Nietzschean Perspectivism: Representation and Values

ABSTRACT

Perspectivism claims that something about the cognitive situation of the knower shapes and limits the content of cognition. In some recent work (Anderson 2023), I proposed a representationalist interpretation of this limitation. Following A.W. Moore in Points of View (1997), I emphasized the notion of perspectival representation, as contrasted against absolute representation. Perspectival representations can be assessed for truth and content only together with some specification of the relevant perspective. For example, ‘San Francisco is 39 miles away’ was true when I was writing this paper in Mountain View, but it was obviously false to say while I was giving a talk on this material in Riverside, 434 miles from San Francisco.¹ For such perspectival representations, truth and content vary across cognitive situations. Absolute representations, by contrast, are not perspectival; they purport to capture the world as it is independently of any perspective, and they exhibit no such variation across situations. Perspectivism is the claim that all our cognition is perspectival, and that absolute representations are impossible for us. It is thus a specific (and controversial) way of cashing out the insight that we are bounded rational agents, whose cognition is essentially shaped by our finite condition.

However illuminating it may be in other ways, this focus on the perspectival character of representations leaves unaddressed one of the most distinctive features of Nietzsche’s own treatment of perspectivism—his widespread claim that our perspectives are somehow rooted in our values, affects, and drives. My goal in the present paper is to develop considerations that begin from the representationalist account of perspectivism, but can help to motivate Nietzsche’s emphasis on the grounding of perspectives in the practical and conative side of our psychology. I hope the argument can raise doubts about the commonly held thesis that our theoretical cognition enjoys purely values-free objective standing. I do not claim, however, that the arguments I offer here suffice to establish Nietzsche’s position, or to definitively rule out the alternatives to it that I will consider.

I will start (section 1) by laying out the representationalist account of perspectives and rehearsing some challenges against Moore’s defense of absolute representations that I mounted in the earlier paper. Since Moore’s account is inspired by the “absolute conception of the world” advanced by Bernard Williams in Ethics and the Limits of Philosophy (1985), I then turn (section 2) to Williams’s distinction between internal and external practical reasons, which turns out to bear significant analogy to the distinction between perspectival and absolute representations. Some of Williams’s own reasons for skepticism about the possibility of genuinely external reasons, when viewed from the standpoint of the analogy, suggest that we might expect to find practical-side considerations (like values, or drives and affects) playing a key role shaping our perspectival representations. I will propose that blocking this plausible expectation would depend on a strict, principled (Kant-style) distinction between theoretical and practical reasoning that is foreign to the spirit of both Williams and Nietzsche. Finally (sections 3 and 4), I will consider some Nietzschean texts that articulate his commitment to an evaluative basis for perspectives and illuminate his reasons for resisting any such strict practical/theoretical distinction. Together with some suggestive examples of the influence of evaluative commitments in cognitive life, inspired by Helen Longino’s account of the importance of a diversity of perspectives for the objectivity of scientific communities, these Nietzschean considerations help to motivate his distinctive commitment to the evaluatively-loaded character of perspectives. In conclusion, I will connect that conception of perspectives to Nietzsche’s genealogical approach to the critique of cognition, and also to the underlying motivations for his suspicion that certain “purified” conceptions of theoretical cognition amount to expressions of the ascetic ideal.

1. Perspectival Representations and Moore’s Leibnizian Argument for Absolute Representations

As the example about the distance to San Francisco shows, perspectival representations are routine and commonplace. Their perspectival character, moreover, poses no impediment whatsoever to their being true, counting as knowledge, and so forth, in just the ordinary ways.² Many are perspectival in an entirely literal sense. Visual perceptions, for example, represent bodies from a specific point of view, and the bodies’ apparent shapes, colors, distances, and relations are all affected by their spatial relation to the viewer and the optics of the visual system. Implicitly indexical truths like the one about San Francisco likewise depend on contextual specification in obvious ways. The role of the system of spatial locations in such examples recalls to mind something similarly perspectival about the representation of motion central to modern physics. Different apparent motions can seem to conflict—for example, the sun appears to us to orbit the earth, but from the point of view of the sun, it is the earth that orbits. Such conflicts are merely apparent; they can be smoothly resolved by recourse to reference frames, which trace each relative motion to its proper perspective. Newton’s achievement was to show how identifying the center of gravity for a system establishes a privileged frame of reference. Relative to that frame, we can fix the true motions within the system, and then show how the various relative motions are related to one another and to the true motions, as well as why the relative motions appear as they do in their local reference frames, whose systematic relations to one another and to the privileged frame are thereby determined. In all these cases, the initial representations are perspectival, but locating them within the background spatial framework allows us to abstract away from that perspectival character.

Nietzsche’s appeal to perspective trades on these literal examples, but extends them to make a broader, more metaphorical appeal to cognitive “perspective,” which is intended to be harder to abstract away from. On this view, the content of our representations is systematically affected by facts about our “situations,” now understood not just in terms of possible spatial locations, but in an extended sense comprising a host of factors that shape cognitive contents (our perceptual system, our particular conceptual tools, the structure of our language, our evaluative outlooks, and so on). But as I already hinted, the literal cases from which we started threaten to undermine the intended force of the metaphor. After all, the fact that the system of spatial relations or reference frames can provide a clear set of standpoints, relative to which we assess the truth of any representations tied to them, entails that the contents of spatially perspectival representations could be systematically restated in terms of underlying non-perspectival facts about how things are in the spatial array itself—subtracting the influence of perspective, and with it the implicit relativity of the original representations. This would trivialize the idea of cognitive perspective.

