An Introduction to Applied Linguistics

particular functions in everyday life, for example, Have you heard the one about 
_____ is commonly used to introduce a joke).
• ‘Formulaic sequences or multi-word items’ (the individual formulaic items).
From a learning perspective, it is useful to classify formulaic language into three 
major categories (Grant and Nation, 2006).
1 Core idioms: These are items where the meaning of the parts bears no obvious 
relationship to the meaning of the whole. The most frequent examples of these in 
English are as well (as), of course, such and such, out of hand, take the piss, and serve 
(someone) right. Surprisingly, there are only just over 100 such items in English.
2 Figuratives: These are items that have both a literal meaning and a figurative 
meaning. For example, We have to make sure we are singing from the same hymn 
sheet has a literal meaning, but it is used here with a figurative meaning – ‘We 
have to make sure we are following the same set of rules’. There are thousands 
of these in English and they are continually being added to. They make up 
most of the entries in idiom dictionaries. Typically the figurative meaning can 
be readily related to the literal meaning of the multi-word unit. Core idioms are 
probably figuratives whose history has been lost.
3 Literals: By far the largest group of formulaic sequences are literals, where the 
meaning of the part clearly makes up the meaning of the whole. Some of the 
highest frequency literals in spoken English are you know, I think, thank you, in fact, 
talk about, and I suppose. Most of what are called collocations are included in literals.
The few core idioms need to be learnt as set phrases, although a large proportion 
of them have a changeable form. Figuratives need to be dealt with using a strategy 
that involves relating the figurative meaning to the literal meaning. Literals do 
not require any interpretive strategy, but may be usefully memorized as a way of 
increasing fluency of access and gaining native-like accuracy. Some literals may 
not have parallel L1 forms.
What Vocabulary Should Be Learned?
What vocabulary to focus on should be determined by two major considerations 
– the needs of the learners and the usefulness of the vocabulary items. The 
traditional way of measuring the usefulness of items is to discover their 

---

36 An Introduction to Applied Linguistics
frequency and range in a relevant corpus. The most striking features of the 
results of a frequency-based study are:
• The very wide spread of frequencies, with some items occurring many, many 
times and some occurring only once.
• The relatively small number of words needed to cover a very large proportion 
of the tokens in a text.
• The very large number of low frequency items that account for a very small 
proportion of the tokens in a text.
These three points are illustrated in Table 3.1 and 3.2. Table 3.1 is the result of 
a frequency count of a 500-token section of this chapter. The 500-word section 
contained 204 different word types which made up 169 word families. Table 3.2 
lists the frequency, the number of words with that frequency and the cumulative 
coverage of the tokens. In Table 3.1 not all the words occurring once or twice are 
listed because there were too many of them to show here. A very large proportion 
of words in even a very big collection of texts occur only once.
By doing frequency counts of large relevant corpora, it is possible to come up 
with lists of words that will be very useful for people in the early stages of learning 
a language. Several such lists exist and they provide a very useful basis for course 
design. The classic list of the most useful words of English is Michael West’s (1953) 
A General Service List of English Words (GSL) which contains 2000 high frequency 
words. There is plenty of evidence that 2000 words is an appropriate size for such 
a list, but the list needs to be based on a corpus where spoken language is well 
represented. The GSL is based on written language, and so needs to be updated by 
a new list based on both spoken and written discourse.
The
Of
To
And
Is
A
That
We
Word
Or
Are
Be
In
As
Count
Counting
Tokens
Family
What
If
Same
Types
Vocabulary
All
Do
22
18
17
16
16
14
12
12
11
10
8
8
8
7
7
7
7
6
6
5
5
5
5
4
4
I
It
Occurs
Related
With
Words
Counted
For
Forms
How
Include
Items
Language
Like
Not
One
Only
Question
Stem
Then
Twice
Would
Answer
Can
Closely
4
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
Conservative
Contains
Deciding
Different
Each
Form
...
About
Affixes
After
Also
Anderson
Answers
Any
Approach
Approached
Approaches
Approaching
Asking
Bauer
Because
Being
...
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1