PHIL 5400 (Spring 2022)

Charlie Kurth

Email: charles.kurth [at] wmich [dot] edu

Office Hours: Wednesdays, 2:00-3:00 pm, and by appointment

Office Location: Moore 3010

Note: While our sessions are planned to be in-person, responding responsibly to the pandemic may require revisions to that plan.

Course Overview

Efforts to make sense of mental content and mental representation generate some of the most central and puzzling topics in the philosophies of mind and cognitive science. Consider: when you believe that your car is in the garage or think that the wet sidewalk indicates that it rained last night, you take your mental states to be have content—to be about your car or the weather. But how do your thoughts come to have the content that they do? A leading proposal maintains that our thoughts have content because they are mental representations. Yet to say that invites a host of challenging philosophical questions. For instance,

• What is a mental representation? If it is just something produced by a biological system like a brain, then how does it come to have meaning? After all, notice that not just any firing of neurons is the basis of a meaningful representation. So what makes the neural activity that underlies your belief about where your car is special—that is, something with meaning?

• What are concepts and what role to they play in mental representation? Must an organism have concepts in order to have representations or thoughts? Are there different types of concepts? If so, how does having a particular kind of concept affect one’s cognitive abilities or status as a thinker?

• How do we represent normative and evaluative content like my thoughts that I should not tell lies, that chocolate is good, or that Darth Vader is evil? Is the story we tell here the same as what we say about descriptive content like your thoughts about where your car is? Or do we need to introduce a new type of representation or additional kinds of concepts?

In this seminar, we will look at one of the most recent and prominent accounts of mental representation: that developed in Nicholas Shea’s Representation in Cognitive Science (2018). To help draw out some of the philosophical significance of Shea’s proposal—and to better position ourselves to critically assess it—we will begin by looking at a core set of related issues in the philosophy of cognitive science. We will also examine an important set debates about the nature of concepts and their role in structuring cognition, including normative and evaluative thought.

Texts

Nicholas Shea, Representation in Cognitive Science (e-book here)

Joulia Smortchkova, Krzysztof Dotega, & Tobias Schlicht, What are Mental Representations? (e-book here)

Selected readings available on the course web site

Assignments and Grading

This course will have three graded components.

1) Class participation (15%). This course is structured as a seminar. So you will be expected to come to class prepared to discuss the readings. To give you some guidance, I will provide weekly reading questions. While you are not required to write up answers to these questions, you should come to class prepared to talk about them.

Additionally, after each class session you will need post a question about that week’s discussion on the course Google Doc (available here). This is an opportunity for you to ask about, say, a lingering issue or new puzzlement that wasn't cleared up in our discussion. Your question must be posted by noon on the Thursday immediately after that week’s class session (so the first one is due on Jan 13), and you should start your question with your name so that we know who is asking what. I will answer your questions after they’ve all been posted.

2) Reading responses (40%). Throughout the course, you will write four brief critical response essays. More specifically, you need to write a response for each of the four parts of the course (the readings for Jan 12 are not eligible for reading responses); and the response paper you write should be for the readings we're looking at that week (e.g., a response paper done in Week 6 should be based on the readings for Week 6, not a previous week's readings). This constraint aside, you are free to choose which reading you write about, and you can use the reading questions as guide for your response papers. These essays should be approximately 500-600 words long (standard formatting).

In your response paper, you should focus on just one of that week's assigned readings. Having picked a reading, you should then do two things. First, you should summarize one argument from the reading you selected (not the entire paper) by putting it into standard form. That is, explicitly lay out your understanding of the argument's premises and conclusions. Second, you should raise an objection to the argument you've presented. As a rough guide, presenting the standard form version of the argument should 30-40% of your discussion with the balance left for your objection.

The essays are to be emailed to me by noon on the Tuesday before the relevant Wednesday class meeting. For graduate students, all four papers will count toward your final grade; for undergraduates, I will use only the top three reading responses.

3) Long paper (45%). The longer paper assignment invites you to explore one of the topics that we will be discussing in the course. For graduate students, your paper should be approximately 15 pages long (double spaced, standard formatting). For undergraduates, your paper should be approximately 12 pages long (double spaced, standard formatting).

