Expressions such as “sharp pain” function metonymically when they describe pain that directly results from physical damage, and metaphorically when no such damage is involved. The latter is the case for both extracts above, and, as I will show, in most of the cases where similar descriptions occur in close proximity to the word pain . In the course of the article, I discuss psycholinguistic and neuroscientific research that suggests that these uses of metaphor may facilitate some form of internal embodied simulation of pain experiences on the part of listeners/readers, which may in turn provide the basis for an empathic response. I point out that different metaphorical descriptions of pain are likely to vary in terms of their potential for eliciting a response involving some form of simulation, and in terms of the nature and intensity of the simulation they may elicit. I argue that the most relevant characteristics of metaphorical descriptions of pain in this respect are their level of detail, creativity and textual complexity. For example, I suggest that, other things being equal, descriptions of pain such as the one in the second extract above are likely to facilitate a richer and more intense simulation of pain experiences than that provided in the first extract.

In this article I consider the most dominant metaphorical tendency for the description of pain experiences in English in the light of converging evidence that some form of embodied simulation is involved in comprehension generally, and in the processing of metaphorical expressions in particular. I begin by showing how the sensations conveyed by the English word pain tend to be described via expressions that refer to potential causes of bodily damage. This is the case, for example, with the use of the adjective sharp and of the simile involving a small garden rake in the two extracts below:

Overall, both the BNC and the MPQ provide evidence of the pervasiveness of metaphorical descriptions of pain in English that realize the source domain I have labelled “ CAUSES OF PHYSICAL DAMAGE ”. I will now turn to neuroscientific and psycholinguistic research that suggests that some form of embodied simulation may be involved in the processing of these expressions, and may indeed be part of the motivation for their frequent use.

Some of the metaphorical descriptors listed above are among the frequent collocates of pain I identified in the BNC, while at least some of the others cannot be regarded as equally frequent or conventional as descriptions of pain experiences. Nonetheless, the authors of the questionnaire state that their list of descriptors was derived from several different ‘authentic’ sources, including the medical literature and patients' language use in medical consultations. In addition, the MPQ itself is also likely to exert some kind of influence over the language used by both doctors and patients, due to its widespread use. As I have already noted with regard to stabbing as a collocate of pain in the BNC, some of the descriptors in the MPQ are to do with types of physical damage that most respondents may well not have experienced directly (e.g., lacerating , torturing , as well as stabbing ). While discussing specifically the use of the expression “ stabbing pain, ” Miller (1978 Miller, J. 1978. The body in question , London: Jonathan Cape. ) suggests that “the patient has abstracted from his idea of stabbing an image of violent penetration which he uses in a metaphorical way to refer to his own pain” ( Miller, 1978 Miller, J. 1978. The body in question , London: Jonathan Cape. , p. 28, quoted in Schott, 2004 Schott, G. D. 2004. Communicating the experience of pain: The role of analogy. Pain , 108: 209–12. , p. 210). I will return to this issue below.

In addition, a further set of descriptors metaphorically evoke a malevolent animate agent whose actions may cause physical damage ( punishing, cruel, vicious, torturing, gnawing, killing ), and two groups of descriptors relate, respectively, to high and low temperatures, which, when extreme, can also result in tissue damage: hot, burning, scalding, searing ; cool, cold, freezing . The MPQ also contains several expressions that have basic meanings to do with movement, which would cause tissue damage if it occurred inside the body: beating, pounding, jumping, shooting . The remaining descriptors primarily convey the emotional or affective dimensions of the pain, as in the case of, for example, wretched and annoying .

A wide range of similar expressions for the description of pain are included in the McGill Pain Questionnaire (hereafter the MPQ), which was devised at McGill University in the 1970s in order to be able to assess the pain experienced by different kinds of patients ( Melzack, 1975 Melzack, R. 1975. The McGill pain questionnaire. major properties and scoring method. Pain , 1: 277–99. ). The MPQ has been used for the assessment of a wide variety of types of pain (from backache through labour pain to oncological pain), and has been translated into several other languages. One of the sections of the questionnaire is concerned with the quality of the pain experience, and requires sufferers to indicate what their present pain ‘feels like’ by choosing among 78 one-word descriptors, divided into 20 groups. At least a third of the 78 descriptors of pain included in the MPQ can be described as instantiations of the source domain “ CAUSES OF PHYSICAL DAMAGE , ” and can be further classified in terms of different types of causes of physical damage:

The eight expressions listed above differ in terms of the kind of cause of physical damage evoked by their basic meanings. Three have basic meanings to do with burning: burning , searing and seared (clearly, only two lemmas are involved here). All the others are broadly to do with pointed or sharp objects that can penetrate the body. Sharp describes a property of objects that can cause cuts. Stabbing , stabbed and lanced refer to the process of penetrating something with a pointed or sharp object. Stinging , in its most basic meaning, involves both penetration via a pointed object and the insertion of a harmful substance, as, for example, in the case of a bee sting. The processes or qualities evoked by the basic meanings of these expressions occur fairly commonly in everyday experience, and the type of pain sensation they may cause is widely familiar. The case where this applies less clearly is stabbing/stabbed , as most people do not have direct experience of being stabbed in the sense of being assaulted with a knife or similar object (see also Pither, 2002 Pither, C. 2002. Finding a visual language for pain. Clinical Medicine , 2(6): 570–571. , and De Souza & Frank, 2000 De Souza, L. H. and Frank, A. O. 2000. Subjective pain experience of people with chronic back pain. Physiotherapy Research International , 5(4): 207–19. , p. 217).

A search for the string pain in the BNC returned 7002 hits in 1387 different texts. Collocates of pain were computed on the basis of log-likelihood 3 3 The log likelihood ratio is a widely used method for calculating statistical significance in corpus linguistics, as it does not assume normal distribution (see Dunning, 1993 Dunning, T. 1993. Accurate Methods for the Statistics of Surprise and Coincidence. Computational Linguistics , 19(1): 61–74. ). The log likelihood value of the collocations mentioned below is above 15.13, which corresponds to p < 0.0001. View all notes and within a window span of one word to the left and one word to the right of the search string. The top 62 collocates of this string include eight expressions that can be subsumed under the “ CAUSES OF PHYSICAL DAMAGE ” domain (NB: the numbers in brackets indicate the rank order of each word in the list of collocates): searing (12), sharp (25), stabbing (33), lanced (46), seared (49), stabbed (50), stinging (59), burning (62). An examination of all individual examples of these collocates revealed that either metonymy or metaphor were involved in all occurrences. More specifically, metaphorical uses accounted for over 85 per cent of the instances of the above expressions in close proximity to pain . The rest of this paper will therefore be concerned primarily with metaphorical uses of “ CAUSES OF PHYSICAL DAMAGE ” expressions in the description of pain experiences.

In this section I provide concrete evidence for the dominance and variety of expressions drawing from the “ CAUSES OF PHYSICAL DAMAGE ” domain in the figurative description of pain in English (see also Deignan, Littlemore, and Semino, forthcoming Deignan, A., Littlemore, J and Semino, E. forthcoming. Figurative language in discourse communities , Cambridge, , UK: Cambridge University Press. ), I rely on two different sources of examples of linguistic descriptions of pain experiences: the 100-million-word BNC and a widely used diagnostic questionnaire for pain sufferers, the McGill Pain Questionnaire ( Melzack, 1975 Melzack, R. 1975. The McGill pain questionnaire. major properties and scoring method. Pain , 1: 277–99. ).

