Conceptual metaphor theory

Full text

Conceptual metaphor theory
Zoltán Kövecses
1. INTRODUCTION AND A DEFINITION
Conceptual metaphor theory (CMT) started with George Lakoff and Mark Johnson’s book,
Metaphors We Live By (1980). The theory goes back a long way and builds on centuries of
scholarship that takes metaphor not simply as an ornamental device in language but as a
conceptual tool for structuring, restructuring and even creating reality. Notable philosophers in
this history include, for instance, Friedrich Nietzsche and, and more recently, Max Black. A
recent overview of theories of metaphor can be found in Gibbs, ed. 2008 and that of CMT in
particular in Kövecses 2010a.
Since the publication of Lakoff and Johnson’s (1980) work, a large amount of research
has been conducted that has confirmed, added to and also modified their original ideas. Often,
the sources of the new ideas were Lakoff and Johnson themselves. Given this situation, it is
obvious that what we know as conceptual metaphor theory today is not equivalent to the
theory of metaphor proposed in Metaphors We Live By. Many of the critics of CMT assume,
incorrectly, that CMT equals Metaphors We Live By. For this reason, I will not deal with this
kind criticism in this introduction to CMT.
The standard definition of conceptual metaphors is this: A conceptual metaphor is
understanding one domain of experience (that is typically abstract) in terms of another (that
is typically concrete). This definition captures conceptual metaphors both as a process and a
product. The cognitive process of understanding a domain is the process aspect of metaphor,
while the resulting conceptual pattern is the product aspect. In this survey of the theory, I will
not distinguish between the two aspects.
2. MAIN CONCEPTS AND DEVELOPMENT OF CMT
In this section, I attempt to spell out the main features of CMT, as I see them. Other
researchers might emphasize different properties of the theory. At the same time, I tried to
select those features on which there is some agreement among practitioners of CMT.
2.1. Metaphors are all-pervasive
In their Metaphors We Live By, Lakoff and Johnson (1980) suggested that metaphors are
pervasive not only in certain genres striving to create some artistic effect (such as literature)
but also in the most neutral, i.e., most non-deliberately used forms of language. CMT
researchers, especially in the early stages of work on conceptual metaphors, collected
linguistic metaphors from a variety of different sources: TV and radio broadcasts, dictionaries,
newspapers and magazines, conversations, their own linguistic repertoires, and several others.
They found an abundance of metaphorical examples, such as “defending an argument”,
“exploding with anger”, “building a theory”, “fire in someone’s eyes”, “foundering
relationship”, “a cold personality”, “a step-by-step process”, “digesting an idea”, “people
passing away”, “wandering aimlessly in life”, and literally thousands of others. Most, if not
all, of such linguistic metaphors are part of native speakers’ mental lexicon. They derive from
more basic senses of words and reflect a high degree of polysemy and idiomaticity in the
structure of the mental lexicon. The magnitude of such cases of polysemy and idiomaticity in
the lexicon was taken to be evidence of the pervasiveness of metaphor. Based on such
examples, they proposed what came to be known as “conceptual metaphors.” However, CMT
does not claim that each and every metaphor we find in discourse belongs to a particular
conceptual metaphor.

Other researchers, however, find the presence of metaphor in real discourse less
pervasive. As noted by Gibbs (2009), different methods produce different results in frequency
counts of metaphors.
2.2. Systematic mappings between two conceptual domains
The standard definition of conceptual metaphors we saw in section 1 can be reformulated
somewhat more technically as follows: A conceptual metaphor is a systematic set of
correspondences between two domains of experience. This is what “understanding one
domain in terms of another” means. Another term that is frequently used in the literature for
“correspondence” is “mapping”. This is because certain elements and the relations between
them are said to be mapped from one domain, the “source domain”, onto the other domain,
the “target.” Let us illustrate how the correspondences, or mappings, work with the conceptual
metaphor ANGER IS FIRE. Before I provide the systematic conceptual mappings that constitute
this metaphor, let us see some linguistic metaphors, as derived by the lexical method, that
make the conceptual metaphor manifest in English:
That kindled my ire.
Those were inflammatory remarks.
Smoke was coming out of his ears.
She was burning with anger.
He was spitting fire.
The incident set the people ablaze with anger.
Given such examples, the following set of correspondences, or mappings, can be proposed:
the cause of fire  the cause of anger
causing the fire  causing the anger
the thing on fire  the angry person
the fire  the anger
the intensity of fire  the intensity of anger
With the help of these mappings, we can explain why the metaphorical expressions listed
above mean what they do: why, for instance, kindle and inflammatory mean causing anger,
and why burning, spitting fire, and being ablaze with anger indicate a high intensity of anger,
with probably fine distinctions of intensity between them.
This set of mappings is systematic in the sense that it captures a coherent view of fire
that is mapped onto anger: There is a thing that is not burning. An event happens (cause of
fire) that causes the fire to come into existence. Now the thing is burning. The fire can burn at
various degrees of intensity.
Similarly for anger: There is a person who is not angry. An event happens that causes
the person to become angry. The person is now in the state of anger. The intensity of the anger
is variable.
The mappings bring into correspondence the elements and the relations between the
elements in the fire domain (source) with elements and the relations between the elements in
the anger domain (target). Indeed, it seems reasonable to suggest that, in a sense, the
mappings from the fire domain actually bring about or create a particular conception of anger
relative to the view of fire we have just seen. This is what it means that a particular source
domain is used to conceptualize a particular target domain. (I will come back to this issue
later.)

