MA Thesis - July, 2018
University of Amsterdam Graduate School of Humanities
MA Philosophy
Metaphor in Interaction
A constructivist and enactivist approach to metaphor in
cognition
Table of Contents
Introduction
Chapter 1 - Language and the socio-cognitive nice 1.1 Social cognition and the 4Es
1.2 Intersubjectivity, social interaction and “we-space” 1.3 Language
Chapter 2 - Theories of metaphor 2.1 Conceptual Metaphor Theory
2.2 Embodied and socially embedded metaphors 2.3 Novel metaphors and creative cognition
Chapter 3 - Cognitive niche construction and social interaction
3.1 Niches, operationally closed and self-regulating autonomous systems 3.2 The symbols-niche of metaphor in social interaction
Conclusion
Introduction
Human beings engage in metaphorical language and reasoning in their everyday communication with others, and in their way of conceptualizing certain aspects of their everyday experience. The capacity for metaphorical abstraction seems to be identifiable cross-linguistically and cross-culturally, even if the actual mappings may differ. In the second half of the last century, much attention has been drawn to this subject matter, and research and debate about it has been emerging in cognitive philosophy, cognitive psychology, cognitive linguistics, argumentation theory, discourse analysis, philosophy of language and mind. By now, many seem to agree that metaphor must no longer be considered as a purely linguistic tool with rhetorical or illustrative purposes; rather, metaphor is considered as a mechanism deeply rooted in our cognitive and conceptual system.
In the field of cognitive linguistics, Lakoff and Johnson have been amongst the first 1 to consider metaphor beyond its mere linguistic use and as a mechanism fundamentally linked to cognition. They believe that the whole conceptual system of human cognitive agents is intrinsically metaphorical, and that this explains how and why we conceptualize the world through metaphorical relations that are mapped onto one another on the basis of their relatable experiential bases. Conceptual Metaphor Theory (CMT), therefore, is aimed at2 showing that the individual’s cognitive schemes are fundamentally linked to metaphor creation, use and understanding, and that metaphorical language arises from the conceptual system of the brain/mind with connections to its neural activity. Lakoff and Johnson believe that the concepts that cognitive agents hold shape and guide the way in which they perceive the world, they act within it and they relate to other agents. Therefore, by stating that the human conceptual system is metaphorical in character, the authors maintain that the way we experience reality and act within it is inextricably linked to metaphorical conceptualization. 3 Metaphorical mappings are understood as sets of correspondences between (elements of) different domains, hence they allow to understand (aspects of) a domain in terms of (aspects of) a different domain. Lakoff and Johnson believe that there are sets of ontological 4 correspondences between the domains involved in the mappings, that allow agents to map their knowledge of one domain onto their knowledge of the other.
Despite its crucial importance in metaphor studies, CMT has been criticized for not considering enough the role that the social environment surrounding the individual - and her 1 Lakoff & Johnson, 1980a; 1980b; 1980c.
2 Lakoff & Johnson, 1980a; 1980b; 1980c. 3 Lakoff & Johnson, 1908b, p. 454. 4 Lakoff, 1993, p. 5.
constant relations with the social other - has in determining, shaping and influencing metaphorical thought and language. Within their approaches to aspects of social and embodied cognition, some scholars have advanced claims regarding the direct influence that5 metaphor bares on socio-cognitive processes and, viceversa, how social cognition can influence metaphor use and understanding. From this broad perspective, CMT has been criticized in its excessive focus on individual cognitive processes as they emerge and unravel internally to the human mind. Embodied and socially embedded theories of metaphor, instead, focus on how social aspects of intersubjectivity and of interaction with the environment play a fundamental role in shaping (social) cognition and metaphor use and understanding.
In the present study, I would like to argue that theories of embodied metaphor and of metaphor in social cognition do not overcome the ‘internalist’ perspective on cognition that 6 they pinpoint as a fundamentally problematic aspect of CMT. As I will discuss throughout this thesis, CMT, the socially enriched social cognition theory and the embodied theory of7 metaphor seem to share a neglect of how the mechanism of metaphor emerges in interaction8 and shapes interaction between individual cognizers and between cognizers and their environment. To maintain an ‘internalist’ perspective on cognition, and to understand metaphorical mappings as occurring on the basis of ontological correspondences between different domains - that exist independently of cognizers’ action - seems grounded in a view of the world as a static and non-modifiable reality that pressures from the outside the organism to evolve, and on which the cognizer has no crucial influence. Instead, the individual should be considered in its active interactions and relations with its surroundings, in its ability to manipulate and be manipulated by the environment and by the social other.
In this study, I would like to explore a new perspective on the workings of metaphor in cognition, in order to account for the fundamental role of interaction in allowing for, and influencing, metaphorical conceptualization altogether. I will proceed by bringing together insights from the enactive approach to cognition, the constructivist approaches to cognitive niche and the interactionist perspectives on metaphor creation and understanding.
Enactive approaches consider cognition to be fundamentally dependent on the action of cognizers . Whilst embedded perspectives consider the tight influence of the environment9 in shaping cognitive processes, the enactive view focuses on the idea that the cognitive agent
5 The authors include Gibbs et al. (2004; 2010;), Landau et al. (2010), Thibodeau & Boroditsky (2011). 6 De Jaegher et al., 2010.
7 Landau et al., 2010.
8 As developed by Gibbs et al. (2004; 2010). 9 Ward & Stapleton, 2012, p. 89.
is one who establishes boundaries and distinctions between itself and the environment.10 Cognitive niche construction theories abandon the idea of the world as an unmodifiable reality and embrace the perspective of the crucial action of the individual in modifying and transforming its cognitive processes and capacities. Finally, interactionist approaches to11 metaphor understand metaphor as a creative mechanism that operates at many different levels of cognition - not only at the linguistic level - through the creation and re-definition of meaning, allowing for new perspectives on the world to emerge continuously. By linking 12 together insights from these theories of metaphor and cognition, I would like to explore the possibility of metaphor as a new cognitive niche, that allows cognitive agents to develop new forms of reasoning, communicating and problem solving in a way that would be inconceivable outside of metaphorical conceptualization.
10 De Jaegher et al., 2010, pp. 441-443. 11 Bertolotti & Magnani, 2016, p. 4796.
Chapter 1
Language and the socio-cognitive niche
In the present chapter I will deal with so-called 4 E approaches to cognition. These approaches have been developed in response to classical cognitivism, in order to account for the role of the body and of the environment on the cognitive performances and properties of human cognizers. In what follows, I will give a brief definition of embodied, embedded, enactive and extended theories of cognition. Further, I will sketch the approach that these theories hold with respect to social cognition, and how they deviate from traditional accounts. With regards to social cognition I will mainly focus on the socio-cognitive process of language use and understanding. Finally, I will focus specifically on aspects of enactive cognition, extended cognition and interaction theory in order to delineate what I believe can be defined as a ‘socio-cognitive niche’.
