Realism and Anti-Realism about Science A Pyrrhonian Stance

international journal for the study of skepticism 5 (2015) 145-167 brill.com/skep Realism and Anti-Realism about Science A Pyrrhonian Stance Otávio Bueno University of Miami otaviobueno@mac.com Abstract Pyrrhonists provide a way of investigating the world in which conflicting views about a given topic are critically compared, assessed, and juxtaposed. Since Pyrrhonists are ultimately unable to decide between these views, they end up suspending judgment about the issues under examination. In this paper, I consider the question of whether Pyrrhonists can be realists or anti-realists about science, focusing, in particular, on contemporary philosophical discussions about it. Although prima facie the answer seems to be negative, I argue that if realism and anti-realism are understood as philosophical stances rather than particular doctrines that is, if they are conceptualized in terms of a mode of engagement, a style of reasoning, and some propositional attitudes the apparent tension between Pyrrhonism, realism, and anti-realism vanishes. The result is a first step in the direction of bringing Pyrrhonism to bear on contemporary debates in the philosophy of science. Keywords realism anti-realism Pyrrhonism stance voluntarism 1 Introduction: Pyrrhonism and Pyrrhonists It is well known that Pyrrhonism is not a philosophical doctrine, and the Pyrrhonist is not someone who defends philosophical views. Rather, Pyrrhonism is a particular strategy of investigation. It can be thought of as an ability to manage opinion and information. In his description of this philosophy in the influential Outlines of Pyrrhonism (ph), Sextus Empiricus (2000) notes: koninklijke brill nv, leiden, 2014 doi 10.1163/22105700-04031176

146 Bueno Skepticism [that is, Pyrrhonism] is an ability to set out oppositions among things which appear and are thought of in any way at all, an ability by which, because of the equipollence in the opposed objects and accounts, we come first to suspension of judgment and afterwards to tranquility. ph I 8; italics added Sextus here makes it clear that Pyrrhonism is an ability of investigation, and the exercise of that ability is a way of implementing inquiry. The Pyrrhonists strategy of investigation consists primarily in the examination of arguments for and against a certain conclusion. Since for them these arguments seem to be equally persuasive, Pyrrhonists are unable to decide between the considerations provided, and as a result suspension of judgment emerges. It is also well known that Pyrrhonism does not involve beliefs; at least beliefs about the ultimate nature of things. On some interpretations, Pyrrhonism allows for beliefs about the appearances, as long as these beliefs are not thought of as involving dogmatic claims about what is really going on.1 Pyrrhonists tell us that they are not committed to establish the way things are (they are not dogmatists). They also tell us that they have no commitment to establish the way things are not (they are not negative dogmatists). Rather, they continue the investigation and, being genuinely unable to decide the issues under consideration, end up suspending judgment. Sextus himself highlights this point. Discussing the different kinds of philosophy, he notes: [I]n the case of philosophical investigations some have said that they have discovered the truth, some have asserted that it cannot be apprehended, and others are still investigating. Those who are called Dogmatists in the proper sense of the word think that they have discovered the truth for example, the schools of Aristotle and Epicurus and the Stoics, and some others. The schools of Academics have asserted that things cannot be apprehended. And the Skeptics are still investigating. Hence the most fundamental kinds of philosophy are reasonably thought to be three: the Dogmatic, the Academic, and the Skeptical. ph I 2 4 A distinctive trait of Pyrrhonism is the fact that it involves continuing the investigation rather than claiming that the truth has been established 1 For different interpretations of this issue within ancient Pyrrhonism, see the papers collected in Burnyeat and Frede (1997).

