Elisabeth Camp

The contributors to this volume all argue that poetry is a tool for epistemic achievement, and that Emily Dickinson uses poetry both to understand the world and to advocate for poetry as a tool of understanding. Many also argue that Dickinson offers a distinctive construal of knowledge, as a continual process of grappling with a world that transcends complete grasp, through daily cognitive, emotional, and practical labor. While some aspects of the resulting portrait fit the stereotype of Dickinson as a reclusive poet observing “small moments” in nature and in her own mental life, the Dickinson we encounter here is decidedly more determined, argumentative, and hopeful than that stereotype allows. This not only re-writes our conception of Dickinson; it provides lessons for philosophy, by offering alternative characterizations of what knowledge is, and of the methodologies through which it can be achieved.

Words are used for many things: to describe, to plan and promise, to invite and command. They are also used to wound—to demean, insult, and exclude. The fact that words can have such potent, pernicious effects is puzzling, because they are, after all, just words. As the schoolyard chant goes, “Sticks and stones may break my bones, but words can never hurt me.” Words do hurt though—not only one’s feelings, but one’s social status, even one’s basic dignity as a human being. How can sounds and shapes do all that? Many philosophers have thought of language as a kind of game. Both games and language are complex, abstract structures that are deployed strategically to achieve serious goals, as well as for fun. Thinking through these similarities can illuminate how something so intangible can have such powerful effects. And seeing how people wield that power for malicious ends can reveal how to turn the tables and fight back.

Philosophers of science in the last half century have emphasized that scientific theories are not sets of transparently interpretable, logically connected true descriptions; rather, they involve implicit appeal to only partially articulated theoretical, practical, and empirical assumptions, and depart from stating the truth in various ways. One influential trend treats scientific theorizing as largely a process of model construction, and analyzes models as fictions. While this chapter embraces the increased role accorded to imagination and interpretation in scientific practice by the models-as-fictions view, it argues that different scientific representations relate to the world in importantly different ways. It distinguishes among a range of distinct representational tropes, or “frames,” all of which function to provide a perspective: an overarching intuitive principle for noticing, explaining, and responding to some subject. Starting with Max Black’s metaphor of metaphor as a pattern of etched lines on smoked glass, the chapter explains what makes frames in general powerful cognitive tools. It then distinguishes metaphor from some of its close cousins, especially telling details, just-so fictions, and analogies, first in the context of ordinary cognition and then in application to science, focusing on the different sorts of gaps that frames or models can open up between scientific representations and reality.

One of America's most celebrated poets, Emily Dickinson was virtually unpublished in her lifetime. When a slim volume of her poems emerged on the American scene in 1890, her work created shockwaves that have not subsided yet. Famously precise and sparse, Emily Dickinson's poetry is often described as philosophical, both because her poetry grapples with philosophical topics like death, spirituality, and the darkening operations of the mind, and because she approaches those topics in a characteristically philosophical manner: analyzing and extrapolating from close observation, exploring alternatives, and connecting thoughts into cumulative demonstrations. But unlike Lucretius or Pope, she cannot be accused of producing versified treatises. Many of her poems are unsettling in their lack of conclusion; their disparate insights often stand in conflict; and her logic turns crucially on imagery, juxtaposition, assonance, slant rhyme, and punctuation. The six chapters of this volume collectively argue that Dickinson is an epistemically ambitious poet, who explores fundamental questions by advancing arguments that are designed to convince. Dickinson exemplifies abstract ideas in tangible form and habituates readers into productive trains of thought--she doesn't just make philosophical claims, but demonstrates how poetry can make a distinct contribution to philosophy. All essays in this volume, drawn from both philosophers and literary theorists, serve as a counterpoint to recent critical work, which has emphasized Dickinson's anguished uncertainty, her nonconventional style, and the unsettled status of her manuscripts. On the view that emerges here, knowing is like cleaning, mending, and lacemakingL a form of hard, ongoing work, but one for which poetry is a powerful, perhaps indispensable, tool.

