Theories developed in the so-called “second-generation” cognitive sciences have permitted significant advances in our understanding of how human beings find linguistic and other forms of symbolic representation to be meaningful.3 In particular, since about 1980, research coming from the “embodiment paradigm” in cognitive psychology and cognitive linguistics has demonstrated just how much people’s ability to make sense of their experience is underwritten by conceptual structures and cognitive processes that emerge from interactions among brain, body, and world. Rejecting any view of cognition as abstract symbol manipulation, embodiment theorists claim that thought – and hence the structure and use of language – is in fact directly grounded in the human body’s sensory and motor capacities. To the extent that Classics considers itself a broadly hermeneutic discipline that aims to shed light on the meanings elaborated by members of Greek and Roman society, it therefore seems crucial for classical scholars not only to have an awareness of the findings of this “embodied” cognitive science, but also to incorporate its insights into their interpretive strategies. For this reason, in this paper I introduce certain theoretical constructs from the cognitive interdiscipline – specifically, image schemas and conceptual metaphor – that I consider key to any psychologically realistic, humanly plausible account of meaning in ancient language and literature (and indeed in ancient culture more generally) and then go on to illustrate their analytic potential through a study of Latin’s metaphorical expressions of courage and cowardice.4
An “Experientalist” Account of Meaning
What, more precisely, does a psychologically realistic, humanly plausible account of meaning look like? In my view, it is one that adopts an explicitly “experientialist” theory of cognition, committed to the idea that, for human beings, our thinking depends fundamentally on the kind of brain we possess functioning in the kind of body we have in the kinds of physical, social, and cultural environments we typically inhabit (or have historically inhabited).5 In other words, it is one that views concepts as embodied mental representations deriving their meaning not through their correspondence to objects in external reality, but through their link to human conceptualizing capacities and psychological functions, which are grounded in and deeply constrained by our bodily nature, as well as by the local and global socio-cultural context.6 It is therefore one that takes a middle ground between the representationalism and functionalism of the “classical” computational theory of mind in philosophy and “good old-fashioned artificial intelligence” (Haugeland 1985, 112) – that is, any notion that human thought consists in the syntactical manipulation of implementation-independent abstract symbol systems that mentally “re-present” the structures of an objectively existing physical world to the mind – and the “radical embodiment” of enaction theorists like Humberto Maturana and Francesco Varela, which proposes a view of cognition as the effect of flat brain-body-action-world systems and which in its strongest forms sees no need for mental representations whatsoever for implementing intelligent behavior.7
Such a moderately embodied theory of cognition (cf. Prinz 2008) implies a very different account of meaning than that found in traditional (formal) philosophical and linguistic semantics. Generally speaking, experientialist approaches reject “truth-conditional” theories of linguistic meaning, which posit that an utterance’s meaning corresponds to the set of conditions in the world (or in any possible world) for which the utterance can said to be true. Moving beyond the view of defining categories by lists of “necessary and sufficient” features, they adopt a theory of categorization that recognizes classes characterized by nonobjective human perceptual, interactional, or purposive properties.8 In many cases, cognitive linguists claim that word meaning may not be reducible at all to symbols expressed in amodal, propositional format and arbitrarily linked to their referents. Rather than being represented in the mind as language-like symbols, the meanings of words very often are said to actually correspond (directly or indirectly through figurative interpretation) to gestalt structures of experience or “image schemas”. In cognitive psychology, an image schema is a highly abstract pre-conceptual structure of cognition that emerges through human perceptual and sensorimotor interaction with the world – as Mark Johnson (1987, xiv) writes, “a recurring dynamic pattern of our perceptual interactions and motor programs that gives coherence and structure to our experience”. Image schemas may therefore be visual in nature, e.g., “long, thin shapes, or containers” (Lakoff 1987, 113‒14), or more abstract representations deriving from the character of human spatial experience, such as UP/DOWN schemas, CENTER/PERIPHERY schemas and MOVEMENT schemas. As cognitive structures that are analogues of (because dependent on the same neural architecture as) sensorimotor experience and thus open to visual and kinesthetic “transformations” in mental space, image schemas provide the inferential patterns that motivate the range of senses typically characterizing the meanings of words.9
