Available online at www.sciencedirect.

com

ScienceDirect
Journal of Pragmatics 62 (2014) 68--76
www.elsevier.com/locate/pragma

Metaphors in the perspective of argumentation

Cihua Xu, Yicheng Wu *
Centre for the Study of Language and Cognition & Department of Philosophy, Zhejiang University,
No. 148 Tian Mu Shan Road, Hangzhou 310028, China
Received 16 September 2013; received in revised form 17 December 2013; accepted 18 December 2013
Available online 16 February 2014

Abstract
Based on his study of the metaphors used in Chilean parliamentarian media participation, Santibáñez (2010) proposes a novel model
which analyses conceptual metaphors from the perspective of argumentation. Santibáñez’s analysis is significant in the sense that it
reveals the vital role of metaphors in daily argumentation, which inspires us to further explore the relationship between metaphor and
argumentation. In this article, we take issue with some points of Santibáñez’s analysis, pointing out that it does not clearly distinguish
between analogy argumentation and arguing by metaphors, nor does it clearly tell us what role conceptual metaphors play in Toulmin
argumentation model, and what the overall structure of arguing by metaphors is. We then address these questions via the cases in
Santibáñez (2010), with reference to Juthe (2005)’s approach to analogy argumentation, Kövecses (2010)’s analysis of conceptual
metaphor and Toulmin ([1958]2003, [1978]1984) model of argumentation model.
© 2013 Elsevier B.V. All rights reserved.

Keywords: Metaphor; Analogy; Mapping; Argumentation; Toulmin argumentation model

1. Introduction

According to Santibáñez (2010:975), although many studies ‘‘have attempted to collate all important rhetorical and
argumentation concepts and figures, the phenomenon of metaphor has not received sufficient attention and it is difficult to
see real efforts to include the interesting analyses that have been proposed in the cognitive sciences, especially in
linguistics.’’ Based on this point, Santibáñez (2010) studies metaphors in Chilean parliamentarian media participation,
proposing a novel way to analyse conceptual metaphor from an argumentative perspective. Inspired by Lakoff’s works on
metaphor (Lakoff, 2004, 2006a,b, 2008), Santibáñez’s analysis is essentially characterized by the integration of
conceptual metaphor theory and Toulmin argumentation model.
Undoubtedly, Santibáñez’s analysis is significant in the sense that it reveals the vital role of metaphors in daily
argumentation, which inspires us to further explore the relationship between metaphor and argumentation. In this article,
we take issue with some points of Santibáñez’s analysis, pointing out that it does not clearly differentiate analogy
argumentation and arguing by metaphors, nor does it clearly tell us what role conceptual metaphors play in Toulmin
argumentation model, and what the overall structure of arguing by metaphors is. We attempt to address these questions
via the cases in Santibáñez (2010), with particular reference to the Juthe (2005)’s proposal on analogy argumentation,
Kövecses (2010)’s analysis of conceptual metaphor and Toulmin ([1958]2003, [1978]1984)’s model of argumentation.
The article is organized as follows: Section 2 discusses the distinction between analogy argumentation and arguing by
metaphors. Section 3 addresses the question of how to define conceptual metaphors. Section 4 analyses the conceptual
metaphors in argumentation schema. Section 5 is a summary of our main points.

* Corresponding author. Tel.: +86 571 8827 3089.

E-mail addresses: xuch@zju.edu.cn (C. Xu), wuyicheng@zju.edu.cn (Y. Wu).