In many cases, however, the two-domain account does not work and must be
supplemented by a model of explanation that relies on four domains, or spaces (see chapter 3
on conceptual integration and metaphor).
Given the metaphorically used set of elements in a domain, we can derive further
knowledge about these elements, and can also map this additional knowledge onto the target.
This additional kind of source-domain knowledge is often called “metaphorical inference,” or
“metaphorical entailment”. For example, to stay with the metaphor above, in somewhat formal
and old-fashioned English we can find sentences like “He took revenge and that quenched his
anger.” Quenching anger can be regarded as a metaphorical inference, given the ANGER IS FIRE
metaphor. If anger is metaphorically viewed as fire, then we can make use of our further
knowledge of anger-as-fire; namely, that the fire can be quenched. CMT provides an elegant
explanation of such cases of extending conceptual metaphors.
At this point, an important question may arise: Can everything be mapped from one
domain to another? Obviously, not. Given a particular conceptual metaphor, there are many
things that cannot be mapped, or carried over, from the source to the target. For example,
given that THEORIES ARE BUILDINGS, the number of rooms or whether the building has a cellar
or an attic are not mapped. Several explanations have been offered to delimit the amount of
knowledge that can be transferred from the source. One of them is the “invariance hypothesis”
developed by Lakoff (1990). It suggests that everything from the source can be mapped onto
the target that does not conflict with the image-schematic structure of the target. Another is
proposed by Grady (1997), who claims, in essence, that those parts of the source domain can
be mapped that are based on “primary metaphors” (on “primary metaphors”, see section 2.5.).
Finally, Kövecses (2000a, 2002) proposed that the source maps conceptual materials that
belong to its main meaning focus or foci. It should be noted that the three suggestions differ
with respect to which part of a conceptual metaphor they rely on in their predictions
concerning what is mapped. The first one relies primarily on the target, the second on the
connection between source and target, and the third on properties of the source. None of these
are entirely satisfactory.
2.3. From concrete domain to abstract domain
As we just saw, CMT makes a distinction between a “source domain” and a “target domain”.
The source domain is a concrete domain, while the target is an abstract one. In the example
conceptual metaphor LIFE IS A JOURNEY, the domain of journey is much more concrete than
the target domain of life (that is much more abstract); hence, JOURNEY is the source (domain).
In general, CMT proposes that more physical domains typically serve as source domains for
more abstract targets, as in the LIFE IS A JOURNEY metaphor.
This observation is based on the examination of hundreds of conceptual metaphors that
have been discovered and analyzed in the literature so far (such as LIFE IS A JOURNEY, ANGER
IS FIRE, THEORIES ARE BUILDINGS). The assumption that most conceptual metaphors involve
more physical domains as sources and more abstract domains as targets makes a lot of
intuitive sense. For example, the notion of life is hard to pin down because of its complexity,
that of anger is an internal feeling that remains largely hidden from us, that of theory is a
sophisticated mental construct, and so on for other cases. In all of them, a less tangible
and thus less easily accessible target concept is conceptualized as and from the
perspective of a more tangible and thus a more easily accessible source concept.
In our effort to understand the world, it makes a lot more sense to move
conceptually in this particular direction: that is, to conceptualize the cognitively less

easily accessible domains in terms of the more easily accessible ones. Notice how odd
and unintuitive it would be to attempt to conceptualize journeys metaphorically as life,
fire as anger, or buildings as theories. We would not find this way of understanding
journey, fire, or building helpful or revealing, simply because we know a lot more
about them than about such concepts as life, anger, or theory. This is not to say that the
reverse direction of conceptualization never occurs. It may occur, but when it does,
there is always some special poetic, stylistic, aesthetic, and so on, purpose or effect
involved. The default direction of metaphorical conceptualization from more tangible
to less tangible applies to the everyday and unmarked cases.
2.4. Metaphors primarily occur in thought
According to CMT, metaphor resides not only in language but also in thought. We use
metaphors not only to speak about certain aspects of the world but also to think about them.
As we saw above, CMT makes a distinction between linguistic metaphors, i.e., linguistic
expressions used metaphorically, and conceptual metaphors, i.e., certain conceptual patterns
we rely on in our daily living, to think about aspects of the world. For example, metaphors
such as LIFE IS A JOURNEY can actually govern the way we think about life: we can set goals
we want to reach, we do our best to reach those goals, we can make careful plans for the
journey, we can prepare ourselves for facing obstacles along the way, we can draw up
alternative plans in the form of choosing a variety of different paths, we can prefer certain
paths to others, and so on. When we entertain such and similar ideas, we actually think about
life in terms of the LIFE IS A JOURNEY conceptual metaphor. And, consequently, we can use
the language of journeys to also talk about life.
The idea that we think about a domain in terms of another can actually mean several
different things. In one sense, as above, people may be guided by a particular conceptual
metaphor in how they conceive of a domain, such as life. In another, given a conceptual
metaphor, they may utilize some of the implications of a particular domain they rely on (such
as JOURNEY) in a conceptual metaphor and apply those implications to the other domain (such
as LIFE) in their reasoning about it (see below for an example). Finally, it can also mean that in
the course of the online process of producing and understanding a linguistic metaphor, the
metaphor activates both the source and the target concept. (This issue is discussed in chapter
35 on metaphor processing.)
A major consequence of the idea that metaphors are conceptual in nature, i.e., that we
conceive of certain things in metaphorical ways, is that, since our conceptual system governs
how we act in the world, we often act metaphorically.
When we conceptualize an intangible or less tangible domain metaphorically as, and
from the perspective of, a more tangible domain, we create a certain metaphorical reality. We
imagine life one way when we think of it as a journey (see above), and in another way when
we think of it as a theatre play, as reflected in Shakespeare’s famous lines “All the world is a
stage / and all men and women are merely players”. The two source domains result in very
different views on life, and in this sense they create very different realities.
Whenever a new source domain is applied to a particular target, we see the target
domain differently than we saw it before. The limiting case of this situation is the one when a
particular target domain does not exist at all, but by the application of one (or several) source
domain(s), it actually gets created. Very often, the etymologies of words for abstract concepts
reflect this early conceptualization. For example, COMPREHENSION (‘understanding’) is clearly
an abstract concept. Given the UNDERSTANDING IS GRASPING conceptual metaphor (as in “I