Comparison to the perspective system of Leibniz’s monads sharpens the point. Recall, monads are differentiated by their internal representations, which differ due to their viewpoints on the world. But this in no way entails that these various perspectival representations are incompatible. Instead, they all mesh seamlessly as reflections of the same reality, each adjusted for the effects of its point of view. It is not that the monadic perspectives fully capture how things are.³ But still, the laws of perspectival “refraction” remain systematic, making it possible to “translate” between monadic perspectives in ways that adjust for point of view. That is the sense in which all the representations “harmonize” with each other.⁴ The situation is analogous to the case of relative motions. There is a systematic background theory that represents underlying reality in absolute terms and simultaneously explains why the contents of perspectival representations are as they are. The effects of perspective on cognitive content are therefore superficial and eliminable, rather than basic and inescapable. Leibnizian perspectives, unlike Nietzschean ones, do not entail any limitation on cognitive representation that would compromise a metaphysics capturing things as they are, absolutely. On the contrary, they find their home in just such a theory.

Moore (1997) exploits fundamentally similar ideas to construct an argument for the possibility of absolute representations. The argument proposes to explain the content of perspectival representations in terms of underlying facts about the cognitive situations that are the source of perspectival effects, similar to the way spatially perspectival contents were explained by appeal to the underlying objective spatial array. Moore concedes that this task will not be as simple in general as it is in the spatial case, because not all sources of perspectival influence form unified systems by which cognitive agents are uniformly affected. For example, some representations are formulated using sophisticated concepts that reflect specific cultural achievements—for example, concepts native to specific scientific frameworks, or to particular social or cultural formations, or thick ethical concepts à la Williams, or what have you.⁵ It may be impossible or unintelligible to apply such concepts or recognize the truths they capture without inhabiting the culturally specific perspective within which they make sense. In that sense, they are inherently perspectival. Moore not only concedes the possibility of such essentially perspectival representations, but is willing to begin his argument from them.

He nevertheless insists that we can recapture in absolute representations whatever is important in such essentially perspectival truths. His strategy is to view the perspectival in light of a “Basic Assumption” about the nature of representation, namely, “that representations are representations of what is there anyway” (Moore 1997, 74). That is, the content of representations is truth conditional; they represent the world as being some way, and their content is fixed by the worldly conditions under which they are true (where the worldly conditions are what is “there anyway,” regardless of how we represent things). In light of that assumption, Moore insists that we can always retreat from a perspectival representation to some wider point of view, from which we can explain not just how the content of that representation relates to the world, but also how its operative perspective works, such that within that perspective the representation counts as true.⁶ Such an explanation will “transcend” the perspective of the initial representation; it does not inhabit the perspective, but describes it and situates its effects in relation to what makes the representation true.

Of course, that explanation may itself also be perspectival, and there is no reason to think that the series of such explanatory standpoints must terminate. But here, Moore deploys a stronger version of his Basic Assumption to infer to the possibility of absolute representation. In Moore’s mind, the idea of the world’s being “there anyway” involves its having the “unity, substantiality, and autonomy of reality,” where the relevant kind of unity entails that “it ought to be possible to give a single account, equipped to mesh with a similar account for any other possible truth, that reveals how” the relevant representations can all be true together, and thereby answer to the world (Moore 1997, 72, my ital.). This version of the thought is supposed to ensure that, for any group of true perspectival representations, there will be one single account that explains how the world makes them true, including in conjunction with true representations from any incompatible perspectives. That account would transcend all the various incompatible perspectives, and would thus be independent from any of them. So it must be absolute. Therefore, absolute representations are possible.

Space does not permit reconstruction of this argument in detail,⁷ but it should be clear that Moore’s Basic Assumption is seriously strong. It involves not merely the commonsensical realist thought that the world is “there anyway,” but a picture of the world as fully determinate and unified, such that its unified character is capable of explaining the truth of whatever ways of representing it we come up with, no matter how diverse and perspectivally limited those ways are—and doing so not just serially, but all together. (The last point rests on the strong unity.) Nietzsche would flatly reject this assumption that the world is unified and determinate.

In the end, therefore, we are left at something of a dialectical impasse between Moore and Nietzsche. Juxtaposing their views is nevertheless illuminating. First, it reveals that Nietzschean perspectivism relies on his denial that we live in a unified determinate world—a denial manifest in passages such as this:

The astral order in which we live is an exception; this order and the relative duration that depends on it have again made possible an exception of exceptions: the formation of the organic. The total character of the world, however, is in all eternity chaos—in the sense not of a lack of necessity but of a lack of order, arrangement, form, beauty, wisdom, and whatever other names there are for our aesthetic anthropomorphisms. (GS 109)

It turns out, then, that pluralist metaphysical intuitions are not separate, detachable ideas of Nietzsche’s, but are essentially connected to his perspectivist outlook. Without that commitment, his view would be vulnerable to the Leibniz/Moore argument for absolute representations, and the influence of perspective would remain superficial and eliminable, in the ways we saw above. I have not offered any independent argument for pluralism, but given the strength of Moore’s Basic Assumption, a Nietzschean perspectivist should feel no pressure to concede it as a premise. And Moore himself has no further argument for the premise. His argument does just assume it. That observation leads to a second point. Moore’s argument is very carefully constructed. Its presupposition of the strong unity of the world thereby reveals how much needs to be assumed in order to rule out Nietzschean perspectivism via the fundamental, broadly Leibnizian strategy of overcoming perspectival influence by tracing it back to, and locating it within, the system of cognitive situations on which the various perspectives depend.

