20and%20synthesis%3F&f=false]
He uses it to work from a text to collocations, and I think it can be adpated to work from collocations to text
ta
mura
Thanks for that Mura. hadn’t seen that before.
Will read later.
Don’t suppose you have a copy of “Multiple Affordances of Language Corpora for Data-driven Learning” you fancy lending me?
All the copies I’ve seen online are ridiculously priced.
hi sorry no don’t have a copy of that book way overpriced! i guess wait for paperback and/or email authors directly for papers 🙂
Have just read through the James Thomas’ paper and it’s an issue I’ve often fretted about myself. That sample of academic writing from the Computer Science researcher that he starts with is very familiar and reminds me very much of many EAP essays I’ve had to plough through over the years. His notion of Collocation Plus, which he defines as “typical co-occuring words” is very much what we’re interested in as well, and what this post is getting at. Andrew has been calling this the collocation of collocations, but maybe Collocation Plus is better.
He’s also depressingly accurate in his observations about how little impact and trickle down of key ideas from corpus studies to teacher education and coursebooks there’s been and how the mainstream has still not yet come to anything remotely like terms with ideas such as colligation or the reconciliation of grammar and vocabulary!
We’re trying to do our bit, but change is mighty slow!
I also liked his idea of word sketches – quite similar to what Bruno Leys was talking about in his recent post here, in a way.
In terms of his approach to “making himself redundant,” I think the first thing that needs to be said is that it’s something that may work with his students, but that they’re a very unusual cadre! They’re obviously already very good, very bright, very driven and independent and have been encouraged to be very aware of language.
That’s not to say that most other students couldn’t get to this stage eventually, but simply to note their unusual nature and to be wary of assuming the route he takes with them is applicable with lower-level, more General English students, who generally need more guidance and support – and input – from teachers.
I think what’s most essential is that teachers actually take on board from the very earliest levels the importance of context, co-text, collocation plus, and so on and pay more attention to the words and grammar that go with the words they’re teaching. The best way to stop students producing the kind of sentences featured in the article is to do more work on both language awareness but more importantly language development – in its fullest sense – from day one. Prevention is better than cure – and it’s easier to build a decent wall from scratch than to have to deconstruct and patch up an already-shaky one!