did not grasp what he said”, “He is slow on the uptake”), it makes sense that the English word
comprehend derives from the word that means ‘grasp’ in Latin.
This kind of “reality construction” is very common in advertising, where, often,
interesting or amusing cases of metaphorical reality get created. When advertisements for, say,
deodorants promise “24-hour protection”, they make us see a deodorant as our helper or ally
in a fight or war against an enemy. The enemy is no other than our own body odour. So if we
did not think of our body odour as our enemy before, i.e., as something we have to be
protected against, the advertisements can easily make us view it as such. In this manner, the
metaphors used in advertisements and elsewhere can create new realities for us. Such realities
are of course metaphorically defined. But this does not make them unimportant for the way
we live. If we think of our body odour as something we need to be protected against and as a
result go and buy a deodorant to overcome the enemy, we are clearly thinking and acting
according to a metaphorically-defined reality. This is a further example of how the
implications of a source domain for a particular target can be utilized (in a process I called
metaphorical inference or entailment above).
Finally, if metaphor is part of the conceptual system, it follows that conceptual
metaphors will also occur in any mode of expression of that system. Research indicates that
the conceptual metaphors identified in language also occur in gestures, visual representations
(such as cartoons), visual arts (such as painting), and others. This does not mean that the
metaphors found in these modes of expression are exactly the same as found in everyday
language and thought, but that a large number of them are (see, e.g., work by Forceville 2008;
Cienki and Müller 2008).
2.5. Conceptual metaphors are grounded
Why is a particular source domain paired with a particular target domain? The most
traditional answer to this question is to say that there is a similarity, or resemblance, between
two things or events. Several different types of similarity are recognized in the literature:
objectively real similarity (as in the roses on one’s cheeks), perceived similarity, and similarity
in generic-level structure. An example for perceived similarity would be a case where certain
actions in life and their consequences are seen as gambles with a win or lose outcome in a
gambling game; cf. LIFE IS A GAMBLING GAME. We can take as an example for the last type of
similarity the conceptual metaphor HUMAN LIFE CYCLE IS THE LIFE CYCLE OF A PLANT. The
two domains share generic-level structure that can be given as follows: In both domains, there
is an entity that comes into existence, it begins to grow, reaches a point in its development
when it is strongest, then it begins to decline, and finally it goes out of existence. Based on
this shared structure, the plant domain can function as a source domain for the human domain.
In other words, the similarity explains the pairing of this particular source with this particular
target; that is, the metaphor is grounded in similarity – though of a very abstract kind.
In many other cases, however, this explanation does not work: The source cannot be
viewed as similar in any way to the target. CMT offers another explanation or justification for
the emergence of these metaphors as well. Let us take the conceptual metaphor in one of the
metaphor systems we examined in the previous section: INTENSITY IS HEAT. This metaphor is
a generic-level version of a number of conceptual metaphors like ANGER IS FIRE, ENTHUSIASM
IS FIRE, CONFLICT IS FIRE, and so on. The specific concepts share an intensity dimension that
is metaphorically conceptualized as heat. The concept of HEAT bears no resemblance to that of
INTENSITY whatsoever. Heat is a physical property of things that we experience with our
bodies, while intensity is a highly abstract subjective notion (on a par with purpose, difficulty,
or as a matter of fact, similarity). What, then, allows the use of HEAT as a source domain for
INTENSITY? CMT suggests that there is a correlation in experience between intensity and heat.

