Publications
Dissertation
Works in Progress
This essay is about a variety of action errors I call “Mistaken Actions” and a puzzle they pose for Causal Theories of Action. I begin with Davidson’s version of the Causal Theory and pose the puzzle as a dilemma: if his theory explains what makes a mistaken action mistaken, then it can’t explain what makes it intentional; and if his theory explains what makes a mistaken action intentional, then it can’t explain what makes it mistaken. I then propose three ways in which more recent Causal Guidance Theories may be able to resolve the dilemma but argue that none yield a definitive solution.
Epistemological orthodoxy has it that memory is, in some sense, a kind of knowing, or as Lai (2022a), quoting Lackey (2008: 251), puts it, “an epistemic source” which “preserves knowledge from one time to another.” Lai dubs this the “Epistemic Conception of Memory" (ECM). But several recent papers argue that ECM falls prey to counterexamples. Of these, Lai (2022a) is the most thorough, as it takes aim at four distinct variants of ECM. The present paper introduces a neglected fifth variant of ECM we call “Capacitism.” According to Capacitism, remembering that p is the exercise of an epistemic capacity that functions to afford knowledge that p. It is argued that Capiticism is an attractive variant of ECM which avoids Lai’s counterexamples as well as a further problem facing other variants of ECM tied to their commitment to memory’s factivity, namely, their resulting commitment to either metaphysical or epistemological disjunctivism. Capacitism avoids these problems by grounding memory’s fundamental link to knowledge in its distinctive teleological, epistemic function for knowledge acquired in the past.
While mental disorders like schizophrenia are paradigmatic human kinds, the going view of human mental kinds like declarative memory is that they are not human kinds. This essay challenges the going view by arguing that declarative memory bears the distinguishing marks of a human kind. We propose that a “constructionist-functionalist” metaphysics for mental kinds like declarative memory can be articulated on the model of existing accounts of other constructed kinds. Finally, we argue that such a metaphysics addresses the more pressing objections to our claim that mental kinds like declarative memory may properly be considered human kinds.
Building off of a recent account of intentionality, the action-forward framework, we introduce a novel account of episodic memory and defend its plausibility as a how-possibly account of such memory. According to our account, episodic remembering consists in acts of performative retelling of the remembering subject’s initial experience of an event where such retelling activates connected capacities for perceptual recognition of the event and its constituents. When such acts are overt, the remembering subject expresses their memory, say, verbally, imagistically, through performance, etc. where doing so reactivates those recognitional capacities. When such acts are covert, the remembering subject merely activates capacities for such expressions without going on to overtly express where doing so reactivates the relevant recognitional capacities. In §2, we introduce two desiderata on accounts of episodic memory. In §3, we give an overview of the action-forward framework for intentionality. In §4, we introduce our account of episodic memory. In §5, we show how the account satisfies the desiderata. We conclude by saying how the account differs from standard accounts in the literature while retaining what we take to be their insights.
According to the traditional approach in philosophy of memory, when all goes well, our episodic memories of particular events in our personal past constitute firsthand knowledge of the who, what, where, and what-was-it-like of those events. That is, according to the traditional approach, episodic memory is at bottom a capacity for a specific kind of knowledge. However, it’s now becoming increasingly common to treat the core epistemic dimension of episodic memories as present but non-essential, that is, as secondary to whatever episodic memory exists to do. What are the grounds for this “non-epistemic turn” in the philosophy of memory? We identify three based on the empirical study of memory: “the argument from construction” appeals to evidence that the contents of episodic memory are constructed rather than stored; “the argument from error” appeals to evidence that episodic memory is highly error-prone; and, finally, “the argument from animals” appeals to evidence that some nonhuman animals can episodically remember. We argue that all three fail.
