2.1 Introduction

Imagine that your friend just told you that ‘we have to move the party to next week’? What lies in common between this rather mundane statement and Romeo’s poetic proclamation that ‘Juliet is the sun’? Both are instances of metaphoric language. Metaphor has captivated the attention of thinkers about language for millennia, although people’s understanding of what metaphor is and how it works has shifted radically over this time. As a case in point, while current theories consider both examples that opened this chapter to be metaphorical, classical thinkers would only have considered Romeo’s proclamation to be. As we will see, a growing understanding of the workings of metaphor has given rise to the insight that, rather than being mere ornament, metaphor underpins many of the ways in which we use language, formally and informally, throughout our daily lives. In this chapter, we will survey the evolution of the theory of metaphor from its beginnings that explained metaphor as a linguistic flourish through to current views of metaphor as a mapping between two concepts. Along the way, we will see accounts of metaphor which fall out of favour, only to resurface years later alongside an expansion of phenomena that fall under metaphor’s umbrella. Included in this expansion we find the productive system of metaphors mapping space to time, to which we will turn as we draw the chapter to a close. As these metaphors are not structured as explicit comparisons, they do not fit the mould of the metaphors discussed in early accounts of the trope, rising in importance only when metaphor comes to be treated as a phenomenon of the mind.

2.2 Metaphor as Linguistic Flourish

For many of us, the notion of a metaphor – which comes from the Greek metapherein, meaning ‘to transfer’ (OED, nd b) – is not an unfamiliar one. On a general level, metaphor is a figure of speech in which one entity is described in terms of another, often quite different, entity, as in Romeo’s declaration that ‘Juliet is the sun’. As such, metaphor operates within the language system, effecting a change from a word’s ‘proper’ meaning to a new, analogically connected one. Yet, as we will see, the notion of transfer alone does not provide a sufficient explanation for how metaphor works, what speakers use it for, and how it arose.

In classical thinking, metaphor was considered to be a departure from ‘ordinary meaning’ (Cicero, ca. 86–86 BCE/Reference Caplan1954, p. 333) and ‘words used properly’ (Quintilian, ca. 90 BCE/Reference Butler1922, p. 301), along with other linguistic tools such as hyperbole, synecdoche, and catachresis. As with these other tropes, metaphor was thought to involve a conscious departure used to achieve some rhetorical – and, typically, aesthetic – end. For example, metaphor may be used to add beauty to language (Quintilian, ca. 90 BCE/Reference Butler1922), to create a vivid mental picture (Cicero, ca. 86–86 BCE/Reference Caplan1954), or to increase our pleasure from language (Cicero, ca. 55 BCE/Reference Rackham1942, p. 121 called metaphor ‘agreeable and entertaining’), thereby elevating the effect of speech. Aristotle (ca. 350 BCE/Reference Roberts2004, p. 122) argued that metaphor ‘gives style clearness, charm, and distinction as nothing else can’.

Because it was seen as the application of a word to a situation for which it was not originally intended, early thinking about metaphor was confined to the purview of individual words, which were in some sense thought to be transferred to something other than that for which they were ‘rightly’ used. The impetus for this transfer is some measure of similarity between the thing to which the word is being applied and the ‘ordinary meaning’ of the word. In this way, metaphor serves to highlight perceived connections between disparate entities, but it does not create new insights. Although metaphors were thought to be translatable to literal language, the origin of the trope was argued to be a lack of words for the precise concepts being communicated. Once metaphors had been used and found to give pleasure, early scholars reasoned, they persisted, coming to be used not just to fill lexical gaps but also for the pleasure that they gave.

Despite some attention to the origin of metaphor and the motivation for its continued use, however, the primary interest in classical thinking (and continuing through rhetorical approaches throughout the nineteenth century) was in prescribing – and, indeed, proscribing – the use of metaphor. As departures from the ‘proper’ or expected meanings of words, tropes such as metaphor were considered to be merely linguistic ornaments, as described above, and both classical thinkers and rhetoricians cautioned that metaphor must be used with care, as overuse may make speech appear artificial (Aristotle, ca. 350 BCE/1984b) or, worse, it may ‘obscure our language and weary our audience’ (Quintilian, ca. 90 BCE/Reference Butler1922, p. 309). Decoration can only be effective when it is used in moderation.

The view that metaphor is solely a linguistic flourish with its basis in the perception of similarity persisted for millennia and underlies discussions of metaphor in much of nineteenth-century rhetoric (e.g., Day, Reference Day1867; Hepburn, Reference Hepburn1875; Whately, Reference Whately1841). These scholars continued to view metaphor as a phenomenon of language, with metaphoric transfer taking place at the level of the word rather than involving a complete, contextualized utterance. Furthermore, as a consequence of their characterization as a transfer from the ‘proper’ use of a word, metaphors continued to be considered ‘deviant’ from literal language and parasitic upon ‘normal’ usage, echoing earlier, philosophical thinking (e.g., Hobbes, Reference Hobbes1651, p. 20 considered metaphor an ‘abuse’ of words; a use ‘in other sense than that they are ordained for’). An additional consequence of the classification of metaphor as an embellishment on normal language use is that there are both good and bad uses of metaphor – a view that likewise dates back to classical thinking. Thus, Aristotle (ca. 350 BCE/Reference Roberts2004, p. 125) cautions that metaphors ‘like other things may be inappropriate. Some are so because they are ridiculous…others are too grand and theatrical; and these, if they are far-fetched, may also be obscure’.

In Aristotle’s view, a fitting metaphor – that is, one that brings genuine insight, inducing an alteration of perspective rather than being misleading or obscure – is remarkably like a code, the deciphering of which brings insight and delight. Developing these ideas further in later writings, Aristotle (ca. 350 BCE/1984b) maintained that mastering metaphor is the ‘greatest thing by far’ and the ‘sign of genius’ (pp. 2334–2335), given that a good metaphor indicates an intuitive perception of the similarity in dissimilars. In this way, metaphor is viewed as a species of analogy that is based on the perception of – and, indeed, the intention to communicate – similarity (Aristotle, ca. 350 BCE/1984b; Cicero, ca. 55 BCE/Reference Rackham1942; Johnson, Reference Johnson and Johnson1981; Quintilian, ca. 90 BCE/Reference Butler1922). The ‘comparison view’ goes on to explain that metaphor, which surfaces in the form ‘X is Y’, is understood by an implicit conversion into the form of a simile, ‘X is like Y’ (e.g., ‘Juliet is like the sun’). The metaphor thus serves to highlight, by analogical extension, the commonalities between the two entities being compared, transferring properties from one to the other.

