Download University of California Santa Cruz Diving into the Wreck and more Study notes Poetry in PDF only on Docsity! University of California Santa Cruz Diving into the Wreck: Embodied Experience in the Interpretation of Allegory A dissertation submitted in partial satisfaction of the requirements for the degree of DOCTOR IN PHILOSOPHY in PSYCHOLOGY by Lacey Okonski June 2015 The dissertation of Lacey Okonski is approved: Raymond Gibbs Jr., chair Alan Kawamoto Nick Davidenko _________________________________ Tyrus Miller Vice Provost and Dean of Graduate Studies v cognition drive allegorical interpretation. The same resources used to reason about action and perception are used to comprehend both everyday language as well as poetic instances of allegory. Keywords: allegory, conceptual metaphor theory, embodied simulation, literary response vi Acknowledgements My grandmother was a pioneer of sorts. She grew up on a farm and was one of the first ladies to serve in the United States Women’s Army Corps (WAC). My mother continued this brave legacy earning a Master’s degree from Stanford in electrical engineering and going on to become a rocket scientist at NASA. These two women passed down to me some of their bravery, intelligence, and kindness. This work is dedicated to them and also to Amelie who has yet to be born. I know she will be equally as great. It takes a village to write a dissertation and my support system has been wonderful: Rodger Hayes, Aunt Cookie, Grandpa Sam, Sandra Bingham, Rachael Behrens, Noelle Nehmer, Sabine and Thomas Blaesi, and Gabriel Molina. Thank you to Sejin Hahn who has been a very understanding partner and is always ready with a smile and a glass of wine. Also, thanks to my goddaughter Aubrie who read some of my dissertation and assured me that it’s not as boring as a lot of things she has read. I hope it inspires her to continue to find her own unique voice as she goes about her studies. Thank you to the following artists: Adrienne Rich, Emily Dickenson, and Robert Frost. To Kate Vrijmoet for her intersemiotic translation of Diving into the Wreck. My whole committee took pause to appreciate this work of art on my introductory slide. Thank you to Gabriel Romero and Victoria Ruskovoloshina, the entire Mambo Romero family, and Bay Area dance community. I am beyond grateful to these artists, poets and dancers for adding so much beauty to my world. vii I would also like to thank the members of my dissertation committee, Nick Davidenko, Alan Kawamoto, and Ben Carson, who were truly a pleasure to work with and who always provided me with intellectual stimulation through their careful consideration of my work. There were some scholars out there who, in spite of their celebrity and busy schedules, took time out to have conversations which greatly inspired me: Rolf Zwaan, Mike Kaschak, Roel Willems, Cornelia Mueller, Jean-Remi Lapaire, Szilvi Csabi, Thomas Schack, Bettina Blaesing, Wlad Godzich, Michael Spivey, Durand Begault, Gerard Steen, Meg Wilson, and Bruce Bridgeman. A special thank you to Litze Hu for sharing her deep knowledge of statistics with me. Ron Rogers and Annabel Prinz from San Jose State University helped me to kick off my career as a statistics lecturer, financing the end of my dissertation work. The Gibbs lab is always a haven of intellectual stimulation. Thank you to my labmates: Julie Lonergan, Marcus Perlman, Laura Morett, Nate Clark, Pat Samermit, and Chris Karzmark. My research assistants, many of whom are making me very proud as they have gone on to very respectable PhD programs of their own: Julie Carranza, Miles Hatfield, Cameron Smith, Kaitlin Beatrix, Giselle Stayerman, Daniel Shubat, Emory Strickland, Grant Glander, Adam Zimmerman, Arun Croll, Claire Williams, Matisse Mozer, Julian Rifkin, Emily Huscher, and Laura Kincaid. Thank you to Nicole Wilson who introduced me to applied linguistics through our collaborations at Open English. I am also lucky to have met some brilliant international scholars from around the world while they were working in the lab as post docs: Marlene Johansson Falck, Luciane Correa Ferreira, and Aneider “Ani” 2 Oh, I kept the first for another day! Yet knowing how way leads on to way, I doubted if I should ever come back. I shall be telling this with a sigh Somewhere ages and ages hence: Two roads diverged in a wood, and I— I took the one less traveled by, And that has made all the difference. In this poem, the metaphorical journey theme comes alive because of the specificity, elaboration, and attention to detail, which invites readers to consider a symbolic message beyond a simple story about taking a walk through the woods. This metaphorical message could be about a job, about a relationship, about buying a house, about going on a diet, or about making decisions in life more generally. How do ordinary readers understand that the poem is allegorical when there is no mention of the larger symbolic theme? Allegory Interpretation “Allegoresis is our imaginative projection into other minds and worlds, and is ‘cognitive’ precisely because much abstract cognition is inherently embodied and imaginative.” (Gibbs, 2011) There is a popular folk notion that many people simply aren’t gifted at poetic interpretation because they cannot understand linguistic symbolism. Frost once commented on having access to these literary gifts when he said: “It's only if you're good at it that I'm talking to you. If you haven't been, I don't want you to be let in on any figures I ever made, don't want any teacher to let you in on me.” (Frost, 1953). Frost might be quite right that some scholars are able to do a more nuanced analysis 3 of the poem, yet his embodied allegory remains one of America’s most loved poems (Kettle, 2001). Robert Pinsky’s “Favorite Poem Project1” caught the attention of over 18,000 Americans (with an impressive age range of 5-97 years old). These ordinary people submitted video readings and explanations about their favorite poems and Robert Frost’s “Road Not Taken” came in first place. This overwhelming popular response to the Frost poem suggests that many people enjoy reading this poetic allegory and find its symbols accessible. In spite of Frost’s idealization of his audience, and contrary to the popular notion that average people can’t understand poetry, this allegory has proven accessible to ordinary readers. Yet the actual interpretation of allegory still remains a mysterious process. Some empirical work has shown that people can use their embodied knowledge to understand allegory. Participants in one study read one of two allegorical poems: Robert Frost’s “The Road Not Taken” and Maxine Kumin’s “Ars Poetica: A Found Poem” (Gibbs & Boers, 2005). Each participant wrote out their interpretations for each three-line segment of the poem, and afterward wrote out their understanding of the complete poem. An analysis of these written protocols showed that participants frequently understood the metaphorical and allegorical meaning of the poems. For instance, one participant interpreted the Frost poem first in terms of the literal surface meaning and then metaphorically: “The traveler, in making this tough decision, contemplates for a long time which path to follow. This could be interpreted as a struggle or challenge in one’s life, where one must decide which is the better path or way to go” (ibid). Participants rarely included personal, 1 http://www.favoritepoem.org/project.html 4 idiosyncratic information in their interpretations. These data suggest that participants were quite good at tapping into the allegorical themes implied by the embodied imagery contained in the poems. In another study that focused on embodied allegory (Gibbs & Blackwell, 2012), participants were asked to read the following allegorical story, taken from a contemporary novel, where a poet/author likens his writer’s block to climbing an insecure ladder: “I wish I could spill forth the wisdom of twenty years of reading and writing poetry. But I am not sure I can... Now it’s like I’m on some infinitely tall ladder. You know the way that old aluminum ladders have the texture, that kind of not too appealing roughness of texture, and that kind of cold gray color? I’m clinging to this telescoping ladder that leads up into the blinding blue. The world is somewhere very far below. I don’t know how I got here. It’s a mystery. When I look up I see people climbing rung by rung. I see Jorie Graham. I see Billy Collins. I see Ted Kooser. They’re all clinging to the ladder too. And above them, I see Auden, Kunitz. Whoa, way up there. Samuel Daniel, Sara Teasdale, Herrick. Tiny figures clambering, climbing. “The wind comes over, whsssew, and its cold, and the ladder vibrates, and I feel very exposed and high up. Off to one side there’s Helen Vendler, in her trusty dirigible, filming our ascent. And I look down and there are many people behind me. They’re hurrying up to where I am. They’re twenty-three-old energetic climbing creatures in their anoraks and goggles, and I’m trying to keep climbing. But my hands are cold and going numb. My arms are tired to tremblement. It’s freezing, and it’s lonely, and there’s nobody to talk to. And what if I just let go? What if I just loosened my grip, and fell to one side, and just — ffffshhhooooww. Let go. “Would that be such a bad thing?” (Baker, 2009) Participants were then asked to answer a series of 10 questions, nine of which were about specific phrases from the poem describing climbing the ladder and one question asking people to describe their own bodily sensations when they were reading the poem. For example, when asked to “Please describe what an ‘infinitely tall ladder’ refers to or represents” 62% of participants described the poet’s path to 7 simulations they were having (e.g. being scared of falling on a shaky ladder) and consistency in the associated feelings these embodied situations were likely to evoke (a sense of fear, stress, or anxiety). Even if some of the participants were not always able to verbalize the particular embodied elements they were using in order to create their abstract allegorical interpretations, they still showed implicit knowledge of the metaphorical mappings between the difficulties of ladder climbing and success/failure in one’s career. Language Comprehension and Embodied Simulation The fact that participants in the previously mentioned studies were able to tap into some of the embodied elements and abstract meanings when reading literature supports the idea that they are using embodied simulations to infer allegorical meanings. Embodied simulation refers to the idea that “we understand language by simulating in our minds what it would be like to experience the things that language describes” (Bergen, 2012, p. 13). Even when people are not engaging in overt action, they automatically and imaginatively employ their bodily resources to create an interpretation of a given text or speech sample (Gibbs, 2006). Psycholinguistic and cognitive neuroscience research provides evidence in support of this idea. For example, one study presented people with sentences such as “The carpenter hammered the nail into the wall.” After reading the sentence, participants were shown a picture of an object, such as a nail or elephant, and asked to quickly judge whether that object was mentioned in the sentence. Of course, people quickly say, “yes” to the picture of a nail and “no” to the elephant. The primary interest, however, was with speeded responses to the nail picture, depending on 8 whether it was shown in a horizontal or vertical orientation. Research indicates that people, on average, were faster to make their “yes” decisions when the picture was in the same spatial orientation as implied by the sentence they just read (Stanfield & Zwaan, 2001). Thus, people are faster to say “yes” when the picture showed the nail in the horizontal orientation than when it was shown upright, or in the vertical position. However, when people first read the sentence “The carpenter hammered the nail into the floor,” they were faster, on average, to say “yes” to the nail picture that presented it in a vertical position than when seen in the horizontal orientation. This pattern of results suggests that participants are simulating the implied orientation of objects when they are interpreting sentences. In a similar study, Zwaan and colleagues found that when participants read sentences about an eagle in a nest they were quicker to verify an eagle with its wings tucked. Likewise when they read a sentence about an eagle in the sky participants were quicker to verify an eagle with its wings spread (Zwaan, Stanfield, & Yaxley, 2002). Once again, people are simulating the shapes of objects when they process language. These results suggest that simulations are dynamic because participants do not prefer any depictions of a nail or eagle equally, they prefer the depictions of the nail and the eagle that match the linguistic context that they just read. In this way, simulations are dynamic and context sensitive. The context provides nuanced information and these contextual details are incorporated in simulations. People do not access static representations of “eagle” or “nail” concepts in their mental lexicon, they create specific simulations based on the details of their current linguistic contexts. 