From Function to Flourishing: Neo-Aristotelian Ethics and the Science of Life

Seyedeh Parisa Moosavi Tabatabaei

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Department of Philosophy University of Toronto

Parisa Moosavi Doctor of Philosophy Department of Philosophy University of Toronto 2019

Abstract

Neo-Aristotelian ethical naturalism purports to place moral virtue in the natural world by showing that it is an instance of natural goodness—a kind of goodness supposedly also found in the biological realm of plants and non-human animals. One of the central issues facing this metaethical view concerns its commitment to a teleological conception of the nature of life that seems radically out of touch with the understanding of life in modern biology. In this dissertation, I aim to mend the relationship between neo-Aristotelian ethics and the science of biology by way of three contributions: First, I argue that contrary to what many contemporary neo-Aristotelians have claimed, the science of biology is relevant to assessing central commitments of neo-Aristotelian naturalism regarding the domain of life. Second, I provide new foundations for neo-Aristotelian naturalism by engaging recent and unexplored work in philosophy of biology on theories of function and the nature of living organisms.

Lastly, I develop and defend a novel account of the neo-Aristotelian concept of natural goodness that is distinctive for incorporating our scientific understanding of the nature of life.

Acknowledgments

I owe the deepest debt of gratitude to Sergio Tenenbaum and Denis Walsh, whose supervision and support was crucial for the completion of this dissertation. Sergio was an exceptional supervisor who was extremely generous with his ideas, time, and energy, not to mention his quick-witted humor. Many of the ideas presented here became clear only after long discussions with him. Denis is the intellectual predecessor to most of the work on philosophy of biology that appears in this dissertation. His work has inspired the direction I took with many of my proposals, and his clarifications and comments were always there when I needed them. I also owe a great deal to the other members of my committee, Phil Clark and Andrew Sepielli, whose suggestions led to critical improvements. Phil has heavily influenced the way I think about many key aspects of neo-Aristotelian naturalism, as I first started thinking about the subject when I took a reading course with him. Andrew also read drafts of all parts of this dissertation and helped me improve them greatly through fruitful discussions. I am also very grateful to Tristram McPherson and Brendan de Kenessy, my external and internal examiners, for reading this dissertation with great care and sharing many insightful questions and comments. In addition, this project has benefited from the feedback of many people who read and commented on drafts of the chapters or essay versions of the material presented here. Comments from John Hacker-Wright, Antoine Dussault, Reid Blackman, Said Saillant, and Benjamin Wald proved especially helpful. While working on this project, I have also been helped, encouraged, and inspired by many people in the philosophical community at the University of Toronto and beyond. I would like to thank Jennifer Nagel, Karolina Hübner, Alex Koo, Mary Frances Ellison, Margaret Opoku-Pare, and Belinda Piercy for superb professional insight and support. Among my fellow graduate students, colleagues, and friends, I would like to thank Johanna Thoma, Jonathan Payton, Zac Irving, Dominic Alford-Duguid, Etye Steinberg, Hasko von Kriegstein, Luke Roelofs, Fermin Fulda, Sean Smith, Jeremy Davis, Jessica Wright, Robyn Forman, Jared Riggs,

iii

Melissa Rees, Jacob Stump, Steve Coyne, Joshua Brandt, Cory Lewis, Alex Djedovic, Benjamin Chin-Yee, Sareh PourYousefi, and Federica Berdini. Finally, this project would not have been possible without the incredible support of my family and friends. I wish to thank my parents Azizeh Ghorbanoghli and Saeed Moosavi for their unconditional love and support and for always believing in me; my brother Farshid Moosavi for being a constant source of emotional support and laughter; my friends Farimah Hakemzadeh, Keiwan Wind, and Sanaz Fesharaki for their friendship and understanding during my most discouraged and frustrated hours; and my partner Steve Abra for making the last stages of writing this dissertation coincide with some of the happiest times of my life.

Contents

Introduction ...... 1

Part I: The Relevance of Biology to Assessing Neo-Aristotelian Naturalism

1 Neo-Aristotelian Naturalism as Ethical Naturalism ...... 4

1.1 Neo-Aristotelian Naturalism: Virtue as Natural Goodness ...... 4 1.2 The Question of Ethical Naturalism ...... 8 1.3 In Defense of a Naturalist Interpretation ...... 12 1.4 The Neo-Aristotelian Argument for Naturalism and Its Limitations ...... 16

2 The Evolutionary Objection to Neo-Aristotelian Naturalism ...... 19

2.1 An Evolutionary-Inspired Dilemma ...... 19 2.2 The Pollyanna Problem ...... 22 2.3 Naturalness as Continuity with Nature...... 26 2.4 The Selfish Gene Problem ...... 29 2.5 Are Flourishing-Based Functions Natural? ...... 33 2.6 The Argument from Representation of Life ...... 36 2.7 Making Peace with the Relevance of Biology ...... 41

Part II: New Biological Foundations for Neo-Aristotelian Naturalism

3 Is the Neo-Aristotelian Concept of Organism Presupposed in Biology?...... 45

3.1 The Need to Examine the Representation of Life in Biology ...... 45 3.2 The Modern Synthesis ...... 50 3.3 The Evo-Devo Approach ...... 56 3.4 In Search of the Concept of Organism in Biology ...... 62

4 From Biological Functions to Natural Goodness ...... 65

4.1 Toward a Novel Defence of the Grounding Premise ...... 65 4.2 Biological Functions Revisited ...... 68 4.3 Organizational Functions as Flourishing-Based Functions ...... 73 4.4 From Organizational Functions to Natural Goodness ...... 81

Part III: A Novel Form of Neo-Aristotelian Naturalism

5 Natural Goodness and Natural-Historical Goodness ...... 89

5.1 The Natural-Historical Conception of Natural Goodness ...... 89 5.2 Natural-Historical Evaluations and the Problem of Anti-Individualism ...... 94 5.3 Thompson’s Case against Individualism ...... 102 5.4 Can Natural Goodness Be Given an Individualistic Account? ...... 108

6 Natural Goodness without Natural History ...... 110

6.1 Teleological Explanation and Patterns of Counterfactual Dependence ...... 110 6.2 Toward an Alternative Account of Natural Goodness ...... 118 6.3 A Modal-Explanatory Account of Natural Goodness ...... 129 6.4 Concluding Remarks: Natural Goodness without Anti-Individualism ...... 136

Bibliography ...... 141

Introduction

This dissertation provides new foundations for thinking about morality in a world that increasingly seems to be the domain of natural science. One of the leading contemporary accounts of how morality fits into the natural world comes from Aristotle. The present-day Aristotelian view in metaethics, known as neo-Aristotelian naturalism, explains morality as based in human nature and thus part of the domain of life. But, in its current form, the view remains objectionably detached from the scientific understanding of life in biology. The aim of my project is to reshape the neo-Aristotelian framework by engaging the concepts and methods of philosophy of biology. I advocate a relation of continuity between our understanding of human and non-human life in ethics and biology, and bring to light the ways in the two domains of inquiry can inform each other. Different versions of neo-Aristotelian naturalism have been put forward by Philippa Foot (2001) and Rosalind Hursthouse (1999) among others, claiming that moral virtue is an instance of natural goodness—a kind of goodness supposedly already present in nature among plants and animals. Common to all versions of the view is a commitment to a teleological conception of the nature of life that seems radically different from the understanding of life in the science of biology. This has given rise to a serious criticism of the view: that it is untenable in light of modern evolutionary biology. The dissertation starts by addressing this problem using a recent and unexplored body of work in philosophy of biology. It then proceeds to develop a novel and revised version of neo-Aristotelian naturalism that is distinctive for incorporating our scientific understanding of the nature of life. The project consists of three parts: The first part of the dissertation establishes the relevance of biology in assessing the metaethical aspirations of the neo-Aristotelian view. I start, in Chapter 1, by arguing that neo- Aristotelian naturalism is a form of ethical naturalism in the standard, metaphysical sense. This would be to reject a recent misinterpretation of the view as merely claiming that morality is based in human nature while remaining neutral between metaphysical naturalism and non- naturalism. I reconstruct the neo-Aristotelians’ argument for naturalism in terms of their

1 central claim that the ethical domain is continuous with the natural domain of life. Then, in Chapter 2, I explain why biology is relevant to assessing this argument. I examine two influential objections to neo-Aristotelian naturalism that appeal to modern evolutionary biology to undermine the central commitments of this view. I frame the evolutionary challenge as a dilemma that seemingly forces neo-Aristotelians to either contradict our considered moral judgment or abandon ethical naturalism. Resolving the dilemma requires showing that a morally plausible conception of human nature can be continuous with the natural domain of non-human life. In order to do this, neo-Aristotelians need to argue that the evaluations that they call ‘natural goodness’ are necessary for understanding the nature of non-human living things. My contention is that making this argument effectively requires consulting the science of biology and its explanatory concepts. The second part of the dissertation offers a solution to the apparent evolutionary dilemma by appealing to two explanatory concepts in biology: the concept of organism and the concept of function. In Chapter 3, I examine the role of the concept of organism in biology. I first argue that the Modern Synthesis theory of evolution, which understands evolution in reductive gene-centric terms, does not give an explanatory role to the concept of organism. But then I explore an alternative view of evolution that has emerged from recent works in evolutionary-developmental biology, where the concept of organism is given a significant explanatory role. I argue that although the place of the concept of organism in evolutionary biology is still contentious, the well-established place of this concept in developmental biology offers a promising starting point for defending the commitments of neo-Aristotelian naturalism. Next, in Chapter 4, I examine the explanatory concept of function in biology. I first argue that the widely-recognized, ‘etiological’ concept of function is not normatively adequate for grounding a suitable account of natural goodness. But then I argue that there is also a different, ‘organizational’ concept of function in biology that can lend support to the neo- Aristotelian project. Based on a version of the organizational account that I defend in this chapter, I argue that the concept of function underlying the neo-Aristotelian norms of natural goodness plays an indispensable explanatory role in biology and is therefore essential to understanding the nature of living things.

The last part of the dissertation draws the implications of consulting biology for the content of the neo-Aristotelian account. First, in Chapter 5, I draw a distinction between the general concept of natural goodness that is needed for the neo-Aristotelian metaethical project and the specific conception of this concept that neo-Aristotelians have proposed based on Michael Thompson’s (2008) work on the natural history of life-forms. I examine this ‘natural- historical’ account of natural goodness by highlighting its commitment to anti-individualism— the thesis that the standards of natural goodness are essentially shared between individual members of a species. I then appeal to the biological phenomenon of phenotypic plasticity to argue that this anti-individualism results in a class of potentially problematic implications for the natural-historical account, where it seems to make the wrong assessment of individuals with unique characteristics resulting from plasticity. After that, in my final chapter, I explore the possibility of an alternative account of natural goodness that is not committed to anti- individualism and does not face this problem. I argue that we can offer such an account by replacing the species-level patterns of regularity in Thompson’s account with a kind of regularity in the modal profile of a goal-directed system that is explanatorily significant and empirically observable in an individual. Thus, I draw from Denis Walsh’s (2012; 2013) account of teleological explanation to offer a novel account of natural goodness that is distinct from the natural-historical account and consistent with individualism. I conclude by arguing that my ‘modal-explanatory’ account of natural goodness is more aligned with our conception of non- human life in biology and better suited to capture the diversity of human form.

Chapter 1

Neo-Aristotelian Naturalism as Ethical Naturalism

Standard forms of ethical naturalism purport to give an account of ethical claims in terms of facts that fall within the scope of natural science. In contrast, neo-Aristotelian naturalism maintains that ethical claims are based on considerations of human nature while denying that human nature falls within the scope of natural science. This has led some commentators to conclude that neo-Aristotelian naturalism does not involve a commitment to metaphysical naturalism, and hence does not amount to a form of ethical naturalism. In this chapter, after giving an overview of neo-Aristotelian naturalism, I argue that it is in fact a form of ethical naturalism in the standard, metaphysical sense. I reconstruct the neo-Aristotelian argument for ethical naturalism in terms of a continuity between the ethical domain and the natural domain of life. I end the chapter by briefly discussing the place of natural science in assessing this argument—which is the focus of Chapter 2. I argue that although the neo-Aristotelians’ departure from a scientific characterization of the natural domain does not preclude them from having a claim to ethical naturalism, it does limit the scope and strength of their argument for this claim.

1.1 Neo-Aristotelian Naturalism: Virtue as Natural Goodness

Neo-Aristotelian ethical naturalism is a modern attempt to place moral virtue in the natural world using ideas from Aristotle’s ethics and metaphysics. Proponents of this view, such as Philippa Foot and Rosalind Hursthouse argue that moral goodness is an instance of what they call natural goodness, a kind of goodness supposedly already present in nature and in the biological realm of plants and non-human animals. On this view, the goodness of moral virtues such as justice and benevolence in humans is akin to the goodness of deep roots in an oak tree. In both cases what is good enables the kind of organism in question to flourish: in the same

4 way that deep roots are good in an oak insofar as they enable it to flourish as an oak, moral virtue is good in a human insofar as it enables them to flourish as a human. Thus, neo-Aristotelian naturalists adopt the Aristotelian doctrine that ethics concerns human life in particular, as opposed to a more general form of being such as rational agency or personhood. They maintain that the criteria for moral evaluation depend on the nature of human beings as living things. On the other hand, they appeal to a teleological view of the nature of living things, which resembles Aristotle’s metaphysics. They appeals to an account of the natural good of living things to argue that moral goodness is an instance of natural goodness. The first step of the neo-Aristotelian project is finding a place for natural goodness in the biological world, i.e., giving an account of what makes the deep roots of an oak tree good. A paradigmatic account of natural goodness is given by Philippa Foot (2001). According to Foot, natural goodness is a form of evaluation that is exclusively attributable in the case of living things in virtue of their nature as living things and according to their form of life. As Foot points out, almost anything can be evaluated in a context that sufficiently relates it to human concerns. What is distinctive about natural goodness is that it is “an intrinsic or autonomous kind of evaluation”—in the sense that it applies to the parts and processes of living things independently of the interests of humans or any other external party. Natural goodness only depends on the relation of an individual organism to its own form of life. This is a feature of living things that is notably different from what is found elsewhere in nature, for instance in rivers or storms (Foot 2001, 27). What is naturally good in the life of oak trees is good independently of how we evaluate them. The evaluation is made based on the nature of the oak itself and what enables it to flourish given its nature. This is why, according to Foot, we say that deep and sturdy roots are good in an oak, but pliability is good in a reed. Deep and sturdy roots are naturally good in an oak because as a tall, heavy tree, the plant needs deep and sturdy roots to flourish qua oak. In contrast, a reed needs pliability to flourish qua reed, because that is what enables the plant to tolerate wind. By the same token, neo-Aristotelians argue that

5 moral virtues like justice and benevolence are naturally good in human beings because they are what enable her to flourish qua human being. Foot’s account of natural goodness relies on Michael Thompson’s work on representations of life. According to Thompson (2008), there is a distinctive form of thought that is used in relation to the domain of life, and can underwrite evaluations of natural goodness and defect. This form of thought is manifested in the kind of generic descriptions that we typically encounter in a nature documentary or a field guide—statements such as “the bobcat has four legs” and “cherries bloom in spring”. Thompson calls these descriptions Aristotelian categoricals and the thoughts expressed in them natural-historical judgments. The general form of these judgments is something like “the S is/has/does F”, “Ss are/have/do F”, or “an S is/has/does F.” Thompson argues that when these judgments are used with the generic meaning, they articulate the characteristic features and activities of the form of life, or the life- form, to which individual living things belong. Moreover, he maintains that these judgments can underwrite evaluations of natural goodness and defect. Thompson argues that natural-historical judgments have a distinctive form of generality that is neither universal nor statistical. They are neither universal generalizations about all instances of a kind nor statistical generalizations about most instances of the kind. The truth of a natural-historical judgment about a life-form S is consistent with some or even most instances of S not matching the general description F that is expressed in the judgment. Consider the natural-historical judgment “the bobcat has four legs”. It concerns the life-form bobcat, which can have any number of bearers. But the judgment does not imply that all or even most of bobcats actually have four legs. The judgment would remain true even if most bobcats lose one of their legs in an accident. Thompson argues that natural-historical judgments have what we may call ‘normative implications’. They ground claims to natural defect by a principle of inference saying: from “the S is F”, and “this S is not F,” we can infer: “this S is defective in that it is not F” (Thompson, 80). For instance, given the natural-historical judgment above, we can infer that a bobcat without four legs is naturally defective and a

6 bobcat with four legs is, to that extent, naturally good. Thus, according to Thompson, the characteristic life of a life-form, as articulated in natural-historical judgments, sets the standard for making evaluations of natural goodness and defect. An important aspect of natural-historical judgments, which is related to their ability to underwrite evaluative inference, is that they have a teleological dimension. As Foot (2001, 30) also observes, not every form of generic statement about a life-form supports inference to evaluative judgments. “Blue tits have a round blue patch on their head” is a generic statement that is superficially similar to “male peacocks have a brightly-colored tail”. But, assuming that the blue tit’s color plays no role in the life of the bird, there is an important difference between these two statements. Unlike the second statement, the first one does not denote an aspect of the life-form that can support evaluative inference. A blue tit’s failure to have a blue patch on the head does not amount to a defect in the same way that a male peacock’s failure to have a brightly-colored tail does. Thus, the fact that natural-historical judgments can give rise to evaluations of natural goodness and defect is not merely in virtue of their being generic judgments, or having what Thompson calls ‘non-Fregean generality’ (Thompson, 77).

According to Thompson, natural-historical judgments are a specific subclass of generics that are—among other things—“teleologically articulable” (Thompson, 79). This means that they can be connected with other natural-historical judgments in teleological relations. It can be said of “male peacocks”, for instance, that they “have a brightly colored tail in order to attract mates”, whereas, in the case of blue tits, the blue patch on the head does not contribute to a further characteristic aspect of the life of the bird in the right way. Thus, the set of natural-historical judgments that capture the characteristic features and activities of a life-form are teleologically related, and together form a special unity or a system, which Thompson calls the natural history of the life-form. It’s this natural history that determines what it is to be a good or defective instance of one’s life-form. What constitutes flourishing for bearers of a life-form, then, is determined based on this natural history. On the neo-Aristotelian view, flourishing is the life-form-specific notion of an organism’s doing well

7 by being a good instance of its life-form, as characterized by natural-historical judgments. We can see that it’s this notion of flourishing that makes inference to evaluative judgments intelligible. It’s because a male peacock’s brightly-coloured tail plays a part in the flourishing of peacocks by attracting mates that failure to have such a tail amounts to a defect. In short, evaluations of natural goodness evaluate parts and aspects of living organisms based on their function in enabling the organism to flourish according to its life-form. The aim of the neo-Aristotelian project is to extend evaluations of natural goodness to the case of human beings and argue that our judgments of goodness and badness in humans— including judgments of moral evaluation—instantiate the same kind of ‘natural’ evaluation. What flourishing consists in differs from one kind of living thing to another. On the neo- Aristotelian view, human flourishing requires the capacity to recognize and act on reasons, where moral virtues constitute exercising this capacity well. Thus, moral evaluation of human character and action is construed as the same kind of flourishing-based evaluation as the natural evaluation of an oak’s roots or a peacock’s tail.

1.2 The Question of Ethical Naturalism

Foot (2001) presents her view as “a naturalistic theory of ethics” (Foot, 5), and contrasts it with Moore’s non-naturalism, various forms of non-cognitivism, and Kantianism. She claims that, rather than predicating a special kind of ‘non-natural’ property, moral judgments share a conceptual structure and status with familiar ‘natural’ judgments of goodness and defect in plants and animals. This presentation is a theme across different versions of the neo- Aristotelian view. Hursthouse (1999) similarly characterizes her view as a form of ethical naturalism. On her formulation, the neo-Aristotelian project is an attempt to ground ethics in considerations of human nature, where human beings are understood to be “part of the natural, biological order of living things” (Hursthouse, 205). Following this presentation, most commentators on neo-Aristotelian naturalism have characterized it as a form of ethical naturalism. Julia Annas (2005), for instance, claims that

8 although virtue ethics is not by definition naturalistic, Aristotelian forms of virtue ethics amount to a form of naturalism that “puts us in our biological place” and bring out continuities between us humans and other living things (Annas, 17). Echoing this view in their Stanford Encyclopedia of Philosophy entry on “Moral Naturalism”, Matthew Lutz and James Lenman (2018) describe neo-Aristotelian naturalism as the official metaethical position of many contemporary virtue ethicists and one of the leading versions of contemporary naturalism next to Cornell realism (Boyd 1988; Brink 1986; Sturgeon 1985; Railton 1986) and Frank Jackson’s (1998) moral functionalism. Yet, several critics have raised questions about whether and the extent to which neo- Aristotelian ethics fits within the general framework of naturalism. The core issue is that neither Foot nor Hursthouse consult scientific findings in their accounts they offer of natural goodness or human nature. Unlike other forms of ethical naturalism that purport to give an account of ethical claims in terms of facts that fall within the scope of natural science, neo- Aristotelian naturalism does not present ethical claims as having any relation to the domain of science. In fact, as we will see in more detail in the next chapter, neo-Aristotelian naturalists intentionally steer clear of scientific accounts of living things. They define natural goodness in terms of what is characteristic of a species, while rejecting that biology can offer an understanding of what is so characteristic. Instead, they appeal to a teleological conception of the nature of a species that seems largely out of touch with modern evolutionary biology. Foot, for instance, explicitly warns against mistaking the notion of function implicit in her account of natural goodness for the notion of a biological adaptation. She points out that, in talking about the function of a brightly-colored tail in the life of a peacock, she is concerned not with the “technical uses” of the term ‘function’ in evolutionary biology, but rather with the “everyday uses” of the term (Foot 2001, ft. 32). This is what some critics find puzzling in light of the neo-Aristotelians’ self-proclaimed commitment to naturalism. Instead of consulting scientific accounts of human nature, neo- Aristotelians rely on an essentialist and teleological conception of the nature of living things

9 that seems untenable in light of modern science. The critics thus appeal to a scientific, evolutionary account of living things to argue that neo-Aristotelians cannot offer an account of virtue that is both naturalistic and plausible (Andreou 2006, Millgram 2009; Woodcock 2009; Millum 2006; Fitzpatrick 2000; Lewens 2010; Odenbaugh 2017). They contend that an evolutionary understanding of human nature does not support the idea that substantial virtues like justice and benevolence are instances of natural goodness in human beings. The assumption underlying this criticism is that ethical naturalism requires an account of morality that is ultimately grounded in empirical science. The idea—as Jay Odenbaugh puts it—is that according to ethical naturalism, “moral properties are natural properties that can be studied by the sciences” (Odenbaugh 2017, 1048). And since neo-Aristotelians deny that what constitutes natural goodness in human beings can be studied scientifically, what they offer is not a form of ethical naturalism. The idea that ethical naturalism should be ultimately understood in terms of a relation to natural science is a prevalent idea in metaethics. In abstract terms, ethical naturalism is often defined as a conjunction of at least two claims: (1) moral realism—the thesis that in making moral judgments we purport to report moral facts, and that there are such facts to report; and (2) metaphysical naturalism—the thesis that all facts and properties are natural facts and properties. But philosophers have generally had a hard time articulating what metaphysical naturalism involves, or what the naturalness of a fact or property consists in. As Russ Shafer- Landau (2003) argues, we may try to understand the natural domain directly by identifying some distinctive feature that is intrinsically possessed by all and only natural properties. But any proposed such property seems to be either too narrow (e.g., being tangible or material, being causally efficacious) or too broad (e.g., being descriptive, being non-conventional).1 As a result, many philosophers approach metaphysical naturalism indirectly and understand it in disciplinary terms, i.e., in terms of the subject matters of scientific disciplines. They characterize natural facts and properties as the kinds of facts and properties that we investigate

1 For various attempts to characterize metaphysical naturalism directly, see also Copp (2003, 37-38).

10 using empirical methods of science, and understand metaphysical naturalism as the view that moral facts and properties are natural in this sense (see, e.g., Shafer-Landau 2003 and Enoch 2011).2 Now, if natural facts and properties are construed as the kinds of facts and properties that we investigate using the empirical methods of science, then ethical naturalism would necessarily involve a commitment to a third claim: (3) epistemic naturalism—the thesis that we know moral claims are true in the same way that we know about claims in the natural sciences. This, however, is not a commitment that neo-Aristotelians would accept. As we saw earlier, neo-Aristotelians argue that moral goodness is an instance of natural goodness, while rejecting that our understanding of natural goodness can come from an empirical science like biology. So if the natural domain is construed as the domain of natural science, neo- Aristotelians don’t seem to be claiming that moral facts and properties are natural at all. Because of this, some commentators have actually interpreted neo-Aristotelian naturalism as not making a claim to ethical naturalism. Jonathan Dancy (2006), for instance, claims that although neo-Aristotelian naturalism calls itself ‘naturalism’—because it holds that moral distinctions are tightly grounded in considerations of human nature, it takes no official stance on the debate between ethical naturalism and non-naturalism. However, in the next section, I argue against this interpretation by spelling out the sense in which neo-Aristotelian naturalism is in fact a form of ethical naturalism. I argue that although neo-Aristotelians characterize the natural domain differently—without making a reference to natural science—their account of morality is nonetheless a naturalist view in the standard, metaphysical sense.

2 Darwall, Gibbard, and Railton (1992) have also characterized ethical naturalism as a placing of ethics with respect to empirical science by offering some sort of methodological or substantive “assimilation”—as opposed to a establishing a contrast (126).

1.3 In Defense of a Naturalist Interpretation

In response to the objection that neo-Aristotelian naturalism is out of touch with modern evolutionary biology, several advocates of the view have tried to articulate what is naturalist about this view (Annas 2005; Hacker-Wright 2009; Thompson 2013; Frey 2018). What is common among these accounts is the formulation of neo-Aristotelian naturalism as the view that the concept human is the central concept of ethics, or the view that ethics is based on considerations of human nature. Initially, it might seem that this is naturalism in a different sense from other views in metaethics that are called ‘naturalism’. What is naturalist about the neo-Aristotelian account, it might seem, is the claim that morality concerns human nature. One might say that the view would be more appropriately called a naturist view. If all that neo- Aristotelians mean by naturalness is having to do with human nature, then their claim would be distinct from metaphysical naturalism, as they would not be concerned with whether morality belongs within the domain of natural facts and properties. This would be in line with Dancy’s interpretation, according to which “neo-Aristotelian naturalists could be ethical non- naturalists” (Dancy 2006, 122). However, it should be noted that the paradigmatic neo-Aristotelian view presented by Foot and Hursthouse also involves a second claim. As Hursthouse puts these two claims explicitly, neo-Aristotelians give an account of ethics in considerations of human nature, where human beings are understood to be “part of the natural, biological order of living things (1999, 205).3 She says, for instance, that this excludes the possibility that humans are understood as creatures with an immortal soul or as persons or rational agents. The question, then, is on what basis neo-Aristotelians claim to understand human beings as part of the natural order if they reject the idea that human nature can be understood scientifically.

3 It’s worth noting that the second claim is not shared by all versions of neo-Aristotelian naturalism. John McDowell’s ‘second-nature’ naturalism, for instance, only claims that morality is natural in the sense that it arises naturally in humans.

I propose that the neo-Aristotelians’ distinctive answer to this question can be found in their attempt to show that there is a continuity between human beings and other living things. The concept of human, which neo-Aristotelians claim to be central to ethics, denotes a life- form or a species. As Thompson points out, he uses the word ‘human’ in a way that is “on a level with words like ‘Norway rat’ and ‘coastal redwood’ (Thompson 2013, 701). Unlike the abstract concept of a person or a rational being, the concept of human is tied to actual material living things with a particular evolutionary history, and is characterized by certain physical features such as having two arms and two legs. Thus, when neo-Aristotelians talk about human nature, they are referring to the nature of human beings as living things—the nature that puts them on the same plane as plants and non-human animals. It is this continuity between human nature and the nature of non-human living things that grounds the neo- Aristotelian’s claim to metaphysical naturalism. The idea is that non-human sub-rational living things uncontroversially belong to the natural domain, and if the human life-form is continuous with these other life-forms, it must belong in the natural order as well. We can trace this continuity most clearly in Foot’s claim that moral evaluations “share a basic logical structure and status” with evaluations of plants and animals (Foot 2001, 27). As we saw in the previous section, Foot’s project is to identify a form of evaluation that applies to plants and animals and then to argue that moral evaluation is an instance of this same kind of evaluation. It is crucial to Foot’s account that evaluations of natural goodness simultaneously apply to the domain of non-human life and the human domain, including moral evaluation. This emphasis on sharedness of the form of evaluation can only be understood in the context of her attempt to give a naturalist account. Note that if, as Dancy suggests, the neo-Aristotelian account was in fact neutral between metaphysical naturalism and non-naturalism, it wouldn’t matter whether evaluations of non-human living things could be shown to have the same structure as moral evaluation, or whether we could make evaluations of non-human living things in virtue of their nature or their flourishing at all. The view would be merely an account of morality in terms of considerations of human nature without making a claim about whether

13 facts about human nature are natural. It would be compatible with such a view that flourishing-based evaluation arises with human beings due to their special, sui generis nature. But, as we have seen, neo-Aristotelians put a lot of emphasis on making sense of flourishing- based evaluations of non-human living things. They articulate an elaborate account of natural goodness, and argue that it applies to evaluations of non-human and human living things alike. The sharedness of the form of evaluation goes toward showing that moral evaluation of human will and character is not a special case. Thus, the continuity between evaluations of humans and non-humans is what ultimately underwrites the neo-Aristotelians’ claim to naturalism. 4 This interpretation is in line with Annas’ claim regarding what is distinctively naturalist about the Aristotelian brand of virtue ethics: that it “puts us in our biological place” by bringing out continuities between us humans and other living things. It shows that our ways of evaluating ourselves morally and otherwise are continuous with “evaluative patterns to be found in the lives of animals and plants” (Annas 2005, 17). A view according to which ethics is based on considerations of human nature but human beings are conceived purely in terms of the abstract notion of rationality would not be naturalist in this way. That is because on such a view, normativity “arises with humans or is produced by humans in a way which owes nothing to the nature which we share with other living things” (Annas, 12). While the abstract concept of a rational agent or a person would cover humans as well as imaginary rational Martians or even angels, it is markedly different and discontinuous from our concepts of non- human sub-rational animals. What is natural about the neo-Aristotelian concept of human is

4 My claim that the neo-Aristotelians’ focus on continuity should be understood as an attempt to defend naturalism echoes McPherson’s (2012) characterization of ethical naturalism and non-naturalism in terms of ‘metaphysical continuity’. In contrast to those who understand the distinction in terms of a difference in epistemology, McPherson gives a direct, metaphysical characterization of the distinction based on whether ethical properties are either reducible to or continuous with other types of property. On this account, the core conviction of ethical non-naturalism that distinguishes it from ethical naturalism is the claim that ethical properties are sui generis, in the sense that they are neither reducible to nor continuous with any other type of property.

14 that it captures a form of life, and the kind of evaluation that it underwrites is already present in the non-human, sub-rational domain of life that we take to be part of the natural order.5 Thus, the basis for naturalization here is not a scientific or value-neutral conception of human life-form, but rather an understanding of the human life-form as continuous with other life-forms, which are claimed to be value-laden in the same way. If the neo-Aristotelian project succeeds, the norms of moral goodness in human life are shown to be natural by showing that they are of the same kind as the norms of natural goodness in non-human life-forms. The assumption here, of course, is that—whatever naturalness consists in—there is no question that plants and non-human animals qua living things are part of the natural order. In fact, if one were to cast doubt on whether non-human living things should be considered natural—say, on physicalist grounds—then it would be fair to say that neo-Aristotelians do not make a claim to metaphysical naturalism in this more strict sense. Their naturalism only goes as far as claiming that moral goodness has the same status as evaluations that are grounded in the nature of non- human living things. However, note that it would be missing the point of their metaethical project to focus on metaphysical debates about whether the domain of life is natural or non- natural. What is interesting about the neo-Aristotelian view is that it gives an account of the

5 In fact, one of the biggest challenges that a later generation of neo-Aristotelians has tried to address is making sense of this continuity despite significant differences that seem to remain between moral evaluations of humans and flourishing-based evaluations of non-humans (Lott 2014; Haase 2018). Foot herself points out that there is a sense in which “human good is sui generis”—a sense that she interprets as having to do with the fact that human flourishing goes beyond self-maintenance and reproduction. Matthias Haase (2018) gives a more in-depth account of what is different about evaluations of human will and character, and locates the difference in the practicality of the latter. Ethical evaluation is action-guiding in a way that merely theoretical evaluations of natural goodness and defect are not. According to Haase, due to the practicality of ethical evaluation, there is an apprehension requirement on ethical goodness, which means that when we evaluate human action as good, we imply that the subject has an understanding or knowledge of what is good to do. This suggests that ethical evaluation—at least prima facie—might have a different logical structure than evaluations of natural goodness in non-humans. Haase characterizes this as a central difficulty for the neo-Aristotelian project, and tries to lay out the shape that a neo- Aristotelian account would have to take to address this difficulty. Although this is not a problem that I aim to solve in this dissertation, I consider the fact that neo-Aristotelians take this problem seriously as further evidence that they are trying to offer a naturalist account.