Often, when we engage in activities at a high intensity (be it physical or emotional), our body
develops body heat. In this sense, intensity is correlated with heat, and this provides the
motivation for the use of HEAT as a source domain for INTENSITY as a target. The generic-level
conceptual metaphor INTENSITY IS HEAT can then be regarded as grounded in a correlation in
experience between a sensory-motor experience and an abstract subjective one.
Conceptual metaphors of this kind are called “primary metaphors” by Lakoff and
Johnson (see, e.g., 1999), who borrowed the term from Joe Grady (1997a, b). Grady proposed
a number of such metaphors in his dissertation (1997a), including SIMILARITY IS CLOSENESS,
PERSISTENCE IS BEING ERECT, and reanalyzed several of the conceptual metaphors in Lakoff
and Johnson’s early work (1980) along the same lines (e.g., MORE IS UP, PURPOSES ARE
DESTINATIONS). He suggested furthermore that several primary metaphors can be put together
to form “compound metaphors.” For example, the PURPOSEFUL LIFE IS A JOURNEY metaphor is
based on the primary metaphors PURPOSES ARE DESTINATIONS, DIFFICULTIES ARE
IMPEDIMENTS, and others.
Many conceptual metaphors (both the similarity-based ones and the primary
metaphors) are based on “image schemas.” These are abstract, preconceptual structures that
emerge from our recurrent experiences of the world (Johnson, 1987; Lakoff, 1987). Such
skeletal preconceptual structures include CONTAINER, SOURCE-PATH-GOAL, FORCE,
VERTICALITY, and several others. For example, the STATES ARE CONTAINERS primary metaphor
derives from the CONTAINER image schema, the LIFE IS A JOURNEY metaphor from the
SOURCE-PATH-GOAL schema, the EMOTIONS ARE FORCES metaphor from the FORCE schema,
and so on.
The research on primary metaphors has intensified the study of metaphors in the brain.
Lakoff (2008) suggested a “neural theory of metaphor”. In it, individual neurons in the brain
form neuronal groups, called “nodes”. There can be different types of neural circuits between
the nodes. In the “mapping circuit” that characterizes metaphor, there are two groups of nodes
corresponding to source and target domain. The circuitry between the two groups of nodes
will correspond to the mappings, or correspondences. In primary metaphors, one group of
nodes represents a sensorimotor experience in the brain, while the other represents an abstract,
subjective experience.
2.6. Provenance of source domains
Since the human body and the brain are predominantly universal, the metaphorical structures
that are based on them will also be predominantly universal. This explains why many
conceptual metaphors, such as KNOWING IS SEEING, can be found in a large number of
genetically unrelated languages. This does not mean, however, that all conceptual metaphors
that are based on primary metaphors will be the same from language/culture to
language/culture. It was recognized early on that the particular culture in which a metaphor
develops is just as significant in shaping the form of the conceptual metaphors in different
languages/cultures as the universal bodily experiences themselves (see, e.g., Taylor and
MacLaury 1995; Yu 1998, 2002; Musolff 2004). Furthermore, several researchers pointed out
that variation in metaphor can also be found within the same language/culture (for a survey of
this research, see, e.g., Kövecses 2005).
As the latest development in this trend, scholars have recognized that it is not only
culture that functions as an important kind of context in shaping the metaphors that emerge.
More and more researchers in this area take into account the tight connection between
metaphorical aspects of our cognitive activities and the varied set of contextual factors that
influence the emergence of metaphors (see, e.g., Cameron 2003; Semino 2008; Goatly 2007;
Gibbs and Cameron 2008; Kövecses 2010b). The overall result is a much richer account of

metaphor. First, it has become possible to account for metaphors that may be completely
everyday but at the same time do not fit any pre-established conceptual metaphors (see, e.g.,
the work by Musolff 2004; Semino 2008; and others). Second, by taking into account the role
of context, we are now in a much better position to see a fuller picture of metaphorical
creativity than before. Indeed, it can be suggested that contextual factors can actually create
novel metaphors that can be referred to as “context-induced” ones (Kövecses 2010b, 2015).
Third, these context-induced metaphors are not limited to the kinds of basic correlations in
experience that form the bases of primary metaphors. Thus, we seem to have a cline of
metaphors, ranging from universal primary metaphors to non-universal context-induced ones.
In other words, metaphors can derive from the body, cultural specificities, and also the more
general context.
3. AN EXAMPLE OF CURRENT RESEARCH IN CMT: INTERLOCKING
METAPHOR HIERARCHIES
As we have seen above, the source domains of conceptual metaphors constitute coherent
organizations of experience, and the mappings from the source onto the target domains create
equally systematically organized target domains. But the question is whether such systematic
source to target mappings are isolated from each other. I suggest that they are likely to belong
to larger, hierarchically organized systems of metaphors.
The principles for the organization of such metaphor systems can be of several distinct
kinds. In one (a), the metaphors are organized in a straightforward hierarchy such that both the
source and the target are specific cases of higher generic-level concepts. In another (b),
different aspects of a given generic-level concept can be differentially conceptualized by
means of conceptual metaphors. In still other cases (c), a single aspect of several different
abstract concepts may organize a large number of subordinated specific-level conceptual
metaphors into a hierarchy. In a fourth (d), the conceptual metaphors form a system because
the target domains are part of an independently existing hierarchy of concepts. In a fifth (e),
what connects the conceptual metaphors and makes them form a system is the fact that a
particular specific level target concept is a special case of a number of different higher-level
concepts that have their own characteristic conceptual metaphors. There are probably
additional ways in which metaphor systems are formed, but for the present purposes it is
sufficient to take these five possibilities into account and briefly describe them.
3.1. Straightforward hierarchies
In this case, both the source and the target are specific-level concepts of generic-level
conceptual metaphors. This is the simplest and most straightforward type of hierarchy, and it
involves a large number of cases. Let us take the well-known ANGER IS A HOT FLUID metaphor.
This is an instance of the generic-level metaphor EMOTIONS ARE FORCES. Actually, the HOT
FLUID source can be further specified, yielding, for example, the concept of STEW as a
potential source domain. We can represent this as follows:
EMOTIONS ARE FORCES
ANGER IS A HOT FLUID IN A CONTAINER (He was boiling with anger.)
ANGER IS A STEW (He was stewing.)
We can find the same situation for love:
EMOTIONS ARE FORCES
LOVE IS A NATURAL FORCE (I was overcome with love.)
LOVE IS THE WIND (It was a whirlwind romance.)
3.2. Different aspects of a single generic concept
What is known as the Event Structure metaphor system presents a more complicated situation
(see Lakoff 1993). Events in general (i.e., the generic-level concept of event) can be actions