Recent theorizing about concepts has been dominated by two general models: crudely speaking, a philosophical one on which concepts are rule-governed atoms, and a psychological one on which they are associative networks. The debate between these two models is often framed in terms of competing answers to the question of "how the mind works" or "the nature of thought". I argue that this is a false dichotomy, because thought operates in both these ways. Human thought utilizes representational structures that function as arbitrary re-combinable bits. This supports a version of the Language of Thought Hypothesis though a significantly more modest one than is typically advanced by advocates of that view. But human thought also employs representational structures that are contextually malleable, intuitive, and holistic; I call these "characterizations". "Dual systems" models of cognition recognize this multiplicity of mental processes, but typically posit largely separate structures, and emphasize conflicts between them. By contrast, I argue that the two forms of representation are more closely integrated, and more symbiotic, than talk of duality suggests.

People speak in different ways within and across contexts; and utterances that are equivalent in their essential content and force can differ in their social significance in virtue of differing phonologically, morphologically, syntactically, and lexically. These dimensions of linguistic variation have not received sustained philosophical attention. We argue that meaningful form contributes significantly to the speech acts that agents perform with their utterances. Speakers communicate social information by navigating fine-grained differences in how they speak, and hearers systematically pick up on and use this information. Moreover, these fine-grained differences performatively alter the conversational context. We review three options for handling these aspects of speech within the philosophy of language—exclusionism, informationalism, and performativism—and sketch prospects and challenges for a theory of sociolinguistic variation going forward.

Humans are inveterate storytellers. We make incessant and insistent narrative sense of the world around us and of our place in it, so much so that some scholars have suggested "homo narrans" as a more appropriate identifying description for our species than "homo sapiens". Indeed, a long-standing tradition holds that our very self-identities have an essentially narrative shape: that who each of us is is determined by the stories of our lives, and that in some sense we create ourselves by crafting those stories. In this essay, I focus on an especially compelling case of narrative self-construction: Wordsworth's Prelude. I argue that we do need rich, substantive selves of the sort delivered by narratives like The Prelude, both in order to evaluate our past actions and to guide future ones. However, the very feature which makes Wordsworth's poem so rhetorically powerful as an autobiography—his invocation of a robust teleological structure, which is imposed on him from infancy by Nature also prevents us from embracing it as a model for our own self-understanding, because it conflicts sharply with modern views about ontology. Contemporary advocates of a narrative conception of the self, such as Jerome Bruner, Alasdair MacIntyre, and Marya Schectman, drop The Prelude's objectionable ontological assumptions. But rather than placing the narrative conception of self on a firm metaphysical foundation, this actually intensifies the threat of fictionalism: the risk that the selves we fashion through stories are mere self-deluding illusions. I conclude by gesturing toward the characters within stories as an alternative literary model that avoids many of these problems.

What is distinctive about literary metaphors? Why do authors use metaphors in literature? I argue that metaphors in literature, like metaphors elsewhere, allow authors to communicate thoughts and stake claims about how the world is. To make this claim plausible, we need, first, to free ourselves of an overly restrictive conception of ordinary discourse, which can be open-ended, nuanced, and imagistically and emotionally evocative, just like literary metaphors; and in particular which can also present contents through perspectives. Second, we need to recognize the interpretive differences that are generated by the literary context: because literary texts are published works of art, literary meaning is constructed as a collaboration between a 'model author' and a 'model reader', each of whom has access to both more and fewer interpretive assumptions than the actual writer and recipient do. Third, we need to attend to the diversity among literary metaphors, which can be laser-focused as well as open-ended, abstract as well as concretely imagistic and emotional, and which can stake truth-evaluable claims while also presenting non-propositional perspectives. Finally, metaphors differ from other perspectival tropes, like exemplification, in presenting one thing through the lens or filter of something else, producing a kind of twofoldness that gives metaphors a distinctive rhetorical and cognitive power.