On the basis of the metaphor identification procedure proposed in Pragglejaz Group (2007 Group, Pragglejaz. 2007. MIP: A method for identifying metaphorically used words in discourse. Metaphor and Symbol , 22(1): 1–39. ), extract (1) can be described as involving the most basic meaning of the adjective sharp : a physical property of solid objects that have a very thin edge or a pointed end. In extract (2), the adjective is used to describe the kind of nociceptive pain sensation that arises in response to being cut with an object that has a very thin point (a drawing pin, in this case). This use is best described as metonymic, since it involves a cause-effect association between the basic meaning of sharp and what the Pragglejaz Group call the “contextual” meaning of the adjective (i.e., the particular kind of pain sensation evoked in the extract). In contrast, the pain experiences described as sharp in extracts (3) and (4) do not arise as responses to injuries inflicted by means of sharp objects. The pain described in (3) is suspected to be caused by damage to the lining of the stomach, while (4) describes a headache that is not associated with any physical damage at all. It can therefore be argued that, in both cases, the contextual meanings (i.e., those particular kinds of pain sensations) are understood via comparison with the basic meaning: pain that does not result from damage inflicted by external entities is described in terms of a property of objects that can cause injuries associated with a widely familiar kind of nociceptive pain. As a consequence, the use of the sharp in both (3) and (4) can be described as metaphorical. Example (5) differs from all previous examples in that the pain that is being described by sharp (as well as by the simile as if stabbed in the gut ) is primarily emotional: the character in question has just discovered that her husband has been unfaithful to her. While it is difficult to claim that her experience does not involve negative physical sensations as well as emotional distress, the metaphoricity of sharp is particularly clear, as the contextual meaning is primarily to do with feelings of abandonment and distress. 2 2 The metonymic basis of metaphorical descriptions of non-nociceptive and emotional pain in terms of different causes of physical damage can be accounted for by Grady's (1997 Grady, J. 1997. Foundations of meaning: Primary metaphors and primary scenes , Berkeley: Unpublished doctoral dissertation, University of California. ) theory of “primary metaphors” (see also Lakoff & Johnson, 1999 Lakoff, G. and Johnson, M. 1999. Philosophy in the flesh: The embodied mind and its challenge to western thought , New York: Basic Books. ; Lakoff, 2008 Lakoff, G. 2008. “The neural theory of metaphor”. In The Cambridge handbook of metaphor and thought , Edited by: Gibbs, R. W. Jr. 17–38. Cambridge, , UK: Cambridge University Press. ). In Grady's terms, the experiential correlation between simple causes of physical damage (e.g., a blade, a flame) and nociceptive pain gives rise to a primary metaphor that can be labelled “ PAIN IS CAUSE OF PHYSICAL DAMAGE ”. This primary metaphor may in turn provide the basis for more complex metaphors for pain involving source domains such as “ TORTURE ”. View all notes

Different scholars use different labels in order to refer to the semantic fields or source domains that account for the most dominant metaphorical patterns in their data, such as “ Attack,”“Torture,” “Alien invasion,” “Abuse , ” “Impalement ” and so on (e.g., Aldrich & Eccleston, 2000 Aldrich, S. and Eccleston, C. 2000. Making sense of everyday pain. Social Science and Medicine. , 50(11): 1631–41. ; De Souza & Frank, 2000 De Souza, L. H. and Frank, A. O. 2000. Subjective pain experience of people with chronic back pain. Physiotherapy Research International , 5(4): 207–19. ; Lascaratou, 2007 Lascaratou, C. 2007. The language of pain: Expression or description , Amsterdam: John Benjamins. ; Söderberg & Norberg, 1995 Söderberg, S. and Norberg, A. 1995. Metaphorical pain language among fibromyalgia patients. Scandinavian Journal of Caring Sciences , 9: 55–9. ). These different formulations can be subsumed under a general source domain that I will label “ CAUSES OF PHYSICAL DAMAGE ”. This is consistent with the wording of the IASP definition quoted above (see also Scarry, 2008 Scarry, E. 2008. “Among school children: The use of body damage to express physical pain”. In Reconstructing pain and joy: Linguistic, literary and cultural perspectives , Edited by: Lascaratou, C., Despotopoulou, A. and Ifantidou, E. 99–134. Cambridge, , UK: Cambridge Scholars Publishing. ). Clearly, the metaphorical patterns that can be related to this broad source domain have a strong basis in metonymy, as they rely on common cause-effect associations for the experience of nociceptive pain in particular (see Lascaratou, 2007 Lascaratou, C. 2007. The language of pain: Expression or description , Amsterdam: John Benjamins. , pp. 164-5). The uses of the adjective sharp in the following extracts from the British National Corpus (hereafter BNC) can help to clarify the interaction between metaphor and metonymy in the description of pain experiences in terms of potential causes of physical damage of (see also Semino, forthcoming Semino, E. forthcoming. “Metaphor, creativity, and the experience of pain across genres”. In Creativity, language, literature: The state of the art , Edited by: Swann, J., Pope, R. and Carter, R. Palgrave, , UK: Basingstoke. ):

The analyses provided in the above studies, and in relevant work from other areas, suggest that there is some variety in the source domains that are conventionally exploited in order to convey pain experiences. However, the findings of research on several different languages are consistent with Kövecses's (2008 Kövecses, Z. 2008. “The conceptual structure of happiness and pain”. In Reconstructing pain and joy: Linguistic, literary and cultural perspectives , Edited by: Lascaratou, C., Despotopoulou, A. and Ifantidou, E. 17–33. Cambridge, , UK: Cambridge Scholars Publishing. ) general observation that ‘the most important metaphors that provide the phenomenological character of pain’ involve source domains that correspond to the most salient causes of pain. Kövecses's (2008 Kövecses, Z. 2008. “The conceptual structure of happiness and pain”. In Reconstructing pain and joy: Linguistic, literary and cultural perspectives , Edited by: Lascaratou, C., Despotopoulou, A. and Ifantidou, E. 17–33. Cambridge, , UK: Cambridge Scholars Publishing. ) list of relevant conceptual metaphors includes the following:

Indeed, pain shares some of the characteristics of target domains that have received considerable attention in the cognitive linguistic literature. Like “ LOVE ,” for example, it is a private, subjective and poorly delineated experience, which cannot be directly observed. In its prototypical form, it is also a common and widely familiar sensation. However, pain is also an embodied experience, and, in this respect, it is therefore more similar to typical source domains such as “ MOTION ” or “ HUNGER ” than to typical target domains such as “ TIME ”. While a great deal of work has been conducted on the metaphorical construction of emotional experiences in general (e.g., Kövecses, 2000 Kövecses, Z. 2000. Metaphor and emotion: Language, culture, and body in human feeling , Cambridge, , UK: Cambridge University Press. ), it is only recently that pain has begun to receive the attention of cognitive linguists ( Lascaratou, 2007 Lascaratou, C. 2007. The language of pain: Expression or description , Amsterdam: John Benjamins. , 2008 Lascaratou, C. 2008. “The function of language in the experience of pain”. In Reconstructing pain and joy: Linguistic, literary and cultural perspectives , Edited by: Lascaratou, C., Despotopoulou, A. and Ifantidou, E. 35–57. Cambridge, , UK: Cambridge Scholars Publishing. ; Kövecses, 2008 Kövecses, Z. 2008. “The conceptual structure of happiness and pain”. In Reconstructing pain and joy: Linguistic, literary and cultural perspectives , Edited by: Lascaratou, C., Despotopoulou, A. and Ifantidou, E. 17–33. Cambridge, , UK: Cambridge Scholars Publishing. ; see also Halliday, 1998 Halliday, M. A. K. 1998. On the grammar of pain. Functions of Language , 5(1): 1–32. and Semino, forthcoming Semino, E. forthcoming. “Metaphor, creativity, and the experience of pain across genres”. In Creativity, language, literature: The state of the art , Edited by: Swann, J., Pope, R. and Carter, R. Palgrave, , UK: Basingstoke. ).

Attempts to truly describe pain indeed appear as difficult as they are frustrating, yet the need to communicate is overwhelming, and I suggest that the only option available is the resort to analogy … (w)hether by means of metaphor or simile. … ( Schott, 2004 Schott, G. D. 2004. Communicating the experience of pain: The role of analogy. Pain , 108: 209–12. , p. 210)

Within this definition, pain is linked to tissue damage, but it is acknowledged that such damage may be potential as well as actual, and may also occur only in the description of the unpleasant experience on the part of the sufferer. In the rest of this section, I show how different types of pain, including non-nociceptive pain, are often conveyed via expressions that evoke different kinds of (causes of) physical damage. This tendency in the description of pain involves both metonymy and metaphor, and may be explained as an attempt to enable others to experience something akin to the sufferer's own sensations.