1.1 Social cognition and the four Es
In the past decades, the study of human cognition has branched out and developed in different approaches within cognitive science and philosophy, but, as of now, this has not led to the development of a single comprehensive approach. The approaches that will be outlined and dealt with in this Chapter are ones that have developed in the past two decades in response to traditional and mainstream theories on cognitive processes, and which fall under the label of 4E cognition theories. These consider cognition to be embodied, embedded, enactive and
extended. The common critical goal underlying the development of 4E cognition theories is to account for the active role that the body, the environment and interactions between agents play in shaping and guiding cognitive processes, against the classical idea of cognition as a process of computation and manipulation of internal representations. It is important to 13 underline that these are not mutually exclusive theories, as they usually draw upon aspects of one another. For example, the extended mind theory of Clark and Chalmers is based on an 14 embedded and embodied view of cognition, and interaction theories are based on an15 embodied and enactive view of cognitive performance. Because of the plurality of perspectives and interpretations, it is useful to provide a very brief overview of the core claims advanced by 4 E cognition theories of mind, before moving on to their positions with regards to social cognition, where more fundamental ideas and implications of these theories will be presented.
13Shepherd, 2012, p. 507. 14 Clark & Chalmers, 1998.
Embodied cognition theories hold, in very general terms, that embodied practices and bodily anatomy are what cognitive performance is based on. The structure and categories 16 that characterize human cognition are shaped by aspects that characterize our existence as bodily agents. This means that bodily anatomy and activity are contended to play a causal17 role in cognitive processes. 18
Embedded approaches, in addition to considering the embodied character of
cognizers, focus on the fact that cognitive agents are situated in their surrounding environment. Therefore, embedded cognition theorists hold that it is our relationships with the environment - and our embodiment - that crucially shape cognitive performance and properties. Cognitive processes - regarding action, thought, perception, communication, and19 so on - are intimately coupled with the surrounding environment, and the very nature of cognition is fixated by the relations holding between the brain, the body and the world. 20 21
Enactive theories of cognition also agree that cognition does not correspond to processes that determine internal representations of states of affairs. Rather, cognition depends entirely on the behaviour and action of the cognizer. So intended, cognitive process 22 are those through which an agent self-regulates and self-identifies in situations that are precarious. These concepts will be defined more clearly shortly ahead in this Chapter. For the present purpose it suffices to say that enactive approaches consider cognition to be fundamentally dependent on the action of cognizers.23 Whilst embedded perspectives consider the tight influence of the environment in shaping cognitive processes, the enactive view focuses on the idea that the cognitive agent is one who establishes boundaries and distinctions between itself and the environment. 24
Finally, the extended cognition theories - sometimes also defined as “active externalism” - hold that elements external to the mind can play an active role in shaping and25 constituting cognitive processes. In some cases, such external elements can also be considered as being cognitive processes themselves. These may be artifacts, institutions, 26 technological tools, and so on, and they are believed to enable action and perception, constitute part of cognitive processes and enhancing our biologically constrained capacities. 27 16 Ward & Stapleton, 2012, p. 82.
17p. 98.
18 Goldman & De Vignemont, 2009, pp. 154-155. 19Ward & Stapleton, 2012, p. 89.
20 The distinction between ‘body’ and ‘brain’ is operated because “letting the brain qualify as part of the body
would trivialize the claim that the body is crucial to mental life” (Goldman & De Vignemont, 2009, p. 154).
21 Kiverstein & Clark, 2009, p. 2. 22 Thompson & Di Paolo, 2014, p. 73. 23 Ward & Stapleton, 2012, p. 89. 24 De Jaegher et al., 2010, pp. 441-443. 25 Clark & Chalmers, 1998, p. 1. 26 p. 2.
As mentioned above, 4E approaches to cognition aim at offering an alternative to classical cognitivism, and to traditional accounts of social cognition. Social cognition identifies with forms of cognition or of action that regard agents, or groups of agents, in their interactional processes, and they can have to do with intentions, emotions, actions, beliefs, social concepts; in other words, with anything that has to do with our stance towards the social other. Social cognition, therefore, is generally accepted to regard the capacity of28 cognitive agents to understand and successfully interact with one another. Just as for 29 cognition in general, also social cognition has not been unanimously defined, and it is no chance that different scholars preferentially focus on different aspects of this phenomenon. The traditional accounts that 4Eapproaches generally try to overcome are those that provide ‘individualist’ or ‘internalist’ explanations. Such perspectives base their explanations about 30 how we understand and interact with others on a separation between internal and individual processes and the body, also in its relations with the surrounding environment. Therefore, what lies outside the boundaries of the mind/brain has no direct involvement with cognitive processes. It is the individual’s cognitive mechanisms and mental states that shape individual and social thoughts and actions, and cognition is understood as the computational manipulation of abstract symbols. 31
With regards to socio-cognitive processes, the traditional consensus among philosophers of mind and cognitive scientists has been the idea that a theory of mind is necessary for understanding and interacting with the social other. Briefly, Theory of Mind (ToM) is the branch of cognitive science that investigates how we attribute mental states to others, in order to explain and predict their behaviour. Such an ability is referred to as 32
mindreading, on the basis of which our socio-cognitive capacities are explained by appealing to the way we grasp mental states and to how we interpret them and attribute them to others. 33 Therefore, the way we deal with others is primarily based on our understanding of them as bearers of mental states and propositional attitudes that we can grasp and use to predict their behaviour or the reasons for their actions. Within ToM, two of the main accounts developed are Theory Theory (TT) and Simulation Theory (ST), of which a very general outline is enough, for the present purpose. ST holds that the mindreading process takes place by predicting behaviour on the basis of our own mental states, on the basis of what mental state we believe we would have if behaving in a certain manner. Therefore, we simulate mental 34 states and attribute to the target individual the state we believe would lead to the observed 28 De Jaegher et al., 2010, p. 441.
29ibidem.