9 In a similar line of inquiry, cognitive neuroscience researchers tested to see whether action verbs create body specific neural simulations (Hauk, Johnsrude & Pulvermueller, 2004). Participants were asked to engage in a passive reading task. Hemodynamic response was collected using event related fMRI. Participants showed increased activation in the motor and premotor cortex in body specific areas when they read body specific action verbs so that when they read words like “lick” the areas showing increased hemodynamic response overlapped with the areas of the motor and premotor area which were previously determined to control mouth related action. This experiment illustrates that simply reading a word can activate embodied simulations even when the body itself is not in action. Previous, entrenched knowledge of the body in action contributes to people’s embodied understandings of action verbs. This brief sample of studies is representative of a new trend in cognitive psychology to view language comprehension as a multi-modal, constructive process that relies on embodied simulations as a means of semantic grounding (Bergen, 2012; Gibbs, 2006). Traditional theories of language comprehension posit that language is an amodal, modular process (Fodor, 1983). Still, the question remains: How do people create embodied simulations for abstract, metaphorical language? Some scholars have rejected the embodied simulation view of language understanding because they doubt that people can ever create embodied simulations of language referring to abstract events (Mahon & Caramazza, 2008; McGlone, 2007; Rakova, 2006). 12 When participants heard metaphorical stories about successful relationships they implicitly imagined that the people involved in the relationship moved in similar trajectories, travelled further, and travelled in a straighter line causing them to walk faster as if they too were on a smooth, successful trajectory. These differences were not found when people heard non-metaphorical stories in which the statement “Your relationship was moving along in a good direction,” was replaced with “Your relationship was becoming very important to you,” an expression that does not imply the same kind of embodied movement. It appears, then, that people are simulating the physical, embodied information associated with the source domain when understanding the linguistic instantiations of conceptual metaphors in a discourse context. The activation of the source domain JOURNEY changed the way participants walked or imagined walking while they were still thinking about the story. Those who were simulating a more challenging journey did not walk or imagine walking as far as those who were simulating a smoother trajectory. Another test of the embodied simulation hypothesis examined people’s speeded understanding of metaphorical phrases such as “grasp the concept” (Wilson & Gibbs, 2007). Participants first learned to associate particular symbols with particular bodily actions. For example, when they saw the symbol (“) they learned to produce a grasping movement. After this training phase, participants then sat in front of a computer, saw an icon, enacted the action, and then were presented with a metaphoric statement to understand as quickly as possible. The results showed that making, or imagining making, a relevant body action (e.g., a grasping motion) 13 facilitated the time it took people to read and understand the metaphorical statements. Even if people are unable to physically grasp a concept, engaging in relevant body actions primes the construction of an embodied simulation to infer the metaphorical meaning of “grasp the concept.” Neuroscience work also showed activation in the motor system of participants’ brains when they read both literal (e.g., “grasped the stick”) or metaphorical (e.g., “grasped the idea”) statements (Desai et al., 2012), which offers additional evidence that embodied simulations may be the foundation of our understanding of metaphorical meanings. These results further suggest that metaphor comprehension is embodied and dynamic. Not only are embodied neural resources activated when metaphorical phrases are accessed as the neuroscience results suggest (ibid.), but this also indicates that there is a dynamic relationship between bodily experiences, mental simulation, and metaphorical comprehension. Real and imagined bodily movements prime metaphor comprehension. It is not the case that the participant discretely activates the conceptual metaphor only once they read the metaphor with no influence from previous context. Their enactment or their imagined enactment of related bodily action primes them to understand the metaphor and increases the fluidity with which they can engage in metaphoric language processing. Again this work suggests that simulations associated with metaphoric thought are not static, discrete representations that we access the same way every time. Rather, context plays a crucial role in the embodied simulation process. These experimental results suggest that embodied conceptual metaphors play an important role in people’s interpretation of metaphorical language. Still, how do 14 people infer metaphorical meanings for language that does not explicitly mention a target concept? Once again, there is no reference to life or life decisions more generally in the Frost poem. Can people still construct embodied metaphorical simulations to interpret allegorical texts? My thesis explored this question through four main experimental studies. These studies specifically examined three different empirical hypotheses. Hypothesis 1: The Traditional View Just as Frost hypothesized that average people aren’t good at interpreting symbols (1953) psychologists have traditionally suggested that people are generally poor at constructing simple logical analogies (Gick & Holyoak, 1983). Only when people are given blatant clues can they construct the analogical relationships that are necessary to map the surface meaning of allegorical text to the deeper symbolic relationships. Based on their experimental studies, Gick and Holyoak conclude, “It is thus no easy matter to spontaneously notice an analogy between two superficially dissimilar situations, even in our highly simplified experimental paradigm.” If constructing logical analogies is the basis for metaphor comprehension (Miller, 1979), then ordinary people should not be good at interpreting allegory. This position is also consistent with many theories of literary interpretation, which maintain that only skilled readers should be capable of understanding complex texts with diverse symbolic and aesthetic messages. Hypothesis 2: The Conceptual Metaphor View This hypothesis suggests that people should be capable of understanding allegorical messages given their tacit knowledge of conceptual metaphors. 17 allegorical interpretation arises from ongoing simulations that are freshly constructed each time a person engages in metaphoric thought. This view can be summarized as follows: “We do not first understand the poem through abstract, purely symbolic means and then react to this (i.e., emotionally and aesthetically). Rather, allegory emerges from simulation processes, and is “soft assembled” in the moment of experience depending on state of person, environment and task” (Gibbs, 2011). In contrast to the traditional perspective, this view asserts that humans naturally have an allegorical impulse whereby they relate abstract concepts to enduring embodied experiences. Furthermore, in contrast to CMT, these metaphorical mappings are assembled in the moment of experience (i.e. soft assembly) instead of accessing the exact same conceptual knowledge, as if it were a static chunk of information. Here I argue that the embodied simulations, custom-made during the interpretation of an allegorical poem, are creative, nuanced, and soft assembled in the moment. Although there may be enduring metaphoric themes, such as LOVE IS A JOURNEY, as suggested by previous empirical work (Gibbs, 2012), more work is needed to look at the way that metaphorical thought emerges in nuanced ways. If this metaphorical thought does rely on embodied simulations, participants should engage in allegorical interpretations of the whole poem, but they may not necessarily organize their interpretations around pre-fabricated conceptual metaphors, one at a time. Participants should show evidence of multiple partially activated, source domains that work dynamically instead of displaying a one-to-one source-to-target mapping pattern. Second, participants should construct a coherent narrative of the entire poem so that, instead of a single conceptual metaphor being activated one at a 18 time, participants are able to construct a blend of source domains that enables them to describe different aspects of the narrative and simultaneously contributes to the overall, emergent, allegorical interpretation. The Studies To test these hypotheses, four studies were conducted. Participants read the following poem: Diving into the Wreck by Adreinne Rich First having read the book of myths, and loaded the camera, and checked the edge of the knife-blade, I put on the body-armor of black rubber the absurd flippers the grave and awkward mask. I am having to do this not like Cousteau with his assiduous team aboard the sun-flooded schooner but here alone. There is a ladder. The ladder is always there hanging innocently close to the side of the schooner. We know what it is for, we who have used it. Otherwise it is a piece of maritime floss some sundry equipment. I go down. Rung after rung and still the oxygen immerses me the blue light the clear atoms of our human air. I go down. 19 My flippers cripple me, I crawl like an insect down the ladder and there is no one to tell me when the ocean will begin. First the air is blue and then it is bluer and then green and then black I am blacking out and yet my mask is powerful it pumps my blood with power the sea is another story the sea is not a question of power I have to learn alone to turn my body without force in the deep element. And now: it is easy to forget what I came for among so many who have always lived here swaying their crenellated fans between the reefs and besides you breathe differently down here. I came to explore the wreck. The words are purposes. The words are maps. I came to see the damage that was done and the treasures that prevail. I stroke the beam of my lamp slowly along the flank of something more permanent than fish or weed the thing I came for: the wreck and not the story of the wreck the thing itself and not the myth the drowned face always staring toward the sun the evidence of damage worn by salt and sway into this threadbare beauty the ribs of the disaster curving their assertion among the tentative haunters. 