Philosophy of Computer Science: Online Resources

Further Readings for Chapter 16:

Programs and the World

Note 1: Many of these items are online; links are given where they are known. Other items may also be online; an internet search should help you find them.

Note 2: In general, works are listed in chronological order. (This makes it easier to follow the historical development of ideas.)

§16.2: Internal vs. External Behavior: Some Examples:

• Other examples, concerning, e.g., the applicability of group theory to physics, are discussed in Frenkel 2013.

• The mathematician Paul Halmos (Halmos 1973, p. 384) points out that, once aspects of Hilbert spaces are seen to be structurally identical to aspects of quantum-mechanical systems, "the difference between a quantum physicist and a mathematical operator-theorist becomes one of language and emphasis only".

• Eugene Wigner's classic essay on "The Unreasonable Effectiveness of Mathematics in the Natural Sciences" (Wigner 1960) is also relevant.
  • Computer scientists should be sure to read R.W. Hamming's response (Hamming 1980).

• On the "indeterminacy of computation", see:
  • Fresco, N., Copeland, B.J., and Wolf, M.J. (2021). The indeterminacy of computatation. Synthese, 199:12753–12775.

  • Curtis-Trudel, A. (2021). The InDeterminacy of computation. Synthese

  • Doherty, F.T.; & Dewhurst, J. (2022), Structuralism, Indiscernibility, and Physical Computation, Synthese 200 (189).

  • Fresco, N. (2022), How Context Can Determine the Identity of Physical Computation, in S. Ioannidis et al. (eds.), Levels of Reality in Science and Philosophy (Cham, Switzerland: Springer):75–96.

§16.3: Two Views of Computation:

• Robin K. Hill's distinction between "Do A" and "In order to accomplish goal G, do A" echoes something that Herbert Simon said three quarters of a century ago. According to:
  • Brown, R. (2004). Consideration of the origin of Herbert Simon's theory of "satisficing" (1933–1947). Management Decision, 42(10):1240–1256.
  "In his exposition of science, Simon … divides it into two kinds: practical and theoretical. Scientific propositions are practical if they are stated in some such form as 'In order to produce such and such a state of affairs, such and such must be done' (Simon, H. A. (1947). Administrative Behavior. Macmillan, New York, p. 248). The equivalent theoretical proposition with the same conditions of verification can be stated in a purely descriptive form: 'Such and such a state of affairs is invariably accompanied by such and such conditions' (Simon 1947, p. 248)." (Brown 2004, p. 1247, my italics)

• Making a distinction akin to Hill's,
  • Putnam, H. (1999). The Threefold Cord: Mind, Body, and World. Columbia University Press, New York, p. 6

  says: "the world is as it is independently of the interests of describers."

  • Similarly, an algorithm (A) "is as it is independently of" any purpose (G) that it might be used for.
  • For further discussion, see Putnam 1999, pp. 45–46.

• Brachman, Ronald J.; & Levesque, Hector J. (2022), Machines Like Us: Toward AI with Common Sense (Cambridge, MA: MIT Press).
  • Suggests that there are circumstances in which "you need an understanding of your task [i.e., G] that goes beyond merely being able to carry it out [i.e., A]" (p. 214).

• Question for the reader:

  Aristotle makes a distinction between a "praxis" (a rational action) and a "kinesis" (a motion). When a dog chases a squirrel up a tree, the dog has a purpose, although its purpose is not the result of deliberative reasoning. A praxis is an action performed for a deliberate goal, what Aristotle called a "telos". In this sense, a praxis is a type of kinesis. (Thanks to William Jacobs for this information.) In:

  Cummins, Robert (1989), Meaning and Mental Representation (Cambridge, MA: MIT Press),

  Cummins suggests "that cognitive behavior is essentially behavior under a semantic interpretation" and that "only behavior tht is conceived as behavior under interpretation (like Aristotle's actions, as opposed to his motions) is naturally explained by appeal to the computation of representations." (p.153).

  Are these distinctions related to the "Do A" (motion) vs. "To G, Do A" (action) distinction?