Realism and Anti-Realism about Science 147 (as dogmatic philosophers do) or asserting that the truth cannot be settled (as academic philosophers insist on). Pyrrhonists do not stop their inquiry for to do so would amount to either a form of dogmatism or simply giving up on investigating. After all, if Pyrrhonists stopped because they have found the truth, they would become dogmatists; if they stopped because they concluded that the truth couldn t be found, they would become negative dogmatists (which is what Academic philosophers amount to in Sextus classification of philosophical investigations above). What if they stopped because they got tired, sick or decided to do something else with their lives? In this case, the Pyrrhonist would simply abandon investigating and would also abandon being a Pyrrhonist. Even a temporary commitment to claiming that things are really a certain way (or that one cannot ever decide how they are) would take Pyrrhonists beyond Pyrrhonism. Suspension of judgment is, thus, an integral part of the Pyrrhonian investigation. But can the Pyrrhonist simply voluntarily decide to suspend judgment? Can one voluntarily decide to withhold belief? To the extent that deciding to believe in something is a puzzling phenomenon, the same seems to go for deciding not to believe and for deciding neither to believe in something nor to believe in its negation hence to suspend belief on the topic under consideration. This raises the issue of the connection between voluntarism and Pyrrhonism. Which degree (if any) of voluntary control over one s beliefs and opinions is possible, including the management of opinion involved in suspending judgment? And once voluntarism is on the table, how far does it extend? Does the Pyrrhonist suspend judgment about everything, or is such suspension somehow restricted? As is well known, Pyrrhonists follow the appearances, and the issue regarding the suspension of judgment about them doesn t emerge, since appearances are not claims about the way things are. Considerations as to whether to suspend one s judgment or not arise only once particular claims are made. These claims may be about a range of issues, including ordinary situations, scientific matters, or philosophical doctrines (among other possibilities). Whenever a claim is made (particularly about the nature of things), Pyrrhonian investigation strategies apply. To make the discussion concrete, in this paper I will examine claims involved in making sense of science, focusing on realist and on anti-realist views about it. In particular, is there a form of realism underlying Pyrrhonism? That is, do Pyrrhonists assume that scientific theories properly describe the world? Or, alternatively, does Pyrrhonism presuppose some form of antirealism? That is, do Pyrrhonists assume that scientific theories need not (or cannot) be true?

148 Bueno In this paper, I will address these topics by examining three issues within Pyrrhonism: (a) (b) (c) Are Pyrrhonists voluntarists? Are Pyrrhonists realists (in particular, about science)? Are Pyrrhonists anti-realists (in particular, about science)? The question (a), about voluntarism within Pyrrhonism, can be raised, of course, quite independently of questions (b) and (c), about the Pyrrhonian attitude toward realism and anti-realism. But I address the voluntarism issue in the context of the realism debate since this issue will pave the way to the discussion, central to my argument, of philosophical stances (van Fraassen 2002). As I use this concept, a stance is a practice of investigation composed of a mode of engagement, a style of reasoning, and some propositional attitudes (see Rowbottom and Bueno 2011). (I ll discuss each of these components in detail below.) And by understanding realism and anti-realism as philosophical stances, I will argue that there is no incompatibility between Pyrrhonism, realism and anti-realism. Thus, in the end, my answers to the three questions above are the same: no and yes (as long as we properly understand what is being asked). But there is no inconsistency in the Pyrrhonian attitude, since these answers, as so much else within Pyrrhonism, are not committing, given a proper use of stances. Before proceeding: a word of warning. In the discussion of Pyrrhonism, I will sometimes raise issues that ancient Pyrrhonists, such as Sextus, have not explicitly discussed. For instance, our contemporary conceptions and theories in science are significantly different from those found in Ancient Greece, although there are important similarities between them. But the hope is that what I have to say about Pyrrhonism captures, or at least is compatible with, the central features of the Pyrrhonian attitude. If not, consider the attitude I discuss as a form of neo-pyrrhonism.2 2 The term neo-pyrrhonism has been used, in different ways, by Robert Fogelin (1994) and Oswaldo Porchat Pereira (2007). Fogelin explores the implications of Pyrrhonism to contemporary debates in epistemology (particularly in accounts of knowledge and justification). Porchat tries to identify, more broadly, positive contributions of Pyrrhonism beyond the usual strategies of suspension of judgment. I am examining here the implications of Pyrrhonism to the framing of the realism/anti-realism debate in general philosophy of science, within broadly Porchatian lines. (I owe to Porchat much of my understanding of ancient Pyrrhonism.)