0378-2166/$ -- see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.pragma.2013.12.004
 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76 69

2. Analogy argumentation vs. arguing by metaphors

Under the title ‘‘Analogy argumentation vs. arguing by metaphors’’, Santibáñez (2010) makes a distinction between
them, claiming that ‘‘[I]t seems that in contemporary argumentation theory the phenomenon of metaphor has been
implicitly investigated in the study of argument from analogy or analogy argumentation. It is generally agreed that analogy
argumentation is based on ‘similarity’’’ (p. 975). This conclusion is reached on the basis of van Eemeren et al. (2007) and
Walton (2006)’s definition of analogy argumentation and schema representation. Walton (2006:96), for example, states
that ‘‘argumentation from analogy is a very commonly used kind of case-based reasoning, where one case is held to be
similar to another case in a particular respect. Since the one case is held to have a certain property, the other case, it is
concluded, also has the same property’’. In the same vein, van Eemeren et al. (2007:138) state that ‘‘characteristic of
analogy argumentation is that in this type of argumentation someone tries to convince someone else by showing that
something is similar to something else.’’ From these words, we can see that the argumentation models defined by Walton
(2006) and van Eemeren et al. (2007) are quite similar to each other. As a consequence, Santibáñez (2010:976)
concludes that ‘‘beyond the specific forms and terms used, both approaches agree in assuming that analogy
argumentation proceeds through similarity and comparison.’’ Furthermore, Santibáñez spells out the real difference
between analogy argumentation and arguing by metaphors, namely, ‘‘in analogy argumentation similarity is a core
dimension, in metaphor it is a secondary aspect’’ (Santibáñez, 2010:976). What is, then, the core dimension in
metaphors? Santibáñez (2010:977), following the ideas of Lakoff and Johnson (1980), states that ‘‘metaphor is not
primarily related to language, but rather to the ‘mapping’ between domains.’’ That is to say, Santibáñez (2010) holds the
stance that the core dimension of analogy argumentation is similarity, whereas the core dimension of arguing by metaphor
is mapping.
We find Santibáñez’s distinction somewhat ineffective. First, ‘similarity’ is a vague and sweeping concept. If the
terminology is used to mean that two objects are similar in many attributes or their internal structures, it would be rather
difficult to distinguish similarity and mapping in nature. Toulmin et al. (1984:216) state that ‘‘in ‘arguing from analogy’, we
assume that there are enough similarities between two things to support the claim that what is true of one is also true of the
other.’’ Santibáñez (2010) intends to use this statement to prove that Toulmin et al. (1984) also believe that analogy
argumentation is grounded on similarity. However, we should notice that Toulmin et al. (1984) use the term ‘similarity’ in its
plural form, which suggests that they have sensed that analogy should be built on possibly mulptile similarities in attribute
and relation between two objects. When discussing the usage of analogy argumentation, Toulmin et al. (1984) employ the
case of Darwin proposing evolution theory via analogy as a good example to show the importance of analogy
argumentation. ‘‘In this case, the appeal to analogy was successful; the points of significant similarity far outweighed the
differences’’ (Toulmin et al., 1984:162). They further claim:
Domestic animals change because farmers select out their breeding stock for desired characteristics, and
presumably something similar accounts for the variation in natural species. Nature, or the environment, must
‘‘select’’ certain members of the species in preference to others, and this selection improves the adaptation of the
species to the environment (Toulmin et al., 1984:162).
Although Toulmin et al. (1984) use the word ‘similar’, it is used to describe parallels in crucial features and internal
relations between two different objects. This kind of parallel relation is akin to the mapping’s connotation in nature.
Therefore, we can use the following mapping diagram to represent the relationship reflected by the word ‘similar’ in
Toulmin et al. (1984):

farmers => nature

breeding stock => certain member of species
changes of domestic animal => adaptation of species

Second, analogy argumentation is based on ‘mapping’, and its basic structure consists of Target-Subject (TS),
Analogue (A), Assigned-Predicate (AP), elements of the Analogue (ε
1 . . . ε
2 ), elements of the Target-Subject (ε1. . .ε2), and
so on (Juthe, 2005:4). Target-Subject (TS) refers to the object that analogy argumentation intends to describe. Analogue
(A) means the object used to be compared with Target-Subject. Assigned-Predicate (AP) is the predicate designated to
Target-Subject derived from Analogue. Elements of the analogue (ε
1 . . . ε
2 ) which determine the Assigned-Predicate
correspond one-to-one with counterpart elements in the Target-Subject (ε1. . .ε2). Accordingly, Juthe (2005:5) defines
analogy as follows:
The Analogue is analogous with the Target-Subject with regard to the Assigned-Predicate if and only if there is a
one-to-one correspondence between the elements of the Analogue which determine the Assigned-Predicate and
the elements of the Target-Subject.
70 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76

The elements of the Analogue and the elements of the Target-Subject are in a one-to-one correspondence if, and
only if, every element of the Analogue which determines the Assigned-Predicate has a counterpart element in the
Target-Subject.