I mentioned that Moore’s defense of absolute representations was inspired by Bernard Williams’s idea of the “absolute conception of the world” (Williams 1985, 132–55). Williams traces that conception to the emergence of the scientific world picture in the European seventeenth century. In the context of Ethics and the Limits of Philosophy, its role is to fund a contrast that pits such an entirely objective, value-free world conception, in light of which we can expect our various efforts to represent the world to converge, against our ethical practices, where (Williams argues) we have no reason to expect any such convergence. But in other contexts, Williams develops views that would seem to side with Nietzsche, and against Moore. Williams’s late work develops genealogical arguments that, for example, trace key components of our notion of truthfulness to the social value and social functions of sincerity and authenticity (see Williams 2002), and he also argues that our “purified” notions of guilt and blame grew out of potentially healthier, shame-connected ancient ideas under (pernicious) influence from (the perspective of) the “morality system” (see Williams 1993a, 1993b). These arguments treat philosophically central concepts as dependent on contingent cultural perspectives, and they work by activating the same kinds of genealogical considerations that Nietzsche employs to trace philosophical commitments to perspectival origins and to insist that they depend on our values and conative attitudes. From the point of view of the dialectical standoff I just sketched, it is surprising—or at least a bit curious—that Williams’s thought seems to enter the ranks on both sides of the main question. How can that be?

2. Williams on Internal Reasons and Practical versus Theoretical Reasoning

In this section, I will suggest that answering that how-question proves illuminating, because it suggests considerations that help motivate Nietzsche’s conviction that representations inherit perspectival limitations from a dependence on values. Before we reach that issue, however, it is helpful to consider the how-question from Williams’s own standpoint, rather than Moore’s or Nietzsche’s.

Our discussion so far identified two opposed streams of thought: one insists that all representations are perspectival, and that genealogies can trace their perspectival content to evaluative roots; the other insists, with Moore, that absolute representations are possible, and that we can arrive at them by isolating perspectives and explaining their effects so as to restate what makes perspectival claims true in independent terms. For Williams, though, these two strands were not only compatible, but mutually reinforcing. Each flows from one horn of a basic distinction that supports a core thesis of Ethics and the Limits of Philosophy⁸—the distinction between, on the one hand, scientific knowledge, which aims to represent things according to an absolute conception of the world and to resolve disagreements by convergence of theory on how things are, and on the other, ethical thought, which is irreducibly perspectival and cannot resolve disagreements by reference to how things are, objectively.⁹ Whereas scientific convergence is a matter of approximation to the absolute conception, ethical thought can be truth-tracking at all only when carried out in terms of “thick concepts” (like <honor> or <courage>) that are irreducibly culturally specific and thus inherently perspectival. We can of course ask philosophical questions about which thick ethical concepts we ought to use, but according to Williams, such efforts will never yield a definitive Ethical Theory.¹⁰ Retreating to those more general questions about our practice does not take us into a truth-tracking inquiry about how things anyway are, but instead to a reflective—but still essentially practical—inquiry about how best to navigate the social world (including the question of what social world we should best make, collectively). Thus, Williams seems to be an absolutist with one hand and a perspectivist with the other because he thinks there are two, basically different kinds of question, one of which puts us on the path toward an absolute conception, while the other is inherently perspectival.

Adherence to this distinction is one key point at which Williams remained resolutely un-Nietzschean, despite the many lessons he drew from Nietzsche for his work on the essentially perspectival side of the ledger. Nietzsche, by contrast, denies any sharp distinction between two foundationally different types of question here. He cast a corrosive, all-consuming skepticism over all philosophical (or, more generally, theoretical) claims to operate from pure cognitive stance, utterly innocent of influence from the evaluative and affective life of the thinker: “Gradually it has become clear to me what every great philosophy so far has been: namely, the personal confession of its author and a kind of involuntary and unconscious memoir; also that the moral (or immoral) intentions in every philosophy constituted the real germ of life from which the whole plant had grown” (BGE 6).¹¹ I will return to the question what kind of argument Williams would need to block this sort of skepticism.

But first, I want to observe that when it comes to our specifically practical reasoning, Williams, too, remains skeptical of taking any purely cognitive, objective stance, cleanly separable from all influence of conative attitudes. That skepticism is encoded in his famous distinction between internal and external reasons, which he deploys in order to raise doubt that we can make any tolerably clear sense of the notion of a strictly external practical reason. As a reminder, internal reasons for an action have some basis in an aim or motive of the agent’s that would be served by the action, whereas external reasons hold regardless of any such basis in the agent’s “subjective motivational set” (= S), and thus in virtue of something about the world itself that stands over against the agent, which could be apprehended by rational deliberation (by anyone) carried out independently from any influence or guidance by the agent’s motivations. Thus, the possibility of external reasons turns on whether practical reason can cognize something objective about the world that holds as a reason, binding in some way independent of anything in S or inferable from it. There is a clear analogy here with Moore’s account of absolute representations, which turns on the possibility of cognition’s capturing something objective about the world “that is there anyway,” in independence from the perspectival concepts that express our cognitive situation.