Though you are free to choose the topic for your long paper, you must get it approved by me before you start working on your draft. I strongly recommend that you begin thinking about your paper topic early and you must have a proposal signed off on by April 1.

You will need to turn in a draft of your paper on Sunday, April 24. I will provide you with feedback by Thursday, April 28 (hopefully earlier). The final version of the paper will be due on Saturday, April 30. You should email both your draft and your final papers to me.

While your draft needn't be a highly polished piece, the more developed it is, the better the comments I will be able to give you. Turning in a cursory draft may result in a grade penalty.

Grading: Final grades will be calculated based on the percentages noted above (using the average for the reading responses). Here is a tentative guide for the grading scale: A=100-92, BA=91-88, B=87-83....

Policies

In order to help ensure a successful class, please heed the following rules and policies:

• Due Dates. Baring unusual circumstances, the due dates on the syllabus are non-negotiable. If you think you have reason to miss an assignment, it is best to inform me well in advance.

• Classroom Environment. Please arrive to class on time. All cell phones must be turned off during class. Texting, web surfing and the like is not permitted. Abuse of these courtesies may lead to penalties.

• Academic Honesty. As a student at WMU, you are responsible for making yourself aware of the University policies and procedures that pertain to Academic Honesty. These policies include cheating, fabrication, falsification and forgery, multiple submission, plagiarism, complicity, and computer misuse. In this class, you will be expected to abide by these obligations. This means that all work presented as original must, in fact, be original; the ideas and contributions of others (be they quotes, summaries, or paraphrases) must be appropriately acknowledged. More information about the WMU Academic Honesty rules as well as the rights of accused students can be found here.

Resources

• Accommodations for Disabilities. I am happy to make accommodations to assist students with documented disabilities (e.g., physical, learning, psychiatric, vision, hearing, etc.). Those wishing to arrange reasonable accommodations must contact Disability Services for Students. A disability determination must be made by this office before any accommodations are provided by the instructor. More information can be found here.

• Mental Health. WMU’s Mental Health Services’ professional staff members work with students to resolve personal and interpersonal difficulties, many of which can affect the academic experience. These include conflicts with or worry about friends or family, concerns about eating or drinking patterns, and feelings of anxiety and depression.

Tentative Schedule of Readings and Assignments

Part 1. Mental Content and the Representational Theory of Mind

Jan 12. Introduction: Philosophy of Cognitive Science and Mental Representation

• Read syllabus

• Rick Grush, “Cognitive Science”

• Joulia Smortchkova et al., What Are Mental Representations? (WAMR), pp. 1-14 only

• Nicholas Shea, Representation in Cognitive Science (RCS), Chap 1.1, 1.2

• Fred Dretske, “Misrepresentation”

  • Qs: What is the problem of mental content as Shea explains it? What conditions does he think an adequate answer must meet? What is a mental representation and how is it accounted for by CRTM, connectionist and anti-represenationalist accounts (Smortchkova et al.)? Why is explaining misrepresentation challenging (Shea, Dretske)? According to Dretske, can appealing to functions help make sense of misrepresentation (think about the bacteria example)? Why does complexity--esp, having multiple routes of detecting--matter for a system's capacity to misrepresent?

• Additional background reading

  • SEP “Mental Representation” (esp. sections 1, 2, 6, 8, 9)

Jan 19. Teleosemantics

• Ruth Millikan, “Biosemantics”

• Jerry Fodor, “A Theory of Content I” (focus on pp. 69-79, though pp. 59-61 will likely be helpful and the whole thing is good because Fodor is very entertaining)

  • Qs: Explain Millikan’s distinction between production and consumption (p 283ff). How does this distinction inform her biosemantic account of mental content (p 286ff)? What does misrepresentation amount to on Millikan’s account? What's the "disjunction problem" and why is it a problem (Fodor, Sec 2)? Why does Fodor think teleological solutions to the disjunction problem like Millikan's are inadequate (see the extended frog example)?

