Speaking about the Brain. Metaphors in popular scientific texts

Table of Contents

1. Introduction

2. Popular scientific texts and Metaphors 2.1 The role of Metaphors 2.2 A short diachronic overview of brain Metaphors

3. Linguistic Framework 3.1 Cognitive Linguistics 3.2 Conceptual Metaphor Theory 3.3 Primary Metaphor Theory 3.4 Blending Theory 3.4.2 What makes a blend metaphoric 3.4.3 Wrap-up: Comparison between CMT and BT 3.5 Metaphor Identification Procedure

4. The Corpus

5. Linguistic analyses of popular scientific Texts 5.1 Category Personification 5.2 Category Machines 5.3 Category Space 5.4 Category Mystery 5.5 Category Nature 5.6 Category Business 5.7 Category Chaos

6. Conclusion

7. References

8. Index

1. Introduction

If language and thought interact and shape each other, then how does our thinking support or limit our understanding of the human brain? Is this process of understanding directed due to the use of specific metaphors? What is the role of metaphors in general? Is the human brain conceptualised as a machine by the use of metaphoric language?

In order to be able to investigate all the mentioned questions, it is necessary to first introduce the instruments needed.

Basically, this thesis deals with metaphor use in popular scientific texts. More precisely, the topic of investigation is how metaphors are used in popular scientific texts, why metaphors are used when speaking about the human brain, and what kind of metaphors are used in this respect. More importantly, this the- sis aims at illuminating the mental processes of meaning construal, i.e. the co- gnitive processes that are triggered by the presence of metaphoric language and that are at work when dealing with popular scientific texts about the human brain.

In order to provide a comprehensive basis for the analysis of popular scientific texts dealing with the human brain, it is important to first give an overview of philosophical considerations about the conceptualisation of the human brain. Therefore, the first chapter gives a summary of philosophical ap- proaches towards speaking about the brain and some basic approaches towards why metaphors are used.

In the following, the necessary linguistic tools for the analysis of the self-compiled corpus of popular scientific texts are defined. The linguistic per- spective taken is the perspective of cognitive linguistics. Hence, the basic claims and assumptions of cognitive linguistics will be defined. From this perspective, the thesis aims at unfolding the mechanisms of the hu- man conceptual system, the rise of ideas, their verbalisation, and the role of metaphors as a vehicle of those ideas, all embedded in the field of metaphor use in popular scientific texts dealing with the human brain.

In terms of the needed theoretical framework for this thesis, linguistic meta- phor theories will be summarised. Most importantly, Lakoff and Johnson’s Conceptual Metaphor Theory (1980) will be addressed, as it gave birth to the view that metaphors are not just tropes or rhetorical devices but rather structure the way we think and therefore are present in everyday language. Deriving from this theory, Grady’s Primary Metaphor Theory (1997) will be briefly mentioned, as it constitutes a refinement of Lakoff and Johnson’s theory.

The explicit theoretical focus of this thesis is on Fauconnier and Turner’s con- ceptual integration networks (1998; 2002), frequently labelled as Blending Theory, because this approach constitutes a cogent method for analysing and displaying general cognitive operations, i.e. it is a reasonable tool for metaphor analysis too. Conceptual Metaphor Theory and Blending Theory will be com- pared in order to reveal similarities, differences, advantages, and disadvantages of both theories concerning the analysis of metaphors. The conclusion of this comparison aims at pointing out the relevance of Blending Theory as the in- strument of choice for analysing metaphors within the scope of this thesis.

As meaning is use (cf. Wittgenstein 1953), the scope of the investigati- on needs to be broadened, so that pragmatics can also be taken into account for answering the initially set research questions. One of the basic assumptions of cognitive linguistics is that cognitive semantics incorporates pragmatics as well, i.e. cognitive linguists regard pragmatics and semantics as a joint field anyway. This will be further illustrated in the outline of the basic assumptions of cognitive linguistics.

After providing a short overview of philosophical approaches towards the conceptualisation of the human brain in language, the use of metaphors, and knowledge structures, the basic assumptions of cognitive linguistics will be presented as well as the aforementioned metaphor theories and Blending Theo- ry.

The next step then is to address the issue of how metaphors can be identified. In this context, the Metaphor Identification Procedure by the Pragglejaz group (2007) will be explained.

After that the existing corpus will be specified, some consequences will

be mentioned, and some difficulties of compiling materials for the creation of a valid corpus for the investigation of the main research question will be outli- ned.

In the next chapter, the main body of the thesis, a linguistic analysis of selected examples of metaphor use in popular scientific texts about the human brain will be conducted by applying Fauconnier and Turner’s Blending Theory. The results of the analyses will be interpreted and summarised.

Finally, the thesis will be completed with a conclusion which picks up on the research question, the philosophical implications, the linguistic theories used, and the results of the analyses.

2. Popular scientific texts and Metaphors

A central question that needs to be asked when trying to grasp the role of metaphors is whether metaphors in use are a reflection of the current predominant scientific paradigm and thus are symbols representing contemporary perception and reasoning in specific fields.

An interesting claim made by Rorty in hisPhilosophy and the Mirror of Nature(1979: 12) is that metaphors rather than statements determine most of our philosophical convictions.

In other words, metaphors are powerful devices which structure our conceptual system. Due to the fact that metaphors are powerful, their role is also of particular importance. Hence, this chapter aims at providing a brief overview of philosophical approaches towards the role of metaphors in general. In the second part of this chapter, the diachronic development of different metaphors for the human brain will be summarised.

2.1 The role of Metaphors

First of all, before providing an overview of philosophical approaches towards the role of metaphor, we need to define what a metaphor is. A more detailed and linguistic definition of metaphor will be provided in the next chapter. For now, a sufficient definition can be found in the Oxford Dictionary of English, which defines a metaphor as “a figure of speech in which a word or phrase is applied to an object or action to which it is not literally applicable” (OED1 2014). Thus, we approach a comprehensive and linguistic definition of meta- phor by starting from a semasiological perspective for now.