22 wreck to “ribs of disaster curving their assertion among tentative hunters.” These details are quite complex and ostensibly require more effort to process as compared to the simpler style utilized by Frost. Many literary critics have looked beyond the surface of this poem to find allegorical interpretations that criticize old myths about gender.2 Related interpretations include confronting disaster as a survivor and not a victim (Whitman, 1975), an exploration of female fantasy set in primal waters (Milford, 1975), and self- recognition, personal change, mythologizing and suppression of difficult emotions for an androgynous protagonist (Nelson, 1981). Yet, the poem is not limited to feminist interpretation. It could easily be about a failed relationship, the loss of a loved one, or a childhood trauma. Many interpretations use the embodied symbols set forth by Rich to find relevant, grounded ways of constructing allegorical meanings. The imagery used in this poem can even be applied metalinguistically to allegory itself: the reader must dive into the poem and search below the surface meaning of the scuba dive, go through these poetic obstacles, to find the deeper symbolic meaning. The hidden treasure amongst the wreck makes the awkward journey worth the dive into the deep, dark world of literary interpretation. The Explicit Primes The hypotheses were tested by offering participants different primes to frame how they understand the poem in particular ways. In one condition, the prime suggests the “literal” interpretation of the poem as a scuba adventure. Participants 2See: http://www.english.illinois.edu/maps/poets/m_r/rich/wreck.htm for excerpts from published analyses of “Diving into the Wreck” 23 were given the following instructions: “This poem is about a person going scuba diving to explore a sunken wreck.” This prompt encourages readers to look only at the surface value of the poem. Under the traditional view, this condition should be the easiest condition for participants to make some sense of the poem because they are not required to make abstract inferences about what the deeper symbolic meaning of the poem might be. Still, it will be interesting to see if participants spontaneously infer allegorical meanings despite the instructions to attend to literal aspects of the poem. In a second prompt condition, the participants were given the idea that the poem is about a failed relationship. Participants read the following prompt: “The poem is about a person reflecting back on a failed romantic relationship.” Under the CMT view, this should be the easiest condition under which to interpret the poem because participants are primed with a singular conceptual metaphor LOVE RELATIONSHPS ARE JOURNEYS. If metaphorical processing is singular and discrete, then participants should do very well in this condition given that they are provided the target domain, which would facilitate the target-to-source mapping. In a third condition, participants were encouraged to consider multiple meanings by reading the following instruction: “Poetry can often convey several different meanings that you should be open to understanding.” This condition suggests that more than one mapping may be appropriate when interpreting the poem and it overtly recognizes the validity of multiple interpretations of this poem without forcing an interpretation strategy. In allowing more freedom to interpret the poem in a variety of ways, this prompt is also the most challenging. It acknowledges 24 additional meaning but gives the participant no additional clues as to what that additional meaning might be. According to the traditional view, participants will perform poorly in this condition. With no specific interpretation suggested, the allegorical meaning will be challenging because the amodal analogies and schemas will be difficult to generate spontaneously. Under CMT, this prompt should be less effective than the prompt where a metaphorical interpretation is supported by giving the participants the target domain because providing the target domain of a failed relationship should facilitate the one-to-one mapping process between a failed relationship and a scuba journey. However, if CMT critics are correct, this prompt might allow for more varied metaphorical interpretations whereas the failed relationship prompt might encourage a more singular perspective. In a fourth control condition, participants are asked to read the poem with no explicit instructions about how they should interpret the poem. There are no suggestions that the poem should be interpreted literally, as in the scuba condition. Likewise, there are also no suggestions that the poem had additional metaphorical meaning(s), as there was in the failed relationship condition and the multiple meanings condition. This condition was implemented to see what participants infer when left to their own devices in the absence of any suggestions and to provide a baseline from which to compare the interpretations elicited from the other prompts. Furthermore, this baseline condition reflects the most natural reading condition: an absence of priming before reading. Yet in many cases students and scholars do not approach a poem of this caliber without some preconceived notions: introduction from a literature teacher/colleague, popular interpretations, knowledge about the 27 wreck and not just a myth. Then there is confusion whether or not he is saying that he was a person apart of the wreck living and searching for his sense of self and following a myth that was never actually written because it was all in his head. Or he is simply just searching a wreck and becomes fascinated with it.” In this allegorical response the participant never clearly mentions what exactly diving to the wreck symbolizes in a clear metaphorical way, but this response does show allegorical elements with strong metaphorical potential. The protagonist is said to be seeking knowledge, submerging his body and soul, being a part of the wreck, and searching for his sense of self. A scuba diver cannot literally chase after knowledge since knowledge is abstract. They could not literally submerge their soul or become one with a sunken vessel or find themselves. These are all metaphorical components in the answer. Yet because the participant does not name a clear metaphorical target domain (such as linking these activities to “a failed relationship” or “reflecting on a negative event from your past”) this response was coded as literal plus. The possibility remains open that the participant is approaching the act of scuba diving as an allegorical activity in itself, filled with meaning and magic, and the metaphors refer to this complex bodily experience. Another possibility is that the participant, if given a longer amount of time to ponder the poem, may have been able to articulate a fully metaphorical response. Part of the magic of allegory is that the target domain is not mentioned even amidst an atmosphere ripe with potential metaphorical meanings. The many allegorical responses that this data set uncovered evade being categorized as either literal or metaphorical, instead they seem to have a glimmer of metaphorical potential. A very strict protocol was followed about the 28 distinction between literal or metaphorical responses and the literal plus responses reflected those that included elements of between-ness or both-ness. Regardless of what prompt participants were given, it is still interesting to see what strategies they used in their explanations. Do these responses show very little embodied imagery and spell out formal analogies, as the traditional view would predict? Do these responses show organization based upon a single conceptual metaphor and map out only the elements within the poem that fit in with that singular metaphor? Or do these responses show evidence that participants were engaging in a multimodal, mixed metaphor approach as they attempted to interpret the narrative of the entire poem in all of its complexity? The approach that a participant takes towards the poem may have a strong relationship with their interpretation strategy. For example, those who chose to interpret the poem at its surface value may have more in common with the other literal oriented interpretations than they do with the other participants that saw their specific prompt. Source Domain & Mini-Allegory Analysis In Stages 2 and 3 of coding the narrative data, the responses were coded to show the types of allegorical themes that were found in the metaphorical and literary plus interpretations. A global interpretation shows which participants explicitly articulated a metaphorical interpretation but these global impressions miss some of the nuances and complexities found in the details of the responses. These nuances, or mini-allegories are instead found at the word, sentence, and paragraph level. A sentence-by-sentence analysis was used to highlight the pervasiveness of allegorical description in the various response types. While some participants may not have 29 overtly recognized a global metaphorical theme they may have implicitly incorporated these mini-allegories into their responses via metalinguistic metaphors (e.g., “diving into the poem”) or metaphors pertaining to the act of scuba diving (“being lost but wanting to be found”). In Study 1 and Study 2 we conducted a source domain analysis using a data centered approach so that instead of trying to force the allegorical language into a one-to-one, source-to-target mapping, we analyzed the data by the presence of language that could be categorized into the seven most pervasive metaphoric themes that were noted in preliminary data analysis: journey, visibility, verticality, discovery, wholeness, containment, and isolation. This approach proved to be suitable for the highly allegorical data and the analysts were able to code in this way successfully with a high degree of consistency. We evaluated the data sentence by sentence to determine whether each sentence contained elements from one or more of the 7 source domains. Many participants’ metaphorical and allegorical sentences had multiple different source domains that seemed to be at least partially abstract in the context. Consider the following example: “So, to me, the metaphoric aspect of his journey are a trip into his own mind, meditation, or something like it, seeking to reflect on events that have passed and opportunities that have gone by, struggling not to get lost in his own deep desires, but gain something valuable by connecting to a part of himself he often ignores or doesn’t have access to” (Study 2, Multiple Meanings Group). In this example, the participant begins with an explicit recognition of the metaphorical elements and an explicit “journey” theme. Then the participant specifies the journey is a “trip” into the speaker’s “own mind” which immediately 32 across conditions completes the exact same Likert scale4 items before formulating their narrative interpretations. This study shows an even greater propensity towards complex, figurative and allegorical interpretations and a lower rate of simplistic, literal explanatory style suggesting that considering the embodied literary experience questionnaire increased the participant’s allegorical impulse as evidenced by the narrative responses. Participants in Study 3 were exposed to a different implicit prime: we asked participants about their embodied simulations during the reading of the poem (prior to collecting their narrative responses) priming them to think more consciously about imagery or simulations they may have experienced. They answered two open-ended questions prompting self-reports about imagining the enactment of the embodied action in the poems as well as an open-ended question asking them to explain the meaning of the poem in as much detail as possible. The motivation for this change in the study design is that Study 3 provides more details about the embodied simulations that participants experienced while reading the poem by including two additional narrative questions regarding the imagined actions people may have experienced while reading the poem and it also explores how embodied simulation reflection specifically primes participants before they interpret the poem. Data analysis suggests that participants were successful at interpreting the poem in an allegorical way and did so through embodied explanations. In addition, responses to the embodied simulation narrative questions suggest that many participants are engaging in 4 A Likert scale is a rating scale often used in psychological research to assess a participant’s level of agreement or disagreement with a statement along a symmetric range of responses. This measure is often seen as a more sensitive option than presenting participants with a forced choice binary decision about whether they agree or disagree. 