and occurrences, and they both involve states, causes and changes. Actions also include long-
term activities, where progress is an issue. Actions are characterized by purposes, potential
difficulty in execution, and manner of performance. These various aspects of events (EVENT 1)
are conceptualized in different ways:
Events: EVENTS ARE MOVEMENTS
Occurrences: OCCURRENCES ARE MOVEMENTS (What’s going on here?)
Actions: ACTIONS ARE SELF-PROPELLED MOVEMENTS (What’s going to be the next
step?)
Cause: CAUSES ARE FORCES (You’re driving me nuts.)
State: STATES ARE LOCATIONS / BOUNDED REGIONS (She’s in love.)
Change: CHANGE IS MOTION (FROM ONE LOCATION TO ANOTHER) (I almost
went crazy.)
Actions: ACTIONS ARE SELF-PROPELLED MOVEMENTS
Purpose: PURPOSES ARE DESTINATIONS (I want to reach my goals.)
Difficulty: DIFFICULTIES ARE IMPEDIMENTS (TO MOTION) (Let’s get around this
problem.)
Manner: MANNER IS PATH (OF MOTION) (We’ll do it in another way.)
Activity: LONG-TERM PURPOSEFUL ACTIVITIES ARE JOURNEYS (We have
a long way to go with this project.)
Progress: EXPECTED PROGRESS IS A TRAVEL SCHEDULE (I am
way behind schedule.)
As we can see, the highest-level metaphor here is related to the overarching category of
events: EVENTS ARE MOVEMENTS. Events come in several forms, and they are characterised by
a variety of different aspects. The various forms and aspects of events are in turn
metaphorically viewed in terms of the source domains of movement, location and force. These
can of course be further elaborated at still more specific levels of concepts.
3.3. A single aspect of several different specific-level concepts
Several conceptual metaphors may belong together by virtue of the fact that they share a
particular aspect that is conceptualized metaphorically by means of the same source domain.
The target domains to which a single source domain is applied is the “scope of a source
domain” (Kövecses, 2000a, 2010a). Thus, the scope of a source can be narrow or wide.
Consider the following conceptual metaphors:
ANGER IS FIRE (He was smouldering with anger.)
LOVE IS FIRE (The fire was gone from their relationship.)
DESIRE IS FIRE (It was his burning ambition to become a lawyer.)
IMAGINATION IS FIRE (The scene set fire to his imagination.)
ENTHUSIASM IS FIRE (He lost the fire.)
CONFLICT IS FIRE (The fire of war burnt down Europe several times in the course of its
history.1)
ENERGY IS FIRE (She’s burning the candle at both ends.)
All of these target domains share the aspect of (degrees of) intensity through the application
of a single source (HEAT OF FIRE). We can suggest that the FIRE source domain has the “main
meaning focus” of intensity (Kövecses, 2000a, 2010a). Thus, one way of metaphorically
understanding intensity is in terms of the heat of fire. This yields the generic metaphor
1 It might be worth mentioning in connection with this example that it has both a literal and a metaphorical
interpretation. Clearly, it is the latter that is intended here.

INTENSITY IS HEAT. Consequently, the specific metaphors above are instances of this generic-
level metaphor. This is a further way in which conceptual metaphors may form a hierarchical
system. As a matter of fact, primary metaphors (see 2.5.) can be seen as forming such systems
in a natural way, since their target domains represent shared aspects (like intensity) of several
different concepts.
3.4. Several different aspects of a single specific-level concept
A specific-level abstract concept may inherit conceptual metaphors from several different
generic-level metaphor systems by virtue of the fact that its prototypical cognitive-cultural
model consists of elements that belong to the different metaphor systems. We can exemplify
this with the specific-level abstract concept of friendship (Kövecses 1995). The model of
friendship conceptually partakes of a number of different metaphor systems. Since according
to the cognitive-cultural model of friendship,
it is a state that two people attribute to each other,
it involves communication between the friends,
it implies mutual interaction with each other,
it consists of the friends and their interactions as a complex system,
it includes participants that feel certain emotions towards each other,
and some others,
the conceptual metaphors that characterize friendship include the following:
State metaphor system:
STATES ARE OBJECTS
ATTRIBUTED STATES ARE POSSESSED OBJECTS (Lakoff 1993)
Communication metaphor system:
THE MIND IS A CONTAINER
LINGUISTIC EXPRESSIONS ARE CONTAINERS
MEANINGS ARE OBJECTS
COMMUNICATION IS SENDING (Reddy 1979)
Interaction metaphor system:
INTERACTIONS ARE ECONOMIC EXCHANGES (Kövecses 1995)
Complex system metaphor system:
ABSTRACT COMPLEX SYSTEMS ARE COMPLEX PHYSICAL SYSTEMS (Kövecses 1995,
2010a)
Emotion metaphor system:
EMOTION IS DISTANCE
EMOTION IS TEMPERATURE (Kövecses 1990, 2000b)
The conceptual metaphors for friendship emerge from these various metaphor systems.
Specifically, we find metaphors such as the following in the descriptions of friendship:
State metaphor system:
FRIENDSHIP IS A POSSESSED OBJECT (My friendship with her did not last long.)
Communication metaphor system:
SHARING (COMMUNICATING) EXPERIENCES IS SHARING OBJECTS: (We share intimate
things with each other.) (Kövecses 1995, 2000b)
The metaphor arises because communication between friends often involves sharing ideas and
feelings.
Interaction metaphor system:
INTERACTIONS IN FRIENDSHIP ARE ECONOMIC EXCHANGES (There is a lot of give and
take in our friendship.) (Kövecses 1995)