So widespread was the influence of their work that the classical approach to metaphor established a precedent in Western thought for centuries to come. Indeed, this widely held view continues to underpin the most common conception of metaphor in the popular mind, as well as in many – albeit not all – scholarly circles today.

2.3 Metaphor Comes into the Mind

Around the end of the nineteenth century, we begin to see a shift in discussions of metaphor from metaphor as language to metaphor as thought. As early as 1886, F. Max Müller began to ask what makes metaphor metaphor, and how it is that metaphor works. In addition, Müller broadened metaphor to include many instances of polysemy and transfer that were treated as separate tropes in earlier works, presaging the expansion of the phenomena considered to be metaphor in twentieth-century discussions. His writings represent a stark departure from the earlier view of metaphor as linguistic flourish.

Whereas Müller (Reference Müller1886) picks up the ancient argument that metaphor arises due to poverty in language, with people ‘strain[ing poetical imagination] to the utmost, in order to find expression for the ever-increasing wants of…mind’ (p. 621), he adds a twist that takes metaphor beyond the realm of language: Müller argues that through metaphor speakers are creating new concepts with old names (p. 622). Over time, Müller explains, words lose details of their meaning so as to become abstract enough to encompass concepts that were formerly separate. As a result, metaphor does not result from the transference that had been a hallmark of earlier accounts of the trope (p. 624) because it does not form a connection between two existent concepts. In Müller’s view, this broadening of meaning suggests that there is not a pre-existing concept to which speakers come to attach a word, but rather that it is the broadening of the word that gives birth to the new, more abstract concept. This account of the metaphoric application of word to concept likewise departs quite strikingly from earlier accounts of metaphor.

An even greater shift from language to thought is seen in the work of Gertrude Buck (Reference Buck1899). Writing at the close of the nineteenth century, Buck was dissatisfied with the view of metaphor as linguistic flourish that characterized the scholarship that preceded her. Drawing on both psychology and evolutionary biology, Buck revisits questions regarding the impetus for metaphor and the functions that metaphor fulfils. Her solutions to the problems posed by metaphor are surprisingly prescient of the work that begins to appear late in the twentieth century.

First, Buck (1899) questions the prevailing account of metaphor as linguistic ornamentation:

Further, whereas the fact that metaphor gives pleasure may explain continued use of metaphor, Buck points out that it can hardly explain its impetus, as there is no means by which a first metaphorizer could know that metaphor would be ornamental and pleasing. Buck goes on to argue that the use of pleasure as a motive is likewise psychologically implausible. Drawing on the writings of William James (‘We feel an impulse…we…act; if hindered, we feel displeasure; and if successful, relief’, cited on p. 21), Buck reasons that before creating a metaphor it cannot be known that this act of creation will give pleasure. The pleasure is an effect of metaphor, not a cause. Because this effect is unknown prior to the creation of the metaphor, it cannot be the case that metaphor arose due to a desire for pleasure, although it may have been perpetuated for this reason.

Buck (Reference Buck1899) further questions the explanation that metaphor arose due to poverty in the linguistic resources available to speakers. If this is the case, she reasoned, why was that poverty of language addressed via the creation of metaphor rather than via coinage or the use of some other figure of speech? More importantly, if metaphor did indeed arise due to poverty in the language, she questions why metaphor would persist even in cases in which there was not a lexical gap. Buck further argues that this account implies both that there was a time in human history when only concrete concepts existed (i.e., that people did not perceive the abstract), and that later, abstract concepts came into being requiring words, thereby giving rise to metaphor. Buck concludes that both implications are unlikely, with the result that this explanation fails to provide a realistic impetus for metaphor.

Drawing heavily on findings from nineteenth-century psychology, Buck (1899) posits instead that concepts grow over time in the human mind rather than appearing fully formed within a culture and then requiring names. Citing children’s overextensions as evidence that early meanings encompass what will later differentiate into smaller categories, Buck argues that ‘first impressions…were likely…large, chaotic, little-differentiated perceptions’ (p. 11). Like children, Buck reasons, metaphor creators do not notice differences between the concepts to which they give the same name. The impetus of metaphor, she reasons, is more immaturity of thought than poverty of language. As such, metaphor is not an instance of transference so much as an example of a meaning that is initially very vague and general but later becomes clearer. Metaphor, thus, emerges ‘from a primitive stage of perception’ (p. 15), a stage that may be referenced and again brought to the fore after the vague, general concept begins to separate into smaller, more specific concepts. As the metaphorizer grows in understanding, separable concepts emerge, although they may continue to be called by the same name. It is at this late stage, with the separation of concepts, that the perception of resemblance between separate concepts may arise. Buck goes on to argue that the conscious creation of metaphor involves a reinvocation of this process for artistic effect, as such reinvocation will invite the reader to trace back to the unified concept, ceasing to see the two things as separate but instead superimposing the one on the other. For example, Buck describes the metaphor ‘the moon rains out her beams’ (p. 50) as first bringing forth the image of rain, followed by a dissolution in the imagination into the image of moonbeams. In laying out her view of metaphor as an evolutionary process, Buck draws metaphor firmly out of its prior categorization as linguistic flourish, leaving it at the doorstep of the mind.