30 De Jaegher et al., 2010. 31 Shepherd, 2012, p. 507.
32 Marraffa, 2011, Internet Encyclopedia of Philosophy. 33 Shepherd, 2012, p. 508.
behaviour, using our own mind as a model. On the other hand, TT claims that we use folk psychological knowledge to infer what mental state the other probably is in from their manifested behaviour. Folk psychological theory, through which we derive our ability to treat others as bearers of mental states such as ourselves, for some is acquired through some sort of innate mechanism or module, for others is a result of enculturation. From these inferences, 35 we then predict subsequent behavioural patterns. 36
The problem that characterizes the debate against these individualistic explanations of social cognition, is the lack of attention they reserve to the role of the body and the environment in shaping the cognitive performances. In order to account for this gap, 4E cognition theories suggest explanations of social cognition that are more ‘interactive’ or 37 ‘collectivist’ . The embodied view on social cognition, for example, is fundamentally action38 and interaction based, according to which our ability to understand others and interact with them depends on non-representational and non-mentalistic processes. However, there is not one comprehensive account on embodied social cognition. One important perspective on these matters is one that focuses on embodied simulation and on the presence of mirror neurons in portions of our brain that underlie simulation processes.
Gallese, for example, believes that mindreading does not play a crucial role in social39 interactions, because in such situations we are able to understand and make sense of the behaviour of others in a direct and almost as a reflex, without a need for an internal interpretation. Gallese admits that surely we can explain behaviour of others by using mentalizing abilities, but often we do not need to, and therefore reflecting on propositional attitudes is not “all there is in social cognition”. In social interactions, then, what occurs 40 most of the times is the firing of mirror neurons, on the basis of which we operate an embodied simulation of the other’s actions. Mirror neurons are those that fire when an agent is executing a certain goal-directed action but also when she is only observing an action, or hearing sounds related to that action . The idea is that the functioning of mirror neurons 41 allows us to model our interactions with others in a direct way, and allows the observer to “know how it feels” to perform such an action. So, for example, while observing someone’s 42 hand gestures, congruent mirror neurons discharge in the observer’s brain, so that she is able to understand whether the gesture is aimed at picking up a cup or at grasping some food,
35 Marraffa, 2011.
36 It is important to keep in mind that various versions of ST and TT exist, but for the present purpose I believe
that this very general outline is enough.
37 De Jaegher et al., 2010. 38 Spaulding, 2017, p. 298. 39 Gallese, 2007, p. 659. 40Ibidem. 41 p. 660. 42 p. 661.
matching action observation and perception to action execution, for which our embeddedness in specific contexts also provides further information for understanding. Gallese suggests that mirror neurons and embodied simulation play a functional role in action recognition but also in ascribing intentions to the observed other, where intentions are understood as predictions of forthcoming new goals.43 What this means is that when observing an action, the subsequent and not yet observed action can be predicted due to the activation of “parietal mirror neurons”, that discharge when the observed actions are embedded in a specific action aimed, overall, at another goal. Therefore, actions with distal goals are recognized as chains of single motor actions, each dependent on the previous one, and this allows the agent to understand the overall intentions of the act on the basis of embodied simulation.
In my opinion, it is certainly fascinating that the understanding of certain actions and - as I will get back to later - certain linguistic expressions is based on a direct neural simulation of what we observe and perceive. However, I believe that this account still remains problematic and not sufficient to explain the performance of socio-cognitive processes. Indeed, this approach can still be interpreted as somehow individualistic, even though it certainly takes into account active influences of bodily activity and aspects of embeddedness on cognitive processes. First of all, Gallese himself claims that, regarding distal goal actions only “simple intentions” can be understood on the basis of embodied44 simulation, hence not accounting for more complex ones, that are mostly those which we deal with every day. Secondly, most of the focus still lies on the internal mechanisms of the individual body and brain. Surely, the presence of the social other is more relevant in this embodied cognition perspective than in cognitivist accounts of social cognition, as without perception of the other’s actions there would be no understanding. However, the social other seems to be only functional for the activation of internal and individual mechanisms, so the active role of social interaction is not considered fully, and the type of explanation provided seems to be provided still from an observational stance. The embodied cognition-based perspectives that I will deal with in the next Section have attempted to account in more depth for the role of intersubjectivity in perception and social cognition.
43 p. 662. 44 Ibidem.
1.2 Intersubjectivity, social interaction and “we-space”
Scholars such as Gallagher , Gallagher and Hutto or De Jaegher, Di Paolo and Gallagher 45 46 47 have suggested accounts of social cognition that are fundamentally based on the role of social interactions. The “embodied Interaction Theory”, for example, is also based on the idea -48 common to 4 E cognition approaches - that our ability to understand others and interact with them depends on non-representational and non-mentalistic processes. Already at the level of perception the cognitive processes involved are not identifiable solely with the processing of sensory information, but, rather, are tightly linked to the presence of others. 49 Before developing the ability to hold propositional attitudes or to ascrie them to others, cognitive agents must develop the embodied practices of embodied primary and secondary intersubjectivity, on the basis of which they learn to grasp intentions through the perception of bodily activities. 50
Based on the ability for embodied simulation, primary intersubjectivity is the capacity to interact with others at the level of the perception of others’ bodily movements and expressions as indicators of what they feel or think. The way we manifest primary intersubjectivity is through simulation of facial expressions, through the capacity of detecting eye movement, understanding emotions through actions, and so on. Therefore, the body of 51 the other opens up possibilities for expressive behaviour of the child, that in this way also differentiates between human beings and inanimate objects. Once having developed this52 form of intersubjectivity, the child can develop secondary intersubjectivity, characterized by the capacity of shared attention, on the basis of which “we start to learn about the world by seeing how others relate to objects in that world”. By observing how others relate to the 53 environment we direct our attention to some aspects of it rather than others; to some individuals rather than others; we interact in different ways with different people; we learn how to communicate with others about objects in the environment. Since actions involve the 54 social other, they acquire meaningfulness, and such meaning and value is derived precisely from actions of others. 55
45 Gallagher, 2008.
46 Gallagher & Hutto, 2008. 47 De Jaegher et al., 2010.
48 Gallagher, 2008; Gallagher & Hutto, 2008. 49 Gallagher, 2008.
50 Gallagher, 2008; Gallagher & Hutto, 2008. 51 Gallagher & Hutto, 2008, pp. 3-4.
52 p. 4.
53 Gallagher, 2008, p. 171. 54 Ibidem.
The embodied interaction theory provides a more interesting perspective on the role of the other in shaping our attitudes, and overcomes the individualist and observational explanations of social cognition mentioned above in a more convincing manner. However, when it comes to understanding the intentions underlying the actions of others on a more complex level - beyond, that is, embodied simulation - embodied interaction theory resorts to the narrative practice hypothesis, from which new problematic aspects arise. According to56 the narrative practice hypothesis, reasons for actions are best understood if seen as part of a storyline that makes such reasons explicit. Therefore, a folk psychological understanding of57 reasons and sense-making of actions is achieved through the intersubjective practice of engaging in story-telling, an interactive process that takes place between the child and the caretaker. Gallagher and Hutto refer to “folk psychological narratives” as ones “about58 agents who act for reasons” and that constitute “objects of joint attention”. These narratives 59 provide understanding of social norms, so that folk psychological expectations are formed with regards to the actions of others.