Williams’s conclusion in “Internal and External Reasons” is that “external reason statements, when definitely isolated as such, are false, or incoherent, or really something else misleadingly expressed” (Williams 1981 [1980], 111). It is important to recall how much of the paper’s argument proceeds simply by making our conception of an internal reason statement more realistic, and then through a burden-shifting gambit that displaces theoretical pressure onto the character Williams calls the “‘external reasons theorist’” (Williams 1981 [1980], 108). The opening moves insist on the inadequacy of a “sub-Humean” model of internal reasons, which holds narrowly that “internal reasons” would have to arise more or less directly, without intervening deliberation, from the elements of S. Williams insists, more realistically, that we are still making internal reasons statements even when the reason we describe is based on what the agent would be motivated to do if her false beliefs were corrected, or if her S itself were corrected by various processes of practical deliberation, broadly understood to include all kinds of things like deliberation that leads her to accept new elements into S, persuasive conversation with other people, imagination that puts the situation in a new light, and so on. On this picture, a person’s motivations are not some fixed bedrock cut off from her own reflective intervention; on the contrary, using deliberation to refine or revise our motivations is one of the most common and obvious ways we decide what to do, and when we act on the resulting motivations, our reasons are still internal, not external. A genuinely external reason, therefore, would be defined by opposition to all this—as a reason the agent has even though there is nothing in her S to support it and no “sound deliberative route” to it from that S (Williams 1995 [1989], 35).¹²

With this conception of an external reason in place, Williams executes the burden shift: an external reason statement captures a reason that would emerge from a “purely rational process” of correct reflection about the objective situation, proceeding in isolation from the agent’s particular S. Therefore, Williams insists, the traditional Humean internalist should not pick up the “onus of proof about what is to count as a ‘purely rational process’ . . . which properly belongs with the critic who wants to oppose Hume’s general conclusion and to make a lot out of external reason statements” (Williams 1981 [1980], 108). After all, it is the external reasons theorist, not the Humean, who needs the notion of “purity” here, as a way to capture the isolation from the agent’s S that allows the reason to count as external in the first place.

This appeal to purity of the rational process is what funds the deep analogy between external reasons, on the side of practical reasoning, and absolute representation, on the cognitive side. In both cases, “purity” is understood as purity from influences arising from the agent’s particular situation. Just as perspectival representations are defined by their dependence on the knower’s cognitive situation, so likewise internal reasons are defined by their connection, via some “sound deliberative route,” to the agent’s motivational situation, defined by S. The perspective or the motivational set thereby defines an “internal” domain in light of which the representation has sense and relevance for the agent. Absolute representations and external reasons, in their turn, are defined by contrast, as those that make their claims on us regardless of our situation. Their “purity” is a matter of making a “purely rational” claim, in the sense that it is radically independent from the more particular, “internal” standpoint of our cognitive or practical agency. External reasons are supposed to be responsive to the considerations that are just “there anyway,” regardless of any motivational perspective, in the same way absolute representations were responsive to the world that is “there anyway” and transcends any cognitive perspective.

What gives Williams’s basic burden-shifting move in “Internal and External Reasons” its force is the very capaciousness of the deliberative moves he is willing to entertain as “sound” in going from a given S to what the agent may legitimately count, in the end, as an internal reason. That is what underlines the strictness of the “purity” just described. Here is a key passage:¹³

But here it may be objected that the account of deliberation is very vague, and has for instance allowed the use of the imagination to extend or restrict the contents of the agent’s S. But if that is so, then it is unclear what the limits are to what an agent may arrive at by rational deliberation from his existing S.

It is unclear, and I regard it as a basically desirable feature of a theory of practical reasoning that it should preserve and account for that unclarity. There is an essential indeterminacy in what can be counted as a rational deliberative process. Practical reasoning is a heuristic process, and an imaginative one, and there are no fixed boundaries on the continuum from rational thought to inspiration and conversion. To someone who thinks that reasons for action are basically to be understood in terms of the internal reasons model, this is not a difficulty . . . . It merely shows that there is a wider range of states, and a less determinate one, than one might have supposed, which can be counted as A’s having a reason to φ. (Williams 1981 [1980], 110)

In light of the analogy between internal reasons and perspectival representations, however, this point about the indeterminacy of good reasoning looks to have implications beyond what Williams envisions in the passage. In particular, the Nietzschean perspectivist should say exactly parallel, catholic-minded things about what counts as good theoretical reasoning within the context of a perspective that shapes one’s cognitive representations. If we accept such a parallel between theoretical and practical reasoning, that would help explain why Nietzsche is willing to countenance a productive role for noncognitive values and conative states in structuring our perspectival representations. Williams concedes these points for the special case of practical reasoning, but (by contrast to Nietzsche) takes theoretical reasoning to be beholden to the absolute conception of the world. What is awkward for Williams’s stance is that such a strict boundary between practical and theoretical reasoning is itself an instance of the same demand for cognitive purity of which he is (with Nietzsche) so skeptical when it comes to internal reasons for action.