15 supposedly special and problematic domain of morality in terms of the familiar facts of life— facts that we have no problem understanding or accepting as real.

1.4 The Neo-Aristotelian Argument for Naturalism and Its Limitations

I have argued that the neo-Aristotelian view is a form of ethical naturalism despite not relying on a scientific characterization of what is natural. What is at issue with the question of ethical naturalism is whether we can show that the domain of morality, with all of its differences and peculiarities, fits within our picture of the natural world. And we have seen that this is exactly what neo-Aristotelian naturalists try to show. They give an account of the moral domain in terms of the familiar domain of life, in order to show that we don’t have to go any further than our intuitive understanding of the nature of life to look for a grounding for morality. What they offer, then, is an instance of ethical naturalism that involves commitments to both moral realism and metaphysical naturalism. One of the reasons this matters is that showing how moral truths fit within the natural world helps to defend morality against certain skeptical arguments. For instance, as Tristram

McPherson (2015) has pointed out, a core strategy of the moral skeptic is to argue that if morality did exist, it would be non-natural, which is to say that it would be inconsistent with a global naturalism—the thesis that everything actual is natural (McPherson, ft. 14). Ethical naturalism offers a defense of morality against this skeptical argument by explaining how the existence of moral truths is consistent with a naturalistic view of the world. But different versions of ethical naturalism can explain this in different ways. And we have seen that neo- Aristotelian naturalism in particular explains how morality fits within the natural world by arguing that it is part of the natural domain of life. Thus, what is distinctive about neo- Aristotelian naturalism is not that it offers an entirely different sense of ethical naturalism (such as what we might call a mere naturism), but that it presents a different argument for ethical naturalism. While most versions of ethical naturalism give an account of the moral domain as reducible or identical to a domain that can be understood in terms of explanatory

16 framework of science, neo-Aristotelian naturalism gives an account of this domain in terms of the natural domain of life. That said, even though the neo-Aristotelians’ lack of offering a scientific grounding does not automatically preclude them from having a claim to ethical naturalism, it does limit the scope and strength of their argument for this claim. Giving an account of the moral domain in terms of our grasp of the more familiar domain of life helps to place it in the fabric of the world by showing that the former is no more mysterious than the latter. But note that there is a question to what extent our commonsense grasp of the domain of life can be trusted. In fact, it’s not clear that our intuitive belief in there being objective moral standards is any weaker than our intuitive grasp of the domain of life, particularly where the neo-Aristotelian, teleological conception of non-human life is concerned.6 Thus, there is a limit to how much the neo-Aristotelian account—in its standard formulation—can contribute toward offering a grounding for the domain of morality and its status as objective and real. As we will see in the next chapter, the evolutionary objection to neo-Aristotelian naturalism questions whether the view really offers such a satisfactory grounding by challenging the neo-Aristotelians’ commitments regarding the domain of life. In other words, the neo-Aristotelians’ departure from a scientific understanding of the domain of life leaves a gap in their argument for ethical naturalism, which needs to be filled if this argument is to succeed. The rest of this dissertation is attempt to fill this gap in the neo-Aristotelian argument for ethical naturalism by consulting the science of biology about the nature of life and tracing the consequences of this approach for the overall shape of the neo-Aristotelian account. In Chapter 2, I discuss the evolutionary objection to neo-Aristotelian naturalism, and give a more nuanced account of the neo-Aristotelian argument for naturalism in response to this objection. Assessing this argument in depth, I argue that even the neo-Aristotelian brand of naturalism needs to look at science to justify some of its claims about the non-human domain of life. Then,

6 The concern here is not that facts about non-human life are somehow non-natural, but that what neo- Aristotelians claim to be objective facts about non-human life are not objective facts.

17 in Chapters 3 and 4, I argue that the neo-Aristotelians’ reservations about consulting the science of biology is based on a false conception of what biology tells us about the nature of life. I offer a novel argument for the neo-Aristotelians’ claim to ethical naturalism by appealing to recent approaches in philosophy of biology, particularly by looking at the evolutionary- developmental concept of organism and the organizational account of biological function. Finally, in Chapters 5 and 6, I argue that paying a closer attention to what biology tells us about living things urges us to revise the standard neo-Aristotelian account. Drawing from philosophical accounts of scientific explanation in biology, I offer a novel account of natural goodness, and argue that this novel account can address some of the concerns that arise in relation to Foot and Thompson’s natural-historical approach to understanding natural goodness.

Chapter 2

The Evolutionary Objection to Neo-Aristotelian Naturalism

Two influential lines of criticism have appealed to an evolutionary understanding of human nature and natural teleology to argue against neo-Aristotelian naturalism. In this chapter, I offer a reconstruction of these two seemingly different lines of criticism as raising instances of the same dilemma, giving neo-Aristotelians a choice between abandoning ethical naturalism and contradicting our considered moral judgment. I draw on the previous chapter’s characterization of neo-Aristotelian naturalism as ethical naturalism to argue that resolving the dilemma requires showing a particular kind of continuity between the norms of moral virtue and norms that are necessary for understanding non-human living things. I also argue that in order to show such a continuity, neo-Aristotelians need to revise the relationship they adopt with empirical science and acknowledge that the latter is relevant to assessing their central commitments regarding living things.

2.1 An Evolutionary-Inspired Dilemma

As we saw in the previous chapter, the central claim of neo-Aristotelian naturalism is that moral virtue is an instance of natural goodness in human beings, where natural goodness is a sort of goodness found in nature among living things. Exploring the critical literature on neo- Aristotelian naturalism, we can recognize two lines of objection that appeal to evolutionary biology to challenge this claim. The first line of objection appeals to an evolutionary understanding of human nature to question whether substantial virtues like justice and benevolence are instances of natural goodness in human beings (Andreou 2006, Millgram 2009; Woodcock 2009; Millum 2006; Odenbaugh 2017). The second line of objection appeals to an evolutionary account of the concept of biological function to undermine the ‘flourishing-based’ concept of function that underlies the neo-Aristotelian concept of natural goodness (Fitzpatrick

2000; Lewens 2010; Odenbaugh 2017). Most critics only focus on one of these objections, and neo-Aristotelians in their responses deal with them separately. But I will argue that these objections share a basic structure and ultimately raise the same problem—a problem that is best understood as a dilemma. The dilemma can be articulated by identifying two desiderata for the neo-Aristotelian project of offering a naturalist account of ethics. There are two parts to the central claim of neo-Aristotelian naturalism mentioned above, which are both essential for the project to succeed: that moral virtue is an instance of what neo-Aristotelians call ‘natural goodness’ in human beings, and that the latter is in fact an aspect of the natural world. Thus, we can interpret neo-Aristotelian naturalism as offering the following argument for ethical naturalism:

Simple Argument for Naturalism

1. Norms of natural goodness are natural. (Naturalness) 2. Norms of moral goodness are instances of norms of natural goodness in

human beings. (Normative Adequacy) C. Therefore, norms of moral goodness are natural.

Let’s look at each of the premises of this argument in turn. The first premise, which I have called Naturalness, states that norms of natural goodness are natural. As we saw in the last chapter, different philosophers have different ways of arguing that a fact or property is natural, and neo-Aristotelians have a distinct way of going about it. In this chapter, we will get a clearer picture of how neo-Aristotelians understand the naturalness of the evaluations that they call ‘natural goodness’. But for now, let’s simply interpret the premise of Naturalness as claiming that norms of natural goodness are part of the natural order. The second premise, which I have called Normative Adequacy, states that norms of moral goodness are instances of norms of natural goodness in human beings. In other words, the premise states that the notion of natural goodness is normatively adequate for capturing the norms of morality, such that the latter can be understood as instances of natural goodness.

Once we reconstruct the neo-Aristotelian argument for ethical naturalism in this way, it becomes apparent that there is a tension between the two premises of the argument. Natural goodness should be defined such that norms of natural goodness are natural, but also such that they coincide with norms of moral goodness in the human case. Each of these requirements seems to pull in a different direction. Even without a full grasp on what it takes to secure each of these premises, we can see that they are difficult to secure at the same time. A straightforward way to ensure Naturalness would be to define natural goodness in terms of biological function or adaptiveness in a strictly evolutionary sense, since evolutionary function is uncontroversially an aspect of the natural world. But it would be hard to expect such evolutionary norms to coincide with our considered judgment about moral virtue and fulfill Normative Adequacy. On the other hand, an easy way to ensure Normative Adequacy would be to stipulate norms of natural goodness such that they coincide with our considered judgment about moral norms. But then it would be far from clear that such stipulated norms meet the requirement of Naturalness. Thus, neo-Aristotelians seem to be facing a difficult choice between abandoning ethical naturalism and contradicting our considered moral judgment. This is how the tension between the two desiderata of neo-Aristotelian naturalism seems to give rise to a dilemma. Although the two lines of evolutionary objection in the literature are not usually presented in this way, I suggest that they both raise instances of this dilemma. The first objection, which is often called the Pollyanna Problem, focuses on natural goodness in human beings. Elijah Milgram (2009), Chrisoula Andreou (2006), and other critics who raise this objection appeal to a scientific understanding of human nature to argue that it is naively optimistic or ‘Polyannish’ to assume that the extension of natural goodness in humans coincides with our considered judgment about moral virtue. The second objection does not share this focus on human beings but has a similar structure. It relies on an understanding of natural goodness in terms of evolutionary functions to question whether the extension of natural goodness in general coincides with our considered judgment about a living thing’s

21 flourishing. William Fitzpatrick (2000) who raises the most elaborate version of this objection primarily focuses on natural goodness in non-humans. Since his argument appeals to Richard Dawkins’ account of evolution in The Selfish Gene (1976), I will call the problem that he raises the Selfish Gene Problem. Despite the similar structure of the two objections, they are often seen as posing very different questions (see, e.g., how they are treated separately in Lott 2012a and 2012b). The Pollyanna Problem is interpreted as granting the neo-Aristotelian account of natural goodness (the premise of Naturalness) and only questioning the attempt to show that moral goodness is an instance of natural goodness (the premise of Normative Adequacy). The Selfish Gene Problem, on the other hand, is interpreted as questioning the neo-Aristotelian account of natural goodness (the premise of Naturalness) in its own right. In the following sections, I argue that both of these characterizations are inaccurate. The Pollyanna Problem cannot be interpreted as granting the neo-Aristotelian account of natural goodness, because the critics’ apparent misrepresentation of the neo-Aristotelian account has to be seen in the context of their attempt to start from a properly naturalist account of natural goodness. Nor can the

Selfish Gene criticism of natural goodness be understood apart from the supposed relation of natural goodness to moral goodness. Instead, both objections are instances of the dilemma articulated above. They start from a scientifically respectable interpretation of natural goodness to raise problems having to do with normative adequacy. And although they raise the problem at different levels, what is ultimately at issue in both cases is the tension between the two premises in the neo-Aristotelians’ argument for ethical naturalism.

2.2 The Pollyanna Problem

Millgram (2009), Andreou (2006), Woodcock (2006), Millum (2006), and Odenbaugh (2015) raise the Pollyanna Problem, which concerns evaluations of natural goodness in human beings. These critics appeal to evolutionary psychology or other empirical sciences for insight into what would be naturally good in in the life of humans, and then point to the objectionable

22 implications of understanding morality in those terms as a reductio of neo-Aristotelian naturalism. What happens if we consult an evolutionary scientific understanding of human nature and try to derive moral virtues from such an understanding? Critics give many examples of problematic results that could follow. Instead of virtues like justice and benevolence, morally objectionable traits like infanticide, rape, and xenophobia may turn out to belong in the life of the human species. Millgram (2009, 561-562), for instance, cites empirical research suggesting that “humans value occupying dominant positions in hierarchies to a degree not compatible with justice of any kind”, or that “human males are fine-tuned by natural selection to rape women in a suitable range of circumstances”. Andreou (2006) similarly argues that sociobiologists seek and find plausible survival-and-reproduction-related functions not only for “nice” phenomena, like maternal love, but also for “nasty” phenomena like sex-selective infanticide by mothers. Woodcock (2006) adds that models of human cooperation reveal that altruistic behavior in humans only exists within significant limits and with considerable side effects. He claims that the most effective mechanisms to prevent free-riders from invading groups of altruists involve xenophobic dispositions and forms of prejudice against people outside one’s interacting group, which are morally objectionable.7 In short, the objection is that given what evolutionary science tells us about human nature, natural goodness in humans does not coincide with our considered judgment about moral virtue. Neo-Aristotelian naturalism thus seems to offer a naturalist account of moral virtue only at the expense of a significant revision of our substantial conception of virtue—a revision that neither neo-Aristotelians nor their opponents are willing to embrace. To put this in terms of an objection to the Simple Argument for Naturalism, the idea is that understanding natural goodness in terms of evolutionary adaptiveness secures Naturalness, but empirical research suggests that it fails at Normative Adequacy.

7 The empirical research cited by these authors includes Frank (1985); Thornhill and Palmer (2000); Hrdy (1999); Hirshleifer and Rasmusen (1989); and Wilson and Dugatkin (1997).

One could respond to this objection by questioning the validity of the empirical research cited by critics. Does evolutionary psychology really show rape and infanticide to be adaptive for human beings? Some of the empirical claims made above are in fact called into question by scholars from various disciplines (see Lloyd 2001; Travis 2003; Kitcher and Vickers 2003). However, this line of response is not fully effective. Even if the cited research fails to show human nature to be at odds with virtues like justice and benevolence, it seems extremely unlikely that better empirical research would deliver results that are in line with our considered moral judgment. Note that our considered moral judgment is formed independently of this kind of empirical research. In fact, the mere possibility that scientific findings could refute our conception of virtue seems implausible. We do not typically consider biological premises to have a privileged position in justifying moral claims. We take the value of justice and benevolence to be independent of the evolutionary benefits they may or may not confer. A view that leaves it up to biology to decide what is morally good already fails to capture this second-order moral judgment. This is why neo-Aristotelians do not respond to the Pollyanna Problem by questioning the validity of the empirical research cited but rather by rejecting its relevance to understanding human natural goodness. Two recent advocates of neo-Aristotelian naturalism, Micah Lott (2012b) and John Hacker-Wright (2009b; 2013) have argued that our understanding of what is naturally good in the life of humans cannot come from empirical science. Lott and Hacker-Wright appeal to the place of practical reason in human life to make this case. Lott (2012b), who offers a more elaborate argument, argues that natural goodness concerns the characteristic way of living and achieving natural ends in a given life-form, and importantly, the human characteristic way of living and achieving natural ends is the way of practical reason. Thus, a proper understanding of human natural goodness involves an understanding of practical reason. However, according to Lott, an understanding of practical reason cannot come from empirical science, because such an understanding can only come from acquiring practical wisdom and ultimately moral virtue. To see this, note that a sound grasp of human

24 practical reason involves substantive judgments regarding what makes a human action rationally justified and what counts as a good reason for action. But this in turn requires a proper sensitivity to considerations about various human ends and values. Lott argues that to the extent that moral virtues characterize sound human practical reason, this is exactly the kind of sensitivity that distinguishes the virtuous person from the non-virtuous. Only the virtuous person can have a full understanding of what she ought to do in any given situation, and this understanding comes through her disposition toward virtuous actions and away from vicious ones. This is why, in Lott’s view, understanding the practical dimension of human flourishing falls outside the scope of empirical science. It is an understanding that is possessed by someone “not qua scientist but qua practically wise person” (Lott 2012b, 409). In a similar vein, Hacker-Wright (2009b; 2013) argues that fully understanding human natural goodness requires a normatively-laden interpretation of our own form of life that cannot be obtained from a biological account of the species Homo sapiens. That is because we can only make such a rational self-interpretation in view of our personhood and our nature as practically reasoning creatures. Thus, Lott and Hacker-Wright both argue that the empirical findings of evolutionary psychology do not give us an account of human natural goodness, and cannot be used to derive conclusions about moral virtue Now, this response clearly helps with showing the normative adequacy of human natural goodness. If our conception of what is naturally good in human life is formed through practical reason, the result is likely to be in line with our considered judgment about moral virtues, which is also acquired through practical reason.8 But if the neo-Aristotelian account of human natural goodness is informed by an understanding of practical reason, it is no longer obvious that it is a naturalist account. Note that the norms of practical reason aren’t any more obviously natural than the norms of moral goodness. Because of this, neo-Aristotelians owe

8 Note that on the neo-Aristotelian conception of practical reason, rationality and morality are not at odds but on the same footing. Practical reason involves not just self-interested considerations but moral considerations as well.

25 their opponents a defense, or at least a clarification, of why we should take human natural goodness to be natural. However, Lott (2012b)—who discusses the Pollyanna Problem in length—does not interpret this objection as targeting the neo-Aristotelians’ claim to naturalism. He rather treats it as an epistemological challenge. He tries to show that although empirical science does not give us knowledge of the human life-form, we have other means of acquiring such knowledge. So he argues that aspects of the human life-form that are not investigated by empirical science can be known through acquiring virtue, because a virtuous person possesses practical knowledge (see Lott 2012b, 423). However, this focus on epistemology is misguided. The question raised by the Pollyanna Problem is not whether one can have knowledge of what neo- Aristotelians are calling ‘natural goodness’, but whether the object of such knowledge is an aspect of the natural world. The critics who raise the Pollyanna Problem are not moral skeptics. Of course if the norms of natural goodness in human life are stipulated such that they coincide with moral virtue, the critics would not deny that a virtuous person has knowledge of these norms.9 But the question remains whether these norms are natural, especially given that our knowledge of them is not empirical knowledge. In short, the neo-Aristotelian response to Pollyanna, which involves rejecting the relevance of empirical findings to our understanding of human natural goodness, saves the premise of Normative Adequacy, but fails to establish the premise of Naturalness.

2.3 Naturalness as Continuity with Nature

The lesson to draw from the Pollyanna objection is that neo-Aristotelian naturalism cannot be interpreted as deriving moral virtues from a scientific, virtue-neutral, account of human nature.10 But if what we take to be naturally good in human life is already informed by an

9 Note that if the critics were moral skeptics, Lott’s argument which starts from the assumption that there are in fact virtuous people would not convince them. 10 See Lott (2012a, 420) for a rejection of a ‘two-stage’ reading of neo-Aristotelian naturalism.

26 understanding of moral virtue, the question remains why tying moral goodness to natural goodness is supposed to amount to an argument for ethical naturalism. Why are the norms of natural goodness in human life considered to be natural in the first place? This is where my account of what is distinctive about the neo-Aristotelian form of ethical naturalism—offered in the previous chapter—comes in. As we saw earlier, neo-Aristotelians attempt to show that what they characterize as human nature is part of the natural world by appealing to the continuity between human beings and other living things. The idea, which I take to be at the core of Foot’s appeal to the concept of natural goodness, is that the human life-form and the norms of natural goodness that it underwrites are continuous with the non- human domain of life, which is uncontroversially natural. Moral evaluations “share a basic logical structure and status” with evaluations of plants and animals (Foot 2001, 27), and to the extent that the latter are natural, so are the former. Thus, the basis for the claim to naturalism is not having a value-neutral conception of the human life-form, but rather the idea that norms of natural goodness in human life are of the same kind as the natural norms in non-human life- forms.

There are two important claims being made here. One says that there is a continuity between natural goodness in humans and natural goodness in non-humans, and the other says that the latter is natural. Integrating these claims into our Simple Argument for Naturalism gives us the following, more refined version:

Refined Argument for Naturalism

1. Norms of natural goodness in non-human life-forms are natural. (Grounding) 2. If norms of natural goodness in non-human life-forms are natural, then norms of

natural goodness in the human life-form are also natural. (Continuity) 3. Norms of moral goodness are instances of norms of natural goodness in the

human life-form. (Normative Adequacy) C. Therefore, norms of moral goodness are natural.

Note that the last premise of the argument has not changed. It still involves what I characterized as the premise of Normative Adequacy. But now two other premises have replaced the premise of Naturalness in the previous version of the argument. The first premise, which I have called Grounding, concerns the norms of natural goodness in non-human life- forms in particular, and states that these norms are natural. The second premise, which I have called Continuity, states that if norms of natural goodness in non-humans are natural, norms of natural goodness in humans are natural as well. Together with the third premise—that moral goodness is an instance of natural goodness in humans—these premises support the thesis that moral goodness is natural. We saw that Neo-Aristotelians secure the premise of Normative Adequacy by arguing that understanding what is naturally good in human life involves acquiring moral virtue. But now we have two new premises to worry about. As it turns out, these two premises introduce a challenge comparable to the dilemma we faced earlier between Naturalness and Normative Adequacy. To secure both of these premises, the norms of natural goodness in non-human life should be characterized such that they are natural, but also such that they are continuous with the norms of natural goodness in human life—i.e., norms that include moral norms. It seems that a version of the old dilemma has reappeared. We can ensure Grounding by characterizing non-human natural goodness in terms of biological functions or adaptiveness. But there would be a question whether such evolutionary norms can be seen as continuous with the relevant set of norms in human life. On the other hand, we can stipulate norms of natural goodness in non-human life such that Continuity is secured, but it would be far from clear that such stipulated norms would be natural.

Interestingly, this brings us to the second line of evolutionary objection against neo- Aristotelian naturalism, which primarily focuses on natural goodness in non-human life-forms (Fitzpatrick 2000; Lewens 2010; Odenbaugh 2017). Here I will focus on the argument offered by Fitzpatrick (2000), which I take to be the most fully developed version of the objection. As we will see below, Fitzpatrick offers an evolutionary account of biological function to argue that

28 biological functions do not coincide with an organism’s flourishing or welfare. I argue that this objection should be understood as arguing that we cannot secure the premise of Grounding by understanding natural goodness in terms of biological functions, because doing so will undermine the premise of Continuity. In the next section, I first present the objection as an instance of the dilemma that forces us to choose between these two premises, and then assess the neo-Aristotelians’ response.

2.4 The Selfish Gene Problem

The objection that I am calling the Selfish Gene Problem is posed as a general problem for the neo-Aristotelian account of natural goodness without particularly focusing on the case of human beings. We saw, in Chapter 1, that Foot characterizes natural goodness in terms of a notion of function that is based on a trait’s characteristic contribution to the organism’s flourishing, where flourishing is understood in terms of natural-historical judgments. On this account, when we say deep roots are naturally good in an oak, we are evaluating the roots in relation to their function in enabling the oak to flourish as an oak, i.e., to live a characteristic oak life. Because of this, the neo-Aristotelian account involves a functional commitment. It is committed to the idea that we can ascribe a function to the parts and aspects of living things that is determined based on the concept of flourishing of the organism. It is this ‘flourishing- based’ account of function that is the focus of Fitzpatrick’s objection. Fitzpatrick gives an evolutionary account of biological function in terms of gene replication, and argues that contrary to what neo-Aristotelians claim, biological functions have nothing to do with an organism’s welfare or flourishing.11 This objection is often taken to be very different from the Pollyanna Problem and attacking natural goodness independently of its relation to moral

11 Fitzpatrick uses “welfare” interchangeably with “flourishing”, which on the neo-Aristotelian view is the life- form-specific notion of an organism’s doing well by being a good instance of its kind. The notion of welfare at issue, then, is not to be interpreted as the subjective notion of well-being in the sense of interest-satisfaction, which plausibly is restricted to sentient organisms.

29 goodness. But here I argue that it is ultimately raising the same dilemma except at a different level. Fitzpatrick’s core idea is that we need to view living things in light of their being products of evolution in order to get a grasp on their functional aspects. He appeals to a genocentric understanding of evolution to develop his evolutionary account of biological function. This view of evolution, which Fitzpatrick adopts from philosophers of biology like Dawkins (1983), and Sterelny and Kitcher (1988), puts genes at the centre of the main processes of evolutionary change. Natural selection is seen as resulting from genes increasing their frequency in the gene pool by exerting phenotypic effects in organisms such that these effects ultimately serve to promote the propagation of the genes. Based on this understanding of natural selection, Fitzpatrick argues that the ultimate function that natural selection has devised for organisms and their parts and features is to increase the frequency of their genes. His thus offers an account of biological function, according to which for an entity to have a biological function is for it to play a non-accidental role in promoting the “ultimate biological end” of replication of its genes (2000, 103-104).

Fitzpatrick argues that once we understand biological functions in this way, we will see that they do not always promote the welfare of an organism. There is no reason to think that gene replication, which is a blind evolutionary force, always promotes organismic welfare, at least according to our intuitive conception of what an organism’s welfare consists in. Fitzpatrick gives many examples of traits that have a biological function but nonetheless do not seem to promote the organism’s welfare. One illustrative example is the case of violent fights in elephant seals. Male elephant seals fight with each other in competition for exclusive control of the harem, and female seals refuse to mate with anyone except the dominant male. Fitzpatrick argues that the violent fights among male elephant seals have a biological function. Looking at evolutionary history reveals that the genes that tend to produce this violent behavior have come to dominate the gene pool by out-propagating rival alleles, because male seals that engage in such fights tend to out-reproduce their peers. However, although this

30 behavior is in line with evolutionary functioning, it often results in the seals’ injury and sometimes even death, which does not seem to be in line with the seals’ welfare in any plausible sense. As Fitzpatrick illustrates with this example and many others, there is no reason to think that what contributes to gene replication always promotes the organism’s welfare or flourishing. This is why, on Fitzpatrick’s account, ascribing biological functions based on contribution to flourishing would be wrong. Note that this, in turn, means that we cannot secure the premise of Grounding by understanding natural goodness in terms of biological functions. One might respond that the neo-Aristotelian notion of flourishing is different from the subjective notion of well-being, and concerns the organism’s doing well by being a good instance of its kind. So it doesn’t necessarily follow from the fact that elephant seal fights involve pain or injury that they cannot be conducive to the flourishing of the organism— especially because reproduction is an aspect of the seals’ flourishing that is promoted by the fights.12 However, even though Fitzpatrick appeals to what he calls “our ordinary, if somewhat fuzzy, conception of organismic welfare” (2000, 69) to make this example illustrative, his overall point does not depend on how exactly we understand the flourishing of elephant seals. Fitzpatrick’s overall point is that biological functions are determined based on gene replication and independently of the organism’s flourishing. So unless we define flourishing in ultimately genetic terms, there is no reason to think that biological functions promote flourishing.13 And of course we cannot define flourishing in genetic terms without violating the premise of Continuity.14 To maintain Continuity, the same concept of flourishing that is the basis of

12 Lott (2012a) actually raises this point in his response to Fitzpatrick. 13 For a more compelling case, see his discussion of the battle over sex ratio between the queen and sterile female workers in ant colonies—a behavior that has an evolutionary function without even increasing the reproductive output of the individual or the colony (Fitzpatrick 2000, 74-77). 14 Fitzpatrick gives the impression that the problem with defining welfare in terms of gene replication is that it would be “a radical departure from intuitive notions of organismic welfare or well-being” (68). But it’s important to clarify that it is not our intuitions regarding the flourishing of animals like elephant seals that keep us from defining flourishing in genetic terms. The reason has to do with the requirement of Continuity.

31 evaluations of natural goodness in non-humans should be the basis of evaluations of natural goodness in human beings. But evaluations of natural goodness in human beings include evaluations of moral goodness, and we have seen that understanding those in genetic terms would violate Normative Adequacy. This is why neo-Aristotelians define flourishing in terms of natural history and not in relation to gene replication. This is how the Selfish Gene Problem targets the tension between the premises of the Refined Argument for Naturalism. If we understand natural goodness in non-humans in terms of biological functions, the premise of Grounding will be uncontroversially secured. But if, as Fitzpatrick argues, biological functions do not concern any suitable conception of flourishing, then evaluations of natural goodness in non-humans will not be flourishing-based evaluations, which means that Continuity will be compromised. The structure of the Selfish Gene Problem is therefore similar to the Pollyanna Problem. It relies on a scientific account of natural goodness to question whether natural goodness can be tied to flourishing and ultimately moral virtue. In response to Fitzpatrick’s objection, Neo-Aristotelians reject the idea that that their claims about natural goodness have anything to do with the concept of function in biology. They argue that their account of natural goodness is based on a different concept of function that is not intended to capture the biologists’ talk of functions. Hacker-Wright (2009) and Lott (2012a) have both responded to Fitzpatrick along these lines, bringing out the fact that Foot herself explicitly says that the type of function she has in mind is distinct from a biological adaptation (see Foot 2001, 32). Lott defends the neo-Aristotelian account of function qua an account of function, arguing that it meets the desiderata of a successful account of function. One of Fitzpatrick’s charges against the flourishing-based account is that because it ignores the history of a trait, it cannot make sense of the distinction between a genuine end served by a function and an accidental benefit (see Fitzpatrick 2000, 185-207). But Lott rightly argues that although the neo- Aristotelian, flourishing-based account of function does not draw this distinction based on

32 evolutionary history, it does draw the distinction on another basis, namely the characteristic life cycle of the life-form (see Lott 2012a, 367-374). In other words, not just anything that contributes to an organism’s natural ends is considered functional on the neo-Aristotelian account. A feature or behavior is considered functional if and only if it plays a part in the characteristic life cycle of the organism’s life-form. However, showing that the flourishing-based account gives a conception of function that is distinct from the evolutionary conception of function is not sufficient to solve the Selfish Gene Problem. It only addresses one horn of the dilemma by securing the premise of Continuity. Surely if natural goodness for plants and animals is characterized in terms of an account of function that is defined based on a suitable conception of flourishing, the continuity across the human and non-human natural goodness is preserved. But are these supposed flourishing-based functions natural? Although Fitzpatrick does not explicitly discuss the requirement of Grounding, it is clear that trying to understand natural goodness as a natural form of evaluation is in the background of his appeal to evolutionary functions. Thus, the neo-Aristotelian response to the Selfish Gene Problem meets the requirement of Continuity by arguing that the flourishing-based conception of function is distinct from the evolutionary conception. But in order to ensure Grounding, neo-Aristotelians need to show that the flourishing-based functions that they ascribe to the parts and aspects of living things are natural. In the next section, I discuss what I take to be the best strategy that is available to neo-Aristotelians for making such an argument.

2.5 Are Flourishing-Based Functions Natural?

How can neo-Aristotelians argue that flourishing-based functions are in fact natural? In the philosophical literature on the concept of function in biology, naturalizing functions is often taken to involve identifying the necessary and sufficient conditions for functional ascription in reductive causal terms. Etiological theories of biological function, which are taken as a successful instance of doing this, identify these conditions in terms of the causal history of a

33 trait (see Wright 1973; Millikan 1989; Neander 1991). Fitzpatrick’s account similarly specifies these conditions in reductive causal terms, particularly in terms of contribution to gene replication.15 In rejecting the flourishing-based account of function, Fitzpatrick seems to assume that specifying the ascription conditions in causal terms is required for the account of function to be naturalistic.16 However, neo-Aristotelians do not seem to be concerned with offering a naturalized account of function in this strict reductive sense.17 They characterize conditions for functional ascription in terms of the concept of flourishing while admitting that what counts as the flourishing of an organism or characteristic of its life-form is itself understood in functional terms. As we saw in Chapter 1, an organism’s flourishing is given in a system of natural-historical judgments, where natural-historical judgments are generic judgments about a life-form that have a teleological dimension and express the characteristic features that “play a part” in the life of that kind of organism. Thus, neo-Aristotelians do not try to specify the conditions for flourishing-based functional ascription in reductive terms (see Lott 2012a, 371-372 for an explicit discussion), and yet they take these functions to be natural. Note, however, that as long as we are dealing with the premise of Grounding, we are only concerned with functional ascriptions in the case of plants and non-human animals, i.e., organisms that we can plausibly take to be part of the natural order. And given that plants and non-human animals are part of the natural order, if there are norms that are grounded in their nature, we can take these norms to be natural as well. Because of this, the flourishing-based functions that neo-Aristotelians ascribe to the parts and aspects of these organisms would be natural if they are grounded in the nature of these organisms as living things.

15 Note that Fitzpatrick denies that his account of function is a standard etiological theory (229-246). 16 Odenbaugh (2017) makes the same assumption when he claims that the etiological account of function is “the only good theory we have of normative natural functions”. He goes as far as claiming that because the neo- Aristotelian account of function is not reducible to our best scientific accounts of functions it is not a naturalistic theory, but a form of vitalism. 17 As we will see later in chapter 4, many philosophers of biology also argue against the reductive approach to understanding biological functions.

That said, not just any notion of function that can be applied to the parts and aspects of living things meets this condition. I could, for instance, ascribe a function to elephants’ tusks based on their use in making ivory for human art and manufacturing. But this function would be ultimately grounded in human ends. If I use this account of function to evaluate an elephant’s tusks as malfunctional on the basis that it is too dense to polish, this would be a standard that is imposed on the elephant tusks from the outside. It would not be grounded in the nature of the elephant as a living thing. The question, then, is whether the neo- Aristotelian, flourishing-based ascriptions of function to the parts and aspects of non-human organisms meet this condition—whether they are based on the nature of plants and animals as living things. Neo-Aristotelians might, for instance, ascribe the function of defence or digging to an elephant’s tusks because this is how the tusks characteristically contribute to the natural history of the elephant life-form. But is this function ascription really based on the nature of elephants as living things or is it something they are imposing on elephants from the outside? Does what neo-Aristotelians describe as the characteristic flourishing of an organism really capture their nature? What threatens the premise of Grounding, then, is not the idea that the characteristic flourishing of an elephant or the flourishing-based function of its tusks are not reducible in causal terms. It is rather the question of whether, as a matter of fact, the nature of an elephant involves having a characteristic flourishing—in the neo-Aristotelian sense—at all. It may be thought that the picture neo-Aristotelians draw of flourishing-based evaluations in the life of plants and animals is just too plausible to deny. Foot says, for instance, that “nobody would . . . take it as other than a plain matter of fact that there is something wrong with the hearing of a gull that cannot distinguish the cry of its own chick, as with the sight of an owl that cannot see in the dark” (2001, 24). However, this intuitive image is challenged by the Selfish Gene objection. A critic like Fitzpatrick would argue that such appearances are explained away by evolutionary biology, which reveals that we can understand living things and their functional aspects without making any reference to the concept of flourishing. On this view, organisms are nothing but gene-replicating machines

35 without a flourishing characteristic of their life-form. Even though we sometimes make evaluations of their parts and aspects, such evaluations are only correct when they line up with evolutionary contribution to gene-replication. In short, the critic would insist that biology is the branch of science that studies living things, and if the neo-Aristotelian, flourishing-based concept of function has no place in biology, we should question whether this concept is essential to our understanding of living things. But if this concept proves to be merely accidental to the understanding of living things, there would be no reason to think that flourishing-based function ascriptions are based on the nature of living things as living. How can neo-Aristotelians argue that flourishing-based functions are necessary for understanding the nature of plants and non-human animals? Thompson (2008) offers a transcendental argument for the life-form concept that has served as the foundation of the neo- Aristotelian account of natural goodness. In his work on representation of life, Thompson argues that the life-form concept, together with its implications of goodness and defect, is necessary for representing a living thing as living. Given that flourishing-based functions are defined in terms of the life-form concept, this transcendental argument is most relevant to the question at hand. If the life-form concept is necessary for understanding living things, the corresponding flourishing-based concept of function would be equally essential. In the next section, I examine this argument to see if it can solve the problem of Grounding.

2.6 The Argument from Representation of Life

According to Thompson (2008), apprehending something as living requires viewing some of its parts as organs like legs and wings, and some of its activities and processes as vital operations like eating and breathing. But what counts as a leg or what counts as eating differs from one kind of organism to another. In fact, the same physical entity or process can amount to different organs or vital operations in different life-forms. For instance, the process of cell division amounts to reproduction in bacteria but constitutes growth in humans. Thompson argues that there is nothing “in the organism considered in its particularity or as occupying a

36 certain region of space” which determines that an organ is there or a vital operation is happening. “That they are there or happening, and thus that we have an organism at all, presupposes the existence of a certain ‘wider context’”, which Thompson takes to be the context of a life-form concept (2008, 56). Thus, he argues that apprehending something as living requires presupposing a life-form concept and its corresponding natural history. What is more, a life-form concept brings with it not just the context required for recognizing organs and operations, but also the related norms of natural goodness and flourishing-based function. For instance, when we recognize an individual living thing as a bat, we already commit ourselves to assessing it against the norms that are implicit in our conception of the bat life- form. No doubt, our conception might be incomplete and some of the natural-historical judgments that we make may be false. We may initially make the natural-historical judgment that “bats are blind” but later find out that “bats can see”. But according to Thompson, however we may revise our conception, it remains the case that presupposing some conception of the life-form and taking some natural-historical judgments to be true is necessary for identifying the bat as a living organism. Thus, to the extent that we do represent and identify living things as living, we are committed to there being norms of natural goodness and flourishing-based function that apply to them. However, the problem with this argument is that—much like Foot’s intuitively plausible remarks—it relies on representations of life in commonsense descriptions of living things rather than state-of-the-art science. Thompson may be right that we ordinarily assume a background of natural-historical judgments about living things when we apprehend them. But it is not clear that the framework of natural-historical judgments and the concept of life-form that they underwrite really provide the best conceptual tool for understanding the nature of the object we perceive. Thompson allows that we may revise our conception of a life-form in light of empirical observations.18 But he takes the logical structure of our representations—the

18 See Thompson’s (2004) vivid discussion of how empirical observations guide us in acquiring knowledge of a novel type of jellyfish.

37 special logical form of natural-historical judgments—to remain intact. In other words, he thinks what a characteristic bat life consists in is subject to revision, but that there is a characteristic bat life that is expressible in the form of natural-historical judgments is not. However, it’s not clear why this aspect of our commonsense understanding of living things cannot be empirically questioned. Why should we assume that individual bats are best understood in relation to characteristic norms that group them together? Thompson draws on the fact that the logical form of natural-historical judgments is distinct from other forms of generality, and that we seem to use these judgments exclusively in relation to living things. Yet, this is not enough to show that we can take any aspect of these judgments for granted. Science often dispenses with our folk understanding of things by offering theories that are superior and explain the relevant phenomena better. This is clear in the case of modern physics, which has replaced our naïve conception of the physical world. There is an extensive literature on our intuitive explanations of motion that suggests they are fundamentally different from Newtonian explanations (Nersessian and Resnick 1989; McCloskey 1983). We seem to commonly apply a model resembling Aristotelian physics, according to which continuous motion requires the sustained force of an internal or external cause, unlike the state of rest, which doesn’t require any causal explanation. In contrast, Newtonian physics maintains that a moving object continues to move until acted upon by some external force. It contradicts the commonsense intuition by suggesting that motion is a state and only changes of state (e.g., accelerated motion) require explanation. Thus, a folk view might help to get a useful grasp on things, but it often loses ground to modern science when we seek a sound understanding of the world and the nature of the things it contains. This is why Thompson’s argument from the representation of life in the folk view is not sufficient to show that the concept of function that is implicit in this view is essential to the understanding of living things. That said, if it turns out that the natural-historical form of thought is not only employed in folk representations of life but also explanatorily indispensable to modern biology, then the fact that it implicitly commits us to the flourishing-based concept of function would be telling.

It would suggest that this concept of function is indeed necessary for understanding living things and accurately represents their nature. In fact, Hacker-Wright and Lott have both suggested as much, claiming that even a science like evolutionary biology would have to presuppose the natural-historical form of thought (Hacker-Wright 2009a; Lott 2012a). However, as we will see in more detail in the next chapter, Lott and Hacker-Wright’s defense of this claim does not add much to Thompson’s transcendental argument. For them, this is simply a consequence of Thompson’s argument that natural-historical judgments are necessary for representing living things. They argue that since biology is the study of living things, biologists need to presuppose the concept of a life-form before they can even start doing biology, because the life-form concept is involved in recognizing living things in the first place. Lott, for instance, claims that “to so much as have a topic for evolutionary explanation, we must rely on Thompson-Foot judgments of life form” (2012, 375). Hacker-Wright similarly claims that a life-form conception “is always in play when we make a judgment of an organism,” regardless of whether we are doing armchair speculation or evolutionary biology (2009, 316). What is troublesome about this argument, however, is that these claims are made without actually looking at the conception of organism in evolutionary biology to see if the relevant concept of a life-form is at work there. Lott and Hacker-Wright seem to assume that that since biology is about living things, biologists are already committed to the life-form concept that is involved in recognizing living things in the first place. However, the mere consideration that the subject matter of a science is pre-scientifically characterized in terms of a concept is not enough to argue that this concept plays an explanatory role in the science. The question is whether the ultimate scientific account of the subject matter remains faithful to the initial pre-scientific characterization, and it is not at all obvious that this is the case with evolutionary biology. In fact, as we will see in the next chapter, the dominant view of evolution today—the Modern Synthesis—does not explain evolution in terms of our pre- scientific conception of an organism at all. It rather defines evolution as change in gene frequencies within a population of genes over time (see Dobzhansky 1937, 11). As many

39 historians and philosophers of biology have noted, this characterization involves a shift in the subject of evolutionary explanation and the ontology of biology—from organisms in Darwin’s theory to genes and populations in the Modern Synthesis.19 Neo-Aristotelians might respond that if a genocentric theory like the Modern Synthesis offers an accurate account of evolutionary biology, this just shows that evolutionary biology is not relevant to the study of living things as living. Evolutionary biology has very specific explanatory aims, and it may turn out that it doesn’t need to take account of all aspects of living things to meet these specific aims. So neo-Aristotelians like Lott and Hacker-Wright might be happy to drop their claim that the life-form concept is presupposed in evolutionary biology, but insist that any relevant field of science that studies the realm of living things as living would inevitably depend on the natural-historical framework. However, without identifying such relevant fields of science, this would be an argument that begs the question against the critic. It’s true that no branch of science captures all aspects of reality. But this doesn’t make folk biology and the natural-historical framework that it underwrites immune to refutation from empirical sciences. Although it may be argued that there are aspects of reality that are not knowable via the methodology of empirical sciences, there is no reason to think that the domain of life is one of them, particularly when it comes to plants and non-human animals. Hacker-Wright calls the neo-Aristotelian account “a logical theory” and distinguishes it from “empirical” theories of living things. He claims that this logical theory is not supposed to be explanatory, but rather plays a different theoretical role in identifying organisms as such (2009a, 316). However, in my view, drawing a distinction between logical and empirical theories of living things and isolating the former from our scientific understanding of the world would be a mistake. The everyday, common-sense representations of living things in natural-historical judgments are not any less empirical than representations of living things in biology. And we shouldn’t assume that the concepts that are implicit in

19 See Daniel Nicholson’s remarks on the disappearance of the organism in evolutionary theory (Nicholson 2014, 1-2).

40 these judgments cannot be subject to further empirical investigation based on science.

Moreover, these representations do not have any privileged position with respect to identifying organisms as such. They are not any more geared toward representing living things as living than are the claims of evolutionary biology.20 The only difference is that the latter have superior epistemological credentials due to the systematic rigor of scientific practice. Thus, even if one is to argue that evolutionary biology fails to capture parts of reality, such an argument needs to be made either by appeal to other branches of biological science such as ecology or ethology, or based on considerations internal to evolutionary biology itself. To sum up, the best strategy for securing the premise of Grounding is to argue that the life-form concept plays an indispensable explanatory role in a biological science. But the first step in making such an argument is acknowledging that the content of empirical science is relevant to answering this question. Such an argument cannot assume that any aspect of our folk biology is immune to refutation. Taking the content of empirical science seriously means actually examining the scientific account of living things to see if the neo-Aristotelian concepts of life-form and flourishing-based function have a role to play in scientific explanation.

2.7 Making Peace with the Relevance of Biology

Neo-Aristotelians have kept their account of natural goodness at a safe distance from the science of biology, and even when they make claims about the presuppositions of evolutionary biology, they treat the question as an a priori matter. In this final section, I discuss some of the reasons neo-Aristotelians might think they should keep biology at arm’s length, and offer some thoughts to counter these reasons. The most obvious reason for neo-Aristotelians to avoid engaging with evolutionary biology would be assuming that the science is not on their side. Critics of neo-Aristotelian naturalism appeal to a genocentric, arguably reductionist, account of evolution that has been

20 When we make a judgment like “cats have four legs” we aren’t primarily concerned with identifying what makes cats living but are simply trying to understand the kind of thing in front of us.

41 very influential in philosophy of biology since the rise of Modern Synthesis in the twentieth century. If this account of evolution is accurate, the prospects of arguing for a suitable account of natural goodness by consulting evolutionary biology would be dim. As we will see in more detail in the next chapter, the concept of an organism as an irreducible entity does not have a place in Modern Synthesis. Genes are seen as the sole units of inheritance and selection, and even the organismic process of development is considered to be a mere execution of a genetic program (see e.g., Williams 1966; Monod 1971). However, what both neo-Aristotelians and their critics seem to overlook is that Modern Synthesis has been seriously challenged from various fronts in recent years. Empirical advances in the understanding of development (Oyama 2000; Oyama, Griffiths, and Gray 2001), evolutionary novelty and selection (West-Eberhard 2003; 2005), and epigenetic inheritance mechanisms (Jablonka and Lamb 1995; 2005) have revealed many theoretical shortcomings of the genocentric approach. Some biologists and philosophers of biology have called for “the return of the organism” (Nicholson 2014), suggesting a very different view of evolution. On this organocentric view, which I will explore more thoroughly in the next chapter, organisms are the primary agents of evolutionary change, and the main processes of evolution are consequences of the distinctive capacities of whole organisms such as their plasticity and robustness (Walsh 2016; Pigliucci and Müller 2010; Huneman 2010). This alternative approach to understanding evolution seems to lend itself well to a neo-Aristotelian argument for a holistic concept of organism. Of course, effectively making such an argument requires a good examination of the relevant work in evolutionary developmental biology.21 But there is reason to think that the most empirically adequate account of evolution and development may in fact yield a conception of organism that can support a suitable account of natural goodness. One may object that there is a more principled reason against relying on the science of biology. On the neo-Aristotelian view, the concept of a living organism is the basis for all

21 An example of the relevant sort of argument can be found in Lubichler and Wagner (2000), who argue that taking the concept of organism to be ontologically prior to its functional structures can solve certain problems of mathematical models in biology with character identification.

42 evaluations of natural goodness including evaluations of moral goodness in humans. The concern is that by making the appropriate concept of organism contingent upon the facts of biology, the kind of necessity that is appropriate to moral goodness will be lost. Even if biologists and philosophers of biology were to decide that the neo-Aristotelian life-form concept is necessary for explaining biological facts, the mere possibility that empirical findings could reveal things to be otherwise seems damaging to the status of ethics as necessary and independent from such biological contingencies. As Thompson puts it, consulting biology to answer the question would be to give a wrong position to biological facts and to “turn ethics into a sub-discipline of biology” (2004, 62). However, my suggestion here is not that we should consult biology for an account of what flourishing and moral goodness in humans consists in. The question is whether the concept of organism that yields the best understanding of living thing—including plants and non-human animals—is suitable for grounding the norms of flourishing and moral goodness in humans. What I am suggesting is that biology is the relevant field to look for the appropriate concept of organism. If biology shows that our most empirically adequate concept of organism is not suitable for grounding the norms of natural goodness, it is not our conception of human goodness, but rather the continuity between humans and non-humans that faces a challenge. Thus, the kind of relationship with science that I am advocating does not threaten the kind of necessity that is appropriately attributed to ethics. We do not look to justify our substantial moral virtues like justice and benevolence ‘from outside’ via virtue-neutral biological facts. It is part of the neo-Aristotelian account of virtue that the justification for substantial virtues comes from practical reason and not through biology. It is rather the justification for the neo-

Aristotelian metaethical position of naturalism that has to come in part from biological facts, which is not surprising given that it is a naturalist position with commitments about biological entities. In summary, I have argued that the evolutionary objection to neo-Aristotelian naturalism raises a dilemma for this view, which ultimately challenges the claim that evaluations of

43 natural goodness capture the nature of living things correctly. As a form of ethical naturalism that relies on there being a continuity between humans and non-humans, neo-Aristotelian naturalism has a set of commitments about the life of plants and non-human animals that are best investigated by empirical science. I have argued that the reason the current responses to both versions of the evolutionary objection are inadequate is that neo-Aristotelians distance their view from empirical science in a way that is unnecessary and unjustified. Thus, in order to move forward in this debate, I propose that neo-Aristotelian naturalism should own its empirical commitments and rethink its relationship with biology. This is exactly what I aim to do in the rest of this dissertation. In the next two chapters, I examine some of the explanatory concepts of biology to see if they can lend support to the commitments of neo-Aristotelian naturalism regarding plants and animals.

Chapter 3

Is the Neo-Aristotelian Concept of Organism Presupposed in Biology?

In this chapter, I examine more thoroughly Lott and Hacker-Wright’s claim that the neo- Aristotelian concept of life-form is presupposed in biology by looking at the concept of organism in modern evolutionary biology. I first argue that the Modern Synthesis theory of evolution, which understands evolution as change in gene frequencies within a population, does not presuppose the relevant concept of organism. I then explore an alternative view of evolution that has emerged in the past twenty years from advances in evolutionary developmental biology. I argue that this so called ‘Evo-Devo’ approach makes room for an explanatory concept of organism that can be reconciled with the neo-Aristotelian view. Moreover, I argue that although the explanatory role of the concept of organism in evolutionary biology is still contentious, the well-established role of this concept in developmental biology can be used to defend the biological commitments of neo-Aristotelian naturalism.

3.1 The Need to Examine the Representation of Life in Biology

Remember the neo-Aristotelian argument for ethical naturalism that we reconstructed in the previous chapter:

Refined Argument for Naturalism

1. Norms of natural goodness in non-human life-forms are natural. (Grounding) 2. If norms of natural goodness in non-human life-forms are natural, then norms of

natural goodness in the human life-form are also natural. (Continuity)

3. Norms of moral goodness are instances of norms of natural goodness in the

human life-form. (Normative Adequacy) C. Therefore, norms of moral goodness are natural.

We saw that the current neo-Aristotelian response to the evolutionary objection does not adequately defend the premise of Grounding in this argument. I argued that defending this premise requires showing that the neo-Aristotelian, flourishing-based concept of function— which is the basis of evaluations of natural goodness—is necessary for understanding non- human living things. I also suggested that the best strategy for showing this would be to argue that the neo-Aristotelian concept of life-form and the corresponding notion of flourishing- based function play an indispensable explanatory role in a biological science. In this chapter as well as the next chapter, I investigate the prospects of making such an argument by looking at some of the explanatory concepts of biology. The focus of this chapter is on the concept of an organism. As I pointed out earlier, some neo-Aristotelians such as Lott (2012a) and Hacker- Wright (2009a) do make the claim that the life-form concept is presupposed in evolutionary biology. But their argument for this claim does not add much to Thompson’s transcendental argument, which takes the representation of life in pre-scientific, natural-historical judgments for granted. So I want to start this chapter by looking more closely at Lott and Hacker- Wright’s argument for the role of the life-form concept in evolutionary biology. My aim is to show that because these authors do not look at the content of evolutionary biology and the concept of organism that is actually at work there, their arguments do not amount to an adequate defence of the relevant neo-Aristotelian concept. There are two considerations that Lott (2012a) and Hacker-Wright (2009a) offer to defend the claim that evolutionary biology presupposes the life-form concept. The first is that living things are the subject matter of evolutionary biology. The second is that some of the explanatory concepts of evolutionary biology presuppose the life-form concept. Let’s call these the argument from subject matter and the argument from explanatory role. In the remainder of this section I examine these arguments in turn.

The argument from subject matter appeals to the fact that biology is the study of living things. Lott and Hacker-Wright argue that since biology is about living things, biologists are already committed to the life-form concept that is involved in recognizing living things in the first place. Lott says, for instance, that “to so much as have a topic for evolutionary explanation, we must rely on Thompson-Foot judgments of life form” (Lott 2012a, 375). Hacker-Wright similarly claims that a life-form conception “is always in play when we make a judgment of an organism,” regardless of whether we are doing armchair speculation or evolutionary biology (Hacker-Wright 2009a, 316). However, this argument is far too quick. Our initial characterization of the subject matter of a science is merely the starting point of our study. The question we should ask is whether the ultimate scientific account of the subject matter under study remains faithful to our initial characterization. Consider organic chemistry—the study of organic compounds, which were initially taken to be compounds found in living organisms. The division between organic and inorganic chemistry was motivated by the fact that compounds derived from plants and animal sources seemed to have distinctive features such as being less stable and more prone to decomposition. These differences were thought to be explained by the vital force theory, the idea that a vital force existed within organic material. However, it turned out that the very compounds that were the focus of organic chemistry were also obtainable from non-living sources. Organic compounds were thus redefined as compounds that contain a significant amount of carbon, even those with a non-biological origin.22 Consequently, modern organic chemistry does not support the vital force theory or any other theory regarding what makes living things distinctive. As far as organic chemistry is concerned there is no distinction between living and non-living things. Thus, the mere consideration that the subject matter of evolutionary biology is pre- scientifically characterized in terms of natural-historical judgments does not lend any support to these judgments or the life-form concept that accompanies them. It is not at all obvious that

22 See Klein (2005) for an account of the shifting ontology of chemistry in the 18th and 19th century.

47 the final scientific account of this subject matter would be best characterized in a way that presupposes the life-form concept. In fact, as we will see in the next section, the dominant view of evolution today—the Modern Synthesis—does not characterize evolution in terms of living organisms at all. It rather defines evolution as change in gene frequencies within a population of genes over time (Dobzhansky 1937, 11). As a result, the subject of evolutionary explanation and the ontology of biology has gone through a shift—from organisms in Darwin’s theory to genes and populations in the Modern Synthesis. So even if living things are initially characterized in terms of the life-form concept, there is a question whether evolutionary biology ultimately remains faithful to this initial characterization. One might think the concept of a gene itself somehow presupposes the concept of an organism, which in turn presupposes a life-form concept. This brings us to the second argument found in Lott and Hacker-Wright’s remarks, namely that some of the explanatory concepts of evolutionary biology rely on the life-form concept. More specifically, Lott suggests that representing something as a gene, or an activity as reproducing, requires the context of a life-form in the same way that representing something as an organ or as a vital operation does

(Lott 2012a, 375).23 If this is correct, not only does evolutionary biology focus on a subject matter that is initially characterized by appeal to life-forms, its theoretical account and explanation of this subject matter is also committed to the life-form concept. However, the central claim in this argument—that the conception of gene or reproduction in evolutionary biology presupposes the life-form concept—is not something that Lott actually argues for. Lott’s idea seems to be that the life-form concept must be in the background of evolutionary concepts for the same reason that it is in the background of everyday descriptions of living things—namely that natural-historical judgments are involved in identifying the domain of life. But note that the conception of gene in evolutionary theory does not have to be the same as the folk conception of gene. Although the folk conception may presuppose the life-form concept, it’s not clear that the scientific conception of gene has any

23 Hacker-Wright made similar remarks about the concept of gene in personal communication.

48 such presuppositions. If it turns out, for instance, that evolutionary biology defines genes in molecular terms by reference to their replicability and their immediate function in protein construction, no reference to the life-form of the organism seems necessary. In fact, at least prima facie, Modern Synthesis biology does not seem to characterize genes by reference to organisms. It is rather organisms that are characterized in terms of genes. As Richard Dawkins famously puts the point, on this view organisms are nothing but “survival machines—robot vehicles blindly programmed to preserve the selfish molecules known as genes” (1976, xxi). Thus, insofar as Lott assumes that the scientific conception of gene must be the same as the folk conception, or that it must somehow presuppose the life-form concept simply because it concerns the domain of life, this argument from explanatory role does not add much to the argument from subject matter. What is missing in Lott and Hacker-Wright’s arguments is looking at the explanatory concepts of evolutionary biology and asking whether they really presuppose the life-form concept. The argument from explanatory role cannot be made a priori. We need to consult the theoretical account of living things offered in biology to see if the relevant concepts actually do play an explanatory role there. This is exactly the question that I aim to investigate in the rest of this chapter. My intention is to first argue that the explanatory role of the relevant concept of organism in evolutionary biology is in fact disputed, and then explore ways in which neo- Aristotelians might be able to argue for the explanatory significance of this concept by appealing to research in evolutionary developmental biology. In section 3.2, I discuss the Modern Synthesis theory of evolution, and argue that the concept of an organism does not have a place in the Modern Synthesis. In Section 3.3, I explore an alternative view of evolution that has emerged in the past twenty years from research in evolutionary developmental biology. I highlight the explanatory role of the concept of organism in the so called ‘Evo-Devo’ account, and argue that this view is more congenial to the claims of neo-Aristotelian regarding the concept of organism. Finally, in section 3.4, I argue that although the explanatory role of the concept of organism in evolutionary biology is still

49 contentious, the well-established role of this concept in developmental biology offers a promising starting point for defending the commitments of neo-Aristotelian naturalism.

3.2 The Modern Synthesis

In this section, I examine the idea that evolutionary biology presupposes the neo-Aristotelian concept of a life-form by looking more closely at the Modern Synthesis theory of evolution. I highlight how this theory differs from Darwin’s original formulation, and argue that the concept of an organism does not play a role in the Modern Synthesis—neither as the subject of explanation nor as an explaining factor. Darwin’s view in the Origin of Species was that organisms arise and develop through the natural selection of small heritable variations that increase their ability to survive and reproduce. His characterization of natural selection was as follows:

[I]f variations useful to any organic being ever do occur, assuredly individuals thus characterised will have the best chance of being preserved in the struggle for life; and from the strong principle of inheritance, these will tend to produce offspring similarly characterised. (Darwin 1859[1872], 102)

Thus, his basic idea was that heritable difference in fitness among organisms results in evolution by natural selection. Darwin understood that evolution by natural selection requires inheritance of parents’ traits by offspring, but he had no good explanation of inheritance. The prevalent idea of inheritance in Darwin’s time was blending inheritance, according to which the traits of offspring would be blends of the traits of their parents. This was problematic for Darwin’s theory, because blending inheritance would cause the whole population to converge to the same average for all traits in a few generations, which would eliminate variation— another essential ingredient of natural selection. This problem was not solved until Mendelian genetics came along and became well- known in the 20th century. Mendel’s theory of particulate inheritance showed how the traits of parents could be passed on to offspring as ‘particles’ or separate entities that would remain

50 unaffected in the offspring. Insofar as the offspring’s traits are a blend of its parents’, that blending happens later, during development. The inherited particles can remain unchanged in a lineage generation after generation, which explains how variation can remain in the population over time. Thus, Mendel’s theory of inheritance filled the gap in Darwin’s account of natural selection, and eventually led to the Modern Synthesis, which integrated Darwin’s theory with genetics. As we will see below, however, this integration with genetics resulted in the disappearance of the concept of organism from evolutionary biology—both as the subject of evolutionary explanation and as an explanatory factor. As Daniel Nicholson describes this shift, in Modern Synthesis biology, organisms have no explanatory role to play because they have no autonomous agency of their own. They are mere “liaisons of evolution; a sort of interface between the phenotypic expression of genes and the selecting role of the environment” (2014, 2). It’s rather the causal capacities of genes that drives evolution forward, which is why evolutionary explanation must be sought at the level of genes. In the rest of this section, I illustrate this shift in the focus of evolutionary explanation by looking at how evolution is portrayed in standard modern texts in biology and philosophy of biology. The first thing that I want to highlight is how the subject of evolutionary explanation has changed from organisms to genes. The Modern Synthesis reconceptualizes evolution as a change in allele frequencies of a population. The locus classicus of this shift is Theodosius Dobzhansky’s Genetics and the Origin of Species, where he defines evolution as “a change in the genetic composition of populations”. According to Dobzhansky, since evolutionary change is essentially genetic change, the study of evolution “falls within the province of population genetics” (Dobzhansky 1937, 11). Similarly, if we look at standard textbook accounts of evolution, we will see that the very idea of evolutionary change is defined in terms of change in the frequency of genes. Douglas J. Futuyma’s Evolution, for instance, says the following:

Because evolution consists of genetic change in populations over time, evolutionary biologists are most interested in those variations that have a genetic basis. (Futuyma 1942, 190)

This characterization of evolutionary change as change in the frequency of genes marks a clear departure from Darwin’s original focus on change in the observed characteristics of organisms (i.e., phenotypes) as the subject of evolutionary explanation. The Modern Synthesis reduction of evolutionary change to genetic change may appear puzzling at first. Even if genes account for inheritance, it still seems that what evolutionary theory has to explain are primarily changes in the phenotype—changes in the anatomy, physiology, and behavior of organisms over time. What we find the most interesting and curious about evolutionary change is how characteristics of organisms change, and particularly how they become adapted to their conditions of existence. There is something very remarkable about living organisms, namely that they fit their environment quite well. Their characteristics seem to have been carefully designed to enable them to appropriate the world around them. We see, for instance, that an aquatic organism like a fish has fins for swimming, and a subterranean organism like a mole has long claws for digging. The fit of organisms to their environment just seems to be a manifest fact about them that evolutionary theory needs to explain. However, the Modern Synthesis shifts away from explaining phenotypic change and toward explaining genetic change for two reasons. The first reason has to do with the Weismann Doctrine—the idea that phenotypes acquired in an organism’s lifetime do not modify the genes that the organism’s offspring inherits. Darwin actually believed that the effects of use and disuse on an organism’s body could be inherited. But this Lamarckian view did not sit well with the genetic account of inheritance and the Weismann Doctrine. Weismann (1893) proposed a strict distinction between the germ line—cells that will be the ancestors of the organism’s sex cells—and the somatic line—cells that form the tissues and other components of the organism’s body. And he argued that while the somatic line originates from the germ line,

52 the germ line remains separate from the somatic line such that no change in the somatic line can be transmitted back into the germ line. This later received confirmation in the so-called ‘central dogma’ of molecular biology, according to which information flows only from DNA to protein molecules and never in reverse. This is the first reason why the Modern Synthesis focuses on genetic, as opposed to phenotypic, change. The ‘Weismann barrier’ between the genotype space and phenotype space does not allow the phenotypic changes that are ‘acquired’, or ‘environmental’, to enter the genotype space. As a result, any phenotypic changes that are not associated with genetic change are considered to be short-lived and not worth focusing on in the study of long-term evolutionary change.24 The kind of change that most interests evolutionary biologists and is considered to constitute evolution, according to the Modern Synthesis, is genetic change. Secondly, although the Modern Synthesis is concerned with explaining adaptations, it does not understand adaptations in terms of how well the characteristics of organisms fit their environment. Instead, since adaptations are viewed as resulting from natural selection, they are simply defined as characteristics that have evolved by natural selection, i.e., by the differential success of genes in replicating themselves (see Futuyma 1942, 247-248; Sober 2000, 84). In fact, the Modern Synthesis denies that there are any design-like characteristics to be explained. There is only an appearance of design, which is explained away by uncovering the mechanism through which it is created. Futuyma’s text, for instance, tells us that the remarkable nest building behavior of weaver ants (Oecophylla)—where workers cooperate and use larval silk to weave together leaves—is merely a genetically determined behavior resulting from many random mutations:

24 As we will see in the next section, the idea that phenotypic changes do not affect the long-term course of evolution is later challenged by more recent research programs such as Evo-Devo (West-Eberhard 2003; 2005) as well as Niche Construction Theory (Odling-Smee et al. 2003). But even though the Moderns Synthesis incorporates some of these criticisms in refining its explanation of evolutionary change, its characterization of evolutionary change as genetic change remains firmly in place.

[T]he weaver ant’s behavior has the appearance of design because among many random genetic variations (mutations) governing the behavior of an ancestral ant species, those displayed by Oecophylla enhanced survival and reproduction under its particular ecological circumstances. (Futuyma 1942, 250)

So the improved survival and reproduction of organisms—which has the appearance of design—is viewed as merely a by-product of molecular processes at the level of genes. Because of this, the adaptations that standard evolutionary theory aims to explain are fully characterized in terms of genes and populations. In other words, it is simply not true that organisms or their observed characteristics constitute the subject of evolutionary explanation in Modern Synthesis biology. Although organisms are not the subjects of evolutionary explanation, one may wonder whether they appear in the content of explanation as explaining factors. So the next thing that I want to look at is the role of organisms in the explanations that are given of evolutionary (genetic) change. Futuyma’s text, which provides a nice illustration of the standard Modern Synthesis account, describes evolution as a two-step process: (i) the origin of genetic variation, and (ii) change in the frequencies of genes caused by genetic drift or natural selection (Futuyma 1942, 270). The first step, which is the origin of genetic variation, is considered to be “the foundation of evolution” and responsible for long-term evolutionary changes (Futuyma, 189). Darwin’s formulation in fact required phenotypic variation to exist in a population, but Darwin did not offer any account of the origin of such variation. He only postulated “a tendency to vary, due to causes of which we are quite ignorant” (Darwin, 107). The Modern Synthesis, in contrast, focuses on genetic variation, and offers two sources for this variation which are both genetic processes: mutation and recombination. In other words, the Modern Synthesis identifies what Darwin characterizes as the organisms’ “tendency to vary” with molecular processes at the level of genes.

The second step in the process of adaptive evolution is change in the frequency of genes in a population due to natural selection. According to Lewontin’s (1970) well-known formulation, evolution by natural selection occurs when there is heritable variation in fitness. This requires that three basic components are in place:

1. Different individuals in a population have different morphologies, physiologies, and behaviors (phenotypic variation). 2. Different phenotypes have different rates of survival and reproduction in different environments (differential fitness). 3. There is a correlation between parents and offspring in the contribution of each to future generations (fitness is heritable). (Lewontin 1970, 1) To see how this ‘recipe’ leads to evolution by natural selection, consider an example from Sober (2000). Suppose we have a herd of zebras in which there is variation in running speed (phenotypic variation). Suppose also that faster zebras are better able to survive because they are better able to evade predators (differential fitness). Further, suppose that running speed is inherited, and offspring take after their parents (fitness is heritable). Given these three conditions in place, we can see that because faster zebras have better success with survival and reproduction and thus replicating their genes, and because their genes are inherited by their offspring, the average running speed in the herd will increase over time. Now, there is no question that organisms are more or less successful in survival and reproduction because of their phenotypes. In our example, some zebras have differential fitness because of variation in their running speed. It is the interaction of the organisms’ phenotypes with the environment that determines which organisms can ultimately pass on their genes to the next generation. So one might think phenotypes do play a causal role in natural selection after all, because they are the factors that are actually screened and selected by natural selection. However, for the same reason that phenotypic change without underlying genetic change is not the subject of evolutionary explanation, phenotypic variation without a genetic

55 basis is also largely considered to be irrelevant as an explanatory factor in the Modern

Synthesis. The textbook account, for instance, claims that although phenotypic variation is sufficient for natural selection to occur, it “can have no evolutionary effect unless phenotypes differ in genotype” (Futuyma, 270). The idea is that if phenotypes vary despite being genetically identical, they will be subject to selection. But assuming that there cannot be inheritance without a genetic basis, this kind of selection cannot have long-term evolutionary consequences.25 Because of this, the Modern Synthesis focuses on the fitness of genotypes, even though “genotypes differ in fitness only because of differences in phenotypes” (Futuyma, 270). In other words, the Modern Synthesis simply abstracts away from the role of phenotypic variation. It explains evolutionary change solely in terms of genetic variation. My discussion of the Modern Synthesis in this section shows that the current paradigm of evolutionary theory understands evolution as change in gene frequencies within a population over time, and explains it in terms of molecular properties of genes and statistical properties of populations. Thus, the neo-Aristotelian concept of an organism does not seem to have a place in the Modern Synthesis, neither as the subject of explanation, nor as an explanatory factor.

3.3 The Evo-Devo Approach

Although the Modern Synthesis is a widely influential view, it is not the only conception of evolution that has been put forward. In fact, in recent years, the Modern Synthesis has been seriously challenged from various fronts. Empirical advances in the understanding of evolutionary novelty and selection (West-Eberhard 2003; 2005), biological development (Oyama 2000; Oyama, Griffiths, and Gray 2001), and epigenetic inheritance mechanisms (Jablonka and Lamb 1995; 2005) have exposed many theoretical problems of the Modern Synthesis and its genocentric approach. The various research programs engaging in these

25 Again, as we will see in the next section, this is one of the tenets of the Modern Synthesis that is later challenged by Evo-Devo among other recent theoretical approaches.

56 studies suggest that organisms and their development cannot be bracketed from the study of evolution quite in the way that is assumed in the Modern Synthesis. As a result, some biologists and philosophers of biology have called for “the return of the organism” (Nicholson 2014), suggesting a very different understanding of evolution. On this organocentric view, organisms are seen as the primary agents of evolutionary change, and the main processes of evolution are seen as consequences of the distinctive capacities of whole organisms such as their plasticity and robustness (Walsh 2016; Pigliucci and Müller 2010; Huneman 2010). In this section, I focus on the evolutionary-developmental account of evolutionary change offered by West-Eberhard (2003; 2005). Evolutionary-developmental biology, or as it is colloquially called ‘Evo-Devo’, is a discipline that is concerned with discovering and understanding the role of developmental mechanisms in evolution (Hall 2003). What is interesting about Evo-Devo is that by identifying the role of developmental mechanisms in evolutionary change, it restores the place of organism in evolutionary theory. West-Eberhard’s evolutionary-developmental account of evolutionary change ascribes a significant explanatory role to the phenotypic and developmental plasticity of organisms in evolution. What is meant by the plasticity of organisms is their ability to change in response to external or internal environmental inputs during their development. An illustrative example of this ability is the real-life case of a two-legged goat described in 1942 by morphologist E. J. Slijper. This goat was born without forelimbs, but it adapted to its condition in unexpected ways, and developed peculiarities such as enlarged hind limbs, a curved spine, and a large neck. As a result, the goat learned to walk and run by using its hind legs alone. As West- Eberhard (2005) explains, the correlated shift in the goat’s morphology and behavior led to “the well coordinated production of a complex and individually advantageous adjustment, producing a novel phenotype with little or no genetic change” (West-Eberhard 2005, 6545). West-Eberhard argues that plasticity is an evolved property of living things that is universal among them and has implications for our understanding of evolution. She claims that evolutionary biologists since the 19th century have been mistaken to dichotomize development

57 and selection. In her account of adaptive evolution, development plays a major causal role, most notably in originating evolutionary novelty. The two steps of the process of adaptive evolution on her account are (i) generation of variation by development, and (ii) screening of variation by selection (West-Eberhard 2003, 139). Thus, West-Eberhard’s Evo-Devo account disagrees with the Modern Synthesis about the origin of variation. The Modern Synthesis was merely concerned with genetic variation, and considered mutation and recombination to be the only sources of variation. As stated earlier, this was due to the fact that variation needs to be inherited in order to be evolutionary and have long-term effects. In contrast, in the Evo-Devo account, the fact that phenotypic variation is what is screened by selection is taken to be more significant, and the role of development in generating phenotypic variation is acknowledged. The fact that West-Eberhard takes phenotypic variation to be significant is not because she neglects the importance of genetic variation for long-term evolution. Unlike some other advocates of the Evo-Devo approach (e.g., Johnston and Gottlieb 1990; Walsh 2015), West- Eberhard does not contest the conception of evolutionary change as genetic change. What she rejects, however, is the claim that genetic variation is the origin of evolutionary change. Her idea is that, just because phenotypic variation needs to be heritable to be evolutionary, it does not follow that it needs to be inherited. In fact, West-Eberhard rejects the distinction between inherited and acquired phenotypes. Following Johnston and Gottlieb (1990), she argues that all phenotypic characters arise in development as a result of an interaction between the organism and its environment. The fact that genes play a role in this interaction does not mean that they directly determine any phenotypes. There is no such thing as a ‘normal’ or merely permissive environment in which genes get to generate phenotypes that are fully inherited. Thus, West-

Eberhard takes it to be misleading to ascribe the origin of variation to the properties of genes. She argues that even though phenotypic variation resulting from development is acquired, it can still be heritable and have long-term evolutionary effects. According to West-Eberhard, the “evolutionary potential of a developmental novelty” depends on two factors: recurrence and heritability. Recurrence refers to the formation of a

58 population of individuals expressing the trait, and is necessary for a developmental novelty to lead to selection. Heritability, on the other hand, is a measure of how much an offspring’s traits resembles the parents’—independently of whether this involves the transmission of genes26— and is required for selection to lead to evolutionary genetic change. West-Eberhard argues that developmental novelties can be recurrent and heritable regardless of whether they are initiated by a genetic or environmental change. Her argument for heritability of environmentally- induced novelties relies on the observation that there is virtually always genetic variation among traits that are phenotypically similar. This hidden genetic variation, which on its own does not constitute any phenotypic variation, can be ‘exposed’ when an environmental change induces phenotypic variation. And since the resulting phenotypic variation that is subject to selection will be accompanied with preexisting genetic variation, it can be inherited. Let me illustrate this with a classic example from C. H. Waddington’s (1953) experiments on Drosophila. In these experiments, Waddington applied a temperature shock during the egg stage of Drosophila, which caused a bithorax-like phenotype in adult flies. This phenotype was exhibited in various degrees among the flies, and responded quickly to artificial selection when

Waddington crossed the bithorax individuals for a few generations. What Waddington observed was that even though the bithorax phenotype was induced environmentally, it was “genetically assimilated” and resulted in long-term evolutionary change. In fact, the novel phenotype ultimately remained in the population even after the temperature shock was removed. Waddington argues that this phenomenon is due to the canalization, or robustness, of the developmental processes in Drosophila, which is their ability to produce the same phenotype regardless of variability of genotype or environment. When developmental pathways are canalized, most individuals phenotypically look the same despite having different genetic make-ups. However, canalization can break down as a result of extreme environmental stress, and this would expose the hidden genetic variation. In this case, the pre-existing genetic

26 Note how this notion of heritability diverges from the genetic conception of inheritance in the Modern Synthesis.

59 variation in the population is so extensive that when canalization breaks down, the bithorax phenotype appears in different degrees. These novel phenotypes then provide the material for selection, which can further change the genetic make-up of the population to the point that the original environmental stress is no longer necessary to trigger the phenotype. Having shown that environmentally-induced phenotypic changes can be heritable in this way, West-Eberhard argues that “the causal chain of adaptive evolution . . . is fundamentally the same whether it starts with genetic or environmental induction” (West-Eberhard 2003, 144). In fact, she argues that environmentally-induced changes are more significant than genetically-induced ones, because they are more likely to be recurrent and so they get incorporated into development more often. But regardless of whether a novel phenotype is induced genetically or environmentally, West-Eberhard distinguishes between the initiation and the origin of novelty. On her account, the origin of novelty requires two events: (i) initiation by a new input in the form of a genetic mutation or an environmental change, and (ii) a developmental response that produces a new phenotype. In other words, genetic or environmental changes merely provide the initial input, to which the organism responds, and it is this response that is the origin of evolutionary novelty. This is a considerable conceptual shift from the Modern Synthesis account. Not only the effects of environment are brought into light, the effects of genes are placed on a par with them as mere inputs. It is ultimately the organism’s response to these internal and external inputs that results in the novel phenotype. West-Eberhard thus proposes that adaptive evolution involves the following events: (i) trait origin, or phenotypic novelty due to genetic or environmental input, (ii) phenotypic accommodation, or the adaptive adjustment of various phenotypes in the organism to accommodate the phenotypic novelty (as seen in the example of the two-legged goat), (iii) initial spread due to the recurrence of the initiating factor, and finally (iv) genetic accommodation, or gene-frequency change due to selection (West-Eberhard 2003, 140). West-Eberhard then draws our attention to what is striking about this account: “gene- frequency change follows, rather than initiates, the evolution of adaptive traits” (West-

Eberhard 2003, 158). Genes are followers rather than leaders, and their most important role is not so much in the origin of novelty as in making a store of genetic variation available for gradual change under selection. What is significant about this result is that on this account, adaptation does not wait for lucky mutations to come along. The novel phenotypes that are available for selection result from developmental plasticity of an already highly adapted organism, which ameliorates many negative effects. This means that novel phenotypes are “not completely random with respect to adaptation,” even though their inducing factors may be (West-Eberhard, 158). Thus, on West-Eberhard’s account, evolutionary novelty does not consist in the random mutation of genes, but is rather biased by the adaptive response of development to the organism’s conditions. This is why, it is the organism—and not its genes— that is viewed as the principal cause of evolutionary novelty and thereby the driver of evolutionary change. This overview of West-Eberhard’s account of evolution is meant to illustrate how recent approaches to understanding evolution challenge the Modern Synthesis in ways that suggest there can be room for an explanatory concept of organism in biology. The Evo-Devo account gives an important explanatory role to phenotypic plasticity, which is a property of whole organisms and cannot be reduced in terms of the molecular properties of genes or statistical properties of populations. Admittedly, the argument I have presented here is schematic, and only includes part of the story. I have not discussed many other empirical and theoretical advances in the understanding of evolution that have called for “the return of the organism as a fundamental explanatory concept in biology” (Nicholson 2014).27 But if neo-Aristotelian naturalists aim to defend the life-form concept by appealing to the explanatory role of the concept of organism in evolutionary biology, the Evo-Devo account of evolution seems to provide a promising starting point.

27 See also Walsh (2016) for a particularly thorough argument for an organocentric view of evolution.

3.4 In Search of the Concept of Organism in Biology

Although Evo-Devo is rapidly growing as a field of biological research, the implications of its findings for evolutionary biology still remain contentious. The exact role of development in evolutionary explanations is currently a hotly debated issue, as can be seen a 2014 exchange in Nature about whether evolutionary biology needs a revision. While some biologists respond by an emphatic “Yes, Urgently,” others still do not feel such an urge and respond “No, All Is Well” (Laland et al. 2014). Some biologists are skeptical of the claims of Evo-Devo for reasons having to do with the insufficiency of empirical evidence. They argue that few cases of phenomena like Waddington’s genetic assimilation have been documented outside the laboratory, and the evidence for the role of phenotypic plasticity in evolution has to be strengthened. Some biologists and philosophers alternatively question the theoretical significance of Evo-Devo for evolutionary explanation. They argue that although development influences a range of traits that natural selection can act on, what matters for evolutionary explanation is ultimately the heritable difference in traits, not the extent of trait variation or how they are caused. Thus, they insist that gene centrism is still “the most powerfully predictive, broadly applicable and empirically validated component of evolutionary theory” (Laland et al., 163). The idea that evolutionary explanation does not have to cite the developmental causes of variation can be traced back to Ernst Mayr’s distinction between proximate and ultimate causes. Proximate causes are the immediate and mechanical causes that mediate between genotypes and phenotypes, while ultimate causes are responsible for long-term change from one population to the next. Mayr (1961) recognizes two distinct sub-disciplines of biology. The first is functional biology, which studies “the operation and interaction of structural elements, from molecules up to organs and whole individuals”. In other words, it is concerned with the ‘how’ question of proximate causes. The second is evolutionary biology, which studies “the causes for the existing characteristics, and particularly adaptations, of organisms”. These are Mayr’s ultimate causes, which answer the ‘why’ question and include natural selection and

62 other evolutionary processes such as drift (Mayr 1961, 1502). The idea is that to the extent that evolutionary explanations are not asking the ‘how’ question, they are not concerned with the proximate causes of development, which are just gory mechanistic detail. Proponents of the organocentric approach have questioned the extent to which the proximate-ultimate distinction and the separation of development from evolutionary explanation can be maintained in light of the findings of Evo-Devo (Laland et al. 2013). Walsh (2015), for instance, explicitly argues against what he calls the ‘fractionation’ of evolution, which consists of thinking of the components of evolution such as development, adaptive change, inheritance, and the generation of novelty as discrete and autonomous. He argues that these processes don’t have separate and distinctive causes, and cannot be observed and studied in isolation from one another (Walsh, 159). But this is an ongoing debate that may need further empirical and theoretical investigation to be resolved. What is far more widely accepted, however, is the explanatory role of phenotypic plasticity and other distinctive properties of organisms in developmental biology. Even if it turns out that organisms can be black-boxed in the study of evolutionary change, research in Developmental Systems Theory suggests that what happens during development cannot be explained solely in terms of the properties of genes (see Oyama 2000; Oyama et al. 2001). It is increasingly accepted is that the developmental process is not merely the unfolding of a genetic program, and the properties of an organism as a unified whole play a significant role in development (Nicholson 2013; 2018). If this view of development is correct, then at the very least, the concept of organism has a place in developmental biology. The evolutionary objection to neo-Aristotelian naturalism assumes that evolutionary biology is the only relevant area of biology where organisms could possibly be found. Lott (2012a) and Hacker-Wright’s (2009a) response to this objection has similarly assumed that evolutionary biology would be the empirical science that presupposes the neo-Aristotelian concept of a life-form. But this focus on evolutionary biology is unnecessary. As I have argued in my discussion of Evo-Devo in section 3.3, there might be a case to make for the role of the concept of organism in evolutionary biology. But even failing

63 that, there is reason to think the organism can be found alive and well in the field of developmental biology, where it has a much more well-established place, and can potentially lend support to the neo-Aristotelian concept of life-form. Of course, arguing that the concept of organism plays an explanatory role in developmental biology is not enough to show that developmental biology presupposes the neo- Aristotelian concept of life-form, which involves a particular conception of living organisms. As we saw in Chapter 1, the concept of life-form denotes the form of life of an organism, and neo-Aristotelians maintain that this form of life can be articulated in a particular, generic form of thought manifested in natural-historical judgments. Because of this, the neo-Aristotelian concept of life-form involves a commitment to the ascription of a characteristic flourishing and flourishing-based functions to an organism and its parts and aspects. So just because the concept of organism plays an explanatory role in developmental biology, it doesn’t immediately follow that the particular, neo-Aristotelian conception of a living organism is correct. There is a further question whether the explanatory concept of organism in developmental biology or other branches of biological science commits us to a suitable notion of flourishing and flourishing-based function. Thus, the arguments of this chapter merely provide a rudimentary starting point for defending the commitments of neo-Aristotelian naturalism by appealing to biology. However, in the next chapter, I offer a more fully- developed defence of these commitments by particularly focusing on the concept of function in biology. I argue that the evolutionary concept of function—that has been at the core of the Selfish Gene objection—does not capture all instances of functional explanation in the various branches of biology. I appeal to an alternative and largely overlooked account of biological function to argue that the neo-Aristotelian, flourishing-based concept of function is explanatorily indispensable to biology and thus necessary for understanding the nature of living things.

Chapter 4

From Biological Functions to Natural Goodness

In Chapter 2, we saw that the commitment of neo-Aristotelian naturalism to function ascriptions based on the concept of the flourishing of an organism has been the target of the Selfish Gene objection. Critics have argued that biological functions are determined based on the evolutionary contribution of genes to natural selection, and have nothing to do with an organism’s welfare or flourishing. In this chapter, I offer a novel defense of the neo- Aristotelian, flourishing-based concept of function against this objection. I appeal to a recent account of biological function that has been overlooked in the critical discussion of neo- Aristotelian naturalism to argue that flourishing-based functions are explanatorily indispensable to biology and therefore essential to the understanding of living things.

4.1 Toward a Novel Defence of the Grounding Premise

Let’s go back to the neo-Aristotelian argument for ethical naturalism from Chapter 2 one more time:

Refined Argument for Naturalism

1. Norms of natural goodness in non-human life-forms are natural. (Grounding) 2. If norms of natural goodness in non-human life-forms are natural, then norms of

natural goodness in the human life-form are also natural. (Continuity) 3. Norms of moral goodness are instances of norms of natural goodness in the

human life-form. (Normative Adequacy) C. Therefore, norms of moral goodness are natural.

We saw that the evolutionary objection ultimately challenges the premise of Grounding in this argument. On the neo-Aristotelian account, evaluations of natural goodness are based on the

65 function or the role an aspect of the organism serves in enabling it to flourish. But critics argue that such flourishing-based ascriptions of function have no place in biology (Fitzpatrick 2000; Lewens 2010; Odenbaugh 2017), and replacing them with the function ascriptions that do have a place in biology leads to an implausible account of moral virtue (Millgram 2009; Andreou 2006; Woodcock 2006; Millum 2006; Odenbaugh 2017). Fitzpatrick (2000), for instance, gives an evolutionary account of biological function in terms of gene replication, and argues that contrary to what neo-Aristotelians claim, biological functions have nothing to do with an organism’s welfare or flourishing. In Chapter 2, I argued that the best strategy for defending the premise of Grounding in the neo-Aristotelian argument for ethical naturalism is showing that the flourishing-based concept of function—which is the basis of evaluations of natural goodness—is necessary for understanding non-human living things. In Chapter 3, I examined Lott (2012a) and Hacker- Wright’s (2009a) attempt at making such an argument by examining the concept of organism in evolutionary biology. I argued that while the explanatory concept of organism in Evo-Devo and developmental biology provides a promising starting point, there is a further question whether it commits us to a suitable notion of flourishing and flourishing-based function. In this chapter, I attempt to answer this question by looking at philosophical accounts of biological function. I argue that there is a kind of functional ascription in biology—based on the contribution of traits to the self-maintenance of an organism—that is best understood as presupposing the concept of the flourishing of an organism. And I suggest that this kind of function ascription instantiates the broader concept of function that underwrites the neo- Aristotelian account of natural goodness.

My arguments in this chapter improve on the existing literature on neo-Aristotelian naturalism in two ways. First, both neo-Aristotelians and their critics have generally assumed that the only respectable account of functional explanation in biology is the etiological account,

66 according to which the function of a trait is given by its historical role in natural selection.28 I argue that the etiological account of function does not capture all cases of functional explanation in biology. I suggest that a more recent alternative—namely, the organizational account—brings out another important concept of function in biology, which affords different prospects for assessing the claims of neo-Aristotelian naturalism. Second, I explain how an argument for neo-Aristotelian naturalism can draw from the explanatory concept of function in biology without giving an undue place to biology in substantive moral reasoning. Authors engaging in this debate have largely assumed that if natural goodness is understood in terms of a concept of function that is used in biology, it must follow that biology can give us the norms of natural goodness, which supposedly include the norms of virtue and reason in the case of human beings. Such a result would, of course, be problematic for neo-Aristotelian naturalism. As we saw in Chapter 2, many critics have argued that the idea that knowledge of moral virtue can come from biology would likely result in intuitively implausible first-order claims about moral virtue (Millgram 2009; Andreou 2006; Woodcock 2006; Millum 2006; Odenbaugh 2017). Moreover, it would imply that we need to radically revise our ordinary methods of moral reasoning, since these methods do not normally consist of deriving moral reasons from biology. However, I argue that although organizational function ascriptions in biology instantiate the concept of function underlying the neo- Aristotelian account of natural goodness, the science of biology is not in a place to give us the full extension of this concept of function. In a nutshell, this is because the organizational concept of function presupposes the concept of the flourishing of an organism. Ascribing organizational functions requires a conception of flourishing or living well for the organism, a full account of which falls outside the scope of biology. Thus, I argue that the explanatory role

28 Fitzpatrick’s account of function, for instance, is essentially an etiological account. Despite having important differences with standard etiological theories (see Fitzpatrick 2000, 234-246), his account is etiological in that he takes gene replication to be the ultimate biological end which determines functions, and defends this choice by appealing to the role of gene replication in explaining natural selection. Odenbaugh’s (2017) version of the objection from biological functions is another example, which explicitly appeals to Godfrey-Smith’s (1994) etiological theory of function.

67 of organizational function ascriptions in biology can support the neo-Aristotelian concept of function without implying that biology can give us a substantive account of moral virtue.

4.2 Biological Functions Revisited

The classic philosophical accounts of biological function fall into two main approaches. The first one is the dispositional approach (Cummins 1975; Adams 1979), according to which functions are determined by the current causal contribution of the function-bearer to a goal or capacity of a larger system. This approach is often put aside on the charge of being too broad and under-specified. On the one hand, it seems to allow for an arbitrary, subjective attribution of functions depending on which capacities of a system interest us. On the other hand, it doesn’t capture the explanatory force of function ascriptions—the idea that the function of a trait explains something about the trait such as why it exists or does what it does. A related criticism of the dispositional approach is that it struggles to make sense of malfunction, and that it fails to differentiate between functioning and mere usefulness. The problem is that the dispositional account defines the function of a trait based on its actual contribution. So when a trait makes no contribution to a goal or capacity of the system, it’s hard to see how the account can detect a malfunction instead of simply not ascribing a function. And when a trait does make such a contribution, there seems to be no way to differentiate between a functional and a merely accidental such contribution.29 The second approach is the etiological view (Wright 1973; Millikan 1989; Neander 1991, Godfrey-Smith 1994), which appeals to the causal history of a trait to overcome the excessive breadth of the dispositional account. On the etiological account, the function of a trait is the

29 That said, introducing the type/token distinction enables dispositional theorists to draw the relevant distinctions. A token trait can malfunction when it doesn’t do what other tokens of its type do, and it can accidentally contribute to the goal of the system when its contribution is not the same as the contribution of normal tokens of its type. There are obviously questions about how types can be individuated, or how often tokens of the type have to contribute to a goal or capacity to ground a function. But these are part of the general problem of under-specification with the dispositional account and do not necessarily amount to a distinct problem for the view.

68 contribution for which it has been evolved by natural selection. In other words, it is the contribution of the trait that historically explains its current presence or prevalence. To use the classic example of the heart, the function of a human heart is to pump blood, because pumping blood is what historically has contributed to the natural selection of humans with a heart, and thus explains the current presence of human hearts. In this way, the etiological account overcomes the main problems of the dispositional account. It specifies natural selection as the proper goal relative to which functions are determined. And it captures the explanatory force of function ascriptions in terms of the explanation of a trait by its natural-selection evolutionary history. Furthermore, since it allows the function of a trait to come apart from its current contribution, it has no problem making sense of malfunction or drawing the distinction between functioning and mere usefulness. However, the etiological account has also faced a lot of criticism. One of the main objections against this account is that it is too narrow to accommodate all instances of functional talk in biology. It defines functions merely based on the past contribution of a trait to natural selection, and makes the current contribution of a trait irrelevant to determining its functions. Although past contribution to natural selection is the relevant explanatory factor in evolutionary biology, there are areas of biology with different explanatory needs. As many philosophers of biology have argued, in areas of biology such as physiology, functional anatomy, and neuroscience—where evolutionary explanation is not the main concern— function ascriptions are primarily related to the current effects of traits rather than their evolutionary past (Walsh and Ariew 2010; Roe and Murphy 2011; and Kraemer 2014). Moreover, even within evolutionary biology, the etiological account might not be adequate for capturing the concept of function in certain more recent research programs. As Krohs (2011) has argued, for instance, the etiological account relies on the theoretical framework of gene deterministic adaptationism. This framework, however, has been rigorously criticized for focusing on natural selection as the sole or primary evolutionary mechanism and ignoring the role of developmental and epigenetic factors in evolutionary change. Present day evolutionary

69 biology features new research programs such as EvoDevo and Eco-EvoDevo, which seem to ascribe functions based on different criteria and regardless of whether a trait has an adaptive origins.30 To address the shortcomings of the dispositional and the etiological views, an alternative account of biological function has recently been proposed. This is the system-theoretic or organizational account of function, which identifies functions based on what contributes to the maintenance of a system’s organization (Schlosser 1998; McLaughlin 2001; Christensen and Bickhard 2002; Weber 2005; Mossio et al. 2009; Moreno and Mossio 2015)31. Interestingly, the organizational approach has emerged out of various attempts to give a scientific account of the nature of life (Kauffman 1993; Barandiaran, Di Paolo, and Rohdel 2009; Di Paolo 2005).32 Most earlier theories of function try to give a unified account that captures not only biological functions, but also other kinds of function such as artifactual or intentional functions. But the organizational account is particularly aimed at the distinctive use of function ascriptions in relation to biological systems, which are characterized as self-organizing and self-maintaining autonomous systems. A self-maintaining system maintains itself by producing at least some of the necessary conditions for its own existence. It consists of parts and processes that contribute to keeping its systemic organization intact and at the same time need to be constantly regenerated and repaired by the organization.33 Living things are clearly self-maintaining systems that maintain themselves via the contribution of their vital organs and operations. On the organizational account of function, it is this self-maintaining character of living things that

30 See Krohs 2011, 129-131 for a discussion of the critique of adaptationism and why the etiological account is inadequate to capture functional ascriptions in post-adaptationist evolutionary biology. 31 I follow Garson (2016; 2017) in grouping this somewhat diverse cluster of theories under the general heading of the organizational approach. 32 See Nicholson (2014, 8-10) for an account of how a revived interest in the nature of life as a genuine scientific problem has led to the organizational approach to understanding biological functions. 33 Different advocates of the organizational approach use different terminology to refer to self-maintaining systems. Here I follow the terminology adopted by Mossio et al. (2009), which I find to be the most appropriate label. But, as it will become clear in the next section, I do not embrace certain other aspects of Mossio et al.’s (2009) version of organizational account.

70 underwrites function ascription. Although different versions of the organizational account flesh out the details in different ways, the general idea is that a part or trait in a self- maintaining system acquires a function by contributing to the working of the system and thereby contributing to its own persistence within the system. Thus, the organizational function of a trait T in a self-maintaining system S is, roughly, the contribution of T to the self- maintenance of S. Consider the example of the heart again. The beating of the heart circulates blood, which carries nutrients to the cells and eliminates waste. In doing so, it helps to repair and regenerate the cells of the body, including the cells that make up the heart itself. Thus, the heart contributes to the maintenance of the whole organism, which in turn contributes to the persistence of the heart within the organism. On the organizational account, this is what grounds the fact that the heart has the function of circulating blood. It is not the contribution of heart tokens to the intergenerational reproduction of their type—as the etiological view would have it—but rather their contribution to their own intragenerational persistence as tokens that determines their function.

An influential defence of several aspects of the organizational approach is offered in McLaughlin’s (2001) account of functional explanation.34 McLaughlin argues that we can capture the explanatory force of function ascriptions without appealing to natural selection. On his view, contrary to what the advocates of the etiological approach typically assume, the historical process of natural selection is not the only mechanism that explains why a functional trait exists and does what it does. There is also the feedback mechanism constitutive of a self- maintaining system, which runs within the individual organism.35 In fact, the explanatory scope of this feedback mechanism extends beyond that of natural selection. Natural selection

34 Although McLaughlin doesn’t use the terminology I have adopted here, his account of function fits very well within the general framework that I am calling the organizational approach. 35 Using a different terminology, McLaughlin refers to self-maintaining systems as “self-reproducing systems.” What he means by “self-reproduction” though is the constant re-production and regeneration of the same token of the system, which I believe is less ambiguously captured by “self-maintenance” (see McLaughlin 2001, 211).

71 can only explain how the effects of a trait token contribute to the nature of future tokens of the same type. In contrast, the constitutive organization of a self-maintaining system can explain how the effect of a trait token can contribute to the nature of that very token (2001, 162-164). A self-maintaining system preserves its identity and integrity by constantly regenerating and repairing itself via a cycle of mutual dependence relationships between the whole system and its parts. This means that the contribution that any part makes to maintaining the system explains the presence of that part itself. We can now see how the organizational account addresses the main issues faced by the two classical approaches. Not only does it capture the explanatory force of functions, it overcomes the problem of under-specification for the dispositional account by identifying a specific goal relative to which the functions are determined. What underwrites ascribing organizational functions is a self-maintaining system’s capacity for self-maintenance, which— according to a growing literature on the subject—is a constitutive feature of living systems essential to their nature as living things (Varela, Maturana, and Uribe 1974; Rosen 1991; Gánti 2003; Nicholson 2013). So the ascription of organizational functions is not arbitrary or merely based on the capacities of a system that interest us, but rather specifically the effect of a trait that explains its nature as part of a self-maintaining system.36 Moreover, unlike the etiological account, the organizational account has no problem capturing the relevance of the current contribution of a trait, since it defines functions based on contribution to the self-maintenance of the current system. It is not committed to the theoretical framework of adaptationism, and does not restrict functional traits to evolutionary adaptations. As such, it offers a

36 It may seem as though by defining functions based on the current contribution of a trait, the organizational account faces the same problems as the dispositional account with respect to capturing malfunctions or making the function/accident distinction. In the next section, I will argue that this is in fact a problem for a certain reductive interpretation of the organizational account. But the version of the organizational account that I will defend does not face this problem, as it presupposes a conception of flourishing of the organism that can be the basis for making the relevant distinctions.

72 complementary account of function that is suitable for post-adaptationist evolutionary biology as well as areas of biology not directly focused on evolutionary explanation.37 While I have not offered a conclusive defense of the organizational account in this section, the preceding review should suffice to motivate the idea that it offers a promising account of an important concept of function in biology. To be clear, it is not my intention to claim that the organizational account exhausts the talk of functions in biology, or that it should replace the etiological account of biological function. My claim is rather that it captures a concept of function in that plays an important role in the scientific study of life, and is—to that extent—necessary for understanding the nature of living things.38

4.3 Organizational Functions as Flourishing-Based Functions

The task now is to see if the organizational account does any better than the etiological account as a basis for evaluations of natural goodness. The problem with understanding natural goodness in terms of the etiological concept of function was that etiological functions are determined based on the contribution a trait makes to natural selection, which can be at odds with the flourishing of an organism. Initially, it may seem that understanding natural goodness in terms of the organizational concept of function yields equally problematic results. Organizational functions are determined based on contribution to self-maintenance, and it may seem that what contributes to self-maintenance can be similarly at odds with the flourishing of an organism. Particularly in the human case, where the neo-Aristotelian conception of flourishing is tied to moral virtue, it’s not clear that what contributes to self-maintenance would always line up with flourishing. It may turn out, for instance, that under certain

37 Nunes-Neto, Moreno, and El-Hani (2014) argue, for instance, that the organizational approach offers a suitable account of function ascriptions in ecology. 38 It’s worth noting that some proponents of the organizational approach do aim to offer an exhaustive account covering all instances of function ascription in biology. As Garson (2017, 1094) has noted, one way to do this would be framing the account as a disjunctive theory, defining functions in terms of the contribution to either the intragenerational persistence or the intergenerational multiplication of the system. For my purposes, however, it doesn’t matter if the account is exhaustive or merely covers one indispensable concept of function in biology.

73 conditions, what best serves a human’s survival would be violence, deception, and selfishness— character traits that are clearly at odds with the neo-Aristotelian conception of human flourishing. However, we will see that this concern is based on a misconception of self-maintenance as merely consisting of the persistence of a certain kind of causal dynamic. I will argue that, despite the best efforts of many organizational theorists, the notion of self-maintenance required for organizational function ascription cannot be causally reduced. On my account, this notion presupposes a conception of the welfare of the self-maintaining system, which cannot be characterized in purely causal terms. Thus, unlike etiological functions—which can be simply determined based on causal facts about natural-selection history—organizational functions cannot be determined without a prior grasp of what constitutes welfare for the organism. And it’s not at all obvious that given the right conception of human welfare, the organizational functions that are instantiated in human beings can conflict with moral virtue. Many advocates of the organizational approach have tried to delineate the criteria for organizational function ascription in purely causal terms. However, as I argue below, a merely causal characterization of the account leads to an overgeneralization problem. Consider the organizational account offered by Mossio et al. (2009) and Moreno and Mossio (2015), which is the most fully developed instance of this reductive approach. On this account, organizational functions are defined based on a trait’s causal contribution to the maintenance of a self- maintaining system, where a self-maintaining system is characterized in terms of two causal conditions: organizational closure and organizational differentiation. Organizational closure is a circular causal relation between some higher-level macroscopic pattern or structure and the lower-level microscopic dynamics and reactions within the system. This circular causality makes the activity of the system a necessary (although not sufficient) condition for the continuous existence of the system itself. Organizational differentiation involves realizing organizational closure in a particularly complex way. An organizationally differentiated system generates different and localizable patterns and structures each making a distinct contribution

74 to the whole organization.39 According to Mossio et al. (2009), biological systems are organizationally closed and differentiated systems, and that is precisely why they are subject to functional analysis. However, we can see that this account extends function ascriptions to domains where functional discourse is not warranted. As Garson (2017) has recently argued, there are structures that meet the conditions of organizational closure and differentiation, and yet do not underwrite the relevant kind of function ascription. Garson offers the example of panic disorder. He argues that panic attacks, which result from catastrophizing misinterpretations of bodily sensations, involve many mechanisms that set the stage for their own recurrence. These are psychological states and behavioral dispositions that are caused by the disorder and that each make distinct contributions to its persistence.40 Nonetheless, it would be a mistake to ascribe organizational functions to these mechanisms in virtue of their contribution to the persistence of panic disorder. As Garson points out, normal medical practice almost universally characterizes panic disorder as involving cognitive or biological dysfunctions (e.g., Ludewig et al. 2005).41

The overgeneralization problem results from the attempt to ground organizational functions in merely causal criteria, which turn out to be instantiated in non-functional domains

39 The condition of organizational closure turns out to be a ubiquitous property in the physical world, manifested in many non-living systems such as hurricanes and candle flames. Mossio et al. introduce the second condition to avoid ascribing organizational functions to non-living systems where the relevant concept of function does not apply (2009, 826). 40 For instance, a single attack makes the person worried about having another and thus more vigilant to future bodily sensations. The person tends to avoid the particular situation that brought on the attack, which results in having fewer chances for disconfirming false beliefs about the experience, and so on. (See Garson 2017, 1096-1098 for a detailed account.) 41 Of course, scientists can choose to use functional analysis to study panic disorder if a certain feature of the phenomenon such as the tendency of the attacks to perpetuate themselves happens to interest them. As I mentioned earlier, the broad dispositional concept of function can be used to ascribe a function to any part of system that contributes to some interesting capacity of the system. But the organizational concept of function is intended to capture a different, explanatory use of function ascription that is distinctive of living systems, and this specific concept of function clearly does not apply in the case of panic disorders.

75 as well.42 This is why philosophers of biology like Mark Bedau (1992) and Peter McLaughlin

(2001) have defended a value-based approach to understanding functions. The reason we don’t ascribe functions to aspects of a panic disorder is not a lack of differentiation or complexity of organization, but rather that no good comes from the behavior of such systems. Bedau effectively argues that to avoid overgeneralization, any account of function needs to implicitly or explicitly include a value criterion, which minimally requires that the function involves a good consequence for someone or something. McLaughlin similarly advocates a value-based, ‘welfare’ approach, according to which function ascription implies the existence of a system that has a good or welfare, and so can benefit from the functions. More specifically in the context of organizational functions, McLaughlin argues that the beneficiary is the self- maintaining system itself, of which the functional trait is a part. He thus incorporates a welfare criterion in his conception of a self-maintaining system. On his account, a self-maintaining system has a welfare of its own, and benefits from its own self-maintenance (2001, 109-204). It might be objected that this reference to welfare disqualifies the organizational account from being a scientifically respectable account of function. However, despite the prevalence of the idea that we must somehow sanitize functions by removing any reference to value, it’s not clear why such a reference is necessarily unacceptable. If we cannot account for functional explanations in biology without such a reference, then commitment to an irreducible notion of value seems inevitable. This is exactly what Bedau and McLaughlin have argued. In order to properly account for functional explanation, we need to presuppose the existence of beneficiaries with a welfare of their own. To put it in McLaughlin’s terms, the commitment to the idea that biological systems have a welfare is “the metaphysical price” of embracing functional explanations in biology. And while philosophers may analyze and clarify this metaphysical cost, it’s ultimately up to empirical scientists to decide whether we ought to “pay the bill.” Philosophers have tried to avoid the supposedly offensive reference to value by giving

42 The problem is not specific to the particular criteria delineated by Mossio et al. (2009). As Garson argues, another prominent reductive version of the organizational account (Schlosser 1998; Weber 2005), which focuses on complexity instead of differentiation, falls prey to the same objection.

76 a stipulative definition of function, e.g., in terms of organizational closure and differentiation.

But it’s not clear what can be gained from such a definition if it doesn’t capture the biologists’ concept. Unless biologists explicitly accept a value-neutral theoretical definition of organizational functions and adhere to it in the face of putative counterexamples, they are committed to the notion of a self-maintaining system as a system with a welfare of its own (McLaughlin 2001, 212). Thus, to avoid the overgeneralization problem, the organizational account needs a conception of the welfare of the system that cannot be reduced to causal dynamics. This element of welfare in the organizational account means that character traits like violence and deception cannot be ascribed an organizational function unless they contribute to human welfare, which, in all likelihood, they do not. One might object that the notion of welfare required for ascribing organizational functions is a minimal, strictly biological notion of welfare that has nothing to do with flourishing in the neo-Aristotelian sense. ‘Biological welfare’ seems to consist in something very specific, namely the system’s self-maintenance in the sense of maintaining its constitutive organization. So although we need to presuppose that a biological system benefits from the organizational functioning of its part, we might be able to give a causal account of what the system’s benefit consists in: the system’s continued existence. And if this is the case, there seems to be no reason to think that organizational functions in human beings have to be consistent with moral virtue. So, again, character traits like violence and deception can turn out to contribute to biological welfare, which would mean that organizational functions are not a suitable basis for a neo-Aristotelian account of natural goodness.

However, a closer look will reveal that what we need to presuppose in organizational function ascription is not just that the system has a welfare, but also a substantive conception of what this welfare consists in. We saw earlier that a purely causal account of organizational functions in terms of criteria like closure and differentiation leads to overgeneralization. To avoid this problem, we need to add a welfare condition to the account, which requires that the

77 system encompassing the functional traits has a welfare and benefits from the functions. But just as the fact that the system has a welfare cannot be deduced from a purely causal description, what constitutes welfare for the system cannot be deduced from causal criteria either. The notion of welfare that is presupposed in organizational function ascription is in fact tied to the self-maintenance of the system. But looking at the organizational concept of malfunction reveals that, even when we limit the domain of discourse to systems that truly have a welfare, we cannot characterize the welfare of these systems in terms of purely causal criteria. As I argue below, the mere fact that a beneficiary system is preserved by a cycle of organizational closure is not sufficient for it to be doing well, and how it maintains itself matters. First, note that the possibility of organizational malfunction means that a functional trait’s contribution to the system’s current cycle of organizational closure is not sufficient for its well-functioning. This implies that the mere persistence of a cycle of organizational closure is not sufficient for the system to be doing well. If all that was required for welfare was maintaining organizational closure, then any trait that contributed to organizational closure would be well-functioning, while any trait that didn’t contribute to organizational closure would not have a function at all.43 But this implication is unacceptable. Surely a heart can malfunction even when it continues to circulate blood, e.g., when it beats irregularly. So to capture the possibility of malfunction, the organizational account needs a conception of welfare that goes beyond the mere maintenance of a causal organization. This conception of welfare has to allow for the system’s constitutive organization to be maintained well or poorly, such that welfare consists in the system’s maintaining itself well.

It might be thought that what it is for an organism to maintain itself well can be understood causally in terms of the effectiveness or stability of maintaining the organization. A malfunctioning heart may not result in a breakdown of the organization, but it probably makes

43 This is because on the organizational account, functionality requires making a contribution to the maintenance of the organization.

78 maintaining the organization less efficient or increases the risk of such a breakdown. This idea is explored in Saborido and Moreno’s (2015) account of organizational malfunction, according to which a malfunctional trait diminishes the overall range of viability of the system.44 However, defining malfunction in terms of its effect on the system’s range of viability requires identifying a benchmark for the range of viability that is expected of the system. In order to see if the activity of a trait reduces the range of viability, we need a prior grasp of the range of circumstances in which the system is supposed to be viable. But it’s not clear how this benchmark can be identified based on merely causal criteria. The system’s range of viability at an earlier time isn’t necessarily a good benchmark, because the system may have been malfunctioning from the beginning. The reference point cannot be an ideal system with an infinite range of viability either, since no organism can be expected to be viable under all possible circumstances. Even when every part of the system is functioning properly, the range of circumstances in which the system can maintain itself will be limited. For instance, a perfectly healthy human is not expected to be able to breathe under water like the fish do. A sloth is not expected to be able to run as fast as a deer, and so on. Thus, a system’s having a range of viability that is narrower than it would have been if its traits were working differently doesn’t necessarily mean that there is any malfunction.45

44 On this account, for each self-maintaining system, there are several specific instances of organizational closure—or regimes of self-maintenance—that it may adopt. The functional traits and organs operate with a specific range of activity in each of these regimes. And switching between these different regimes by way of a regulatory subsystem enables the system to maintain its viability in the face of internal and external change. A functional trait isn’t necessarily indispensable for the organism’s life, because not every functional trait contributes to all possible regimes of self-maintenance. When a component of the system breaks down, the regulatory subsystem can compensate by shifting to a different regime that doesn’t require the contribution of that component. But the system’s overall range of viability might be reduced as a result, because the alternative regime might not be viable in as many circumstances. This, on Saborido and Moreno’s account, is when a malfunction occurs. 45 It may be thought that the benchmark should be the maximum range of viability that could result from the working of each trait given the constraints posed by the material structure of the rest of the system. But several parts of the system could be malfunctioning at the same time, and the current structure of the rest of the system could already be defective. This means for any constraint posed by the material structure of the system, it’s underdetermined whether it is a legitimate constraint or one that involves a malfunction.

Moreover, the issue is not merely identifying a threshold for the minimum required degree of viability for the organism to count as doing well. Having a higher overall degree of viability does not always mean that the organism is maintaining itself better. In fact, if the higher degree of viability comes at the cost of loss of certain abilities, the organism may very well be doing worse overall. An interesting example can be found in Laron Syndrome, an autosomal recessive disorder that is characterized by an insensitivity to growth hormone resulting in dwarfism and obesity. Although individuals with Laron Syndrome are prone to a number of neurological and metabolic problems, studies suggest that they not only have a normal life span, but also a significantly reduced risk of cancer and type 2 diabetes (Shevah and Laron 2007; Guevara-Aguirre et al. 2011; Bartke, Sun, and Longo 2013; Laron et al. 2017). It’s precisely the insensitivity to growth hormone that offers protection from major age-related diseases, which arguably makes a positive contribution to viability. Nonetheless, it is not considered a well-functioning of the receptor proteins to resist binding to growth hormone and protect body from aging. The receptor’s inability to bind to growth hormone is unequivocally treated as a pathological condition and a malfunction. This example shows that viability is not the only consideration that is relevant in determining whether a trait is functioning well or poorly. What it is for the organism to maintain itself well seems to go beyond achieving a high degree of viability. Other considerations having to do with what the organism is supposed to be able to do seem to put constraints on what counts as a good manner of achieving viability for a given kind of organism. This helps to see why the relevant notion of welfare would not be reducible in terms of the abstract idea of effectiveness or stability of maintaining an organization.

Absent a successful reductive account, I propose that organizational function ascription presupposes a substantive conception of what it is for the organism to maintain itself well, which in turn presupposes a conception of what it is for the organism to do well in general. Organizational functions are self-maintaining functions. They specifically track the contribution of traits to the self-maintenance of the organism. But what constitutes self-

80 maintenance for an organism depends on what constitutes doing well for the kind of organism that it is. I propose that the neo-Aristotelian notion of flourishing captures this kind-specific conception of doing well. Thus, I suggest that self-maintenance should be understood as an aspect of the organism’s flourishing, and in the context of other aspects of flourishing such as development, reproduction, and in the human case, rational agency. The example of Laron Syndrome demonstrates how these other aspects of flourishing can put constraints on what is an acceptable manner of self-maintenance for the organism. It’s because the resistance of growth hormone receptors in individuals with Laron Syndrome hinders other aspects of their flourishing such as growth and development that its positive effect on viability does not amount to better organizational functioning. Similarly, even if character traits such as violence or deception increase viability, it doesn’t immediately follow that they properly contribute to self-maintenance or are well-functioning character traits. Organizational functions depend on what constitutes flourishing for the organism, which essentially makes them instances of flourishing-based functions. So, organizational functions in human beings cannot be determined without a prior grasp of human flourishing. And if, as neo-Aristotelians tell us, human flourishing requires a life of virtue, no character trait that conflicts with such a life would be well-functioning on the organizational account.

4.4 From Organizational Functions to Natural Goodness

We saw, in Chapter 2, that the evolutionary objection challenges the idea that the flourishing- based concept of function is necessary for understanding living things, which in turn undermines the thesis that flourishing-based evaluations of natural goodness apply to living things on account of their nature. Earlier in this chapter, I argued that this objection is based on the assumption that the only respectable account of functional explanation in biology is the etiological account. And I showed that the organizational account captures a class of biological functions distinct from etiological functions that have been overlooked in the conversation about neo-Aristotelian naturalism. Then, I argued that organizational functions are best

81 understood as instances of flourishing-based functions. Now we can see how neo-Aristotelians can respond to the evolutionary objection by appealing to the organizational concept of function in biology. To the extent that organizational functions are part of a full scientific account of living things, the flourishing-based concept of function is necessary for understanding living things. This means that the neo-Aristotelian evaluations of natural goodness are based on a concept of function that is essential for grasping the nature of living things. I have thus offered a novel defense of the premise of Grounding by arguing that the concept of function underlying evaluations of natural goodness is not only used in folk representations of life but also explanatorily indispensable to modern biology. In a way, this is what neo-Aristotelians like Lott and Hacker-Wright have been claiming all along. As we saw earlier, Lott (2012a) and Hacker-Wright (2009a) claim that even the science of biology has to presuppose neo-Aristotelian function ascriptions. But for Hacker-Wright and Lott, this is an a priori claim made independently of the actual content or explanatory needs of empirical biological science. In contrast, my argument shows that we can in fact locate the neo-

Aristotelian concept of function in a class of functional explanations in biology. It may be objected that if natural goodness is understood in terms of a concept of function that is used in biology, then biologists would be in a position to study evaluations of natural goodness, including the norms of virtue and reason. But as we saw earlier, the implication that knowledge of moral virtue can come from biology would be problematic. On one hand, as several critics have pointed out, it would likely result in intuitively implausible first-order claims about what moral virtue consists in (Millgram 2009; Andreou 2006; Woodcock 2006;

Millum 2006; Odenbaugh 2017). On the other hand, the idea that we can derive moral conclusions from strictly biological premises would imply that we need to radically revise our ordinary modes of reasoning about morality. Note that we do not typically consider biological premises to be in a privileged position to justify moral claims. Doing so, in Thompson’s words, would be to “turn ethics into a sub-discipline of biology”, and to “deny what is legitimately

82 called the ‘autonomy of ethics’” (2004, 62).46 If this autonomy of moral reasoning from biology is worth preserving, then it might seem that the only way to accept that biology can study evaluations of natural goodness would be to suggest that biologists need prior knowledge of moral virtue before engaging in biological inquiry. But this would be an even less palatable implication. However, the idea that the concept of function underlying evaluations of natural goodness has a place in biology does not imply that biology can offer knowledge of moral virtue. Just because a concept plays an explanatory role in a domain of inquiry, it doesn’t follow that this domain of inquiry can give us the full extension of the concept. I have argued that the organizational functions ascribed in biology are instances of flourishing-based functions. But this is not to say that flourishing can be understood solely based on organizational functioning. Organizational functions are merely a subclass of flourishing-based functions. They are the subclass of flourishing-based functions that trace contribution to self-maintenance, which is only one aspect of an organism’s flourishing. There are other aspects of the flourishing of an organism that go beyond self-maintenance, and so are not captured by organizational functions. On Foot’s account, for instance, in the case of plants and non-human animals, flourishing consists of self-maintenance, development, and reproduction (2001, 33). On Hursthouse’s account, the list is actually longer for some non-human animals, and at the very least includes enjoyment and freedom from pain as well (1999, 200). The flourishing of an organism thus goes beyond self-maintenance, and involves other dimensions, depending of the kind of organism in question. Flourishing-based functions specify how each of these different

46 What Thompson is concerned with here is only what we may call epistemic autonomy, and more specifically, epistemic autonomy of ethics from the domain of biology. This is, roughly, the thesis that biological evidence is not relevant to the epistemic justification of “pure”, or “non-mixed”, ethical claims. (See Maguire 2017 for a discussion of different types of autonomy that might be attributed to the ethical domain.) Although some forms of ethical naturalism might deny the epistemic autonomy of ethics from other domains, neo-Aristotelian naturalists have frequently presented their view as compatible with the epistemic autonomy of ethics from biology. They have explicitly rejected the interpretation of their view as deriving substantive norms of virtue from biology, and have said again and again that they do not attempt to give a revisionist, ‘biologistic’ account of moral reasoning and deliberation (see Thompson 2004, 72; Lott 2012b, 418-421).

83 aspects of flourishing are achieved in the life of a kind of organism—e.g., how organisms of a given kind find nutrition, grow, reproduce themselves, and so on. While any of these functions can be the basis of an evaluation of natural goodness, only the ones concerning self- maintenance fall under organizational functions. Thus, although organizational functions can be the basis of evaluations of natural goodness, not all evaluations of natural goodness are based on organizational functions. In the case of human beings, of course, flourishing is even more complex. As Foot points out, the elements that make up good human lives are not only more diverse, but also highly interdependent. For example, the good of reproduction in human beings is related to other goods such as the love and ambition of parents for their children, the special role of grandparents, and many other considerations that do not apply in the case of non-human organisms. As a result of this interdependence, other elements of a good human life can line up such that abandoning a good like having children is not necessarily a defective choice for a human being. The demands of work to be done, for instance, might give a person a reason to forgo family life altogether (Foot 2001, 42). What makes the human case yet more complex is that human life is characterized by practical reason, i.e., the capacity to recognize, respond to, and act in light of reasons. In other words, there is an aspect of human flourishing—manifested in human action and character—that cannot be understood without a grasp of practical reason. Because of this, as we saw in Chapter 2, an understanding of human flourishing cannot come from biology or any other empirical science (Lott 2012b; Hacker-Wright 2009b; 2013). As Hacker-Wright frames the issue, fully understanding human flourishing requires a normatively-laden interpretation of our own form of life that cannot be obtained from a biological account of the species Homo sapiens. It’s rather a rational self-interpretation that can only be made in virtue of our personhood and our nature as practically reasoning creatures. In Lott’s words, such an understanding is possessed by someone “not qua scientist but qua practically wise person” (Lott 2012b, 409).

At this point, it is worth reminding ourselves of the reason why the neo-Aristotelian view is a form of naturalism despite the claim that our conception of human flourishing is formed through practical reason not through empirical science. As I argued in Chapter 1, there is something rather distinctive about the way neo-Aristotelians argue for ethical naturalism. Other forms of ethical naturalism such as Cornell realism (Boyd 1988; Brink 1986; Sturgeon 1985; Railton 1986) construe the natural domain as the domain of natural science, and aim to give an account of morality in terms of facts and properties that fall within the scope of natural science. In contrast, neo-Aristotelian naturalism does not understand the natural domain in epistemic terms or in relation to natural science. Instead, it starts from the idea that the domain of non-human life is uncontroversially natural, and purports to show that the domain of morality is continuous with this natural domain. Thus, neo-Aristotelians argue that flourishing-based evaluations of human action and character are natural because they belong to the broader evaluative patterns that are found in the lives of animals and plants. There is no commitment to the idea that the epistemology of human flourishing is empirical or scientific. It’s only the form of evaluation that is claimed to be continuous across the human and non- human domain. Thus, my suggestion that we need to examine the explanatory concepts of biology should be understood in the context of assessing this specific claim about the continuity of the form of evaluation. The appeal to the concept of function in biology is not due to an epistemological commitment to acquiring knowledge of human form through biology or any other empirical science. The objective is not to assess a substantive account of human flourishing by consulting empirical science. It’s rather the claim to continuity, and particularly its implications regarding non-human living things, that I have set out to defend by examining the explanatory concepts of biology. In other words, what I have tried to show is that norms of natural goodness in non- human life-forms are natural (the premise of Grounding). As I pointed out earlier in Chapter 2, when it comes to understanding the domain of non-human life, the science of biology does have a privileged position over folk views due to its rigorous and systematic study of the

85 subject. This is why the neo-Aristotelians’ ascription of flourishing-based functions to the parts and aspects of plants and animals should be assessed by consulting biology. What the explanatory role of organizational functions in biology tells us is that the flourishing-based concept of function has a respectable place in our understanding of non-human life. That said, a full account of natural goodness for any given kind of organism would go beyond the narrow scope of organizational functions. And we have seen that in the case of human beings, certain aspects of human flourishing fall outside the scope of empirical science altogether.47 It might be objected that although the argument of the chapter does not undermine the autonomy of our moral reasoning from biology, it implies something implausible about biology itself, namely that biologists need a prior knowledge of moral virtue in order to ascribe biological functions correctly. Even though organizational functions specifically concern self- maintenance, as long as they are instances of flourishing-based functions, identifying them correctly requires making assumptions about other aspects of flourishing, which can put constraints on what it is for the organism to maintain itself well. We saw in the previous section, for instance, that in the case of Laron Syndrome the positive effect of the resistance of growth hormone receptors on viability does not amount to organizational well-functioning, due to its other, negative effects on growth and development. So it might seem that in the human case, the dimension of flourishing that concerns practical reason and moral virtue can similarly put constraints on organizational functions. However, it should be noted that although there is a certain degree of interdependence among the different aspects of an organism’s flourishing, there is also a good degree of independence. Grasping what it is for the organism to maintain itself well does not require a complete understanding of other aspects of the organism’s flourishing. Depending on what part of an organization is being studied, only certain aspects of flourishing would be relevant, and

47 It’s worth noting that my argument here does not amount to a defense of other forms of ethical naturalism such as Cornel realism, not only because these views don’t rely on the same premise about the nature of living things, but also because they typically involve a commitment to the idea that the epistemology of human flourishing is empirical or scientific.

86 in most cases only certain general facts need to be presupposed about those aspects. Obviously, a manner of self-maintenance cannot count as maintaining oneself well if it makes it impossible for the organism to realize other aspects of its flourishing. Biologists need to know if a manner of self-maintenance is inconsistent with other aspects of the organism’s flourishing. But this only presupposes a partial knowledge of those other aspects, particularly knowledge of whether and how they are affected by what the organism does to maintain itself. A physiological study of human bodily functions, for instance, does not require much of a grasp of substantive norms of practical reason. And to the extent that it does, it’s only in terms of general requirements such as the capacity for consciousness or memory. If a branch of empirical science engages in a study of organizational functions of human emotions and other psychological dispositions, such a study might need to make further assumptions about what is needed for the proper functioning of our practical reason. But this is not an implausible suggestion. A psychological disposition that hinders the higher functioning of our practical reason cannot be ascribed an organizational function, even if it has a positive effect on viability. If it turns out, for instance, that the capacity for empathy is essential to moral development, a psychopath’s lack of empathy cannot be viewed as proper organizational functioning, regardless of how it affects the individual’s viability. And it’s not implausible to claim that a study of psychopathology needs to be attuned to what philosophers have to say about moral development. Thus, my argument for the role of the concept of flourishing in biology does not involve unpalatable commitments regarding the relation between ethics and empirical science. To summarize, I have argued that the concept of function underlying evaluations of natural goodness is indispensable to biology, and is therefore necessary for understanding living things. This result in turn supports the thesis that natural goodness is an evaluation of living things based on their nature as living things, which is an essential part of the neo- Aristotelian account of moral virtue in terms of human nature. My argument therefore amounts to a defence of the premise of Grounding in the Refined Argument for Naturalism. I

87 have argued that we can consider evaluations of natural goodness in non-human life-forms to be natural because these evaluations are based on the nature of plants and animals as living things. Of course, my argument does not amount to a defense of any particular substantive conception of human flourishing or moral virtue. As we saw earlier, organizational function ascriptions in biology only involve limited commitments regarding the flourishing of an organism, and aspects of human flourishing related to moral virtue largely fall outside such commitments. But the idea is not to provide a vindication of substantive accounts of moral virtue by appealing to biology. It’s rather the neo-Aristotelian metaethical characterization of moral virtue as continuous with the domain of life that I have tried to defend. The central claim of neo-Aristotelian naturalism is that moral evaluations “share a basic logical structure and status” with evaluations of plants and non-human animals (Foot 2001, 27). Moral virtue is claimed to be good in a human being in the same way that deep roots are good in an oak, in that both evaluations are based on the organism’s nature and characteristic flourishing. The objection from biological functions challenges this claim by denying that we can make flourishing-based evaluations of oaks and other non-humans in view of their nature as living things. This is what I have tried to respond to by revealing the role of flourishing-based function ascriptions in biology. As we have seen, organizational function ascriptions presuppose a notion of doing well for the organism that accords with the neo-Aristotelian concept of flourishing and flourishing-based evaluation. Thus, there is an evaluation of living things based on their characteristic flourishing that has the same conceptual structure as evaluations of human beings, including evaluations of human action and character. In Foot’s words, “there is no change in the meaning of ‘good’ between the word as it appears in ‘good roots’ and as it appears in ‘good dispositions of the human will’”(2001, 39).

Chapter 5

Natural Goodness and Natural-Historical Goodness

In this chapter, I raise a new objection to Foot and Thompson’s account of natural goodness in order to motivate seeking an alternative to this account. I start by drawing a distinction between the general concept of natural goodness that is required for the neo-Aristotelian account of moral virtue and the specific, natural-historical conception of this concept that has been advocated by Foot and Thompson. I critically examine the natural-historical account of natural goodness by highlighting its commitment to anti-individualism about life-forms. I argue that this anti-individualism results in a class of potentially problematic implications for the natural-historical account, where it makes the wrong assessment of individuals with unique characteristics. I then argue that neo-Aristotelians have no independent motivation for anti-individualism aside from the assumption that the natural-historical account is the only viable conception of natural goodness. Thus, I suggest that given the problematic implications of anti-individualism, neo-Aristotelians have reason to seek an alternative account of natural goodness that is consistent with individualism.

5.1 The Natural-Historical Conception of Natural Goodness

Let’s start with a review of the general shape of the neo-Aristotelian account of moral virtue and the place of the concept of natural goodness in this account. As we have seen in the preceding chapters, neo-Aristotelian ethical naturalism attempts to place morality in the natural world by offering an account of virtue in terms of human nature. Proponents of this view argue that moral goodness is an instance of natural goodness in human beings, where natural goodness denotes a kind of evaluation based on the concept of the flourishing of an organism that is applicable not only in the case of humans but also in plants and non-human animals. On the neo-Aristotelian account, the goodness of moral virtues such as justice and

89 benevolence in humans is akin to the goodness of deep roots in an oak tree. In both cases, the evaluation is made based on the organism’s nature and what enables it to flourish according to its form of life. Philippa Foot, who introduces the notion of natural goodness, defines it as a form of evaluation that is exclusively attributable in the case of living things in virtue of their nature and according to their form of life. She points out that anything can be evaluated in a context that sufficiently relates it to human concerns. But natural goodness is distinctive in that it is “an intrinsic or autonomous kind of evaluation” (Foot 2001, 27). Being intrinsic means that evaluations of natural goodness apply to the parts and processes of living things independently of the interests of humans or any other external party. They only depend on the relation of an individual organism to its own form of life. This is a feature of living things that is notably different from what is found elsewhere in nature. What is naturally good in the life of an oak is good independently of how we evaluate it. Unlike rivers or storms, oaks and beavers can be evaluated based on standards that are constitutive of their own nature. When we say that deep and sturdy roots are naturally good in an oak, this evaluation is based on the nature of the oak itself and what enables it to flourish given this nature. As a tall and heavy tree, an oak needs deep and sturdy roots to flourish qua oak. By the same token, Foot argues that human beings need moral virtues like justice and benevolence in order to flourish qua human being. We have also seen that Foot relies on Michael Thompson’s (2008) work on the natural history of life-forms to flesh out her account of natural goodness further. Thompson’s work suggests that members of a species share a life-form and a natural history—manifested in a set of natural-historical judgments about the life-form. Incorporating this suggestion into her account, Foot specifies that when we make evaluations of natural goodness, it’s the natural history of the species or the life-form that sets the standard for evaluation. To lay out this account more clearly, let’s briefly recap Thompson’s account of life-forms and their natural history from Chapter 1. We saw that according to Thompson (2008), the natural-historical form of thought, which is the distinctive form of thought manifested in

90 natural-historical judgments, is exclusively used in relation to living things, and can direct us toward an understanding of their nature. Natural-historical judgments are generic statements like “the bobcat has four legs” and “cherries bloom in spring” that articulate the characteristic features and activities of different forms of life, or life-forms, to which individual living things belong. Thompson argues that these judgments have peculiar logical features that set them apart from other forms of generality such as universal or statistical generality. The truth of a natural-historical judgment is consistent with there being instances of the life-form that do not match the general description expressed in the judgment. For instance, it would remain true that “the bobcat has four legs” even if most bobcats lose one of their legs in an accident. Thompson argues that what we can infer about such non-conforming instances is that there is something defective about them. We can say that a bobcat with only three legs is defective in that it doesn’t have four legs. Thus, according to Thompson, natural-historical judgments underwrite evaluative inferences to natural goodness and defect. They offer an account of why a bobcat without four legs is naturally defective and a bobcat with four legs is, to that extent, naturally good.

An important aspect of natural-historical judgments—which is pertinent to their implications of goodness and defect—is that they have a teleological dimension. As Foot also observes, not every form of generic statement about a life-form supports inference to evaluative judgments. “Blue tits have a round blue patch on their head” is superficially similar to “male peacocks have a brightly-colored tail”. But assuming that the blue tit’s color plays no role in the life of the bird, lacking the blue patch does not seem to amount to a defect in a blue tit in the same way that failure to have a brightly-colored tail does in the case of a peacock

(Foot 2001, 30). Thompson argues that natural-historical judgments are a specific subclass of generics that are “teleologically articulable” (2008, 79). This means that they can be connected with each other in teleological relations, such that we can say, e.g., that “male peacocks have a

91 brightly colored tail in order to attract mates”.48 Thus, the set of natural-historical judgments that capture the characteristic features and activities of a life-form are teleologically related, and together form a special unity or a system, which Thompson calls the natural history of the life-form. It’s this natural history that determines what it is to be a good or defective instance of one’s life-form. It’s because a male peacock’s brightly-coloured tail plays a part in the natural history of peacocks by attracting mates that failure to have such a tail amounts to a defect. Thus, on the natural-historical account, evaluations of natural goodness evaluate parts and aspects of living organisms based on their role in enabling the organism to realize the characteristic life of the life-form according to its natural history. The resulting natural-historical understanding of natural goodness is the focus of the present chapter. Although neo-Aristotelians typically use ‘natural goodness’ as synonymous with this natural-historical account, I propose that we distinguish between the general concept of natural goodness and the specific, natural-historical conception of this concept that is presented by Foot and Thompson. So I use ‘natural goodness’ to refer to the general notion of the kind of intrinsic evaluation that is constitutive of living things, and use ‘natural-historical goodness’ to refer to Foot and Thompson’s specific conception of this kind of evaluation in terms of the natural history a life-forms.49 My intention is to examine the natural-historical account of natural goodness in particular and highlight a potentially problematic implication of this account that does not necessarily extend to the general notion of natural goodness. Thus, I suggest that we distinguish between the two following claims:

48 As we will see in the next chapter, there is a question about how exactly the teleological connective ‘in order to’ is to be interpreted in this formulation. In fact, there is also a question about whether Foot and Thompson agree on this interpretation. However, my discussion of the natural-historical conception of natural goodness in this chapter does not depend on the answer to these questions. The objection that I will raise to the natural-historical account of natural goodness applies to this account regardless of how exactly the element of teleology is interpreted. So to keep things simple in this chapter I speak of “Foot and Thompson’s natural-historical account” even though Foot and Thompson might actually turn out to hold different versions of the natural-historical account. 49 Note that this diverges from the way Foot uses ‘natural goodness’ throughout her work. While I preserve the term ‘natural goodness’ for the general concept of a kind of intrinsic evaluation that is constitutive of living things, Foot often uses the term to refer to her own natural-historical conception of this concept.

(NG) There is a distinctive kind of evaluation that applies to living things and their parts

and features in virtue of their nature as living things and independently of the interests of any external beneficiary.

(NHG) There is a distinctive kind of evaluation based on the natural history of life-forms that applies to living things and their parts and features in virtue of their nature as living things and independently of the interests of any external beneficiary.

My critical discussion of Foot and Thompson’s view in next section only concerns their commitment to NHG and does not target NG. In fact, I believe that NG is essentially correct: representing living things as living involves interpreting them in a way that licenses intrinsic and constitutive evaluations. The argument that I presented in Chapter 4 for the premise of Grounding was a defence of NG. I appealed to the explanatory role of the flourishing-based concept of function in biology to argue that the flourishing-based evaluations of natural goodness are necessary for understanding the nature of living things. Note, however, that this defence of NG does not commit me to accepting NHG. My argument was not tied to any particular interpretation of the flourishing of an organism. The idea was that ascribing organizational functions requires presupposing a conception of living well for the organism that cannot be causally reduced. It was left open whether this conception of living well depends on natural-historical judgments about the organism’s life-form or is rather based on other considerations. In the next section, I argue against NHG by raising a problem for the natural-historical interpretation of flourishing and natural goodness. I argue that although in many familiar cases the extension of natural-historical goodness coheres with our intuitions regarding the evaluations of natural goodness, there are also cases where it seems to offer the wrong evaluation.

5.2 Natural-Historical Evaluations and the Problem of Anti- Individualism

The natural-historical account of natural goodness has a lot of initial appeal. It seems true that we typically use natural-historical judgments to describe and understand living things. And familiar, paradigmatic instances of such judgments like cats having four legs, owls seeing in the dark, lionesses teaching their cubs to kill, etc., seem intuitively plausible. It’s also plausible to say that we typically use these judgments as the basis of making evaluations of individual organisms and what goes well or badly in their life. This is why Foot claims that “nobody would . . . take it as other than a plain matter of fact that there is something wrong with the hearing of a gull that cannot distinguish the cry of its own chick, as with the sight of an owl that cannot see in the dark” (2001, 24). However, even if we frequently rely on generic judgments about life-forms to make evaluations of living things, it is not clear that these judgments are more than useful approximations or rules of thumb that we employ in our folk understanding of living things.

Even if there is a sense in which many of these judgments are true, it is not obvious that they are true under Thompson’s interpretation of them as natural-historical judgments. Note that to provide an account of natural goodness as the kind of intrinsic evaluation that is constitutive of living things, natural-historical judgments need to identify aspects of a kind of organism that capture its nature as a living thing. It’s not enough for these judgments to be merely true of living thing without being crucial to grasping their nature. Consider, for instance, a generalization like “elephants are easy to train”. We might make this judgment simply because elephants for the most part exhibit this feature and it is a feature that happens to interests us, without necessarily thinking that it’s anything like a constitutive feature of an elephant. We might even use this judgment to evaluate elephants based on whether they are easy to train. But note that such an evaluation would not be an evaluation of natural goodness. If being easy to train is not constitutive of the nature of an elephant, then there is nothing naturally defective

94 about an elephant that is not easy to train.50 Thus, generic judgments about a species can be true without necessarily identifying constitutive features that can be the basis of evaluations of natural goodness. My contention is that although in a great many cases the judgments that Thompson classifies as natural-historical judgments do identify aspects of living things that can underwrite evaluations of natural goodness, this is not always the case. In what follows, I cast doubt on the natural-historical account by introducing a class of cases where natural- historical goodness and defect does not capture what intuitively seems to be the right intrinsic assessment of a living organism. Admittedly, determining what the natural-historical account says about a particular case can be tricky. The epistemology of natural-historical judgments is not a straightforward matter. We may disagree about what natural-historical judgments are true about a life-form, and it may be hard to resolves such a disagreement. As we saw earlier, the truth of a natural- historical judgment like “peacocks have a brightly-colored tail” depends on whether having a brightly-colored tail is a characteristic feature of peacocks that plays a role in other characteristic aspects of the life of peacocks. This means that to see whether this natural- historical judgment is true, we need to know other natural-historical facts about peacocks, such as how they characteristically attract mates and reproduce. Similarly, in order to grasp these other natural-historical facts, we need to make further assumptions about how the features that they identify fit in the natural history of peacocks. In short, there is a recursiveness involved in the truth conditions of natural-historical judgments, as the truth of each judgment depends on the truth of other judgments about the life-form. Although this recursiveness doesn’t necessarily amount to a problematic or vicious circularity for the natural-historical account, it does complicate my task of offering counterexamples for this account.

50 I am not claiming that either Foot or Thompson are committed to interpreting “elephants are easy to train” as a true natural-historical judgment about elephants. I am merely using the example to demonstrate that a generic judgment about a species can be true without being true under a natural-historical interpretation.

Moreover, even aside from their teleological dimension, natural-historical judgments are instances of generics, and there is much debate about the truth conditions of generics among philosophers of language. Nonetheless, ordinary speakers of English seem to have clear intuitions about the truth or falsity of familiar and paradigmatic examples of these judgments. There seems to be little disagreement about the truth of generic judgments such as “cats have four legs” and “dogs bark”, or the falsity of “cats are female” and “dogs don’t bark”. Thus, in my attempt to show that natural-historical judgments do not provide a suitable basis for an account of natural goodness, I rely on intuitive assumptions about what natural-historical judgments could be true. Now, consider the judgment “goats have four legs”, which seems like a good candidate for a true natural-historical judgment about goats. Not only is having four legs a characteristic feature that is generally observed in goats, it plays a role in other aspects of the goats’ natural history such as how they stand and how they walk. According to the natural-historical account of natural goodness, what follows from this natural-historical judgment is that a goat without four legs is naturally defective. This is an assessment that is intuitively plausible in the case of most goats without four legs. However, consider now the case of Slijper’s two-legged goat, which I discussed earlier in Chapter 3. West-Eberhard (2005) writes about this real-life case in her discussion of the role of developmental plasticity in the origin of species differences. The goat, which was described in 1942 by morphologist E. J. Slijper, was born without forelimbs. But it adapted to its condition in unexpected ways, and developed several behavioral and morphological specializations similar to those of kangaroos and other bipedal mammals, including enlarged hind limbs, a curved spine, and an unusually large neck. As a result, the goat learned to walk and run by using its hind legs alone. As West-Eberhard’s points out, the correlated shift in the goat’s morphology and behavior led to “the well coordinated production of a complex and individually advantageous adjustment, producing a novel phenotype with little or no genetic change” (2005, 6545). Slijper’s goat lived for a year before it died in an accident unrelated to its physical condition. And I would argue that the case of this goat

96 presents a challenge to the natural-historical account. Since Slijper’s goat didn’t have four legs, the natural-historical account evaluates it as naturally defective. But given that the goat seems to have been able to manage perfectly well without four legs, it’s not clear why we should accept this evaluation. One might try to defend the natural-historical evaluation of Slijper’s goat as defective by arguing that even though the goat could hop on two legs, its way of moving around was not as effective as normal goats. It seems highly plausible that a goat that has to hop on two legs would not be able to get around as fast or as easily as normal goats walking on four. However, note that regardless of whether this was in fact the case with Slijper’s goat, Thompson’s natural-historical evaluation is not sensitive to how fast or easily a two-legged goat can move around. Even if Slijper’s goat was able to move with the same or even higher speed and ease than ordinary goats, the account would evaluate it as defective on the basis of its difference from other goats. Note that on Thompson’s account, what makes “goats have four legs” a true natural-historical judgment and a basis for natural-historical evaluation is that having four legs is a characteristic feature of goats that characteristically contributes to other characteristic aspects of a goat’s life such as movement and finding nutrition. Even if it was possible for Slijper’s goat to realize these other characteristic aspects of a goat’s life with only two legs, its way of achieving them would not be the characteristic way. It would still remain the case that goats characteristically achieve movement and nutrition by walking on four legs. So the natural-historical account would evaluate Slijper’s goat as naturally defective regardless of how fast or easily it can move around and realize other aspects of a goat’s life. Moreover, even if as a matter of fact Slijper’s goat was slower in its movements compared to other goats, it is not obvious that this difference in speed should be enough to make it defective. Presumably ordinary, four-legged goats also vary in their speed based on other factors such as their size. And yet, as long as a goat’s speed is within the characteristic range—as defined by the range that is observed with some degree of regularity among goats— having a lower speed does not amount to a natural-historical defect. Why, then, should we

97 consider Slijper’s goat defective for falling short of the speed characteristically exhibited by other goats? Note, further, that we do not consider ordinary goats defective for falling short of the speed exhibited by other species like cheetahs and lions. Even if the ability to run as fast as a lion would be an advantage in the life of a goat, we do not consider goats as a species to be naturally defective or naturally worse off than faster species. The idea of natural goodness is to evaluate living things based on standards internal to their own form of life and not based on what would have been an ideal or optimal way for them to be. This is why a cheetah’s ability to run fast or an eagle’s ability to fly would be external and irrelevant standards for evaluating a goat. But, by the same token, an ordinary four-legged goat’s ability to run with a certain speed also seems like an irrelevant standard to use for assessing Slijper’s goat. Given that Slijper’s goat diverges significantly from ordinary goats in its morphology and method of locomotion, it is plausible to think that it merits a different standard of evaluation based on its own modified form and not the way other goats happen to move around. In fact, the problem with the natural-historical evaluation of Slijper’s goat is not merely that it misidentifies the legs as defective, but also that it fails to correctly identify the good of many aspects of the goat’s morphology. Slijper’s goat had a dorsoventrally compressed ribcage, a swayed back, and dorsally located clavicles and scapulae, which together enabled the goat to keep its upright posture and walk on two legs. Any of these features would likely be a defect in an ordinary goat, but in Slijper’s goat, they make a crucial contribution to the goat’s way of achieving movement and maintaining its way of life. So it seems plausible to say that the presence of these features in Slijper’s goat is naturally good, and their loss would be a natural defect. The natural-historical account, however, is not able to make this intuitively plausible assessment, and would have to say that all the peculiar aspects of Slijper’s goat are instances of natural defect—just as they would be instances of natural defect in an ordinary goat. In response to this objection, a defender of the natural-historical account might contest my original assumption that “goats have four legs” is a true natural-historical judgment. They might suggest that once we encounter a case like Slijper’s goat, we need to revise our initial

98 natural-historical judgment because we learn that having two legs is also an acceptable way for goats to be. In other words, we find out that the life-form goat actually consists of two naturally sound sub-kinds that have different characteristics and are assessed based on different standards. Polyphenism—the development of alternative phenotypes from a single genotype as a result of differing environmental conditions—is a commonly observed phenomenon among living things. From sex determination in crocodiles to the development of division of labor in honeybees, environmental conditions can give rise to distinct phenotypes among the members of a species. And there is no reason to think that the natural-historical account cannot handle these polyphenic traits. We can make different natural-historical judgments about the various sub-kinds within a life-form and arrive at different standards for assessing instances of each sub-kind. In the case of goats, for instance, upon encountering Slijper’s goat, we can recognize a new sub-kind of bipedal goats that differ from other goats not only in the number of their legs but also in other characteristics like the shape of their spine and how they walk. Thus, we can make different natural-historical judgments about this newly discovered phenotypic variant of goats and say, e.g., that “bipedal goats have two legs,” “have a curved spine,” “move by hopping around,” etc., while “quadrupedal goats have four legs” and so on. However, the problem with this response is that it ignores an important aspect of natural-historical judgments as generic judgments. Although we can make different natural- historical judgments about different sub-kinds within a life-form, the natural-historical account does not allow for having a sub-kind with only one actual instance. That’s because we only make natural-historical judgments when there is a class of real and present entities that we want to describe. Natural-historical judgments are not about hypothetical kinds that have no actual instances or are abstracted from one isolated instance. Their truth implies the existence of multiple actual instances of the kind or sub-kind that they describe. Julius Moravcsik (1994) takes notice of this “existential import” of generic judgments about biological kinds (Moravcsik 232). He argues that these generics require that there must exist something that corresponds to

99 their subject expression. A generic judgment like “beavers build dams” implies that there must be or have been some actual beavers. What the judgment describes is not some Platonic idea of beaverhood that may or may not be instantiated. It rather describes “a set of elements, some of which are actual and have a common causal origin and are linked via other causal links, while the non-actual ones are projectible from the actual cases” (234). Thus, a natural-historical judgment like “bipedal goats have two legs” cannot be true unless there are multiple actual instances of bipedal goats that share a causal origin and are linked via causal links. In other words, we need more than one actual instance of the kind to make it the subject of a true natural-historical judgment—a condition that is not met if all we have is a single bipedal goat with its unique set of traits. Thus, even though the natural-historical account can handle polyphonic traits that are established and occur with regularity among the members of a species, it faces a problem in the case of individuals that are unprecedented and unique. This problematic implication of the natural-historical account results from its commitment to the rejection of a position that Thompson calls individualism. In his discussion of the concept of life, Thompson (2008) rejects the idea that whether an individual entity amounts to a living thing and whether its parts and processes amount to organs and vital operations is determined by lower-level facts about the region of space-time occupied by the individual. Instead, he argues that the representation of the individual as a living thing depends on the ‘wider context’ of its life-form, which goes beyond the individual and “is not fixed or determined by anything in the individual itself” (2008, 50-51). As we saw earlier, Thompson defines a life-form or a species as the sort of kind that can be the subject of a natural-historical judgment (2008, 48). And as I explained above, the generic character of natural-historical judgments requires them to identify features that are characteristically shared between the instances of the life-form. Thus, on Thompson’s account, a life-form is a particular kind of kind that is essentially shared between multiple instances. A life-form must have (or have had or be going to have) more than one actual instance such that it can be the subject of true natural- historical judgments. This is why, on the natural-historical account of natural goodness, an

100 individual organism cannot be subject to a standard of evaluation that is not shared by other actual organisms. Thus, the natural-historical account offers a conception of natural goodness that is committed to anti-individualism—it maintains that the standards for evaluation of an individual are not determined by lower-level facts about the region of space-time occupied by the individual itself. It requires that the standards of natural goodness are essentially shared between individuals. And I have argued that this commitment of the natural-historical account to the idea that the standards of evaluation are essentially shared results in a class of potentially problematic implications for this account, where it makes the wrong evaluation of individuals with a unique form. It should be noted that although I have focused on one illustrative example with the case of the two-legged goat, the development of unique forms as a result of phenotypic plasticity is by no means a rare occurrence among living things. In fact, West-Eberhard argues that “the two-legged goat effect” is a widespread phenomenon (2003, 53). As we saw in Chapter 3, developmental plasticity is a distinctive feature of living things that plays an important role in originating evolutionary novelty. When a new phenotype arises as a result of genetic or environmental input, the plasticity of living organisms enables them to respond by developing other novel and highly adaptive phenotypes via phenotypic accommodation—i.e., the adaptive adjustment of various phenotypes in the organism to accommodate the new phenotype. If the genetic or environmental factor initiating the change is recurrent, the new phenotypes will spread and ultimately get genetically accommodated via natural selection. But clearly, whether a novel phenotype in a given organism is naturally good or defective does not have to depend on whether the phenotype gets spread in the population later on. And we have seen that the natural-historical account is unable to identify an adaptive novel phenotype as naturally good until it is multiplied across the species. So now we may ask if it’s possible for neo-Aristotelians to drop the commitment to anti- individualism about natural goodness in order to handle such cases. In the rest of this chapter, I

101 assess the neo-Aristotelians’ motivation for this commitment to anti-individualism by discussing an explicit argument that Thompson has made against individualism.

5.3 Thompson’s Case against Individualism

Thompson offers an explicit argument against individualism as part of his rejection of what he calls a “list-making” approach to understanding the nature of life (2008, 43). In his view, the list-making approach—which involves understanding life in terms of a list of properties that are claimed to constitute living things—has two misguided ambitions. The first ambition is the idea that life can be given a real definition—i.e., a constitutive account in terms of independently-recognizable properties that all and only living things have. The second ambition is that it can be given an individualistic account—i.e., an account in terms of the properties of the region of space-time occupied by an individual organism. In the rest of this section, I will look at Thompson’s argument against each of these ambitions in turn. I will argue that Thompson’s case against individualism depends on the assumption that his own natural-historical account, which is not consistent with individualism, is the only viable account of natural goodness.

Can Life be given a real definition?

The first misguided ambition of the list-making approach to understanding life, according to Thompson, is the idea that life can be given a real definition—i.e., a constitutive account in terms of properties shared by all and only living. As an example of the list-making approach, Thompson discusses the definitions of life frequently given in the beginning of biology textbooks. These definitions typically give a list of criteria that we can use to recognize living things. They point out, for instance, that living things are highly organized and homeostatic, take energy from the environment and convert it into other forms, respond to stimuli, grow and develop, are adapted, and reproduce themselves (2008, 34). We may think that such lists can offer an account of what life consist in. But Thompson argues that these lists fail to

102 contribute toward an understanding of life, because none of the criteria that get repeated in such lists can be properly understood without already possessing a grasp on life. Take being organized, for instance. It’s very common for list-making accounts of life to mention organization as one of the essential properties of living things. But as Thompson argues, it’s not clear how the intended notion of organization is supposed to work in a constitutive account of life. We may try to understand organization in abstract terms, say as complexity of form. But organization in this general sense is not a distinguishing mark of life. Many non-living structures, from crystals to car engines, also exhibit high degrees of organization in this sense. According to Thompson, what distinguishes living things is a more definite, special kind of organization, which we may call vital organization. It is not just that they have parts; they have members and organs. But if we try to clarify what amounts to a member or an organ and what kind of organization amounts to vital organization, we realize that we cannot understand these concepts independently of our grasp on the concept of life. The same can be said of the other criteria in the list, such as responding to stimuli or taking and converting energy. Interpreting the criteria in abstract terms makes them too permissive, while specifying the special sense they take in living things requires presupposing an understanding of life. For instance, we may interpret responding to stimuli as a process whereby one thing follows another or reacts to it in the chemical sense. But in this abstract interpretation of the term, a non-living thing such as an asphalt road getting warmer due to sunlight also counts as responding to stimuli. So again, Thompson argues that what is distinctive about the way in which living things respond to stimuli in life-processes cannot be recognized from the point of view of physics or chemistry.

In short, Thompson argues that the list of criteria fails to offer a non-circular account or a real definition of life. As he puts it, the abstract categories mentioned in the definition are all “thrown into a higher gear” to yield the concepts necessary for understanding life. But we run into a circle trying to understand each of these concepts in terms of another.

103

A number of abstract categories—that of a concrete individual; of a thing’s being a part of something; of order or organization; of one thing’s following another in a process; of a thing’s doing something—are all together determined or specified, or thrown into a higher gear, to yield the concepts: organism; organ, ‘part’ or ‘member’; vital order or organization; life-process; and vital operation.

These concepts, the vital categories, together form a sort of solid block, and we run into a kind of circle in attempting to elucidate any of them. (Thompson 2008, 47)

The conclusion Thompson draws from this argument is that no reductive constitutive account of life can be given. The nature of the properties that appear in a constitutive account of life cannot be explained without making reference to these properties themselves.51 Therefore, to the extent that the list-making approach attempts to give a non-circular—i.e., reductive—account, it involves a misguided ambition.

Can life be given an individualistic account?

The other false ambition of the list-making approach, according to Thompson, is what he calls “an extreme individualism”, which he describes as the idea that an account of life should tell us what a region of space-time needs to be like if it is to be occupied by an individual organism (2008, 49). To argue against this ambition, Thompson assumes that although the criteria repeated in textbook definitions are not recognizable without a prior grasp on life, they do in fact capture the characteristic features of living things. But he argues that the fact that an individual organism meets these criteria does not supervene on lower-level facts about the individual. Take homeostasis and reproduction, for instance. Homeostasis simply means staying the same, while reproduction means making a copy of oneself. But, of course, neither homeostasis nor reproduction requires that all aspects of the organism are preserved. Homeostasis requires

51 As Thompson argues, trying to give a “holistic” account of these vital concepts in terms of their interrelations with each other will not work either, because the same interrelations hold between the abstract categories that are instantiated in non-living things (2008, 48).

104 that a living thing stays the same in some, but not all, respects. And reproduction similarly requires that some, but not all, aspects of the organism are copied. Thompson suggests that what is preserved in homeostasis and reproduction is the form of the organism, and argues that this form is not determined by individualistic facts about the organism.

If we call the relevant sameness sameness of form, then that a thing has a given such form will not be an ‘individualistically’ determinable fact about the thing; it will not, for example, be simply equivalent to any collection of physical or phenomenal facts about the thing itself or the region of space-time occupied by its perduring double. (Thompson 2008, 51)

The idea is that the respect of sameness, or the form of the organism, shows an independence from the lower-level facts about the individual organism, such as its physical structure or its phenomenal character. And since criteria such as homeostasis or reproduction depend on the form of the organism, whether an entity meets these criteria is not individually determined either. To illustrate this point, Thompson gives the example of metamorphosis in butterflies. Metamorphosis involves radical changes in a butterfly’s physical and phenomenal properties. But these changes do not violate the sameness of form. The form of the butterfly is preserved despite these radical changes. In fact, we can imagine a kind of poly-metamorphic butterfly that goes through metamorphosis a hundred times and in a hundred different ways in its lifetime without violating the sameness of form. Thompson argues that what it takes for the form of the butterfly to stay the same changes each time, and is not associated with any particular arrangement of physical or phenomenal facts. Note, however, that the fact that the form of an individual can stay the same while physical or phenomenal facts change is not enough to rule out individualism. Just because materially different things can add up to the same form it doesn’t mean that the form is not determined by the underlying material. To show that the form is not determined by the underlying individualistic material, Thompson needs to show that the same individualistic material can add up to different forms depending on the context. His view is that what amounts

105 to homeostasis, reproduction, or in fact any of the vital operations characteristic of life, is not determined by “the individual lump of stuff” in front of us, but rather a ‘wider context’ that goes beyond the individual (2008, 54). So in rejecting individualism, he is claiming that the form of an individual can change while the lower-level physical or phenomenal facts about the individual stay the same. Thompson discusses a number of examples to argue that the same set of lower-level facts can amount to different vital operations in different contexts. One example is the case of mitosis, which constitutes different vital operations in amoeba and human beings. Thompson argues that the same process of cell division that amounts to reproduction in amoeba constitutes growth in human beings. In another example, he invites us to imagine a novel kind of shark that is nourished not like normal sharks by smaller fish, but rather by microscopic organisms such as plankton. We can imagine that these quasi-sharks do something that looks very much like hunting, munching, and swallowing smaller fish. They chase after fish and incorporate them just like normal sharks do. But let’s suppose that instead of extracting the nutritious elements from the fish, these quasi-sharks make a special brew that they occasionally spit out to frighten predators. Thomson argues that in this example the same process that amounts to eating in normal sharks does not add up to eating in quasi-sharks. Even though the lower-level individualistic processes are the same in both cases, the resulting vital operation differs depending on the context of the life-form. Thus, Thompson argues that for any given material process, “philosophers can arrange” that it does or does not amount to a certain vital operation by changing the context (2008, 54- 55). If this is in fact the case, then vital operations characteristic of life do not supervene on lower-level facts about the individual entity, which means that life cannot be given an individualistic account. However, note that all that Thompson’s examples show is that vital operations do not supervene on local lower-level facts about isolated processes. None of the examples rules out supervenience on the entire set of lower-level facts about the individual organism. We can

106 accept that what constitutes growth in humans does not supervene on the isolated process of mitosis in a cell, or that eating in a shark does not supervene on the isolated sequence of processes that take place in chasing and incorporating fish. But these examples leave out many other aspects and properties of an individual human or a shark. The quasi-shark, for instance, will not have the exact same material constitution as a normal shark if it doesn’t digest fish in the same way, and will materially differ from normal sharks despite certain isolated processes being the same. Thus, what constitutes eating in a shark may very well supervene on the complete set of individualistic facts about the shark, which also includes facts about what happens to the fish after it is incorporated in the shark’s body. In other words, it is not obvious that the context determining whether the set of underlying processes amount to eating has to go beyond an individual shark. We may accept that the vital organs and operations of an organism don’t supervene on isolated lower-level parts and processes within the individual, while still maintaining that they supervene on the entire set of lower-level facts about the individual organism. The question, then, is whether the form of an individual can change despite the entire set of lower-level facts about the individual remaining the same. Can a living organism be made into a different kind of living thing—or even a non-living thing—by merely changing its surrounding context? Thompson discusses the case of a swamp creature to answer this question in the affirmative. He argues that a spontaneously-generated swampman that is a molecule-for-molecule replica of Thompson himself but has appeared as a result of a cosmic accident is neither a human nor a living thing (2008, 60). If Thompson’s analysis of the swampman is correct, then clearly the form of an individual is not determined by lower-level facts about the individual. However, as we will see below, Thompson’s argument for this claim is not exactly conclusive. Thompson’s grounds for rejecting the idea that the swampman is a living thing is that this creature has no causal connection with other human beings, which means that there is no wider context for linking the individual to other individuals of the same form. According to

107

Thompson, in the absence of such a context, it is underdetermined what organs or vital operations the parts and processes of the swamp amount to. It’s unclear, for instance, whether the part that is the molecular equivalent of Thompson’s left arm is really an arm or, say, a horribly deformed wing. An individual living thing can instantiate its form well or poorly, and if all we have to go by is one isolated individual, we seem to have no basis for making the distinction (2008, 60-61). Thus, Thompson’s argument against individualism ultimately comes down to the claim that there is no basis for distinguishing between goodness and defect and interpreting the parts and processes of an individual as vital organs and operations without the extra-individualistic context. As we saw earlier, Thompson’s own account of the basis for indentifying the form of living things appeals to natural-historical judgments, which have anti-individualistic implications. On his view, a life-form is defined as the subject of natural-historical judgments, which provide a basis for interpreting the parts and processes of individual living things and evaluating them as naturally good or defective. And given that natural-historical judgments are generic judgments identifying characteristic features of a class of individuals, they cannot be made merely based on lower-level facts about an isolated individual. In short, what prompts Thompson to reject individualism is that he believes the only way to make sense of the distinction between natural goodness and defect is in terms of natural- historical judgments, which result in an account that is not consistent with individualism. In other words, anti-individualism is not an independently motivated commitment that Thompson thinks we should try to preserve, but simply a result of his natural-historical account of the nature of life, which he takes to be our best shot at understanding the distinction between natural goodness and defect.

5.4 Can Natural Goodness Be Given an Individualistic Account?

I started the chapter by suggesting that we need to distinguish between the general concept of natural goodness needed for the neo-Aristotelian account of moral goodness and the natural-

108 historical conception of this concept offered by Foot and Thompson. The general concept of natural goodness is simply the notion of a kind of evaluation that applies to living things and their parts and features in virtue of their nature as living things and independently of the interests of any external beneficiary. The more specific, natural-historical conception of natural goodness, however, ties these evaluations to natural-historical judgments about life-forms. And I have argued that this connection to natural-historical judgments commits Foot and Thompson’s account to anti-individualism about evaluations of natural goodness. Of course, if the natural-historical account is the only viable account of natural goodness, then this commitment to anti-individualism would be inevitable. But I have argued that anti- individualism comes with a cost. It results in a class of potentially problematic implications for our account of natural goodness, where at least intuitively the account seems to make the wrong assessment of individuals with unique characteristics. We have seen that, being anti- individualistic, the natural-historical account makes it a priori impossible for a living thing to be evaluated by a standard that is not shared by any other actual living thing. This is while our intuitive grasp on what is good or defective in the life of an organism does not seem to require the standards of evaluation to be essentially shared. And we have seen that the phenotypic plasticity of living organisms can give rise to cases like the two-legged goat, where the correct evaluation of the organism’s parts and aspects seems to diverge from the evaluations of the natural-historical account. I have also argued that neo-Aristotelians have no independent motivation for anti- individualism aside from the assumption that there is no viable alternative to the natural- historical account of natural goodness. Thus, by highlighting the problem with anti- individualism, I have argued that neo-Aristotelians have reason to seek an alternative account of natural goodness that is consistent with individualism. In the next chapter, I explore the possibility of avoiding the problem of anti-individualism by offering such an alternative account.

109

Chapter 6

Natural Goodness without Natural History

In this final chapter, I sketch a novel account of natural goodness that is distinct from the natural-historical account and consistent with individualism. This alternative account of natural goodness does not rely on generic judgments about a species, and instead grounds the distinction between goodness and defect in explanatory generalizations that describe an individual organism. In a nutshell, the account determines whether a part or an aspect of an organism is naturally good or defective depends on the best explanation of the organism’s various parts and aspects. In offering this account, I draw from Denis Walsh’s account of teleological explanation, which identifies a kind of regularity in the modal profile of a purposive system that is empirically observable and explanatorily significant. I argue that this kind of regularity can replace the generic, species-level patterns of regularity in the natural- historical account to take us toward an alternative, modal-explanatory account of natural goodness that is consistent with individualism.

6.1 Teleological Explanation and Patterns of Counterfactual Dependence

My account of natural goodness draws heavily from Walsh’s (2012; 2013) approach to teleological explanation. So let’s start with an overview of Walsh’s approach. As we will see in this section, Walsh defends the status of teleological explanation as scientifically respectable by identifying a distinct kind of regularity that is only captured by teleological explanation. Later in sections 6.2 and 6.3, we will see how my account appropriates this distinct kind of regularity as a basis for evaluations of natural goodness. A teleological explanation explains the occurrence or the nature of some event or entity by citing the goal, end, or purpose to which it contributes (Walsh 2012, 173). We regularly offer and accept teleological explanations in a variety of contexts, most predominantly in folk

110 psychology, but also very commonly in folk biology. We explain, for instance, an agent’s going to the store by citing her goal of buying groceries. We also explain a mammal’s sweating by citing the goal of its thermoregulatory system to maintain proper body temperature. But despite the prevalence of teleological notions in biology, they are often considered problematic and associated with the pre-Darwinian view of life as the product of God’s conscious design. As Walsh (2012; 2013) also points out, modern scientific methodology is predominantly mechanistic. On the mechanistic worldview, every natural phenomenon has a productive cause, and to explain the phenomenon one adverts to the causal mechanisms that produce the phenomenon. Moreover, causal-mechanistic explanations are, in principle, exhaustive and complete: they apply to the occurrence of every event, and once completed, they leave no unexplained residuum. Hence, the mechanistic worldview seems to leave no room for teleological explanation. In fact, if scientific explanation requires citing causes, teleological explanation does not seem to be able to explain anything at all. Teleological explanation involves citing goals. And at the time of the occurrence of the event that is to be explained as a means toward a goal, the goal is an unactualised, future state of affairs that cannot cause the occurrence of the means. Walsh aims to defend the status of teleological explanation as a legitimate, non- redundant form of scientific explanation that is both compatible with and autonomous from causal explanation. On his view, for some events, fully understanding their place in the natural world requires that they are explained “twice over: once by appeal to the mechanisms that cause them and once by appeal to the purposes they subserve” (Walsh 2012, 174). Walsh’s account of teleological explanation is adapted from a prominent approach to the understanding of causal explanation that is given by James Woodward (2001; 2005). Walsh’s strategy is to pinpoint the elements that make causal explanations explanatory on Woodward’s account, and to argue that teleological explanations have these elements too. According to Woodward, giving a causal explanation of a natural phenomenon requires identifying a set of conditions that make the difference between its occurrence and its non-

111 occurrence. This involves identifying the phenomenon to be explained as part of what he calls an invariance relation, which is essentially a robust relation of counterfactual dependence between two states or events. A relation is invariant or stable across certain changes if it holds—up to some appropriate level of approximation—across those changes. By contrast, a relation is not invariant across certain changes if it “breaks down” or fails to hold—even approximately—under those changes (Woodward 2005, 239). Woodward argues that if xc is genuinely a cause of xe, we would expect the generalization describing how changes in xc are 52 correlated with changes in xe to be invariant under at least some change in xc. This means that, for a range of circumstances, if we were to intervene to change the value of xc, the value of xe would also change in a systematic way, such that the generalization describing the relation between xc and xe would continue to hold. Note, for instance, that this is the case for 2 Newton’s law of universal gravitation. The generalization F = Gm1m2/r describes the relation between the magnitudes of two masses m1 or m2, the distance d between them, and the value of the gravitational force they exert on each other. Not only does this generalization hold for the actual values of the variables in a given instance, it also continues to hold in a range of counterfactual circumstances under variations in the value of m1, m2, d, or F.