An element ε
1 of the Analogue is a counterpart of an element ε1 of the Target-Subject if and only if element ε
1 has
relation R to another element ε
2 in the Analogue and element ε1 has relation R to another element ε2 in the Target-
Subject.
The ‘‘one-to-one correspondence’’ in the above definition is, as a matter of fact, the mapping. Juthe (2005:24) points out
that his definition of analogy is inspired by Steinhart (2001). In The Logic of Metaphor------Analogous Parts of Possible
Worlds, Steinhart (2001:93) directly uses the mapping function fM to represent the one-to-one correspondence relation in
two different cognitive domains. We can see that the ideas in Steinhart (2001) are mainly from Holyoak and Thagard
(1995) and Gentner (1983). Gentner unequivocally states that ‘‘analogy is characterized by the mapping of relations
between objects, rather than attributes of objects, from base to target’’ (Gentner, 1983:168). Based on the above analysis,
we argue that Santibáñez (2010)’s claim (i.e., ‘‘the core dimension of analogy argumentation is similarity, whereas the
core dimension of arguing by metaphor is mapping’’) is untenable.
Since the core dimensions of analogy and metaphor are the same, namely ‘mapping’, another question that naturally
arises is what is the difference between analogy and metaphor, if any? In our view, the ‘‘same-domain-analogy’’ and
‘‘different-domain-analogy’’ (Juthe, 2005:5) can tell the distinction between analogy and metaphor. Juthe (2005:5) defines
them as follows:

(a) A same-domain-analogy is an analogy where not only the relations between the elements of the different objects are
the same but also the elements are from the same domain.
(b) In a different-domain-analogy, the analogy comes only in virtue of having the same relation between the elements of
the different objects; the elements of the two objects belong to wholly different domains.

Starting from the distinction presented above, Juthe (2005:5) further points out that ‘‘the distinction between these
kinds of analogy also explains what metaphors are. If a different-domain-analogy is between domains that are very
distant, then the elements will be very different and the analogy tends to become a metaphor or parable.’’ This is similar to
the claim by Gentner that ‘‘Metaphors are a special kind of analogy in that the source and target domains are always
semantically distant’’ (Holyoak, 2005:120). We can use the following figure to show a continuum with two ends.
The dot in the solid line represents a certain kind of target domain; while the dot in the dotted line represents the
corresponding source domain. At the left end of the continuum, source and target domains are cognitively very close. In
the right part of the continuum, source and target domains’ distance gradually increases. Typical analogy argumentation
tends to be located in the left, whereas typical metaphor tends to be located in the right. From Fig. 1, we can say that there
are relations and differentiations between analogy argumentation and arguing by metaphors. Both of them are in one and
the same continuum, one in the left and the other in the right. Certainly, there are many cases going in between. The
vagueness of them makes it difficult to distinguish and classify them under a universally accepted criterion.
Additionally, Fig. 1 also illustrates another question. In the left part, the probability of grasping their internal relevance is
relatively higher because source and the target domains lie in close cognitive domains. As a result, we can find out
sufficient correspondences to provide support for an inferred conclusion. Whereas in the right part, as the distance
between source and the target domains increases, there are more differences than resemblances between them. As a
consequence, the probability of committing mistakes in arguing by metaphors (which are based on limited mapping
relations) would increase. For instance, the employees in a company intend to quit their jobs when the company is in a
financial crisis. The owner of the company tries hard to convince them by saying ‘‘our company is only a three-year-old

Fig. 1. The continuum of analogy and metaphor.