Contemporary philosophical orthodoxy treats names as universally accessible, arbitrary tags that track referents across space, time and possibility. I argue that paradigmatic nicknames, like ‘Shrimpy’, ‘Crooked Hillary’ and ‘Bubblegum’, do track referents, but are marked by contrast with proper names by enforcing restrictions on who they can be used by and with, and when they can be used; and by framing their referents under affectively valenced social identities. While proper names can also carry social information, it is part of nicknames’ characteristic function to manage face, in ways that an adequate overall theory of meaning needs to explain.

The Politics of Language is a significant advance in the nascent theory of social meaning. It departs from orthodox theories of meaning in prioritizing audience uptake over speaker production and the alignment of emotional affect and social identity over the rational exchange of information. However, in their zeal to offer a radical, ‘non-ideal’ alternative to orthodox philosophy of language, Beaver and Stanley the orthodoxy’s genuine insights about aspects of communication that stem from our treating each other as rational agents, and aspects of language that stem from its role as a system for building complex structures out of discrete parts. Together, these produce a distorted description of the role that literal meaning plays in communal negotiations about communicative responsibility, in turn limiting their ability to diagnose and combat the malign speech practices they are most concerned to elucidate.

Most philosophical and linguistic theorizing about meaning focuses on cooperative forms of communication. However, much verbal communication involves parties whose interests are not fully aligned, or who do not know their degree of alignment. In such contexts, speakers sometimes turn to insinuation: implicatures that permit deniability about risky attitudes and contents. I argue that insinuation is a form of speaker's meaning in which speakers communicate potentially risky attitudes and contents without adding them to the conversational record, or sometimes even to the common ground.

We do many things with words. We describe, we plan and promise, we invite and command, we hint and intimate. We also use words to wound – to demean, insult, and exclude. The fact that words can have such potent, pernicious effects is puzzling, because they are, after all, just words. As the schoolyard chant goes, “Sticks and stones may break my bones, but words can never hurt me.” Words do hurt though–not only our feelings, but our social status, even our basic dignity as human beings. How can sounds and shapes do all that? Many philosophers have thought of language as a kind of game. Both games and language are complex, abstract structures that we deploy strategically to achieve serious goals, as well as for fun. Thinking through some of these similarities can illuminate how something so intangible can have such powerful effects, and seeing how people wield that power for malicious ends can reveal how to turn the tables and fight back.

I distinguish among a range of distinct representational devices, which I call "frames", all of which have the function of providing a perspective on a subject: an overarching intuitive principle or for noticing, explaining, and responding to it. Starting with Max Black's metaphor of metaphor as etched lines on smoked glass, I explain what makes frames in general powerful cognitive tools. I distinguish metaphor from some of its close cousins, especially telling details, just-so stories, and analogies, in ordinary cognition and communication. And I use these distinctions to illustrate different sorts of gaps that frames or models can open up between representation and reality.

Philosophers have long debated the relative priority of thought and language, both at the deepest level, in asking what makes us distinctively human, and more superficially, in explaining why we find it so natural to communicate with words. The “linguistic turn” in analytic philosophy accorded pride of place to language in the order of investigation, but only because it treated language as a window onto thought, which it took to be fundamental in the order of explanation. The Chomskian linguistic program tips the balance further toward language, by construing the language faculty as an independent, distinctively human biological mechanism. Devitt (Ignorance of language. Clarendon Press, Oxford, 2006) attempts to swing the pendulum back toward the other extreme, by proposing that thought itself is fundamentally sentential, and that there is little or nothing for language to do beyond reflecting the structure and content of thought. I argue that both thought and language involve a greater diversity of function and form than either the Chomskian model or Devitt’s antithesis acknowledge. Both thought and language are better seen as complex, mutually supporting suites of interacting abilities.