According to “orthodox” representationalism, perceptual states possess constitutive veridicality (truth, accuracy, or satisfaction) conditions. Typically, philosophers who deny orthodox representationalism endorse some variety of anti-representationalism. But we argue that these haven’t always been, and needn’t continue to be, the only options. Philosophers including Descartes, Malebranche and Helmholtz, appear to have rejected orthodox representationalism while nonetheless endorsing perceptual representations of a fundamentally practical kind not captured by orthodox representationalism. Moreover, we argue that the perceptual science called on by contemporary philosophers to defend orthodox representationalism instead motivates a return to this older view. Finally, we suggest that contemporary philosophers may conceptualize fundamentally practical perceptual representations as “de agendo” representations, a species of representation that has constitutive appropriateness rather than veridicality conditions.
Public discourse implicitly appeals to what we call the “Traumatic Untrustworthiness Argument” (TUA). To motivate, articulate, and assess the TUA, we appeal to Hawley’s (2019) commitment account of trust and trustworthiness. On Hawley’s account, being trustworthy consists in the successful avoidance of unfulfilled commitments and involves three components: the actual avoidance of unfulfilled commitments, sincerity in one’s taking on elective commitments, and competence in fulfilling commitments one has incurred. In contexts of testimony, what’s at issue is the speaker’s competence and sincere intention to speak truthfully. The TUA targets trauma victims’ competence rather than their sincerity. According to the TUA, empirical evidence shows that trauma undermines victims’ trustworthiness with regard to speaking truthfully about their trauma by undermining their competence to remember the event. We argue that what the evidence shows is rather that remembering traumatic events involves a distinct “mode of manifesting” the competence to remember particular events from the personal past. Trauma victims are competent to speak truthfully about their trauma and ought to be trusted at least with regard to the central details of the event. By suggesting otherwise, the TUA threatens an insidious form of epistemic injustice which Hawley’s account helps us locate.
Prima facie, accounts of scientific representation should illuminate how models support justified surrogative reasoning while remaining neutral on the nature of inductive inference. We argue that doing both at once is harder than it first appears. Accounts like “DEKI,” which distinguish justified and unjustified surrogative inferences by appealing to a distinction between derivational and factual correctness, cannot accommodate non-formal, non-rule-based accounts of inference such as John Norton’s material theory of induction. In contrast, a recent expressivist-inferentialist account appears compatible with material inference, but at the cost of abandoning inductive neutrality.
There’s reason to think that doubts about the veracity of witness testimony are at least sometimes underwritten by sexist and racist (perhaps implicit) biases. When this is the case, distrusting victim testimony constitutes a distinct form of testimonial epistemic injustice. However, the extent to which epistemic injustice is in play depends on the extent to which doubts about the veracity of witness testimony might be based on legitimate (unbiased) evidential reasoning. In this paper, we consider what we call the "Traumatic Untrustworthiness Argument" as a potentially legitimate basis for doubting victim testimony. The argument is as follows: (P1) When people are particularly liable to misremember events, we should be skeptical about their reports of those events. (P2) People are particularly liable to misremember traumatic events. (C) So, we should be skeptical about their reports of traumatic events. For our purposes, the key premise in the TUA is P2. According to P2, trauma negatively impacts the veracity of episodic memories. We argue that evidence from the psychology and neuroscience of memory speaks against P2. Indeed, the evidence suggests that, in certain respects, traumatic memories may actually be more vivid and accurate than non traumatic memories. Rather than revealing traumatic memories to be untrustworthy, we suggest that the psychology and neuroscience of traumatic and non-traumatic memory highlight the need for particular forms of care in the way traumatic cases are investigated and tried.
In this paper, we ask the question: what (if anything) is the use of thinking about phenomenology in the context of AI, and in particular machine learning? We will isolate one sense of “phenomenology”, namely the sense in which it is commonly understood within analytic philosophy of perception. Then, we will give examples of projects within sensory substitution and restoration science that rely heavily on machine learning and to which, according to us, phenomenology in the sense specified makes a relevant contribution. Finally, we will shed some light on what this contribution looks like and why it is important.