 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76 71

child. Over the past three years, we have been raising this child together, after having gone through a lot of difficult times.
So, all of you are actually her parents. How can we abandon our child when she is in trouble?’’ Apparently, what the owner
says is an arguing by metaphors, which should be backed by a mapping relation:

child => company

parents => employees
trouble in life => difficulty in business operation
abandon => job-hopping

The target domain of this particular mapping is running a company, whereas the source domain is raising a child.
Running a company and raising a child are two highly different cognitive domains, which means that they are semantically
very distant. This mapping may serve as a needed wake-up call for the employees’ sense of duty, which may then prompt
them to stay with the company. Nevertheless, it could be rather misleading. Running a young company can be
analogously compared to raising a small child. But, strictly speaking, this kind of metaphor is ONLY effective for the owner/
‘parent’ of a company, but NOT for the employees of a company, who are actually ‘babysitters’ of a three-year-old child/
company. Parents/owners would never abandon a child/company at any situation. But employees could have good
reason to leave a company under a specific situation. If a company should be in an economic crisis, job-hopping could be
a reasonable choice. Thus, the owner’s misuse of metaphors argumentatively as such could possibly lead the employees
to make false reasoning which might be followed by a wrong choice (i.e., staying with the company). Therefore, we agree
with Santibáñez (2010)’s claim in his final remarks, that is, we need a good framework to ‘‘evaluate the discourse by
analyzing when and why a conceptural metaphor becomes a fallacious move.’’

3. Confirmation of conceptual metaphor

In the third part, Santibáñez (2010) reiterates some of the claims on metaphors by cognitive linguists like George Lakoff
and Marker Turner. He states that metaphor is a mental process, which occurs before linguistic expressions are
constructed. That is to say, the nature of metaphor is conceptual. Then, Santibáñez (2010:978) summaries six basic steps
to understand metaphors:

(1) obtaining linguistic expressions;

(2) distinguishing conceptual domains;
(3) naming the conceptual metaphor contained;
(4) describing the situational logic of the source domain;
(5) characterizing aspects of the domain source;
(6) putting forwarding correspondences or mappings between the domains.

Santibáñez (2010:979) analyses the following text (see Table 1) via the six steps:

Table 1
Example 1 revisited.

Original Spanish version English translation

Miércoles 16 de mayo 2007 Wednesday 16 May 2007

Según Cardemil, in Chile la ‘‘dueña de casa no manda, no
tiene claro qué decir; los niños llegan a cualquier hora, se
atrasa el almuerzo, no hay plata para la comida, se extravía
el presuesto en materias de gestión que no estaban
consideradas, y el dueño de casa sale a emborracharse,
cegando las posibilidades de futuro de la misma familia’’.
According to Cardemil, in Chile ‘‘the lady of the house does not give
orders, does not know what to say; the kids come home at all hours,
dinner is late, there is no money for food, the budget is wasted on
activities that had not been considered previously, and the husband
goes out to get drunk, wasting the chances of the family for the future’’.

A parliamentarian of the right wing coalition in Chile uses this text to attack the then President Bachelet. According to
Santibáñez (2010:979), the conceptual metaphor for this text is ‘THE PRESIDENTS ARE MOTHERS/FATHERS’, in
which ‘mother’ is the source domain and ‘president’ is the target domain.
We find Santibáñez’s analysis hardly acceptable. According to the analysis model (Kövecses, 2010), the conceptual
metaphor in this text should be ‘NATION IS A FAMILY’, rather than otherwise. ‘THE PRESIDENTS ARE MOTHERS/
FATHERS’ is only one corresponding relation in ‘NATION IS A FAMILY’. To illustrate this point, we can use ‘LOVE IS A
JOURNEY’ as an example. Kövecses (2010:4--9) analyses it via the following steps. First, we should collect and define
the relevant metaphorical expressions, such as ‘Look how far we’ve come’, ‘We’re at a crossroads’, ‘We’ll just have to go
72 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76

