Chapter 1 the study of collocations

particular community, or if it is an "arbitrary choice, in terms of linguistic 
structure, for the role of standard expression".  For example, 'it's twenty to six' 
is a standard expression in English since it is a convention that one tells 'to 
[Hour]' rather than 'preceding [Hour]' or 'before [Hour]', and 'I want to marry 
you' is an arbitrarily established standard usage, compared to a less standard 
paraphrase such as 'I wish to be wedded to you', which could be used in a 
formal letter or a satirical speech (Pawley & Syder 1983:198).  As with most of 
the theories examined so far, Pawley and Syder do not define the notion of 
lexicalised sentence stems any further, and they did not offer an explicit list of 
sentence stems that could be used as a framework in the study of collocations.   
 
The view that language consists of blocks or 'chunks' was also supported 
by Nattinger and DeCarrico (1992), who proposed the compilation of a lexical 
phrase dictionary for L2 learners.  Nattinger and DeCarrico give the following 
examples of lexical phrases for inclusion in the dictionary: 
 
 
Conversational Maintenance (regularities of conversational interaction 
that describe how conversations begin, continue and end).  Summoning: 
 
153

---

Excuse/pardon me (sustained intonation); Hey/hi/hello, (Name); How 
are you (doing)? I didn't catch/get your name; Do you live around here? 
Hello, I'm + NAME; Good morning/afternoon/evening, (how are you) 
What's up? (Nattinger & DeCarrico 1993). 
 
 
From the examples of lexical phrases, as these were presented by 
Nattinger and DeCarrico, it appears that lexical phrases are not the same as 
collocations or lexicalised sentence stems.  Lexical phrases appear to be more 
general than collocations and less systematic than lexicalised sentence stems. 
Also, Nattinger and DeCarrico are not concerned with providing explanations 
about why certain lexical phrases are put together, which would be more useful 
for the study of collocations. 
 
A set of criteria for examining whether a combination of words is a 
collocation or not is outlined by Kjellmer (1984), who also suggests the study of 
collocations in a grammatical framework.  Kjellmer defines collocations as 
"lexically determined and grammatically restricted sequences of words" 
(Kjellmer 1984:163).  According to this definition, only recurring sequences that 
are grammatically well-formed can be considered as collocations.  For example, 
during a search of the Brown Corpus, Kjellmer found the following sequences: 
'green ideas', 'try to', 'hall to'.  From these strings, it is only 'hall to' and 'try to' 
that recur, and from these two, only 'try to' that is grammatically well-formed.  
Therefore, only 'try to' is a collocation (Kjellmer 1984:163).  Kjellmer also tries to 
establish a set of rules for assessing 'collocational distinctiveness'.  According to 
these, a sequence is highly distinctive when it appears frequently in many and 
 
154

---

different categories of texts; it is long (minimum length is two words); and it is 
structurally complex.  
 
On the other hand, Renouf and Sinclair (1991) applied their theory of 
studying collocations to 'frameworks' consisting of discontinuous sequences of 
two words, whose grammatical well-formedness depends on what intervenes, 
e.g. 'a + ? + of', 'too + ? + to' (Renouf & Sinclair 1991:128).  They found out that 
in some cases there seems to be a stronger collocational pull exerted by one of 
the pair on some items rather than on others, e.g. in the framework 'too + ? +to', 
'to' would be able to collocate with 'easy', 'hard', 'good' and 'proud' even in the 
absence of 'too', e.g. ‘easy to do’, ‘good to do’, but not with 'much' or 'tired' 
which require the presence of 'too', e.g. ‘too tired to dance’, ‘too full to eat’, 
(Renouf & Sinclair 1991:133).  Thus, Renouf and Sinclair demonstrated that the 
collocations of grammatical words offer an appropriate basis for studying 
collocations, since "co-occurrences in the language most commonly occur 
among grammatical words" (Renouf & Sinclair 1991:128). 
 
The importance of grammatical words for the study of collocations was 
also confirmed by Jones and Sinclair (1974).  Even though their study on 
English lexical collocations was based on a relatively small corpus (147,000 
running words), it yielded some interesting results concerning the study of 
collocation: the influence of the node does not extend beyond span position 
Node (N) + 4 (see also Berry-Rogghe 1973).  Grammatical words are not 
collocationally neutral (unlike Haskel 1971).  Even though grammatical words 
are weak at predicting their environment, they do show ability to predict word 
 
155

---

classes at specific span positions, e.g. the collocates of the word 'the' in position 
N-1 are mainly verbs and prepositions, while in position N+1 they are nouns 
and adjectives.  The significance of a collocation takes into account the overall 
frequency of the two items concerned, the number of times they occur together, 
and the length of the text.  Collocations can appear to be 'text dependent'.  
Verbs tend to collocate with grammatical items, e.g. 'put' and 'take' collocate 
with a great number of prepositions to form phrasal verbs.  Association 
between lexical items is subject to grammatical influence, e.g. the adjective 
'good' is preceded by adverbs and followed by nouns as significant collocates.  
Significant collocations show a considerable amount of position dependence, 
e.g. in a span of 4, significant collocations most frequently occur in the span 
positions immediately next to the node, N-1 and N+1, while very little occurs at 
the two extremes of the span, N-4 and N+4.  Finally, collocation was found to 
be an organising principle that influences the construction and interpretation of 
utterances (Jones & Sinclair 1974:48; Leitner 1992). 
 
The study of collocations in structural patterns was also suggested by 
Aisenstadt (1979).  Aisenstadt distinguishes collocability restrictions as part of 
the wide field of collocability.  Word combinations whose constituents are 
restricted in their 'commutability', i.e. their ability to combine with other words, 
are called restricted collocations (Aisenstadt 1979:71).  Restricted collocations 
are defined as combinations of two or more words used in one of their regular, 
non-idiomatic meanings, following certain structural patterns (e.g. 
V+(art)+(A)+N), and restricted in their commutability not only by grammatical 
 
156

---

and semantic valency (e.g. in the restricted collocation 'shrug one's shoulders' 
both components have a narrow semantic valency), but also by usage (e.g. we 
can 'bear a grudge' but we cannot *'bear hatred/ scorn') (Aisenstadt 1979:71, 
1981:54).  Restricted collocations are different from free word-combinations.  
For example 'carry' can enter a large number of free word-combinations when it 
means 'to support the weight of something' like 'carry a 
book/bag/chair/torch/table/etc.', but it may also enter a restricted collocation 
pattern 'carry conviction', 'carry persuasion', 'carry weight' when it is used to 
denote 'being convincing' or 'winning the argument' (Aisenstadt 1979:72).  
Some of the structural patterns of restricted collocations in English listed by 
Aisenstadt are given below in Table 2: