A functional naturalism

We assert that Nature operates for the sake of an end, and that this end is a good (Aristotle, PN 455b17).


I provide two arguments against value-free naturalism. Both are based on considerations concerning biological teleology. Value-free naturalism is the thesis that both (1) everything is, at least in principle, under the purview of the sciences and (2) all scientific facts are purely non-evaluative. First, I advance a counterexample to any analysis on which natural selection is necessary to biological teleology. This should concern the value-free naturalist, since most value-free analyses of biological teleology appeal to natural selection. My counterexample is unique in that it is likely to actually occur. It concerns the creation of synthetic life. Recent developments in synthetic biology suggest scientists will eventually be able to develop synthetic life. Such life, however, would not have any of its traits naturally selected for. Second, I develop a simple argument that biological teleology is a scientific but value-laden notion. Consequently, value-free naturalism is false. I end with some concluding remarks on the implications for naturalism, the thesis that (1). Naturalism may be salvaged only if we reject (2). (2) is a dogma that unnecessarily constrains our conception of the sciences. Only a naturalism that recognizes value-laden notions as scientifically respectable can be true. Such a naturalism is a functional naturalism.

This is a preview of subscription content, access via your institution.


  1. 1.

    Of course, this teleological explanation is not the only sort of explanation that could be offered in reply to the question at hand. One may also offer a mechanistic explanation that describes the causal processes that allows a watch to tell time. My point is merely that, in ordinary life, we would—at least sometimes—happily accept a teleological explanation.

  2. 2.

    For helpful introductions to the literature on biological functions, see Bedau (1993), Garson (2016) and Wouters (2005).

  3. 3.

    And as I will soon discuss, biological functions give rise to good effects for their possessors. Artefactual functions need not do this.

  4. 4.

    McShea (2012) has argued that “goal-directedness [or teleology] is a function of the perceived complexity of the system” (p. 682, my emphasis). One advantage of this view is that it makes teleology naturalistically respectable whole avoiding the objection I raise in Sect. 2 against the etiological view. But, strikingly, this account makes it up to us whether something is teleological. For critical discussion of such “mentalism,” see Bedau (1990). Here, I will assume the following: Whether or not something has a biological function is independent of our cognitive activity.

  5. 5.

    This statement should perhaps be tempered so as to not offend the tastes of scientific antirealists. Scientific antirealists who are also “pro-teleology”—teleological realists—may accept that teleological explanations play some role in our best scientific theories. That is, such antirealists will say that our best theories appeal to some teleological explanations.

  6. 6.

    For just one influential naturalistic analysis in this vein, see Millikan (1984, p. 28).

  7. 7.

    Perhaps ironically, Bedau (2010) defends a view of biological life on which crystals count as alive. This view, however, is extremely controversial. I will assume that crystals are not alive.

  8. 8.

    It is a common thought that natural selection suffices for teleology. For instance, Basl and Sandler conclude that “selection etiologies are sufficient for genuine teleological organization” (2013, p. 699) after only a brief discussion of the claim.

  9. 9.

    One might imagine God creating a human out of nothing or a lightning strike hitting a tree and causing particles to be rearranged so that a man, Swampman, arises from the ashes. For the original Swampman case, see Davidson (1987, p. 443). It seems—at least to many—that these cases, while perhaps metaphysically possible, are too fantastic to bear much weight. Value-free naturalists generally “have little patience with purely science-fiction counterexamples” (Garson 2016, p. 10). On the other hand, the case I develop in the main text is likely to actually occur.

  10. 10.

    At the very least, this is true of relatively simple lifeforms like amoebas. This is all that is needed for the case that I will develop. For all I will say, conscious organisms like humans may be more than merely complex, organized collections of physical matter.

  11. 11.

    I am only considering animal eukaryotic cells here.

  12. 12.

    If life is a matter of degree—so that, at least in principle, some things may be more alive than others—rather than a binary matter, then we may be justified in claiming that synthetic biologists have already developed entities that enjoy an intermediate status on the “liveliness” scale.

  13. 13.

    Since the system integrates the program, metabolism, and container roles, each of these functional roles is such that some part of the system performs it.

  14. 14.

    Where p is a part of organism o, the inference from ‘p does something F to keep o alive’ to ‘p has the function of F-ing’ is admittedly defeasible. But I find it intuitive enough that, by default, we should accept the inference. This default, of course, can be overridden.

  15. 15.

    As is common knowledge among those who have studied biology in high school, mitochondria are the powerhouse of a cell!

  16. 16.

    Holm (2012) and Holm (2013) have also argued that bottom-up synthetic biology presents a problem for accounts of teleology that appeal to natural selection. But my modal argument involving Eve and Steve, which I develop below in the main text, is novel.

  17. 17.

    My argument is inspired by the (in)famous metaphysical argument that a statue and its constituent clay are distinct objects, since the clay may exist even if the statue does not. For a few defenses of the attendant view of material constitution, see Fine (2003), Oderberg (1996) and Wiggins (1968).

  18. 18.

    This seems clearly possible. I might know how to build a house even if I mistakenly believe that doors are for letting light in and windows are for allowing people to enter, and leave, a house. I might put together a perfectly normal house and intend the doors to function as windows and the windows to function as doors.

  19. 19.

    It is important to note that Alice has a de re intention, of what is in fact Steve’s metabolizer, that it performs the program role. This means that, whatever the metabolizer actually is, Alice intends, of it, that it perform the program role. After all, after Alice hits her head, she begins to believe, of whatever actually plays the metabolism role, that it plays the program role. Alice may have the de dicto intention that whatever plays the metabolism role plays the metabolism role, but this is unproblematic. I only require Alice’s de re intention.

    To see the distinction between de re and de dicto intentions, consider another case. A philosophy student may have the de dicto intention that her conclusion not be a premise, but still have the de re intention, of what is in fact her conclusion, that it be a premise. When this happens, we say that the student has (unwittingly) made a question-begging argument.

  20. 20.

    Why not claim that Steve’s metabolizer fails to have the function of ensuring metabolism occurs, but instead functions as if it ensures metabolism occur? A rock may not have the function of being a chair, but it may function as if it is a chair. A relevant difference between the rock and Steve’s metabolizer and the rock, however, is that someone must intend, of the rock itself that it function as a chair. On the other hand, Alice—and we may suppose, everyone else—never intended, of Steve’s metabolizer, that it ensures that metabolism occurs. The rock has something like an artefactual function to serve as a chair, whereas Steve’s metabolizer does not have an artefactual intention to ensure that metabolism occurs. A second relevant difference is that the rock, being inanimate, has no interests in any sense. There is nothing that is literally good for it. However, Steve, being alive, does have interests in some sense. There are states of affairs—like being in a nutrient-rich environment—that are literally good for Steve. Ensuring metabolism is literally good for Steve, whereas serving as a chair is not literally good for a rock.

  21. 21.

    Ignore haecceities like the property of being Steve. Eve could have had all the non-haecceitistic properties of Steve. This is all I require.

  22. 22.

    This is an application of Leibniz’s Law. If x and y have different (modal) properties, then x ≠ y.

  23. 23.

    After all, if my heart has the function of pumping blood, then it is eminently plausible that my physical duplicate’s heart has the function of pumping blood. Just as I would die if my heart failed to pump blood, my physical duplicate would die if his heart failed to pump blood.

  24. 24.

    So as Sandler notes, a “synthesised organism still has a good, and it is still a good of its own” (2012, p. 52).

  25. 25.

    Why, after all, do we say that my heart has the biological function of pumping blood? It seems that its being a part of me and its regularly generating good effects for me are jointly sufficient for its having a function.

  26. 26.

    Basl (2012) advances this objection. He would be happy to accept that a synthetic organism like Eve is not the product of natural selection. Basl denies that “the only aetiologies capable of grounding teleology…are natural selection etiologies” (2012, p. 544). But Basl insists that the relevant artefactual etiologies “involve intentions on the part of the designer/user” (2012, p. 544).

  27. 27.

    Therefore, as an anonymous reviewer put the point, appealing to Darwin’s (1859) distinction between artificial and natural selection would not help the value-free naturalist here.

  28. 28.

    As an anonymous reviewer pointed out to me, it is hard to reason with someone who will bite any bullet. I cannot convince anyone who is so committed to an etiological view on which teleology is reducible to selection that she is eager to say that Eve and Steve’s metabolizers do not have functions. But this is not a problem specific to my argument. In the face of any genuine counterexample to her view, a dogmatist can always accept an absurd claim.

  29. 29.

    That is, it would be question-begging to deny that Steve’s or Eve’s metabolizer lacks a biological function solely on the grounds that some value-free analysis of teleology (that appeals to natural selection) is true. After all, I am currently arguing that no such analysis of teleology is true.

  30. 30.

    Bedau similarly argues that “value plays a role in…teleological explanations” (1992b, p. 805). Bedau is the most prominent defender of this view.

  31. 31.

    This Teleological Conditional bears some similarity to Bedau’s “first grade of teleology” (1992b, p. 787). But there are some differences. Bedau, for instance, states his first grade of teleology with a biconditional. Ayala states something close to the Teleological Conditional when he tells us that a “feature of a system will be teleological…if the feature has utility for the system…and if such utility explains the presence of the feature in the systems” (1970, p. 13). However, Ayala’s claim seems to constitute a weakened variant of the converse of the Teleological Conditional.

  32. 32.

    The name ‘Teleological Conditional’ is a misnomer since the statement is, strictly speaking, a universal generalization, not a conditional. But alas, the name ‘Teleological Generalization’ does not roll off the tongue as well.

  33. 33.

    For instance, introductory biology textbooks tell us facts like this: “[B]lood delivers nutrients and removes wastes throughout an animal’s body. These functions are made possible by the circulatory system” (Campbell and Reece 2005, p. 874, my emphasis).

  34. 34.

    If it is impossible for death to constitute an all-things-considered good, then this case is no counterexample to the Teleological Conditional. Of course, I would welcome this result.

  35. 35.

    Thanks to an anonymous referee for suggesting this seeming counterexample to me.

  36. 36.

    Bedau, a teleological realist, also claims that the goodness present in teleology is pro tanto, not all-things-considered: “[T]he goodness of Cing implies merely that Cing confers a good, not that Cing is best overall…So the value analysis requires, not that Cing confers…the best good, but only that Cing confers some good” (Bedau 1992b, p. 791, his emphasis).

  37. 37.

    Just to be clear, Basl and Sandler (2013) do not explicitly raise this objection against teleological realism in particular.

  38. 38.

    Thanks to an anonymous reviewer for pressing me to clarify the notion of an organism’s good.

  39. 39.

    For a discussion of why it is unproblematic for the teleological realist if the notion of goodness is vague, see Bedau (1992b, pp. 792–793).

  40. 40.

    Regan (1976, p. 487) carefully distinguished between having an interest in something and taking an interest in something. The latter requires a mind, whereas the former—so I claim—does not. A plant can have an interest in sunlight even if it does not take an interest in sunlight. It is in a plant’s interests that it receive enough sunlight, but plants plausibly do not have minds.

  41. 41.

    Some value-free naturalists, like Basl and Sandler (2013), may be tempted to analyze facts about what is good for an organism in terms of facts about selection. But given my arguments in Sect. 2, this strategy is incompatible with teleological realism. Biological functions are not to be understood in terms of selection. But given teleological realism, biological functions are to be understood in terms of an organism’s good.

  42. 42.

    Here, I use subscripts in order to make it unambiguous what I use pronouns to refer to. For example, in ‘John and James3 ate cake, but he3 would have preferred pie’, the pronoun ‘he’ refers to James, not John.

  43. 43.

    I intend the de re reading of premise (2), which states that if value-free naturalism is true, then there is no biological fact stating that something has a property that, as a matter of fact, happens to be an evaluative property.

  44. 44.

    I assume that biological functions are properties. It is not clear that this is strictly speaking correct, but the simplifying assumption that functions are properties makes it easy to state this argument. Nothing of substance rests on this simplifying assumption. If functions are not properties, premises (2) and (3) of the Value Argument need only be slightly reworded.

  45. 45.

    Cameron (2004) argues for a similar conclusion, but he appeals to strong emergence. I wish to remain neutral as to whether strong emergence exists. And I can: Perhaps some complex microphysical facts count as evaluative. For a vision of a naturalism that can, in principle, accommodate this, see Sect. 4.

  46. 46.

    And it seems right to be skeptical of any “disjunctive” analysis on which natural selection plays a role in only one disjunct in the analysans. Why would natural selection only sometimes contribute to teleology?

  47. 47.

    This is Wright’s (1976, p. 39) analysis of biological teleology. I have reworded it for simplicity’s sake.

  48. 48.

    Bedau (1991, p. 648 fn8).attributes the case to Robert Van Gulick.

  49. 49.

    A slight variant of the case is a counterexample to organizational accounts of teleology, whereby “self-maintenance is sufficient for teleology” (Holm 2012, p. 538). Here, “self-maintenance is characterised as a property of systems that are able to exert a causal influence on their surroundings in order to maintain…the boundary conditions required for their own existence” (Holm 2012, p. 537). Suppose the stick would be destroyed if it were not pinned to the rock. So, the stick is a self-maintaining system. But then it seems organizational accounts are committed to the absurd claim that the stick is a teleological system. But neither the stick nor any of its parts has any biological functions. Relatedly, Basl (2012, p. 546, fn 10) points out that Holm’s organizational account is committed to the—in my view, absurd—claim that hurricanes and candle flames count as teleological systems.

  50. 50.

    And we must weigh the advantages of accepting any such analysis against the intuitive appeal of the Teleological Conditional. The Teleological Conditional supports premise (3) of the Value Argument. The value-free analyses we are considering are posed as objections to (3). Accepting such an analysis requires rejecting the Teleological Conditional, and this is, in my view, a substantial cost to pay.

  51. 51.

    For critical discussion of “mentalist” analyses of biological teleology by a teleological realist, see Bedau (1990). For critical discussion of “systems” analyses by a teleological realist, see Bedau (1992a).

  52. 52.

    Bedau (1991) seems to defend a teleological realism on which value is irreducible. While I am sympathetic to this view, I do not wish to commit myself to it here.

  53. 53.

    For one helpful discussion of Aristotle’s views on teleology, see Cameron (2010).

  54. 54.

    One interesting upshot of such a naturalism concerns Teleological Individualism, “the view that organisms…are…goal-oriented systems while biological collectives…are mere assemblages of organisms” (Basl 2017, p. 1058). Basl (2017) argues that Teleological Individualism is incompatible with etiological accounts of teleology. So, it may be good news for Teleological Individualism that etiological accounts are false. Proponents of Teleological Individualism are better off accepting teleological realism.


  1. Aristotle. (1908). Parva Naturalia. In J. A. Smith & W. D. Ross (Trans.), The works of Aristotle. Oxford: Clarendon Press.

  2. Ayala, F. (1970). Teleological explanations in evolutionary biology. Philosophy of Science, 37(1), 1–15.

    Article  Google Scholar 

  3. Basl, J. (2012). Nothing good will come from giving up on aetiological accounts of teleology. Philosophy and Technology, 25(4), 543–546.

    Article  Google Scholar 

  4. Basl, J. (2017). A trilemma for teleological individualism. Synthese, 194(4), 1057–1074.

    Article  Google Scholar 

  5. Basl, J., & Sandler, R. (2013). The good of non-sentient entities: Organisms, artifacts, and synthetic biology. Studies in History and Philosophy of Biological and Biomedical Sciences, 44(4), 697–705.

    Article  Google Scholar 

  6. Bedau, M. (1990). Against mentalism in teleology. American Philosophical Quarterly, 27(1), 61–70.

    Google Scholar 

  7. Bedau, M. (1991). Can biological teleology be naturalized? Journal of Philosophy, 88(11), 647–655.

    Article  Google Scholar 

  8. Bedau, M. (1992a). Goal-directed systems and the good. The Monist, 75(1), 34–51.

    Article  Google Scholar 

  9. Bedau, M. (1992b). Where’s the good in teleology? Philosophy and Phenomenological Research, 52(4), 781–806.

    Article  Google Scholar 

  10. Bedau, M. (1993). Naturalism and teleology. In S. Wagner & R. Warner (Eds.), Naturalism: A critical appraisal (pp. 23–51). Notre Dame: University of Notre Dame Press.

    Google Scholar 

  11. Bedau, M. (2010). Four puzzles about life. In M. Bedau & C. Cleland (Eds.), The nature of life (pp. 393–404). Cambridge: Cambridge University Press.

    Google Scholar 

  12. Cameron, R. (2004). How to be a realist about “sui generis” teleology yet feel at home in the 21st century. The Monist, 87(1), 72–96.

