80-300 Minds, Machines, and Knowledge

Instructor: Kevin T. Kelly, Professor

Office: 135 K BH

Office hours: M, T 11:00-12:00

Contact: X 8567, kk3n@andrew.cmu.edu

Room: BH 150

Time: M 4:30-6:50

Text: articles available for download in pdf format from this page.

Requirements:

Undergraduates: short reading questions and two short (5 pgs) essays related to the course.

Graduates: reading outlines and two longer (9 pgs) essays related to the course.

Grading standard:

I am looking for concise, incisive prose. Therefore, length limitations will be taken very seriously. Write a paper of double length and cut it down to size. Also, the reading questions are intended to give you credit for the time you spent on timely class preparation so that the class discussions are smooth and rewarding for all. Therefore, late submissions, for any reason, will be penalized by 20%.

Description: This course investigates some basic relations between epistemology, philosophy of mind, and the concept of computation. In Plato and Descartes, a dualistic conception of mind is motivated by a quest for mathematical certainty. Behaviorists were motivated by epistemological ideas drawn from logical positivism. The concept of computation was developed, in part, to ground the consistency of mathematical method but also came to provide a new, “cognitive” conception of mind as the operation of a high-level computer programming running in the “wet-ware” of the brain, which led to the development of artificial intelligence programs alleged to think the way humans do. Finally, philosophical theories of rationality are heavily informed by behaviorism and are also interestingly constrained by computability.

Rationalism and Mental Substance

Plato’s Phaedo excerpts. Full text.

Aristotle’s On the Soul, excerpts from book ii. Full text.

Discussion of Aristotle’s view

Descartes’ Principles excerpts. Full text.

Reading questions (answers due Sept. 8):

How does Plato argue that the mind is distinct from the body?
How does Descartes argue that the mind is distinct from the body?
Where does Plato come closest to discussing Aristotle’s view?
Were you surprised by Aristotle’s reference to the pilot of a ship? Explain.

Empiricism, Pragmatism, and Positivism: Psychological Philosophies

D. Kalupahana, “A Buddhist Tract on Empiricism”.

J. Locke, “An Essay Concerning Human Understanding”, Book II Chapter I, 1-5, 24-25, chapter XII 1-8, XXIII 1-5.

D. Hume “Enquiry Concerning Human Understanding”, Sections I-VI and Section XIII, Part III.

C. S. Peirce, “How to Make our Ideas Clear”

A. J. Ayer “The Elimination of Metaphysics”

Reading questions (answers due Sept. 15):

The Buddha is famous for the thesis that there is “no self”, so there is no coherent basis for selfishness. It seems also that the Buddha agreed with the basic outline of Empiricism. How might John Locke argue that there is “no self”?
Hume praises skeptics, but how is empiricism a general solution to skeptical doubts?
How does Hume assign an empiricist sense to causation? Do you agree?
Peirce’s “pragmatism” is evidently quite similar to the empiricism of Locke and Hume. How is it different?

Behaviorism: An Empiricist Psychology

Gilbert Ryle, The Concept of Mind, chapters 1 and 2.

R. Carnap, “Psychology in Physical Language” excerpt.

Frank Ramsey, “Truth and Probability”, up to section 4.

John B. Watson, Behaviorism, chapter 1.

Reading questions (answers due Sept. 22):

Re-run Descartes’ argument for the real distinction between mind and body on one of Ryle’s “category mistake” examples.
Compare Carnap’s argument for behaviorism with Locke’s critique of substance in general.
How is Watson’s behaviorism is stricter than what empiricism would require?
Ramsey is considered to be the inventor of the subjective Bayesian approach to statistics. Compare Ramsey’s definition of belief with Hume’s.

Indeterminacy and Holism: Quine’s Behaviorist Philosophy of Language

R. Carnap, “Meaning Postulates”, Meaning and Necessity, 1947.

W.V.O. Quine, “Two Dogmas of Empiricism”, Philosophical Review 1951.

W.V.O. Quine, Word and Object, chapter 2.

Reading Questions (answers due Sept. 29):

For Carnap, what distinguishes a meaning postulate from a statement that is not a meaning postulate? What makes a statement analytic?
What is Quine’s indeterminacy thesis and how does it depend upon behaviorism? (Refer to your answer to the previous question).
Why is identical stimulus meaning across speakers a defective account of synonymy in the respective speaker’s languages?
Outline the versions of the analyticity concept considered by Quine, along with his response to each.

The Rise of Scientific Realism and the Mind-Brain Identity Theory

Epicurus’ Letter to Herodotus.

Hobbes’ Leviathan, Part One, Chapter I.

Herbert Feigl, “Logical Reconstruction, Realism and Pure Semiotic”, Philosophy of Science, 17:1950. start at p. 194, first full paragraph beginning: “At the risk of provoking intense controversy…” (!)

Hilary Putnam, “What is Mathematical Truth”, in Mathematics, Matter, and Method, 1979. Read the short selection indicated between arrows.

U. T. Place, “Is consciousness a brain process?”, in The Mind-Brain Identity Theory, C. V. Borst, ed., 1970.

J. J. C. Smart, “Sensations and brain processes”, in The Mind-Brain Identity Theory, C. V. Borst, ed., 1970.

Reading Questions (answers due Oct. 6):

For Epicurus and for Hobbes, what is a mental state (e.g., a sensation)? Relate this idea to Aristotle’s theory of soul and give an obvious Cartesian objection.
What is realism and how do Feigl and Putnam argue in favor of it over empiricism?
Contrast Quine’s analysis of examples like Everest and the Morning star with those of Place and Smart.
How would Smart respond to Descartes’ argument for the real distinction between mind and body?