If Williams were an orthodox Kantian, the resources needed to construct a defense for such a stance would be ready to hand. For a Kantian insists that there is a principled, logico-transcendental distinction, grounded in the nature of reason itself, between the principles of theoretical reason and those of practical reason. This distinction is supposed to be an a priori precondition of the possibility of empirical knowledge, including any knowledge (e.g., from genealogies) about the sources and methods we use to arrive at our judgments. But of course, the Kantian account does not come for free or go without saying.¹⁴ The transcendental arguments that fund it are deeply controversial, and tellingly for the present context, entirely rejected by Williams. In Ethics and the Limits of Philosophy, he rests the separation of scientific from ethical thought on considerations of an entirely different kind—contingent, multiply interpretable considerations like the emergence of a tradition of purely objective theorizing together with an absolute conception of the world in seventeenth-century Europe, or the hoped-for convergence of scientific theorizing compared to the plausibility of cultural specificity in ethical thought and practice. These are not even the sort of reasons that could establish the strict, logical and principled boundary Kantians defend. Indeed, the very historical sophistication exemplified by Williams’s specificity in tracing the absolute conception to seventeenth century Europe tends to reinforce, rather than silence, worries about the cultural peculiarity of the key ideas.

3. Thick Concepts, Values, and Perspectives

Naturally, Williams is not without recourse in the face of such doubts. The most obvious move to shore up the wanted distinction between science and ethics would be to point back to the key role Williams identified for “thick ethical concepts” in securing such objectivity as we can attain in ethical thought, and to insist that such concepts have no parallel role in science. What is distinctive about thick concepts is the way they combine descriptive content (which underwrites objectivity and truth claims for the judgments involving them) with evaluative content that gives those judgments normative force. The facts of Socrates’ conduct at trial, for example, meet the truth conditions for his being courageous and honest, and thereby fund objective judgments that he was. The truth of those judgments can exert some rational pressure on his critics, even if they would prefer to assess him using different thick concepts. (Perhaps they saw the same conduct to be rash, offensive, and impious.) Williams’s thought, then, might be that scientific concepts do not have this mixed character combining descriptive and evaluative content, and so are suitable for being purified and put into the service of approximating the absolute conception in a way that the concepts we need to manage ethical life are not.

Nietzsche, however, would be highly dubious that scientific (or other theoretical) concepts actually do exhibit this kind of purity, and a good deal of work in the recent philosophy of science lends support to his stance. For example, Helen Longino (1990, 2002) has taught us to see scientific objectivity itself not as an individual level property achieved through the elimination of bias and evaluative commitments by each researcher, one by one, but instead as a community-level property that emerges from a healthy, well-functioning research community, within which a diverse multiplicity of perspectives is actively maintained, and all claims and presuppositions of the scientific enterprise are subject to legitimate criticism from those points of view. Although Longino deserves the credit for articulating the bases for this view, almost all of us are committed to it in practice in our day-to-day life as researchers. Just recall, for example, the role of Longino-style considerations in funding the standard argument we rely upon to claim that diversity is epistemically valuable to the university’s mission of knowledge production and transmission.¹⁵ The fundamental bases on which Longino grounds her view, meanwhile, are deeply Nietzschean. The core argument is, first, that theoretical virtues like simplicity, plausibility, and even empirical adequacy, or the like, cannot overcome the underdetermination of theory by evidence, and thereby guide theory choice, without some concrete specification of what is to count as simplicity (etc.) in the relevant context, and then, second, such specifications are essentially never free of social value commitments (Longino 1990). So just as Nietzsche is keen to suggest, disciplined theoretical belief formation in real life practice involves and presupposes the activation of (perspectival) values.

One might be tempted to doubt that such heuristic norms of science—or at least, the virtue of empirical adequacy, in particular—could really always depend on social values. Relatedly, many may doubt that science involves the kind of thick concepts Williams characterized. Such doubts can arise from quite commonsensical views about the autonomy of cognitive norms. I have space here only for an outline suggestion of how a Nietzschean argument against such doubts might go. Fuller articulation of the argument pattern will be the subject of future work. One interesting strategy would be this: 1) first, there are some clear cases in which uncontroversially cognitive scientific judgments (or concepts) are shaped by social values; then 2) Nietzsche would insist (more controversially) that certain key concepts central to even highly abstract sciences are likewise value-laden, in ways that are more covert but would nevertheless be revealed by genealogical inquiry pressed with sufficient determination and suspicion; and then 3) the best explanation of the observations collected under 1) and 2) is the assumption of a basic continuity between the two kinds of cases, grounded on a general underlying penetration of evaluative considerations into cognition. Nietzsche’s claim would then be that such evaluative influence is mediated through the perspectives that shape cognition.

To indicate the potential of this argument pattern in a preliminary way, I will consider one relatively clear case of social values’ influencing cognition. I will then briefly point to some Nietzschean texts that establish his own commitment to the more controversial second step. To be clear, I choose my initial case not because I think it is typical of all science, but exactly because it is an obvious and undeniable case of evaluative penetration, where apparently value-free, directly empirical concepts turn out to be impossible to insulate from deeply evaluative commitments. Sustained examination of Nietzsche’s more controversial claims that similar penetration is present even in the far reaches of physics will have to await another occasion.

In a recent PLOS ONE study that received significant press, Anderson, et al. (2023) reviewed the ethnographic literature on women’s contributions to hunting in so-called “hunter/gatherer” social groups. Their review showed that the data support widespread contributions by women to hunting activity across the large majority of subsistence societies considered, and across most game types, including big game. In societies where hunting was the most important subsistence activity, women participated 100% of the time. Because this survey explodes a very widespread assumption in the literature, the paper was titled “The Myth of Man the Hunter.”

