Springer Nature is making SARS-CoV-2 and COVID-19 research free. View research | View latest news | Sign up for updates

How Frogs See the World: Putting Millikan’s Teleosemantics to the Test


How do frogs represent their prey? This question has been the focus of many debates among proponents of naturalistic theories of content, especially among proponents of teleosemantics. This is because alternative versions of the teleosemantic approach have different implications for the content of frog representations, and it is still controversial which of these content ascriptions (if any) is the most adequate. Theorists often appeal to intuitions here, but this is a dubious strategy. In this paper, I suggest an alternative, empirical test for theories of content. I propose that we should examine whether a theory generates content ascriptions that fit with our best scientific explanations of animal behavior. I then focus on the most prominent version of teleosemantics, Ruth Millikan’s consumer-oriented approach, and argue that it fails the empirical test in the frog case, since it yields a content ascription that (i) does not include properties that should be included (namely, being small, dark and moving) and (ii) includes a property that should not be included (namely, being frog food). This is an important result in itself, but it also demonstrates by way of example how progress can be made in the complex debate about theories of content.

This is a preview of subscription content, log in to check access.


  1. 1.

    Among the optimists are the proponents of teleosemantics mentioned below, but also e.g. Allen (1992) and Bermúdez (2003). The paradigmatic representative of the opposing, “pessimistic” position is Donald Davidson (1982).

  2. 2.

    This story was originally inspired by Lettvin et al.’s (1959) experiments on leopard frogs (Rana pipiens), but has since developed a life of its own.

  3. 3.

    As Karen Neander rightly points out, the content ascription fly at location l, which some theorists have proposed, is incompatible with this basic empirical fact (Neander 2006: 170).

  4. 4.

    Millikan’s seminal work is, of course, Millikan (1984). A useful summary of her main ideas can be found in Millikan (1989a). For a more detailed introduction, cf. Shea (2005).

  5. 5.

    One may also describe the situation as one where the N-state has several different consumers. But to make things easier, I treat all the mechanisms influenced by the N-state as part of one big subsystem, which enables me to speak of the consumer of the N-state.

  6. 6.

    Millikan develops this notion of function in Millikan (1984), chapters 1 and 2, and defends it against objections in Millikan (1989b).

  7. 7.

    The “mapping relation” is introduced in Millikan (1984), chapter 6. For a more recent discussion, cf. Millikan (2005).

  8. 8.

    More precisely, Millikans theory implies that frog food at location l is the N-state’s descriptive content. As Millikan (1995) emphasizes, representational states with descriptive content can also have additional directive content. Millikan calls these states “pushmi-pullyu representations”. The N-state might be a pushmi-pullyu representations (with the additional directive content: catch the item at location l!), but I will not pursue this question further and focus exclusively on the N-state’s descriptive content.

  9. 9.

    It should be noted that the localization content of the N-state, rendered in all these ascriptions as “at location l”, can also give rise to controversies. How large is the area that is singled out by the localization content? Is the position of the item in question represented in an allocentric or in an egocentric coordinate system? Cf. Neander (2006: 180–183), and Bermudez (2003: 95–104), for interesting discussions of these issues.

  10. 10.

    To be fair to Pietroski, his switch from “this content ascription is implausible” to “the intentional explanation employing this content ascription is implausible” does constitute a significant step forward, cf. the discussion in the section “An Objection” below.

  11. 11.

    Another, less obvious instance of this argumentative strategy is Neander’s (1995) objection against Millikan which is based on the intuition that fertile female that will survive long enough to reproduce is implausible as a content ascription for perceptual states of male hoverflies (Neander 1995: 127). Of course, I agree that such a content ascription is implausible, but I contend that mere implausibility is not a sufficient reason to reject it.

  12. 12.

    This holds for explanations in scientific psychology as well as for folk psychological explanations, which – pace Paul and Patricia Churchland – are also very successful in explaining and predicting human behavior; cf. Horgan and Woodward (1985).

  13. 13.

    Shettleworth (2010) provides a comprehensive overview of this field. The fact that many ethological explanations are intentional is emphasized by Allen and Bekoff (1997), Bermudez (2003) and Cheney and Seyfarth (1990).

  14. 14.

    Karen Neander pursues a similar strategy: she argues that “some candidate contents serve the purposes of mainstream cognitive science better than others do” (Neander 2006: 168), and that this favors certain theories of content over others. Her arguments, though different from mine, are in some respects congenial to the claims that I defend in this paper. But on the whole, it seems to me that Neander relies too heavily on the intentional or quasi-intentional characterizations of brain processes by scientists, and on their judgments concerning the “appropriateness” of certain behavioral reactions, despite the fact that the scientists’ statements on these matters are quite heterogeneous and often unclear (as Neander herself admits at some point, cf. Neander 2006: 183f).

  15. 15.

    Cf. Price (1998: 71), who suggests that “catchable flies” might be a more appropriate content ascription than just “flies”.

  16. 16.