2.4 Metaphor as Interaction

A more contemporary influence on the study of metaphor was that of literary critic I. A. Richards in his work The Philosophy of Rhetoric (Reference Richards1936). In contrast with the traditional theory, which limited its application of the term metaphor to a small subset of those which are exhibited by language, Richards’ proposal followed that metaphor is fundamentally ‘a borrowing between and intercourse of thoughts, a transaction between contexts’ (p. 94). More concretely, in Richards’ view, rather than being a transference, metaphor is ‘two thoughts of different things active together and supported by a single word, or phrase, whose meaning is a resultant of their interaction’ (p. 93). To distinguish between the different constituents of metaphor, Richards proposed the terms vehicle (what is transferred) and tenor (what it is transferred to), which interact to create new meaning – one that is more than simply an amalgam of the two halves (p. 93). The metaphor, then, is the ‘whole double unit’ (p. 96), with the interaction between vehicle and tenor fundamental to the definition of what is metaphor. As such, the resultant meaning ‘is not attainable without their interaction’ (p. 100).

In addition to laying out this account of metaphor as an interaction of thoughts, Richards (1936) notes that the omnipresence of metaphor in language permeates ordinary fluid discourse (p. 60). As a result, Richards eschews the ornamental nature of metaphor purported in the Aristotelian tradition to argue that metaphor is language’s constitutive form. He goes on to draw connections between the interactions that characterize metaphor and the nature of meaning itself, which he argues is the use of signs to bring together abstract concepts which are each other’s (and the context’s) missing parts. These connections are particularly evident in his characterization of a word as ‘a substitute for…not one discrete past impression but a combination of general aspects’ (p. 100). Hence, the interactions that underlie metaphor are akin to those that underlie the meanings attached to individual words and, further, all our perceptions, our lived world, are ‘a product of earlier or unwitting metaphor’ (p. 109). According to Richards, ‘we all live, and speak, only through our eye for resemblances’ (p. 89), with the result that metaphor is not a mark of genius (as per Aristotle), but a mark of humanity.

Developing further Richards’ (1936) ideas, particularly the notion that metaphoric meaning derives from the interaction of ‘two thoughts of different things active together’ (p. 93), the philosopher Max Black (Reference Black1962) expounded an ‘interactional view’ of metaphor. According to this view, the duality of reference in a metaphoric statement is signalled by the contrast between the focus (the non-literal word or words) and the surrounding frame (which is understood literally). As exemplified by Black, in the utterance ‘the chairman ploughed through the discussion’, the focus is denoted by the word ‘ploughed’ and the rest of the sentence constitutes the frame, with the ultimate meaning resulting from an interaction of focus and frame. It is this combination of focus and frame which constitutes a metaphor.

Black (Reference Black1955, Reference Black and Ortony1979) argues that in metaphor the focal word obtains a new meaning imposed by the frame, and that the reader is forced to connect the two ideas in order to interpret the utterance. Hence, there are two subjects: the principal (later, primary) subject, about which the speaker is communicating (corresponding to Richards’ tenor), and the subsidiary subject, which the speaker uses to communicate about the primary one (corresponding to Richards’ vehicle). In uttering the metaphor, the speaker communicates about both. To illustrate, in Wallace Steven’s assertion ‘society is a sea’, the primary subject is denoted by ‘society’ and the secondary subject by ‘sea’. In addition to introducing information about ‘society’, the speaker, in uttering this metaphor, has augmented our knowledge about the sea. In this way, metaphor differs from literal language as there is a layering of subjects about which a speaker is communicating.

To interpret the metaphor, readers rely on a ‘system of associated commonplaces’: things that are generally known about the subsidiary subject that can easily be transferred to the primary subject. In the context of a particular metaphorical statement, the primary and secondary subjects interact in the following ways (Black, Reference Black and Ortony1979, p. 29): ‘(a) the presence of the primary subject incites the hearer to select some of the secondary subject’s properties; and (b) invites him to construct a parallel implication-complex that can fit the primary subject; and (c) reciprocally induces parallel changes in the secondary subject’.

Thus, in Black’s (1955, 1979) view, metaphor results from an interaction of words within a particular context, such that the new context (the frame of the metaphor) imposes an extension of meaning upon the focal word (and, often, also upon the frame). To this end, the successful application of metaphor relies on the cognizer’s awareness of the extension of meaning: they must attend concurrently to both the new and old meanings of the words in use (Black, Reference Black1955). The metaphor highlights some traits and backgrounds others, organizing the comprehender’s view of the entity about which the speaker is communicating. In this way, metaphor can bring about a change in the structures of the associated concepts.

2.5 Metaphor as Cognitive Mapping

Following on from Black (Reference Black1962), the latter part of the twentieth century gave rise to a movement in which metaphor was positioned as a cognitive phenomenon (Lakoff & Johnson, Reference Lakoff and Johnson1980; Ortony, Reference Ortony1979). Conceptual Metaphor Theory (CMT), which was pioneered by George Lakoff and Mark Johnson in their formative book Metaphors We Live By (Reference Lakoff and Johnson1980), was founded on the notion that our conceptual system is largely metaphorical and that the way we think, what we experience, and what we do every day is structured by metaphors. In this way, metaphors are not only confined to certain genres and the attempt to create some kind of artistic effect but, instead, emerge in ordinary, everyday language. This duality of role for metaphor echoes Müller’s (Reference Müller1886) distinction between novel metaphors, which ‘we use with a conscious feeling that they are our own work or the work of someone else, and that they require a kind of excuse, or even an interpretation’ (p. 618), and metaphors that are completely conventionalized, ‘so completely absorbed into the blood of a language as no longer to be felt as a metaphor’ (p. 618).

From the cognitive linguistic perspective, a conceptual metaphor consists of two domains and a mapping between them, whereby one domain is understood in terms of another. For example, in the metaphor love is a journey,Footnote ¹ love, which is the target (i.e., the domain being described), is understood in terms of a journey, which is the source (i.e., the domain being used to describe the target). To accomplish this, we draw on a fixed set of conceptual correspondences – technically called mappings – between the source and target domains, in which constituent conceptual elements of the source domain correspond systematically to constituent elements of the target domain. In the metaphor love is a journey, for instance, this set of mappings captures a coherent view of journeys that is mapped onto love: lovers may be conceived as travellers (‘they’ve come so far’), who proceed along a certain route (‘a bright future lies ahead of them’) but may encounter obstacles along the way (‘the course of true love never did run smooth’) (NOW Corpus). In this respect, the mappings bring into correspondence the elements and the relations between the elements in the journey domain (source) with the elements and the relations between the elements in the love domain (target), thereby shaping the conceptualization of love (see Table 2.1). As such, conceptual metaphors could be described as understanding one domain of experience in terms of another via a systematic set of correspondences between the two domains (Kövecses, Reference Kövecses2002).