What is striking, however, is that most of the recent scientific pressure against the “man the hunter” picture had come not from ethnographic studies of contemporary populations, but instead from archaeology, where many studies identified irresistible evidence of women’s engagement in hunting and war in prehistoric societies, even though those archaeological studies typically acknowledged (rashly!) that among recent hunter-gatherers, these activities do show a marked gendered division of labor. A paper by Haas, et al. (2020) offers a telling case. The point of the paper is to document widespread evidence of female hunting from American archaeological sites, even though (per the first sentence of the paper) “Big-game hunting is an overwhelmingly male-biased behavior among recent hunter-gatherer societies” (Haas, et al. 2020, citing Binford [1971] and Kelly [2013] for the point). Except it just isn’t, as Anderson, et al. (2023) show through their survey. The best part is a telling data point from Haas, et al. (2020). They note that their own study had to overcome and displace a long history of misinterpretation of the archaeological data themselves: for example (and it is one example among many), “Concerning the Paleoindian Gordon Creek burial, Breternitz, et al. (1971) grappled, ‘Since the burial has been determined to be a female, the inclusion of a projectile point preform has been difficult to explain. However, if the artifact had been used as a knife or scraper, typically women’s tools, then its inclusion with the burial is a more consistent association’” (quoted in Haas, et al. [2020, 1]). That is, Breternitz and colleagues were looking directly at the tool, which was clearly a projectile point (for hunting), but since it was buried with a woman, they classified it as a tool for food preparation.

For my purposes, what is important is that core observational concepts, like <kitchen tool> and <hunting tool>, turn out to be deeply bound up with evaluative preconceptions about the appropriate behavior of people with different genders—so much so that Breternitz and colleagues cannot even see the tool in front of them, “since the burial has been determined to be a female.” That is to say, even these observational concepts approximate “thick concepts” in the Williams-style sense. They therefore provide a route through which perspectival social values affect our cognition. As Longino has taught us, eliminating such social values is a fool’s errand; it will only result in their being suppressed and therefore operating uncontrolled. The contrary strategy is required. To achieve scientific objectivity, we must own our reliance on social values, and control them through the careful deployment of multiple perspectives in the scientific community.¹⁶ This should sound familiar to Nietzsche scholars (compare GM III 12).

Nietzsche himself controversially—perhaps one could even say, “notoriously”—holds that this sort of evaluative penetration is not limited to sciences like anthropology, where valuing creatures are the objects of study. As we saw, he insists that it extends into the most abstract and avowedly “disinterested” reaches of philosophy (BGE 6, quoted above), and even highly mathematized sciences like physics are not immune. The clearest locus is BGE 22, where Nietzsche is concerned to interrogate the notion of natural law in physics:

Forgive me as an old philologist who cannot desist from the malice of putting his finger on bad modes of interpretation: but “nature’s conformity to law,” of which you physicists talk so proudly, as though—why, it exists only owing to your interpretation and bad “philology.” It is no matter of fact, no “text,” but rather only a naïvely humanitarian emendation and perversion of meaning, with which you make abundant concessions to the democratic instincts of the modern soul! “Everywhere equality before the law; nature is no different in that respect, no better off than we are”—a fine instance of ulterior motivation, in which the plebeian antagonism to everything privileged and autocratic as well as a second and more refined atheism are disguised once more. “Ni Dieu, ni maître”—that is what you, too, want, and therefore “cheers for the law of nature!”—is it not so? (BGE 22)

Nietzsche’s idea that the concept of natural law is actually a product of underlying democratic values is remarkable, and I will make no attempt to defend it or render it more plausible here. What matters for immediate purposes is the evidence this claim provides that Nietzsche in fact contemplates something like the strategy of argument I sketched above. Of course, as the passage moves on Nietzsche is keen to propose his own “will to power”-based alternative to the core concept of natural law, but (consistently) he is very far from claiming that his alternative would be “pure,” or value-free: “Supposing that this also is only interpretation?—and you will be eager enough to make this objection—well, so much the better” (BGE 22). It follows that the allegation that the natural-law concept expresses democratic values is not supposed to count as an objection because that would compromise the value-free ideal of science. Instead, the intended objection rests on which values the concept embodies. (See also BGE 14, along similar lines.) The Nietzschean position, then, must be that evaluative penetration into cognition is inevitable, because (even though they do not at first seem to be) scientific concepts turn out to be “thick concepts” that combine descriptive and evaluative content in a densely entangled whole.

At this point, the Nietzschean can confront Williams with the following line of thought. If we start the argument from within the side of practical reason, and if we reject any a priori Kantian distinction between practical and theoretical reasoning as logically different in kind, then we should expect the same considerations inclining to skepticism about a special class of external (practical) reasons to extend outward from specifically practical reasons to include all our reasons. Given the indeterminacy of the kinds of reasoning that provide “sound deliberative routes,” it would seem to be impossible to draw any principled boundary isolating just the practical reasons as those for which the ideal of cognitive purity, and consequently a meaningful notion of external reasons, fails. This raises quite general doubt about the purity of cognition and its reasons from all conation, and thence also about the absolute conception and absolute representations. After all, theory choice is always, in the last analysis, choice, and if we attend to the ways theoreticians operate in practice, it is hard to maintain seriously that they are not normally invested (in the full evaluative and affective sense) in the positions for which they argue. And again, as Williams (1981 [1980], 110) himself insists, our (practical) reasoning is very capacious indeed. It is, in fact, “essentially indeterminate” in what kinds of moves it licenses; all kinds of thinking, including imagining, “feeling one’s way into a situation,” and the rest, may be included in the methods by which a sound deliberative route might expand the membership of S, and thereby open up a reasoned connection to the relevant choice.¹⁷ Once we take that point on board, our reasoning’s capaciousness seems capable of explaining quite a bit about our choices, including theoretical choices. From this standpoint, it is unpersuasive to hold it out as obvious that we cannot understand the practice of science in similar, essentially practical terms, operating with its own thick concepts, whose applicability and inferential potential would be unintelligible without inhabiting the perspective they constitute. After all, the scientific way of thinking serves its role by helping us to navigate the world in general, just as surely as the way tied to our thick ethical concepts helps us to navigate the socio-ethical world.