Woodward argues that a generalization relating xc and xe that is invariant in this way enables us to explain xe in terms of xc. An explanation shows how changes in explanandum counterfactually depend on changes in the factors cited in the explanans, i.e., how the explanandum would have been different if the factors cited in its explanans had been different.

So we may think of the explanation as telling us how to manipulate xe if it were possible to intervene to change xc. In this sense, Woodward’s account involves a “manipulationist” conception of explanation (2001, 5). Importantly, this account of explanation does not limit explanatory generalizations to what are often regarded as laws of nature. The invariance relation does not require that the

52 More accurately, we would expect them to be invariant under some interventions on xc, where an intervention on xc with respect to xe is a change in the value of xc that changes xe, if at all, only via a route that goes through xc and not in some other way.

112 generalization describing the relation between the two states is exceptionless or that it meets other traditional criteria for lawfulness such as breadth of scope and theoretical integration. All that is required for a generalization to be invariant is that it is stable under some range of interventions.53 Because of this, Woodward’s account of causal explanation promises a more satisfactory treatment of explanation in the special sciences, where the generalizations underwriting explanations are often non-strict, qualified, or ceteris paribus. Woodward’s counterfactual theory of causal explanation is the basis for Walsh’s account of teleological explanation. According to Walsh, just as a relation of counterfactual dependence holds between a cause and its effect, a distinct relation of the same form holds between a goal and the means to its attainment. He argues that although Woodward’s invariance relation— which we may call causal invariance—always holds between a cause and effect, when the effect is also a goal and the cause is a means to attaining that goal, there is an additional invariance relation between the two, which holds in the reverse direction and underwrites a teleological explanation of the means in terms of the goal. Walsh (2015, 198) calls this additional invariance relation purposive invariance. He argues that while the causal invariance relation captures the counterfactual dependence of an effect on its cause, the purposive invariance relation captures the counterfactual dependence of cause (as a means) on the effect (as a goal).

The idea is that a means-end relation between xc and xe involves a different pattern of counterfactual dependence between them than a mere cause-effect relation. Let me demonstrate the difference between the two patterns in an example. Suppose you are walking in a market when you notice that a store actually carries sweet lemons, a childhood favorite that you have had a hard time finding anywhere in town. Happy with this incident, you buy

53 According to Woodward, generalizations that are legimitately regarded as laws of nature are invariant under a particularly wide range of interventions. But it is still their invariance and not whether, independently of this, they satisfy the standard criteria for lawfulness that gives them explanatory import. A generalization can be robust under a much narrower range of interventions than paradigmatic laws and yet still count as invariant in a way that enables it to underwrite explanations (2005, 240).

113 some lemons.54 Now, contrast this scenario with a different case where you already know that the store carries sweet lemons and you go to the market specifically to buy some. Although in both cases you buy sweet lemons as a consequence of going to the market, the patterns of counterfactual dependence are not the same. Both cases get the same causal explanation. In both cases, buying lemons is causally explained by going to the market. This is because the relation between going to the market and buying lemons exhibits causal invariance and continues to hold under certain changes in the conditions of your going to the store. For instance, if you had gone to the market wearing something else, you would still have bought the lemons. Or if you had gone to the market a little earlier in the day, you would have bought the lemons a little earlier in the day. However, in the second case, there is also a different invariance relation that supports additional counterfactuals. For instance, in the second case, if lemons were kept somewhere in the back not immediately noticeable to someone walking in the market, you would still have bought some. Perhaps even if it was another store carrying sweet lemons you would have gone to the other store instead and would still have bought them. Or if you were busy during the day and worried that the store might close before you could walk there, you would have driven your car to the market instead. This additional pattern of invariance, which is absent in the first case, is due to the fact that in the second case there is a means-end relationship between going to the market and buying lemons. On Walsh’s account, this different pattern of invariance underwrites a distinct form of explanation that is teleological explanation. That is why, in the second case, we can explain your going to the market by pointing out that you go there in order to buy sweet lemons. Thus, Walsh articulates the distinct relation of counterfactual dependence that holds between a means and a goal as follows:

The relation between a goal and its means is change-involving and invariant in that if xe is the

system’s goal and (under a set of conditions) xc is the means to the attainment of xe, then,

54 Walsh discusses a similar example adapted from Aristotle’s Physics II.5 involving a man collecting subscriptions for a feast. In Walsh’s example, the protagonist meets a debtor at the market as a chance encounter and collects some money owed to him (Walsh 2015, 193).

114

given xe as a goal, xc will occur as a matter of regularity (under those conditions).

Furthermore, an intervention on xe—one that changes the goal—would bring about a

difference in xc (the means). But not just any difference; generally, the new value of xc will be

one that is conducive to the new goal, xe. (Walsh 2013, 52)

On this account, the purposive invariance relation between a goal xe and a means xc is marked by its stability under two kinds of changes. First, the relation between xc and xe is not disturbed by at least some changes in the initial conditions. Under a range of conditions, as long as xe is a goal, xc will occur as a matter of regularity such that xe occurs as a result. Second, under a * range of conditions, if the goal xe is changed to xe , the means would correspondingly change, * to xc , such that the new value of the means is conducive to the new value of the goal. In the example discussed above, we can see these two patterns of stability in the counterfactual condition in which you are too busy to walk to the market and the one in which a different store carries the lemons respectively. Since your action of going to the market is directed toward the goal of buying lemons from the store, under certain conditions—such as being busy during the day—you would adjust your action—e.g., by driving instead of walking—such that you would still buy the lemons. Similarly, if certain aspects of your goal—such as which stores the lemons are to be bought at—change, you would adjust your action—e.g., by going to a different store instead—such that you would still buy the lemons. As we can see above, Walsh’s account of purposive invariance makes reference to the notion of a goal. It’s worth noting that, in offering an account of teleological explanation, Walsh is not concerned with giving an account of what it is for something to be goal-directed, but is rather trying to identify the aspects of goal-directness that make it explanatory. Here, we may simply understand goals based on the notion of goal-directedness in Cybernetics and

Systems Theory (Rosenblueth, Wiener and Bigelow 1943; Bertanlanffy 1969; Braithwaite 1964; Nagel 1961; 1977; Sommerhoff 1950).55 On this account, having a goal is a complex, empirically

55 Although Walsh offers a different, non-reductive account of goals—in terms of an interdefineable cluster of concepts consisting of goal, affordance, and repertoire—elsewhere (2015, 211), in his discussion of teleological

115 observable property of a goal-directed system that results from the system’s architecture and causal capacities. A goal-directed system is, roughly, a system that has the capacity to attain and maintain an end-state in the face of external and internal perturbations. When changes in the external or internal environment threaten to deflect the system from this end-state, the system typically alters its behavior in a way that redirects the system toward this state. The stable end-state is the system’s goal. Since goal-directed systems don’t always maintain their goal states successfully, the account only requires that the system oscillates around its goals or approaches them asymptotically. Thus, goal-directedness can be characterized in terms of behavioral signs such as persistence and/or plasticity. Persistence involves maintaining the end-state across a wide range of conditions; and plasticity means doing so by responding differently to conditions and in a manner appropriate to attaining the end-state in each. On any occasion, a plastic system has a ‘repertoire’ of responses that it is capable of producing, some of which are conducive to the attainment of the system’s goals. And it exhibits a bias toward the goal-conducive elements of its repertoire (Walsh 2012, 177). Note that although the paradigm cases of successful teleological explanation appeal to the intentions of an agent, having a goal in Walsh’s intended sense does not presuppose intentionality. Non-human organisms and other systems without intentional states can exhibit goal-directed behavior just as rational agents do. In fact, Walsh considers many biological subsystems such as the endocrine, immune, and thermoregulatory systems to be goal-directed systems, as they involve persistent and/or plastic attainment of stable end-states (2015, 202). Moreover, although this is not immediately clear in Walsh’s formulation, it’s worth noting that the invariance relation between a system’s goal xe and a means xc to achieving it does not require that the goal state is achieved successfully. The relation of counterfactual dependence between xe and xc does not require that they actually occur, but rather that there are a range of actual or counterfactual conditions in which xc occurs as a matter of regularity in a way that is

explanation, he appeals to the system-theoretical notion of goal-directedness as a “natural and observable” feature of a system’s dynamics (2015, 195).

116 conducive to xe. To see an example, suppose I have the goal of buying sweet lemons, but ultimately fail to buy them because the store runs out of them before I get there. Even though my goal state of buying lemons is not actually realized, we can see that it still has an invariant relation to my action of going to the market. There are a range of counterfactual conditions in which the store does not run out of lemons before I get there, and in which I buy lemons. And my behavior in these counterfactual conditions shows persistence and plasticity toward the goal of buying lemons such that in each condition I adjust my action in such a way that it is conducive to the goal of buying lemons. Walsh argues that although the conditions that make teleological explanations explanatory can be given causal terms, the content of a teleological explanation cannot be reduced to causal explanations without explanatory loss. Unlike causal explanations, teleological explanations feature the end-state as a goal in the explanans. And this enables them to capture a different aspect of the modal profile of the system that is not captured by causal explanations. On his view, causal and teleological explanations are complementary and noncompeting, in that they explain different facets of the same phenomenon. One explains

“how” the phenomenon happened, while the other explains “why” (Walsh 2015, 199). Consider, for instance, a mammal’s lowering of its body temperature. We may explain how the mammal’s thermoregulatory system does this by citing the process of vasodilation—the dilation of blood vessels to the skin—which increases the rate of heat exchange with the environment. Once we offer this causal explanation, there is a further and quite different question why the mammal does this. To explain why the mammal lowers its body temperature, we need to cite the goal of its thermoregulatory system to maintain its temperature within a certain range. We can see that the latter explanation captures a different aspect of the system’s modal profile that is not captured by the causal explanation. It allows us to infer, for instance, that if the mammal had been cold rather than warm, the thermoregulatory system would have triggered a different response—such as vasoconstriction or the shrinking of the blood vessel—to increase the body temperature instead.

117

Walsh’s account of teleological explanation is compelling. It offers a plausible view of what makes teleological explanation explanatory, and shows how causes and goals explain in the same way—namely by uncovering patterns of counterfactual dependence—despite offering distinct explanations. Causal explanations highlight the counterfactual dependence of effects on causes, while teleological explanations highlight the counterfactual dependence of means on goals. Lastly, Walsh’s account identifies a kind of regularity that is explanatorily relevant and empirically observable in an individual system. This particular kind of regularity in the modal profile of a goal-directed system—which Walsh calls purposive invariance—will be the core of my account of natural goodness. As I will argue in the next section, patterns of purposive invariance can replace the generic patterns of regularity in the natural-historical account of natural goodness, and give us an alternative account of natural goodness that does not involve a commitment to anti-individualism.

6.2 Toward an Alternative Account of Natural Goodness

Before I can offer my account of natural goodness, I need to address a concern that arises in relation to my use of Walsh’s approach to teleological explanation. I said earlier that we can understand the notion of goal that is implicit in Walsh’s account of teleological explanation based on the systems-theoretic notion of goal-directedness. But we also saw that the system- theoretic notion of goal-directedness is given in causal terms. This means that on this system- theoretic account, any physical system that tends toward some steady state of equilibrium in the right way would be characterized as a goal-directed system. As a result, the system- theoretic account seems to extend the concept of goal to systems that nobody would consider to be genuinely goal-directed. As Bedau (1992) has argued, for instance, many equilibrium systems such as a damped pendulum or a marble inside a bowl tend toward a steady state of

118 rest in a way that meets the criteria of the systems-theoretic account, and yet we do not consider such systems to be goal-directed.56 Specifying criteria that exclude non-goal-directed equilibrium systems turns is in fact a serious problem for the systems approach, and has led many to conclude that this approach ultimately fails to differentiate between genuine goal-directed systems and systems that merely behave as if they were goal-directed. Bedau (1992), for instance, argues that the causal dynamics of a system do not determine whether the system is goal-directed. He proposes that a distinctive feature of goal-directed systems is that their behavior is of value to something or someone, which is why the systems approach fails to capture goal-directedness in terms of merely causal, quantitative criteria. Now, if the systems approach fails to exclude non-goal-directed systems that merely behave as if they were goal-directed, our account of teleological explanation similarly extends these explanation to systems that merely exhibit the behavioral signs of a goal-directed system. So the modal profile that we called purposive invariance is not a sufficient condition for genuine purposive behavior. It merely marks a persistent and/or plastic tendency toward an end-state, which only amounts to a necessary condition for purposiveness at best. Hence, Walsh’s notion of purposive invariance—which we may call perseverance to avoid the implication of genuine purposiveness—is clearly not sufficient for giving an account of natural goodness. Just because a system perseveres toward an end-state, it doesn’t follow that it can be evaluated based on whether this end-state is achieved. Clearly, a pendulum whose bob is kept from returning to the state of rest is not thereby failing or defective, since there is no value associated with the state of rest to begin with. This means that if we are to give an account of natural goodness based on the notion of perseverance, we will have to supplement

56 As I mentioned earlier, it is not the aim of Walsh’s account of teleological explanation to offer an understanding of genuine goal-directedness, but rather to identify the aspects of goal-directedness that make it explanatory. So the concern that I am raising here is not necessarily a problem for Walsh’s view. It’s rather a concern about whether we can use Walsh’s view regarding teleological explanation as a basis for an account of natural goodness.

119 the account with additional criteria that can mark the difference between genuinely purposive and non-purposive cases. However, as I mentioned earlier, my aim is to use the patterns of regularity underlying teleological explanation to replace the generic patterns of regularity in the natural-historical account. And we can see that generic patterns of regularity—i.e., the patterns of regularity underlying generic judgments—are not inherently evaluative either. Broadly speaking, generic judgments pick out features that characterize instances of a kind, without necessarily implying that there is anything good about exhibiting these characteristic features. Note, for instance, that judgments such as “Toronto winters are cold”, “blue tits have a blue patch on their head” and “mosquitoes carry the West Nile virus” qualify as true generics, without having evaluative implications. This is why the natural-historical account of natural goodness, which we discussed in the previous chapter, is not solely based on genericity. As we saw earlier, Foot and Thompson’s account of natural goodness is based on natural-historical judgments, which are a particular kind of generic judgments and meet several other criteria beside genericity. Most importantly, natural-historical judgments have a teleological dimension. They can be connected with each other in teleological relations, such that we can say, e.g., that “male peacocks have a brightly colored tail in order to attract mates”. In other words, each natural-historical judgment describes a characteristic feature that is a means toward some other characteristic, natural- historical aspect of the life-form. And one can cite these other characteristic aspects of the life- form in answer to the question “why” the organism has the various parts and aspects that it has. We saw that a statement like “blue tits have a round blue patch on their head” does not amount to a true natural-historical judgment even though it is quite plausibly a true generic statement about a life-form. This is because unlike the brightly-colored tail of peacocks, the color of blue tits’ head is not a means toward other characteristic, natural-historical aspects of their life. So one cannot give an answer to the question “Why do blue tits have a round blue patch on their head?” by citing a further characteristic aspect of their life.

120

Moreover, the set of natural-historical judgments that capture the characteristic features and activities of a life-form together form a special unity or a system. This system of judgments describes the characteristic life-cycle or the natural history of the life-form. The elements of a natural history are teleologically related in such a way that their teleological relations together form a cycle. Each element of the natural history is teleologically linked to other elements of the system not just as a means, but also ultimately as an end. Take, for instance, a tiger’s ability to hunt prey. Not only is this characteristic a means toward further ends such as nourishment, it is also an end toward which other characteristics such as the tiger’s ability to run fast are a means. So each natural-historical judgment about a life-form is teleologically connected to other natural-historical judgments about the life-form in both directions. Thus, in order for a judgment like “The S is/does/has F” to be a true natural-historical judgment about a life-form S, describing a natural-historical characteristic F, the following conditions must be met:

(1) Genericity: “The S is/does/has F” is a true generic statement.

(2) Teleology: F is a means toward G, where G is another natural-historical characteristic of S. This is the case iff “The S is/does/has F in order to be/do/have G” is a true generic statement. (3) Reverse-Teleology: Another natural-historical characteristic of S, H, is a means toward F. This is the case iff “The S is/does/has H in order to be/do/have F” is a true generic statement. In fact, this characterization is still too broad to isolate the kind of judgment that

Thompson calls natural-historical judgment. A generic statement can be teleologically linked to other generic statements without describing a life-form. As Thompson points out, statements about artifacts and crafts can also meet the conditions specified above (2008, 79-80). We can say about cars, for instance, that “they have a carburetor in order to mix air and fuel”, “they mix air and fuel in order to burn the fuel”, and so on. But these generic statements are not

121 natural-historical judgment, and cannot be the basis of evaluations of natural goodness. As we saw in the previous chapter, natural goodness is an intrinsic kind of evaluation that applies to the parts and aspects of living things independently of the interest of humans or any external party. In the case of artifacts and crafts, the basis for evaluation of their parts and mechanisms is ultimately something about us. Cars are ultimately for our transportation, and if we evaluate a car’s carburetor based on how it mixes air and fuel, the basis for this evaluation will be our own ends and interests. In order to distinguish natural-historical judgments from artifactual judgments, Thompson suggests that there is a kind of “partial idealism” about artifacts and crafts in the following sense. The truth of a teleological statement about an artifact “presupposes that someone makes or has made the corresponding judgment, or at least some others belonging to the same system of judgments”. It simply cannot be true that cars are a means to transportation, or that a car’s carburetor is a means to mixing fuel and air, unless someone has made these judgments or at least other judgments regarding cars. As Thompson puts it, an unrecognized artifact is not really an artifact but “a merely possible one” (2008, 80). In contrast, natural-historical judgments can be true without anyone knowing about them or making any kind of judgment about the life-form in question. In fact, unrecognized life-forms are common. Thus, we may add a fourth condition to our list of conditions that natural-historical judgments must meet: (4) Independence: The truth of “The S is/does/has F in order to be/do/have G” and “The S is/does/has H in order to be/do/have F” does not presuppose that anyone makes the corresponding judgment or any other judgment about the life-form S.

Now, there are different ways in which we may interpret the teleological connective in conditions (2)-(4). What Thompson says about this connective is that, far from having anything to do with “the category of intention and psychic teleology”, it is simply used to organize the elements of a natural history. Natural teleological judgments, i.e., judgments of the form “The S is/does/has F in order to be/do/have G”, articulate the relations of dependence that hold among

122 the elements and aspects of a given life-form (2008, 78). They specify which elements and aspects of a life-form provide the conditions required for other elements and aspects to arise. The teleological connective, then, “simply expresses the concept that is converse to this conception of dependence” (2008, 79). So, to say that “male peacocks have a brightly colored tail in order to attract mates” simply means that having a brightly-colored tail is a condition that is required for peacocks to attract mates. One way to interpret this dependence relation among the elements of a natural history is as a mere causal dependence relation. On this interpretation, “The S is/does/has F in order to be/do/have G” simply means that F and G are two characteristic features of life-form S, where F causally leads to G. The idea would be that the various aspects and phases in the life of a life- form causally depend on, and contribute to, one another. And that the teleological connective ‘in order to’ merely denotes this causal relation, without implying anything further. However, this interpretation of Thompson’s account makes it inadequate as a basis for evaluations of natural goodness. If all that is required for a generic statement to count as a natural-historical judgment is that it describes a characteristic feature that causally contributes to other characteristic features of the type of entity in question, many non-living systems can also be the subject of true natural-historical judgments. Volcanoes, for instance, can be described by generic statements such as “volcanoes are formed when magma from within the earth’s upper mantle works its way to the surface” and “volcanoes erupt when pressure builds up”.57 Moreover, the characteristic features and phases described in these generic statements causally depend on each other. In fact, geologists talk about the ‘life-cycle’ of a volcano, describing the stages volcanoes go through as they form, become active, erupt, and become extinct. Nonetheless, it would be a mistake to evaluate volcanoes as good or defective based on how well they exhibit this characteristic life-cycle. The causal dependence relation between the

57 Note that these statements will remain true even if as a result of climate change volcanoes do not always exhibit this pattern perfectly.

123 different stages that they go through does not amount to a teleological relation of the right type, and it would be misleading to say “volcanoes are formed in order to erupt” and the like. Another example of a system with a characteristic set of causally interconnected stages that do not underwrite evaluations of natural goodness is the case of panic disorder, which we discussed in Chapter 4. We saw that panic attacks result from psychological states and dispositions that are themselves caused or strengthened by the attacks. But this recurrent cycle of causally linked processes is not enough to view a panic disorder as a basis for evaluations of natural goodness. It would be a mistake to evaluate the elements of a panic disorder based on how well they make a characteristic contribution toward the recurrence of attacks and the persistence of the cycle. Thus, if natural-historical judgments are to serve as the basis for evaluations of natural goodness, the dependence relation among the elements of a natural history cannot be interpreted as a mere causal relation. To correctly issue evaluations of natural goodness, the relevant notion of teleological dependence has to imply that the characteristic feature described in a natural-historical judgment contributes to something that is a genuine good of the life-form. What is missing in the case of volcanoes and panic attacks is that the contribution of the characteristic elements and stages of the system cannot be plausibly viewed as good. This is why the causal dependence relations that they instantiate do not amount to the relevant kind of teleological relation. The truth of “S is/does/has F in order to be/do/have G” requires not only that F causally contribute to G, but also that F and G are good in the life of S. So I propose that we incorporate this interpretation of the teleological dependence relation in our conditions (2) and (3) above, and offer the following account of natural- historical judgments: The statement “The S is/does/has F” is true natural-historical judgment about life-form S iff: (1) Genericity: “The S is/does/has F” is a true generic statement. (2) Teleology: “The S is/does/has F in order to be/do/have G” is a true generic statement, where G is a natural good in S.

124

(3) Reverse-Teleology: “The S is/does/has H in order to be/do/have F” is a true generic

statement, where H is a natural good in S. (4) Independence: The truth of “The S is/does/has F in order to be/do/have G” and “The S is/does/has H in order to be/do/have F” does not presuppose that anyone makes the corresponding judgment or any other judgment about the life-form S. It is worth noting that Foot’s discussion of the teleological dimension of natural- historical judgments actually suggests a similar interpretation of teleology. In her discussion of what she sees as “a gap” in Thompson’s account, Foot argues that in order to support inference to evaluations of natural goodness, natural-historical judgments must describe a feature that serves a function, or plays a part, in the life characteristic of the life-form. What counts as “the life” of a kind of living thing, however, is understood in terms of essential goods such as survival and reproduction. Foot believes that for plants and non-human animals, “the life” characteristic of the life-form has to do, directly or indirectly, with development, self- maintenance, and reproduction. In her view, this is why a statement like “blue tits have a round blue patch on their head” does not support inference to evaluative judgments. The color of the head plays no part in any of these aspects of “the life” of the bird. In the case of human beings, of course, essential goods go beyond this list and are much more diverse. In fact, one could argue that even in the case of non-human living things, there can be essential goods other than the three goods that Foot considers.58 But regardless of what exactly the goods of a given kind of living thing consist in, it’s important to see that the teleological dependence relation in conditions (2) and (3) has to be understood in terms of these goods. Unless a feature makes a characteristic contribution toward these goods, it does not have a place in the natural history of the life-form. In Foot’s words, for any element of the natural history, we must be able to ask: “what’s the point of it?” and “what good does it do?” (Foot 2001, 30-31).

58 As I mentioned in chapter 4, for instance, Hursthouse considers enjoyment and freedom from pain among the characteristic goods in the life of sentient animals.

125

Now, if the teleological component of the natural-historical account cannot be understood without presupposing an evaluative conception of “the life” of living things, then this account does not offer a reduction of natural goodness. Purely non-evaluative criteria such as patterns of genericity or relations of causal dependence are not enough to determine the natural good of a life-form on the natural-historical account. We rather need to presuppose an already evaluative conception of “the life” of a given kind of organism before we can make the natural-historical judgments that are the basis of evaluations of natural goodness. However, the fact that the natural-historical account does not reduce natural goodness in non-evaluative terms does not mean that it is uninformative or that it contributes nothing to our understanding of natural goodness. What is distinctive about the natural-historical account is that it identifies genericity as a necessary condition for standards that can underwrite evaluations of natural goodness. Note that the general idea that survival and reproduction are good in the life of plants and animals is not by itself enough to determine specific evaluations of their parts and aspects. Almost any feature or aspect of an organism might contribute to a good like survival under the right conditions. But we don’t consider just any contribution to survival as an instance of natural goodness, and we don’t consider just any failure to contribute to survival as an instance of natural defect. By making genericity a necessary condition, the natural-historical account identifies a basis for turning general plausible assumptions about the good of living things (e.g., that survival and reproduction are generally good) into specific standards of evaluation that apply in the case of a given kind of organism. Consider the following example from Micah Lott. A particular arrangement of fur on a tiger’s face might discourage a hunter from shooting the tiger if the pattern happens to remind him of his mother’s face (2012a, 371). But we do not thereby evaluate this pattern of fur as naturally good. The natural-historical account explains why such an inference is not warranted by appealing to patterns of genericity: it’s neither characteristic of tigers to have patterns of this particular shape, nor characteristic of such patterns to contribute to a tigers’ survival. In contrast, a tiger’s sharp eyesight or the ability to run fast not only make a contribution toward

126 the tiger’s survival, but also do so systematically and characteristically, by enabling the tiger to engage in other characteristic behaviors such as hunting prey. We can make true generic statements saying “tigers run fast” or “tigers run fast in order to hunt prey”. On the natural- historical account, this is why characteristics like running fast and having sharp eyesight are naturally good in the life of a tiger, and lacking them is a defect. Thus, although the account is not reductive, it identifies a basis for interpreting the parts and aspects of living things and making evaluations of natural goodness and defect. Admittedly, there is a recursiveness in the way the natural-historical basis of evaluation is understood. The truth-value of natural-historical judgments depends on the good of the organism, while what counts as the good of the organism is itself determined by the natural- historical judgments that are true of it. However, as Lott (2012a) has argued, this doesn’t mean that this natural good is not answerable to empirical observation or natural facts. What it means is that in trying to formulate true natural-historical judgments, “our thought does not go ‘outside’ the system of natural-historical judgments that represent the life-form”. Our conception of the natural history of a life-form shapes our observations of individual living things, but these observations in turn revise and refine our conception of the life-form. In Thompson’s words, there is a “reciprocal mutual interdependence” (2004, 52) between observations of individual living things and natural-historical judgments about their life- form.59 So the fact that certain general aspects of our conception of the good of a life-form are not derived from observation does not mean that this concept is entirely immune from revision based on empirical evidence. And as we saw in the case of the tiger life-form, it’s the condition of genericity that determines how empirical observations inform our conception of the life- form under investigation. Now, I argued in the previous chapter that this component of genericity in the natural- historical account leads to an anti-individualistic conception of life-forms, which in turn results

59 See Thompson (2004, 47-51) for an example of how our conception of a life-form and our empirical observations of individual living things mutually inform each other.

127 in a class of potentially problematic implications for the account where it makes the wrong evaluation of individuals with a unique form. This is why I set out to offer an alternative account of natural goodness in this chapter. To avoid the implications of anti-individualism, then, my account needs to drop the component of genericity in the natural-historical account. But this means that I cannot appeal to patterns of genericity across instances of a given life- form to explain what determines specific evaluations of natural goodness and defect. In other words, I cannot explain the difference between the arrangement of fur in a tiger’s face and its ability to run fast by appealing to generic patterns found across members of the species tiger. I need to offer a different basis for evaluations of natural goodness. And that’s where the patterns of invariance that I called perseverance come in. I started this section with acknowledging that the modal profile of perseverance is not sufficient for genuine purposiveness. A non-purposive system such as a pendulum can persevere to an end-state without being purposive or underwriting evaluative judgments based on its patterns of perseverance. This means that patterns of perseverance do not amount to a sufficient basis for an account of evaluations of natural goodness. However, as I argued above, neither are the patterns of genericity that are employed in the natural-historical account. Foot and Thompson supplement the criterion of genericity with additional, arguably evaluative criteria, which results in a non-reductive account of natural goodness. As we will see in the next section, my alternative account of natural goodness is similarly non-reductive, and involves evaluative criteria. I will argue that my account also gives a basis for evaluations of natural goodness that enables us to turn general plausible assumptions about the good of living things into specific standards of evaluation, and explains the relation between these evaluations and empirical observation. Nonetheless, I will show that my account does all of this without making any reference to genericity, and therefore does not share the problematic anti-individualistic implications of the natural-historical account.

128

6.3 A Modal-Explanatory Account of Natural Goodness

In a nutshell, my account of natural goodness replaces the criterion of genericity in the natural-historical account with perseverance, which—as we have seen—is the kind of modal profile underlying teleological explanation. Instead of grounding the standards of evaluation in generic patterns found among the instances of a life-form, my account grounds these standards in patterns that would be found in the modal profile of a single individual. In other words, instead of relying on what “the S is/does/have” (in the generic sense), my account will rely on what a given instance of S perseveres toward being/doing/having. To see how my account does this, let’s first put together the elements of the natural-historical account of natural goodness. As we saw in the previous section, a true natural-historical judgment is a statement that meets four conditions: (1) Genericity, (2) Teleology, (3) Reverse-Teleology, and (4) Independence. We have also seen that on the natural-historical account, evaluations of natural goodness are determined based on natural-historical judgments, such that a given feature in an individual organism is naturally good when it matches the characteristic feature expressed in a true natural-historical judgment about the individual’s life-form, and it is naturally defective when it doesn’t so match. Thus, we can say that the natural-historical account defines natural good as follows: Natural-Historical Goodness For any individual organism x that is an instance of the life-form S, being/doing/having F is a natural good iff: (1) Genericity: “The S is/does/has F” is a true generic statement. (2) Teleology: “The S is/does/has F in order to be/do/have G” is a true generic statement,

where G is a natural good in S. (3) Reverse-Teleology: “The S is/does/has H in order to be/do/have F” is a true generic statement, where H is a natural good in S.

129

(4) Independence: The truth of “The S is/does/has F in order to be/do/have G” and “The S

is/does/has H in order to be/do/have F” does not presuppose that anyone makes the corresponding judgment or any other judgment about the life-form S. Now, a closer look at the elements of the account above reveals that condition (1) is actually redundant, because it is implied in conditions (2) and (3). As Thompson has argued, the truth of the generic statement “P in order to Q” entails both P and Q (2008, 78). Thus, we can eliminate the first condition, and simplify the natural-historical account of natural goodness as follows: Natural-Historical Goodness For any individual organism x that is an instance of the life-form S, being/doing/having F is a natural good iff:

(1) Teleology(GEN): “The S is/does/has F in order to be/do/have G” is a true generic statement, where G is a natural good in S.

(2) Reverse-Teleology(GEN): “The S is/does/has H in order to be/do/have F” is a true generic statement, where H is a natural good in S.

(3) Independence(GEN): The truth of “The S is/does/has F in order to be/do/have G” and “The S is/does/has H in order to be/do/have F” does not presuppose that anyone makes the corresponding judgment or any other judgment about the life-form S.

Note that I have used the subscript (GEN) to indicate that all the three conditions are defined in terms of generic statements. Now, my account of natural goodness similarly involves conditions that capture the criteria of teleology, reverse teleology, and independence. But on my account, these conditions are defined differently, and based on the explanatory modal profile of a naturally purposive system. This account is as follows: Modal-Explanatory Goodness For any individual organism x, being/doing/having F is a natural good iff:

(1) Teleology(EXP): If x were to be/do/have F, x’s being/doing/having F would be teleologically explained by G, where G is a natural good in x.

130

(2) Reverse-Teleology(EXP): If x were to be/do/have F, x’s being/doing/having F would teleologically explain H, where H is a natural good in x.

(3) Independence(EXP): The teleological explanation of F by G and the teleological explanation of H by F would be independent of the goods of any entity other than x. Let’s go over the three components of the account in turn. The first two conditions,

Teleology(EXP) and Reverse-Teleology(EXP), capture the idea that the natural goods of a living organism are teleologically interconnected. F is a natural good of the organism only if, on one hand, there is another natural good of the organism, G, such that F would be a means toward G, and on the other hand, there is another natural good of the organism, H, that would be a means toward F. Much like the conditions of Teleology(GEN) and Reverse-Teleology(GEN) in the natural-historical account, Teleology(EXP) and Reverse-Teleology(EXP) presupposes a prior, evaluative conception of the life of the organism. Recognizing a natural good of organism x requires presupposing at least some other natural goods of x. But here, instead of understanding the means-end relation between the natural goods in terms of generic statements featuring the ‘in order to’ clause, this relation is understood in terms of teleological explanation and the kind of modal profile that it involves. On my account, if P is a means toward Q, then P is teleologically explained by Q, where teleological explanation is understood in terms of Walsh’s account discussed in section 6.1. Thus, the conditions of Teleology(EXP) and

Reverse-Teleology(EXP) require that there would be a counterfactual dependence relation of perseverance between F and G and between H and F respectively.