While all kinds of pain tend to be associated with affective responses, for chronic sufferers the experience of negative emotions is often inextricably linked with the experience of negative physical sensations. 1 1 See Damasio (1999 Damasio, A. 1999. The feeling of what happens , London: William Heinemann. , pp. 71-9) for a discussion of the distinction between “pain sensation” and “pain affect”. Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/hmet . View all notes This is recognized in the definition of pain provided by the International Association for the Study of Pain (IASP), which also underscores the importance of language:

Pain is also the kind of subjective and poorly delineated experience that is difficult to express satisfactorily in language, especially when the pain is both chronic and at least partly neuropathic (see Scarry, 1985 Scarry, E. 1985. The body in pain: The making and unmaking of the world , Oxford, , UK: Oxford University Press. ). In such cases, there is no visible sign of physical damage, and investigations via X-rays and CAT scans may also fail to detect an obvious cause. Sufferers therefore have to rely primarily on language to communicate their experiences, and to obtain both emotional support and professional help. These are the circumstances in which both patients and doctors report communicative problems, and in which patients tend to feel misunderstood and misbelieved (e.g., Kugelmann, 1999 Kugelmann, R. 1999. Complaining about chronic pain. Social Science and Medicine , 49: 1663–76. ; Lascaratou, 2007 Lascaratou, C. 2007. The language of pain: Expression or description , Amsterdam: John Benjamins. , pp. 174-7).

Pain is a basic and essential human experience. In its prototypical form, it occurs as a response to tissue damage, and constitutes a crucial warning mechanism whose function is to prevent harm to our bodies: the pain experienced when coming into contact with a flame, for example, is due to the damage that the flame causes to our skin, and triggers a reaction (instinctively moving away from the flame) that prevents further damage. This kind of pain (known as “nociceptive” pain) contrasts with less prototypical kinds of pain that are not, or not simply, explainable as a response to tissue damage. “Phantom limb” pain, for example, is felt in parts of the body that have been amputated, and is an example of “non-nociceptive” or “neuropathic” pain. This kind of pain is due to problems within the nervous system that are often difficult to diagnose and treat. More common pain experiences such as migraine and backache tend to have both nociceptive and neuropathic components. This frequently applies to pain that becomes chronic: typically, the pain starts as a result of an episode involving tissue damage, such as an accident, but then persists long after the injuries have healed, for weeks, months, or years.

In sum, a discussion of metaphorical descriptions of pain as possible triggers of embodied simulations needs to take into account their linguistic characteristics, their degree of conventionality or novelty, and the relevant co-text and context. In the next section, I return to the linguistic expression of pain experiences and I propose an approach to their analysis as potential triggers of embodied simulations that attempts to take into account the variety and complexity of actual metaphor use.

In other words, according to this classification, the kind of rich simulation discussed by Gibbs (2006b Gibbs, R.W. Jr. 2006b. Metaphor interpretation as embodied simulation. Mind and Language , 21: 434–458. ) in relation to metaphorical expressions (i.e., option 3 above) results from the activation of complex source domains as discussed in Conceptual Metaphor theory ( Lakoff & Johnson, 1999 Lakoff, G. and Johnson, M. 1999. Philosophy in the flesh: The embodied mind and its challenge to western thought , New York: Basic Books. ). In contrast, the more limited perceptual simulation that involves “only a few related perceptions” (i.e., option 2 above) does not require the activation of the relevant source domain. It may, however, be more appropriate to think of Ritchie's (2009 Ritchie, D. 2009. Relevance and simulation in metaphor. Metaphor and Symbol , 24(4): 249–62. ) three types of expressions and responses as representing different points on a continuum 7 7 This is consistent with Ritchie's latest thinking (personal communication). View all notes , as the distinction between 2 and 3 in particular is overly clear-cut. The rich embodied simulations described by Gibbs do rely on the basic, non-metaphorical meanings of metaphorical expressions, but the activation of these meanings does not necessarily involve the complete activation of complex source domains or whole conceptual metaphors. On the other hand, in several cases the perceptual simulators discussed by Ritchie in relation to conventional metaphorical expressions correspond rather closely to the source domain meanings of those expressions (or, in his own terms to primary perceptual simulators), as when he suggests, for example that interpreting “ A chilly reception ” may involve “the sensation of cold” as well as “the emotion of rejection” (see also the discussion of “ My lawyer is a shark ” above).

Ritchie (2008 Ritchie, D. 2008. X IS A JOURNEY: Embodied simulation in metaphor interpretation. Metaphor and Symbol , 23(3): 174–99. , 2009) usefully attempts to distinguish between different degrees and modes of involvement of “simulation” in the processing of different metaphorical (and non-metaphorical) expressions. His Context-Limited Simulation theory ( Ritchie, 2006 Ritchie, D. L. 2006. Context and connection in metaphor , Basingstoke, , UK: Palgrave Macmillan. ) combines Barsalou's perceptual simulation model of cognition with Sperber and Wilson's (1995 Sperber, D. and Wilson, D. 1995. Relevance: Communication and cognition , Oxford, , UK: Blackwell. ) Relevance Theory. Within Ritchie's theory, all linguistic expressions may activate links to other linguistic expressions as well as a variety of “perceptual simulators”. These include simulations of sensori-motor experiences (e.g., visual characteristics such as shape), proprioceptive experiences (i.e., internal body states such as warmth), and introspective experiences (i.e., cognitive and emotional experiences such as fear). Ritchie's approach to metaphor relies on a (fuzzy) distinction between “primary” and “secondary” simulators associated with words. Primary simulators correspond to the defining characteristics of the relevant concept, while secondary simulators correspond to further, looser associations. For example, the word shark may activate links to other words and phrases (e.g., predator ) and to primary simulators for size, shape, colour, and so on. In addition, the word may activate links to a wider set of linguistic expressions (e.g., surfer ) and secondary simulators for bloodied water, scenes from horror films, emotions such as fear, and so on. ( Ritchie, 2006 Ritchie, D. L. 2006. Context and connection in metaphor , Basingstoke, , UK: Palgrave Macmillan. , pp. 110-13). In Ritchie's definition, words are used metaphorically when, in context, most-- if not all--primary simulators are suppressed as irrelevant, and only some of the secondary simulators remain active. For example, Ritchie argues, the metaphorical use of “ shark ” in “ My lawyer is a shark ” is processed primarily in terms of secondary simulators for introspective emotional states such as fear and distrust. In addition, however, some primary perceptual simulators may nonetheless be weakly activated, such as images of sharp teeth ( Ritchie, 2006 Ritchie, D. L. 2006. Context and connection in metaphor , Basingstoke, , UK: Palgrave Macmillan. . p. 130). Ritchie (2006 Ritchie, D. L. 2006. Context and connection in metaphor , Basingstoke, , UK: Palgrave Macmillan. , p. 215) also recognizes that context-irrelevant simulators may not be completely suppressed if they are “salient” in Giora's (2003 Giora, R. 2003. On our mind: Salience, context, and figurative language , Oxford, , UK: Oxford University Press. ) sense.

Steen (2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol , 23(4): 213–241. ) builds on the work by Gentner and Bowdle in particular in order to distinguish between “deliberate” metaphors, that are processed via some form of comparison involving a cross-domain mapping, and “nondeliberate” metaphors, that are not. Deliberate uses of metaphors are characterized by explicit linguistic devices that aim to shift the recipient's attention towards the source domain, as in the case of “A is B” metaphors, similes, and novel metaphors. Nondeliberate metaphors, in contrast, tend to be conventional, and involve no textual indication of the need to activate knowledge from the source domain. Both Steen and Gentner and Bowdle recognize, however, that the processing of metaphorical expressions is likely to be affected by the textual context (e.g., Gentner & Bowdle, 2001 Gentner, D. and Bowdle, B. 2001. Convention, form, and figurative language processing. Metaphor and Symbol , 16(3-4): 223–247. , p. 233). Steen (2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol , 23(4): 213–241. , pp. 222-3), in particular, points out that the use of conventional metaphorical expressions may be described as deliberate when several words from the source domain are used in close proximity to one another and acknowledges that ‘the full formal range of linguistic and rhetorical construction types for deliberate metaphor is an urgent issue for further research’ ( Steen, 2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol , 23(4): 213–241. , p. 225).