Conceptual metaphors typically employ a more concrete or perceptually rich concept as the source domain and a more abstract concept as the target domain. That is, we conceptualize cognitively less accessible domains in terms of domains that are more easily accessible (Kövecses, Reference Kövecses, Semino and Demjén2017). This, in turn, allows for the conceptualization, expression, and communication of more abstract and complex topics in more concrete and simpler terms (Gibbs, Reference Gibbs2008); hence, we see that love, theories, ideas, and social organizations may be understood in terms of journeys, buildings, food, and plants. The reason for this, CMT proposes, is that the sensory and motor representations that derive from constant successful functioning in the natural environment are recycled in order to support abstract thought. That is, conceptual metaphors draw upon an experiential basis and are grounded in embodied experiences.

Kövecses (Reference Kövecses2002, p. 7) argues that our experiences with the physical world serve as a ‘natural and logical foundation’ for the comprehension of more abstract domains. Hence, conceptual metaphors tend to be asymmetric in terms of their directionality. For instance, although we may talk about love as a journey (e.g., ‘look how far we’ve come’; ‘I don’t think this relationship is going anywhere’) or ideas as food (e.g., ‘just food for thought’; ‘here are a few facts to digest’) (GloWbE CorpusFootnote ²), it would be unconventional and nonsensical to refer to journeys as love or food as ideas. Moreover, by drawing on additional knowledge from the source domain, we can generate new understanding of the target domain via metaphorical entailment. For instance, in the metaphor love is a journey, our knowledge about journeys can be used to infer the meaning of expressions about love, such as ‘we’ve veered off course’ or ‘our relationship is on the rocks’, or to generate novel expressions, such as ‘ain’t we ridin’ on the freeway of love’ (NOW Corpus). These expressions and other similar ones reveal the existence of an underlying system of conceptual correspondences between love and journeys (Lakoff, Reference Lakoff and Ortony1993).

While more concrete domains typically act as source domains for more abstract target domains, we do see exceptions to this tendency. However, such instances tend to occur in service of a special poetic, stylistic, or aesthetic effect (Kövecses, Reference Kövecses, Semino and Demjén2017). For instance, in Dylan Thomas’ (Reference Thomas1936) poem, A Grief Ago, an unexpected lexical item, grief, has been inserted into the slot of a fixed construction that is usually used to talk about times past (e.g., ‘a year/month/short while ago’; cf., Nattinger & DeCarrico, Reference Nattinger and DeCarrico1992). As a result, our understanding of the abstract target, grief, comes via metaphorical extension of an abstract source, time, to bring in a new meaning for grief as a measurement of time. We also see instances in which concrete source domains are deployed as a means of metaphorically describing concrete target domains. As a case in point, in Lakoff and Turner’s (Reference Lakoff and Turner1989) Great Chain of Being metaphor, concrete entities, such as humans, animals, plants, and inanimate objects, can be metaphorically conceptualized in terms of other concrete entities. In these metaphors, we see both that lower-level concepts can function as a source domain for higher-level target domains (e.g., people are animals: ‘the Big Tech fat cats got fatter and slurped down all the market’s extra Cat Chow’ (NOW Corpus)), and that higher-level concepts can provide a source domain for lower-level target domains (e.g., animals are people: ‘the Cheshire Cat is smiling’ (NOW Corpus)).

In developing an account of metaphor as a mapping between concepts, scholars have noted that, in many cases, there are aspects of our knowledge about the source domain that fail to be transferred to the target domain and, furthermore, there are aspects of our knowledge of the target domain that do not have their origins in the source domain. For example, while love is like a journey in some respects, it is not a journey in other respects – it does not take place on a train or a boat and we do not buy tickets to embark on it (Ruiz de Mendoza, Reference Ruiz de Mendoza2009; Vega Moreno, Reference Vega Moreno2007). These gaps in the mapping of concepts suggest that there are constraints in terms of what can be mapped systematically from source to target domain . To account for this, Lakoff (Reference Lakoff1990, Reference Lakoff and Ortony1993) proposed the Invariance Principle, which stipulates that the metaphorical mappings between source and target domain will preserve the image schematicFootnote ³ structure of the source domain in a way that is topologically consistent with the structure of the target domain. In this way, when paths are metaphorically mapped, sources will be mapped onto beginnings, goals onto end points, trajectories onto progressions through change, etc. Thus, if we state that ‘the meeting went from 8.30pm to 10.00pm’ (GloWbE Corpus), the path’s source is mapped to the beginning of the temporal period and the goal is mapped to the end.

A corollary of the Invariance Principle is that inherent target domain structure cannot be violated and, thus, automatically limits both what can be mapped from the source domain and how the mapping will play out (Lakoff, Reference Lakoff and Ortony1993). This explains why you may give someone information while retaining it for yourself or why you can give a person a kick without them possessing it afterwards. In these examples, which instantiate the metaphor actions are transfers, an action, such as a kick, is conceptualized as an object transferred from an agent to a patient. While the source domain dictates that once an object has been given, the recipient possesses it, the target domain dictates that an action ceases to exist after it has occurred. As such, the inherent structure of the target domain precludes the existence of the action once it has taken place. In this regard, the Invariance Principle constrains the mapping whereby an action is seen as a transfer of possession, such that the possession component, lacking a corresponding element in the target domain, is not itself transferred. However, as noted by Ruiz de Mendoza (Reference Ruiz de Mendoza, Croitoru, Tuchel and Praisler2005), the consequence of giving someone a kick is that the person is affected by the action. To convey this meaning, ‘kick’ (the figurative object of the transfer of possession) must be developed metonymically into ‘effects of the kick’; thus, the receiver of the kick figuratively possesses the effects of the kick rather than the kick itself after the action has taken place. The underlying principle here, which has been termed the Mapping Enforcement Principle, entails that no item in either domain of a metaphoric mapping will be discarded if there is a meaningful way to find a corresponding item in the other domain.