This definition of ‘folk psychological narratives’, however, is not particularly clear in distinguishing these from other types of narratives that do not provide understanding of social norms and of intentions for action. Indeed, it could be argued that most narratives depict agents that act for reasons and constitute objects of shared attention, in which case almost any sort of narrative could be relevant for this social and educational purpose. Also, folk psychological narratives could be different in different cultures or communities, hence expectations and understanding of intentions may be also dependent on the cultural and social embedding of the agents involved in the interaction, and this is not accounted for in Gallagher and Hutto’s approach. Finally, the narrative practice hypothesis is meant to suggest an alternative to the traditional perspectives that believe mindreading to be necessary for social cognition. But, it seems, also the attribution of reasons for action to others is a form of60 attribution of mental states, and it is not clear how the authors differentiate this process from the attribution of propositional attitudes through mindreading.
I believe that the enactive perspective on social interaction can better account for the active role of intersubjectivity in shaping social cognition and in overcoming traditional individualistic accounts. The important aspect that the enactive cognition approach - in the version developed by Thompson and Di Paolo - add to embodied account of (social)61 cognition is the concept of autonomy, that is considered to apply to individual cognitive
56 Gallagher & Hutto, 2008. 57 p. 9.
58 ibidem. 59 pp. 10-11. 60 p. 13.
systems as well as to social interactive processes. These scholars focus on the idea that cognition is constitutively dependent on the body, intended as an autonomous system. First 62 of all, the concept of autonomy is tightly linked to the notion of individuation, that characterizes the body as making itself distinct from the surrounding environment, and as maintaining such a distinction. The process of the body’s self-individuation is explained by its autonomy, and these notions together allow for the distinction between what is a cognitive process or system, and what is not. 63
Taking it one notion at the time, an autonomous system is intended as “a network of co-dependent, precarious processes able to sustain itself and define an identity as self-determined system”. An autonomous system, in other words, is one constituted by a64 series of processes, each one of which enables and is enabled by another process within the system: a system of this sort is defined as operationally closed. Because of the fact that an 65 autonomous system is constituted by a network of enabling relations, it is also understood as precarious. This means that if any of the enabling conditions that belong to an operationally closed network is removed, then the process will likely stop, as it does not exist independently and it sustains itself mainly due to its own constitutive processes. Moreover, 66 an autonomous system is characterized by the capacity of adaptivity and sense-making, through which it establishes and selects the interactions with the environment that will facilitate and maintain its autonomy; interactions that, in this way, acquire meaning and a normative status - being the best conditions for maintaining its equilibrium. Following from 67 all of this, a cognitive system is not intended as one that represents internally states of affairs, but, rather, is one that behaves in a way as to maintain its existence and flourishing. In the words of Thompson and Di Paolo, “basic cognition is more a matter of adaptive self-regulation in precarious conditions than abstract problem solving”. 68
When autonomous systems interact with each other and maintain their own autonomy in the process we have social interaction. In social interactions, the actions and intentions of69 the participating agents does not fully determine its outcome because the interaction is “taking on a life of its own”, and the outcome will depend also on the dynamics that the 70 interaction itself takes on. Social interaction, therefore, is defined as 71
62 p. 68. 63 p. 69.
64 De Jaegher et al., 2010, p. 441.
65 Thompson & Di Paolo, 2014, pp. 70-71. 66 p. 72.
67 p. 73. 68 Ibidem.
69 De Jaegher et al., 2010, p. 441. 70 p. 442.
“a co-regulated coupling between at least two autonomous agents, where: (i) the co-regulation and the coupling mutually affect each other, constituting an autonomous self-sustaining organization in the domain of relational dynamics and (ii) the autonomy of the agents involved is not destroyed (although its scope can be augmented or reduced)”.72
In the case of using a tool, of observing a social scene, of being in the sole physical presence of another, social interaction does not occur. It is important to underline that the authors do not exclude that situations of this sort are, in a sense, social and may influence our self and socially directed attitudes. However, these do not count as social interactions, because there is no mutuality involved and the agents are not co-regulating. Therefore, social cognition is73 not only restricted to understanding others, as it extends to understanding together with others through acting and interacting with each other. It is not denied that social cognition has, in74 fact, some non-interactional aspects, as mentioned above. The interest on social interactions so intended is, for the present purpose, that these enable and constitute wholly new autonomous systems that are sustained solely by the activity of the agents involved. In these cases, sense making - as described above - becomes participatory, as the relations with each other and with the environment depends on their mutual activity and on patterns of bodily coordination, resulting in a system that would not have emerged on its own.
This conception of social interaction is compatible with Krueger’s perspective on 75 social cognition, that brings together embodied, enactive and extended cognition views. In his perspective, the first mode of agents’ social interaction is grounded in embodiment, and it is identified as the face-to-face interaction between agents that become available to one another in a shared contexts of interaction. When this physical co-presence of the other becomes 76 socially relevant, then the shared context of interaction can be defined as ‘ we-space’. This is77 a space that has a role in driving socio-cognitive processes, as it is a social space in which interaction is co-coordinated. This means that ‘we-space’ does not stem from the action of a single agent toward the other, but it is co-regulated by the “ongoing engagement of social agents”. 78
It is the notion of engagement that, I believe, marks the important difference between socially relevant and non-relevant presence of the other. As mentioned above, physical co-presence can be intended, on some level, as social, but it does not involve interaction. In 79
72 De Jaegher et al., 2010, pp. 442-443. 73 p. 443. 74 p. 442. 75 Krueger, 2011. 76 p. 643. 77 p. 644. 78 Ibidem. 79 De Jaegher et al. 2010, pp. 442-443.
‘we-space’, gestures and expressions are fundamental in the creation of shared meaning and emotional attunement. It is gestures - or expressive behaviour - that are relevant in social cognition, because they are what drives interpersonal understanding, contrarily to non-expressive actions - such as instrumental, locomotive or reflexive actions. Therefore, 80 mere bodily co-presence is to be distinguished from actual interpersonal engagement. 81 Instead, co-regulation takes place when the behaviour of agents toward each other actually affect each other’s own behaviours. Therefore, this kind of interaction results in the emerging of information that was not available prior to their engagement.