our separate ways’, ‘We can’t turn back now’, ‘I don’t think this relationship is going anywhere’, ‘We’re stuck’, ‘It’s been a
long, bumpy road’, ‘This relationship is a dead-end street’, ‘We’re just spinning our wheels’, ‘Our marriage is on the rocks’,
‘We’ve gotten off the track’, etc. Then, we should distinguish source domain and target domain, and name them
accordingly. In the above-mentioned expressions, ‘LOVE’ is the target domain, whereas ‘JOURNEY’ is the source
domain. These metaphorical expressions about ‘LOVE’ show that there is a conceptual metaphor ‘LOVE IS A JOURNEY’
in our conceptual system. Finally, we need to establish the following mapping relations in order to clearly illustrate the
contents of conceptual metaphor:
Source: JOURNEY Target: LOVE
the travelers => the lovers
the vehicle => the love relationship itself
the journey => events in the relationship
the distance covered => the progress made
the obstacles encountered => the difficulties experienced
decisions about which way to go => choices about what to do
the destination of the journey => the goal(s) of the relationship

In what follows, let us reconsider the text analysed by Santibáñez. We can also establish the mapping relations with
reference to the analysis in the fourth part of Santibáñez (2010):

the lady of the house => the president

the kids => people
the dinner => the projects of the Parliament
the use of money => the national budget
the husband => the secretary of the Administration

A question can be raised as to what the name for this mapping relation is. In our view, the name should be ‘NATION IS
A FAMILY’, in which ‘NATION’ is the source domain and ‘FAMILY’ is the target domain. If we adopt Santibáñez’s analysis,
‘THE PRESIDENTS ARE MOTHERS/FATHERS’ should be a conceptual metaphor. We can equally argue that the
following ones, such as ‘the people are the kids’, ‘the projects of the Parliament are the dinner’, ‘the national budget are the
use of money’, ‘the secretary of the Administration are the husband’, can be the conceptual metaphors as well for this text.
Obviously, we cannot accept this analysis in Santibáñez (2010).

4. Conceptual metaphor in argumentation model

In Section 4, Santibáñez (2010:981--982) analyses the case (see Table 1) within the framework of Toulmin
argumentation model. He divides the arguing by metaphors into two interrelated parts. The first part focuses on source
domain:
The basic frame in Fig. 2 is Toulmin’s framework of argumentation. According to Tulmin ([1958]2003, [1978]1984),
‘Claim’ in the figure refers to the goal that one argumentation intends to achieve, and every ‘Claim’ is built on ‘Ground’. In
order to infer ‘Claim’ from ‘Ground’, we need ‘Warrant’ to guarantee that ‘Ground’ can prop up ‘Claim’. When ‘Warrant’ is
questioned, we need ‘Backing’ to strengthen it. ‘Claim’ is classified in terms of strong and weak ones. Modal qualifiers can
mark strength or the weakness of ‘Claim’. At last, the validity of one argumentation is still reliant on the exceptions. Thus,
‘Rebuttal’ includes the exceptions. If there should be any exception, ‘Claim’ would be invalid. In Fig. 2, ‘Claim’ of this
argumentation is ‘the mother is not a good mother’, whereas ‘Ground’ is the metaphorical expression of the
parliamentarian. ‘Claim’ and ‘Ground’ are connected by ‘Warrant’ (i.e., All good mothers have to show authority in the
house to be a good mother.). Conceptual metaphor (‘PRESIDENTS ARE MOTHERS’) is ‘Backing’ to prop up ‘Warrant’.
The second part concentrates on target domain. Santibáñez (2010:981--982) claims that the target domain is
reconstructed through the mapping system in conceptual metaphor, as shown in Fig. 3 below:
Santibáñez (2010:981) argues that an argumentation in target domain can be established based on the above steps.
‘Ground’ in Fig. 3 is derived via mapping. In general, the basic idea of Santibáñez is that we can conclude what we want
through reconstructing a new argumentation in target domain, based on the established argumentation structure in source
domain via the mapping system.
We think that there are two more demerits in Santibáñez’s analysis of the issue concerning arguing by metaphors.
First, in the sub-argumentation in Fig. 2, Santibáñez claims that the conceptual metaphor is ‘Backing’. One cannot help
asking, what is ‘Backing’? According to Toulmin et al. (1984), ‘Warrant’ cannot be convinced at any time; hence, we need
‘Backing’ to prop up it. ‘Backing’ is the presupposition of ‘Warrant’ in argumentation. That is to say, ‘Backing’ props up
 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76 73