One limitation of the Invariance Principle is that it accounts solely for constraints on metaphorical mappings that are structured by image schemas, thus giving rise to the Extended Invariance Principle (Ruiz de Mendoza, Reference Ruiz de Mendoza1998). According to this principle, the generic-level structure of the source domain, and not merely its image schematic structure, is available for mapping and must be preserved in a way that is consistent with the inherent structure of the target domain. For example, with strength is durability, attributes of a material are often mapped onto corresponding human attributes (e.g., ‘she has nerves of steel’; ‘I also have asbestos fingers’) (iWeb CorpusFootnote ⁴). Similarly, with people are animals, characteristic animal behaviour is often mapped onto corresponding human behaviour (e.g., ‘he is an absolute shark in the boardroom’; ‘I am nothing but a snake in the grass’) (GloWbE Corpus). Neither metaphor invokes image schematic structure, yet the generic-level structure of the source and target domains is preserved such that attributes and behaviour in the source domain map onto attributes and behaviour in the target domain (Peña Cervel & Ruiz de Mendoza, Reference Peña Cervel, Ruiz de Mendoza, Panther, Thornburg and Barcelona2009; Ruiz de Mendoza & Pérez-Hernández, Reference Ruiz de Mendoza and Pérez-Hernández2011).

One further constraining factor in metaphoric mappings, termed the Correlation Principle (Ruiz de Mendoza & Santibáñez Sáenz, Reference Ruiz de Mendoza and Santibáñez Sáenz2003), specifies that in order for a source domain concept to qualify as a target element counterpart, it must share all the relevant implicational structure of the target element in the context in which it is produced. For example, in the case of a marriage, the notion of a journey is an appropriate source if focusing on the success of a relationship in terms of reaching common goals but not if focusing on the exclusivity of the relationship. In this way, we might say that to know a metaphor means to unconsciously know the systematic mappings between a source and a target (Kövecses, Reference Kövecses2002), including their constraints.

2.6 Metaphor as Conceptual Blending

Metaphoric mappings alone, however, cannot fully account for all possible interpretations available to the language user. Building on Fauconnier’s (Reference Fauconnier1994) Theory of Mental Spaces, Blending Theory (BT) is an approach to language and communication which maintains that the construction of meaning involves the conceptual integration of the source and target domains into one; a new ‘blended’ mental space (Fauconnier & Turner, Reference Fauconnier and Turner1998, Reference Fauconnier and Turner2002; cf., Black, Reference Black1962; Richards, Reference Richards1936). Whereas in CMT, metaphors are positioned as stable and systematic relationships between two conceptual domains (the source and target), in BT, the basic unit of cognitive organization is not the domain but the mental space; a partial and transitory representational structure that is constructed online when people think and talk. As these structures are created in particular situations for the purpose of dynamic meaning construction, they are smaller and more specific than conceptual domains. Thus, while CMT tends to be more focused on conventional patterns of metaphorical conceptualization and their motivated extensions, BT tends to be more concerned with novel conceptualizations, such as those occurring in jokes, newly coined words, words we use in unconventional ways, etc., which do not arise from entrenched cross-domain relationships and may be short-lived.

Rather than modelling metaphor as a mapping between two conceptual domains, BT employs a four-space model: two input spaces (which, from a CMT perspective, are associated with the source and target domains); a generic space, which contains the conceptual structure shared by both inputs; and the blend space, which contains material from the input spaces that combines and interacts. Thus, whereas CMT posits a simple, unidirectional projection of mappings between source and target domains, the four-space model posited by BT involves the projection of material from both the source and target spaces to the blend.

Blending also involves three basic processes: composition, which involves the projection of content from each of the inputs into the blended space; completion, which involves filling in patterns in the blend to achieve a complete scenario by drawing on knowledge from our conceptual system; and elaboration, which is the simulation of the event in the blend. In addition to inheriting structure from each input space, the blend may develop emergent content of its own from the juxtaposition of elements from the inputs at any point in the process (Fauconnier & Turner, Reference Fauconnier and Turner1998), thus lending dynamicity and flexibility to the process of creating or interpreting a metaphor. In this way, blending can account for the emergence of features that are not present in either of the input spaces. For example, in the sentence, ‘my surgeon is a butcher’ (GloWbE Corpus), the implicature of incompetence emerges, despite not being a feature of surgery or butchery. This stems not from the projection of source to target but from the juxtaposition of the means of butchery with the goal of surgery that happens in the blend. Thus, in addition to offering a new account of the mappings involved in metaphor, BT aims to offer a general cognitive model for meaning-making and, specifically, the emergence of novel conceptualizations such as those seen in metaphor.

2.7 Metaphor Identification

With the introduction of theoretical treatments of metaphor such as CMT, a wide range of commonly used metaphorical language has been identified. However, some critics of CMT view many of the examples cited in expositions of the theory (such as ‘I don’t think this relationship is going anywhere’ and ‘a bright future lies ahead for them’ (GloWbE Corpus)) as conventional expressions that are understood literally rather than via metaphor (e.g., Jackendoff & Aaron, Reference Jackendoff and Aaron1991). Indeed, experimental research shows that people’s judgements of potentially metaphorical meanings are highly variable and dependent on the specific judgement task and the contexts within which these utterances are embedded (Gibbs et al., Reference Gibbs, Buchalter, Moise and Farrar1993; Gibbs & Colston, Reference Gibbs and Colston2012; cf., Julich-Warpakowski, Reference Julich-Warpakowski2022). Taken together, these findings suggest that it may be quite difficult to draw general conclusions about people’s use and understanding of ‘metaphorical’ language as opposed to other types of linguistic meaning (Gibbs, Reference Gibbs2017a). Indeed, rather than being categorically different, literal and figurative language may lie along a continuum. Müller (Reference Müller2008) offers a related view, centred on the idea that metaphors oscillate between various degrees of sleeping and waking, with their status changing depending on context and intention.