The Polish microbiologist Ludwik Fleck contributed to the understanding of the role of metaphors by addressing the question of how scientific facts emerge in hisGenesis and Development of a Scientific Fact(1979). According to Fleck, “scientific knowledge is a function of what certain people believe at a certain time in history” (van Tongeren, 1997)2.

Thus, the human conceptual system is based on social and cultural compon- ents.

In his complex book The Archaeology of Knowledge (1969) Michel Foucault explains his methodological approach by making use of the term ar- chaeology in the sense that specific systems of knowledge, e.g. a priori know- ledge, are bound to rules which define the boundaries of a specific discourse and these systems have to be ‘archaeologically’ uncovered. Foucault provides an extensive definition of knowledge, which according to him, is, among other things, “defined by the possibilities of use and appropriation offered by dis- course” (Foucault 1969: 190). The notion of knowledge is central to the analy- sis of metaphors, as metaphors are vehicles for the transfer of knowledge. As For the sake of brevity, the acronymOEDwill be used from now for denoting the Oxford Dictionary of English.

InMetaphors in medical Texts(1997) Geraldine W. van Rijn-van Tongeren refers to Fleck in order to support Lakoff and Johnson’s claim that knowledge and understanding are based on experience, i.e. they are based on the interaction of humans with their environment, which is also a basic assumption of cognitive linguistics as it will be pointed out in the following chapter. the notion of knowledge is also central to cognitive linguistics, this point will be further elaborated in the following chapter.

Foucault, similar to Fleck, claims that knowledge is governed by socio-cultural frames, more precisely predominant paradigms, which are subject to change in the course of time.

In his bookGenesis and Development of a Scientific Fact(1979), Fleck discusses the concept of syphilis and the discovery and development of the Wassermann reaction3 in order to test his theory that scientific concepts and theories depend on social and cultural conditions.

He notices that, over the course of time, the concept of syphilis has changed and that this transformation is accompanied by the emergence of new fields of knowledge (cf. van Tongeren 1997: 40). These findings are in accordance with Foucault’s claims that a specific discourse changes over the course of time in the sense that a predominant paradigm is abandoned in order to be replaced by a new one due to the gain of knowledge.

Furthermore, Fleck claims “that a scientific concept should be viewed as a re- sult of the historical development of several lines of collective thought” (van Tongeren 1997: 41). His notion of collective thought is crucial to his theory. He provides a definition of what he calls a thought collective which is “a commu- nity of persons mutually exchanging ideas or maintaining intellectual interac- tion” (Fleck 1979: 39) [original italics]. Also, he makes use of the term thought style in the sense that the thought collective is “the special ‘carrier’ for the thought style, the historical development of any field of thought, as well as the given stock of knowledge and level of culture” (Fleck 1979: 39) [original ita- lics]. It is important to emphasise that in Fleck’s theory thought style is deter- mined by language and in this respect, of course, by metaphors, as language is a vehicle for the transmission and preservation of knowledge within a society.

So, how exactly can Fleck’s theory contribute to investigating metaphor use in popular scientific texts about the human brain? Basically, as will be shown in the following chapter, Fleck’s theory contains ideas which support

The Wassermann reaction is the complement-fixing reaction which can be found in a positive complement-fixation test for syphilis.

Lakoff and Johnson’s theory of metaphor as a conceptual phenomenon which

is entrenched in everyday language and thought. Fleck does not provide a lin- guistic theory of metaphor, but he points out that concepts and theories are go- verned by the social and cultural environment they are embedded in. Further- more, Fleck uses the phrase “neglected relations” (cf. Fleck 1979: 22) to argue that the choice of words leads to highlighting or backgrounding ideas which a theory consists of (cf. van Tongeren 1997: 41). As will be pointed out in the next chapter, this claim is also in accordance with the basic assumptions of co- gnitive linguistics and Lakoff and Johnson’s Conceptual Metaphor Theory (1980).

Interestingly, Fleck’s theory also incorporates the notion of cognition. He claims that cognition “means to ascertain the results following from certain preconditions” (van Tongeren 1997: 43). According to his view, these precon- ditions equal active elements of knowledge, i.e. they are actively assumed by a specific thought collective (cf. van Tongeren 1997: 43). The contrary concept, passive aspects of knowledge, are, according to his theory, the consequential results of the aforementioned assumptions. Hence, choosing a metaphor for establishing a theory, or more generally an idea, is an active element of know- ledge, whereas the entailments of the metaphor are the passive elements of knowledge. As this view of cognition refers more to a mutual mental process of the whole thought collective, it is important to mention that Fleck’s understan- ding of the role of the individual within a collective is that “the act of ascertai- ning is the contribution of the individual” (Fleck 1979: 40).

Fleck argues that cognition is “the most socially-conditioned activity of man” (ibid.: 43). He assigns great importance to words, especially when they are adopted by the collective and taken away from the individual. This process of collectively using specific terms in reference to an item or a phenomenon obviously relates to the process of conventionalisation, i.e. the cultural ent- renchment of the use of a word in a specific context. Fleck also states that words may even “acquire a magical power and exert a mental influence simply by being used” (ibid.). Despite the non-scientific term magical, this phrase car- ries a strong claim: (1) once established in a collective, e.g. a group of experts or, in a broader sense, a speech community, words become very powerful; (2) words exert a mental influence on those who use them. This clearly relates to Lakoff and Johnson’s Conceptual Metaphor Theory (1980), as will be demonstrated in the next chapter.

As a consequence of this claim, individuals within a collective “cannot think in any other way” (ibid.: 47) [original italics] due to the social consolidation of cognition (cf. van Tongeren 1997: 44).