33 embodied simulations and that they explicitly report simulations beginning at specific places in the poem. Study 3 results show a relationship between the explicit and implicit prime so that participants who engaged in simulation were more likely to craft a figurative response in the relationship condition than in any other condition. Studies 1-3 showed many metaphorical and allegorical responses and few literal responses. Study 4 sought to strengthen the scuba priming to see if the rate of literal responses could be increased with a stronger prime. Extended stories, crafted to read like magazine articles, were created to be compatible with the scuba explicit prime and the failed relationship explicit prime. The results of study 4 suggest that strengthened primes caused an increase in allegorical responses in the scuba condition and also an increase in metaphorical responses in the failed relationship condition, but no increase in literal interpretations. Study 1 First, participants were asked to read the poem prefaced with one of three primes (literal, metaphorical, or multiple meanings) or with no prime (control condition). Participants were explicitly urged to interpret the poem as being about scuba diving, about a failed relationship, or to examine the multiple possible meanings that the poem may have. In addition to the 3 explicit primes we also included a control group in order to establish a baseline of how participants read the poem in the absence of explicit prompting to take on a particular type of interpretation. Then participants were asked to explain the poem’s meaning in as much detail as possible. Method 34 Participants Sixty-four Psychology students at UC Santa Cruz participated in the study (42 females and 22 males), who received partial course credit for their participation in the experiment. A signed informed consent was obtained from all participants prior to the start of the experiment. Materials The poem was presented on a computer in a private experiment room. Participants read the poem and then completed the rest of the 4-page questionnaire. The questionnaire was presented in a Microsoft Word document. They typed out their responses to the narrative prompt in the word document. Stimuli & Design. Participants were informed that this study examines people’s understanding of poetry. Then they were asked to complete two tasks in a one-hour time slot. First, they were asked to read the poem with one of four prompts. They were encouraged to read the poem slowly so they could answer the subsequent questions regarding the poem. Then they read the prompt and the poem at their own speed. This was a between subjects variable so that any participant would see only one prompt: 1. Control Condition (Control): No prompt N= 14 2. Surface Meaning Condition (Scuba): “This poem is about a person going scuba diving to explore a sunken wreck.” N=17 37 or they implicitly recognized the symbolism by using allegorical language in their response. This manipulation (although created to encourage literal responses) did not increase the tendency to adopt a strictly literal interpretation of the poem. Paired z tests revealed that there were no significant differences in the proportions of literal responses from any group of participants, regardless of what prompt they read. Consider the following comments from a participant who was urged to think about the surface meaning of the poem, “However, beneath the surface the poem is about exploring new, unknown experiences on your own and learning from them, essentially being your own teacher but at the same time being a student. These unknown experiences can help a person grow and learn how to rely on themselves and no one else.” Even though this participant was explicitly primed to interpret the poem in terms of the surface value they still proceed to dig deeper and articulate a metaphorical interpretation. Under Hypothesis 1, we would predict a majority of short, simple, literal responses, especially in the group of participants explicitly primed to think the poem was a simple poem about scuba diving. Instead, this type of literal response is the exception and not the rule. In fact, the responses that articulate one or more metaphorical theme seemed to be a much more typical response type even after reading the scuba prime. This finding suggests that allegorical interpretation is not a special mode of cognition that requires abstract, logical symbol manipulation but rather, allegorical interpretation utilizes a fundamental propensity of human cognition: metaphorical thought. 38 The data from study one results also reveal a second main finding, suggesting that an explicit metaphorical prime may increase people’s ability to interpret metaphorical meaning but also may hinder the range of responses a participant will ponder. In the relationship condition, participants read a prompt that suggested the poem was about a failed relationship to prime participants to facilitate a metaphorical interpretation of the poem. Again, there was only a single participant who adopted a strictly literal interpretation of the poem. However, unlike in the other three conditions, these participants were more likely to adopt a singular metaphorical interpretation (88%). Consider the interpretation of the following participant who was exposed to the failed relationship prime, “A person is reminiscing and revisiting an old relationship and they are unsure of what to expect. He/she attempts to revisit old thoughts to see how things happened/unfolded and see what caused the wreck/fall out in the relationship. Everything that once had life and thrived in the relationship is now gone. The actual ‘wreck’ that he visits can be viewed as the relationship. At the end, he/she then makes it seem like he/she is the bigger person for having revisit the relationship and trying to pinpoint what went wrong.” This explanation is representative of the responses that the failed relationship prime elicited: singular, metaphorical, and focusing on a failed relationship. In spite of possible pigeonholing effects, the metaphorical failed relationship prime did have advantages. Paired z tests, conducted to compare differences in proportions, supported this finding. The failed relationship prompt yielded significantly higher metaphorical responses compared to the control condition (z = 1.97, p<0.01) and significantly higher metaphorical responses compared to the 39 multiple meanings condition (z = 2.71, p<0.001). The failed relationship prompt also yielded a higher proportion of metaphorical responses when compared to the scuba diving prompt but this difference was only marginally significant (z=1.93, p=0.05). None of the other primes were significantly different when compared for the metaphorical responses. Likewise, the failed relationship prime yielded fewer literal plus results than the control condition (z = -2.10, p<0.05), the scuba prime (z = -2.12, p<0.05), or the multiple meanings prime (z = -2.22, p<0.05). These results suggest that the failed relationship prompt did yield a greater proportion of metaphorical responses and a lesser proportion of literal plus responses compared to the other primes. The conceptual metaphor prompt primed participants, with few exceptions, to make a singular metaphorical interpretation. It could be that these effects were two- fold: First, participants who might have otherwise made an allegorical interpretation may have had an easier time articulating the metaphorical connection, often adopting the interpretation that was given in the prompt; likewise participants who may have had multiple or unique interpretations of the poem may have been more likely to adopt the singular explanation that they read as a prime. Thus it facilitated articulating a global, coherent metaphorical interpretation but it may have also pigeonholed the participants from thinking of multiple possible interpretations. Although giving participants this prime may have pigeonholed their responses to some extent, it is important to also note the efficacy of using a conceptual metaphor prompt. This prime did offer participants an advantage that helped them to interpret 42 Journey was found to be the most prevalent source domain amongst the responses. This highly generative figurative theme was found in 47% of all sentences analyzed and was used extensively regardless of prime. One participant making use of this theme stated, “If I take it a bit figuratively, then it could be symbolizing or otherwise referencing a person’s journey in life (or perhaps that of humanity overall…).” The italicized segments highlight the source domain elements utilized in the response. Themes relating to journey like this were found sprinkled throughout the data set highlighting the dominance of the journey related imagery used in interpreting this poem. Containment was another popular source domain as noted by this participant, “The poem projects feelings of drifting away or dying, with the author’s description of aloneness and imagery of being submerged in water.” Again the italicized portion highlights the containment elements as well as a sense of verticality and isolation. The least frequently mentioned source domain was isolation mentioned in only 14% of all sentences analyzed. Although participants may use journey elements more frequently than isolation elements in constructing their responses, these 7 source domains were used with a high degree of consistency regardless of which prime a participant read before crafting their responses and even in cases of literal or literal plus responses. While an individual participant may not use every element in every sentence, they certainly make use of a subset in nearly every response. This high degree of consistency found in the responses, suggests that these embodied elements are primed by the poem. They are active cognitive devices that provide the necessary conceptual structures participants need to construct both literal and allegorical meanings in crafting their responses. 