In addition to experimental research showing variability in judgements of metaphoricity, evidence for this continuum may be seen in other communicative behaviours. For instance, the conventionality of referring to a person who is extremely sad as ‘depressed’ may suggest that the association between literal depression (i.e., downward motion) and sadness is opaque or dead. However, the existence of co-speech gestures depicting a slow and downward motion accompanying this use of ‘depression’ may be viewed as reflecting a conceptual understanding of the metaphor sadness is down (Gibbs & Colston, Reference Gibbs and Colston2012). In this way, metaphorical thinking may really be a matter of cognitive activation for specific individuals in particular moments of speaking and listening (Gibbs & Santa Cruz, Reference Gibbs and Santa Cruz2012; Müller, Reference Müller2008) rather than a purely linguistic phenomenon. As a result, a full understanding of metaphor must go beyond the utterance to seek the full range of ways in which the characterization of one domain may be structured by another. This dynamic view of metaphor offers an account of the metaphoricity of so-called dead metaphors, which people may use without awareness of the metaphorical source domain, while still displaying evidence of retaining critical metaphorical knowledge motivating a word or phrase’s use in context. Thus, determining the extent to which any metaphor is dead, sleeping, or alive necessitates an examination of an individual’s language, speech, gestures, and other bodily actions in a given moment, as opposed to language alone (Gibbs & Colston, Reference Gibbs and Colston2012).

One challenge highlighted by the continuum from the literal to the metaphorical concerns the identification and exposition of metaphoric language. Variability in intuitions and a lack of agreed criteria for metaphor identification are thought to complicate any evaluation of theoretical claims about the possible relations between metaphoric language and metaphoric thought, as well as its frequency and organization in discourse (Pragglejaz Group, 2007). Metaphor identification is further complicated by the fact that metaphorical concepts, such as allegory, are not always evidenced through overtly metaphorical language (Pragglejaz Group, 2007). The traditional and familiar ‘X is Y’ format, known as the resemblance metaphor (Ureña & Faber, Reference Ureña and Faber2010), has been shown to account for just a small percentage of the metaphors that appear in discourse (Cameron, Reference Cameron2003) and, as we have seen, does not account for the majority of the metaphors discussed in the development of CMT, including many that form the basis for the studies reviewed throughout this book.

2.8 Metaphorical Realities

In light of both the connections between domains and the resultant language interpretations characterized by metaphor, conceptual metaphors represent both a process and a product: the cognitive processes involved in understanding a domain form the process, and the resulting conceptual patterns form the product. This duality provides a natural explanation for the range of metaphoricity that we see along the continuum from literal to metaphorical language. Furthermore, according to Kövecses (Reference Kövecses, Semino and Demjén2017), when we conceptualize a less tangible domain from the perspective of a more tangible domain, we create a certain metaphorical reality. We imagine love in one way when we think of it as a journey (cf., Table 2.1) and in other ways when we conceive of it as a war (‘they’re not just fighting for their relationship’), as a nutrient (‘he was starved of affection’), or as an economic exchange (‘I’m invested in this relationship’) (COCAFootnote ⁵). In essence, the deployment of these different source domains gives rise to the creation of different metaphorical realities of love. Consequently, some aspects of the target domain become highlighted while other aspects are de-emphasized or hidden through the choice of a specific metaphor, thereby encouraging specific patterns of inference (Lakoff & Johnson, Reference Lakoff and Johnson1980; cf., Black, Reference Black and Ortony1979). For example, love is war highlights aspects of the conflicts encountered with love and hides aspects of cooperation. In this way, we see that metaphors provide a vital conceptual tool for creating realities within the social world; reflecting and influencing how people understand different types of experiences as well as how they consequently behave (as shown below). This bias in the process of conceptualization constitutes the ‘framing’ power of metaphor (Semino et al., Reference Semino, Demjén and Demmen2018). The notions of frame and framing are widespread across various fields of scientific enquiry, ranging from sociology (Goffman, Reference Goffman1974) and cognitive semantics (Fillmore, Reference Fillmore1976), to media studies (e.g., Dimitrova & Stromback, Reference Dimitrova and Stromback2005; Druckman, Reference Druckman2001). While one single definition is not agreed upon, a frame is broadly construed as a portion of background knowledge that ‘(i) concerns a particular aspect of the world, (ii) generates expectations and inferences in communication and action, and (iii) tends to be associated with particular lexical and grammatical choices in language’ (Semino et al., Reference Semino, Demjén and Demmen2018, p. 3).

As these examples show, many abstract domains may act as targets for multiple metaphors, with subtle differences in meaning between them. Building on this insight, it has been suggested that metaphors can be employed as framing devices to affect reasoning. The effects of metaphorical framing on reasoning have been the focus of a variety of experimental studies (e.g., Gibbs, Reference Gibbs2006; Hartman, Reference Hartman2012; Landau et al., Reference Landau, Sullivan and Greenberg2009; Sopory & Dillard, Reference Sopory and Dillard2002; Thibodeau & Boroditsky, Reference Thibodeau and Boroditsky2011, Reference Lai and Boroditsky2013). For instance, metaphorically framing crime as a beast or as a virus has been found to lead people to propose different solutions to a city’s crime problem in ways that are consistent with the entailments of the metaphors. As a case in point, Thibodeau and Boroditsky (Reference Thibodeau and Boroditsky2011, Reference Lai and Boroditsky2013) found that while crime is a virus metaphors encouraged preferences for social reform, crime is a beast metaphors encouraged preferences for enforcement laws and punishment measures (see also Steen et al., Reference Steen, Reijnierse and Burgers2014). Similarly, using the term guardian instead of warrior to metaphorically characterize the role of police officers has been shown to encourage more positive attitudes towards law enforcement (Thibodeau et al., Reference Thibodeau, Crow and Flusberg2017). These findings suggest that metaphorical framing may be used to shape people’s reasoning as well as influence their solutions to different societal issues (see Boeynaems et al., Reference Boeynaems, Burgers, Konijn and Steen2017 for an overview). Moving beyond the lab, discourse-focused metaphor research has provided important insights into the real-world effects of framing via metaphor choices on a wide range of social and political issues, such as climate change (Nerlich, Reference Nerlich2012), crime (Cotterill, Reference Cotterill2003), and health emergencies (Olza et al., Reference Olza, Koller, Ibarretxe-Antuñano, Pérez-Sobrino and Semino2021), as well as migration (Charteris-Black, Reference Charteris-Black2006), taxation (Lakoff, Reference Lakoff2004), and terrorism (Pinelli, Reference Pinelli2016). Like the earlier laboratory studies, this research has shown how metaphor choices may frame issues in particular ways, revealing evaluative stances and ideological meanings that may not be overtly expressed in language and constructing particular courses of action as natural or necessary (Bogetić, Reference Bogetić2019). Because a single target domain may be structured via a variety of source domains, metaphor can facilitate different ways of framing experience, and a deeper understanding of metaphor is critical to building our understanding of those domains of experience that surface as the targets in metaphorical conceptualizations.