Jan 26. Maps and Cartographic Representation

• Elizabeth Camp, “Thinking with Maps”

• Michael Rescorla, “Predication and Cartographic Representation”

• Ben Blumson, “Mental Maps” [Recommended. Contra Camp and Rescorla, Blumson argues that cartographic and linguistic representation aren’t substantively different]

  • Qs: Why do so many philosophers believe thought must be language-like? What is the difference between weak- and strong-LOT? What’s isomorphism? How are maps different from photos or Venn diagrams? In what ways does Camp think that maps and languages involve different combinatorial principles? Do you agree? In light of the Camp, Rescorla, and Blumson, how much of language-based thought could (in principle) be done in a map-based format?

Feb 2. Explanation

• Andy Clark & Josefa Torbio, “Doing without Representing?” (Sec 5-6 can be skimmed)

• Farid Zahnoun, “On Representation Hungry Cognition (and Why We Should Stop Feeding It)”

  • Qs.What is the distinction between the classicist’s explicit notion of representation and the connectionist’s distributed notion (e.g., what do they have in common, where do they disagree)? What is a mobot and why might it suggest that internal representation isn’t necessary for complex behavior (hint: what’s “subsumption architecture”)? According Beers, when are we allowed to take (i) a theoretical description of a system’s behavior that relies on representations as (ii) an accurate account of how the system actually operates? Why are Clark and Toribio unhappy with Beers’ proposal (hint: what’s the distinction between representationality and symbols)? What is a Centrifugal Governor and why might it suggest complex cognition needn’t be computational? What is the idea of a “representation-hungry” domain and how does this show the limits of the Centrifugal Governor example? Explain the difference between online and offline cognition (with examples). Why is offline cognition supposed to be special? Which of the three issues that Zahnoun raises to Clark & Tornio do you take to be most significant? Why?

Part 2. Shea’s Varitel Semantics

Feb 9. Foundations and Functions

• Shea, RCS, Chap 2 (Sections 2.3, 2.4, 2.6 can be skimmed), Chap 3 (3.7 can be skipped)

• Bence Nanay, “A Modal Theory of Function” (Sections VII-IX can be skimmed)

  • Qs: According to Shea, when do we have reason to posit a representation in our explanation of a given phenomenon (see the rifle and bacteria examples in 2.2)? What is a representational vehicle? What is a task function and how does it differ from the notions of function used by Dretske and Millikan? What are clustering, robustness, stabilizing, and exploitable relations? Why are these ideas important for understanding task functions? What is the point of the example in 3.6a? How does it work? What is Nanay’s concern with standard accounts of biological function (hint: in what sense does he think they are circular)? Is Nanay’s worry a problem for Shea—explain?

Feb 16. Exploitable Relations

• Shea, RCS, Chap 4.1-4.6, Chap 5 (focus on Sec 5.1-5.4)

  • Qs: Ch4. What is exploitable correlational information and why is it relevant to Shea’s account? How is this relevance illustrated by the toy example of 4.1b? What is an unmediated explanation, why is it important, and how is it illustrated in the frog example (4.2)? What is the evidential test introduced in 4.2 and how is it illustrated in the example of 4.3? How does Shea’s account differ from Millikan’s consumer-based proposal? Consider the examples of 4.6. In the numerosity example, what are the two possible explanations and what does Shea take his account to help us understand? Similarly, what does Shea take the macaque example to show? Ch5. What are place cells and what function do they appear to have in rats (e.g., what’s the significance of offline activity in these cells)? What is exploitable structural correspondence and why is it significant for Shea’s proposal? How does the idea of exploitable structural correspondence apply to the rat case and why does it allow us to take activation of place cells as the vehicles for (something like) cartographic representations (hint: robustness and stabilization play a role)?

Feb 23. Determinate content?