There is evidence from both psycholinguistic and neuroscientific research that the linguistic realization and degree of conventionality of metaphorical expressions affect the way in which they are processed. A series of studies by Gentner and Bowdle suggest that similes and novel metaphors are processed via comparison involving “structure mapping” across domains, while conventional metaphors are processed via categorization--namely, by placing the target concept within an abstract category evoked by the conventionalized metaphorical meaning of the expressions that is used metaphorically ( Gentner & Bowdle, 2001 Gentner, D. and Bowdle, B. 2001. Convention, form, and figurative language processing. Metaphor and Symbol , 16(3-4): 223–247. , 2008; Bowdle & Gentner, 2005 Bowdle, B. and Gentner, D. 2005. The career of metaphor. Psychological Review , 112(1): 193–216. ). These claims are broadly consistent with Giora's (2003 Giora, R. 2003. On our mind: Salience, context, and figurative language , Oxford, , UK: Oxford University Press. ) Graded Salience Hypothesis, according to which the most salient meanings of linguistic expressions are always activated first, regardless of context. As salient meanings may be literal or non-literal, highly conventionalized metaphorical expressions will be processed by accessing the metaphorical meanings directly, while novel metaphorical expressions may require a two-stage process. Indeed, recent brain-imaging research does not suggest a clear-cut distinction between metaphorical and non-metaphorical language, but points to differences between the processing of metaphorical expressions that have lexicalized and “salient” metaphorical meanings, as opposed to the processing of novel metaphorical expressions, where the metaphorical meaning is not salient. Several studies have shown that the brain's right hemisphere is involved in the processing of expressions that realize novel conceptual metaphors, but not in the processing of expressions that have conventional metaphorical meanings ( Ahrens et al., 2007 Ahrens, K., Ho-Ling, L., Chia-Ying, L., Shu-Ping, G., Shin-Yi, F. and Yuan-Yu, H. 2007. Functional MRI of conventional and anomalous metaphors in Mandarin Chinese. Brain and Language , 100(2): 163–171. ; Giora, 2007 Giora, R. 2007. Is metaphor special?. Brain and Language , 100: 111–114. ).

Although the findings of these psycholinguistic studies cannot be straightforwardly extended to the processing of metaphors for pain, they do suggest that some form of embodied simulation may potentially be triggered by metaphorical descriptions such as those involving the “ CAUSES OF PHYSICAL DAMAGE ” source domain. However this claim relies on the problematic assumption that the basic, nonmetaphorical meanings of expressions such as stabbing or burning are involved in the processing of descriptions such as “ a stabbing/burning pain ”. 6 6 For example, it has been suggested that mirroring mechanisms involving the motor areas of the brain are involved in the processing of metaphorical expressions such as “ grasping a concept ” (e.g., Lakoff, 2008 Lakoff, G. 2008. “The neural theory of metaphor”. In The Cambridge handbook of metaphor and thought , Edited by: Gibbs, R. W. Jr. 17–38. Cambridge, , UK: Cambridge University Press. ; Gallese & Lakoff, 2005 Gallese, V. and Lakoff, G. 2005. The brain's concepts: The role of the sensory-motor system in conceptual knowledge. Cognitive Neuropsychology , 22(3/4): 455–79. ), but the experimental evidence is inconclusive (see Aziz-Zadeh et al., 2006 Aziz-Zadeh, L., Wilson, S. M., Rizzolatti, G. and Iacoboni, M. 2006. Congruent embodied representations for visually presented actions and linguistic phrases describing actions. Current Biology , 16: 1–6. ). View all notes Although Gibbs and Matlock (2008 Gibbs, R. W. Jr. and Matlock, T. 2008. “Metaphor, imagination, and simulation: Psycholinguistic evidence”. In The Cambridge handbook of metaphor and thought , Edited by: Gibbs, R. W. Jr. 161–76. Cambridge, , UK: Cambridge University Press. , p. 168) point out that this claim does not necessarily imply a two-stage model of metaphor comprehension, it does raise the issue as to whether all metaphorical expressions are processed in the same way. Gibbs (2006b Gibbs, R.W. Jr. 2006b. Metaphor interpretation as embodied simulation. Mind and Language , 21: 434–458. ) recognizes that his account of metaphor understanding in terms of embodied simulation is not intended to explain all instances of metaphor use, due to the complexity and variety of the ways in which metaphor can manifest itself in communication.

The view of embodied simulation that is relevant to this claim is broader than the approach that is adopted in mirror neuron research (see Gibbs, 2006b Gibbs, R.W. Jr. 2006b. Metaphor interpretation as embodied simulation. Mind and Language , 21: 434–458. ), and relies on the findings of a variety of psycholinguistic experiments. For example, Gibbs, Gould, and Andric (2006 Gibbs, R. W., Gould, J. J. and Andric, M. 2006. Imagining metaphorical actions: Embodied simulations make the impossible plausible. Imagination, Cognition and Personality , 25: 221–238. ) investigated whether people's descriptions of the mental images they formed when reacting to metaphorical expressions such as “ stretch for understanding ” were affected by watching, imitating or imagining the relevant physical action (e.g., physically stretching). They found that, under all three conditions, the majority of informants (78%) talked about performing the relevant action when verbalizing their reactions to the metaphorical expressions. Further evidence in support of the hypothesis that embodied simulation is involved in the processing of metaphorical expressions comes from studies that investigated priming effects. Wilson and Gibbs (2007 Wilson, N. and Gibbs, R. W. Jr. 2007. Real and imagined body movement primes metaphor comprehension. Cognitive Science , 31: 721–31. ) found that informants recognized expressions such as “ grasp the concept ” faster if they had just performed or imagined the relevant physical movement (e.g., grasping an object). Similar priming effects were found by Matlock (2004 Matlock, T. 2004. Fictive motion as cognitive simulation. Memory and Cognition , 32: 1389–1400. ) for sentences involving fictive motion, namely metaphorical expressions drawing from the source domain of “ MOVEMENT, ” such as “ The road goes through the desert ”. Matlock found that informants read and recognized this kind of sentence faster when they had previously read about fast, long-distance travel over an easy terrain, as opposed to slow, short-distance travel over a difficult terrain (see also Matlock, Ramscar & Boroditsky, 2005 Matlock, T., Ramscar, M. and Boroditsky, L. 2005. The experiential link between spatial and temporal language. Cognitive Science , 29: 655–64. ).

I therefore propose that the variety of metaphorical descriptions of pain experiences that draw from the “ CAUSES OF PHYSICAL DAMAGE ” source domain can be seen as potential triggers of embodied simulations of similar experiences. This raises the question of how different metaphorical descriptions of pain differ in terms of the nature of the simulation they may facilitate, or, indeed, in their potential for triggering a simulation at all. This issue is particularly important if we consider that, by and large, we are not exposed to the causes of other people's pain, but only to their verbal descriptions of their pain experiences. The findings of Osaka et al. (2004 Osaka, N., Osaka, M., Morishita, M., Kondo, H. and Fukuyama, H. 2004. A word expressing affective pain activates the anterior cingulate cortex in the human brain: An fMRI study. Behavioural Brain Research , 153: 123–7. ) provide some tentative evidence for the potential of metaphorical descriptions of pain to cause a partial simulation of others' pain experiences. As other existing experimental evidence involves visual stimuli, my discussion is inevitably speculative. It will, however, build on relevant work on metaphor processing, which is briefly discussed next. 5 5 Following Steen, 1994 Steen, G. 1994. Understanding metaphor in literature , London: Longman. , p. 44, I use the terms “metaphor processing,” “metaphor comprehension,” or “processing of metaphorical expressions” to refer to “any psychological process relating to linguistic metaphors.” View all notes

Nonetheless, the relatively basic form of empathy that is mediated by embodied simulation is an important phenomenon in responses to others' pain. In Gallese et al.'s (2004 Gallese, V., Keysers, C. and Rizzolatti, G. 2004. A unifying view of the basis of social cognition. Trends in Cognitive Sciences , 8: 396–403. ) terms, embodied simulation allows an experiential as opposed to a conceptual understanding of another person's experiences. A conceptual understanding of another's experience is achieved when, for example, visual stimuli are interpreted as evidence that someone is performing particular actions or experiencing particular sensations or emotions. In the case of pain, this involves knowing that someone is in pain. An experiential understanding of another's experience, in contrast, is achieved when, for example, visual stimuli activate visceromotor structures that provide an albeit partial first-person simulation of the actions, sensations or emotions that someone else is going through. In the case of pain, this involves experiencing sensory and emotional states that are similar to those one would directly experience in the other person's situation. I suggest that the metaphorical patterns for the description of pain experiences I discussed in the previous section are motivated, at least in part, by the urge to convey one's pain sensations in a way that allows others to experience something that approximates as closely as possible what those sensations feel like. As pain sensations are difficult to put into words, we tend to describe them in terms of situations involving something that causes the most basic kind of physical, nociceptive pain.