According to Fleck, every thought collective of science consists of two concentric circles: (1) a smaller esoteric circle consisting of experts and (2) a larger exoteric circle consisting of laymen (cf. Fleck 1979: 105). As he also points out, the idea of thought collectives can be expanded to thought commu- nities, e.g. religion, politics, commerce, art, or even fashion (cf. ibid.: 102 ff.). What all these thought communities or collectives have in common is the way the two aforementioned circles, or more precisely the members of these circles, communicate with each other (cf. van Tongeren 1997: 47). Furthermore, Fleck describes this communication process in more detail, comparing it to the relati- on of the elite to the masses known in sociology (cf. Fleck 1979: 105). By using the example of a fashion community, Fleck points out that “[t]he most dedicated followers of fashion are found far out in the exoteric circle”, as they do not have direct contact with the powerful members of the esoteric circle (ibid.: 108). This points out the social interrelations and shows that a specific thought style is subject to the power relations between the two circles and, in addition to that, even to the power relations within the circles.

However, according to Fleck, the inherent form of science of each circle is expressed in different types of scientific literature; he distinguishes between four types: (1) journal science, (2) vademecum science, (3) popular science, (4) textbook science (cf. ibid.: 112). As this thesis aims at analysing metaphors in popular scientific texts, the focus is on the third category, while the others will go unnoticed for the sake of brevity.

A characteristic of popular science is that through the omission of detail and controversial opinions “an artificial simplification” is produced (ibid.: 112). In addition to that, “[p]opular science in the strict sense is science for nonexperts, that is, for the large circle of adult, generally educated amateurs” (ibid.: 112).

This simplification of knowledge not only makes science attractive and lively, but also establishes scientific facts as if they were in no way controversial. However, there is also the element of interrelation between expert knowledge and popular knowledge involved. Fleck points this out by giving the example of an economist who “speaks of the organism of the economy” (ibid.: 112), which is a demonstration of how metaphors used by the experts are turned into certainties in the exoteric circle. Nevertheless, these metaphors reflect back on the esoteric circle due to the feedback of the exoteric circle and influence the experts as well (cf. ibid.: 112 ff.). According to Fleck, “[p]opular exoteric knowledge stems from specialised esoteric knowledge” (ibid.: 113).

In short, popular exoteric knowledge “shapes specific public opinion as well as the Weltanschauung and in this form reacts in turn upon the expert” (ibid.: 113).

The essence of the above briefly summarised approaches is that (1) knowledge is subject to processes of transformation in the course of time, (2) cognition as the mental process of acquiring knowledge through thought or ex- perience is governed by social and cultural principles, (3) in popular science there is an interaction between professionals and laymen, (4) popular science is a simplification of knowledge.

As this thesis focuses on the linguistic analysis of metaphors in popular scientific texts about the human brain, certain aspects of Ludwik Fleck’s (1979) ideas have been left out. Also, other epistemological theories such as Kuhn’sThe Structure of Scientific Revolutions(1962) will not be used for the sake of brevity and due to the focus on the analysis of metaphors from the perspective of cognitive linguistics.

2.2 A short diachronic overview of brain Metaphors

As the focus of this thesis is on analysing current popular scientific texts about the human brain and the metaphorical expressions used in reference to the brain, this thesis cannot provide a full diachronic account of brain metaphors and paradigm shifts over the course of time. It needs to be emphasised that this thesis constitutes a study on the current use of brain metaphors. Nevertheless, a short diachronic overview of brain metaphors provides a better understanding of the current conceptualisation of the human brain in language. Therefore, the following paragraphs will provide a short sketch of the most im- portant conceptualisations of the human brain over the course of time. In his book The Metaphorical Brain (1972) the neuroscientist Michael

A. Arbib claims that much of modern thought about the brain can be summari- sed in two metaphors: (1) the cybernetic metaphor (“HUMANS ARE MA- CHINES”4), and (2) the evolutionary metaphor (“HUMANS ARE ANIMALS”) (cf. Daugman, 2001). This simple categorisation may be proven wrong for popular scientific texts in the analysis chapter of this thesis, as the evolutionary metaphor may be bound to other contexts than popular science, but it may also be proven right.

However, the cybernetic metaphor seems to be the predominant meta- phor as will be shown in a short chronological overview of brain metaphors. The assumably earliest model of the mind is ascribed to Hippocrates who, in the fifth century BC, proposed an early hydraulic model of the mind which was based on the four humours (cf. Daugman 2001: 4).

In ancient Greece, the water wheel was already known6 which may be seen as the motif for linking humans and machines in the way that water drives the

The use of capitalised letters is due to Lakoff and Johnson’s conventional use of mnemonics for labelling conceptual metaphors.

For a more detailed account of ancient Greek medicine see Jouanna, Jacques, and Ph. J. van der Eijk. Greek medicine from Hippocrates to Galen selected papers. Leiden: Brill, 2012. C.f., for example, Lucas, Adam. “Waterwheel”. Encyclopaedia of Technology andInvention. New York: Basic Books, 2011. wheel and blood drives the human body, which constitutes a cause-effect rela- tion in both cases.

It seems that the technological environment already influenced ancient Greek philosophers in their conceptualisation of abstract ideas and theories. This hy- draulic model of the mind was carried on by the Romans with some modifica- tions.

The probably first mechanical metaphor for the brain is attributed to Descartes. He proposed that “animals, as distinct from men, were pure automa- ta” (Daugman 2001: 4) [original italics]. Due to the fact that Descartes was a deist and religion played a major political and social role at that time, he limi- ted this mechanical brain metaphor in its application only to animals. The En- lightenment opened the door for even more progressive ideas; one of them is certainly La Mettrie’s L’ Homme machine (1748) in which the human body and the human brain are explicitly described as machines (cf. ibid.: 5).

As technology advanced, brain metaphors adapted accordingly. The telegraph provided a nerve metaphor in Hodgin and Huxley’s studies on membrane current in nerves (cf. ibid.: 7). Even earlier John von Neumann published his bookThe Computer and the Brain(1958) in which he explored the idea that a logical calculus is the foundation for nervous activity, which is an equation of computers and human brains (cf. ibid.: 9).

To cut the matter short, it seems that humans have for a very long time understood the human brain in terms of machines. As technology advances, the metaphors used also advance, marking a significant progression not only in exploring how the human brain works but at the same time, as new technology is conceptualised as a symbol of the advance of the human race, postulating the equation of humans and machines.

The resulting question is whether the mental marriage of humans and machines is a liberation or a prison.