It’s worth noting that the conditions expressed in Teleology(EXP) and Reverse-

Teleology(EXP) do not require that x actually is/does/has F. In other words, F can be a natural good for x even if it is not achieved in x. By the same token, G or H can be a natural good for x even if they are not achieved in x. What the conditions require is rather that if x were to be/do/have F, its being/doing/having F would fit in patterns of perseverance involving other natural goods like G and H. This is how the account allows for the occurrence of natural defect in an organism, i.e., cases where the organism lacks something that is a natural good. On the

131 natural-historical account, this possibility simply follows from the logic of generic statements.

It can be true that “tigers run fast in order to hunt prey” even if a particular tiger is unable to run fast. On the modal-explanatory account, the possibility of defect is captured by the idea that even if x lacks F, it can be true that the modal profile of x is such that, on the one hand being/doing/having F would make a persistent and/or plastic contribution toward G, and on the other hand H would make a persistent and/or plastic toward being/doing/having F. In a tiger that actually cannot run fast due to an injury, for instance, it can be true that if it were to run fast, doing so would make a persistent and/or plastic contribution toward hunting prey, such that we would teleologically explain the tiger’s running fast as a means toward hunting prey. Similarly, it can be true that in such a tiger other aspects of the life of such a tiger, like having four legs, would make a persistent and/or plastic contribution toward running fast, even if running fast is not actually achieved.

The second condition, Independence(EXP), captures the idea that evaluations of natural goodness are distinct from evaluations of artifacts and crafts in that they do not depend on the interests of us or any beneficiaries external to the system. We saw that on the natural- historical account this criterion is formulated in terms of whether anyone needs to actually make the relevant natural-historical judgments. On the modal-explanatory account, the idea is formulated in terms of teleological explanation. In the case of artifacts and crafts, the teleological explanation of a given feature or part of the system ultimately goes beyond the system itself. That is because what ultimately underwrites the teleological relations that hold between the parts and mechanisms of an artifact is the fact that it is a means toward our ends. A car’s carburetor, for instance, can be teleologically explained as a means toward the end of mixing air and fuel in the proper air–fuel ratio for combustion.60 The mixing of air and fuel is in turn teleologically explained as a means toward the combustion of the fuel, which is itself a means toward moving the car. But this explanation ultimately depends on what is good for us.

60 Note that the carburetor has the right modal profile for teleological explanation. Under a range of conditions, it mixes air and fuel in the ratio that is appropriate for combustion regardless of, for instance, how much fuel is in the tank.

132

What underwrites the status of the moving of the car as an end is the fact that it is a means toward our end of transportation. In other words, the complete teleological explanation of the carburetor will have to make reference to our goods and interests. In contrast, in the case of living organisms, the teleological explanation of the system’s parts and processes will be independent of the goods of any entity other than the system itself. To use our earlier example, a tiger’s ability to run fast can be explained as a means toward the goal of hunting prey. This goal in turn can be teleologically explained in terms of further goods such as nutrition and survival. But the chain of teleological relations does not have to go beyond the goods of the tiger itself for the explanation to be complete.61 As we saw earlier, the natural goods of the organism instantiate teleological relations that form a cycle. So even though each natural good is teleologically explained in terms of another natural good, we don’t need to leave the cycle of the goods of the organism to complete the explanation.62 Now we can see how the modal-explanatory account compares to the natural-historical account in making evaluations of natural goodness and defect. Consider again the case of a tiger with an unusual arrangement of fur on its face that happens to resemble a hunter’s mother’s face. The natural-historical account explains why this feature is not a natural good by appealing to the fact that it is neither characteristic of tigers in general to have it, nor characteristic of this feature to contribute to tigers’ survival. In contrast, the modal- explanatory account explains the fact that this feature is not a natural good in a very different way, by appealing to the modal profile of the very tiger that has the unusual face. Although the

61 Note that the claim here is not about the causal explanations that apply in the case of an organism. The causal explanation of an organism’s parts and aspects can very much depend on the goods of external entities—for instance, if it is an organism that we have artificially bred to serve interests of our own. The claim is rather that the teleological explanation of these parts and aspects as means towards the natural goods of the organism is independent of the goods of external entities. 62 Note that this doesn’t require that all the goods of an organism are placed on a single cycle, but rather that there is at least one cycle, exclusively consisting of the goods of the organism, that they are part of. Because of this, not every natural good of the organism has to be a means toward survival or self-maintenance in order to be part of a cycle. Human character traits that constitute different virtues, for instance, can be part of a cycle of virtuous traits that are interdependent without necessarily contributing to survival.

133 arrangement of fur on the face of the tiger happens to makes a contribution toward survival, this contribution is not part of a persistent and/or plastic pattern of contribution toward survival. Looking at the modal profile of the tiger reveals that the arrangement of fur on the tiger’s face does not persevere toward the good of survival, and is not teleologically explained by it. The relation between the arrangement of fur and the hunter’s being discouraged from shooting does not show a pattern of counterfactual dependence beyond the invariance relation between a cause and effect. If there was a means-end relation between the two, there would be a different and additional pattern of counterfactual dependence between them, such that in a range of conditions, the pattern of fur on the tiger’s face would make a persistent and/or plastic contribution toward discouraging a hunter from shooting. But this is not the case. The relation between the two is contingent and merely causal. In other words, it’s similar to the case of buying sweet lemons as a result of accidentally finding them in the store as opposed to going to the store with the goal of buying sweet lemons. This is why, on the causal- explanatory account, the particular arrangement of fur on the tiger’s face is not a natural good. The tiger’s modal profile can also explain why certain other features such as running fast and having sharp eyesight are natural goods in a tiger’s life. Consider a tiger’s ability to run fast, which contributes to hunting prey and obtaining food among other things. In contrast to the pattern of fur on the tiger’s face, running fast can be viewed a genuine means toward the good of survival, because the contribution it makes toward survival is persistent and plastic. It is not just that running fast enables the tiger to obtain food in a one-off case under highly contingent circumstances. There are a range of actual and counterfactual conditions in which this ability would contribute to the tiger’s hunting prey and obtaining food. In fact, as I said above, the modal profile has no problem identifying this good in a tiger that actually lacks the ability to run fast. Even if a tiger lacks this ability due to an injury or a congenital problem, it can still be the case if the tiger were to have the ability to run fast, this would be a means toward the goods of hunting prey and obtaining food. This is why, on the modal-explanatory

134 account, running fast is a natural good in a tiger regardless of whether or not the tiger actually has the ability. Moreover, although the account allows us to identify natural goods that are missing in an organism due to a defect, it does not characterize just any potentially beneficial feature as naturally good. In the case of a tiger, for instance, having wings and the ability to fly might be beneficial and enable the organism to achieve its ends more successfully. But this doesn’t mean that this is a natural good in the life of a tiger, or that a tiger is defective for not having wings. So a suitable account of natural goodness should be able to explain why a tiger without legs is defective while a tiger without wings is not. On the modal-explanatory account, the difference between the two cases is explained in terms of modal profile of an individual tiger. What prevents us from ascribing a natural good to having wings in the life of a tiger is the condition of Reverse-Teleology(EXP). If having wings was a natural good in the life of a tiger, there would be other aspects of a tiger’s life that persevere toward having wings and would be teleologically explained as a means toward this end. But a tiger’s anatomy and morphology simply do not have the right modal profile to meet this condition. The parts and aspects of a tiger’s anatomy and morphology are best explained as means toward locomotion by walking and running as opposed to flight. Finally, note that unlike the natural-historical account, the modal-explanatory account is not committed to the idea that members of a species need to be evaluated based on the same standard. It is possible for different individuals to have different modal profiles resulting in different evaluations. To see this, consider again the case of Slijper’s two-legged goat discussed in Chapter 5. Suppose that the goat has adapted to its condition by developing enlarged hind limbs, a curved spine, and an unusually large neck, which enable it to walk and run by using its hind legs alone. As a result of these adjustments, the modal profile of the goat is geared toward moving around by walking on two legs. It is not that the two legs make a one-off contribution to the goat’s moving around under highly contingent circumstances. Rather, due to the goat’s anatomy, having two legs fits into a system of locomotion that makes a persistent and plastic

135 contribution toward moving with two legs. This is why, on the modal-explanatory account, having two legs can be characterized as a natural good in the life of this goat, even though it might amount to a natural defect in the life of other goats. Note that on this account having four legs is actually not a natural good in the life of Slijper’s goat, for the same reason that having wings or fins is not a natural good in its life. The modal profile of this goat is not geared toward having such characteristics and using them as a means toward obtaining its goods. The goat’s anatomy is simply not suitable for walking on four legs in the same way that it is not suitable for flying or swimming. It’s rather suitable for walking and hopping on two legs, which is why its natural good consists in having two legs rather than four. We may contrast Slijper’s goat with a more ordinary goat that has lost two legs in an injury, or is born with two legs due to a congenital condition without having been able to adjust to having two legs during development. Such a goat does not have enlarged hind limbs or a curved spite, and is not capable of hopping on two legs. The modal profile of such a goat is actually geared toward having four legs, even though it only has two. Thus, unlike Slijper’s goat, this goat is defective in that it fails to have four legs. What is naturally good in the life of this goat is having four legs rather than two. Thus, the modal-explanatory account makes the same evaluations as the natural- historical account in cases where these evaluations are intuitively plausible. But it makes different evaluations in the case of individuals with unique characteristics where the natural- historical account issues potentially problematic evaluations due to its commitment to anti- individualism. In other words, the modal-explanatory account is a viable alternative to the natural-historical account that does not share its problematic, anti-individualistic implications.

6.4 Concluding Remarks: Natural Goodness without Anti-Individualism

In the previous chapter, I highlighted the distinction between the general concept of natural goodness and Thompson’s specific, natural-historical conception of this concept. I argued that the natural-historical conception of natural goodness is anti-individualistic, and issues

136 implausible evaluations of individuals with a unique from. I also argued that neo-Aristotelians have no independent basis for anti-individualism aside from the assumption that the natural- historical account is the only viable conception of natural goodness. As we saw in section 5.3, Thompson’s argument against individualism ultimately came down to the claim that without the extra-individualistic context, there would be no basis for distinguishing between goodness and defect. He argued that in the case of a swamp creature, there is no basis for determining whether its molecular equivalents of human arms really are arms or rather are horribly deformed wings. In this chapter, however, I have offered just such a basis. The modal- explanatory account of natural goodness is a viable alternative to the natural-historical account that can draw a plausible distinction between goodness and defect even in the case of an isolated individual. It can explain why a swamp creature’s molecular equivalents of human arms are really arms and not deformed wings: the modal profile of the swamp creature is exactly like that of a human. In the same way that a human’s anatomy and structure is geared toward using arms rather than wings, the swamp creature is structured in a way that is suitable for using arms rather than wings.

Not only does the modal-explanatory account give an intuitive verdict regarding the various cases discussed above, it also has desirable implications regarding our understanding of human form. On one hand, the account allows for a pluralistic conception of human form, where different individual human beings can potentially have different forms that give rise to different standards of evaluation. On the other hand, the standard of evaluation for each individual is grounded in facts about the individual itself rather than facts about other members of the species that the individual happens to be born into. This pluralist and individualist conception of human form is desirable for many reasons, but a particularly salient one is that it enables us to present a more nuanced view of human disability. One of the objections raised against Foot’s version of neo-Aristotelian naturalism is that it labels the lack of any characteristic ability in human beings as a defect, while it’s not clear at all that departure from a characteristic ability necessarily makes a disabled person defective

137

(Woodcock 2006). Many philosophers defend a value-neutral view of disability, according to which the lack of a physical or mental ability that other people have is a neutral feature that that is not in itself bad. Elizabeth Barnes (2014; 2016), for instance, argues that having a disability is a mere difference from others. It is a difference that makes you a minority, but it does not necessary make you worse off because of the difference. Defenders of the value- neutral view don’t deny that having a disability—in the sense of lacking an ability that ‘normal’ people have—can involve the loss of some goods or that it can ultimately result in being worse off. But they argue that the relation between having the disability and being worse off is highly contingent and socially constructed. Moreover, they argue that precisely what makes you lose out on some goods can allow you to participate in other goods that are unique to disability. It might be true, for instance, that the ability to hear is a good that deaf people lack. But there are other goods, such as the unique experience of a signed language as first language, which deaf people enjoy instead.63 The natural-historical conception of natural goodness makes it difficult to accommodate such a view of disability due to its commitment to the idea that the standards of goodness are necessarily shared across the members of a life-form. Consider a person with a disability such as deafness from birth, who is perfectly happy as a deaf person and embraces being deaf and participates in the unique set of goods that come with this identity. On the natural-historical account, lacking the ability to hear amounts to a defect, because “humans have a sense of hearing” is, quite plausibly, a true natural-historical judgment. Yet, it’s not obvious that the deaf person in our example should be viewed as naturally defective merely because “humans have a sense of hearing”. In contrast to the natural-historical account, my modal-explanatory account can avoid systematically treating every case of deafness as a defect, because on my account the basis for evaluation is the modal profile of the individual person. And it’s at least in principle possible for the physical and social aspects of the life of a deaf person to be

63 In fact, over the last few decades, disability rights activists have advocated an affirmative conception of disability (Swain and French 2000), according to which disability is a positive social identity that can be actively embraced and celebrated.

138 arranged in such a way that their deafness does not constitute a natural defect in their life. Of course, this is merely scratching the surface on the topic of disability and human form. I have neither offered a conclusive defence of the value-neutral view of disability, nor shown that Foot and Thompson’s natural-historical account of natural goodness has no way of accommodating such a view. I have only presented a few rudimentary considerations to illustrate how the fact that the modal-explanatory account is consistent with individualism can work in its favor when it comes to offering an understanding of human form. In summary, I have argued for a novel account of natural goodness that is distinct from Foot and Thompson’s natural-historical account. On my modal-explanatory account, what determines evaluations of natural goodness are the patterns of counterfactual dependence among the parts and aspects of an organism that underwrite teleological explanations of these parts and aspects. What is distinctive about this account is that instead of relying on generic judgments about a species, it grounds the distinction between goodness and defect in an explanatorily significant type of modal profile that is instantiated in an individual organism. Thus, I have argued that we can have a conception of natural goodness that is consistent with individualism and in line with the diversity and plasticity that is manifest everywhere in realm of living things. I started this dissertation by noting that contemporary neo-Aristotelian naturalism explains morality as continuous with the domain of life while remaining objectionably detached from the scientific understanding of life in biology. I set out to mend the relationship between neo-Aristotelian ethics and the science of biology by engaging the concepts and methods of philosophy of biology. In the first two chapters of the dissertation, I articulated how neo-Aristotelian naturalists argue for ethical naturalism, and explained why biology is relevant to assessing this argument. I proceeded in the next two chapters to offer a novel defence of some of the commitments of neo-Aristotelian naturalism by appealing to the explanatory concepts of biology. Particularly, I argued that the concepts of flourishing and flourishing-based function are necessary for understanding the nature of living things, and can

139 underwrite a notion of natural goodness that is continuous across human and non-human life.

Finally, in the last two chapters, I have demonstrated how paying a closer attention to what biology tells us about living things can help us revise certain aspects of the standard neo- Aristotelian account. Drawing from philosophical accounts of scientific explanation in biology, I have offered a novel account of the flourishing-based evaluations of natural goodness that can better accommodate the developmentally and evolutionarily significant phenomenon of phenotypic plasticity. The metaethical framework I have presented in this dissertation is therefore distinctive for incorporating our scientific understanding of living things, not only in the way it argues for the continuity between ethics and the domain of life, but also in its account of what this continuity consists in.

140

Bibliography

Adams, F. (1979). A goal-state theory of function attributions. Canadian Journal of Philosophy, 9, 493–518.

Andreou, C. (2006). Getting on in a varied world. Social Theory and Practice, 32(1), 61–73.

Annas, J. (2005). Virtue ethics: What kind of naturalism?. In: S. M. Gardiner (Ed.) Virtue Ethics, Old and New (pp. 11-29). Cornell University Press.

Barandiaran, X., Di Paolo, E., & Rohde, M. (2009). Defining agency: individuality, normativity, asymmetry and spatio-temporality in action. Journal of Adaptive Behavior, 17(5), 367–386.

Barnes, E. (2014). Valuing disability, causing disability. Ethics, 125(1), 88-113.

———. (2016). The minority body: a theory of disability. Oxford University Press.

Bartke, A., Sun, L., & Longo, V. (2013). Somatotropic signaling: trade-offs between growth, reproductive development, and longevity. Physiological reviews, 93 (2), 571-598.

Bedau, M. (1991). Can biological teleology be naturalized? Journal of Philosophy, 88, 647–55.

———. (1992). Where’s the good in teleology? Philosophy and Phenomenological Research, 52, 781–806.

Bertanlanffy, L. (1969). General systems theory. New York: Braziller.

Bigelow, J., & Pargetter, R. (1987). Functions. Journal of Philosophy, 84, 181-97.

Boorse, C. (1976). Wright on functions. Philosophical Review, 85, 70–86.

Boyd, R. (1988). How to be a moral realist. In: J. Sayre-McCord (Ed.) Essays on Moral Realism (pp. 187–228). Cornell University Press.

Braithwaite, R. B. (1964). Scientific explanation: a study of the function of theory, probability and law in science. Cambridge University Press.

Brink, D. O. (1986). Externalist moral realism. Southern Journal of Philosophy, 24 (Supplement): 23–40.

141

Clark, D. (1986). A cognitive approach to panic disorder. Behaviour Research and Therapy, 24,

461–70.

———. (1997). Panic disorder and social phobia. In D. M. Clark, & C. G. Fairburn (Ed.) Science and practice of cognitive behaviour therapy (pp. 119–53). Oxford University Press.

Christensen, W., & Bickhard, M. (2002). The process dynamics of normative function. The Monist, 85(1), 3–28.

Cummins, R. (1975). Functional analysis. Journal of Philosophy, 72, 741-65.

Dancy, J. (2006). Nonnaturalism. In: D. Copp (Ed.) The Oxford handbook of ethical theory. Oxford University Press.

Darwall, S., Gibbard, A., & Railton, P. (1992). Toward fin de siecle ethics: some trends. The Philosophical Review, 101(1), 115-189.

Darwin, C. (1859[1872]). The origin of species by means of natural selection. Modern library.

Dawkins, R. (1976). The selfish gene. New York: Oxford University Press.

———. (1983). The extended phenotype: the long reach of the gene. New York: Oxford University Press.

Di Paolo, E. (2005). Autopoiesis, adaptivity, teleology, agency. Phenomenology and the Cognitive Sciences, 4, 429–452.

Dobzhansky, T. (1937). Genetics and the origin of species. New York: Columbia University Press.

Fitzpatrick, W. J. (2000). Teleology and the norms of nature. Taylor & Francis.

Enoch, D. (2011). Taking Morality Seriously: A Defense of Robust Realism. Oxford University Press on Demand.

Foot, P. (2001). Natural goodness. Oxford University Press.

Frank, R. H. (1985). Choosing the right pond: human behavior and the quest for status. Oxford University Press.

142

Frey, J. (2018). How to be an ethical naturalist. In: J, Hacker-Wright (Ed.) Philippa Foot on

Goodness and Virtue (pp. 47-84), Palgrave.

Futuyma, D. J. (1954) Evolution. Sunderland.

Gánti, T., (2003). The principles of life. Oxford University Press.

Garson, J. (2016). A critical overview of biological functions. Springer.

———. (2017). Against organizational functions. Philosophy of Science, 84, 1093–1103.

Godfrey-Smith, P. (1994). A modern history theory of functions. Noûs, 28, 344-62.

Guevara-Aguirre, J., et al. (2011). Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans. Science translational medicine, 70(3), 70ra13-70ra13.

Hacker‐Wright, J. (2009a). What is natural about Foot's ethical naturalism? Ratio, 22(3), 308- 321.

———. (2009b). Human nature, personhood, and ethical naturalism. Philosophy, 84 (329), 413- 427.

———. (2013). Human nature, virtue, and rationality. In: J. Peters (Ed.), Aristotelian ethics in contemporary perspective (pp. 83-96).

Hall, B. K. (2003). Evo-Devo: evolutionary developmental mechanisms. International Journal of Developmental Biology, 47(7/8), pp. 491-496.

Hirshleifer, D., & Rasmusen, E. (1989). Cooperation in a repeated prisoners’ dilemma with ostracism. Journal of Economic Behavior and Organization, 12(1), 87–106.

Hrdy, S. B. (1999). Mother nature: a history of mothers, infants, and natural selection. New York:

Pantheon Books.

Hursthouse, R. (1999). On virtue ethics. Oxford University Press.

Huneman, P. (2010). Assessing the prospects for a return of organisms in evolutionary biology. History and Philosophy of the Life Sciences, 32(2/3), 341–371.

143

Jablonka, E. (1995). Epigenetic inheritance and evolution : the Lamarckian dimension. New York:

Oxford University Press.

———. (2005). Evolution in four dimensions : genetic, epigenetic, behavioral, and symbolic variation in the history of life. Cambridge, Mass.: MIT Press.

Jackson, F. (1998). From metaphysics to ethics: a defence of conceptual analysis. Clarendon Press.

Johnston, T. D., & Gottlieb, G. (1990). Neophenogenesis: A developmental theory of phenotypic evolution. Journal of Theoretical Biology, 147(4), 471-495.

Kauffman, S. (1993). The origins of order: self-organization and selection in evolution. Oxford: Oxford University Press.

Kitcher, P., & Vickers, L. A. (2003). Pop socio-biology reborn: The evolutionary psychology of rape and violence. In P. Kitcher (Ed.), In Mendel’s Mirror: Philosophical Reflections on Biology (pp. 333–355). Oxford: Oxford University Press.

Klein, U. (2005). Shifting ontologies, changing classifications: plant materials from 1700 to 1830. Studies in History and Philosophy of Science Part A, 36(2), 261-329.

Kraemer, D. M. (2014). Revisiting recent etiological theories of functions. Biology and Philosophy, 29, 747–759.

Krohs, U., (2011). Functions and fixed types: Biological and other functions in the post- adaptationist era. Applied Ontology, 6, 125-139.

Laland, K. N., Odling-Smee, J., Hoppitt, W., & Uller, T. (2013). More on how and why: cause and effect in biology revisited. Biology & Philosophy, 28(5), 719-745.

Laland, K. N., Uller, T., Feldman, M., Sterelny, K., Müller, G. B., Moczek, A., ... & Futuyma, D. J.

(2014). Does evolutionary theory need a rethink? Nature, 514(7521), 161.

Laron, Z., et al. (2017). IGF-I deficiency, longevity and cancer protection of patients with Laron syndrome. Mutation Research/Reviews in Mutation Research, 772, 123-133.

144

Laubichler, M. D., & Wagner, G. P. (2000). Organism and character decomposition: Steps

towards an integrative theory of biology. Philosophy of Science, 67, 289-300.

Leslie, S. J. (2008). Generics: cognition and acquisition. Philosophical Review, 117(1), 1-47.

Lewens, T. (2010). Foot note. Analysis, 70, 468-473.

Lewontin R. (1970). The units of selection. Annual Review of Ecology and Systematics, 1, 1-14.

Liem, K. F. (1973). Evolutionary strategies and morphological innovations: cichlid pharyngeal jaws. Systematic Zoology, 22(4), 425-441.

Lloyd, E. A. (2001). Science gone astray: evolution and rape. Michigan Law Review, 99(6), 1536– 1559.

Lott, M. (2012a). Have elephant seals refuted Aristotle? Nature, function, and moral goodness. Journal of Moral Philosophy, 9(3), 353-375.

———. (2012b). Moral virtue as knowledge of human form. Social Theory and Practice, 38(3), 407–431.

Ludewig, S., Geyer, M. A., Ramseier, M., Vollenweider, F. X., Rechsteiner, E., & Cattapan- Ludewig, K. (2005). Information-processing deficits and cognitive dysfunction in panic disorder. Journal of Psychiatry and Neuroscience, 30, 37–43.

Lutz, M., & Lenman, J. (2018). Moral naturalism. The Stanford encyclopedia of philosophy (Fall 2018 Edition), Edward N. Zalta (ed.), URL = https://plato.stanford.edu/archives/fall2018/entries/naturalism-moral/

Maguire, B. (2017). The autonomy of ethics. In: T. McPherson, D. Plunkett (Ed.) The Routledge handbook of metaethics (pp. 431-442). Routledge.

Mayr, E. (1961). Cause and effect in biology. Science, 134(3489), pp. 1501-1506.

McCloskey, M. (1983). Naive theories of motion. In D. Gentner & A. L. Stevens (Eds.), Mental models (pp. 299–324). L. Erlbaum Associates.

145

McLaughlin, P. (2001). What functions explain: functional explanation and self-reproducing

systems. Cambridge University Press.

McPherson, T. (2012). Ethical non-naturalism and the metaphysics of supervenience. Oxford studies in metaethics, 7, 205-234.

———. (2015). What is at stake in debates among normative realists? Noûs, 49(1), 123-146.

Millgram, E. (2009). Life and action. Analysis, 69(3), 557–564.

Millikan, R. G. (1989). In defense of proper functions. Philosophy of Science, 56(2), 288-302.

Millum, J. (2006). Natural goodness and natural evil. Ratio 19(2), 199-213.

Monod, J. (1971). Chance and necessity: an essay on the natural philosophy of modern biology. (A. Wainhouse, Trans.). New York: Knopf.

Moravcsik, J. (1994). Essences, powers, and generic propositions. In: T. Scaltsas, D. Charles, & M. L. Gill (Eds.), Unity, identity, and explanation in Aristotle's metaphysics (pp. 229-244). Oxford University Press.

Moreno, A., and Mossio, M. (2015). Biological autonomy. Springer.

Mossio, M., Saborido, C., & Moreno, A. (2009). An organizational account of biological functions. The British Journal for the Philosophy of Science, 60(4), 813-841.

Nagel, E. (1961). The structure of science: problems in the logic of scientific explanation. Harcourt,

Brace & World.

———. (1977). Teleology revisited. Journal of Philosophy, 84, 261-301.

Neander, K. (1991). Functions as selected effects: the conceptual analyst’s defense. Philosophy of Science, 58, 168–84.

Nersessian, N. & Resnick, L. (1989). Comparing historical and intuitive explanations of motion: Does naive physics have a structure. In Proceedings of the 11th Annual Conference Cognitive Science Society. Psychology Press.

146

Nicholson, D. (2013). Organisms ≠ machines. Studies in the History and Philosophy of Biological

and Biomedical Sciences 44, 669–678.

———. (2014). The return of the organism as a fundamental explanatory concept in biology. Philosophy Compass 9(5), 347–359.

———. (2018). Reconceptualizing the organism: from complex machine to flowing stream. In: J. Dupre & D. Nicholson (Eds) A manifesto for a processual philosophy of biology (pp. 139- 166). Oxford University Press.

Nunes-Neto, N., Moreno, A., & El-Hani, C. (2014). Function in ecology: an organizational approach. Biology & Philosophy 29(1), 123-141.

Odenbaugh, J. (2017). Nothing in ethics makes sense except in the light of evolution? Natural goodness, normativity, and naturalism. Synthese, 194(4), 1031-1055.

Odling-Smee, J., K. Laland, and M. Feldman. (2003). Niche construction: the neglected process in evolution. Princeton: Princeton University Press.

Oyama, S. (2000). The ontogeny of information. Durham, N.C.: Duke University Press.

Oyama, S., Griffiths, P. E., & Gray, R. D. (Eds.). (2001). Cycles of contingency: developmental systems and evolution. Cambridge, Mass: MIT Press.

Pigden, C. (1991). Naturalism. In: P. Singer (Ed.) A companion to ethics (pp. 421-431). Blackwell Publishers.

Pigliucci, M., & Müller, G. B. (Eds.). (2010). Evolution, the extended synthesis. Cambridge, MA: MIT Press.

Railton, P. (1986). Moral realism. The Philosophical Review, 95(2), 163–207.

Roe, K., & Murphy, D. (2011). Function, dysfunction and adaptation? In: P. Adriaens, A. De Block (Ed.) Maladapting minds: philosophy, psychiatry, and evolutionary theory (pp. 216– 237). Oxford University Press.

147

Rosen, R. (1991). Life itself: a comprehensive inquiry into the nature, origin, and fabrication of

life. Columbia University Press.

Rosenblueth, A., Wiener, N., & Bigelow, J. (1943). Behavior, purpose and teleology. Philosophy of science, 10(1), 18-24.

Saborido, C., & Moreno, A., (2015). Biological pathology from an organizational perspective. Journal of Theoretical Medicine and Bioethics, 36(1), 86–95.

Schlosser, G. (1998). Self-re-production and functionality: a systems-theoretical approach to teleological explanation. Synthese, 116, 303–54.

Scholl, R., & Pigliucci, M. (2015). The proximate–ultimate distinction and evolutionary developmental biology: causal irrelevance versus explanatory abstraction. Biology & Philosophy, 30(5), pp. 653-670.

Shafer-Landau, R. (2003). Moral realism: a defence. Oxford University Press on Demand.

Shevah, O., & Laron, Z. (2007). Patients with congenital deficiency of IGF-I seem protected from the development of malignancies: a preliminary report. Growth Hormone & IGF Research, 17(1), 54-57.

Sober, E. (2000) Philosophy of biology, 2nd Ed. Westview Press.

Sommerhoff, G. (1950). Analytic biology. Cambridge: Cambridge University Press.

Sterelny, K., & Kitcher, P. (1988). The return of the gene. The Journal of Philosophy, 85(7), 339– 361.

Sturgeon, N. (1985). Moral explanations. In: D. Copp & D. Zimmerman (Eds.) Morality, reason, and truth (pp. 49-78). Rowman and Allanheld.

Swain, J., & French, S. (2000). Towards an affirmation model of disability. Disability & society, 15(4), 569-582.

Thompson, M. (2004). Apprehending human form. In: A. O’Hear (Ed.), Modern moral philosophy (pp. 47-74) Cambridge: Cambridge University Press.

148

———. (2008). Life and action: elementary structures of practice and practical thought. Harvard

University Press.

———. (2013). Forms of nature. In: G. Hindrichs & A. Honneth (Eds.) Freiheit: Stuttgarter Hegel- Kongress 2011 (pp. 701-735). Frankfurt am Main: Vittorio Klostermann.

Thornhill, R., & Palmer, C. T. (2000). A natural history of rape: biological bases of sexual coercion. Cambridge, Mass.: MIT Press.

Travis, C. B. (Ed.). (2003). Evolution, gender, and rape. Cambridge, Mass.: MIT Press.

Varela, F., Maturana, H., & Uribe, R. (1974). Autopoiesis: the organization of living systems, its characterization and a model. Biosystems, 5, 187–196.

Waddington, C. H. (1953). Genetic assimilation of an acquired character. Evolution, 118-126.

Walsh, D. (1996). Fitness and function. The British Journal for the Philosophy of Science, 47(4), 553-574.

———. (2012). Mechanism and purpose: a case for natural teleology. Studies in History and Philosophy of Biological and Biomedical Sciences, 43, 173–181.

———. (2013). Mechanism, emergence, and miscibility: the autonomy of evo-devo. In: P. Huneman (Ed.) Functions: Selection and Mechanisms (pp. 43-65). Dordrecht: Springer.

———. (2015) Organisms, agency, and evolution. Cambridge University Press.

Walsh, D., Ariew, A. (1996). A taxonomy of functions. Canadian Journal of Philosophy, 26(4), 493–514.

Weber, M. (2005). Philosophy of experimental biology. Cambridge University Press.

Weismann, A. (1893). The germ-plasm: a theory of heredity. C. Scribner's sons.

West-Eberhard, M. J. (2003). Developmental plasticity and evolution. New York: Oxford University Press.

———. (2005). Developmental plasticity and the origin of species differences. Proceedings of the National Academy of Sciences, 102 (suppl. 1), 6543-6549.

149

Williams, G. C. (1966). Adaptation and natural selection. Princeton, N.J: Princeton University

Press.

Wilson, D. S., & Dugatkin, L. A. (1997). Group selection and assortative interactions. The American Naturalist, 149(2), 336–351.

Woodcock, S. (2006). Philippa Foot’s virtue ethics has an Achilles’ heel. Dialogue, 45(3), 445– 468.

Woodward, J. (2001). Law and explanation in biology: invariance is the kind of stability that matters. Philosophy of Science, 68(1), 1-20.

———. (2005). Making things happen: a theory of causal explanation. Oxford university press.

Wright, L. (1973). Functions. The Philosophical Review, 82(2), 139-168.

150