Both of the empathic phenomena mentioned by Avenanti et al. need to be distinguished, however, from the conscious attribution of mental states (e.g., beliefs) to others ( Goldman, 2006 Goldman, A. I. 2006. Simulating minds: The philosophy, psychology, and neuroscience of mindreading , Oxford, , UK: Oxford University Press. ), and from the more complex phenomenon known as ‘perspective-taking’—the ability to consider the world from someone else's viewpoint ( Galinsky, Maddux, Gilin, & White, 2008 Galinsky, A. D., Maddux, W. W., Gilin, D. and White, J. B. 2008. Why it pays to get inside the head of your opponent: The differential effects of perspective taking and empathy in negotiations. Psychological Science , 19(4): 378–381. ). In addition, the ‘mirroring’ phenomena observed in the above studies do not exhaust the complexity of the reactions we can have to others' pain. The experience of feelings of compassion for others in pain depends in large part on a range of further factors, such as our relationship with the person in question, our openness to sharing their experiences, our sense of responsibility for their well-being, and so on (see Rizzolatti & Sinigaglia, 2008 Rizzolatti, G. and Sinigaglia, C. 2008. Mirrors in the brain: How our minds share actions and emotions (F. Anderson, Trans.) , Oxford, , UK: Oxford University Press. , pp. 190-2; Cameron, 2010 Cameron, L. 2010. Metaphor and reconciliation: The discourse dynamics of empathy in post-conflict conversations , London: Routledge. ).

It may thus be possible to think of at least two forms of empathy linked to one another in an evolutionary and developmental perspective. A comparatively simple form of empathy, based on somatic resonance, may be primarily concerned with mapping external stimuli onto one's body. A more complex form of empathy, based on affective resonance, may deal with emotional sharing and with the evaluation of social bonds and interpersonal relations. ( Avenanti et al., 2005 Avenanti, A., Bueti, D., Galati, G. and Aglioti, S. M. 2005. Transcranial magnetic stimulation highlights the sensorimotor side of empathy for pain. Nature Neuroscience , 8: 955–960. , p. 958)

The kind of empathy that is mediated by embodied simulation in Gallese's sense is a relatively automatic, unconscious process: it involves a similarity between the sensory and/or emotional states we experience when involved in particular activities, and those we experience when watching others involved in those activities. 4 4 The use of the term “simulation” should not therefore be taken to suggest an exact match between the internal states of self and other, whether in terms of the nature or the intensity of these internal states. Indeed, Gallese (2009 Gallese, V. 2009. Mirror neurons, embodied simulation, and the neural basis of social identification. Psychoanalytic Dialogues , 19: 519–536. , p. 231) acknowledges that the “mirror” metaphor in “mirror neuron” research “is perhaps misleading.” View all notes Avenanti et al. (2005 Avenanti, A., Bueti, D., Galati, G. and Aglioti, S. M. 2005. Transcranial magnetic stimulation highlights the sensorimotor side of empathy for pain. Nature Neuroscience , 8: 955–960. ) make sense of the differences between their findings and those of studies such as Singer et al. (2004 Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R. J. and Frith, C. D. 2004. Empathy for pain involves the affective but not sensory components of pain. Science , 303: 1157–1162. ) and Jackson et al. (2005 Jackson, P. L., Meltzoff, A. N. and Decety, J. 2005. How do we perceive the pain of others? A window into the neural processes involved in empathy. Neuroimage , 24: 771–779. ) by further distinguishing between sensory and affective reactions to others' experiences:

Cumulatively, these studies suggest that some form of simulation observable at the neural level occurs in response to others' sensory and emotional experiences, including particularly pain experiences, and that this simulation may provide the basis for empathic responses to others' experiences. While the observation of neural processes in the research on empathy for pain may suggest a brain-body opposition, Gallese (2009 Gallese, V. 2009. Mirror neurons, embodied simulation, and the neural basis of social identification. Psychoanalytic Dialogues , 19: 519–536. ) emphasizes that the activation of shared neural circuits provides evidence for “embodied” simulation, which he defines as “a crucial functional mechanism of intersubjectivity by means of which the actions, emotions, and sensations of others are mapped by the same neural mechanisms that are normally activated when we act or experience similar emotions and sensations” ( Gallese 2009 Gallese, V. 2009. Mirror neurons, embodied simulation, and the neural basis of social identification. Psychoanalytic Dialogues , 19: 519–536. , p. 520). He adds that:

Several of the above studies also investigated the potential relationship between ‘mirroring’ neural patterns on the one hand, and, on the other hand, the subjects' empathic tendencies and their conscious attribution of pain sensation to others under experimental conditions. Singer et al. (2004 Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R. J. and Frith, C. D. 2004. Empathy for pain involves the affective but not sensory components of pain. Science , 303: 1157–1162. ) found a correlation between amount of activation in the affective areas of the subjects' pain matrix while observing their loved ones in a pain-inducing situation and their scores on a questionnaire that measured their empathic tendencies. Jackson et al. (2005 Jackson, P. L., Meltzoff, A. N. and Decety, J. 2005. How do we perceive the pain of others? A window into the neural processes involved in empathy. Neuroimage , 24: 771–779. ) and Avenanti et al. (2005 Avenanti, A., Bueti, D., Galati, G. and Aglioti, S. M. 2005. Transcranial magnetic stimulation highlights the sensorimotor side of empathy for pain. Nature Neuroscience , 8: 955–960. ) reported a correlation between degree of activity in the relevant areas of the pain matrix during exposure to others' pain and the subjects' ratings of the intensity of the pain experienced by the people who received the painful stimulation (see also Avenanti, Paluello, Bufalari, & Aglioti, 2006 Avenanti, A., Paluello, I. M., Bufalari, I. and Aglioti, S. M. 2006. Stimulus-driven modulation of motor-evoked potentials during observation of others' pain. NeuroImage , 32: 316–324. ). A more recent study using the same method as Avenanti et al. (2005 Avenanti, A., Bueti, D., Galati, G. and Aglioti, S. M. 2005. Transcranial magnetic stimulation highlights the sensorimotor side of empathy for pain. Nature Neuroscience , 8: 955–960. ) interestingly found no mirroring effects in subjects with Asperger syndrome watching videos involving the painful stimulation of hands ( Minio-Paluello, Baron-Cohen, Avenanti, Walsh, & Aglioti, 2009 Minio-Paluello, I., Baron-Cohen, S., Avenanti, A., Walsh, V. and Aglioti, S. M. 2009. Absence of embodied empathy during pain observation in Asperger Syndrome. Biological Psychiatry , 65: 55–62. ), while Xu et al. ( Xu, Zuo, Wang, & Han, 2009 Xu, X., Zuo, X., Wang, X. and Han, S. 2009. Do you feel my pain? Racial group membership modulates empathic neural responses. Journal of Neuroscience , 29(3): 8525–8529. ) provide evidence that the amount of activation in the affective areas of the pain matrix during exposure to others' pain may be lower when the stimuli involve members of a different racial group from that of the subjects.