Realism and Anti-Realism about Science 149 2 Pyrrhonism and Voluntarism Voluntarism is the view according to which one can have some level of voluntary control over one s beliefs. It is common to distinguish the scope of voluntarism relative to the kinds of beliefs that are involved. (i) Basic perceptual beliefs are typically not included in the scope of voluntarism. We cannot simply decide to believe, given a table and our normal perceptual modalities, that there is no such table in front of us. We would normally need to have some defeaters in order to rule out the belief that the table is there (such as that we are hallucinating it). And even then, typically, the table will still seem to be in front of us if we have the corresponding visual experience. (ii) Inferred, theoretical beliefs are the usual candidates for beliefs under voluntary control. Suppose someone asks you: How many empty sets exist? Perhaps someone (say, a nominalist) answers: sets don t exist; therefore, there are no empty sets. But someone else (say, a platonist) may think that the proper answer may require stronger ontological commitments (which, it is thought, the question presupposes). A different answer is then provided: sets do exist, and there is only one empty set (the one referred to in the correct set theory). Alternatively, someone else (say, a more thoroughgoing platonist) may give yet another answer: sets exist, and there are several, in fact, infinitely many, empty sets (one for each rank in type theory; one for each possible set theory etc.). Which of these answers should you give? After examining the available arguments, it s by no means obvious that any one of them is ultimately telling none of them forces you to believe in a particular answer. It is not as though a single answer is the only viable alternative and needs to be adopted. As a result, you may then decide to believe one way or the other.3 Thus, there is an important difference between perceptual and theoretical beliefs. The former seem involuntary given your visual experiences whereas the latter seem to be within the scope of our voluntary control. I can t avoid forming the belief that I m typing on a laptop computer at this moment, just as presumably you can t avoid forming the belief that you are reading these words. In contrast, theoretical beliefs provide significantly more room for adjustment and control. 3 This seems to be a typical situation in philosophical debates. Usually there are several arguments in support of a certain philosophical view, and several arguments against it. But nothing forces one to believe that particular position. As a consequence, one can decide to believe one way or the other. Needless to say, this is precisely the kind of scenario explored by the Pyrrhonist.

150 Bueno There are those who deny that this is the case. Some theoretical beliefs force themselves upon us just as perceptual beliefs do. When Kurt Gödel writes that we have something like a perception of the objects of set theory (1964: 485), he is ultimately noting that, on his view, we are able to perceive these objects as having certain properties and lacking others, just as we perceive physical objects in our environment: this process also involves recognizing the objects in question as having or lacking certain properties. Gödel emphasizes that the perception of set-theoretic objects can be seen from the fact that the axioms [of set theory] force themselves upon us as being true (1964: 485). On his view, we have direct intuition of the truth of basic mathematical axioms, including those axioms of set theory (and arithmetic). As a result, if the relevant axioms force themselves upon us as being true, it seems that the resulting theoretical beliefs are not open to our voluntary control, just as perceptual beliefs are not. Since Gödel s account highlights a putative similarity between mathematical intuition and perception, it putatively supports an important analogy between perceptual and (at least some) theoretical beliefs. One could then use such a Gödelian account to resist voluntarism about some theoretical beliefs.4 How about Pyrrhonists, do they adopt voluntarism? There is one sense in which the answer is clearly No. Suppose that voluntarism is thought of as a particular philosophical view about the nature of belief formation. The view insists that, at least for a range of theoretical, inferred beliefs, it is under our voluntary control whether we are going to form or not a particular belief of this kind. Understood in this way, this is precisely the sort of view about which Pyrrhonists suspend judgment. They will point out that there is disagreement about the truth of voluntarism: some defend it; others reject it. On the one hand, a voluntarist epistemology is an important feature of Bas van Fraassen s empiricist account of science, including, in particular, theoretical beliefs (van Fraassen 1989; 2002). On the other hand, we saw that even regarding theoretical beliefs a Gödelian view could be invoked to resist voluntarism about them. Pyrrhonists, however, are unable to decide the issue. There are good arguments to resist voluntarism about theoretical beliefs, given the similarities between intuition and perception, but there are also good arguments to adopt voluntarism about such beliefs, given that they generally allow for adjustments under our control. It is genuinely unclear for Pyrrhonists how to decide 4 For further discussion of Gödel s view on this issue, see Maddy (1990) and Parsons (2008). Eli Chudnoff (2013) develops a fascinating account of intuition according to which, similarly to perceptions, intuitions have presentational phenomenology. On his account, just as in Gödel s, it is less clear whether there is room for voluntarism about at least some theoretical beliefs.