So intended, ‘we-space’ is understood as a cognitive niche, and the structure of this 82 particular cognitive niche is altered by agents due to the interactive nature of their social engagements.83 Precisely for its interactive nature and its emerging solely through engagement between cognitive agents that act together, I believe that ‘we-space’ can be intended as a socio-cognitive niche, of which interactions are the enabling and constitutive feature.
Krueger focuses on the importance of gestures as scaffolds and motivators for social interaction in the sense that they have a public and embodied nature and they are externalized practices, and affect the agent that performs them, and she to whom they are directed to. In 84 other words, in his account of social cognition, Krueger wants to draw attention to the importance of engaged interaction and the emergence of coordinated behaviour, therefore presenting an account of socio-cognitive practices that is, at one time, embodied, enactive and extended. At a socio-cognitive level, coordination allows agents to share intentions through movements, gestures, verbal language in such a way to influence each other’s behavioural responses within a context of shared interaction - ‘we-space’. Through this kind of interaction and coordination, the subjects involved actively create meaning, that evolves, aligns and changes on the basis of further interactions. In we-space, therefore, agents engage in the “co-creation of shared meanings”. The 85 socio-cognitive niche, therefore, becomes the space in which the possibility of language use and understanding is co-created and emerges solely because of the participatory engagement that the agents are involved in.
The fundamental contribution of the extended mind theory insights is that it challenges the idea of the internal nature of mental states and processes for a more externalized account. Moreover, it is relevant because it explains how the body and the environment are tightly coupled, and this coupling is exploited and manipulated in the
80 Krueger, 2011, p. 644. 81 p. 645. 82 Ibidem. 83 p. 646. 84Ibidem. 85 p. 649.
understanding of others as externalized and extended resources that are not confined within the boundaries of the mind. This is why Clark defines the cognitive agent as “plastic, environmentally exploitative and ecologically efficient” , and why Krueger intends social86 cognition as a form of extended cognition, and social interaction is intended as a form of extended social cognition . 87
1.3 Language
One of the most distinctive aspects of human cognition and of the domain of human interactions is the socio-cognitive process of language. In my opinion, the account of language provided by the embodied simulation and mirror neurons theory presented, for example, by Gallese88 does not provide a satisfactory account of this process. In fact, according to this view, the observation of speech-related, therefore communicative, mouth actions leads to the activation of the motor system involved with the execution of such actions.89 Therefore, during language processing, the motor system is activated, hence understanding the meaning of sentences does not rely primarily on symbolic mental representation but on embodiment. In action-related sentences, in fact, the neural structures 90 in the premotor cortex that preside over the action execution play a role in understanding the semantic content of sentences that describe such actions - such as sentences about actions performed with hands, mouth and feet. Yet, the question that remains is what happens when91 processing abstract semantic content - or, simply, non-action related content - that cannot be explained through embodied simulation theory.
Clark and Chalmers and Clark focus, instead, on language in its materiality and in92 93 its role in scaffolding cognitive processes in a way that is able to transform the mind. This 94 socio-cognitive process, it is claimed, is cognition-extending because it allows for types of computational processes impossible without it, it allows to discover and fixate abstract patterns of reality, and it allows agents to reflect on our thoughts and on our capacity of actively guiding and shaping our thoughts.95 Linguistic symbols that are charged with
86 Clark, 2008, p. 44. 87 Krueger, 2011, p. 643. 88 Gallese, 2007. 89 p. 660. 90 p. 664. 91 Ibidem.
92 Clark & Chalmers, 1998. 93 Clark, 2006; 2008. 94 Clark, 2008, p. 44. 95 Ibidem.
meaning are, thus, externalized and materialized, and they become available intersubjectively.
“Language, thus construed, is not a mirror of our inner states but a complement to them. It serves as a tool whose role is to extend cognition in ways that on-board devices cannot. Indeed, it may be that the intellectual explosion in recent evolutionary time is due as much to this linguistically-enabled extension of cognition as to any independent development in our inner cognitive resources”. 96
The extended mind approach, therefore, investigates the nature of language as a socio-cultural cognitive practice and as a cognitive super-niche and as a socially and 97 culturally endowed practice that aims at enhancing our basic biological capacities for purposes that aid and enable higher levels of cognition.98 As a cognitive super-niche, language enables us to objectify our own thoughts and engage with them, hence stabilizing and fixating certain modes of thought that are intrinsically fluid.
More than the fact that the environment we are situated in is stable, I believe that through interactions we create new spaces - or, in some sense - environments that are maintained stable by the constant interaction and by the network of enabling relations that constitute it. In the words of Thompson and Di Paolo, “a precarious, operationally closed 99 system is literally self-enabling, and thus it sustains itself in time partly due to the activity of its own constitutive processes”, so it requires constant relations with the outside world. In my opinion this definition can be applied to communicative and linguistic interactions, and the
socio-cognitive niche can be intended as an operationally closed network constituted by relations between enabling conditions. These are culturally and socially embedded conditions and embodied conditions that, through materiality and verbalizing extend beyond the boundaries of the mind. The enactive cognition perspectives outlined in the previous section, together with Kruger’s account of we-space and Clark’s view on language, I believe allow to consider language in a more comprehensive manner, as a form of social interaction that constitutes an autonomous system that emerges in the socio-cognitive niche generated through human engaged interactions.
96 Clark & Chalmers, 1998, p. 6. 97 Clark, 2006.
98 Menary et al., 2003.
Chapter 2
Theories of metaphor
In this chapter I will present and analyse the core concepts underlying the phenomenon of metaphor according to some of the influential and contemporary theories. Firstly, I will present the fundamental ideas underlying Conceptual Metaphor Theory (CMT) as presented by Lakoff and Johnson,100 and provide a critical assessment of such concepts. I will then move on to perspectives that aim at enriching CMT by analysing the phenomenon of metaphor within accounts of social cognition and embodied cognition theories. The accounts presented in this Chapter will be analysed and evaluated in their aim to overcome certain problematic aspects of Lakoff and Johnson’s theory of metaphor in cognition. Finally, I will focus on some theoretical issues revolving around the notion of novel metaphors and discuss important insights on the subject matter developed by interactionist approaches to metaphor.