At least three of the words ( ghan-ghan , kiri-kiri and chiku-chiku ) are similar to the English expressions discussed in the previous section: They evoke situations involving physical damage, and, from the glosses provided by Osaka et al., appear to be used metaphorically to describe pain experiences that do not result from that kind of tissue damage. Osaka et al.'s findings are similar to those of Singer et al. (2004 Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R. J. and Frith, C. D. 2004. Empathy for pain involves the affective but not sensory components of pain. Science , 303: 1157–1162. ) and Jackson et al. (2005 Jackson, P. L., Meltzoff, A. N. and Decety, J. 2005. How do we perceive the pain of others? A window into the neural processes involved in empathy. Neuroimage , 24: 771–779. ): Activity in the anterior cingulate cortex was observed in response to the pain-evoking words, but not in response to the nonsense syllables. In other words, an internal simulation involving the affective component of the neural system for pain seems to be possible in response to linguistic descriptions of pain, as well as in the response to the perception of pain in others. Due to Osaka et al.'s rather general use of the notion of “onomatopoeia,” it is difficult to draw strong conclusions concerning the precise aspects of the pain-evoking expressions that might have been particularly responsible for the effects that were observed.

“(Z)uki-zuki” for throbbing pain with a pulsing sensation, “ghan-ghan” for splitting headache as if being continuously struck, “kiri-kiri” for stabbing pain with a feeling of being drilled into with something sharp, “chiku-chiku” for an intermittent pain akin to being struck by thorns, “hiri-hiri” for a lingering feeling of pain, “zukin-zukin” for continuous throbbing pain. ( Osaka et al., 2004 Osaka, N., Osaka, M., Morishita, M., Kondo, H. and Fukuyama, H. 2004. A word expressing affective pain activates the anterior cingulate cortex in the human brain: An fMRI study. Behavioural Brain Research , 153: 123–7. . p. 124)

A series of studies have shown that some parts of the neural network for the representation of painful experiences (the “pain matrix”) become active both when someone experiences a painful stimulus and when they observe someone else in a pain-inducing situation. More specifically, fMRI brain imagining has revealed activity in the areas of the pain matrix associated with the affective qualities of pain (the bilateral anterior insula and the rostral anterior cingulate cortex) when subjects observed their own partner experience a familiar pain stimulus ( Singer et al., 2004 Singer, T., Seymour, B., O'Doherty, J., Kaube, H., Dolan, R. J. and Frith, C. D. 2004. Empathy for pain involves the affective but not sensory components of pain. Science , 303: 1157–1162. ) and when they watched photographs representing limbs in pain-inducing situations ( Jackson, Meltzoff, & Decety, 2005 Jackson, P. L., Meltzoff, A. N. and Decety, J. 2005. How do we perceive the pain of others? A window into the neural processes involved in empathy. Neuroimage , 24: 771–779. ). In contrast, Avenanti et al. ( Avenanti, Bueti, Galati, & Aglioti, 2005 Avenanti, A., Bueti, D., Galati, G. and Aglioti, S. M. 2005. Transcranial magnetic stimulation highlights the sensorimotor side of empathy for pain. Nature Neuroscience , 8: 955–960. ) found evidence of mirroring effects in the sensori-motor areas of the pain matrix, which is responsible for representing, for example, the localization and intensity of painful experiences. Using single-scope transcranial stimulation, Avenanti et al. noted that, during the observation of painful stimuli, the excitability of the subjects' hand muscles was substantially reduced. This is consistent with what happens when pain is directly experienced in one's hands, as the inhibition of muscle activity contributes to self-preservation.

Over the last two decades, different lines of research in neuroscience, psychology and psycholinguistics have suggested that internal, embodied simulation is involved in a variety of cognitive activities, including imagination and the comprehension of action and language. More specifically, it has been suggested that some form of simulation may be involved in empathic responses to others' pain on the one hand, and in the processing of metaphorical expressions on the other.

VARIATION IN METAPHORICAL DESCRIPTIONS OF PAIN AND EMBODIED SIMULATION

In this section I build on the work discussed in the previous sections in order to suggest that different kinds of metaphorical descriptions of pain drawing from the “CAUSES OF PHYSICAL DAMAGE” source domain may facilitate different kinds of simulation processes. I propose that the nature and intensity of the simulation that may be involved in the processing of different metaphorical descriptions will depend primarily on (a) the property, entity or process that corresponds to the most basic, nonmetaphorical meaning of the metaphorically used word(s); (b) the degree of conventionality or novelty of the metaphorical uses of the relevant expressions, both in relation to pain experiences and in relation to other target domains; and (c) the presence of local metaphorical patterns, which may contribute to the evocation of detailed metaphorical scenarios.

I begin by returning to the adjective sharp, which is used to describe a pain sensation in the first extract from the BNC I quoted at the beginning of the paper. The extract, which is reproduced again below, occurs during an informal conversation among a group of students:

6. Just had a sharp pain go right down the bottom of my leg! (From the spoken demographic section of the BNC, file KWC)

This utterance elicits a response from another speaker (Oh, don't worry about it!), but no other references to pain occur within the part of the interaction that is included in the relevant file in the BNC. As I mentioned earlier, there is ample evidence of the conventionality of the metaphorical use of sharp to describe non-nociceptive pain sensations. Sharp is one of the descriptors for pain that are included in the MPQ. In the BNC, sharp is the 25th most frequent collocate of pain: it occurs 26 times immediately before the word pain. 21 of these occurrences are metaphorical as in example (6) above, amounting to approximately 4.5 occurrences out of 1,000 citations of sharp. This far exceeds the threshold suggested by Deignan (2005 Deignan, A. 2005. Metaphor and corpus linguistics, Amsterdam: John Benjamins. ) for establishing the distinction between conventional or novel metaphorical uses of words on the basis of corpus evidence: Deignan suggests that ‘any sense of a word that is found less than once in every thousand citations can be considered either innovative or rare’ (Deignan, 2005 Deignan, A. 2005. Metaphor and corpus linguistics, Amsterdam: John Benjamins. , p. 40). In addition, the entry for sharp in the Macmillan English Dictionary for Advanced Learners (which is based on a different corpus of contemporary English) includes, amongst others, a meaning of the adjective that is explained as follows: “a sharp pain is sudden and severe”. The adjective also has a range of other similarly frequent metaphorical uses, some of which can be broadly related to the use I am discussing here, as they are to do with sudden, intense and unpleasant experiences (e.g., “a sharp taste,” “a sharp noise,” “a sharp look”). In Giora's (2003 Giora, R. 2003. On our mind: Salience, context, and figurative language, Oxford, , UK: Oxford University Press. ) terms, it can be argued that the salient meanings of sharp are likely to include the sudden and severe quality of some sensations and experiences, including pain sensations. In Steen's (2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol, 23(4): 213–241. ) terms, the metaphorical use of sharp in examples such as (6) can be described as a nondeliberate use of metaphor.

It is possible, in principle, that the use of sharp in descriptions such as (6) may be processed via a simulation that involves the most basic meaning of the adjective. In Ritchie's (2006 Ritchie, D. L. 2006. Context and connection in metaphor, Basingstoke, , UK: Palgrave Macmillan. ) terms, this would involve (primary) sensory perceptual simulators for the visual and tactile characteristics of sharp objects, as well as proprioceptive simulators for the sensation of being cut by a sharp object, and introspective simulators for the resulting emotions of distress and anxiety. However, such a rich simulation is rather unlikely, due to the conventionality of sharp as a metaphorical description of pain and of other negative experiences, and to the fact that no other metaphorical expressions are used by the speaker to evoke a scenario involving physical damage (the other metaphorical expression in the same utterance, “go right down” involves the source domain of “MOVEMENT” and is also rather conventional). Hence, it is more likely that sharp will be processed by accessing directly an appropriate conventional metaphorical meaning. In Ritchie's terms, this may involve the activation of secondary proprioceptive simulators for nociceptive pain and/or introspective simulators for pain-related distress and anxiety. Due to the semantic bleaching caused by the adjective's frequent and varied metaphorical uses, it is even more likely that the use of sharp in the extract above may only activate simulators for generically unpleasant proprioceptive and introspective experiences, or that it may not facilitate a perceptual simulation at all, but simply give access to other relevant words that are associated with it, such as severe or terrible.