As current technology focuses on computers and associated networks, the computer metaphor for the brain today remains very popular, even ubiqui- tous. Nonetheless, even the most established paradigms are bound to eventual- ly become obsolete and to be replaced. Hence, the computer metaphor should not be taken for granted.

A linguistic investigation of the current use of brain metaphors in popu- lar scientific texts will reveal which mappings, or conceptualisations, are pre- dominant.

Before we can turn to the analysis of brain metaphors in popular scientific texts, it is necessary to explicitly define the linguistic perspective taken and the instruments or, more precisely, the theories needed for the analysis. This is the aim of the following chapter.

3. Linguistic Framework

To be able to see the bigger picture, the previous chapter provided an overview of some early and more recent metaphorical conceptualisations of the human brain in scientific discourse. In addition, an approximation towards the nature of popular scientific texts was provided. In this context, the mechanisms of establishing certain perspectives and of spreading specific knowledge through the use of metaphors were briefly stated.

This chapter provides the foundation for the linguistic analysis of brain metaphors in popular scientific texts by defining the linguistic perspective of this thesis and by summarising the most important and most influential meta- phor theories.

3.1 Cognitive Linguistics

According to Fauconnier (2003), cognitive linguistics is “a powerful approach to the study of language, conceptual systems, human cognition, and general meaning construction” (Fauconnier 2003: 1). Fauconnier lists several fields within cognitive linguistics, e.g. cognitive grammar (cf. Langacker 1991), me- taphor theory, mental spaces, and conceptual integration (cf. Fauconnier 2003). As this thesis deals with brain metaphors in popular scientific texts and there- fore analyses word meanings in specific contexts, the focus will be on cogniti- ve semantics, metaphor theory, and conceptual integration. In this thesis, the relations between cognitive semantics and other semantics theories will not be discussed.

First, the guiding principles of cognitive semantics will be summarised. According to Talmy, “[r]esearch on cognitive semantics is research on conceptual content and its organisation in language” (Talmy 2000: 4). Talmy’s proposition is definitely a valid summary of the research interest and aim of cognitive semantics, but it needs to be specified in greater detail.

Evans and Green (2006) provide a comprehensive basic overview of the general assumptions of cognitive semantics.

They list four guiding principles of cognitive semantics:

1. Conceptual structure is embodied (the‘embodied cognition thesis’). 2. Semantic structure is conceptual structure. 3. Meaning representation is encyclopaedic. 4. Meaning construction is conceptualisation.

(Evans & Green 2006: 157)

In the following, these four guiding principles will be explained in more detail. The embodied cognition thesis, mentioned in the first principle, holds that conceptual organisation is embodied, i.e. bodily experience or, in other words, our interaction with the environment gives rise to conceptual structure and makes it meaningful (cf. ibid.). Evans and Green illustrate this point by giving an example which refers to image schemas.

An image schema is a “relatively abstract conceptual representation that arises directly from our everyday interaction with and observation of the world around us” (Evans 2007: 106). In other words, an image schema derives from embodied experience. A frequently used example in cognitive linguistics literature for the illustration of image schemas is the containment (CONTAINER) image schema in which the “directly embodied experience of interacting with bounded landmarks” (Evans & Green 2006: 158) gives rise to meaningful concepts, i.e. the human body, for example, is often conceptualised as a container: “He’sinlove.” (Lakoff & Johnson 1980: 32).

These points will be illustrated in more detail in the discussion of Lakoff and Johnson’s Conceptual Metaphor Theory.

The second guiding principle claims that “language refers to concepts in the mind of the speaker rather than to objects in the external world” (Evans & Green 2006: 158), i.e. meanings are basically concepts. However, it is im- portant to note that although semantic structure, i.e. the meanings that are asso- ciated with linguistic units (e.g. words), is put on a level with conceptual struc- ture, both are not identical. Meanings rather “form only a subset of possible concepts” (ibid.: 159).

The third guiding principle claims that semantic structure is encyclopa- edic. This principle holds that “word meaning cannot be understood indepen- dently of the vast repository of encyclopaedic knowledge to which it is lin- ked” (ibid.: 206) [original emphasis]. Secondly, “this encyclopaedic knowledge is grounded in human interaction with others (social experience) and the world around us (physical experience)” (ibid.: 206). Thus, cognitive semantics rejects the dictionary view of word meaning that words only “represent neatly packa- ged bundles of meaning (ibid.: 160). Instead, cognitive semantics rather re- gards conventional meaning “just a ‘prompt’ for the process of meaning con- struction: the ‘selection’ of an appropriate interpretation against the context of the utterance” (ibid.: 161) [original emphasis].

The fourth guiding principle listed by Evans and Green asserts that meaning is constructed at the conceptual level: meaning construction is equated withconceptualisation, a dynamic process whereby linguistic unitsserve as prompts for an array of conceptual operations and the recruitment of background knowledge.

(ibid.: 162) [original emphasis]

It is important to emphasise that these four guiding principles, of course, are interrelated. In the course of this thesis, these principles will be illuminated in more detail.

As already mentioned, the focus of this thesis in terms of the applied theory regarding the analysis of metaphors will be on Fauconnier and Turner’s Blending Theory. Their theory strives to reveal the mental processes that are involved in the comprehension of language. Therefore,meaningandknowledgeare two central aspects that need further consideration.

The encyclopaedic view postulated by cognitive linguistics holds that encyclo- paedic knowledge, which is accessed via language, “provides simulations of perceptual experience” (ibid.: 206). The traditional view within the field of le- xical semantics, the dictionary view, holds that the mental lexicon represents the knowledge of word meaning (cf. ibid.: 208). According to this view, “the core meaning of a word is the information contained in the word’s definition” (ibid.: 208). Thus, the dictionary view distinguishes between lingu- istic knowledge (word meaning) and encyclopaedic knowledge, which is con- sidered external to linguistic knowledge and is categorised as a subset of world knowledge, e.g. a word’s connotations (cf. ibid.: 207 ff.).