43 While participants in a particular condition may prefer certain source domains to others, all 7 metaphoric themes were utilized frequently. These results support the idea that participants were using dynamic strategies to access metaphoric thought and it casts doubt upon the idea that participants might be accessing metaphors singularly or discretely with clean, neat one-to-one source-to-target mappings. Rather the meaning appears to be emerging from the ability to engage multiple source domains simultaneously. These emergent themes are explored further in the mini-allegory analysis below. Mini-Allegory Data The most prevalent mini-allegory themes and the overall proportion of sentences with each mini-allegory are listed below in Table 3 by source domain. Table 3 Study 1 Mini-Allegory Proportion Data Source Domain Mini-Allegories Journey Search/Exploration (34%) Progress/Impediments (19%) Time (13%) Downward Journey (7%) Inner Journey (6%) Verticality Negative Affect (33%) Depth (24%) Positive Affect (12%) Obstacles (11%) Repression (3%) Visibility Search (45%) Lighting (20%) Contemplation (17%) Appearances (11%) Containment Immersion (22%) Hidden Objects (15%) Knowledge Acquisition (11%) Inner journeys (10%) Time (8%) Fullness (3%) Discovery Loss (18%) Authenticity (15%) 44 Knowledge (13%) Memory (9%) Protection/Coping (3%) Wholeness Broken (14%) Parts/Subcomponents (4%) Isolation Negative Outcomes (30%) Positive Outcomes (15%) In Stage 2 of the source domain analysis, each sentence was analyzed for allegorical themes.7 Participants constructed mini-allegories to support their interpretations of the poem as a whole. Mini-allegories have been previously noted in other discourse settings as a way to tap into larger allegorical themes (Gibbs, Okonski & Hatfield, 2013). For example, there were themes within verticality that showed the ways that participants were allegorically interpreting the downward and upward movement from the poem in various ways. Some participants noted a sense of failure or a damaged psyche/inner-self associated with downward vertical expressions while others used verticality as an obstacle where you might “fall” in love or feel the need to “get over it fast” after being hurt in a relationship. These nuanced patterns that helped to structure the source domain can be organized into mini-allegories. Mini allegories may incorporate more than one source domain. For example, inner journey is a blend of 2 source domains: containment and journey. These mini- allegories sometimes relate to one or more targets, but they do not do so in isolation. They emerge in conjunction with the other source domains and in addition to the other within-source domain mini-allegories. For example, the following participant noted: “The poet is descending into their damaged psyche to examine the failed 7 This analysis was based on whether each sentence contained at least one instance of a source domain being used in an allegorical or metaphoric context in the overall data set. There were 377 sentences considered overall in this analysis. See the appendix for a full set of examples from the data of each mini allegory. 47 priming whatsoever (as in the control condition of Study 1) we can expect few strictly literal responses, the majority of responses being classified as literal plus or with a metaphorical interpretation. This low rate of strictly literal responses suggest that participants do not necessarily need additional priming to make complex metaphorical mappings after reading an allegorical poem. A huge proportion of participants alluded to deeper meanings in the literal plus and metaphorical responses and few adopted a strictly literal interpretation. This pattern of results does not support traditional theories of metaphor cognition. If traditional theories of metaphor were correct there would be a pattern of literal analysis even when participants were given explicit priming to facilitate the metaphorical interpretation (such as the failed relationship prime). Instead the data shows the exact opposite pattern of results. The participants were able to articulate allegorical and metaphorical responses in every condition at a very high rate. Even in the conditions where participants had no clues to the metaphorical meaning (control condition) or when they were explicitly prompted to consider the literal interpretation (as in the scuba condition) participants were still more likely to construct an allegorical or literal interpretation. The results found in Study 1 support the CMT view of metaphor cognition to some extent. The failed relationship prime that was created to promote a conceptual metaphor interpretation of the poem did show an overall increase in metaphorical responses with 88% of the responses utilizing a singular metaphorical general interpretation strategy (all the other conditions fell below 60% for metaphorical responses). This prime was strong enough to redistribute participants who may have 48 otherwise articulated an allegorical response to instead articulate a singular metaphorical response. These findings support the idea that conceptual metaphors are devices with cognitive advantages. Participants who were urged to think of the sunken ship as a failed past relationship were able to draw upon their rich embodied knowledge around journeys and their previous experience with metaphorical love journeys to interpret the poem metaphorically. Yet, these results also suggest that all of the assumptions of CMT may not be completely correct, particularly with respect to the singularity and discreteness assumptions. The prevalence of mini-allegories suggests that participants do not think about metaphors discretely or singularly but instead they blend various mini- allegories to make sense of such a complex poem. This strategy seemed to help participants unite their more global interpretation with their localized interpretations based on individual symbols or stanzas from the poem. These results taken together with the source domain analysis suggest that while conceptual metaphors are powerful cognitive devices, there are much more complex mapping dynamics that cannot be explained with a singular, discrete metaphor. Instead, participants seem to blend various source domains to create a global interpretation of the complex poem they were given to interpret. Study 2 In the second study, participants were exposed to the same explicit primes suggesting a particular interpretation strategy as was used in Study 1. They were also exposed to an implicit prime where they were asked to rate a series of statements that measured embodied literary response ratings on a four-point Likert scale. These 49 items were implemented to urge participants to tap into their subjective literary reactions, their narrative responses, and their embodied simulation processes that they may have experienced during the reading of the poem. Given the results from Study 1, which showed pervasive use of embodied source domains to construct allegorical interpretations, tapping into these literary responses should yield even more robust allegorical and metaphorical responses from participants. First, participants were asked to read the poem prefaced with one of three explicit primes (literal, metaphorical, or multiple meanings) or with no prime (control condition) just as in Study 1. A second task was added to implicitly prime participant’s consideration of their embodied and subjective literary reactions to help them process the poem more deeply before writing out their answers. The literary response task involved reading a series of statements and rating them along a 4-point Likert scale (where 0 = “strongly disagree with this statement” and 4 = “strongly agree with this statement”). Participants were asked to rate 17 statements8. A series of 3 items were designed in order to assess individual interpretation strategies: allegorical meaning, surface meaning, and multiple meaning interpretations. For example, the surface meaning question explicitly assessed the degree to which a participant simulated the source domains in the poem in a literal, scuba context by asking them to rate their level of agreement or disagreement with the following statement: “While reading the poem, I understood that the poem referred to actual experiences with scuba diving.” An additional series of 14 questions were adapted from previous research on embodied literary experiences (Kuiken et al., in press). 8 See Appendix section for the exact items that participants were asked to rate in each experiment 52 Stimuli & Design Participants were informed that this study examines people’s understanding of poetry. Then they were asked to complete three tasks in a one-hour time slot. First, they were asked to read the poem with one of four prompts. They were asked to read the poem slowly so they could answer the subsequent questions regarding the poem. Then they read the prompt and the poem at their own speed. This was a between subjects variable so that any participant would see only one prompt: 1. Control Condition (Control): No prompt N=9 2. Surface Meaning Condition (Scuba): “This poem is about a person going scuba diving to explore a sunken wreck.” N=8 3. Open-Ended Metaphorical Interpretation Condition (Multiple Meanings): “Poetry can often convey several different meanings that you should be open to understanding.” N=8 4. Specific Metaphorical Interpretation Condition (Failed Relationship): “The poem is about a person reflecting back on a failed romantic relationship.” N=9 Second, participants were asked to give their ratings for the 17 Likert scale items. Finally, participants wrote out their narrative responses. The ratings and 53 narrative items were within subjects as all participants were asked to respond to both the ratings and the narrative questions. Results The global interpretation strategy data will be presented first as it reveals the global approach that participants adopted when crafting their responses. A paired z test was conducted to compare the proportions for each prime by global interpretation strategy. Table 5 shows the proportions of global interpretation styles for each of the three primes and for the control condition. Global Interpretation Data Table 5 Study 2: Literary Response Prime Proportion Data Control (N = 8 ) Scuba (N = 9) Relationship (N = 8) Multiple (N =9 ) Metaphorical 75 % 100 % 100 % 100 % Literal Plus 25 % 0 % 0 % 0 % Literal 0 % 0 % 0 % 0 % Compared to Study 1, participants showed an even stronger tendency to interpret the poem allegorically and metaphorically across all conditions. Regardless of what prompt they read, all of the participants (100%) interpreted the poem with allegorical elements (literal plus) or fully articulated the greater metaphorical meaning (metaphorical). For example, one participant who was prompted to read this poem literally as if it were just about a scuba dive concluded, “The poem may also refer to the idea that the narrator did not follow the life path that he intended, and that 54 his life ended up being the wreck. The poem could suggest that he was trying to find his way back or trying to find where he made the mistake that led him along this different path.” These results are representative of the responses from the scuba condition. Every participant who was prompted to consider the poem in this literal way went on to interpret the poem metaphorically (100%). This supports Hypotheses 2 and 3 that claim that participants should be able to reason metaphorically and interpret the poem, even if they are not prompted to interpret the poem metaphorically, because much of cognition is metaphorical. In the control condition, participants were still able to extract metaphorical themes (75%), as exemplified by this response “The poem is obviously discussing a scuba diving experience, in which the narrator is looking for something in the depths of the ocean. But, in a deeper, not so obvious way, the poem is referring to the soul- searches individuals go on through life.” There were no