    The following account is based on Ingle (1998), Nishikawa (1989), p. 475f, and Stebbins and Cohen (1995), chapter 7.

  17. 17.

    To be precise, frogs respond to the real size of objects only within a certain range, up to approximately 15 cm, which is the maximal ‘striking distance’ for a frog (Ingle 1998: 173f). For further limitations of the frogs size-distance constancy mechanisms, cf. Ingle (1998: 176f).

  18. 18.

    The optic tectum is a major part of the midbrain. It seems to be “the highest brain ‘center’ for visual motor control mechanisms [in amphibians; PS] and is involved in triggering prey-catching behavior” (Stebbins and Cohen 1995: 48).

  19. 19.

    Visual illusions in frogs are reported by Bastakov (1997, 2008). For a description of malfunctioning visual systems in toads, cf. Ewert (1968).

  20. 20.

    Many ethologists who study amphibians shy away from explicitly intentional expressions like “seeing” and “perceiving” to avoid allegations of anthropomorphism, but not all of them do (cf. Cobas and Arbib 1992; Stebbins and Cohen 1995; Stull and Gruberg 1998). In any case, my main point is not that scientists explicitly use intentional terms, but that their use is warranted by the empirical facts.

  21. 21.

    On the other hand, sensory input explanations have a specificity that is not to be found in intentional explanation: they explain not only why the frog showed a feeding response, but also why it oriented before snapping, why it responded exactly 2 s after spotting the prey, and so on.

  22. 22.

    One question remains: what if neurobiologists identify the “N-state” – the state of the optic tectum that causes the feeding response? Can’t we then replace the intentional explanation “the frog exhibits a feeding response because it has a perception as of a small, dark, moving object” by the neural-level explanations “the frog exhibits a feeding response because its optic tectum is in such-and-such a state”? This is a very fundamental worry, which applies to all (putatively) intentional systems, including humans, and it is not directly relevant here, since my aim in this paper is not to give a general argument for the indispensability of intentionality. My reply, however, is the familiar one: I think that such a neural-level explanation is also less general than the intentional explanation, because the perceptual state is multiply realizable. There can be many different “N-states” for different species of frogs (toads, salamanders, lizards, etc.), but all of them can be described as “perceptions as of small, dark, moving objects”, and fall under the same (suitably general) psychological laws. Again, the neural-level explanation has the advantage of specificity (it explains the particular behavior in more detail), while the intentional explanation has the advantage of greater generality (it places the behavior in a more general, species-transcending pattern).

  23. 23.

    I speak of “deep” causal/functional properties here, since it might be the case that some surface properties, e.g. colors, are also causal powers – powers to affect our sensory system in certain ways. These “superficial” causal/functional properties need to be treated differently, but a discussion of this issue would take us too far afield.

  24. 24.

    For a brief summary of the most well-known studies, cf. Shettleworth (2010), chapter 11.

  25. 25.

    I use the expression “tasty substance” in this sentence only as a placeholder. It is, of course, an interesting question how the piece of meat is represented by the crow, but my focus here is solely on the representation of the trap-tube/trap-table.

  26. 26.

    Cf. Dretske’s (1988) “structuring causes” of behavior.

  27. 27.

    Thanks to Colin Allen, Fabian Hundertmark and Albert Newen for bringing to my attention that this was an objection that I needed to address.

  28. 28.

    That such descriptions are mistaken is conventional wisdom among psychologists (see e.g. Pinker 1997: 400f).

  29. 29.

    Cf. Pietroski (1992: 278–281), who comes to a similar conclusion, though through a completely different route.


  1. Allen, C. (1992). Mental content. British Journal for the Philosophy of Science, 43, 537–553.

  2. Allen, C. (2001). A tale of two froggies. Canadian Journal of Philosophy (Supplement), 27, 105–115.

  3. Allen, C., & Bekoff, M. (1997). Species of Mind. Cambridge, MA: MIT Press.

  4. Agar, N. (1993). What do frogs really believe? Australasian Journal of Philosophy, 71, 1–12.

  5. Bastakov, V. A. (1997). Visual illusions in frogs and toads. Perception, 26. ECVP Abstract Supplement.

  6. Bastakov, V. A. (2008). Size constancy mechanism and spatial illusions in an anuran. Perception, 37, 158. ECVP Abstract Supplement.

  7. Bermúdez, J. L. (2003). Thinking without Words. Oxford: Oxford University Press.

  8. Burge, T. (2009). Perceptual objectivity. Philosophical Review, 118, 285–324.

  9. Burge, T. (2010). Origins of Objectivity. Oxford: Oxford University Press.

  10. Cheney, D. L., & Seyfarth, R. M. (1990). How Monkeys See the World. Chicago: University of Chicago Press.

  11. Cobas, A., & Arbib, M. (1992). Prey-catching and predator-avoidance in frog and toad: defining the schemas. Journal of Theoretical Biology, 157, 271–304.