A consequence of this distinction is that lexical semanticists who take the dictionary view treat semantics and pragmatics as two different disciplines; word meaning (sense) falls into the domain of semantics, whereas encyclopaedic knowledge (reference) incorporates contextual factors, e.g. cultural knowledge, which influence meaning and therefore falls into the domain of pragmatics (cf. ibid.: 209). There are several problems addressed by Evans and Green (2006) which arise from this dichotomy (cf. ibid.: 210-13).

To cut a long story short, the essence is that cognitive semanticists re- ject this arbitrary distinction between dictionary knowledge and encyclopaedic knowledge due to the observation that a word’s meaning is always dependent on the context of its use, e.g. deictic expressions always need a contextual reference (cf. ibid.: 212).

Cognitive semanticists do not deny that “words do have relatively well-entren- ched meanings stored in long-term memory (the coded meaning)” (ibid.: 213), but because word meaning depends on the exact contextual use, the construction of word meaning is always an on-line process (cf. ibid.: 213-14). Hence, cognitive semantics does not distinguish between dictionary knowledge and encyclopaedic knowledge; it rather views dictionary knowledge as a subset of encyclopaedic knowledge (cf. ibid.: 207, 216).

Moreover, cognitive semanticists consider encyclopaedic knowledge to be an organised hierarchical network structure (cf. ibid.: 216).

In this respect, there are certain aspects of knowledge which are more central than others. According to Langacker (1987), salience determines the centrality of specific aspects of knowledge (cf. ibid.: 117).

Furthermore, Langacker (1987) divides the aforementioned encyclopaedic network into four different but overlapping types of knowledge.

The first type is conventional knowledge which refers to widely known information within a speech community. Because this information is well established within a speech community, it is “more central to the mental representation of a particular lexical concept” (ibid.: 217).

The second type, generic knowledge, is different from specific know- ledge, as it refers to a more comprehensive category, i.e. it “applies to many instances of a particular category” (ibid.: 218). As the name already implies, it refers to more general knowledge and is very likely to be conventional. One example for this distinction is that it is well known that lemons usually taste bitter (generic knowledge), whereas the knowledge about the rotten lemon that you wanted to dunk into your cup of tea this morning falls into the category of specific knowledge (cf. ibid.).

Intrinsic knowledge, which constitutes the third type, refers to the spe- cific properties of an entity that are internal to it. The opposite concept, extrin- sic knowledge, is knowledge which is external to this entity. For example, the shape of a banana is intrinsic knowledge, whereas the knowledge that Andy Warhol’s iconic pop art banana is on the cover of The Velvet Underground & Nico’s debut album is extrinsic knowledge. This example also illustrates that these knowledge configurations overlap, i.e. the well-known iconic LP cover is common cultural knowledge which is extrinsic knowledge in reference to the banana but at the same time one aspect of conventional knowledge.

The fourth type of Langacker’s (1987) encyclopaedic knowledge net- work is characteristic knowledge which indicates “the degree to which know- ledge is unique to a particular class of entities” (Evans & Green 2006: 218). For example, in Formula One racing, the metonymic and metaphorical expres- sion Silver Arrow refers to the Mercedes racing team, as the colour silver has been characteristic of German racing teams for decades. Interestingly, this ex- ample again provides evidence for the overlapping of the different knowledge categories, as the expression Silver Arrow is an established phrase for a specific kind of race cars. It is important to note that these knowledge categories repre- sent continua (cf. ibid.: 219).

So, although the encyclopaedic view entails that meaning depends on context in opposition to the dictionary view which splits core meanings from non-core meanings, i.e. semantics from pragmatics, encyclopaedic meaning does not equal contextual meaning (cf. ibid.: 220). It is the interaction of the aforemen- tioned four kinds of knowledge that gives rise to encyclopaedic meaning, i.e. the specific context activates a selection process in terms of choosing the right meaning in a given context.

Cruse (1986) dubs the phenomenon of assigning more importance to a specific aspect of the encyclopaedic knowledge due to the discourse contextcontextualmodulation(cf. ibid.).

This section provided a short summary of the central assumptions of cognitive linguistics and especially a definition of the encyclopaedic view as a knowledge configuration.

Next, cognitive linguistic theories of metaphor will be summarised, starting with Lakoff and Johnson’s Conceptual Metaphor Theory (1980) before turning to Grady’s Primary Metaphor Theory (1997).

3.2 Conceptual Metaphor Theory

Although similes and metaphors are related and similar concepts, it is important to distinguish between them. A good distinction can be found in Croft and Cruse (2004) who claim that in similes there is a restricted mapping between two domains, whereas in prototypical metaphors there is an open mapping (cf. Croft & Cruse 2004: 213) between both domains. Additionally, two domains are presented as a blend in metaphors, whereas in a simile, the two domains are presented as separate (cf. ibid.).

The traditional view of metaphor holds that two kinds of meaning can be clear- ly distinguished: literal and figurative meaning (cf. Evans & Green 2006: 287). The central idea of this view is that while literal language is the conventional way of communicating meaning, figurative language is rather literary (cf. ibid.). Thus, most everyday language is literal according to this position.

This chapter deals with Lakoff and Johnson’s Conceptual Metaphor Theory (1980), as the theory is commonly regarded as the cornerstone of ap- proaches to metaphor from the perspective of cognitive linguistics. Indeed, the- re are also earlier approaches to metaphor which, to some extent, include co- gnitive notions, e.g. Richards (1936) who considered metaphors a technique for the facilitation of comprehension and claimed that meaning has an experiential basis. In addition, he proposed that cognitive processes such as categorisation are involved in the comprehension of metaphors. Another example is Black (1962) who pointed out that a literal paraphrase of a metaphorical expressions leads to a loss of cognitive content.

Lakoff (1993) himself pays tribute to Reddy (1979) who, according to Lakoff, demonstrated that everyday English is largely metaphorical and there- fore contributed to the abandonment of the traditional view of metaphor (cf. Lakoff 1993).

However, within the scope of this thesis only Lakoff and Johnson’s CMT7 will be considered the take-off of a metaphor theory from the perspective of cognitive linguistics.