The interactions are conceptualized as “economic” exchanges because people often mention a
fifty-fifty basis in their friendship interactions, which indicates not just a physical exchange of
objects.
Complex system metaphor system:
FRIENDSHIP IS A COMPLEX PHYSICAL SYSTEM (BUILDING, MACHINE, PLANT) (We have
built a strong friendship over the years.) (Kövecses 1995, 2010a)
Emotion metaphor system:
AN EMOTIONAL RELATIONSHIP IS A DISTANCE (They have a close friendship.)
EMOTION IS TEMPERATURE (They have a warm friendship.) (Kövecses 1990, 2000b)
The last two conceptual metaphors have to do with the notion of intimacy that characterize
several emotions, yielding the metaphors INTIMACY IS CLOSENESS and INTIMACY IS WARMTH
(both of which are primary metaphors).
More generally, since aspects of friendship constitute a part of these metaphor systems, the
hybrid concept of friendship will share with them the specific metaphors.
3.5. The target concepts form a hierarchical system of concepts
The best example for this kind of metaphor system is what is called the Great Chain of Being
(Lakoff and Turner 1989). This is a hierarchical system of concepts corresponding to objects
and entities in the world, such as humans, animals, physical things, and so on. The extended
version of this hierarchy consists of the following (Lakoff and Turner 1989; Kövecses 2010a):
God
Complex systems (universe, society, mind, theories, company, friendship, etc.)
Humans
Animals
Plants
Complex physical objects
Inanimate objects
The hierarchy becomes a metaphor system when things on a particular level are
conceptualized as things on another level. Notice that this can happen in both directions.
Lower-level concepts can function as source domains for higher-level ones as target (e.g.,
PEOPLE ARE ANIMALS) and higher-level ones can function as source domains for lower-level
ones as target (e.g., ANIMALS ARE PEOPLE). Furthermore, the HUMAN, ANIMAL and PLANT
categories are often graded internally – a conceptualization that can lead to racist language
(e.g., an “inferior race”).
In summary, conceptual metaphors are not isolated conceptual patterns in the mind but
seem to cluster together to form a variety of interlocking hierarchical relationships with each
other.
4. CRITIQUE OF CMT
In the world of academia, CMT is in a curious situation: despite its many undeniable
achievements, its obvious usefulness in and popularity across several disciplines, each and
every aspect of it has come under criticism in the past thirty years. Indeed, several scholars
have expressed their skepticism regarding the very existence of conceptual metaphors (e.g.,
Cameron and Maslen, eds., 2010).
A further curious aspect of the situation is that a considerable body of the criticism is
based on Lakoff and Johnson’s (1980) work exclusively, which represents only the initial stage
of CMT, ignoring much of the later work in CMT. Since this chapter has described, or at least
briefly mentioned, some of that work subsequent to Metaphors We Live By, I will not take up
the criticisms that relate to these features of CMT. I will not discuss issues regarding the

processing of conceptual metaphors either, since these are described in another chapter of the
present volume (see chapter 35).
A charge that is sometimes levelled at CMT is that it works with the concept of
domain (as in the idea that conceptual metaphors involve two domains) and that is itself not a
well-defined concept and that it probably cannot be defined precisely at all. But, as a matter of
fact, CMT works with a fairly clear definition of a domain that goes back to Fillmore’s
definition of a frame: A domain, or frame, is a coherent organization of human experience.
This definition makes do in most cases.
Another criticism maintains that CMT is based on circular reasoning. Here the claim is
that scholars in CMT use linguistic metaphors to identify conceptual metaphors, on the one
hand, and that at the same time they suggest that the linguistic metaphors exist because of the
already present conceptual ones, on the other. One cannot base the existence on conceptual
metaphors on linguistic metaphors and at the same time explain the presence of linguistic
metaphors on the basis of conceptual metaphors. However, this criticism ceased to be valid
after several experiments that did not involve language or linguistic metaphors (beginning
with Gibbs’ work in the early 1990s) unambiguously confirmed the existence of conceptual
metaphors. If conceptual metaphors have been proven to have psychological reality by
psycholinguistic experiments, linguists should not deny their existence; they should work to
see how they appear and function in language (and other modalities). (For summaries of these
experiments, see Gibbs 1994, 2006; Gibbs and Colston 2012.)
But the most commonly and strongly expressed criticism concerns methodological
issues; namely, how to identify metaphors in discourse, how the study of metaphor should be
based on real data (rather than just lexical or intuitive data), and others (see, e.g., Deignan
2005; Steen, et al. 2007). As I indicated above, we should now take these developments as an
integral part of CMT. However, the issue of the need to use of real data for metaphor analysis
reveals an apparently real weakness of CMT: it is that CMT researchers do not pay sufficient
attention to the discourse and social-pragmatic functions of metaphor in real discourse. This
sounds like a valid point. However, I do not think that CMT should be thought of as a view of
metaphor whose only job is to collect metaphorical expressions, set up conceptual metaphors
based on the expressions, lay out the mappings that constitute those conceptual metaphors,
and see how the particular conceptual metaphors form larger systematic groups. A large
further part of the mission of conceptual metaphors is to describe the particular syntactic,
discursive, social, pragmatic, rhetorical, aesthetic, etc. behavior and function of the metaphors
in real data. And this is precisely something that is currently conducted by a great number of
researchers (e.g., Low, et al. eds. 2010). But, to my mind, these researchers are not competing
with more “traditional” CMT scholars; instead, they are working out an aspect of CMT that
was “neglected” by CMT scholars. The addition is necessary and more than welcome. This
kind of work is just as much part of CMT as other aspects of the theory. In other words, I find
that the “neglect” was not really neglect. The lack of sufficient attention to the syntactic,
pragmatic, etc. features of metaphors resulted from CMT scholars’ effort to add a cognitive
dimension to metaphor that was mostly lacking in previous work. This was, and still is, the
mission of CMT, in collaboration with other metaphor researchers. Without pursuing that
mission, we would know much less about metaphor today.
5. THE PRESENT STATE AND FUTURE DIRECTIONS OF CMT
In my view, CMT is a complex and coherent theory of metaphor. As even the sketchy picture
above reveals, CMT is a theory of metaphor that is capable of explaining a variety of issues
concerning metaphor. In particular, it can explain:

-why we use language from one domain of experience systematically to talk about
another domain of experience;
-why the polysemy of words in the lexicon follows the patterns it does;
-why the senses of words are extended in the concrete-to-abstract direction;
-why children acquire metaphors in the sequence they do;
-why the meanings of words emerge historically in the sequence they do;
-why many conceptual metaphors are near-universal or potentially universal;
-why many other conceptual metaphors are variable cross-culturally and
intraculturally;
-why many conceptual metaphors are shared in a variety of different modes of
expression (verbal and visual);
-why many metaphor-based folk and expert theories of a particular subject matter
are often based on the same conceptual metaphors;
-why so many conceptual metaphors are shared between everyday language and
literature (and other forms of non-everyday uses of language);
-why and how novel metaphors can, and do, constantly emerge;
-etc.
No other theory of metaphor is capable of explaining all of these issues. This does not mean,
however, that CMT has achieved a “state of perfection”, and that is has no room to develop
further. I have pointed out several issues where CMT scholars need to do much more to
explain the facts. One such issue was the discrepancy resulting from making use of different
methodologies in establishing the frequency of metaphors in discourse. Another outstanding
issue that was mentioned is which conceptual materials are carried over from one domain to
another. These are just some of the difficult questions that await answers, but there are
additional ones that need to be answered in the future.
On a more positive note, there are also several new research directions that promise an
even better understanding of metaphor than what we have today. Lakoff and his colleague’s
work on the neural theory of metaphor is one of them (see, e.g., Lakoff 2008). What
complicates research on the neural aspects of metaphor, which is itself extremely complex, is
that metaphor use is taking place in a variety of different types of context that are constantly
monitored by the brain in the course of metaphorical conceptualization. These contextual
factors can be regarded as actually priming the use of particular linguistic metaphors that may
or may not belong to conventional primary or compound conceptual metaphors (see Kövecses
2015). The result is an extremely complex situation that challenges, and calls for the
cooperation of, researchers from a variety of different disciplines, such as neuroscience,
metaphor theory and pragmatics, just to mention a few. This is a research project that will
surely take several years to complete.
Finally, in section 3 we have seen that conceptual metaphors occur not in isolation but
in a variety of different and interlocking hierarchical structures. This poses several challenges
to researchers. First, how do such metaphorical hierarchies emerge in social cognition? And
more specifically, how do they emerge and how are they represented in the brain? Second,
how and on what basis do the users of metaphors select the appropriate level at which they
formulate their metaphors in discourse? Third, how can “context-induced” metaphors be
integrated into such hierarchical systems? Or, possibly, should we suppose a larger system that
would accommodate both the body-based and the non-body-based metaphors? These are just
some of the research issues for the future study of the hierarchical organization of metaphors.
Even more generally, it can be suggested that CMT will continue to play a key role in
the development of cognitive linguistics as a general study of language (as well as several