Let me now compare the above use of sharp with the use of drilling to describe the pain caused by a headache in the example below, from the novel Regularly Scheduled Life by K. A. Mitchell. The extract occurs after a short stretch of dialogue that takes place in the playing fields of the school where Sean, one of the novel's main characters, is a teacher:

7. The drilling pain started up on the left side of Sean's head again. He couldn't remember ever getting headaches like this before. (Mitchell, 2009, p. 95 Mitchell, K. A. 2009. Regularly scheduled life, Macon, GA: Samhain Publishing. )

The metaphorical use of drilling as a description of non-nociceptive pain is similar to that of sharp in so far as it is part of the same broad linguistic pattern that I have captured in terms of the conventional conceptual metaphor “PAIN IS CAUSE OF PHYSICAL DAMAGE”. Like sharp, drilling is also one of the descriptors for pain that are included in the MPQ. Indeed, the description of the character's pain as drilling in the extract above is unlikely to be perceived as particularly creative, let alone as a one-off. On the other hand, however, this use of drilling is much less conventional than the use of sharp I discussed earlier. In the BNC, drilling never occurs within five words of pain, and I also found no metaphorical collocations with pains, ache, hurt and head. The Macmillan English Dictionary reports no sense of the word that relates to pain. In Giora's (2003 Giora, R. 2003. On our mind: Salience, context, and figurative language, Oxford, , UK: Oxford University Press. ) terms, the description of a particular type of pain sensation is unlikely to be one of the salient meanings of drilling, so that, in context, the use of this expression to describe pain is more likely to be intended and perceived as deliberate (Steen 2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol, 23(4): 213–241. ). In addition, while the basic meaning of sharp is a property of objects, the basic meaning of drilling is the process of making a hole using a very specific kind of tool, which involves both penetration and rotation. Most people are also likely to be familiar with the sensation of having a tooth drilled, which, even under local anaesthetic, can produce a powerful and unpleasant physical sensation. Hence, other things being equal, there is greater potential that the processing of “drilling pain” in example (7) may involve the basic meaning of the adjective, and the activation of some elements of a scenario in which a tool capable of drilling is applied to someone's body. In Ritchie's terms, this would result in a simulation that could involve (primary) sensory simulators for the shape and feel of a drill, proprioceptive simulators for the sensation of being penetrated by a drill, and introspective simulators for the anxiety and distress that would accompany such a situation. While such a rich simulation is, in my view, more plausible than in the case of sharp, it needs to be taken into account that the use of drilling in relation to pain is not entirely novel, and that no other linguistic expression in the co-text contributes to the evocation of a scenario involving physical damage. It is therefore possible that the processing of drilling in the extract above may activate a more limited simulation, involving only secondary proprioceptive simulators for unpleasant physical pressure and/or introspective simulators for emotional distress. In both extracts (6) and (7) any simulation triggered by “sharp pain” or “drilling pain” is also likely to be relatively fleeting, as neither description of pain is elaborated in the relevant local (spoken or written) co-text.

Let me now quote again the extract from the BNC I introduced in section 2 as example 4, which is extracted from a novel:

8. She swallowed again and tried to ignore the terrible sharp pain that was twisting viciously into the side of her head. (From the Imaginative Writing section of the BNC, file HGT)

The pain described here is experienced by a character called Robyn while she is having an awkward conversation with a man with whom she unexpectedly had sex the night before. Robyn's headache has first been mentioned eight paragraphs before the extract above, where it was described as “hanging excruciatingly over one side of her face”. In extract (8) the pain that is first described as terrible and sharp is subsequently presented as “twisting viciously into” the side of the head of the character. In other words, sharp is the first element in a local textual pattern that involves two further metaphorically used words, namely: the verb twist, which suggests a particularly painful way of penetrating the body with a sharp object, and the adverb viciously, which personifies the pain by attributing a particular attitude to it. There is no evidence in the BNC or in corpus-based dictionaries that twisting or viciously are conventionally used metaphorically in relation to pain, although they are consistent with the conventional description of pain in terms of physical damage and malevolent aggression (in fact, vicious is one of the descriptors included in the MPQ). In other words, within the co-text, sharp contributes to a textual pattern that involves some degree of metaphorical creativity, and that, in Steen's (2008 Steen, G. J. 2008. The paradox of metaphor: Why we need a three-dimensional model for metaphor. Metaphor and Symbol, 23(4): 213–241. ) terms, may be described as a case of deliberate metaphor. In addition, the local metaphorical pattern provides enough detail to imagine a specific scenario of physical aggression, or even torture. This arguably creates the conditions for a rich simulation of the kind described by Gibbs (2006b Gibbs, R.W. Jr. 2006b. Metaphor interpretation as embodied simulation. Mind and Language, 21: 434–458. ), for example, whereby a reader may imagine going through the experience of being subjected to a protracted physical attack. This may involve sensory simulators for the shape and feel of an object capable of being twisted into one's head, and possibly of the body of an attacker brandishing this object. In addition, further proprioceptive simulators may be activated for the sensations of internal pressure and pain resulting from that kind of physical attack, as well as introspective simulators for the fear, distress and helplessness associated with being assaulted by a malevolent agent. Any such simulation would be more sustained than in the case of the previous two examples, as the description of the character's headache occupies more textual space, and involves several words contributing to evoke a single pain-inducing scenario.

I will finish by considering two examples where both the creativity and the deliberateness of the metaphorical descriptions are more obvious. Example (9) is a part of a quotation attributed to a migraine sufferer in a factsheet produced by the City of London Migraine Clinic, and was briefly introduced at the beginning of the paper:

9. The pain was like a small garden rake over my eyes and top of my head, digging in and scraping away. (Migraine patient quoted in factsheet produced by the City of London Migraine Clinic)

Here the description of the sufferer's migraine pain involves a simile of the form “A is like B”. Similes have been found to favour processing via comparison involving mappings from source to target domain (see Gentner & Bowdle, 2008 Gentner, D. and Bowdle, B. 2008. “Metaphor as structure-mapping”. In The Cambridge handbook of metaphor and thought, Edited by: Gibbs, R. W. Jr. 109–28. Cambridge, , UK: Cambridge University Press. ): in other words, the basic, physical meanings of the expressions that make up the simile are activated in processing. 8 8In terms of the metaphor identification procedure proposed in Pragglejaz Group (2007 Group, Pragglejaz. 2007. MIP: A method for identifying metaphorically used words in discourse. Metaphor and Symbol, 22(1): 1–39. ), the expressions that are part of similes are used in their basic meanings, and are therefore not used metaphorically (see also Semino, 2008 Semino, E. 2008. Metaphor in discourse, Cambridge, , UK: Cambridge University Press. , pp. 16-17). The metaphoricity of some similes lies in the comparison between the basic meanings of these expressions and the aspects of the topic or target domain that the simile is used to describe. View all notes The description provided via the simile is consistent with the conventional pattern that I have related to the source domain causes of physical damage, but is both more creative and richer in detail than the examples above, and evokes a very specific and vivid metaphorical scenario. This scenario includes both a very precise kind of object that can cause physical damage (“a small garden rake”), and two actions that involve both movement and penetration (“digging in and scraping away”) and that are performed on two very precise parts of the sufferer's head (“over my eyes and top of my head”). None of the expressions that form the local textual pattern in this extract have conventional metaphorical uses to do with the sensation of pain: in fact, garden rake, and any scenario normally associated with it, would normally be unrelated to physical harm but rather have positive emotional associations. For all these reasons, I would argue, this example is likely to facilitate a rich and intense simulation of what it would be like to be in that very specific pain-inducing situation. The words that make up the simile can activate sensory simulators for the size of the rake and its normal setting and function, as well as further sensory associations to do with colour, weight, and so on (depending, in part, on the reader's familiarity with gardening and garden rakes). The level of detail and creativity of the description can further activate proprioceptive simulators for very specific sensations of physical pressure and pain, and introspective simulators for the acute distress and helplessness that would arise when experiencing that kind of pain. Indeed, the quotation from which extract (9) is taken was included in the London Migraine clinic factsheet in order to help convey how a migraine attack is different from the milder headaches that most people are familiar with. Arguably, this is best achieved by putting readers in a position to feel what it is like to experience a kind of pain that they are unlikely to have ever experienced directly.