  12. Comer, C., & Grobstein, P. (1981). Tactually elicited prey acquisition behavior in the frog, Rana pipiens, and a comparison with visually elicited behavior. Journal of Comparative Physiology, 142, 141–150.

  13. Davidson, D. (1982). Rational animals. Dialectica, 36, 317–327.

  14. Dretske, F. (1988). Explaining Behavior. Cambridge, MA: MIT Press.

  15. Ewert, J.-P. (1968). Der Einfluss von Zwischenhirndefekten auf das Beute- und Fluchtverhalten der Erdkröte (Bufo Bufo L.). Zeitschrift für vergleichende Physiologie, 61, 41–70.

  16. Horgan, T., & Woodward, J. (1985). Folk psychology is here to stay. Philosophical Review, 94, 197–226.

  17. Ingle, D. (1968). Visual releasers of prey-catching behavior in frogs and toads. Brain, Behavior and Evolution, 1, 500–518.

  18. Ingle, D., & Cook, J. (1977). The effect of viewing distance upon size-preference of frogs for prey. Vision Research, 17, 1009–1014.

  19. Ingle, D. (1998). Perceptual constancies in lower vertebrates. In V. Walsh & J. Kulikowski (Eds.), Perceptual Constancy (pp. 173–191). Cambridge: Cambridge University Press.

  20. Lettvin, J. Y., Maturana, H. R., McCulloch, W. S., & Pitts, W. H. (1959). What the frog’s eye tells the frog’s brain. Proceedings of the Institute of Radio Engineers, 1940–1951.

  21. Linzey, D. (1967). Food of the leopard frog, Rana p. pipiens, in Central New York. Herpetologica, 23, 11–17.

  22. Maturana, H. R., Lettvin, J. Y., McCulloch, W. S., & Pitts, W. H. (1960). Anatomy and physiology of vision in the frog (Rana pipiens). Journal of General Physiology, 47, 129–175.

  23. Millikan, R. (1984). Language, Thought, and Other Biological Categories. Cambridge, MA: MIT Press.

  24. Millikan, R. (1989a). Biosemantics. Journal of Philosophy, 86, 281–297.

  25. Millikan, R. (1989b). In defense of proper functions. Philosophy of Science, 56, 288–302.

  26. Millikan, R. (1991). Speaking up for darwin. In B. Loewer & G. Rey (Eds.), Meaning in mind. Fodor and his critics (pp. 151–164). Oxford: Blackwell.

  27. Millikan, R. (1995). Pushmi-Pullyu representations. Philosophical Perspectives, 9, 185–200.

  28. Millikan, R. (2004). Varieties of Meaning. Cambridge, MA: MIT Press.

  29. Millikan, R. (2005). On Meaning, Meaning, and Meaning. In R. Millikan (Ed.), Language: A Biological Model (pp. 53–76). Oxford: Clarendon.

  30. Neander, K. (1995). Misrepresenting & malfunctioning. Philosophical Studies, 79, 109–141.

  31. Neander, K. (2006). Content for Cognitive Science. In G. MacDonald & D. Papineau (Eds.), Teleosemantics (pp. 167–194). Oxford: Oxford University Press.

  32. Nishikawa, K. (1989). Organismal” vs. “Mechanistic” biology. Herpetologica, 45, 473–479.

  33. Papineau, D. (1984). Representation and explanation. Philosophy of Science, 51, 550–572.

  34. Papineau, D. (1998). Teleosemantics and indeterminacy. Australasian Journal of Philosophy, 76, 1–14.

  35. Pietroski, P. (1992). Intentionality and teleological error. Pacific Philosophical Quarterly, 73, 267–282.

  36. Pinker, S. (1997). How the Mind Works. New York: Norton.

  37. Price, C. (1998). Determinate functions. Nous, 32, 54–75.

  38. Shea, N. (2005). On Millikan. Belmont: Wadsworth.

  39. Shettleworth, S. (2010). Cognition, Evolution, and Behavior. Oxford: Oxford University Press.

  40. Stebbins, R., & Cohen, N. (1995). A Natural History of Amphibians. Princeton: Princeton University Press.

  41. Stull, A., & Gruberg, E. (1998). Prey-selection in the leopard frog: choosing in biased and unbiased situations. Brain, Behavior and Evolution, 52, 37–45.

  42. Taylor, A. H., Hunt, G. R., Medina, F. S., & Gray, R. D. (2009). Do new caledonian crows solve physical problems through causal reasoning? Proceedings of the Royal Society B, 276, 247–254.

Download references


I would like to thank Colin Allen, Hannah Altehenger, Fabian Hundertmark, Albert Newen, Christian Nimtz, and an anonymous referee for this journal for many helpful and challenging comments on previous versions of this paper.

Author information

Correspondence to Peter Schulte.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Schulte, P. How Frogs See the World: Putting Millikan’s Teleosemantics to the Test. Philosophia 40, 483–496 (2012). https://doi.org/10.1007/s11406-011-9358-x

Download citation


  • Teleosemantics
  • Animal cognition
  • Perception
  • Mental content
  • Intentionality