participants who interpreted this poem literally although there were a few responses (25%) that were coded as literal plus that were not explicitly developed into a metaphorical interpretation. The following is an example of a response coded as literal plus, which alludes to metaphorical meaning but does not explicitly articulate a metaphorical interpretation about what the greater meaning of the poem is: “The poem was very metaphorical and can be used to discuss or explain many experiences in life, including scuba diving or something in every day life. It was a very deep poem that could just be about scuba diving, but it depends on how you view it personally. Anyone can take the words and apply it to whatever they are feeling or experiencing at the time.” This response explicitly notes the metaphorical nature of the poem and contains several 57 of the poem. This strengthens the findings from Study 1 that the traditional view is inadequate to explain the trends in this data. Providing participants with a single target domain in the prompt did not necessarily offer an advantage over participants who received open-ended prompts. If interpreting the poem as allegorical was reliant upon a single conceptual metaphor being discretely activated, the failed relationship prime should have offered a clear advantage. This was clearly not the case, at least when participants were primed by the literary response items. Providing participants with a targeted prompt (as in the scuba and failed relationship condition) did seem to limit their interpretations. They did not consider multiple interpretations as much as the participants in the open ended conditions. Critics of CMT might be right that priming with conceptual metaphors encourages stock responses (Tsur, 1999). In some ways, offering this prime was actually limiting instead of productive. These results do not support Hypothesis 2, that participants discretely access prefabricated conceptual metaphors one at a time. Participants across primes displayed an allegorical impulse in every group. They displayed the ability to take the unmapped source domains and create a coherent narrative by grounding and repurposing the embodied features (Gibbs, 2011). The idea that allegorical interpretation is an impulse seems particularly apt in analyzing this data because even when the poem was explicitly marked as literal in the scuba condition, participants were still invited to unpack the metaphoric meanings below the surface. This was true for every condition10. This pattern of results suggests that this allegorical impulse is indeed a very strong cognitive propensity and best supports 10 See Appendix for example responses from participants in each condition 58 Hypothesis 3, that we utilize dynamic simulations to engage in metaphorical thought. The literary response items helped to facilitate our participant’s allegorical impulse and further process the poem’s additional meaning. After interacting with the implicit prime, there were no significant differences in proportions of metaphorical or literal plus response rates across the 4 conditions. Source Domain Data Table 6 displays the proportion of sentences that included at least one element from these 7 themes by prime type and the overall proportion of each source domain in the entire data set: Table 6 Study 2: Participants Use of Source Domains by Prompt Type Condition Journey Visibility Verticality Discovery Containment Isolation Wholeness Scuba 37% 42% 43% 45% 28% 15% 30% Failed Relationship 40% 38% 31% 27% 48% 8% 23% Multiple Meanings 69% 31% 44% 50% 72% 14% 28% Control 37% 32% 22% 34% 39% 17% 22% Overall 42% 35% 33% 38% 43% 13% 24% Participants consistently referred to embodied imagery in the poem when crafting their interpretations. They used the source domains found in the poem as allegorical themes in their responses shifting from the literal use of the source domain to the abstract themes that constituted their allegorical interpretations. The main themes found the data were identical to the themes in Study 1: journey, visibility, verticality, discovery, containment, wholeness, and isolation. Table 6 shows the proportion of participants who used each source domain at least once in their 59 response11. Although there are some preferences to utilize a particular source domain depending upon prompt type, there is also some consistency in that, regardless of prompt type, participants utilized multiple source domains in their responses. For example, while participants in the control group were less likely to use isolation imagery (17%), they still used it to some extent and showed higher rates of using journey (37%), visibility (32%), discovery (34%), and containment (39%). Source domain reference was the most consistent for journey and containment, suggesting that these are important explanatory devices utilized widely when conceptualizing the poem’s allegorical meaning. Participants were likely to incorporate several source domains in crafting their allegorical interpretations although they don’t use every source domain in every sentence. This allows them to blend source domains dynamically, within the sentence and across the response. Consider, for example, the following sentence, “He dove into the relationship and swam around, but it all seems very sad, which can be applied to the relationship that failed.” This participant blends together multiple source domain elements with multiple target domains. Source domains include journey, verticality, and containment and this blend is presented in conjunction with feelings of sadness and failure. This is a complex mapping that can best be accounted for by an embodied simulation approach where metaphorical meanings can be soft assembled on the fly to create nuanced meanings using multiple source domains. Mini-Allegory Data 11 This analysis was based on whether each sentence contained at least one instance of a source domain being used in an allegorical or metaphoric context in the overall data set. There were 178 sentences considered overall in this analysis. 62 This indicates that there are nuanced, context specific simulations of the source domain where different metaphoric themes are blended in order to make sense of this complex poem in it’s entirety. Participants constructed mini-allegories to support their interpretations of the poem as a whole. The same mini-allegory themes from Study 1 were found in this data set. This finding very much challenges Hypothesis 1 and Hypothesis 2. The traditional view would not have predicted this type of rich metaphorical entailments to arise so consistently in the data. Instead this hypothesis would expect great difficulty articulating the meanings behind the poem. These results also call into question CMT because although they suggest some consistency with regards to the metaphorical regularities found in the data, the complexity and nuance of these mini- allegory themes are better explained by a dynamical account of metaphorical cognition. Once again we see that in order to explain and conceptualize a very complex poem participants utilized embodied metaphorical themes but they did so in a nuanced way via mini-allegory, not via formal logic nor via singular and discrete metaphor activation. Study 3 The results from Study 2 suggest that the embodied and subjective literary experiences have a strong priming effect on subsequent interpretations and yield more metaphorical interpretations across conditions. In Study 3, we sought to further explore the embodied literary responses in particular. The same poem and the same 4 prompts were taken from study one, but we added several questions to assess 63 embodied simulations experienced while reading the poem. Then we asked the participants to explain the meaning of the poem: 1. Did you imagine yourself diving at any point while you were trying to understand the poem? 2. If you answered yes, what sensations did you imagine? Please describe in as much detail as possible. 3. Please now write out in your own words what you believe to be the poem’s meaning. Please feel free to write as much as you would like. This exercise worked as a device to gain insight about participant’s awareness of simulations experienced while reading the poem as well as an implicit prime to test the idea that elaborating on embodied simulations directly after reading the poem may facilitate metaphorical thought modes which in turn might prime metaphoric responses. Participants were randomly assigned to one of the four explicit primes that suggested an overall interpretation strategy (control, scuba diving, failed relationship, or multiple meanings). They read the explicit prime, read the poem, completed the embodied simulation questionnaire and then wrote out their interpretations of the poems meaning. Method Participants 52 Psychology students at UC Santa Cruz participated in the study (46 females and 6 males). They received partial course credit for their participation in the experiment. A signed informed consent was obtained from all participants prior to the start of the experiment. 64 Materials The poem was presented on a computer in a private experiment room. Participants read the poem and then wrote out their response to one open ended question. The questionnaire was presented in a Microsoft Word document. They typed out their responses to the narrative prompt in the word document. Stimuli & Design Participants were informed that this study examines people’s understanding of poetry. Then they were asked to complete three tasks in a one-hour time slot. First, they were asked to read the poem with one of four prompts. They were asked to read the poem slowly so they could answer the subsequent questions regarding the poem. Then they read the prompt and the poem at their own speed. This was a between subjects variable so that any participant would see only one prompt: 1. Control Condition (Control): No prompt N= 13 2. Surface Meaning Condition (Scuba): “This poem is about a person going scuba diving to explore a sunken wreck.” N= 13 3. Open-Ended Metaphorical Interpretation Condition (Multiple Meanings): “Poetry can often convey several different meanings that you should be open to understanding.” N= 13 67 Paired comparison z tests12, conducted to compare proportions of the types of global interpretation strategies utilized by the participants in the 4 conditions, revealed that the failed relationship prime elicited a greater proportion of metaphorical responses than the control condition (z = 2.47, p<0.05) or the scuba prime (z = 2.88, p<0.01) and a lower proportion of literal plus responses than the control condition (z = -2.16, p<0.05) or scuba condition (z = -2.57, p<0.05). There were no significant differences amongst the groups for the rate of literal responses and the multiple meanings prime did not yield significantly different proportions from any of the other groups for any response type. These results confirm the efficacy of the combination of the metaphorical prompt and the embodied simulation priming in eliciting metaphorical responses. Embodied Simulation Data A second main finding from Study 3 was that many participants reported experiencing embodied simulations and/or imagery when reading the poem. The data collected from the additional simulation questions revealed a high degree of consistency in having these experiences while reading the poem, having the introspection to be able to report these experiences, and being able to note exactly when the onset of these imaginative experiences began in the poem. The following participant is a very typical example of the self-reports obtained in this experiment: “It happened right after the awkward phase of going down the ladder and into the water. I felt being engulfed by the water all around me, the cold particularly. Also the blacking out before realizing the mask helps to breathe, almost a suffocating feeling.” 