2.1 Conceptual Metaphor Theory
The phenomenon of metaphor has been extensively explored and analysed for centuries. Aristotle, for example, offers one of the most ancient accounts of metaphor as an instance of novel poetic language, in which names or terms are used in a way different from their ordinary sense.101 In the classical perspectives, following from Aristotle and ancient rhetoricians, metaphor was regarded mainly as a rhetorical and poetic tool that allows for a deviation from literal sense. Hence, a word that ordinarily is used to describe something, is used to describe something else. 102 However, in the second half of the last century, much attention has been drawn to this subject matter, and research and debate about it has been characterizing cognitive philosophy, cognitive psychology, cognitive linguistics, argumentation theory, discourse analysis, philosophy of language and mind. Today, many seem to agree that metaphor must no longer be considered as a purely linguistic tool with rhetorical or illustrative purposes. As Lakoff puts it, “metaphor is not just a matter of language, but of thought and reason. Language is secondary”. 103
Lakoff and Johnson have been among the first cognitive linguists to dedicate an extensive analysis of metaphor, no longer understood as simple linguistic tool, but, rather, as a fundamental part of the human conceptual system, considered as being fundamentally
100 Lakoff & Johnson, 1980a; 1980b; 1980c; Lakoff, 1993; Johnson, 2010. 101 Lakoff, 1993, p. 1.
102 Johnson, 2010, p. 401. 103 Lakoff, 1993, p. 7.
metaphorical in character. 104 According to these scholars, since concepts govern both our thought and the way we ordinarily function in reality, our conceptual system defines how we perceive, how we act and how we relate to the social other. The very same conceptual system is also at the basis of language and communication, hence linguistic evidence is seen as a tool that uncovers its fundamental aspect, namely its metaphorical nature.105 Therefore, if the human conceptual system is intrinsically metaphorical, the authors suggest that the way we experience reality, we act within reality and we think about reality “is very much a matter of metaphor”.106 Although understanding and talking about concepts are based on the same underlying system, CMT highlights the importance of not confusing metaphorical
expressions with metaphorical mappings, that constitute metaphors themselves. In fact, the former constitute the names of the metaphorical mappings, they have propositional form and exist in a variety of different linguistic expressions that instantiate the metaphor. 107The latter, instead, do not have propositional form, rather, they are sets of correspondences between (elements of) different domains.108 In the words of George Lakoff, with regards to the conceptual metaphor “LOVE IS A JOURNEY”,
“The metaphor involves understanding of one domain of experience, love, in terms of a very different domain of experience, journeys. More technically, the metaphor can be understood as a mapping (in the mathematical sense) from a source domain (in this case, journeys) to a target domain (in this case, love). The mapping is tightly structured. There are ontological correspondences, according to which entities in the domain of love (e.g. the lovers, their common goals, their difficulties, the love relationship, etc.) correspond systematically to entities in the domain of a journey (the travelers, the vehicle, destinations, etc.)”.109
The metaphorical mapping, therefore, provides the understanding of one domain of experience - a target domain - in terms of another domain of experience - a source domain. As Indurkhya puts it, “the object of description is the target, and the object that is being used to unconventionally describe the target is the source”. 110
Linguistic expressions, however, can still be important evidence to expose the metaphorical structuring of our thought and behaviour. Following the authors, we can consider as an example, the “ARGUMENT IS WAR” metaphor, that is instantiated in everyday language in a variety of different propositional forms, such as “I won the 104 Lakoff & Johnson, 1980a, p. 195.
105Lakoff & Johnson, 1908b, p. 454. 106 Ibidem.
107 Lakoff & Johnson, 1980a, p. 195; Lakoff, 1993, p. 6. 108 ibidem.
109 Lakoff, 1993, p. 5. 110 Indurkhya, 1992, p. 246.
argument”, “sheattackedmy argument”, “his argument was based on a good strategy” and so on. Because of the fact that arguments are metaphorically conceptualised - and understood - in terms of war, the way we will experience argumentative behaviour and the way we will act in an argument will be guided by such a mapping. 111 Therefore, the ontological mappings across conceptual domains also allow for epistemic correspondences between elements of the domain involved, so that knowledge about war will systematically be mapped onto knowledge about arguments.112 The fact that also the linguistic expressions used in talking about arguments are metaphorical in character, is a mere consequence of the metaphorical structure of the concept of argument in terms of war.
According to Lakoff and Johnson, 113 the human conceptual system contains both non-metaphorical and metaphorical concepts. The former are those that “emerge directly from our experience and are structured in their own terms”, such as spatial orientations, ontological concepts that arise from physical experience - entity, container, person -, or structured experiences and activities - eating, moving, transferring; the latter, instead, are those that are “understood and structured not merely in their own terms but rather in terms of other concepts. This involves conceptualizing one kind of object or experience in terms of a
different kind of object or experience”. There are roughly three types of metaphorical concepts that account for what is involved in metaphorical structuring. Firstly, we have orientational metaphors, which structure concepts by orientating them relatively to their non-metaphorical orientation, as in the “GOOD IS UP” metaphor. Secondly, we find ontological metaphors, namely those that map the ontological status of entity, or container, and so on, onto something that does not inherently have such a status, as in the “IDEAS ARE ENTITIES” metaphor. Finally, we hold structural metaphors, which structure one type of experience or activity in terms of another type of experience of activity, as in the “UNDERSTANDING IS SEEING” metaphor.114 Thus, what occurs in metaphorical mappings is that entailments which characterize non-metaphorical concepts are transferred onto the corresponding metaphorical ones. So, what the authors suggest, is that abstract concepts are generally defined and understood by cognizers in terms of more concrete ones that are non-metaphorical. However, it is sometimes confusing to identify on what basis Lakoff and Johnson categorize certain concepts as abstract or as more concrete. In fact, concepts like space and substance are understood as being concrete and non-metaphorical, whereas time, life and mind are understood as abstract. It can be argued that, for example, space is a concept very often understood, perceived and talked about not in its own terms, but rather in terms of a container or an object, as something we occupy and we possess (“ stay in
111 Lakoff & Johnson, 1980b, pp. 454-456. 112 Lakoff, 1993, pp. 6-7.
113 Lakoff & Johnson, 1980a, p. 195. 114 pp. 195-196.
your space”, “I’m taking some space”, “she invaded my space”). A similar objection can be made with regards to the concept of substance. The concept of time, according to the definitions provided by the authors, belongs to the domain of abstract concepts and can be understood in terms of, for example, a valuable commodity. But the concepts of ‘valuable’ and of ‘commodity’ also seem quite abstract, if we stick to the idea that non-metaphorical is what is defined in its own terms. 115The same can be said of many other concepts suggested by the authors as source domains.