    Article  Google Scholar 

  13. Cameron, R. (2010). Aristotle’s teleology. Philosophy Compass, 5(12), 1096–1106.

    Article  Google Scholar 

  14. Campbell, N., & Reece, J. (2005). Biology (7th ed.). San Francisco: Pearson Education Publishing.

    Google Scholar 

  15. Darwin, C. (1859). On the origin of species by means of natural selection. London: John Murray.

    Google Scholar 

  16. Davidson, D. (1987). Knowing one’s own mind. Proceedings and Addresses of the American Philosophical Association, 60(3), 441–458.

    Article  Google Scholar 

  17. Fine, K. (2003). On the non-identity of a thing and its matter. Mind, 112(446), 195–234.

    Article  Google Scholar 

  18. Garson, J. (2016). A critical overview of biological functions. New York: Springer.

    Google Scholar 

  19. Holm, S. (2012). Biological interests, normative functions, and synthetic biology. Philosophy & Technology, 25(4), 525–541.

    Article  Google Scholar 

  20. Holm, S. (2013). Organism and artifact: Proper functions in Paley organisms. Studies in History and Philosophy of Biological and Biomedical Sciences, 44(4), 706–713.

    Article  Google Scholar 

  21. Kripke, S. (1980). Naming and necessity. Cambridge: Harvard University Press.

    Google Scholar 

  22. Mayr, E. (1988). Toward a new philosophy of biology. Cambridge: Harvard University Press.

    Google Scholar 

  23. McShea, D. (2012). Upper-directed systems: A new approach to teleology in biology. Biology and Philosophy, 27(5), 663–684.

    Article  Google Scholar 

  24. Millikan, R. (1984). Language, thought, and other biological categories. Cambridge: MIT Press.

    Google Scholar 

  25. Oakwood, M., Bradley, A., & Cockburn, A. (2001). Semelparity in a large marsupial. Proceedings: Biological Sciences, 268(1465), 407–411.

    Google Scholar 

  26. Oderberg, D. (1996). Coincidence under a sortal. The Philosophical Review, 105(2), 145–171.

    Article  Google Scholar 

  27. Rasmussen, S., Constantinescu, A., & Svaneborg, C. (2016). Generating minimal living systems from non-living materials and increasing their evolutionary abilities. Philosophical Transactions Royal Society B, 371(1701), 1–10.

    Article  Google Scholar 

  28. Rasmussen, S., Liaohai, C., David, D., Krakauer, D., Packard, N., Stadler, P., et al. (2004). Transition from nonliving matter to living matter. Science, 303(5660), 963–965.

    Article  Google Scholar 

  29. Regan, T. (1976). Feinberg on what sorts of beings can have rights. The Southern Journal of Philosophy, 14(4), 485–498.

    Article  Google Scholar 

  30. Romero, L., & Butler, L. (2007). Endocrinology of stress. International Journal of Comparative Psychology, 20(2), 89–95.

    Google Scholar 

  31. Sandler, R. (2012). The value of artefactual organisms. Environmental Values, 21(1), 43–61.

    Article  Google Scholar 

  32. Venter, C., Gibson, D., Glass, J., Lartigue, C., Noskov, V., Chuang, R., et al. (2010). Creation of a bacterial cell controlled by a chemically synthesized genome. Science, 329(5987), 52–56.

    Article  Google Scholar 

  33. Wiggins, D. (1968). On being in the same place at the same time. The Philosophical Review, 77(1), 90–95.

    Article  Google Scholar 

  34. Wouters, A. (2005). The function debate in philosophy. Acta Biotheoretica, 53(2), 123–151.

    Article  Google Scholar 

  35. Wright, L. (1976). Teleological explanations. Berkeley: University of California Press.

    Google Scholar 

Download references


For helpful discussion on the topic of this paper and for awesome philosophical tutelage in general, I wish to thank Mark Bedau. I also wish to thank several anonymous reviewers at Synthese for their helpful comments.

Author information



Corresponding author

Correspondence to Anthony Nguyen.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Nguyen, A. A functional naturalism. Synthese 198, 295–313 (2021). https://doi.org/10.1007/s11229-018-02002-x

Download citation


  • Philosophy of biology
  • Biological teleology
  • Teleological realism
  • Naturalism
  • Synthetic biology
  • Value