Realism and Anti-Realism about Science 151 between the two opposing views. This leads them to suspend judgment about the issue, and as a result, it cannot be said that they embrace voluntarism. There is, however, a sense in which a Pyrrhonist seem to be a voluntarist. Perhaps the best way to understand the Pyrrhonian attitude regarding voluntarism is to consider it as a form of voluntary agnosticism: voluntarism regarding suspension of judgment. It is under one s voluntary control whether one suspends judgment or not about a given issue. But this point needs to be understood carefully. The Pyrrhonist is genuinely unable to decide between rival accounts of a given issue. In this respect, suspension of judgment simply emerges from the inability to decide, and is not a voluntary action. However, Pyrrhonists can voluntarily produce the conditions under which they will be led to such suspension. In this way, they systematically explore a form of voluntarism as part of their investigation: they actively look for counterarguments to balance the weight of evidence offered by the particular view under examination. The active process of search for counterarguments creates a situation in which Pyrrhonists end up being struck by the equal force of arguments, pro and con, regarding the issue at hand the opposing evidence that is generated from this search. The result is then suspension of judgment.5 But it is important to note that voluntary agnosticism is not a philosophical view. Pyrrhonists would suspend judgment about any such proposal. Rather, the idea is simply that Pyrrhonists seem to exhibit some voluntary control over their suspension of judgment. To insist on this point: Pyrrhonists voluntarily engage with the process of continually examining the evidence for and against a certain claim until their ability to suspend judgment is manifest being unable to decide the issue (given the evidence), they suspend judgment and then recurrently repeat this process, again and again. This suspension of judgment may be thought of as a form of voluntarism; it does not lead to the commitment to a particular belief, but rather to the lack of commitment, since no belief is acquired. What we have here is the suspension of judgment typical of the agnostic who holds no view about the nature of the matter under consideration. However, there may be a problem for Pyrrhonism if it is conceptualized in terms of voluntary agnosticism. It may be objected that, if voluntary agnosticism is invoked, one does not obtain a serious form of skepticism (Lammenranta 2008: 15). In order to obtain suspension of judgment, Pyrrhonists may need to choose the evidence selectively. After all, it is only on extremely rare occasions that the evidence will point equally strongly in opposite directions. If the 5 See Alston (1988) for an examination of different kinds of control of belief (and other propositional attitudes).

152 Bueno evidence is collected responsibly, and if it is taken properly into account, the argument goes, suspension of judgment will rarely emerge. As a result, Pyrrhonists do not seem to be serious inquirers interested in the truth, and the corresponding form of skepticism, if it involves choosing selectively the evidence, is not very serious either. As Lammenranta notes: If the skeptic really finds such equally convincing oppositions everywhere, this requires that she attend only to arguments that are equally strong on both sides of the issue, and manage to forget arguments that do not balance in this way. But why should anybody try to do this? If we aim at truth, we should rather take all arguments on both sides of the issue into account, and these arguments typically do not balance. 2008: 15 In response, three points should be made. (a) Pyrrhonists do not require that the evidence point equally strongly in opposite directions. As Lammenranta presumably agrees, no one should require that either. If the total evidence available supported P and if it also supported not-p (thus supporting P and not-p), it would be seriously misleading as evidence. The total evidence is not required to support, equally well, contradictory (or even contrary) claims. Rather, the evidence is such that one is unable to decide whether P or not-p is the case. There may be evidence for P and some other evidence for not-p (not for both!), and it may be unclear how to rule out the evidence in each case. It may be argued that if a part of the evidence supports P and some other part supports not-p, then the evidence as a whole supports P and not-p equally strongly, since it would give both, say, a probability around 0.5. But in this case it seems that one would not be in a position of favoring P over not-p (or vice versa), which is precisely the situation the Pyrrhonian skeptic is in. Of course, Lammenranta is not suggesting that the evidence is contradictory, but that skeptics selectively choose the evidence, and somehow manage to forget arguments that do not balance in this way. But is this really the case? (b) I don t think Pyrrhonists selectively choose the evidence. If they did that, it would be perfectly appropriate to ignore them. But the situation is very different. Suppose that the available evidence supports one party to a dispute. In response, Pyrrhonists will question to what extent the evidence in fact provides the relevant support. However, this is done dialectically. For example, Pyrrhonists may indicate considerations that seem to question, if not completely undermine, the alleged evidence. Or they may suggest considerations that seem to favor the negation of the conclusion supported so far. As a result, they will note, one cannot conclude just yet that the evidence favors one party to the debate.