Fig. 2. Argument in terms of source domain.

‘Warrant’ and solves the problem of reliability of ‘Warrant’. The arrow between ‘Backing’ and ‘Warrant’ in Fig. 2 tells us that
the conceptual metaphor (‘PRESIDENTS ARE MOTHERS’) in the sense of Santibáñez’s props up ‘All good mothers have
to show authority in the house to be a good mother’. Logically, the former presents the relationship between the concept
‘PRESIDENT’ and the concept ‘MOTHER’; while the latter presents ‘the important features shared by all good mothers’.
The former does not prop up the latter. Based on this analysis, we thus conclude that it is improper to take conceptual
metaphor as ‘Backing’ to prop up ‘Warrant’ in source domain, as Santibáñez (2010) does.

Fig. 3. Argument in terms of target domain by mapping process.

74 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76

Warrant

Farmers => Nature

Breeding stock => Certain members of the species
Changes of domestic animal => Adaptation of Species

Grounds Claim
Domestic breeders succeed in
Environment must act upon
producing improved strains of
animal populations selectively, in
domestic animals-for example,
a way that explains why only the
stronger or bigger ones-by controlling
best-adapted variants within any
the breeding of their animals and
species are able to survive.
selecting out preferred strains.

Fig. 4. Analogy argumentation of Darwin’s evolution theory.

Second, the method of dividing arguing by metaphors into two Toulmin argumentation models plus one mapping
relation in Santibáñez (2010) reveals that the mapping relations in conceptual metaphors can prop up the final conclusion.
However, this method causes great damage to Toulmin argumentation model. If we use Toulmin argumentation model to
analyze arguing by metaphors, we should add a mapping relation outside Toulmin argumentation model, which means
that Toulmin argumentation model is inadequate in analyzing arguing by metaphors. There are some elements beyond
Toulmin argumentation model. Nevertheless, another question remains to be answered, can Toulmin argumentation
model integrate conceptual metaphor and its mapping relations? Our answer is positive, for two reasons. One is that
Toulmin et al. (1984) present an analysis of the analogy argumentation in An Introduction to Reasoning, saying that
Darwin’s evolution theory is a successful analogy. In Section 1, we have demonstrated that the core dimension of analogy
argumentation and arguing by metaphors are both a kind of mapping. Thus, we can transfer ‘Warrant’ represented by
‘similar’ into a structural mapping relation, which can reflect the core dimensions of arguing by metaphors in Toulmin
argumentation model. The structure can be depicted in Fig. 4.
The other reason is that Toulmin et al. (1984) also propose the concept of argumentation chain: ‘‘any argument is liable
to become the starting point for a further argument; this second argument tends to become the starting point for a third
argument, and so on. In this way, arguments become connected together in chains’’ (Toulmin et al., 1984:73).
Based on the above analysis, we can reanalyse the case in Santibáñez (2010) within the framework given in Fig. 5.
Fig. 5 presents an overall structure of arguing by metaphors. It consists of two sub-argumentations. In the first sub-
argumentation as represented in solid line box, ‘Claim (1)’ is ‘The mother is not good’, and ‘Grounds (1)’ is (ε1. . .ε7), which
refers to the 7 items in the ‘Grounds’ of Fig. 2. In the second sub-argumentation as represented in dotted line box, ‘Claim
(2)’ is ‘The president is not good’, which is our final conclusion. The final conclusion takes ‘Claim (1)’ as its ‘Grounds’.