Cognitive semanticists consider metaphor “fundamental to the structure of the conceptual system rather than superficial linguistic ‘devices’” (Evans & Green 2006: 293).

In their book Metaphors We Live By (1980) Lakoff and Johnson made three claims: (1) metaphor is ubiquitous in everyday language, i.e. metaphor is not just a rhetorical figure as the traditional view of metaphor proposes; (2) there is a high degree of coherence and systematicity in these entrenched meta- phors which are commonly used in everyday language; (3) metaphor is a mode of thought, i.e. metaphors structure certain concepts (cf. Taylor 2002: 487).

The key assumption of Lakoff and Johnson (1980) is that metaphor is basically a cognitive operation in which elements from a source domain are mapped onto elements of a target domain. The mapping is unidirectional and these do- mains are not arbitrarily chosen but rather are in some way coherent. Thus, the essence of metaphor is not the single involved elements of the distinct domains but their interaction. In other words, conceptual metaphors feature cross-do- main correspondences (cf. Taylor 2002: 489). In general, conceptual operations such as mappings from a source to a target domain are known as conceptual projection (cf. Evans & Green 2006: 286).

By convention, Lakoff and Johnson (1980) use mnemonics, i.e. small capitals to indicate the domain mappings, e.g. “THE MIND IS A BODY” (Lakoff & Johnson 1999) in which the first domain (MIND) corresponds to the target domain and the second (BODY) to the source domain (cf. Taylor 2002: 489-90). In conceptual metaphors the source domain “is concrete, and can be experienced or perceived ‘directly’, while the target domain is more abstract or concerns ‘subjective’ experience” (ibid.: 491).

According to Lakoff and Johnson (1999), a consequence of this fixed direction of mapping from source to target is that abstract domains need to be conceptua-

From now on, the acronymCMTwill be used in reference to Lakoff and Johnson’s Conceptual Metaphor Theory (1980) for the sake of brevity. lised through more basic domains, i.e. they need to be accessed through meta- phor (cf. ibid.).

Langacker defines domains as “necessarily cognitive entities: mental experiences, representational spaces, concepts, or conceptual complexes” (Langacker 1987: 147). So, before returning to CMT, a further cla- rification of domain is necessary. We need to touch upon Langacker’s theory of domains (1987) and to a minor extent also upon Fillmore’s notion of frames (1975; 1977; 1982; 1985; 1992) in order to establish a positive base for the un- derstanding of CMT, Grady’s Primary Metaphor Theory (1997), and Faucon- nier and Turner’s Blending Theory.

Firstly, Langacker’s theory of domains presupposes that meaning is encyclopa- edic, and that “lexical concepts cannot be understood independently of larger knowledge structures” (Evans & Green 2006: 230). Furthermore, Langacker (1987) argues that domains hold a hierarchical structure, i.e. they differ in terms of complexity and organisation (cf. ibid.: 230-31). In other words, some domains are more basic than others. He refers to this hierarchical network of domains as domain matrices (cf. Langacker 1987: 150-52). Basic domains such as SPACE are directly experienced through the bodily interaction with the en- vironment; thus, they are derived from embodied experience (cf. Evans & Green 2006: 231). Furthermore, Langacker (1987) argues that “basic domains derive from pre-conceptual experience, such as sensory-perceptual experience, which forms the basis of more complex knowledge domains” (Evans & Green 2006: 232).

Evans and Green (2006) also note that Langacker’s theory of domains focuses on “conceptual ontology: the structure and organisation of knowledge, and the way in which concepts are related to and understood in terms of others” (ibid.: 231) [original emphasis].

Another knowledge structure which is related to Langacker’s domains is a frame. A frame can be defined as “a schematisation of experience (a know- ledge structure), which is represented at the conceptual level and held in long- term memory” (ibid.: 222). Basically, frames enable the understanding of words, as they are complex knowledge structures. Fillmore (1982) notes that a frame provides “the background and motivation for the categories which these words represent” (Fillmore 1982: 116-17). The terms domains and frames are often used interchangeably, although they differ in certain respects. According to Langacker (1987), concepts are rather structured by multiple domains (cf. Evans & Green 2006: 230-31). Also, Fillmore did not explicitly point out the distinction between basic and abstract domains, which is based on embodi- ment. While Fillmore’s frames primarily address abstract domains, Langa- cker’s theory addresses both types (cf. ibid.: 231). Another difference is that Fillmore (1992) rather links frames to grammatical implications, e.g. valence relations, whereas Langacker’s domains are concerned with knowledge structu- re and organisation (cf. ibid.).

Finally, it is important to note that basic domains do not equal image schemas. A good overview can be found in Evans and Green (2006). Basically, although basic domains and image schemas both “are conceptual representati- ons that are directly tied to pre-conceptual experience” (ibid.: 233), image schemas are more complex knowledge structures, compared to basic domains. Thus, image schemas “are likely to contribute to the domain matrices of a wide range of concepts” (ibid.: 234), whereas basic domains are not necessarily bound to a range of domain matrices. Evans and Green (2006) even observe a third distinction between image schemas and basic domains; they argue that image schemas “derive from sensory experience and therefore have image content” (ibid.: 235) [original emphasis]. Accordingly, basic domains, e.g. SPACE, are less complex and not imagistic in nature.

After this definition of domains, frames and their relations, we now come back to Lakoff and Johnson’s Conceptual Metaphor Theory (1980). So far, it has been pointed out that metaphor is a mapping of elements from a source domain to elements of a target domain. This mapping is a cognitive pro- cess which structures the way we think; hence, it is conceptual. Lakoff and Johnson (1980) address three different metaphor types: (1) structu- ral metaphors, (2) orientational metaphors, (3) ontological metaphors. Accor- ding to Lakoff and Johnson (1980), structural metaphors are “cases where one concept is metaphorically structured in terms of another” (Lakoff & Johnson 1980: 14). Orientational metaphors organise systems of concepts with respect to one another. In most cases, orientational metaphors give a concept spatial orientation, e.g. HAPPY IS UP (cf. ibid.). Ontological metaphors are based on our experience of psychical objects and substances and let us structure abstract concepts in terms of these experiences (cf. ibid.: 25). Personification and con- tainment are examples for ontological metaphors. All three types of metaphor have an experiential basis.