Image schemas may also be metaphorically interpreted, as a way of extending meaning further into abstract domains. According to the theory of conceptual metaphor, it is through the regular metaphorical mapping of bodily-based image schematic structure onto concepts not directly grounded in experience that human abstract thought is in fact possible. Recognizing the all-pervasive character of certain metaphorical patterns in language, George Lakoff and Mark Johnson argued (1980, 1999) that the clustering of metaphorical linguistic expressions around many (mostly abstract) concepts in fact reflects inherently metaphorical understandings that speakers of a language possess of those concepts.10 Speakers of a language talk about abstract domains of experience metaphorically, that is, because they actually conceive of them metaphorically in terms of other (mostly concrete) experiences. In this embodied view of cognition, literal concepts are those formed through bodily interaction with the world, and metaphors are regular projections or mappings of conceptual content – concepts or whole structured domains of knowledge – that occur as a way of mentally representing and reasoning about abstract concepts not directly grounded in physical experience. An important claim of this theory is that metaphorical mappings are not arbitrary and unconstrained, but experientially motivated, typically by their grounding in systematic correlations in phenomenal experience. Image schemas and their metaphorical projections therefore provide a solution to the problem of how linguistic expressions and other symbols acquire their meanings, since in this view all abstracta are grounded, at some level, in structures of cognition that emerge from bodily experiences that are directly meaningful to human beings.
It is in this sense, then, that I take cognition and language to be embodied: namely, that human conceptualization, and thus the inferential processes guiding semantic extension, depends in large part on cognitive structures – i.e., images schemas – and construal operations that arise naturalistically from (indeed are analogues of) recurring perceptual and kinesthetic experiences. Through unidirectional mappings of image schematic structure to domains not directly grounded in experience, literal (physicospatial) understanding comes to be extended to abstract reasoning. Because they emerge from, or are grounded in, repeated human bodily movements through space, perceptual interactions, and ways of manipulating objects, image schemas – unlike Fodorian representations (amodal abstract symbols) – are thus directly meaningful representational structures. At the same time, because they are gestalt patterns of experience which capture the structural contours of our bodily interactions with the world (only secondarily imageable “in the mind’s eye”), image schemas – unlike enactivist couplings, which do away with mental representations altogether – are inherently multimodal structures that operate beneath consciousness. Cognitive and linguistic embodiment therefore pertains to the fact that figurative as well as literal understanding is based on at least partial activations of the same sensorimotor areas of the brain.11
Consider, for example, that in many languages what we call anger is conceptualized as heated fluid in a container. In English, this conceptualization is captured in idiomatic phrases such as blow one’s stack, flip one’s lid, and let off steam, where the notion of emotional intensity is mapped to that of the liquid’s temperature, and anger’s effects on the body to the pressurization of the liquid.12 A version of this metaphor also appears in Latin, as indicated by expressions like Quinctius quidem adeo exarsit ira (“Quinctius so ‘blazed forth’ in anger”, Liv. AUC. 35.31.13), mortis fraternae feruidus ira (“‘Seething’ with anger at his brother’s death”, Verg. Aen. 9.736), or ardet et iram / non capit … / … exaestuat ira (She (sc. Procne) ‘burns’ and cannot ‘contain’ her anger … she ‘boils over’ with anger”, Ov. Met. 6.610–11, 623)13. It is the fact that such mappings involve the transfer of an organized system of knowledge from concrete physical experience (namely, how fluids behave in heated, pressurized conditions) to abstract emotional experience (namely, anger) that allows English and Latin speakers to think, and thus talk, coherently about an aspect of human life that may be difficult to comprehend in and of itself. Moreover, talking about anger in terms of hot fluid is immediately meaningful to speakers of these languages, since the metaphor is grounded in the apparently universal human experience of feeling hot and pressurized when angry. Experimental studies have shown that the occurrence of anger coincides with objectively measurable increases in skin temperature and blood pressure.14 In this way, (part of) the physiology of anger itself affords a ready image for conceptualizing such experiences in the abstract.15