Fig. 5. An overall structure of arguing by metaphors.

 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76 75

(ε
1 . . . ε
7 ) in ‘Warrant (2)’ refers to the (ε1. . .ε7)’s correspondence established via mapping relations in target domain, which
is the ‘Grounds’ in Fig. 3. In the first sub-argumentation, ‘Warrant (1)’ guarantees that we can reasonably infer ‘Claim (1)’
from ‘Grounds (1)’. In the second sub-argumentation, ‘Warrant (2)’ propping up conclusion is a mapping structure, which is
strengthened with the assistance of the conceptual metaphor (‘NATION IS A FAMILY’).
As illustrated by the argumentation structure, we need two vital conditions to infer a reliable conclusion in arguing by
metaphors. The first condition is that we can infer ‘Claim (1)’ from ‘Grounds (1)’; the second condition is that the mapping
relations in ‘Warrant (2)’ should be strong enough to effectively prop up the derivation of a final conclusion. Since the first
condition can be satisfied with ease, the strength of argumentation in arguing by metaphors crucially relies on the
powerfulness of the mapping relations. In the case of Santibáñez (2010), hearer can infer the meaning ‘‘the mother is not
good’’ without trouble from the following statement ‘‘ε1: the lady of the house (mother) does not give orders; ε2: the lady of
the house (mother) does not know what to say; ε3: the kids come home at all hours; ε4: the dinner is late; ε5: there is no
money for food; ε6: the budget gets wasted on activities not considered previously; ε7: the husband goes out to get drunk’’.
This process is the first sub-argumentation in Fig. 5. With respect to the discourse context, speaker can infer that the
words are intended to verbally attack the president at that time. Therefore, we can establish a group of correspondences:
ε1 ¼> ε
1 ; ε2 ¼> ε
2 ; ε3 ¼> ε
3 ; ε4 ¼> ε
4 ; ε5 ¼> ε
5 ; ε6 ¼> ε
6 ; ε7 ¼> ε
7 . This is also a group of mapping relations, whose
strength is built on the correspondences within itself and is backed via a conceptual metaphor: NATION IS A FAMILY.
‘‘NATION IS A FAMILY’’ is a highly conventionalized conceptual metaphor in western culture (Lakoff, 2002). Based on the
correspondences and the conceptual metaphor, it should not be difficult for hearer to reach a conclusion that the president
is not good. If the conversation is in a culture which is in lack of the conceptual metaphor ‘‘NATION IS A FAMILY’’, we
believe that the strength of its argumentation would be greatly weakened without such a powerful backing.
Compared with Figs. 2 and 3, Fig. 5 is obviously advantageous. Apart from explicating the conceptual metaphor, Fig. 5
integrates the mapping relations in arguing by metaphors into Toulmin argumentation model. This method supplements
Toulmin argumentation model in arguing by metaphors. Furthermore, it can assist us to have a better analysis on the issue
of metaphors.

5. Summary and conclusion

In this article we have discussed the demerits of Santibáñez (2010), with a special focus on his analysis of conceptual
metaphors. From the perspective of argumentation, we argue that the real difference between analogy argumentation and
arguing by metaphors lies in the semantic distance between the source and the target domain, rather than that analogy
argumentation is based on the similarities and arguing by metaphors is based on the mapping, as claimed in Santibáñez
(2010). For us, both types of argumentation are based on the mapping. Specifically, we use the method proposed by
Kövecses (2010) to analyse the first case in Santibáñez (2010), pointing out that ‘NATION IS A FAMILY’ is the real
conceptual metaphor for this case, rather than ‘THE PRESIDENTS ARE MOTHERS/FATHERS’. Additionally, we have
established a unified Toulmin argumentation model and have systematically integrated the mapping relations in arguing
by metaphors into this model, based on Toulmin et al. (1984)’s concepts of analogy argumentation and argumentation
chains. Inadequate as Santibáñez (2010)’s analysis is, we should acknowledge his contribution to the study of metaphors
and argumentation. Metaphors’ argumentation doubtless has some positive effects in various fields such as political
discourse, commercial discourse and educational discourse, which can be represented not only by a single-modal
method but by multi-modal method. We thus believe that a refined Santibáñez model can help us to have a better
understanding of the mechanism as well as the value of a wide range of arguing by metaphors.