Intellectualists hold that knowledge-how is a species of knowledge-that, and consequently that the knowledge involved in skill is propositional. In support of this view, the intentional action argument holds that since skills manifest in intentional action and since intentional action necessarily depends on propositional knowledge, skills necessarily depend on propositional knowledge. We challenge this argument, and suggest that instructive representations, as opposed to propositional attitudes, can better account for an agent’s reasons for action. While a propositional-causal theory of action, according to which intentional action must be causally produced “in the right way” by an agent’s proposition-involving reasons, has long held sway, we draw on Elizabeth Anscombe’s insights in sketching an alternative theory of action. In so doing, we reject the implicitly Cartesian conception of knowledge at the core of the intentional action argument, while hanging on to the idea that mental states are representations of a certain kind. Our argument provides theoretical support for anti-intellectualism by developing an account of nonpropositional, practical content.
According to orthodox representationalism, perceptual states are representational in the sense of having constitutive truth, veridicality, or accuracy conditions. Burge (2010) justifies realism about perceptual representations posited by perceptual science on the basis of their explanatory ineliminability. I clarify his argument, including the controversial role of the constancies and their relationship to representational function. I argue that the constancies don’t do the work Burge wants them to, but that his realist strategy may vindicate an unorthodox version of representationalism.
Intellectualists hold that knowledge-how is a kind of propositional knowledge, while anti-intellectualists hold that knowledge-how and propositional knowledge differ in kind. What resources or methods may philosophers legitimately and fruitfully employ to adjudicate this debate? A number of recent anti-intellectualists (e.g. Adams 2009; Brown 2013; Devitt 2011; Glick 2011; Noe 2005; Wallis 2008) have appealed to findings in the cognitive neuroscience of memory, and in particular, the dissociation between “procedural” and “declarative” memory. However, according to Stanley and Krakauer (2013), both cognitive neuroscientists and anti-intellectualist philosophers have been guilty of “misunderstanding” scientific findings due to “two overly hasty and ultimately incorrect identifications between categories in neuroscience and mental kinds.” Specifically, they (I) misidentify motor skill with “procedural knowledge,” and (II) misidentify propositional knowledge or knowledge of facts with “declarative knowledge.” I clarify two arguments S&K appear to make in support of their claim and argue that neither succeeds.
The goal of this paper is to illuminate the connections between agency, perception, subjectivity, space and the body. Such connections have been the subject matter of much philosophical work. For example, the importance of the body and bodily action on perception is a growth area in philosophy of mind. Nevertheless, there are some key relations that, as will become clear, have not been adequately explored. We start by examining the relation between embodiment and agency, especially the dependence of agency on perception and the dependence of perception on agency. We also consider the nature of subjectivity itself: In virtue of what do humans and animals but not rocks and pencils have genuine perceptual and agentive intentional contents? We sketch a hylomorphic account of subjects and subjectivity, which highlights connections between the conclusions argued for in the previous sections and some basic principles of teleosemantics.
Most representationalist accounts of visual perception hold that visual representations are belief- or judgement-like sensory states with veridicality or accuracy conditions. In a word: visually perceiving is seeing something as being some way. I propose an alternative account on which visual representations are intention-like and have “appropriateness” rather than veridicality or accuracy conditions. In a word: visually perceiving is seeing what to do. According to this view, visual representations are a species of “embodied instructive representation.” After sketching the general account of embodied instructive representations and saying a bit about why they deserve to be called representations, I’ll argue that visual perceptual representations may be fruitfully understood as embodied instructive representations.
2025. “A Defense of the Epistemic Nature of Episodic Memory” with Seth Goldwasser for the Epistemic Injustice in Healthcare blog: https://epistemicinjusticeinhealthcareproject.blogspot.com/2025/10/a-defence-of-epistemic-nature-of.html
2025. Comments on Ruth Millikan’s Post 3 for the BrainsBlog: https://philosophyofbrains.com/
Title: “Intentionality: A Problem-Solving Approach”
Institution: University of Pittsburgh (Philosophy)
Committee: Rick Grush (UCSD Philosophy), James Shaw (Pitt Philosophy), John McDowell (Pitt Philosophy), Robert Brandom (Pitt Philosophy), Susanna Schellenberg (Rutgers Philosophy).