other disciplines outside linguistics), as we keep discovering its extensive presence at all levels
of linguistic description and its important contribution to connecting mind with the body,
language with culture, body with culture, and language with the brain.
Further readings:
Barcelona, A. (ed) (2000) Metaphor and Metonymy at the Crossroads, Berlin: Mouton.
Cameron, L. and Low, G. (eds) (1999) Researching and Applying Metaphor, Cambridge:
Cambridge University Press.
Gibbs, R.W. (ed) (2008) The Cambridge Handbook of Metaphor and Thought, Cambridge and
New York: Cambridge University Press.
Gibbs, R.W. and Steen, G. (eds) (1999) Metaphor in Cognitive Linguistics, Amsterdam:
Benjamins.
Katz, A.N., Cacciari, C., Gibbs, R.W. and Turner, M. (1998) Figurative Language and
Thought, New York and Oxford: Oxford University Press.
References
Cameron, L. (2003) Metaphor in Educational Discourse, London: Continuum.
Cameron, L. and Maslen, R. (eds) (2010) Research Practice in Applied Linguistics, Social
Sciences and the Humanities, London: Equinox.
Cienki, A. and Müller, C. (2008) ’Metaphor, gesture, and thought,’ In R.W. Gibbs Jr.
(ed.) The Cambridge Handbook of Metaphor and Thought, 483–501.
Cambridge: Cambridge University Press.
Deignan, A. (2005) Metaphor and Corpus Linguistics, Amsterdam: John Benjamins.
Forceville, C. (2008) ’Metaphors in pictures and multimodal representations,’ In R.W.
Gibbs Jr. (ed.), The Cambridge Handbook of Metaphor and Thought, 462–482.
Cambridge: Cambridge University Press.
Gibbs, R.W. (1994) The Poetics of Mind, Cambridge and New York: Cambridge University
Press.
Gibbs, R.W. (2006) Embodiment and Cognitive Science, Cambridge and New York:
Cambridge University Press.
Gibbs, R.W. (2009) ’Why do some people dislike conceptual metaphor theory,’ Journal of
Cognitive Semiotics, Vols. 1-2, 14-36.
Gibbs, R.W. and Coulston, H. (2012) Interpreting Figurative Meaning, Cambridge and New
York: Cambridge University Press.
Gibbs, R.W. and Cameron, L. (2007) ‘Social-cognitive dynamics of metaphor performance’,
Cognitive Systems Research, 9: 64-75.
Goatly, Andrew. (2008). Washing the Brain. Metaphor and Hidden Ideology. Amsterdam:
John Benjamins.
Grady, J. (1997a) ’Foundations of meaning: primary metaphors and primary scenes’ unpublished
Ph.D. diss. Department of Linguistics, University of California at Berkeley.
Grady, J. (1997b) THEORIES ARE BUILDINGS revisited. Cognitive Linguistics 8: 267-290.
Johnson, M. (1987) The Body in the Mind, Chicago: The University of Chicago Press.
Kövecses, Z. (1990) Emotion Concepts, Berlin and New York: Springer-Verlag.
Kövecses, Z. (1995) ‘American friendship and the scope of metaphor’, Cognitive Linguistics
6: 315–346.
Kövecses, Z. (2000a) ’The scope of metaphor’, in A. Barcelona (ed.) Metaphor and
Metonymy at the Crossroads, 79-92. Berlin: Mouton.

Kövecses, Z. (2000b) Metaphor and Emotion, Cambridge and New York: Cambridge
University Press.
Kövecses, Z. (2002) Metaphor. A Practical Introduction, Oxford and New York: Oxford
University Press.
Kövecses, Z. (2005) Metaphor in Culture. Universality and Variation, Cambridge and New
York: Cambridge University Press.
Kövecses, Z. (2010a) Metaphor. A Practical Introduction, 2nd edn, Oxford and New York:
Oxford University Press.
Kövecses, Z. (2010b) ‘A new look at metaphorical creativity in cognitive linguistics’,
Cognitive Linguistics, 21-4: 663-697.
Kövecses, Z. (2015) Where Metaphors Come From. Reconsidering Context in Metaphor,
Oxford and New York: Oxford University Press.
Lakoff, G. (1987) Women, Fire, and Dangerous Things, Chicago: The University of Chicago
Press.
Lakoff, G. (1990) ‘The invariance hypothesis: Is abstract reason based on image schemas?’,
Cognitive Linguistics, 1-1: 39-74.
Lakoff, George. (1993). The contemporary theory of metaphor. In A. Ortony, ed., Metaphor
and Thought. Second edition. 202-251. Cambridge and New York: Cambridge
University Press.
Lakoff, G. (2008) ‘The neural theory of metaphor’, In R.W. Gibbs, (ed.) The Cambridge
Handbook of Metaphor and Thought, 17-38, Cambridge and New York: Cambridge
University Press.
Lakoff, G. and Johnson, M. (1980) Metaphors We Live By, Chicago: The University of
Chicago Press.
Lakoff, G. and Johnson, M. (1999) Philosophy in the Flesh, New York: Basic Books.
Lakoff, G. and Turner, M. (1989) More Than Cool Reason. A Field Guide to Poetic Metaphor,
Chicago: The University of Chicago Press.
Low, G., Todd, Z., Deignan, A. and Cameron, L. (eds) (2010) Researching and Applying
Metaphor in the Real World, Amsterdam: Benjamins.
Musolff, A. (2004) Metaphor and Political Discourse: Analogical Reasoning in Debates
about Europe, Basingstoke, U.K.: Palgrave Macmillan.
Reddy, M. (1979) ‘The conduit metaphor’, in A. Ortony (ed.) Metaphor and Thought,
284–324, Cambridge: Cambridge University Press.
Semino, E. (2008) Metaphor in Discourse, Cambridge: Cambridge University Press.
Steen, G, et al. (2007) ‘MIP: A method for identifying metaphorically used words in
discourse’, Metaphor and Symbol 22-1: 1-39.
Taylor, J.R. and MacLaury, R. (eds) (1995) Language and the Cognitive Construal of the
World, Berlin: Mouton de Gruyter.
Yu, N. (1998) The Contemporary Theory of Metaphor. A Perspective from Chinese,
Amsterdam: John Benjamins.
Yu, N. (2002) ‘Body and emotion: body parts in Chinese expression of emotion’, Pragmatics
& Cognition 10-1: 341-367.