Another crucial aspect characterizing the phenomenon of conceptual metaphor is that each metaphorical expression provides only a partial definition and understanding of the concept under discussion. The reason for this is that the metaphorical mapping underlying the linguistic expression tends to highlight certain correspondences between the domains involved, and hide others.116 Therefore, different linguistic expressions that instantiate the same generalized, conceptual metaphor, will focus on certain entailments and not others. For example, the “IDEAS ARE PEOPLE” metaphor highlights correspondences such as those of coming into existence, dying, developing, and so on, whereas the “IDEAS ARE COMMODITIES” metaphor stresses aspects such as those of being valuable, that can be bought and sold. 117
Lakoff and Johnson believe that metaphors have a strong experiential grounding, and are based on what they define as “complex experiential gestalts”. 118An experiential gestalt is understood as an experience that, when considered as a whole, has qualities that go beyond the simple total of its parts. It constitutes a multidimensional organisational structure of types of experiences, “(a) a structure within a person’s experience that identifies that experience as being of a certain kind; or (b) a structure in terms of which a person understands some external occurrence and that identifies that occurrence as being of a certain type”. 119 The dimensions that form the structure of a certain experience can be superimposed onto the dimensions of structure of a different experience, hence resulting in the complex experiential gestalt of metaphor. For example, the experiential gestalt for ‘war’ and ‘conversation’ both involve participants, stages, a linear sequence, causation relations and purpose. If this experiential basis is not understood or picked up on, the correspondences between the domains will not be understood either, and the metaphor itself will not be grasped. 120 This risk is even more present when the same source domain is used in different mappings; the
115 Lakoff & Johnson, 1980a, p. 195. 116p. 200.
117 p. 201.
118 Lakoff & Johnson, 1980a, pp. 201-206; 1980b, pp. 478-480. 119 Lakoff & Johnson, 1980a, p. 205.
fact that it relates to different domains of the structure of a certain experience is what allows for the metaphor to be correctly understood.
The reason why I find this aspect of CMT problematic is its tight link to the role that image schemas are believed to play in structuring experience in cognition. In fact, experiential gestalts are meant as a generalization of the concept of schema 121 - that I find is already enough of a generalization. What CMT suggests is that inferences regarding elements or aspects of our experience follow from the image schemas that we have formed regarding such experiences. Schemas are understood as cognitive building blocks for thought and action, based on the classification of stimuli into categories. Since it is not possible for humans to process every aspect of the environment at a given time, the schema view suggest that we access the knowledge we have classified to interpret a given situation. 122 The underlying mechanism of conceptual metaphor, therefore, is that the image schema - or experiential gestalt - of a source domain is mapped onto a target domain, and this occurs in line with the Invariance Principle: “metaphorical mappingspreserve the cognitive topology (that is, the image-schema structure) of the source domain in a way consistent with the inherent structure of the target domain”. 123 Therefore, each domain has its own inherent and fixed structure that must be preserved by the metaphorical mapping in order for it to be accepted and understood. Consequently, certain mappings will not be possible to instantiate if they violate the inherent structure of one of the domains, a structure that automatically limits and constraints metaphor possibilities. 124
This idea seems to be defending an “internalist” account of cognition, in a way that, perhaps, slightly differs from that criticized by De Jaegher et al .125 What I mean here by ‘internalist’ is that the process of abstracting features of experience and categorizing them in order to then access them in specific situations seems to suggest a conception of the mind and cognition as simply extrapolating information from the environment to then process it and use it on the basis of its own mechanisms. The separation between environment and mind/brain is slightly less extreme that the one attacked by De Jaegher et al., but it still remains. In a 4 E cognition perspective and, as will be dealt with further, in the Interaction 126 Theory of metaphor and in cognitive niche construction theories, image schemas - if this term could even be acceptable - should be intended as being constantly re-defined by the cognizer on the basis of interactions with the environment and with others, and on the basis of metaphorical conceptualization itself. CMT’s claim that metaphors must - and do - preserve
121 p. 202.
122 Landau et al., 2010, p. 1047. 123 Lakoff, 1993, p. 13; emphasis added. 124 p. 13.
125De Jaegher et al., 2010. See Chapter 2, p.3 of the present study. 126 See Chapter 3 of the present study.
the image-schema structure of the domains in the mapping is based on the idea that such structures are inherent to the domains of experience and the correspondences between them are fixedand pre-existing. This perspective also operates a clear-cut separation between the127 individual cognizer, her cognitive processes and her surrounding environment. The idea that certain mappings are not possible due to the constraints operated by the image-schema structure of the domains involved is to consider concepts, language and meaning as ultimately fixed and non-modifiable. I believe that there are two ways of interpreting this claim suggested by CMT, and both of them are problematic. Firstly, it can seem as if the ontological correspondences between (elements of) certain domains pre-exist the agency of the cognizer on the environment, and this is what allows for metaphorical cognition. This externalist point of view has been dealt with extensively in Chapter 1 of the present study and will be dealt with again in Chapter 3. Secondly, the correspondances can be intended as fixed within the image-schematic structure of concepts that the agent abstracts from her experience. In this case, the problem would be to explain how it could be possible that different agents happen to abstract the same common properties of experiences and categorize them in the same way, so to understand and accept certain metaphorical mappings rather than others. An important question that follows from this problem revolves around the possibility for novel metaphors. As a matter of fact, if metaphorical cognition is based on a fixed set of correspondences, the possibility for novel metaphorical ascriptions of meaning is extremely limited. The relation between conventional and novel metaphors will be dealt within Section 2.3 of the present Chapter.
2.2 Embodied and socially embedded metaphors
As mentioned in the previous section, Conceptual Metaphor Theory (CMT) as developed by Lakoff and Johnson considers the human conceptual system to be intrinsically metaphorical. According to this perspective, people ordinarily use metaphors to understand and reason about abstract and complex concepts in terms of more concrete and simpler ones. Metaphors operate as conceptual mappings between source concepts - represented by familiar, concrete notions, often derived from the physical domain - and target concepts - represented by more abstract notions that are difficult to grasp and conceptualize. However, despite its crucial importance and influence in metaphor theory from a cognitive perspective, CMT has also been criticized and rethought, especially because it has been said to give too little attention to the social aspects of cognitive agents’ lives. Within their approaches to aspects of social and
embodied cognition, some scholars128have advance claims regarding the direct influence that metaphor bares on socio-cognitive processes and, viceversa, how social cognition can influence metaphor use and understanding.
