Multiword Units: Definitions and Distinctions
The term multiword unit refers to a sequence of words that is retrieved from memory and perceived or produced as a single chunk. Such a unit may consist of only two words (e.g. take care), or several (e.g. on the other hand); it may form an entire sentence (e.g. all roads lead to Rome.). For the purposes of this chapter, the term includes not only units that function as lexical items, but also phrases or sentences that occur very frequently in spoken or written communication and therefore are commonly perceived or produced as a single unit (e.g. I donât know). It also includes units that may appear in variant grammatical forms (e.g. take into account, took into account, havenât taken into account etc.)
An associated concept is collocation: the tendency of certain words to associate with certain other words in discourse, but not necessarily next to one another: for example, the adjective rapid with the noun change in a sentence like The changes were rapid and dramatic (compare: the changes were speedy/quick/fast and dramatic). When a particular collocation frequently appears as a sequence of adjacent words (e.g. rapid change, see later in this chapter), it may be identified as a multiword unit.
Multiword units have been classified in various ways. One distinction is between âtransparentâ and âopaqueâ units, sometimes called âcompositionalâ and ânon-compositionalâ: the first relating to those which could be understood from knowing the component words (like at all times), and the second those which could not (like at all) (Martinez & Schmitt, 2012). They have also been categorized in terms of what function they fulfill in a sentence or utterance: for example, whether they simply convey referential meaning (e.g. wage war on, try out), whether they have a discourse function (to cut a long story short), or convey the speakerâs attitude (Iâm afraid, if I may). Sometimes certain types of units are grouped according to their structural features: phrasal verbs like give up, for example, or binomials like more or less. All these types of multiword units need to be learned. They are not normally to be presented for the first time as sets within any of the categories described above (see further discussion of this point below, at the end of the section In the Classroom 1), but rather as they come up in context and are perceived to be useful to teach. The distinctions and categories, however, are useful to the teacher, and may help to support teaching decisions. For example, the non-compositional ones (whose meaning could not be guessed from their component words) will need more careful explanation than the compositional.
The Importance of Teaching Multiword Units
Vocabulary
The learning of multiword units is important for the language learner primarily because most of them function as lexical items in their own right, in the same way as single words do; few today would dispute the place of vocabulary as the crucial factor in language acquisition.
The number of single-word lexical items that a student needs in order to understand natural oral discourse has been estimated at between 5,000 and 7,000 word families; and even more (8,000 to 9,000) for unsimplified written text (Schmitt, 2008). If we assume that multiword items account for between a half and a third of natural discourse (Conklin & Schmitt, 2012), then the number of these that need to be mastered is also formidable, even taking into account the fact that many of them are compositional, and will be understood â though not necessarily produced â through knowledge of the component words. Moreover, research has shown fairly conclusively that learners within an instructional setting of three to four hours a week where English is not normally encountered outside the classroom cannot acquire even the smaller numbers of items indicated above through incidental encounter during extensive reading or listening (Huckin & Coady, 1999; Laufer, 2005; Brown & Waring, 2008). It is necessary, therefore, to supplement such incidental learning through a substantial amount of deliberate classroom teaching of both single- and multiword items.
Basic multiword units such as of course, look for are common and convey meanings that could not be guessed through knowledge of the component words, so should be taught in the same way as single words. Unfortunately, vocabulary lists in coursebooks still tend to focus on single words, as do traditional frequency lists of lexical items such as the General Service List (West & West, 1953), or even more modern corpus-based ones such as the Academic Word List (Coxhead, 2000), or the Academic Vocabulary List (Gardner & Davies, 2013). In most cases today, the identification and teaching of multiword units are the responsibility of the teacher.
There are, however, more reasons, besides general vocabulary acquisition, why it is important to teach these items.
Fluency
In a seminal article, Pawley and Syder (1983) suggested the idea, now widely accepted, that the use of memorized sequences is the basis of native-like fluency in a language. They included in their definition of âmemorized sequencesâ not only lexical items such as those suggested above, but also â and mainly â common lexicalized sentences or stems such as Will you marry me? (be) sorry to âŚ, I donât think ⌠Clearly, such sequences could be composed by a learner by fitting single words into a grammatical structure according to rules, but there are two reasons why this would lower fluency. First, it takes a lot longer to formulate a sentence by choosing single words and then working out what the correct grammatical form and sequence should be than it does to retrieve a single memorized sequence. Second, the ready-made sequence is likely to be idiomatic and commonly used, and therefore more easily processed by the listener or reader. To take a simple example: the conventional question âWhatâs the time?â could also be expressed as âWhatâs the hour?â, âWhich is the hour?â, âHow late is it?â and so on, as it is in other languages; but the learner needs to know how it is conventionally expressed in English in order to function fluently in conversation.
Grammatical Accuracy
Many multiword units display grammatical features, and could be analyzed according to a grammatical rule: I donât know, for example, or if I were you. The first is the negative form of the present simple, using the auxiliary do; the second is a âsecond conditionalâ with a subjunctive form of the verb be. But why bother working it out? The first is the conventional way of conveying that you are ignorant of something, the second a preface to a suggestion: and these are the meanings we need to explain as we teach them. However, the fact that they are exemplars of regular grammatical features of the language is a bonus: it means that they provide the learner with reliable models of that feature which they can later use, by analogy, as a basis to build similar utterances themselves. The sentence I donât know, for example, can help the learner perceive and formulate other correct combinations like I donât eat, I donât go, I donât mean âŚ, without needing to ponder over the use of the auxiliary do in a negative present simple: they simply âsound rightâ. There is some evidence that young learners of their first language process memorized sequences in this way: they learn them first as global âchunksâ, and only later are able to understand that these actually have a regular structure which can be used in composing further similar utterances (Tomasello, 2003).
This is not to imply that grammatical rules as such do not need to be taught and practiced. Learners who are taught English through a limited number of hoursâ instruction a week in a country where English is little used outside the classroom do not have the exposure time necessary in order to encounter the huge number of exemplars that would enable them to acquire an intuitive knowledge of grammatical forms through communicative input only. They need the short cut that is provided by conscious knowledge and application of explicit rules. The learning of multiword sequences that exemplify a grammatical feature complements the intellectual grasp of the rule, and helps move the learner faster towards an intuitive feel for what is acceptable.
Motivation
Beginner learners do not have to wait to express themselves in English until after they have acquired a basic vocabulary of single words and the rules that enable them to put these together to make meanings. If they master a repertoire of basic interactional expressions early on, they will be quickly enabled to express ideas fluently and accurately in English: a huge boost to motivation (âHey, look at me, Iâm already speaking English!â). Some of the ways students can be helped to acquire such a repertoire are presented below under Simple Oral Repetition.
Classroom Management
This is perhaps a minor issue compared to the ones discussed above, but nevertheless worth noting, particularly in the context of beginner classes. Multiword units used for classroom management include both teacher commands, comments or questions such as âOpen your books at page âŚâ, âWell done!â, âDo you understand?â, or âAre you ready?â and common student utterances such as âI donât understandâ, âHow do you say ⌠in English?â, or âPlease say it againâ. Such items need to be taught from the very early stages, particularly for multilingual classes. Many teachers teaching in monolingual classes where they share their studentsâ mother tongue use the L1 for such classroom-management language, and in some situations this may be inevitable. But in general, teaching and thereafter using regularly such phrases in any class will enable the teacher to run the lesson in English and is likely to lead to learning o...