Abstract: Language, maps, pictures, and models all exhibit “intentionality” — they’re “about” something. What does it mean to be “about”something in the relevant sense? And in virtue of what is something about or intentionally directed at something else? In the wake of Grice (1957) and Searle (1980), many philosophers hold that the intentionality of public representations— language, maps, pictures, and models of the sort agents produce and employ in communication— can only be understood against the background of a more fundamental kind of intentionality, namely, the intentionality of mental states like beliefs, desires, intentions. Accordingly, the intentionality of mental states is said to be “original,” while the intentionality of public representations is said to be “derivative.” So, by the lights of many philosophers, illuminating intentionality means illuminating original intentionality.
Alas, original intentionality has proven an elusive creature. Attempts to analyze it seem to either deflate psychological intentionality to the point of changing the subject, or else tend to go in a circle or generate a regress by appealing to “interpreters” or other notions that seem to presuppose psychological representations. As a result, some (most notably Burge 2010) have opted to treat original representations as an unanalyzable primitive. I consider this an unhappy state of affairs, and in my dissertation, I diagnose two widespread assumptions as the underlying cause. The first assumption is that original representations must be like derived representations in that they must have constitutive truth, veridicality, accuracy, or satisfaction conditions. The second assumption is that the intentionality of action will have to be explained in terms of the intentionality of mental states.
My dissertation aims not merely to critique these assumptions but to articulate an alternative conception of the nature of and relationships between action and aboutness. To this end, I develop and defend a “practical epistemic access” analysis of representation and a “problem-solving” account of intentional action. On the view I propose, the directedness of actions does not derive from the aboutness of mental representations. Rather, the aboutness of representations is fundamentally rooted in the intentional directedness of actions.
The problem-solving account of intentional action is a kind of teleological alternative to causal theories of intentional action. Causal theories of action define intentional action in terms of the way they are caused or causally guided and controlled by intentional states. I argue that causal theories ultimately fail because they cannot account for the means-end form of intentional actions. In contrast, teleological theories define intentional actions in terms of their means-end form-- they deny that this form could causally derive from intentional states . However, most teleological accounts analyze the teleological form of action in terms of teleological explanation, and I argue that this is not entirely satisfying. The problem-solving account analyzes the means-end form of intentional action in terms of the notion of an “application-accountable solution to a practical problem.” On this view, an intentional action is a process that functions to non-accidentally satisfy a need for flourishing. Needs for flourishing define “practical problems,” and intentional actions are the processes that function to solve them.
On the practical epistemic access analysis, representations constitutively function to position agents to appropriately intentionally respond to facts. So representations are a kind of ability or capacity for intentional actions. Original representations are “embodied instructive representations” — abilities to intentionally respond to facts that do not depend on any further such abilities. They are themselves potentiations of intentional actions that are directed at facts, and in this way they provide agents with direct practical epistemic access to facts. They succeed iff they are potentiations of appropriate actions. Hence, on my view, original representations have constitutive appropriateness conditions rather than truth, veridicality, accuracy, or satisfaction conditions. The latter may be associated with embodied instructive representational success conditions, but they are not constitutive of them. Derived representations correspond to a “surrogative” species of representation that functions to provide an indirect form of practical epistemic access. Since surrogative representations fulfill their representational function by virtue of and to the extent that they have certain properties qua stand-ins, truth, veridicality, accuracy, and satisfaction conditions may be understood as a constitutive aspect of appropriateness conditions for surrogative representations.
I argue that embodied instructive representations deal elegantly with a variety of psychological phenomena that are not well-modeled by propositional attitudes. What’s more, they let us have our cake and eat it too because familiar propositional attitudes like belief can be subsumed by embodied instructive representations that relate agents to surrogative representations. Overall, I argue that my practically-oriented and unified accounts of intentional actions and representations offer a theoretically powerful alternative to the standard approach to (and its problematic assumptions about) intentional phenomena.