4. Nietzsche on Perspectives and Values

Nietzsche himself clearly held that our perspectives are normally, and perhaps always, grounded in evaluative considerations. No comprehensive engagement with the Nietzschean texts is possible here, but as a way of recalling to short-term memory a few key texts that underwrite the present interpretation, let me take up explicitly the main piece of textual counterevidence against the way I have been reading Nietzsche, and explain how I would propose to disarm its apparent implications.

On my reading, Nietzschean perspectivism holds that all representations are perspectival, and further, that our perspectives—including cognitive perspectives—are never entirely pure, but always show some dependence on values and conative attitudes. It is easy to find (more or less) casual statements of such a position in Nietzsche: for example, criticisms of metaphysicians who “deny the role of impulses in knowledge” (GS 110), or his claim that Plato’s dogmatism about pure knowledge denies “perspective, the basic condition of all life” (BGE P), or his famous insistence in the Genealogy that disconnecting the affects from cognition would be “to castrate the intellect” and so render cognition impossible (GM III 12). But perhaps all this is Nietzsche’s typical hyperbole? At least one texts suggests as much, indicating that in more sober-minded moments, Nietzsche was willing to concede the possibility of a purified, affect-free mode of cognition, precisely in science. In a context contrasting science against perspectival, value-driven philosophy, he remarks,

To be sure: among scholars who are really scientific men, things might be different—“better,” if you like—there you may really find something like a drive for knowledge, some small, independent clockwork that, once well wound, works on vigorously without any essential participation from all the other drives of the scholar. The real “interests” of the scholar therefore lie usually somewhere else—say, in his family, or in making money, or in politics. Indeed, it is almost a matter of total indifference whether his little machine is placed at this or that spot in science, and whether the “promising” young worker turns himself into a philologist, or an expert on fungi, or a chemist: it does not characterize him that he becomes this or that. In the philosopher, conversely, there is nothing whatever that is impersonal; and above all, his morality bears decisive witness to who he is—that is, in what order of rank his inmost drives stand in relation to each other. (BGE 6)

As an initial point, we should separate two lines of thought here—one crucial to the questions I have been concerned with, and the other potentially relevant, but more orthogonal. The first strand concerns the objective shape of certain bodies of cognition; it would claim that the knowledge produced by the scholar’s “independent clockwork” is independent of perspective, and must therefore be composed of absolute representations (or be suitable for restatement in such terms via Leibniz/Moore-type reconstruction). The second strand consists of commentary on the psychological profile of the scholar as a type, and it focuses on the idea that there is a domain within her psychology that is “purified,” in the sense of being systematically detached from her noncognitive affects and interests (which “therefore lie usually somewhere else”).

In my view, the key to interpreting the passage comes from the basically instrumental characterization of Nietzsche’s description of science here. The scientist is defined by a psychological configuration that enables the development of some skillset (it hardly matters which one), which can function like a “little machine,” and is suitable for being deployed at this or that place within science, where it can perform its “independent clockwork.” This is the same picture of a psychologically “objective” scientist to which Nietzsche returns in BGE 207, where he complains against the “‘unselfing’ and depersonalization” involved in such an “objective spirit,” and then continues:

the ideal scholar in whom the scientific instinct, after thousands of total and semi-failures, for once blossoms and blooms to the end, is certainly one of the most precious instruments there are; but he belongs in the hand of one more powerful. He is only an instrument [my ital.]; . . . he is a mirror . . . he is accustomed to submit before whatever wants to be known, without any other pleasure than that to be found in knowing and “mirroring”; he waits until something comes and then spreads himself out tenderly lest light footsteps and the quick passage of spirit-like beings should be lost on his plane and skin. Whatever still remains of him as a “person” strikes him as accidental, often arbitrary, still more often disturbing . . . . (BGE 207)

Nietzsche goes on to argue that this precious instrument belongs in the hands of a philosopher, who is capable of providing a framework within which the knowledge the instrument produces can be given sense and point. That philosophical framework, meanwhile, will definitely not be perspective- or value-free. On the contrary, the very same passage that culminates with his talk of the scholar’s “independent clockwork” began (as I quoted earlier) with this: “Gradually it has become clear to me what every great philosophy so far has been: namely, the personal confession of its author and a kind of involuntary and unconscious memoir; also that the moral (or immoral) intentions in every philosophy constituted the real germ of life from which the whole plant had grown” (BGE 6).