• Francis Egan (2019), "Content is pragmatic"

• Francis Egan (2020), WAMR, “A Deflationary Account of Mental Representation” (focus on pp 26-35)

• Shea, RCS, Chap 8.1-8.2c, 8.5a

  • Qs. (2019) What does Shea take the content of the frog’s representation to be? Why does Egan think this fails to be a naturalistic account—that is, in what sense is Shea’s account pragmatic and why is that a problem? Why does Shea think his account identifies a necessary explanatory role for representations? Why does Egan disagree? (2020) What’s a “tracking” theory and why does Egan think they fail? What is Egan’s “pragmatic” account? For instance, what does she mean in saying, “explanatory focus resolves indeterminacy” (p 33)? And what is the difference between “cognitive content” and “computational theory proper”? On her account, what’s the point of positing representational content? Why does Egan think her account is naturalistic? Do you agree? In what sense is her account “deflationary” (see p 41)?

Mar 2. Real Explanatory Work?

• Dan Burnston, “Contents, Vehicles, and Complex Data Analysis in Neuroscience”

  • Qs: What are the “injective mapping” and “causal isomorphism” claims that algorithmic homuncularism (AH) is committed to? How do these claims relate to Shea’s account of “vehicle realism”? What does Burnston take the problematic assumption in Shea’s account to be? Why is looking to the data analytic methods of cognitive scientists important for Burnson’s objection—that is, what does data analytics suggest about how cognitive processing works in the brain and why is this a problem for Shea’s account (see, eg, p 1626)? How does all of this play out in Burnston’s alternative explanation of the monkey “colored dots” experiment? What is Burnston’s alternative account of representation?

Mar 9. Spring break, no class

Mar 16. Feminist and Anti-Representationalist Alternatives

• Naomi Scheman, “Individualism and the Objects of Psychology”

• Daniel Hutto & Glenda Satne, “The Natural Origins of Content”

  • Qs. NS: What is the “individualist assumption” that Scheman is challenging? Is this assumption something that, say, Shea is committed to (e.g, how does the assumption relate to question about indeterminacy)? In what sense is the individualist assumption a piece of ideology and why does Scheman think that this is problematic? What, for instance, does Scheman mean in saying that “questions of meaning and interpretation [of psychological states] cannot be answered in abstraction from a social setting” (229)? Similarly, what does she mean in saying the “individualist assumption stem[s] from the ideology of liberal individualism” (230)? What does Scheman’s alternative, social constructivist account look like? H&S: Explain the neo- Cartesian, Behaviorist, and Pragmatist positions. What is the distinction between (i) a primitive, contentless forms of intentionality and (ii) a content-based form of intentionality? Similarly, what’s the distinction between naturalizing content and providing an explanation of the natural origins of content? What is Ur-intentionality and why is recognizing it important to understanding what the “infield” should do? How does all this come together—that is, what’s the proposal of pp. 534-5?

Part 3. Concepts

Mar 23. What are Concepts?

• SEP: Concepts (focus on Sec 1.1, 2 & 4)

• Edouard Machery, “Precis of Doing without Concepts”

• Collin Rice, “Concepts as Pluralistic Hybrids” (focus on Sec 1-4)

  • Qs. What does it mean to say that concepts are things (e.g., mental states) with “internal structure”? What are some of the competing ways that the internal structure of concepts has been understood by philosophers and cognitive scientists? What is the “received view” of concepts (Machery, Sec 3)? What is the “heterogeneity view” that Machery endorses? What are the three types of concepts Machary focuses on and how do they support his heterogeneity view? What is “concept eliminativism” and when is it appropriate to be an eliminativist? Explain Rice’s two objections to Machery. What does it mean to say that concepts are pluralistic hybrids?

Mar 30. What Role Do Concepts Play in Cognition?

• Elizabeth Camp, “Putting Thoughts to Work: Concepts, Systematicity, and Stimulus-Independence”

• Peter Carruthers, “On Being Simple Minded” [Recommended, esp Sec 1, 4-6]

  • Qs. According to Camp, in what ways are structural and systematical aspects central to conceptual thought? What is “minimalism” and why might we think it’s indicative of having concepts? What is “intellectualism,” how does it go beyond minimalism, and why might these differences be important? What is “stimulus independence” and what difference (if any) is there between it and what Clark is talking about with his notion of “representation hungry” problems? Camp and Carruthers disagree about what meeting the generality constraint involves and what kinds of critters (bees, chimps, humans, etc) have the capacity to meet the constraint. What’s at issue in this debate and whose arguments do you find more compelling?