However, Taylor (2002) critically reflects Lakoff and Johnson’s (1980) claim that due to the asymmetrical mapping from a source to a target domain, i.e. the unidirectionality of the mapping operations, and the ubiquity of metaphor in everyday language, an abstract domain always must be accessed through meta- phor (Taylor 2002: 491). He claims that the target domain needs to have an in- itial structure before a metaphorical mapping is performed, as the knowledge of which elements of the source domain can be mapped to which elements of the target domain is an essential precondition (cf. ibid.). In other words, Taylor (2002) doubts that “it is metaphor that creates our conceptions of reasoning, time, morality, and so on” (ibid.) [original italics].

However, the main objective of Lakoff and Johnson’s CMT is to uncover the metaphorical mappings between conceptual domains (cf. Croft & Cruse 2004: 194). Furthermore, Lakoff and Johnson distinguish between novel metaphors, which they call ‘image metaphors’, and conventionalised metaphors (Lakoff & Turner 1989: 99; Lakoff 1993: 229). Lakoff and Johnson (1980) argue that tho- se metaphors which become conventionalised are grounded in human experi- ence and therefore are cognitively significant. For that reason, CMT focuses on conventional metaphors, not novel metaphors (cf. Croft & Cruse 2004: 195). One example for conventional metaphors has already been provided in refe- rence to image schemas (see above); “He’s in love.” (Lakoff & Johnson 1980: 32) is a metaphorical phrase for the underlying conceptual metaphor STATES ARE CONTAINERS (cf. ibid.: 31-32), i.e. it is an ontological metaphor. This example illustrates that a conventional metaphor basically is a conceptual mapping between elements of a source domain to elements of a target domain. Moreover, “a conceptual metaphor is schematic for the metaphorical expressi- on which instantiate it” (Taylor 2002: 493). A schema can be defined as a “symbolic assembly viewed from the perspective of the usage-based thesis” (Evans 2007: 187)8. These schemas are arranged as a continuum; there are high-level schemas and low-level schemas (cf. Taylor 2002: 493-94). Low- level schemas “are closer to fully specified instances”, whereas “[h]igh-level schemas capture commonalities across a large range of examples” (ibid.). As already mentioned in the comparison of domains and image schemas, an image schema is emergent and derives from sensory and perceptual interaction with the world, i.e. pre-conceptual embodied experience (cf. Evans & Green 2006: 301).

Furthermore, Croft and Cruse (2004) point out that “[t]he mapping between source and target domains involves two sorts of correspondences, epistemic and ontological” (Croft & Cruse 2004: 196). While there are ontological corre- spondences between elements of a source domain and elements of a target do- main, epistemic correspondences refer to “relations between elements in one domain and relations between elements in the other domain” (ibid.).

However, given the conceptual nature of metaphor, the metaphorical mappings carry rich inferences, i.e. “certain patterns of reasoning” (cf. ibid.: 197) which may be transferred from the source to the target domain. These rich inferences are calledmetaphoricalentailments.

Another central aspect of Lakoff and Johnson’s theory is the Invariance Hypothesis proposed by Lakoff (1990: 54). Basically, given the unidirectionality of mapping operations from source to target domain, it aims to explain which elements of the source domain are mapped onto the target domain. Hence, the Invariance Hypothesis is “a constraint on metaphorical mapping” (Croft & Cruse 2004: 201), i.e. “it constrains potentially incompatible mappings” (Evans & Green 2006: 302).

Lakoff (1990) argues that “[m]etaphorical mappings preserve the cognitive to- pology (that is, image-schematic structure) of the source domain” (Lakoff “The usage-based thesis holds that the mental grammar of the language user (his or her knowledge of language) is formed by the abstraction of symbolic units from situ- ated instances of language use: an utterance” (Evans 2007: 216-17) [original empha- sis]. 1990: 54). In addition, Lakoff claims that “what is mapped preserves image- schematic structure, though not all image-schematic structure need be mapped” (ibid.: 72). Furthermore, Lakoff adds that the image-schematic structure of the source domain determines reasoning in the target domain (cf. Croft & Cruse 2004: 201).

Due to an important constraint on the Invariance Hypothesis proposed by Turner (1990), namely that the image-schematic structure of the target domain should not be violated when using metaphors (cf. Turner 1990: 252), Lakoff (1993) introduces the notion of target domain overrides (Lakoff 1993: 216). Lakoff (1993) illustrates this idea by using the example of give a kick versus give an idea, in which only the latter allows that the transmitted force or energy is kept in the target domain (cf. Croft & Cruse 2004: 201). So, domain over- rides are basically processes that prevent metaphorical entailments from pro- jecting to the target domain (cf. Evans & Green 2006: 303).

Lakoff and Johnson (1980) argue that although target domains are structured to some extent, “they are not clearly enough delineated in their own terms to satisfy the purposes of our day-to-day functioning” (Lakoff & Johnson 1980: 118). Also, due to the asymmetry of source and target domain, conceptual metaphors are needed in everyday language to structure concepts (cf. Croft & Cruse 2004: 202). In fact, the Invariance Hypothesis and the target domain override cannot fully account for which specific elements from a source domain are mapped onto a target domain and how the resulting emergent structure in the target domain is determined (cf. ibid.).

However, as it will be shown later, Fauconnier and Turner’s Blending Theory (1998; 2002) provides a more precise account for the selection of certain elements in the mapping process.

As already mentioned before, Lakoff and Johnson’s CMT focuses on conventional metaphors rather than on novel metaphors. Yet, they embed also novel metaphors (Lakoff & Johnson 1980: 139-43) and argue that some novel metaphors are created through the extension of already existing conventional metaphors (ibid.: 52-3), while other novel metaphors, i.e. image metaphors9, are more restricted (cf. Lakoff & Turner 1989: 90). The reason for the restriction of image metaphors is that there is no involvement of rich knowledge mapping and no inferential structure of conventional image-schematic metaphors in these metaphors (ibid.).