With that, we are in a position to address the first and more decisive strand of thought. Is the scientific knowledge generated by the scholar rightly to be understood as part of a body of cognition converging on an absolute conception, and therefore as being (or at least pointing toward) an absolute representation that marks a limit on Nietzsche’s doctrine of affective perspectivism? The answer is clearly no. Such “clockwork” science belongs within, and draws essential, shaping content from, some philosophical framework or other. The philosophical concepts it depends on in the background are themselves related to one another holistically—abstract philosophical concepts, for Nietzsche, always participate in some system and draw part of their contents from it (BGE 20)—and then, too, all of them, together with the system they form, are impossible to isolate from “the moral (or immoral) intentions” that “constitute the real germ of life from which the whole plant had grown.” This conclusion supplies an important motivating explanation for Nietzsche’s conviction that scientific concepts will ultimately be revealed to carry some influence from evaluative commitments; they inherit it from the philosophical frameworks that give them a sense.¹⁸

To conclude otherwise, we would have to read BGE 6’s contrast between the impersonal scholar and the personally invested philosopher as an invidious comparison directed against the philosopher, criticizing her lack of objectivity. But that is clearly not the spirit of the passage. It is the impersonal scholar whose role is merely instrumental, in the service of the personally invested philosopher. When it comes time to assess the cognitive content of the resulting theoretical achievements, therefore, impersonal scientific results are to be located within a fully affectively loaded and perspectival philosophical framework, by which they are shaped. In the end, therefore, Nietzsche’s view has to be that all the representations involved are perspectival, and indeed, that their objectivity is of the sort grounded in, not insulated from, the affects. Their objectivity is perspectivist, “understood not as ‘disinterested contemplation’ (which is a nonconcept and absurdity), but rather as the capacity to have one’s pro and contra in one’s power and to shift them in and out, so that one knows how to make precisely the difference in perspectives and affective interpretations useful for knowledge” (GM III 12). This is the only legitimate version of objectivity because, in the words of the famous central passage on perspectivism:

There is only a perspective seeing, only a perspective “knowing”; and the more affects we allow to speak about a matter, the more eyes, different eyes, we know how to bring to bear on one and the same matter, the more complete will our “concept” of that matter, our “objectivity,” be. But to eliminate the will altogether, to disconnect the affects one and all, supposing that we were capable of this: what? would that not be to castrate the intellect? (GM III 12)

Even the second, psychological strand of the passage from BGE 6 reinforces the same moral, through its unmistakable suggestions that there is something psychologically wrong with the scholar. The spirit of objectivity she exemplifies may be instrumentally “precious” (BGE 207), but Nietzsche leaves us in no doubt that it is intrinsically and personally alienating, or even destructive for the scholar/scientist herself. It cuts the scholar off from her own needs, corrupts her self-understanding, and divides her against herself. The purity ideal she mistakenly imposes on her cognitive conception of knowledge ends by also being imposed upon her own psychology and devolving into a form of ascetic self-alienation.

5. Conclusion: Against Cognitive Purity

We are in a position to conclude. What I hope to have shown by means of the excursus through Williams is how centrally the idea (or ideal) of cognitive purity figures in both motivating and characterizing the conception of absolute representation upon which Moore, following Williams, rests the denial of perspectivism. Williams himself is (in my opinion, rightly) deeply skeptical of such claims to purity when it comes to practical reasoning. But his own conception of the essential indeterminacy of practical reasoning gives us little reason to believe in a sharp distinction between our reasoning in its practical mode and our reasoning more generally. Without such a sharp distinction, the analogy between external reasons on the practical side and absolute representations on the cognitive side takes on striking salience, and Williams’s persuasive doubts about cognitive purity in access to external reasons seem hard to resist when they are extended over to the cognitive side and applied to absolute representations. Nietzschean perspectivism, with its insistence on the importance of the conative roots of our cognitive lives, receives some compelling motivation.

In this context, two of the most characteristic, but otherwise confusing, aspects of Nietzsche’s core commitments about theoretical philosophy are likewise illuminated with fairly clear and straightforward motivations. Nietzsche’s genealogical approach to the criticism of alternative philosophical theories can be made to appear blatantly fallacious—so many mere ad hominem distractions which refuse to engage directly with the underlying theoretical questions. But from the point of view advanced here, such genealogical reasoning is well motivated. Nietzschean perspectivism holds that underlying value commitments will always be playing a role in shaping cognitive commitments, and if they are not brought to the surface so that they can be articulated, controlled, and potentially challenged, we cannot make any progress toward (what we should call) Longino objectivity. Conversely, such genealogies themselves are needed to provide further inductive and evidential support in turn for Nietzschean perspectivism, by exploding, case by case, our pretensions to cognitive purity.

To Nietzsche’s eyes, moreover, those very pretensions to purity are themselves symptomatic expressions of asceticism, so our result also clarifies the stakes of Nietzsche’s broader suggestions that the will to truth can become positively ascetic. Those suggestions directly target the ideal of purity that is bound up with our striving after absolute representations. What is at stake in this debate is not just a theoretical question about whether in fact absolute representations are possible, but also an evaluative question about whether the ideal of cognitive purity that erects the demand for them into a standard can be good for us, given our condition as finite, bounded rational agents. Nietzsche wants to convince us that the pursuit of such an ideal, and the associated effort to extricate ourselves entirely from the influence of our affectively shaped and socially embedded perspectives, is a form of deep self-denial—an inauthentic rejection of our inescapable way of being as cognitive agents.¹⁹

REFERENCES

Footnotes