Realism and Anti-Realism about Science 153 But are there always undermining considerations that favor the negation of the conclusion at issue? Lammenranta s point is that often there are not. Pyrrhonists disagree. In fact, their practice of investigation consists in keep searching for such undermining considerations to counterbalance the dogmatists claims. And by uncovering counterarguments, Pyrrhonists can resist the excesses of those who claim to have found the true nature of things. However, what would happen if the undermining considerations couldn t be found, and Pyrrhonists were unable to provide suitable counterarguments to a given claim? If after searching further with no success (perhaps the person who could provide the counterargument hasn t been born yet?), Pyrrhonists might eventually convince themselves that there were no such counterarguments. In that case, they would no longer be Pyrrhonists, and would become dogmatic philosophers. That s a possibility that Pyrrhonists don t rule out, and this fact underscores the seriousness of the Pyrrhonian investigation. (c) In light of these considerations, we can see why Pyrrhonism does not produce an irresponsible form of inquiry. On the contrary, Pyrrhonists are exceedingly careful and systematic in their adjudication of disputes, and refrain from making judgments in the absence of suitable evidence. If, in the end, it turns out that the evidence does not clearly favor either side of the issue under consideration the arguments Pyrrhonists consider seem to be equally persuasive to them suspension of judgment results. 3 Pyrrhonism and Realism In the contemporary literature, there are many forms of realism about science. I will mention three prominent ones: (i) Standard scientific realism (Putnam 1979; Boyd 1990; Psillos 1999): According to the scientific realist, scientific theories are true (or approximately true), and the terms of these theories refer (to relevant objects in the world). Moreover, scientific theories are empirically successful because they provide true (or approximately true) accounts of the objects that compose the world. However, concerns about how to accommodate reference in the context of scientific change have motivated an alternative realist view that does not rely on such reference. On this view, rather than securing reference to particular entities, the realist should be realist about the relevant structures (characterized, in particular, by suitable properties and relations). The resulting view is a structural form of realism. (ii) Structural realism (Worrall 1989, Ladyman 1998; French and Ladyman 2003a, 2003b; French 2006): For the structural realist, scientific theories are

154 Bueno true (or approximately true); however, the truth of these theories does not depend on the referential success of their terms, but on their ability to capture the relevant structure of the world. Scientific theories are empirically successful because they provide true (or approximately true) accounts of such structure. But how should we make sense of what a structure is if it is not characterized (or constituted) by underlying objects? Is it intelligible for relations to be primary than objects? Concerns of this kind have motivated an alternative realist view that incorporates both structural and object-oriented features. The resulting view is semirealism. (iii) Semirealism (Chakravartty 2007): According to the semirealist, scientific theories are true (or approximately true), but the terms of these theories refer to both objects (that is, relevant entities in the world) and properties (in particular, properties that can be detected). The cluster of these properties and the particular relations that hold among the objects under consideration yield the relevant structures. Scientific theories are empirically successful because they provide true (or approximately true) accounts of the detection properties in the world, and thus they provide proper accounts of the relevant objects and the appropriate structures in which they feature. In a certain respect, semirealism inhabits a common ground between an entity-based scientific realism and structural realism. On the one hand, the role of objects in scientific success is emphasized, given that objects need to be properly identified; on the other hand, the role of structural considerations is also stressed by noting the importance of detection properties and the relations among the objects in question. Are Pyrrhonists realists? There is one sense in which the answer is No. Scientific realism, structural realism and semirealism are all philosophical views about the nature of science. They are precisely the sorts of thing about which Pyrrhonists suspend judgment. There is, of course, disagreement about which of these views is correct. Structural realists argue that scientific realism is inadequate given its commitment to the metaphysics of objects. Scientific realists complain about the intelligibility of the structural realist conception of a structure without any objects. Semirealists point out the shortcomings in both views, finding a common ground that acknowledges some role for objects and structures (determined, in part, as noted, by detection properties). In turn, scientific realists will complain that the semirealist picture concedes too much to the structural realist by acknowledging the role of structures. And structural realists will complain that the semirealist picture concedes too much to the scientific realist by acknowledging the role of objects. Pyrrhonists, being unable to decide these issues without begging