Apr 6. Distinctly Practical and Characterological Concepts?

• Carlotta Pavese, “Practical Concepts” (Chap 18 in Routledge Handbook of Philosophy of Skill and Expertise)

• Evan Westra, “Character and Theory of Mind: An Integrative Approach”

• Inês Hipólito et al., “Embodied Skillful Performance: Where the Action Is” [Recommended: presents an alternative to the Pavese proposal]

  • Qs. CP: What does it mean to say we can know a fact through different modes of presentation, and how might we distinguish between different modes of presentation? Give some examples. What do you take the key point(s) of the Casio example to be? Do you agree with it? What does Pavese take showing the explanatory need for (practical) representation to involve? Is it different from Shea’s proposal? Does that matter? What is Ideomotor Apraxia and why is it important for Pavese’s argument? How does Pavase’s appeal to motor schemas help her make an argument practical representation is conceptual?

EW: What is mindreading and what role is it thought to play in human cognition? Why does Westra think that standard accounts of mind reading are inadequate? What is character-trait attribution and how does it differ from attribution of beliefs and desires? What is the point of the discussion of empirical findings in Sec 2? What limitations does Westra see in recent “pluralist” accounts of mind reading? What are hierarchical predictive coding models of the mind and how do they help us understand the role that character traits might play in cognition? Do questions about whether there are character traits, or whether character-trait attribution is correct, matter for Westra’s account?

Part 4. Evaluative and Deontic Content

Apr 13. Moral Rules: Innate or Value-based?

• Gilbert Harman, “Moral Philosophy and Linguistics”

• Peter Railton, “Moral Learning: Conceptual Foundations and Normative Relevance” [sec. 5.5-6.0 can be skimmed]

• Fiery Cushman, “Action, Outcome, and Value: A Dual-System Framework for Morality” [Recommended: companion to Railton from the perspective of a psychologist]

  • Qs. GH: What is moral nativism? In what ways do societal and individual differences put pressure on the idea that morality is undergirded by a common set of (general) rules or principles? What is an idiolect? What core features of language and linguistic competence does Harman focus on? What is the “principles and parameters” theory of language? Why does it seem plausible to many? What is a “bimoral” individual? Why might bimorals seem strange or problematic (as Gendler seems to think they do)? Why are trolley problems interesting in the context of Harman’s investigation? Are moral innate? Explain your answer.

PR: How does Railton understand moral learning—eg, how does it contrast with moral development and social learning? What is distinctive of moral learning? What is ‘expectation-based action-guidance’? How is ‘expectation-based action-guidance’ related to Humean projection and what might it suggest about moral nativsim? What’s distinctive of ‘deep learning systems’? What should moral learning look like if it’s the product of domain-general, deep learning mechanisms? What evidence supports this picture? What is the distinction between empathetic distress and empathetic concern, and why does it matter for Railton’s account? What do “moral intuitions” amount to on Railton’s account and how does his account of their authority differ from what dual-process accounts say?

Apr 20. Moral Rules: A Learning Model

• Shaun Nichols, “Moral Learning and Moral Representation”

  • Qs. What are the core issues that Nichols is investigating in his paper? What’s the contrast between low-level (value-based) and structured (rule-based) accounts of mental content? What is the point of the scuba diving example (3.1)? What is Nichols’ concern with Cushman’s model-free account? How does Railton’s account differ from Cushman’s (e.g., how does it explain judgments of wrongness)? Why is Nichols nonetheless dissatisfied with Railton’s proposal? How does Nichols’ proposal capture the connection between moral judgment and motivation? Do you find it plausible? What is the rule learning challenge and how does Nichols address it? What lessons for debates between rationalists and sentimentalists does Nichols draw from his observation about statistical learning?

April 24. Drafts due

April 30. Final papers due