My final example is taken from a book that resulted from a project involving chronic pain sufferers attending a residential course held at the INPUT Pain Management Unit at St. Thomas's hospital in London (Padfield, 2003 Padfield, D. 2003. Perceptions of pain, Stockport, , UK: Dewi Lewis Publishing. ). Eleven sufferers agreed to work with an artist, Deborah Padfield, to produce photographs that conveyed their experience of pain. The photographs were accompanied by verbal descriptions of what the images represented, which are written in strongly personal and autobiographical terms. Example (10) below is part of the verbal description produced by Frances Tenbeth, who, at the time of the project, had been suffering from chronic pain for 42 years:

10. I am constantly battling with the physical pain. You could possibly describe it as swords on fire. It is as if they are ripping out my leg all the time. Red hot swords. They move. They start in my back and move down relentlessly, like an escalator. […] I think it is probably one rod and a million swords. (Frances Tenbeth in Padfield, 2003 Padfield, D. 2003. Perceptions of pain, Stockport, , UK: Dewi Lewis Publishing. , p. 60)

This extract opens with a metaphorical use of the verb battle to describe the sufferer's relationship with her pain. Frances then explicitly introduces a figurative comparison between her pain and “swords on fire,” and goes on to use metaphorical expressions that are to do with at least two kinds of causes of physical damage: sharp metal objects cutting the flesh (“ripping out,” “rod,” “a million swords”), and heat (“hot,” “on fire”). The hotness of the metaphorical swords is also conveyed metonymically by references to colour (“red”), while the description of the swords includes movement up and down Frances's back and legs (“move down relentlessly,” “like an escalator”). In other words, while all the various figurative expressions can be subsumed under the “CAUSES OF PHYSICAL DAMAGE” source domain, there is creativity in the choice of at least some of the specific expressions that are used (e.g., “swords,” “ripping out”), and in the establishment of a tight textual pattern that involves the combination of different kinds of causes of physical damage (application of sharp objects, movement of sharp objects, and heat), resulting in a rich metaphorical scenario. Frances's description is also very detailed in terms of references to very specific objects, their characteristics and their (hyperbolic) number (“a million swords”).

The description from which this extract is taken is followed by three photographs: a close-up of a red, apparently incandescent spear-like object against a black background, and two images involving human legs covered in several groups of small daggers forming what Frances describes as a “herring bone” pattern. The blades of the daggers have a red shading that suggests heat, and the background is black in both cases. In other words, the interaction between the text and the photographs results in what Forceville (2008 Forceville, C. 2008. “Metaphor in pictures and multimodal representations”. In The Cambridge handbook of metaphor and thought, Edited by: Gibbs, R. W. Jr. 462–82. Cambridge, , UK: Cambridge University Press. ) calls a “multimodal metaphor”:

As a first approximation, I will define multimodal metaphors as metaphors in which target, source, and/or mappable features are represented or suggested by at least two different sign systems (one of which may be language) or modes of perception. (Forceville, 2008 Forceville, C. 2008. “Metaphor in pictures and multimodal representations”. In The Cambridge handbook of metaphor and thought, Edited by: Gibbs, R. W. Jr. 462–82. Cambridge, , UK: Cambridge University Press. , p. 463)

Frances's verbal description makes fully explicit the figurative function of the daggers in the visual images, as well as the fact that their redness is meant to represent the perception of (metaphorical) heat. The multimodal interaction of the verbal text with the visual image adds to the overall deliberateness of Frances's description of her pain.

Overall, Frances's multimodal metaphorical description of her pain has the potential to facilitate a rich and intense simulation including multiple sensory, proprioceptive and introspective simulators. The extract contains a variety of lexical choices that can activate sensory simulators for the colour, size, shape and feel of specific physical objects (“red hot swords,” “one rod”), as well as proprioceptive simulators for the intense physical sensations that would result from repeated penetration and burning via a large number of incandescent sharp objects. The associated introspective simulators for negative emotions are potentially much more powerful and overwhelming than in the case of the more conventional and less detailed descriptions I have discussed.

More specifically, in the case of brief and conventional metaphorical descriptions of pain such as “sharp pain” or “drilling pain” a particularly high degree of involvement is required on the part of the listener/reader in order for processing to involve a rich and intense simulation, or even any simulation at all (see also Cameron's 2010 Cameron, L. 2010. Metaphor and reconciliation: The discourse dynamics of empathy in post-conflict conversations, London: Routledge. notion of ‘being prepared’ for empathy). With examples such as (9) and (10), the opposite is the case. The level of detail, complexity and creativity of these metaphorical descriptions of pain is such that only a deliberately low degree of involvement (or positive resistance to empathy) would prevent a reader/listener from gaining what Gallese et al. (2004 Gallese, V., Keysers, C. and Rizzolatti, G. 2004. A unifying view of the basis of social cognition. Trends in Cognitive Sciences, 8: 396–403. ) call an experiential understanding of the other person's pain, i.e. going through at least part of the sensory, proprioceptive and introspective sensations that one would experience in the situation that is being described. Indeed, examples (9) and (10) were selected for inclusion in different types of publications as particularly powerful descriptions of the experiences of chronic pain sufferers. Example (10), was part of an exhibition which many visitors described as extremely moving and effective. I can also add anecdotally that some members of the audience for a talk including several examples such as (10) (and the accompanying photographs), reported feelings of emotional and physical discomfort. The multimodal nature of the project which gave rise to example (10) is likely to be particularly crucial, as it combines the kinds of effects that are usually treated as separate experimental conditions in neuroscientific research (e.g. Osaka et al., 2004 Osaka, N., Osaka, M., Morishita, M., Kondo, H. and Fukuyama, H. 2004. A word expressing affective pain activates the anterior cingulate cortex in the human brain: An fMRI study. Behavioural Brain Research, 153: 123–7. ; Jackson et al., 2005 Jackson, P. L., Meltzoff, A. N. and Decety, J. 2005. How do we perceive the pain of others? A window into the neural processes involved in empathy. Neuroimage, 24: 771–779. ).

The form of simulation that is likely to be involved in processing the more creative and deliberate metaphorical descriptions would, in Ritchie's (2009 Ritchie, D. 2009. Relevance and simulation in metaphor. Metaphor and Symbol, 24(4): 249–62. ) tripartite classification, fall under the rich, complex and intense embodied simulations of the kind suggested by Gibbs (2006b Gibbs, R.W. Jr. 2006b. Metaphor interpretation as embodied simulation. Mind and Language, 21: 434–458. ). In my discussion, however, I have not adopted strict distinctions between different types of simulations, but I have described potential simulations in terms of their degrees of intensity and richness, and of the range of simulators they may involve. Even the most complex simulations, would not, in my view, involve the activation of complete source domains, but rather the imaginative construction of detailed and specific scenarios, which draw from generic conceptual domains, but do not necessarily correspond to frequent or familiar pain-inducing situations. Indeed, the last two examples show even more clearly than expressions such as “stabbing pain” that we often hyperbolically describe pain sensations in terms of scenarios that we have not experienced directly: the “garden rake” scenario in (9) is implausible but possible, while the “million swords” scenario in (10) is downright impossible. Even the most unrealistic scenarios, however, involve combinations of familiar sensations (e.g., being burned or being cut), or, in some cases, of extreme versions of familiar sensations (e.g., being stabbed or being cut with a million swords). Arguably, therefore, these scenarios can be simulated by integrating different (and relatively familiar) component elements into a single imaginable whole. 9 9The imaginative production and interpretation of scenarios such as that evoked by Frances Tenbeth is also likely to rely on previous responses to descriptions and images of torture and injury in fiction and the media, which make such experiences familiar even to people who have no first-hand knowledge of them (I am grateful to David Ritchie for this observation). View all notes