According to Croft and Cruse (2004), “it seems plausible that even conventional metaphors draw on the full richness of our encyclopaedic knowl- edge of our bodily and cultural experience, especially when they are first coined” (Croft & Cruse 2004: 204). In other words, they suggest that even conventional metaphors feature a richer blending than only image-schematic structure (cf. ibid.). Furthermore, Croft and Cruse (2004) argue that conven- tional metaphors have lost some of their original properties and thus, novel metaphors need to be studied to get to the core of the underlying cognitive mechanisms involved in the interpretation of these metaphors (cf. ibid.). They describe several stages in the life of a metaphor, i.e. they describe how a metaphor becomes entrenched in everyday language and therefore becomes conventionalised and equal to other literal expressions (cf. ibid.: 205).

Lakoff (1987) himself claims that it is important to distinguish between con- ventional metaphors and dead metaphors. He points out that only words like pedigree are dead metaphors, as a normal contemporary speaker of English is no longer aware of the word’s derivation from the Old French pie de grue, which meant “foot of a crane” (cf. Lakoff 1987: 143). The word pedigree is stored in the mental lexicon as an independent lexical unit without a connec- tion to its original source domain ANIMAL BODY PARTS. As the contempo- rary speaker of English is not aware of the etymological derivation of pedigree, the term in its original meaning cannot act as a literal basis for further metaphorical extension, which is also not possible for completely novel metaphors, as they are at the beginning not entrenched within a speech com- munity (cf. Croft & Cruse 2004: 204-5). Once a metaphorical expression is Metaphors like ‘He is a lion.’ are called resemblance metaphors (Grady 1999), as they describe the perceived resemblance (ibid.) of the metaphorical comparison. Basically, image metaphors are resemblance metaphors based on physical resemblance (cf. Lakoff & Turner 1989). ‘dead’, it is conceivable that it becomes the base for newly created metaphori- cal expansions; thus,life cyclemay be a proper label for what Croft and Cruse callthe life history of a metaphor(cf. ibid.). A central question is also how a metaphorical mapping is motivated. Grady (1997a) notes that the initial motivation for metaphorical mapping may not be the need to understand the target domain, but rather that people need to build symbols of their own conceptualisations of the target domain that others can understand (cf. Taylor 2002: 491). Taylor (2002) points out that “[b]y talk- ing about the abstract in terms of the concrete, we create the illusion of objec- tivity, and thereby facilitate communication about the abstract” (ibid.: 491-2). Abstract does not mean that certain concepts lack bodily experience, but rather that individuals experience things differently, so that communication needs a conventionalised way of speaking about more abstract concepts such as, for example, love. Thus, creating an illusion of objectivity facilitates communica- tion in such a way that it enables the verbalisation of senses (meaning) among speakers of a speech community. Croft and Cruse (2004) note that “a speaker uses an expression figuratively when he/she feels that no literal use will pro- duce the same effect” (Croft & Cruse 2004: 193). This point illustrates also that meaning depends on the context of an utterance. At the same time, it points out one aspect which is also addressed by Lakoff and Johnson (1980) in their CMT. Structuring a target domain in terms of a specific source domain results in the highlighting of certain aspects of the target domain while others are backgrounded at the same time (Lakoff & Johnson 1980: 10-4).

All in all, Lakoff and Johnson’s CMT can be summarised in the follow- ing way: (1) it deals with conventional metaphorical expressions which occur in everyday language; (2) these conventional metaphorical expressions are manifestations of the human conceptual system; (3) the metaphorical mapping between two semantic domains is unidirectional: only elements from the source domain are mapped on to the target domain and provide structure in the target domain; (4) metaphorical reasoning about concepts in the target domain is possible through the metaphorical mapping (cf. Croft & Cruse 2004: 198). In order to account for the question which elements are mapped from the source to the target domain, Lakoff (1990) proposed the Invariance Hypothesis.

Metaphorical mappings differ in their level of schematicity, i.e. some are more basic than others. Generally, in CMT it is assumed that more basic concepts, or source domains, give structure to more abstract target domains. Also, metaphors can interact with each other and in that way give rise to complex metaphor systems (Evans & Green 2006: 299).

However, Lakoff and Johnson’s CMT has been criticised in many ways. One crucial point concerns the selective projection from the source to the tar- get. Gibbs (1994) points out that the mapping theory fails to provide explana- tion for metaphors such as That surgeon is a butcher. The notion of incompe- tence is not mapped from the source domain BUTCHERY (cf. Gibbs 1994: 217). This point will be discussed in more detail in the context of Blending Theory.

One point which has not been mentioned yet is that Lakoff and Johnson (1980: 35-9) also address metonymy as a conceptual mechanism of human cognition. While in metaphor the formula is ‘A is B’ (‘B’ stands for source do- main, ‘A’ stands for target domain), in metonymy the formula is ‘B stands for A’, where ‘B’ refers to the vehicle and ‘A’ to the target (cf. Evans & Green 2006: 312). Kövecses and Radden (1998) provide a good definition of metonymy. According to them, “[m]etonymy is a cognitive process in which one conceptual entity, the vehicle, provides mental access to another conceptu- al entity, the target, within the same domain, or ICM” (Kövecses & Radden 1998: 39). An idealised cognitive model (ICM) is a “relatively stable mental representation that represents a ‘theory’ about some aspect of the world and to which words and other linguistic units can be relativised” (Evans 2007: 104).

As ICMs relate to Rosch’s Prototype Theory10 , they will not be illustrated in more detail, because this would extend the scope of this thesis. Of course, there are also cases in which metaphor and metonymy interact, Goossens (1990), for example, calls a type of metaphor and metonymy cooperation metaphtonymy. In such cases, “metaphorical and metonymic processes are recruited in the con- For example, cf. Rosch, Eleanor. “Cognitive representations of semantic categories.” In Journal of Experimental Psychology: General, Vol. 104(3), Sep. 1975, 192-233.

[...]