HTML Preview Mixed Metaphor page number 1.
Mixing Metaphors
Mark G. Lee and John A. Barnden
School of Computer Science,
University of Birmingham
Birmingham, B15 2TT
United Kingdom
Abstract
Mixed metaphors have been neglected in recent metaphor research. This paper suggests that such
neglect is short-sighted. Though mixing is a more complex phenomenon than straight metaphors, the
same kinds of reasoning and knowledge structures are required. This paper provides an analysis of both
parallel and serial mixed metaphors within the framework of an AI system which is already capable of
reasoning about straight metaphorical manifestations and argues that the processes underlying mixing
are central to metaphorical meaning. Therefore, any theory of metaphors must be able to account for
mixing.
Introduction
The phenomenon of mixed metaphors has been largely neglected by previous research in metaphor understanding.
This has been due to two prevalent assumptions. First, mixed metaphors are often regarded as examples of (at worst)
pathological language use or (at best) poor style. Secondly, it is clear that the understanding of a mixed metaphor is
more difficult that of a single metaphor, since a mix requires reasoning about several vehicle domains.
In this paper, we wish to argue that the former assumption is wrong: mixed metaphors are common in mundane
everyday discourse and can be understood by hearers without recourse to specialised reasoning. In addition, the sec-
ond assumption is detrimental to long term progress since mixed metaphorical manifestations rely on straight meta-
phors. More specifically, this paper makes the following claim: the reasoning processes and data structures involved
in understanding mixed metaphors are identical to those used in understanding straight metaphors. Therefore, cur-
rent research on metaphor processing should be capable of being extended to deal with mixed phenomena and mixing
can provide valuable insight into the processes underlying straight metaphors.
To this end, this paper describes some initial work done with ATT-Meta [Barnden, 1997] to handle various types
of mixing and reprises an earlier claim for the need for within-vehicle reasoning and the use of conversion rules to fil-
ter the relevant connotations of a particular metaphor.
Mixed metaphors are often regarded as humourous or cases of defective speech. Consider the following patholog-
ical sentence, quoted by Fowler [Fowler, 1908]:
1. “This, as you know, was a burning question; and its unseasonable introduction threw a
chill on the spirits of all our party.”
In example 1, the question is metaphorically “hot”. However, its introduction makes the party’s spirits “cold”. Despite
this contradiction, the sentence can be understood to mean that the question was somehow controversial and its inap-
propriate introduction saddened the emotions of the party members. Fowler criticised such examples as poor style.
However, despite the conflict between “hot” questions and “cold” emotions, the connotation of the sentence can be
easily understood since it alludes to two well-known metaphors, i.e. “DIFFICULT QUESTIONS ARE HOT
OBJECTS” and “SAD EMOTIONS ARE COLD OBJECTS”. Furthermore, it is unlikely that most native speakers
would even consider the disparity of “hot” questions causing “cold” reactions. This is because, in each case what is
mapped is not an instance of temperature change, but a connotation with direct relevance to the tenor domain.
In this paper, we will argue that it is often necessary to do extended reasoning prior to mapping from vehicle to
tenor. Therefore, a capacity for within-vehicle-reasoning is essential and any conversion must also act as a strict filter
to limit the range of metaphorical meaning.
The paper is structured as follows: in Section 2, we will outline and distinguish two key types of mixed metaphor:
serial and parallel. In Section 3, we will briefly outline ATT-Meta and provide an analysis of each type of mixed met-
aphor which our program is capable of dealing with and then in Section 4 extend the discussion to other types of