2.9 Time in the Words of Space

How do people talk – and potentially think – about time? Speakers of English, as well as many other languages, make use of a wide range of metaphors to talk about time, such as time is money (‘how do you spend your spare time?’), time is a resource (‘we have limited time to act’), and time is a valuable commodity (‘cherish the time you have’) (NOW Corpus; Kövecses, Reference Kövecses2002; Lakoff & Johnson, Reference Lakoff and Johnson1980). Time is also regularly personified as a thief (‘time, the subtle thief of youth’; Milton, Sonnet 7), a pursuer (‘at my back I always hear time’s winged chariot hurrying near’; Marvell, To His Coy Mistress), and a devourer (‘time, the devourer of everything’; Ovid, Metamorphoses 15) (cited in Kövecses, Reference Kövecses2002, pp. 55–56). However, one system of metaphors for time that is particularly prominent is the time is space system (e.g., Clark, Reference Clark and Moore1973; Evans, Reference Evans2004; Filipović & Jaszczolt, Reference Filipović and Jaszczolt2012; Haspelmath, Reference Haspelmath1997; Lakoff & Johnson, Reference Lakoff and Johnson1999; Núñez & Sweetser, Reference Núñez and Sweetser2006; Traugott, Reference Traugott, Greenberg, Ferguson and Moravcsik1978; Yu, Reference Yu1998), whereby representations in the domain of time are thought to arise from the human experience of navigating through, orienting within, and observing motion and location in space.

As seen in Table 2.2, examples demonstrating that language from the relatively concrete and perceptually rich domain of space is recruited to talk about the abstract domain of time abound. These linguistic expressions reveal numerous regularities in spatial metaphors for time (see, among others, Clark, Reference Clark and Moore1973; Kranjec, Reference Kranjec, Forbus, Gentner and Regier2006; Lakoff & Johnson, Reference Lakoff and Johnson1999; Moore, Reference Moore2006, Reference Moore2014; Núñez et al., Reference Núñez, Motz and Teuscher2006), suggesting systematic conceptual relations between the two domains. For instance, moments can be conceptualized as points (at midday; on Saturday) and durations as distances (a long meeting; a short journey), events can be moved (the wedding was brought forward; the interview was pushed back), and so forth.

Despite the wide-ranging use of spatial language to talk about time, researchers have noted important limitations. Most importantly, space involves three dimensions – length, width, and height – and spatial terms may presuppose one (e.g., ‘long’/’short’ and ‘in front’/’behind’), two (e.g., ‘wide’/’narrow’, ‘beside’), or all three (e.g., ‘tall’/’short’, ‘above’/’below’) (Clark, Reference Clark and Moore1973). However, it is those that presuppose one dimension, such as in front/behind or long/short, that tend to be extended from space to time (Clark, Reference Clark and Moore1973; Evans, Reference Evans2004, Reference Evans2013; Lakoff & Johnson, Reference Lakoff and Johnson1999; Tenbrink, Reference Tenbrink2007; Traugott, Reference Traugott, Greenberg, Ferguson and Moravcsik1978),Footnote ⁶ suggesting that time is conceptualized, at least in English, as a directed axis.

Among those metaphors that characterize time as a directed axis, particular attention in the existing literature has been paid to two dominant spatial metaphors that are used in the representation of time: Moving Time and Moving Ego. From the Moving Time perspective, which derives from the metaphor time passing is motion of an object, time is conceptualized as a conveyor belt on which events move, from the future to the past, relative to a stationary Ego. The conceptual correspondences suggested by this metaphor are depicted in Table 2.3.

In contrast, from the Moving Ego perspective, which stems from the metaphor time passing is motion over a landscape, time is construed as a stationary landscape that the active Ego moves across. The conceptual correspondences suggested by this metaphor can be captured in the form shown in Table 2.4.

As indicated by these mappings, for both metaphors, the Ego’s location provides the deictic reference point – a position which correlates with the present at any given time; hence, we talk of the ‘here and now’.Footnote ⁷ Similarly, in both metaphors the space in front of the Ego maps to the future, while the space behind the Ego maps to the past. This alignment of time along a directed axis, which underlies both the Moving Ego and the Moving Time metaphors, is applicable not only to dynamic predications but also to static ones, such as ‘Christmas is in front of us’ (Tenbrink, Reference Tenbrink2011).

Linguistically, both the Moving Time and the Moving Ego metaphors can be used to depict events in the past, present, and future:

1. (1)

   a.	Summer is approaching [us].
   b.	The moment of truth has arrived.
   c.	The deadline has passed [us].

1. (2)

   a.	We’re approaching summer.
   b.	We’ve arrived at the moment of truth.
   c.	We’ve passed the deadline.

