Phraseology and Culture in English

Doris Schönefeld 
primitive in the sense of not being characterizable in terms of other
domains still more basic”, or, in more general terms, “cognitively irreduci-
ble representational spaces or fields of conceptual potential.” (Langacker 
1991b: 4). Palmer (1996: 46) lists temperature as one type of imagery / im-
ages along with visual, auditory, kinesthetic, olfactory, and affective im-
agery of feeling states, with images being understood as “mental represen-
tations that begin as conceptual analogs of immediate, perceptual experi-
ence from the peripheral sensory organs. ... they are also, therefore, indirect 
conceptual analogs of the environment” (ibid: 47). This puts temperature in 
a close relation to what Johnson (1987: 28) calls “embodied” or “image 
schemas”.
5
Due to the basicness of temperature sensation to human beings, I ex-
pected the default sense to be the same in all three languages. This implies 
also similarity in the sensation of the temperature denoted by 
HOT
as less or 
more comfortable, a fact which is guided by what object sends off the heat 
and in what situation. Compare such examples as hot fat vs. hot sand vs. 
hot bath vs. hot chocolate to illustrate this point. 
Other senses, bound to other mental models, must be induced by the 
contextual (and co-textual) embedding of the word, which means that we 
understand
HOT
in one of its other senses only when there is a contextual 
clue to the respective base / underlying frame. 
The dependence of the selection of a word’s sense on particular frames 
might perhaps escape the language user’s attention, since we always use 
language in particular situations or contexts and – more often than not – are 
unaware of the guidance context gives for the interpretation of an expres-
sion’s meaning. The need for contextual embedding will, however, catch 
the eye of the dictionary user, since there, sense specification is usually (at 
least in dictionaries of a later date) signalled by an explicit reference to the 
underlying model by means of context words. Thus, E hot is defined in a 
dictionary (OALD 1992: 438) as follows “... 1(a) having a relatively or 
noticeably high temperature; giving off heat. ...” The entry contains another 
nine subentries, each giving another sense the word may take when it co-
occurs with particular nouns, such as spices, news etc. These nouns are the 
clues to the frames in which the word at issue has a meaning deviant from 
the default reading. 
It might be asked here why I then, in my argument, do not simply make 
use of the information provided by dictionaries on the senses of the word/s 
under analysis. One point against that is that I am interested in the senses 
that can be identified as occurring in actual language use, and secondly – as 

---

Hot, heiß, and gorjachij
143
will become evident in the data analyses – I think that dictionaries follow a 
strategy of sense differentiation that is meant to give their users the most 
direct access to the meaning of a word. As a result of this, commonalities 
between senses may however be overlooked, just as well as the registered 
senses may turn out to be contextual interpretations of one, though more 
general, common sense. 
From my set of data, the “non-default” readings or senses of 
HOT
are the 
first point to consider in my analysis, since cross-linguistic differences, if 
there are any, can be expected to show up here. 
Recall that the base evoked by an expression is also understood in the 
sense of ‘domain’ (cf. note 2), and it is only a small step to recognizing the 
principle underlying the extensions from the default sense: it is metaphor, a 
conceptual process that is understood as a cross-domain mapping of an ex-
pression’s profile, so that one (usually more abstract) concept can be under-
stood in terms of another (usually more concrete one).
6
As cognitive-lin-
guistic research has amply demonstrated (cf. Gibbs 1994; Lakoff 1987; 
Gibbs 1999; Cienky 1999; Kövecses 1999, 2000; to name just a few), it is 
here that a speech community’s thought patterns leave noticeable traces: 
metaphor is culture-specific in that, for example, the metaphors found to be 
employed in the conceptualization of particular (more or less abstract) con-
cepts may differ in different speech communities, thus highlighting particu-
lar aspects of these concepts, or in that conceptual metaphors can be spe-
cially elaborated in one or the other culture, whereas others are scarcely 
employed.
7
In other words, 
HOT
(as source domain language), like any 
other metaphorically used expression, may take part in different metaphoric 
mappings, or it may be found to occur in more or less variable expressions 
reflecting one and the same conceptual metaphor, and thus exhibit cultur-
ally motivated cross-linguistic differences. 
The clues to the conceptual metaphorical mappings are usually found in 
the linguistic expressions reflecting them. Metaphorically used words re-
veal the source domain of the metaphor by their literal meaning, and the 
linguistic environment will give hints at the target domain of the mapping.
8
Thus it makes sense to have a look at a source-domain word (in my case 
HOT
) in its typical linguistic environment, that is at collocations of 
HOT
.
From the perspective of collocational analysis, it has been shown that 
polysemous senses of one word usually go with specific contexts, i.e. co-
occurring words (cf. Biber et al. 2000: 26f; Moon 1998: 189, for example). 
That means that the words a node word attracts in a collocation, i.e. a node 
word’s collocates, differ depending on the sense it is meant to render. For 

---

144
Doris Schönefeld 
my analysis, it follows that the collocates of polysemous 
HOT
will reveal its 
intended sense, which – if not the literal one – is an extension by metaphor 
or metonymy. To close the argument, if speech communities employ differ-
ent metaphors (as an effect of conventionalised ways of construing phe-
nomena and events), these will show in and simultaneously motivate the 
occurrence of different collocations. 
In what follows, I will present the corpus data which are meant to sup-
port the claims just made. 
3.2. Corpus data 
The analyses were carried out in a parallel way for all three languages.
9
The 
corpora available were searched for the words at issue, with the (compati-
ble) concordance programmes producing lists of KWIC (key-word-in-con-
text)-concordances, which could then be processed in various ways. The 
first ordering principle applied was the search word’s grammatical func-
tions, so that the data output could be analysed separately for 
1. attributive 
usage 
2. 
predicative usage, and 
3. adverbial 
usage 
In the course of the compilation of the data, it turned out to be useful to 
have a further category, cross-cutting the other three: the usage of 
HOT
in
4. 
idioms (as the most invariable and opaque type of collocation). 
3.2.1. Hot in English usage 
The corpus used is the British National Corpus: World Edition (December 
2000, SARA Version 0.98), which is a corpus of approximately 100 million 
words of running text of various text types (written (books, periodicals, 
magazines etc.) and spoken). 
Table 1 gives an overview of the overall number of occurrences and the 
functional distribution of hot.

---

Hot, heiß, and gorjachij
145
Table 1. Number of occurrences of hot in the BNC 
Function of hot 
Number of occurrences 
Total 8,733
10
Attributive usage 
3,997 (of a frequency of N > 5) 
Predicative usage 
1,512 (of a frequency of V > 5) 
Function of hotly 
Number of occurrences 
Adverbial usage 
235