To truly understand a word, you need to know more than just the definition. You need to know how to actually use it.
Teachers will tell you that children often cram a lot of new vocabulary into one sentence just to sound smart, when it’s clear that they don’t understand how to use those words.
Word pairs – or collocations – are combinations of words that are statistically likely to appear together in text. They give children the contexts they need to start using new words correctly within minutes of encountering them.
“Word pairs are words that are statistically likely to appear together in text.”
You might think of word pairs almost as set phrases, such as bulging eyes or bulging wallet, or crave attention and crave coffee. But where do the statistics come from? Well, from various corpora – or databases of collected texts – like books, magazines, news stories, transcribed conversations and blog posts. All of that content, from writers all over the world, is fed into a computer and analyzed for frequency and patterns of use.
“English can be mapped more scientifically and more accurately than ever before”
Thanks to modern technology, bodies of texts – anything from entire novels to 140-character tweets – can be scrutinized and mined for the patterns of how English behaves ‘in the wild’; the world of words, dictionaries and communication can be mapped more scientifically and more accurately than ever before.
Today, if you search for a word in a corpus, a computer can tell you how often it is used, which other words it is used with, and how often those two words are used together. Words that are often used together get a high score.
“Words pairings become part of our collective consciousness.”
If blundering is used alongside politician enough times, that pairing becomes part of our collective consciousness. So the numbers behind word pairs are a direct reflection of how writers have used words in practice, a direct reflection of the patterns of language that surround us.
Encountering a new word alongside its word pairs allows children to reach a meaningful understanding of the various contexts, quickly. That’s why it’s an effective way to increase the speed and efficiency of word learning.
Knowledge of word pairs is proven to increase the quality of children’s academic written work. A study by Dr George McCulley found that readers perceive texts containing many common word pairs to be more readable and logical, and this translates to better scores in written tests. Using vocabulary in its most typical contexts shows an overall awareness of and familiarity with English usage, and gives writing real clarity and flow.
“Word pairs are essential to creative writing.”
Word pairs are also essential to colourful, creative writing, but for different reasons. That’s because it’s only once children know the accepted conventions of language that they can break them to artistic effect.
Research by Beata Beigman Klebanov for ETS showed that “higher scoring essays tend to have higher percentages of both highly associated and dis-associated pairs, and lower percentages of mildly associated pairs.” In other words, students who can use words in surprising, unusual ways – i.e. with low scoring word pairs – also receive the best grades, since this kind of writing feels inventive and original.
Interestingly, in her research Klebanov found that the information provided by studying word pair frequency could be used to effectively improve automated essay-scoring systems. In other words, computers can reliably score tests, aided by analysis of word pairs.
Learning word pairs fast-tracks children to gaining a writer’s intuition and honing their command of language. It’s a fast, focused way to learn, that goes right to the heart of what makes good writing.
Interested in word learning for your child?
- McCulley, G. (1985) Writing Quality, Coherence, and Cohesion. Research in the Teaching of English. 35 (3), pp. 305-327.
- Beigman Klebanov, B. and Flor, M. (2013) Word Association Profiles and their Use for Automated Scoring of Essays. Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics, pp. 1148-58.