Phraseology and Culture in English

particular syntactic categories they attract – a phenomenon that Firth termed 
“colligation” and that is perhaps better known as “grammatical collocation” 
(cf. Benson, Benson & Ilson 1986: x). Thus also such units as consider sth 
Adj (important, wise, worthless...) or consider sb NP (a fool, an expert, a 
traitor...) are members of the category of collocation (for a broad discus-
sion of the term, see Schönefeld 2001: 228–237). 
2.1. Language-specificity 
One feature of collocations, especially obvious to the non-native speaker of 
a language, is their language-specificity, which makes them unpredictable 
from the learner’s perspective. So it is not unusual that the non-native 
speaker transfers the collocational knowledge of his/her mother tongue to 
the language s/he is learning, thereby running the risk of being immediately 
recognized as a “foreigner”. This is because collocations contribute to a lan-
guage’s or rather a language user’s semantic accent, a term that Lucy (2000: 
xiiif) suggests (by analogy with a speaker’s accent in pronunciation) to 
cover the fact that speakers in their encounter of other languages will ex-
pect to find their own semantic categories.
At first sight, the language-specific nature of collocations seems coun-
terintuitive. For the existence of patterns or clusters (both structural and 
lexical) in language use is assumed to “reflect the recurrence of similar situa-
tions in human affairs; ...” (Sinclair 1991: 110), and it can further be as-
sumed that recurrent situations are typically those that are basic and, hence, 
common, or at least similar, to all societies. So, if human beings experience 
similar situations, why should they be induced to understand and verbalize 
them differently? On closer examination, two things will, however, become 
obvious: (a) human affairs differ to quite an extent in different parts of the 
world, and (b) similar or comparable situations are not necessarily experi-
enced, seen and understood in the same way. Thus, although it is true to 
consider the recurrent character of many phenomena, situations, and events 
which we experience and talk about to be the eventual motivation for collo-
cations to arise, this does not imply that people end up with the same kinds 
of collocations. This lack of identity is only too natural for situation (a), 
where people are simply involved in and concerned with different affairs. 
Disparity in situation (b) follows from other considerations: experiencing 

---

Hot, heiß, and gorjachij
139
or recognizing reality as a particular event (state, phenomenon, etc.), i.e., 
structuring reality, is not universally given. The recognition of structure in 
the world (or the projection of structure onto the world) is influenced by 
what people already know about the world, by the (mental) models they 
have constructed for the sake of making sense of the world. These models 
are basically of two kinds: personal and cultural, the latter of which I would 
understand to arise from experience shared in and by a community. Shore 
(1996: 47) describes them as conventional mental models that “have been 
externalized as shared institutions as well as internalized by individuals 
[and] ... are a community’s conventional resources for meaning making.” 
They differ from personal models in that they “are constructed as mental 
representations in the same way as any mental models with the important 
exception that the internalization of cultural models is based on more so-
cially constrained experiences than is the case for idiosyncratic models.” 
From this it follows that language-specific, conventional ways of word-
ing – i.e. also collocations – can be considered an indication of specific un-
derlying cultural models, or, conversely, cultural knowledge can be assumed 
to play a part in linguistic usage (cf. also Quinn & Holland 1987: 23f). That 
means that linguistic usage data can be employed as an informant on under-
lying conventional mental models, just as well as these models can be ex-
ploited to account for what people talk about and how they do it. 
From the latter perspective, the language-specificity exhibited by collo-
cations can be attributed to the fact that they do not necessarily arise from 
general (logical) principles of semantic compatibility, but that they emerge 
as a result of experiencing and conceptualizing particular situations (events, 
things and properties) in ways that are culturally determined. 
The interrelation between verbal form and underlying cultural model can 
be seen as a consequence of frame evocation: an expression’s meaning can 
be fully understood only when the understander has recourse to the back-
ground structure against which the concept named has been shaped
2
(cf. 
also Fillmore & Atkins 1992: 75) These frames are always culture-bound, 
with the “cultural load” ranging from considerable (and thus noticeable, 
especially for the non-native speaker) to very subtle.
3
What I call the “cultural load” becomes especially obvious in a language’s 
verbal formulae, since they, as Shore (1996: 58) notes, “encode traditional 
wisdom, specialized knowledge, or techniques in highly conventional forms 
of speech. Examples of verbal formulae are proverbs, sayings, traditional 
narratives, prayers, spells, and nursery rhymes.” These types of expressions 
render in a condensed form socially constrained practices and experiences 

---

140
Doris Schönefeld 
of the people speaking the language at issue. Turning to collocations, I first 
have to explicate the coverage of the term. Collocations, as defined above, 
naturally include such fixed expressions as proverbs, sayings, idioms etc., 
but they also include (complex) expressions allowing for some variation, 
though still exhibiting some preference in their make-up. This subgroup of 
collocations gives the phenomenon of collocation its probabilistic flavour: 
words as constituents of a collocation tend to be used together, in a more or 
less rigid form and cannot, therefore, be understood as a language user’s 
free selections on the basis of logical considerations. Apart from the feature 
of (limited) variability, definitions of collocation also draw on that of se-
mantic opaqueness. My reading comprises the whole spectrum: expressions 
whose meaning is opaque (many idioms) and expressions whose meaning 
is “compositional” in that it can be gathered from the meanings of the par-
taking words.
4
The evocation of cultural knowledge (via frames, for exam-
ple) can be understood to be variously strong with different types of collo-
cations: it will be strong and obvious in the case of fixed, formulaic ex-
pressions, such as proverbs and spells, whereas collocations at the more 
“open” end of the scale (i.e. less fixed, less opaque expressions) will proba-
bly make their cultural load less explicit. Still, as my sample analysis will 
show, also these expressions are not culture-free, and a cross-linguistic 
analysis (even one of closely related languages) can help to identify cultural 
traits that might go unnoticed otherwise.