Acknowledgements

We are grateful to Xiaolong Yang and the Journal of Pragmatics reviewer for their helpful comments on the draft of this
article and their constructive suggestions. We also gratefully acknowledge funding support from National Humanities and
Social Science Foundation of China (No. 11CZX052), Humanities and Social Science Project of the Ministry of Education,
China (09YJCZH109), Zhijiang Youth Scholar Program of Social Sciences in Zhejiang Province of China
(11ZJQN035YB), and the Major Program of National Social Science Foundation of China (No. 11&ZD088).

References

Gentner, Dedre, 1983. Structure-mapping: a theoretical framework for analogy. Cognitive Science 7, 155--170.
Holyoak, Keith J., 2005. Analogy. In: Holyoak, Keith, J., Morrison Robert, G. (Eds.), The Cambridge Handbook of Thinking and Reasoning.
Cambridge University Press, Cambridge.
Holyoak, Keith J., Thagard, Paul, 1995. Mental Leaps: Analogy in Creative Thought. MIT Press, MA.
Juthe, A., 2005. Argument by analogy. Argumentation 19, 1--27.
76 C. Xu, Y. Wu / Journal of Pragmatics 62 (2014) 68--76

Kövecses, Zoltán, 2010. Metaphor: A Practical Introduction. Oxford University Press, New York.
Lakoff, George, 2002. Moral Politics: How Liberals and Conservatives Think. University of Chicago Press, Chicago.
Lakoff, George, 2004. Don’t Think of an Elephant. Know your Values and Frame the Debate. Chelsea Green Publishing, Vermont.
Lakoff, George, 2006a. Thinking Points: Communicating Our American Values and Vision. Farrar, Straus and Giroux, New York.
Lakoff, George, 2006b. Whose Freedom? The Battle over America’s Most Important Idea. Farrar, Straus and Giroux, New York.
Lakoff, George, 2008. The Political Mind: Why You Can’t Understand 21st-century American Politics with an 18th-Century Brain. Viking, New
York.
Lakoff, George, Johnson, Mark, 1980. Metaphors We Live By. University of Chicago Press, Chicago.
Santibáñez, Cristián, 2010. Metaphors and argumentation: The case of Chilean Parliamentarian media participation. Journal of Pragmatics 42,
973--989.
Steinhart, EricC., 2001. The Logic of Metaphor: Analogous Parts of Possible worlds. Kluwer Academic Publisher, Dordrecht.
Toulmin, Stephen, 2003. The Uses of Argumentation. Cambridge University Press, Cambridge.
Toulmin, Stephen, Rieke, Richard, Janik, Allan, 1984. An Introduction to Reasoning. Macmillan Publishing Co., Inc., New York.
van Eemeren, Frans, Houtlosser, Peter, Snoeck Henkemans, Francisca, 2007. Argumentative Indicators in Discourse: A Pragma-Dialectical
Study. Springer, Dordrecht.
Walton, Douglas, 2006. Fundamentals of Critical Argumentation. Cambridge University Press, New York.

Cihua Xu is Associate Professor at the Center for the Study of Language and Cognition (CSLC), Zhejiang University, China. His research
interests include logic, pragmatics and cognitive linguistics.

Yicheng Wu is Professor of Linguistics at the Center for the Study of Language and Cognition, Zhejiang University, China. His research interests
include semantics, syntax, pragmatics and the interface between them. He has published many articles in scholarly journals such as Linguistic
Inquiry, Linguistics, Lingua, Journal of Pragmatics, Cognitive Linguistics and Journal of Chinese Linguistics.