12 See Appendix Section C for a complete table of paired comparison z tests for proportions by group 68 This participant from the control condition notes that the simulations happen while going down the ladder and then goes on to explain details about temperature, the proprioceptive experience of being in the water, blacking out, breathing, and suffocating as sensations imagined while reading the poem. Note that this very same onset was found in many other responses, including the following participant who read the failed relationship prime prior to reading the poem: “I imagined myself submerging into the water in a careful manner and cautiously letting go of the last rung of the ladder and releasing myself into all of this dark, blue, black, cold emptiness.” Participants seem to start the imagined sensations while going down from the ladder, noting various visual and proprioceptive sensations from releasing into the water, the darkening light conditions and the increasingly cold temperatures. Table 10 displays the proportions of participants who reported having imagined sensations while reading the poem. Table 10 Study 3 Self-Reported Imagined Sensations Imagined Scuba Diving Scuba Failed Relationship Multiple Meanings Control Yes 69% 85% 85% 92% No 31% 15% 15% 8% As indicated in Table 10, self-reports of experiencing sensations while reading the poem (or perhaps more aptly referred to as embodied simulations) were quite frequent. The scuba group had the lowest report of simulations at a proportion of 69% of participants stating that they did experience some simulations, the 69 metaphorical primes both elicited a proportion of 85% of participants reporting such simulations, and the control group had the highest proportion of participants reporting such simulations with 92% of all participants. This suggests that being primed in a literal way does not necessarily prompt you to imagine the actions described in the poem to a greater extent as the control and figurative primes actually elicited higher rates of self-reported simulations. Paired comparison z tests were conducted to compare the proportion of participants who reported imagining sensations while reading the poem.13 These tests revealed no significant differences reaffirming that participants in every group reported a high degree of embodied simulation in this study. Table 11 Study 3: Simulation Onset Location Onset Location Stanza 1: Scuba Gear Stanza 2-4: Ladder & Water Immersion Generic Diving Responses Proportion 22% 69% 9% Table 11 displays the onset location for self reported simulation in Study 3. Note that there is a high degree of consistency, not at the very start of the action in the poem in Stanza 1 but rather at the moment where the protagonist goes down the ladder and immerses in the water in Stanzas 2-4. Assuming that simulations are not available for conscious inspection, this would lead participants to be unsure of where the imagined sensations began. Unsure or eager to please participants might report simulation onset when the protagonist first acts in the poem (putting on the scuba gear 13 See appendix for a complete table listing the results of the paired comparison z tests. 72 simulations. These results speak to the instrumental role that embodied cognition plays in metaphorical thought and allegorical interpretation. Instead of suppressing literal meaning to think about failed relationship metaphors in an amodal way, participants use this embodied simulation task to creatively aid in metaphorical and allegorical tasks. CMT asserts that a substantial way that people conceptualize abstract language is through embodied grounding and these results are consistent with the idea that conceptual metaphors are productive cognitive devices that rely heavily on deeply entrenched embodied source domains. However, CMT falls short of explaining these results in their entirety. The deeply entrenched source domains necessary to interpret this poem must be crafted in dynamic ways that can only be explained by Hypothesis 3, the Embodied Simulation view of metaphor cognition. The descriptions of simulations during the reading of poem do not show that participants began simulation right at the start of the poem, activated a single conceptual metaphor, and continued full simulation until the end of the poem. Instead, participants seem to have a gradual onset of the simulations that often peaks at the second stanza where the protagonist is going down the ladder. They blend various source domains to create an allegorical ambiance, often switching from sentence to sentence, or even within sentence, from literal language, to metaphorical language, to allegorical language, to metalinguistic diving metaphors that describe exploring the poem itself, to the allegorical symbolism that might be experienced in the actual act of scuba diving. Indeed the poem is so complex that imaginative, dynamic simulations are necessary to construct an explanation that can accommodate local and global interpretations. The static activation of a single 73 metaphor would not suffice. Although embodied source domains and the metaphorical failed relationship prime clearly offer cognitive advantages, these are utilized and described in dynamic way supporting Hypothesis 3. Study 4 The previous results showed very few participants who chose to adopt a strictly literal interpretation of the poem and one possible account for these results is that the scuba prime was too weak. To rule out this possibility, participants were exposed to stronger primes. This experiment sought to push people even further towards a literal interpretation by providing them with a magazine article about the literal act of scuba diving in order to increase the implicit priming of the poem’s surface meaning. In the scuba condition, participants were primed with a story about scuba diving that was modified from an online article used to promote scuba diving tourism. This same article was modified in critical places to create a version that would relate to the metaphorical prompt about relationships. The article was carefully modified so that the syntax and word choice remain almost identical except for key phrases where the second article would instead present metaphorical language describing online dating using scuba imagery.14 For example, in the scuba article participants would read sentences that related to scuba diving such as: “There’s nothing that matches diving into the reef for combining relaxation, fascination and exhilaration all in one activity.” In the relationship article, the participants would see a nearly identical sentence modified to relate to dating: “There’s nothing that matches diving into love for combining 14 See appendix for the entire extended prime articles. 74 relaxation, fascination and exhilaration all in one activity.” Participants were told that they would be participating in two separate studies, the first would examine reader’s understandings of magazine articles and the second would examine reader’s understanding of poems. Then they read and answered a few questions about the so-called magazine article before beginning the poem task. Following the article task participants always read a matching explicit prime: participants who read a scuba article also read a scuba prime before reading the poem and those that read a dating article also read a failed relationship prime before reading the poem. Finally, participant’s narrative responses were collected as they explained the poem in as much detail as possible. Method Participants Sixty-six Psychology students at UC Santa Cruz participated in the study (44 females and 22 males). They received partial course credit for their participation in the experiment. A signed informed consent was obtained from all participants prior to the start of the experiment. Materials The poem was presented on a computer in a private experiment room. Participants read the poem and then wrote out their response to one open ended question. The questionnaire was presented in a Microsoft Word document. They typed out their responses to the narrative prompt in the word document. 77 In the scuba condition, only 2 of the participants took a strictly literal interpretation. More than half of the participants took a literal plus interpretation and nearly half of the participants still adopted a metaphorical interpretation. The rate of literal interpretations does not appear to be noticeably higher than in previous studies in spite of the increased priming. Additionally, the rate of literal responses is exactly equal to the failed relationship condition where 2 out of the 33 participants noted a strictly literal interpretation of the poem. However, it does appear that more responses fell into the literal plus category rather than taking an explicit metaphorical interpretation. We can compare these results with the relationship condition. Again, we see that only 2 participants took a strictly literal interpretation. One single participant took an allegorical interpretation and the vast majority of participants took a singular metaphorical approach (91%). These results show an increase in literal plus responses for those that were primed with a scuba story and a scuba prime. A paired z test to compare this to the proportion obtained using the metaphorical prime revealed that this difference was highly significant (z = 4.42, p<0.001). Likewise, the participants that read a matched article about online dating and then saw the failed relationship prime showed a higher proportion of metaphorical responses. A paired comparison z test revealed that this relationship was also highly significant (z = 4.18, p<0.001). Study 4 tested whether we could strengthen the literal priming and obtain a greater proportion of literal answers. Yet still only 6% of the participants in the scuba condition provided strictly literal responses. This was not statistically different from 78 the proportion of literal responses obtained with the metaphorical primes. In fact, the metaphorical prime yielded an identical rate of literal responses and this was an extremely low proportion compared to the proportion of participants who elected to use literal plus or metaphorical interpretations. However, we did see a rise in literal plus responses from participants in the scuba condition. The scuba prime in Study 1 yielded a rate of about 35% of the responses with literal plus interpretations while Study 4 yielded a rate of about 52% of the responses with a literal plus interpretation. The increased literal priming did not increase the rate of strictly literal responses although it did decrease the likelihood that participants would take an explicitly metaphorical interpretation and increased the likelihood that they would take a literal plus perspective. These results were found to be statistically significant. In this way, the priming seemed to work to inhibit full-blown metaphorical interpretations but it still did not yield a higher rate of literal responses. Did the strengthened dating priming increase the strength of the failed relationship prime? The extended failed relationship prime did increase metaphorical responses compared to the scuba condition. The most common response in this condition focused on the failed relationship scenario given in the prime as exemplified by this participant’s response: “ ‘there is a ladder….I go down.. I crawl like an insect…there is no one to tell me when the ocean will begin’ ^ may be about how one works hard to build a relationship up (up the ladder) but no one ever told them that the relationship may not last forever, the love may sink (down the ladder). The person will be let down, appear weak (like an insect) ‘first the