According to Setiya (2011), a desideratum on a theory of reasons for action is that it explain the necessary truth of conditionals like the following:
(1) If A is doing φ on the ground that p, A is doing φ because she believes that p.
Setiya argues that only the causal-psychological theory (CPT) of acting for reasons can meet this desideratum. According to CPT, reasons are contents of psychological attitudes like belief and desire, and one acts because or on the grounds of a reason p iff psychological attitudes represent p cause one’s action non-deviantly. However, Setiya also acknowledges one “formidable” anti-causalist contender: a “revamped” behaviorist position he constructs out of Hyman’s (1999) theory of knowledge. Setiya’s behaviorist holds:
(K) To know that p is to be capable of doing things becauseR p.
From which Setiya derives the following rendering of (1):
(1k) If A is doing φ becauseR p, A is doing φ because she knows that p
Given that what one does, one has the capacity to do, and (K), it follows from the antecedent of this conditional that A knows that p. Given that what one does, one has and has exercised the capacity to do, if A is doing φ becauseR p, she is doing φ in exercise of a capacity that constitutes her knowing that p. So the behaviorist can explain the conditional so long as there is a sense or use of “because” that’s tied to the exercise of capacities. Still, they can’t quite satisfy the desideratum. After clarifying Setiya’s objection to the behaviorist account, I show how the basic behaviorist insight can be developed to meet it via my “Intentional Action Acorn” (IAA) account of epistemic subject reasons or “takings.”
Borrowing from Williamson’s (2000) criticism of Hyman’s theory of knowledge, Setiya objects that (K) is inconsistent with the truth of explanations like: “He can’t go outside becauseR it is raining, since he doesn’t know about the rain. Nor can he stay in becauseR it is raining, for the very same reason.” The reason is that “For the advocate of (K), these are two separate incapacities, no more explained by lack of knowledge than an accidental generalization explains its instances.” (151) Setiya’s objection to the behaviorist account of acting for reasons is that it’s likewise inconsistent with the truth of such explanations. Such explanations presuppose that if one acts for a reason that p, one’s action manifests some way one is taking things to be, and p corresponds to that taking. But in treating acting for reasons as a primitive, the behaviorist offers no account of such takings. And it’s not clear how they could so long as they identify beliefs with abilities to act for reasons and exercises of such abilities with action performances. Because the CPT treats takings as psychological representations, it doesn’t have this problem. A further issue is that such explanations suggest that takings map to the abilities whose exercises manifest them one-to-many. Hence, a single taking can explain multiple incapacities. But if the behaviorist were to somehow ground takings in abilities for actions, takings will map to abilities one-to-one. This issue is less pressing, I argue, because the behaviorist can explain the one-to-many intuition by appealing to the single event that would occasion many discrete takings.
Enter the IAA account of takings. Following Thompson (2007) and Stout (1996), the IAA treats psychological states as aspects of intentional actions understood as developing processes. Then, using an acorn metaphor, the IAA develops McDowell’s (2011) idea that “An intention for the future stands to the acting one engages in when one starts to execute it [...] as a caterpillar stands to the butterfly it becomes in metamorphosis.” (16) On the IAA, psychological representations are potentials for and the first phase in the life of an action; they stand to fully performed actions as acorns stand to fully developed oak trees. The IAA has the behaviorist’s abilities double as demonstrative concepts that directly (causally) apply to facts. The resulting acorns are non-syllogistic practical judgements that constitute an agent’s practical grasp of those facts. Acorns--i.e., potentiated actions-- are correct representations (takings) when they “practically fit” (are appropriate to) the facts. Appropriate acorns that mature into performances are agents’ pro-tanto justifying reasons for acting. The facts that would have made a matured misrepresenting acorn appropriate are an agent’s pro-tanto excusing reasons. Finally, intentional action acorns compete for performance, and the winners constitute agents’ implicit “all things considered” judgments.
*Draft available upon request.