Sometime in the middle of the last century, Benjamin Whorf, famous for his contention that language shapes thought, made a controversial statement about the Eskimo language having seven words for snow. Frequently quoted or, rather misquoted overtime, Whorf’s number of snow words was inflated to nine, twenty, fifty, and even one hundred. A similar snowball effect seems to be taking place with the contentious issue of how many exposures a vocabulary item needs for its retention.
About 20 years ago, researchers originally proposed that a second language (L2) learner needs six exposures to a word in order to retain it. However, with the emergence of more and more research into vocabulary acquisition, the number swelled to 10, then 15 and has now reached 20. Do we really need 20 exposures to the word? How realistic is it? What constitutes an exposure?
How many or how well?
First, my personal experience as a language learner as well as a language teacher shows that you can well do with fewer than six encounters if you have an acute communicative need. For example, if you are desperately groping for a word in L2, and a teacher (or a competent language speaker) supplies it on the spot you are quite likely to remember it. I have learnt many words in French and Spanish exactly this way, by “echoing” what my conversation partner or teacher has said.
Evidently, it is not only “how many times” but also “how well” that should matter. If you come across a lexical item many times but do not do much with it (known as incidental learning), you probably need countless repetitions before it is committed to memory. Conversely, if you are more actively involved in processing the language you meet (known as intentional learning) you are more likely to remember it. It is therefore up to teachers, to create situations and provide meaningful contexts in which learners can have the opportunity to recycle and reuse the language they have learnt.
In or out of context?
Decontextualised language practice was frowned upon during the heyday of Communicative Language Teaching. However, more and more evidence suggests that decontextualised vocabulary learning is perfectly justified (see for example Laufer, 2006). Students on EAP courses require many academic words such as fundamental, evolution, and welfare to understand dense scientific texts. In order to comprehend them you need to reach a certain level (according to Nation, usually 6000 – 8000 words), and decontextualised rote learning can often be the only way. However, how can you grasp, let alone use, such expressions as come to think of it or the next thing I know… without meeting them over and over again in meaningful contexts?
Testing the optimal number of encounters is also extremely difficult because there are an infinite number of external factors. One such factor is learnability. Some words are inherently difficult to learn. For example, buy more or less corresponds to the Spanish comprar and should not pose much difficulty for a Spanish-speaking learner of English while many other verbs cannot be mapped onto the same concepts in your first language (L1). For example, pursue does not have a direct word-for-word correspondence with Hebrew. Join can be rendered into French as rejoindre, s'engager, adhérer or devenir membre. It depends on what you want to join in order to choose the right French word, i.e. on the collocations of the verb.
Teaching or researching?
To combat this and other intervening factors researchers often resort to using made-up words (known as “non-words”) in their studies, which I personally have ethical reservations about. Another problem with the studies attempting to count the number of encounters is that they mainly focus on the passive knowledge of vocabulary. For a practitioner teaching Communicative courses where the main aim is so get students to speak and use new language such research is neither valid nor of much use.
When I teach a new lexical chunk in class, I attempt to push the learners’ output as soon as possible by encouraging them to experiment with the new language in different situations. If I were to put my students under experimental conditions, I would have to deliberately stall them while busy counting the number of times a particular chunk was encountered before the learners could produce it.
Individual words versus multi-word items
Most researchers investigating the number of encounters needed to remember new vocabulary unfortunately overlook collocations. It is ironic that the same linguists who strongly propound the importance of learning vocabulary in chunks are preoccupied with counting exposures with individual words.
Establishing how many encounters a learner needs with a new collocation is even trickier What constitutes a new collocation? An intermediate level learner may be:
a) familiar with the words meet and requirement but not know that they can collocate (meet the requirements);
b) familiar with one of the words (reach) in a collocation (reach a compromise)
c) unfamiliar with any of the words (e.g. bear resemblance)
Somewhat paradoxically, evidence suggests that type A collocations may present more difficulty for learners than entirely new collocations (type C). It could be the novelty effect that makes learners pay more attention to new collocations and overlook partially familiar ones, hence frequent errors such as *did a mistake or *made homework.
Furthermore, some research suggests that lexical chunks, which alliterate (e.g. slippery slope, prim and proper), tend to be learnt faster than the ones which do not display such a pattern (see Boers & Lindstromberg 2005).
Coursebooks & schools
Unfortunately, course books do not provide enough encounters with lexical items. Nowadays most course books are organised thematically. While learners may be exposed to the same lexis within a particular unit, few course books ensure the same lexis is recycled across the textbook, i.e. over a series of units. It is therefore the teachers’ responsibility to ensure that the language our students come across is revised in subsequent lessons and regularly recycled.
I once came across an article, which suggested optimum revision intervals as follows:
10 minutes after the initial encounter
1 day after the initial encounter
1 week after the initial encounter
1 month after the initial encounter
6 months after the initial encounter
Is it feasible in a school setting? Most EFL classes take place once a week, while in secondary schools there are a lot of timetabling constraints to ensure that the above intervals are adhered to. I would therefore recommend revising new lexis as often as possible whenever an opportunity arises. Teachers can make a good habit of finishing every lesson with a review of the language they have collected on the board during the lesson and starting every new lesson with a quick revision of the language covered earlier on the course.
As regards productive knowledge, the teacher should be able to spot when learners are trying to retrieve a partially learnt item and help them by eliciting it. The teacher may aid the students with prompts and questions or, in other words, provide the necessary scaffolding.
While researchers continue to argue how many encounters with a word are necessary in order for the learner to retain it and what the optimal conditions for retention are, most agree that frequent recycling is essential for effective vocabulary learning. It is our responsibility as teachers to make sure that regular revisiting of the lexis is part of a language course.
- Boers, F. & Lindstromberg, S. (2005). Finding ways to make phrase-learning feasible: The mnemonic effect of alliteration. System, 33(2), 225-238
- Laufer, B. (2006). Comparing focus on Form and Focus on FormS in second language vocabulary learning. Canadian Modern Language Review 63(1), 149-166
- Nation, I.S.P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press.
By Leo Selivan
This article was first published in September 2010