Our ordinary and theoretical talk are rife with “framing devices”: expressions that function, not just to communicate factual information, but to suggest an intuitive way of thinking about their subjects. Framing devices can also play an important role in individual cognition, as slogans, precepts, and models that guide inquiry, explanation, and memory. At the same time, however, framing devices are double-edged swords. Communicatively, they can mold our minds into a shared pattern, even when we would rather resist. Cognitively, the intuitive power of a frame can blind us both to known features that don’t fit easily within the frame, and also to “unknown unknowns” we have not yet encountered. Thus, perhaps Locke is right to disavow such “eloquent inventions” as “perfect cheats” that “insinuate wrong ideas, move the passions, and thereby mislead the judgment.” Against this, I argue that while the metaphor of double-edged swords is indeed apt, this is because frames are tools for thought. Like any tool, they can be used well or badly; but they do not fall outside the realm of rationality altogether. I describe how framing devices express open-ended perspectives, which produce structured intuitive characterizations of particular subjects. I argue that frames can make effective, distinctive epistemic contributions in the course of inquiry, and that the cognitive structures that frames produce can contribute to, and constitute, epistemic achievements in their own right, even in highly idealized circumstances at the nominal end of inquiry. Throughout, I focus especially on scientific understanding, because it serves as a paradigm case of rational inquiry, from which frames and perspectives are most likely to be excluded.

Identity labels like “woman”, “Black,” “mother,” and “evangelical” are pervasive in both political and personal life, and in both formal and informal classification and communication. They are also widely thought to undermine agency by essentializing groups, flattening individual distinctiveness, and enforcing discrimination. While we take these worries to be well-founded, we argue that they result from a particular practice of using labels to rigidly label others. We identify an alternative practice of playful self-labelling, and argue that it can function as a tool for combating oppression by expressing and enhancing individual and collective agency.

I argue that stories are ‘equipment for living’ in two senses: retrospectively, they provide ‘configurational comprehension’ of a temporal sequence of events; prospectively, they offer templates for action. Narrative conceptions of the self appear well poised to leverage these functional roles for stories into an intuitively compelling view of self-construction as self-construal. However, the narrative conception defines selves in terms of the lives they live: a self is the protagonist in a lifelong story. And narrative structure is itself defined by ‘retrospective necessity’: the meaning of events within a story is given by their contribution to that story's ending. Together, this entails that life stories hold selves metaphysically, epistemically, and practically hostage to their ends. Fortunately, narratives are just one species of interpretive frame. I suggest some alternative types of frames, including identity labels and metaphors, that support configurational comprehension, action guidance, and self construction without shackling selves to their lives’ ends.

Slurs are incendiary terms so much that many ordinary speakers and theorists deny that sentences containing them can ever be true, and utterances where they occur embedded within normally "quarantining" contexts, like conditionals and indirect reports, are still typically offensive. At the same time, however, many speakers and theorists also find it obvious that sentences containing slurs can be true; and there are clear cases where embedding does inoculate a speaker from the slur's offensiveness. I argue that four standard accounts of the "other" element that differentiates slurs from their more neutral counterparts semantic content, perlocutionary effect, presupposition, and conventional implicature all fail to account for this puzzling mixture of intuitions about truth, and for this mixture of projection and quarantining. Instead, I propose that slurs make two distinct, coordinated contributions to a sentence's conventional communicative role: predication of group membership and endorsement of a derogating perspective on the group. Predication of group membership is "at issue" by default, but different semantic and conversational contexts can alter the relative prominence and scope of the two contributions.

I consider what it might mean for an utterance to "express" an attitude in a way that differentiates expressive from descriptive speech, and how individual words might have the conventional function of performing such acts of expression. Drawing on recent work in formal semantics, I argue that while there are promising models for implementing the intuition that certain expressions and constructions, for instance, deontic and epistemic modals, pejoratives like "damn", and slurs, express non-cognitive states and/or have dynamic non-truth-conditional effects, these models are not easily extended to the classic case of "thin" ethical terms.