§16.4.3: Are Inputs Needed?

• Besides the basic philosophical "theorem" that, for any X, there is a philosophy of X (§2.7), there is another that says:
  "For any possibility you can name, there exists a philosopher who turned it into a theory"
  (Casati, R. (2000). Shadows: Unlocking Their Secrets, from Plato to Our Time. Vintage/Random House, 2004, New York. Translated by Abigail Asher; p. 65).

• Concerning the relation of an output to a non-existent input, there are philosophical theories about relations to non-existent entities; see:
  • Grossmann, R. (1974). Meinong. Routledge and Kegan Paul, London, p. 109.

  • Rapaport, W.J. (1985-1986). Non-existent objects and epistemological ontology. Grazer Philosophische Studien, 25/26:61–95, §4.5.

§16.7: Algorithms Don't Need a Purpose:

• The philosopher Frances Egan takes the mathematical functional view, focusing (in our terminology) on A, not G:
  • Egan, F. (1991). Must psychology be individualistic? Philosophical Review, 100(2):179–203, esp. pp. 196–197.

  • Egan, F. (1995). Computation and content. Philosophical Review, 104(2):181–203, esp. p. 185.

  On that view, Marr's "computational" level is not teleological (Egan 1991, p. 201).

• Shagrir, O. and Bechtel, W. (2015). Marr's computational-level theories and delineating phenomena. In Kaplan, D., editor, Integrating Psychology and Neuroscience: Prospects and Problems. Oxford University Press, Oxford, §2.2
  • Suggests that Marr's "computational" level conflates two separate, albeit related, questions: "why" and "what".

• On mathematical understanding, see:
  • Rapaport, W.J. (1988a). Syntactic semantics: Foundations of computational natural-language understanding. In Fetzer, J.H., editor, Aspects of Artificial Intelligence, pages 81–131. Kluwer Academic Publishers, Dordrecht, The Netherlands.

  • Rapaport 2012

  and Albert Goldfain's work on how to get AI computer systems to understand mathematics in addition to merely doing it:

  • Goldfain, A. (2006). Embodied enumeration: Appealing to activities for mathematical explanation. In Beetz, M., Rajan, K., Thielscher, M., and Rusu, R.B., editors, Cognitive Robotics: Papers from the AAAI Workshop (CogRob2006), Technical Report WS-06-03, pages 69–76. AAAI Press, Menlo Park, CA.

  • Goldfain, A. (2008). A computational theory of early mathematical cognition. PhD dissertation (Buffalo: SUNY Buffalo Department of Computer Science & Engineering)

§16.8: Algorithms and Goals:

• On "succinct ways of saying what an algorithm is for", see Chaitin 2002 and Chaitin 2006b.

• Turner, R. (2019). Correctness, explanation and intention. In Manea, F., Martin, B., Paulusma, D., and Primiero, G., editors, Computing with Foresight and Industry. CiE 2019, pages 62–71. Springer, Cham, Switzerland.
  • A useful discussion of the intentionality or purposefulness of computer programs in the context of program verification.

• Harbecke, J. and Shagrir, O. (2019). The role of the environment in computational explanations. European Journal for Philosophy of Science, 9(37), §§1, 2.
  • Offers an excellent summary of the Do A/To G, do A distinction.

§16.9: Computing with Symbols or Their Meanings:

• The distinction between symbols and meanings also appears in the philosophy of mathematics concerning "structuralism":
  • Is the pattern, or structure, of the natural numbers all that matters?
  • Or does it also matter what the natural numbers in the pattern "really" are?

  For discussion, see:

  • Benacerraf, P. (1965). What numbers could not be. Philosophical Review, 74(1):47–73.

  • Horsten, Leon, Philosophy of Mathematics, The Stanford Encyclopedia of Philosophy (Spring 2022 Edition), Edward N. Zalta (ed.), §4.