Make sure you remember the arguments for realism and Smart’s objections and replies.

Functionalism, Cognitivism, and Strong AI

H. Putnam, originally in Art, Mind, and Religion, W. H. Capitan and D.D. Merrill, Pittsburgh: University of Pittsburgh Press, 1967.

A. M. Turing, “Computing Machinery and Intelligence”, in Mind 59: 433-460.

A. Newell and H. A. Simon, “Computer Science as Empirical Inquiry: Symbols and Search”, in Mind Design, J. Haugeland ed., Cambridge: MIT Press, 1996.

David Marr, Vision, San Francisco: Morgan Freeman 1980.

Supplementary text on Turing machines:

M. Sipser, Introduction to the Theory of Computation, Boston: P.W.S. Publishing Company, 1997.

Reading Questions (answers due Oct. 13):

For Putnam, what is a mental state? Why is the mind-brain identity thesis unlikely?
When Newell and Simon define designation, they have in mind the value assigned to a variable in a computer program. If X := a then the program can access a to alter it to f(a) or can enter a conditional branch “If X = a then do blah”. Do you think their definition of designation works for natural language (think about Quine and collateral information).
Do Putnam and Turing agree about the nature of the mental? Explain.
Compare Putnam’s view of mental states with Marr’s three levels. In figure 1-4, is there a question Marr fails to ask that Putnam would take to be crucial?

Mini-Symposium

Mid-term papers due Oct 20.

Undergraduates: 4 pages + footnotes and references.

Graduates: 5 pages + footnotes and references.

Critically compare and contrast at least two topics covered in class.

Examine at least 4 outside sources on the topic (use Google and Jstor to help).

Emphasize clarity, accuracy, and succinctness over originality on this paper.

Searle’s Chinese Room

G. W. F Leibniz, Monadology, selection.

John Searle, “Minds, Brains, and Programs”, Behavioral and Brain Sciences 23: 417-457.

Supplementary reading:

John Preston, “Introduction”, in Views Into the Chinese Room, Johyn Preston and Mark Bishop, eds, Oxford: Oxford University Press, 2002.

Reading Questions (answers due Oct. 27):

What is Searle’s Chinese room argument? Compare it to Leibniz’ mill argument. Does Searle agree with Leibniz’ conclusion?
Have either Searle or strong AI proved their respective points? Why or why not?
Do you think a Turing machine can “realize” or “instantiate” a massively parallel computation (in the sense relevant to “multiple realizability”? Relate your answer to Searle’s discussion of Weizenbaum’s construction.
What would Quine say about Searle’s argument? Would Searle care?

Qualia and the Inverted Spectrum

Sydney Shoemaker, “The Inverted Spectrum”, Journal of Philosophy 79: 357-381.

Reading Assignment (answers due Nov 3):

What is the “metaphysical” problem and what is Shoemaker’s response to it?
What is the “epistemological” problem and what is Shoemaker’s response to it?

Consciousness, Parallel Computation and Connectionism

Tim Maudlin (1989), “Computation and Consciousness”, Journal of Philosophy: 407-432.

David Rumelhart (1989), “Connectionist Modeling: Neural Computatoin/Mental Connections”, in Mind Design, J. Haugeland ed., Cambridge: M.I.T. Press.

Reading Assignment (answers due Nov 11):

1. What is the supervenience thesis and how does Maudlin strengthen it?

2. Succinctly summarize Maudlin’s argument and distinguish it from “funny instantiation” arguments.

3. What is connectionism and how might it relate to Maudlin’s objection?

4. What are the alleged advantages of connectionist models?

The Cartesian Theater

This is a fun and spooky article that provides one possible escape from Maudlin’s argument against serial AI.

Daniel Dennett (1992), “Time and the Observer”, Behavioral and Brain Sciences 15, 183-247.

What are the Orwellian and Stalinesque versions of the Cartesian Theater and what is wrong with them?
What is the Phi phenomenon and what is Goodman’s explanation?
What is the multiple drafts model and why is it more parsimonious than the Cartesian Theater or Goodman’s explanation?
Relate the multiple drafts model to Maudlin’s strong supervenience thesis. Does the multiple drafts model provide an escape route for serial processing?

Realism and Ockham’s Razor

Anti-realists like empiricists and behaviorists think that hidden states are inscrutable or meaningless because alternative internal arrangements are compatible with experience. Realists respond that, nonetheless, some arrangements are simpler and therefore better supported by the observations than others (think of Dennett’s article last time, which claimed that multiple drafts is more parsimonious than the Cartesian theater). That is an appeal to Ockham’s razor or inference to the simplest explanation. It is therefore a pivotal issue in the realism debate to explain how a fixed bias toward simplicity could help science find the truth better than some other fixed bias unless it is assumed at the outset that the truth is simple. We will look at a novel answer to that question.

Kevin Kelly (2008), “Ockham’s Razor, Truth, and Information”, in The Philosophy of Information, P. Adriaans and J. Van Benthem eds., Dordrecht: Elsevier.

What is the Bayesian explanation of Ockham’s razor?
How can fixed advice help one get to an undisclosed destination?
How does Ockham’s razor minimize retractions?
What is the proposed account of empirical simplicity?