As mentioned in the previous Chapter, 4E cognition theories all share a critical stance towards classical cognitivism, believed to focus too much on individual cognitive processes and on how they unravel internally to the agents' mind. Instead, individuals are believed to have a more participatory attitude, and social aspects of intersubjectivity and social embeddedness are believed to play a fundamental and active role in shaping (socio) cognitive workings. Similar claims can be identified in embodied and embedded cognition theories with regards to metaphorical cognition, advanced in order to argue that metaphor use in communication and in cognition are deeply and fundamentally linked to human social activities. 129
These perspectives embrace CMT’s core claim that metaphors represent a fundamental aspect of human cognition, however criticizing a highly individualistic attitude in the explanation of how conceptual metaphors operate within humans’ cognitive systems. Gibbs et al., 130 focus on the role of embodiment in grounding metaphorical cognition and allowing metaphorical understanding. Other perspectives131 highlight how metaphorical mappings have the power to actively influence the way we process socially relevant information by shaping our socially-directed thoughts and attitudes,132 or by influencing the way we deal with social issues, problem solving and decision making strategies.133 These approaches seem to agree with each other, and with CMT, on the fact that the use of conceptual metaphors enable cognitive agents to conceptualize, understand and communicate complex or abstract concepts in a more concrete manner. Conceptual metaphors achieve this task at various levels of cognitive activity, such as in communication, in reasoning, in individual cognition or in socially directed activity.
Landau, Keefer and Meier,134 for example, suggest a metaphor-enriched social cognition theory, in order to show that, often, the cognitive events that underlie social thoughts and attitudes are forms of metaphorical cognition. In fact, they criticize the two viewpoints on social cognition that they consider to be dominating: the schema view and the embodied cognition view. Landau et al. believe that according to the former, since people lack the capacity to process all aspects of the environment at a precise moment, they tend to 128 The authors that will be reviewed in this Chapter include Gibbs et al. (2004; 2010;), Landau et al. (2010),
Thibodeau & Boroditsky (2011).
129 Gibbs, 2013, p. 55. 130 Gibbs et al., 2004; 2010.
131 Such as Landau et al., 2010, and Thibodeau & Boroditsky, 2011. 132 Landau et al., 2010.
133 Thibodeau & Boroditsky, 2011. 134 Landau et al., 2010.
operate a classification of stimuli into categories of knowledge that are accessed when relevant to the situation at hand. Therefore, schemas are intended as “mental structures that contain abstract representations of accumulated knowledge about categories of similar stimuli”135 that are believed to underlie and guide social information processing and socially relevant attitudes and beliefs. With regards to embodied cognition theories on metaphor, instead, Landau et al. believe that what is claimed is that knowledge is not processed in this abstract manner, as it rests upon people’s representations of their bodily states . 136 Nevertheless, Landau et al. believe that these two perspectives share the idea that people process social stimuli on the basis of their knowledge of similar stimuli.
The metaphor-enriched social cognition approach is grounded within the theoretical framework of CMT, as it aims at highlighting how people tend to establish relations between social and abstract concepts “in terms of superficially dissimilar concepts that are relatively easier to comprehend”. 137 However, according to these scholars, CMT does not take into consideration the whole range of social influences that actively play a role in, and are actively influenced by, metaphorical cognition. Landau et al. do not want to deny that cognitive agents rely on schemas in order to force a simple structure upon their highly complex social environment and to conceive of abstract concepts in isolation. Rather, through a revision of CMT in a social key, they want to add that many social concepts that are intrinsically abstract and naturally difficult to understand are mapped onto (apparently) dissimilar concepts to facilitate interpretation and evaluation. 138
Therefore, the main argument against the embodied approach to metaphor put forward by the aforementioned scholars is that with regards to metaphor theory, embodied cognition perspectives brake down solely to embodied simulation theory, and that they characterize metaphorical mappings as based on associations of similar bodily states. 139 Instead, Landau et. al. - in line with Thibodeau and Boroditsky - suggest a distinction between embodied cognition and conceptual metaphors “to highlight how conceptual metaphor can use representations of bodily states that is qualitatively distinct from the way embodied simulations are currently characterized to use such representations”. 140Conceptual metaphors, therefore, shape social information processing in a way that is distinct from the way embodied simulation works.
135 p. 1045.
136 Whether this is a correct interpretation of embodied metaphor theory will be dealt with below. For now, it is
important to understand how the social environment is believed to influence - and be influenced by - metaphorical cognition.
137 p. 1046. 138 p. 1047.
139 Landau et al., 2010, p. 1046. 140 p. 1054.
It is important to point out that the embodied metaphor theories advanced, among others, by Gibbs and Macedo141 and by Gibbs, Lima and Francozo142 do not constrain metaphor use and understanding solely to embodied simulation. As mentioned in the previous Chapter, it is true that embodied cognition theorists argue for the role of mirror neurons in language processing related to action, 143 and these ideas are compatible with findings on metaphor processing that have recorded reaction time of participants in understanding metaphors associated to kinesthetic experiences: expressions such as “to grasp a concept” were understood more rapidly if preceded by a (imagined) gesture of grasping by the subject. Nevertheless, the authors underline that “embodied simulations playsomerole in people’s 144
immediate processing of verbal metaphors, and language more generally”,145 therefore rejecting that embodiment - and much less embodied simulation - is the exclusive base for metaphorical cognition. The possibility that is being explored is that cognitive agents understand at least some abstract concepts in terms of embodied metaphors, meaning “in terms of metaphorical mappings where the source domains are deeply rooted in recurring aspects of embodied experiences”.146 However, Gibbs et al. do not seems to offer an explanation regarding the cases in which such embodied grounding is absent, or what are the other phenomena at the basis of metaphor besides embodiment. They allude to their existence, but seem to go no further.
As for the metaphor enriched social cognition theory, the emphasis that Landau et al. stress on the idea that metaphorical mappings often occur between dissimilar concepts and not between similar stimuli - as the embodied cognition approach is believed to hold - 147does not seem to actually deal with the influence of social embeddedness in cognition, and does not seem to overcome the internalist standpoint of CMT pointed out in the previous Section. Landau et al. seem to defend CMT’s claim without taking into account that conceptual metaphors are considered as being grounded on experiential gestalts and on the image-schematic structure of concepts. The very possibility of mapping dissimilar concepts onto one another, through metaphorical mappings, is based on fixed ontological correspondences between the domains of two (or more) experiential gestalts of the target and source concepts. As seen in the previous Section, in understanding conceptual metaphors the agent picks up on the similarities existing between the domains involved in the mapping, therefore the dissimilarity between the target and source domain is only apparent. Surely, the stimuli or the image-schematic knowledge of the stimuli are different, but at the basis of the 141 Gibbs & Macedo 2010.
142 Gibbs et al., 2004.
143 Goldman & de Vignemont, 2009, p. 156; Gibbs & Macedo, 2010, p. 690. 144 Gibbs & Macedo, 2010, p. 691.
145 p. 691, emphasis added. 146 p. 690.