• There are debates in the philosophy of mathematics concerning "pure" or "abstract" math vs "applied" math; see, e.g.:
  • Marshall, R. (2019). Philosophy, maths, logic and computers. 3:16

  These debates are also relevant to the "abstract" Do A vs. the more "applied" To G, do A.
  They may also be related to the distinction between "pure" syntax vs "applied" semantic interpretation.

§16.10: Syntactic, Internal, and Indigenous Semantics

• On syntactic semantics, see:

  • Rapaport 1986a

  • Rapaport 1988a

  • Rapaport, W.J. (1995). Understanding understanding: Syntactic semantics and computational cognition. In Tomberlin, J. E., editor, AI, Connectionism, and Philosophical Psychology (Philosophical Perspectives, Vol. 9), pages 49–88. Ridgeview, Atascadero, CA.
    • Reprinted in Toribio, Josefa, & Clark, Andy (eds.) (1998), Language and Meaning in Cognitive Science: Cognitive Issues and Semantic Theory (Artificial Intelligence and Cognitive Science: Conceptual Issues, Vol. 4) (New York: Garland).

  • Rapaport, W.J. (2000). How to pass a Turing test: Syntactic semantics, natural-language understanding, and first-person cognition. Journal of Logic, Language, and Information, 9(4):467–490.
    • Reprinted in James H. Moor, The Turing Test: The Elusive Standard of Artificial Intelligence (Dordrecht, The Netherlands: Kluwer Academic, 2003): 161–184.

  • Kay, K. (2001). Machines and the mind. The Harvard Brain, 8(Spring).

  • Rapaport, W.J. (2003). What did you mean by that? Misunderstanding, negotiation, and syntactic semantics. Minds and Machines, 13(3):397–427.

  • Rapaport, W.J. (2006a). How Helen Keller used syntactic semantics to escape from a Chinese room. Minds and Machines, 16:381–436

  • Rapaport 2012

  • Rapaport 2017

• On Rescorla's "indigenous semantics", see:
  • Rescorla, M. (2012). Are computational transitions sensitive to semantics? Australian Journal of Philosophy, 90(4):703–721.

  • Rescorla, M. (2014). The causal relevance of content to computation. Philosophy and Phenomenological Research, 88(1):173–208.

• Indigenous semantics emphasizes causal relations, whereas syntactic semantics emphasizes the importance of conceptual-role semantics:
  • Rapaport, W.J. (2002). Holism, conceptual-role semantics, and syntactic semantics. Minds and Machines, 12(1):3–59

• On an "internal" semantics in Shannon's theory of information, related to Turing's work during World War II on decoding, see:
  • Isaac, Alistair M.C. (2019), "The Semantics Latent in Shannon Information", British Journal for the Philosophy of Science 70(1):103–125, esp. p. 106.

§16.11.2: Symbols: Marks vs. Meanings

• On content:
  • Here is one standard way of discussing the nature of "content":
    "The defining characteristic of representations is their aboutness, that is to say, the fact that representations are about something other than themselves. … What representations are about or refer to are their contents … ."
    — Vernazzani, Alfredo and Coelho Mollo, Dimitri (2023), "The Formats of Cognitive Representation: A Computational Account", Philosophy of Science (forthcoming), §1, p. 3; my italics.

    On this view, representations are semantic in the sense of §13.2.2.3: They are relations between two things, (1) that which does the representing and (2) that which is represented. Could something represent itself? If it could, it would be syntactic, in the sense of §13.2.2.2. In syntactic semantics (§16.10.2), one part of a system represents, or is about, another part of that same system.

    Next, note that if the content of a representation is what it is about or refers to, then its content is external to the representation. But later (§2.1, p. 7), the authors just quoted say this:

    "Coloured pieces of paper, binary code stored in a memory drive, and patterns of neural activation in the brain can all carry representational content: they can all be representations, say, of [a person]’s face. As carriers of content, these internal states and processes are called representational vehicles."
    But, if those internal states "carry" content, then what they carry would seem to have to be internal to the representation.

    And that is why I find the term 'content' confusing. At best, there would seem to be two kinds of content: an internal content carried by the representing vehicle and an external content (which, like Santa Claus, need not exist!)