Ernie Lepore and Herman Cappelen (2005) argue that contextual influences on semantic content are much more restricted than most theorists assume, by presenting three tests for semantic context-sensitivity and concluding that only a very restricted class of expressions pass them. They combine this extreme semantic minimalism with an even more extreme speech-act pluralism, according to which a speaker has said anything that she can be reported as having said. I argue that because Lepore and Cappelen refuse to distinguish what is said from what is claimed, their tests wrongly classify metaphor as semantically context-sensitive. I then argue that our ordinary linguistic practices support a distinction between what is said and what is claimed, and that underwrites a much more moderate form of speech act pluralism.

Mobsters and others engaged in risky forms of social coordination and coercion often communicate by saying something that is overtly innocuous but transmits another message ‘off record’. In both ordinary conversation and political discourse, insinuation and other forms of indirection, like joking, offer significant protection from liability. However, they do not confer blanket immunity: speakers can be held to account for an ‘off record’ message, if the only reasonable interpreta- tions of their utterance involve a commitment to it. Legal liability for speech in the service of criminal behavior displays a similar profile of significant protection from indirection along with potential liability for reasonable interpretations. Specifically, in both ordinary and legal contexts, liability depends on how a reasonable speaker would expect a reasonable hearer to interpret their utterance in the context of utter- ance, rather than on the actual speaker’s claimed communicative intentions.

I compare two of Davidson's main discussions of metaphor. I argue, first, that despite some puzzling inconsistencies, the overall thrust of “What Metaphors Mean” is a radical form of noncogitivism, on which speakers of metaphors merely cause their hearers to perceive certain features in the world, but do not claim or implicate that things are any particular way. By contrast, in “A Nice Derangement of Epitaphs,” Davidson endorses a neo‐Gricean account of metaphor as a form of speaker's meaning. However, he also now advocates a radical anticonventionalism about linguistic meaning in general, which threatens to undermine his new view of metaphor. Finally, I articulate and evaluate a neo‐Davidsonian view of metaphor, which retains as much as possible from both papers.

The papers collected in this volume were initially presented at the Third International Symposium of Cognition, Logic, and Communication: “A Figure of Speech.” This was a conference dedicated to philosophical aspects of metaphor, held in Riga, Latvia in December 2007. I believe that these papers nicely reflect both the range of specific topics, and the range of positions on those topics, that have received the most attention in studies of metaphor within philosophy and related fields like linguistics and cognitive science. Here, I offer a rapid overview of the theory of metaphor, in order to situate the papers in relation to one another and within the field more generally.

Category mistakes sentences like Julius Caesar is a prime number, Colourless green ideas sleep furiously, or Saturday is in bed are theoretically interesting precisely because they are marginal: as by-products of our linguistic and conceptual systems lacking any obvious function, they reveal the limits of, and interactions among, those systems. Do syntactic or semantic restrictions block is green from taking "Two" as a subject? Does the compositional machinery proceed smoothly, but fail to generate a coherent proposition or delimit a coherent possibility? Or is the proposition it produces simply one that our paltry minds cannot grasp, or that fails to arouse our interest? One's answers to these questions depend on, and constrain, one's conceptions of syntax, semantics, and pragmatics, of language and thought, and of the relations among them and between them and the world.

Nick Shea’s Representation in Cognitive Science (OUP 2018) places a broadly teleosemantic account of mental representation on a realist footing by stressing the need to explain *how* a representational system tracks information, with different formats exploiting structural correspondences between vehicles and contents in different ways. I explore how representational realism interacts with ecumenicalism about format, and more specifically how different representational systems distribute the burden of representation between systemic and local levels. Variations in systematicity and structure matter functionally, by affecting systems’ representational capacities and vulnerabilities. And they matter theoretically, by affecting where and how we posit and test for representational mechanisms.

In “Games and the Art of Agency,” Thi Nguyen argues that games both highlight and foster a profound complexity in human motivation, in the form of “purposeful and managed agential disunity.” I agree that human agency is “fluid and fleeting” rather than stable and unified; but I argue that Nguyen’s analysis itself relies on a traditional conception of selves as enduring goal-driven agents which his discussion calls into question. Without this conception, games look more like life, and both look riskier, than we might otherwise hope.