   (NOW Corpus)

Moreover, as demonstrated by the expressions above, the Moving Time (1) and Moving Ego (2) perspectives are inverse to one another: either the temporal event is depicted as moving relative to a stationary Ego (which is not always explicitly encoded in natural speech), or the Ego is depicted as moving relative to a fixed temporal event. As a result of their metaphorical movement, temporal events have implied fronts and backs in the Moving Time metaphor, but they are unfeatured in the Moving Ego metaphor, thus giving rise to the following mappings:

Following on from the fact that the two metaphors differ with regards to which entity is in motion, the trajectories of motion in the two metaphors are the reverse of one another: while Moving Ego is metaphorically directed towards the future, Moving Time is metaphorically directed towards the past. In this way, the Moving Ego perspective more closely aligns with our sensorimotor experience of self-motion: as we move forward in space towards objects in the distance, we move forward in time towards events in the future (Radden, Reference Radden, Baumgarten, Böttger, Motz and Probst2004).Footnote ⁸

In addition to descriptions of temporal events as located relative to Ego, there are a multitude of descriptions that locate temporal events relative to one another with no reference to the ‘now’. Rather, these descriptions merely depict an earlier/later relationship between the events:

1. (3)

   a.	Spring comes before/is before summer.
   b.	My birthday comes after/is after yours.

   (Google Books)

The differences between views of time that make reference to Ego and views of time that are perspective-free can be traced back to the philosopher John McTaggart (Reference McTaggart1908), who proposed the fundamental distinction between these two basic classifications of time. The first classification, ‘A-series’ (or ‘deictic’ time), characterizes temporal events that are established in relation to the Ego by virtue of being in the future, present, or past; thus, deictic time depicts a constant change in the status of temporal events. The second classification, ‘B-series’ (or ‘sequential’ time), characterizes temporal events that are established in relation to one another by virtue of being earlier or later; thus, sequential time depicts temporal events as a static chain, like beads strung together on a necklace, whose statuses never change.

Importantly, while deictic time facilitates a past/future temporal relationship, this relationship is absent without a deictically grounded ‘now’. In contrast, sequential time exhibits an earlier/later relationship that cannot be accounted for by the Moving Time and Moving Ego metaphors. To capture this important distinction, a fundamental reclassification was devised. Instead of classifying temporal metaphors on the basis of which entity is moving – time or Ego – a classification based on the relevant reference point was proposed as a means of differentiating past/future from earlier/later relationships, resulting in Ego-Reference-Point (Ego-RP) metaphors (of which the Moving Time and Moving Ego metaphors are subcases) and Time-Reference-Point (Time-RP) metaphors (Moore, Reference Moore2006; Núñez & Sweetser, Reference Núñez and Sweetser2006; Núñez et al., Reference Núñez, Motz and Teuscher2006; see also Evans, Reference Evans2004). An additional Ego-free representation, extrinsic time, was further proposed by Kranjec and colleagues (Kranjec, Reference Kranjec, Forbus, Gentner and Regier2006; Kranjec & McDonough, Reference Kranjec and McDonough2011). Unlike in sequential time, in which the intrinsic fronts of events are facing towards earlier times (with earlier events positioned in front of later ones), extrinsic time is argued to encompass a path leading from earlier to later times (with later events positioned in front of earlier ones), akin to reading a calendar (see Figure 2.1).Footnote ⁹ As we shall see in Chapter 3, these metaphoric spatializations of time have inspired studies probing the conceptual connections between spatial and temporal reasoning.

Figure 2.1 Deictic and non-deictic spatial construals of time

(adapted from Kranjec & McDonough, Reference Kranjec and McDonough2011, p. 738)

2.10 Chapter Conclusion

The changing conception of metaphor over two millennia of scholarship has led to a concomitant shift in the questions being asked about this trope. Early work positioning metaphor as merely a linguistic flourish focused on questions of why people created metaphor, how metaphor gives pleasure, and how to prescribe and proscribe its usage. With the movement of metaphor into the mind, questions have likewise shifted to focus on the intercourse between conceptual domains thought to be the hallmark of metaphor, and to cognitive and behavioural consequences of choosing a particular metaphorical framing for a concept. The introduction of accounts treating metaphor as a cognitive phenomenon rather than merely a linguistic juxtaposition of dissimilars has likewise opened the door to the study of conceptual mappings that surface across a wide array of linguistic constructions. One result is that many metaphorical systems have been identified in which there are few expressions of the form X is Y, including the particularly productive system of correspondences by which time is conceptualized and described in terms of space reviewed in Section 2.9. Taken together, these expansions have given rise to a large and active field of metaphor studies.

The vast quantity, extraordinary variety, and richness of recent research on metaphor has carved a field that has often appeared fragmented and overwhelming (Semino & Demjén, Reference Semino and Demjén2017). Indeed, in more recent years, the dominant framework of CMT has been developed and modified (e.g., Grady, Reference Grady1997; Kövecses, Reference Kövecses2015), tested (e.g., Gibbs, Reference Gibbs1994), critiqued (e.g., Murphy, Reference Murphy1996), and applied in a variety of ways (e.g., Charteris-Black, Reference Charteris-Black2004). This continual revisiting of the question of what exactly metaphor is adheres to a long-standing tradition in the study of metaphor not only in the twentieth century (e.g., Black, Reference Black1962; Richards, Reference Richards1936) but also in the earlier works of philosophers and rhetoricians dating back to classical times (see Jäkel, Reference Jäkel, Gibbs and Steen1999). As observed by Jäkel (Reference Jäkel, Gibbs and Steen1999), who provides an overview of some of the forgotten contributions to CMT, many of the central tenets are not new but, rather, draw from an extensive ancestry of the cognitive approach (e.g., Blumenberg, Reference Blumenberg1960, Reference Blumenberg1971; Kant, 1790/Reference Kant and Bernard1951; Weinrich, Reference Weinrich and Weinrich1958, Reference Weinrich1963, Reference Weinrich and Weinrich1964, Reference Weinrich and Weinrich1967, Reference Weinrich and Weinrich1976).

While metaphor has been defined, theorized, and applied in different and often mutually incompatible ways, there is, however, a generally broad consensus that metaphor involves the perception of correspondences between dissimilar entities and processes that enables us to see, experience, think, and communicate about one thing in terms of another (Semino & Demjén Reference Semino and Demjén2017, p. 1). If these connections are indeed conceptual, however, there should be evidence beyond the appearance of metaphorical statements. Thus, one particularly important direction that the study of metaphor has taken in recent years is the search for evidence of the psychological reality of the connections resulting from these correspondences. This line of research, in turn, has expanded the focus of study to include not only the producer of metaphor but also the receiver. It is to these directions that we now turn.