air is blue, and then it is bluer, then green, then black’ ^ at first the love seemed perfect then it gradually got worse and worse ‘whose breast still bear the stress’ ^person still in much pain” 79 This participant takes a singular metaphorical approach to the poem, embraces the failed relationship prime, and very clearly maps out the source domain elements from the poem and the metaphorical entailments of how they relate to a failed relationship. These types of responses underline the strength of the failed relationship prime in that participants do seem to have clearer metaphorical mappings under this condition. Being given this metaphorical interpretation strategy lessens the cognitive demands of both adopting a unique interpretation and also mapping out the meaning of the specific actions from the poem in a very limited amount of time. The strength of these responses provide some support for CMT yet the nuances and blending of the source domain to create mini-allegories suggest more complex dynamics are going into the cognitive process to make these interpretations work both globally for the entire poem as well as line by line. Mini-Allegory Table 14 displays the overall proportions of responses containing mini- allegory in Study 3. Participants included mini-allegories in their interpretation of the poem and the participants over 94% of the time, regardless of whether they saw the scuba primes or the relationship primes. Thus, even when we strengthened the prime to elicit greater proportions of literal responses we still find that participants are incorporating mini-allegory in their responses. This trend was confirmed in all four studies. Mini-allegories are ubiquitous elements when interpreting an embodied allegorical poem. This suggests that participants are not conducting a literal gloss translation of the poem without tapping into the embodied meanings behind the 82 metaphors singularly and discretely. Instead it suggests that participants consider a variety of complexities when weaving together their interpretations. It also alludes to the fact that not every participant took a singular interpretation. Even those who had an explicitly articulated, metaphorical global interpretation often referred to the polysemy of the source domains and incorporated mini-allegories in addition to the main theme. The poem that inspired these studies is far more complex than previous metaphor stimuli and also more complex than the poems used in previous allegory research such as the Robert Frost poem. The data collected from open-ended responses to this complex stimulus reflects an increased complexity and highlights the spectrum of language use that participants have at their disposal from literal to allegorical to metaphorical. Allegorical thought and language holds a special in- between place that floats back and forth from the literal embodied source domains to the abstract target domains without fully committing to either side. Analyzing these responses was a particular challenge as the allegorical nature of the responses evaded clear categorization as either metaphorical or literal. They seldom showed a clear one-to-one mapping between a particular source to a particular target. Instead the dynamics and fluidity of the mapping patterns came to life as allegorical constellations crafted from selecting a subset of mini-allegories. This action creates an allegorical ambiance that serves as a bridge from the literal to the metaphorical. The strategy of many to many mappings allowed participants to tackle such a complex interpretation task. It should be recognized that metaphorical meaning takes place on a continuum, often with many ambiguous cases. 83 The results from the control condition in Study 1 are of particular importance as they reflect the most natural conditions for reading the poem: with no priming whatsoever. When readers appreciate a poem from home or when literary scholars take a critical approach to a poem they are not provided with a pre-formulated interpretation. Instead they approach it on their own without such priming. The prevalence of metaphorical responses and literal plus responses that contained allegorical and metaphorical elements even with such a complex stimulus and even with absolutely no prompting reveals a great deal about cognition. This finding alone builds on previous research looking at allegorical cognition and supports the idea that humans are symbolic creatures endowed with an allegorical impulse to connect enduring embodied experiences with abstract concepts. Hypothesis 1, which suggests a literal default mode of thought, was clearly not supported by any of the studies presented here and this was also not the case even when participants completed the task in the absence of any priming whatsoever. In fact, even when explicitly primed to think of the poem via its surface value, as a poem simply about a scuba trip, many participants were still drawn to full or partial metaphorical explanations. Study 4 sought to strengthen literal primes and increase the proportion of literal responses. While this was partly successful in increasing the literal plus responses there was still a very low rate of strictly literal responses and a nearly unanimous tendency to include mini-allegorical themes inspired from the source domain descriptions given in the poem. The difficulty in eliciting strictly literal responses, even given the strengthened primes in study 4, reaffirms the strength of the allegorical impulse. 84 The overall efficacy of the failed relationship prime, and it’s increased advantages when combined with a brief reflection on embodied simulations experienced while reading through the poem, provide convincing evidence that metaphorical thought is deeply rooted in embodied cognition. Mary Gick and Keith Holyoak (1983), based on the data they had available at the time, assumed that participants have a very hard time making sense of analogies, even with obvious clues. This data demonstrates that even in the absence of obvious clues participants can still interpret embodied allegory (as in the control condition from Study 1) and furthermore, when they are given metaphorical clues these impulses become even stronger. Other traditional views of language assume that the default mode of cognition is literal, that metaphors are simply understood as paraphrased literal expressions stored amodally in the mental lexicon. Yet these results reveal the pervasive use of metaphorical language and the additional advantage of reflecting back on embodied simulations (as in Study 3) or reflecting on embodied and subjective literary responses (as in Study 2). Traditional theories are incapable of explaining the data collected in these four studies. The multiple meanings prompt that suggested the poem may have multiple interpretations and urging participants to remain open to considering these various possibilities seemed to overwhelm participants when presented without any implicit primes, as in Study 1. This explicit prime led participants in Study 1 to display a lower proportion of full blown metaphorical interpretations and an increased proportion of literal plus or strictly literal interpretations compared to the other primes. This was not the case in Studies 2 and 3 where participants were also 87 type but also a general trend to consistently use the 7 source domains in every condition. Finally, source domains were used in dynamic ways supporting Hypothesis 3. There didn’t seem to be one-to-one, source-to-target mappings revolving around a singular conceptual metaphor. Instead, in an allegorical context, there is a network of partially articulated source domains that work together to create an allegorical ambiance, like a constellation of partially developed source domains that cluster together in meaningful ways. The allegorical interpretation emerges from this momentary soft-assembly so that there is not just a set of arbitrary scuba details but a purposeful, symbolic tale with deeper meanings woven into the embodied details. The analysis of mini-allegories further supported this notion that participants are using multiple source domains and multiple mini-allegories to create an allegorical ambiance derived from dynamic, embodied cognitive tendencies. Rather than mapping out metaphorical relationships in a linear fashion, meaning emerged from participant’s construction of their own allegorical ambiance via mini-allegories and multiple source domain references. It is also worth recognizing that the highly prevalent literal plus responses that were allegorical in nature may serve to acknowledge the complexity of the concrete source domain experience through describing that experience allegorically. Many times the source domain is explained as such a simple, basic, easy to describe experience that it can be articulated without conscious effort. Yet this data reveals how the simple source domain elements like “up” or “dark” can be woven into a very complex physical experience. Part of the challenge of unpacking Rich’s poem is in 88 integrating the many details of the source domain to comprehend the overall experience. There is a beauty and richness to this experience alone. While there is a strong cognitive tendency to structure abstract target domains via embodied source domains perhaps it is time to acknowledge that this relationship is more dynamic than previously expected. At times the source domain may be highly complex and partly understood via abstract target domains. There may be a much stronger tendency to use the source domain to conceptualize the target but this relationship may not be strictly unidirectional. In conclusion, this work provides data that helps broaden the scope of current empirical work. Metaphors are so generative that they can be hard to capture in a laboratory setting, both in terms of creating figurative stimuli and also in terms of measuring participant’s linguistic responses to such language. Several decades of scholarship focusing on metaphor suggest that figurative thought shapes much of cognition, yet there has been little focus on extended forms of metaphor in the literature. There are certainly benefits to running clean, controlled studies yet there are still many questions that cannot be addressed by noting participants responses to isolated, sanitized metaphors. How do people reason metaphorically when they experience allegorical interpretations in extended, richer contexts? How do mixed metaphorical themes work together, build upon each other, and blend in meaningful ways? How do people respond differently to language that was created by poets who are focused on imagery, style, and aesthetics rather than laboratory stimuli that was created to cleanly isolate single metaphorical themes? How do participants appreciate the source domain itself as a complex and allegorical experience? More 89 research needs to be done to study people’s complex, messy, beautiful interpretations to uncover these complex, embodied, interpretative processes. The four studies presented in this thesis build upon and extend the existing literature by looking at how people utilize embodied simulation in abstract language processing when presented with an extended, complex allegorical poem. These results help provide a glimpse into the world of metaphor cognition, to move beyond crudely asserting that a single conceptual metaphor is activated, and reveal the mapping dynamics that flexibly accommodate multiple metaphoric themes as the cognitive system works towards interpreting a complex allegorical poem. Traditional theories of metaphor cognition and CMT are inadequate to explain the data presented in this dissertation on embodied allegory. This is undeniable evidence that metaphorical cognition is much more dynamic than previously suspected.