Realism and Anti-Realism about Science 155 the question against some of these views, suspend judgment. In this sense, Pyrrhonists are not realists, and they could not be, given the dogmatic characterization of science these views provide (in the sense that they advance defenses of what scientific knowledge ultimately is). There is, however, a sense in which Pyrrhonists can be thought of as being realists. They are, after all, searching for the truth, which is, as should be clear by now, the central feature shared by all realist accounts of science. In fact, searching for the truth is perhaps the only feature among realist views about which there is some agreement. But in what sense are realists searching for the truth? There are at least two ways of answering this question.6 One is to understand it as a question about the proper interpretation of science (the question is, thus, raised at the metalevel): Scientific realists are trying to articulate the correct interpretation of scientific theories; they are seeking the truth about which epistemology of science is right. In this sense of truth seeking, there seems to be no difference with antirealism. But there is another way of understanding the question as addressing something about the world (the question is, thus, raised at the firstorder level): Realists are searching for truths about how the world is beyond the appearances, by interpreting scientific theories as they do. In this way of understanding the question, realism and anti-realism are significantly different. One could imagine Pyrrhonists seeking both sorts of truth. They could be trying to determine what is the proper interpretation of scientific theories, but they could just as well be searching for truths about the world that go beyond the appearances. It just happens that Pyrrhonists haven t found these truths yet. But that doesn t prevent them from persisting in trying to find them, and thus continuing the search. As Sextus tells us: Men of talent, troubled by the anomaly of things and puzzled as to which of them they should rather assent to, came to investigate what in things is true and what false, thinking that by deciding these issues they would become tranquil. ph I 12 Here we have a description of the way in which a person of talent eventually became a Pyrrhonist: it was by searching for such truths, and by systematically failing to be able to decide them, that this person eventually noted the rashness of deciding the issues under consideration. The difficulty to make a 6 Thanks, here, to Anjan Chakravartty.

156 Bueno decision emerged from the fact that to every account an equal account could be opposed (ph i 12), and Pyrrhonists were unable to decide between the conflicting accounts. In fact, it s precisely the ability to make such oppositions that Pyrrhonists explore in their inquiry, and the oppositions include, I suggest, different interpretations of scientific theories and conflicting claims about the way the world is beyond the appearances. In fact, it is this ability to set out oppositions among things which appear and are thought of in any way at all (ph i 8; italics added) that Pyrrhonists explore. Among those items that are thought and opposed in any way at all, I suggest, we have the two different ways of interpreting the realist s search for truth. As a result of their investigation, however, being unable to decide between opposed and equally persuasive accounts, Pyrrhonists eventually come to hold no beliefs (ph i 12). Hence, since they are still searching for the truth in either of the two senses just identified they can be thought to be realists, without, however, being committed to realism as a philosophical doctrine. But how exactly should realism be understood in this case? Clearly, for Pyrrhonists to be realists, realism cannot be conceived of as a particular doctrine to be believed (as noted, Pyrrhonists will just suspend judgment about such a view). But realism can be thought of in a different way: it can be conceived of as a practice of investigation (with associated goals, norms, methods, and techniques); in other words, it can be conceived of in terms of what I will refer to as a stance. In the contemporary philosophical literature about science, Bas van Fraassen has been responsible, perhaps more than anyone else, for rehabilitating the concept of stance. On his view: A philosophical position can consist in something other than a belief in what the world is like. We can, for instance, take the empiricist s attitude toward science rather than his or her beliefs about it as the more crucial characteristic. A philosophical position can consist in a stance (attitude, commitment, approach, a cluster of such possibly including some propositional attitudes such as beliefs as well). Such a stance can of course be expressed, and may involve or presuppose some beliefs as well, but cannot be simply equated with having beliefs or making assertions about what there is. van Fraassen 2002: 47 48 The central feature of a stance is the attitude one has toward a domain of inquiry (say, a certain portion of the world) rather than the particular beliefs one may have about it. To talk about a stance is to emphasize the relevant attitude.

Realism and Anti-Realism about Science 157 But this doesn t yet answer the question of what a stance is. In particular, what is involved in having a certain attitude toward a given domain? Although van Fraassen does not elaborate on this point, a number of authors have considered this very issue. In particular, a stance can be thought of as having three components (see Rowbottom and Bueno 2011): (a) A stance has a mode of engagement, that is, it involves a particular way of approaching the world. For example, in considering a certain situation, someone can be more or less active, more or less contemplative; the person can also be more investigative or more dogmatic in examining the domain. (b) A stance also has a style of reasoning, that is, it involves certain patterns of inference that are used to obtain the relevant results. Different patterns of inference employ different inferential devices: some are logical (different logical principles are at stake, depending on the particular logic that is used); other devices are of a non-logical sort (for instance, diagrams, templates, models, simulations; each can be used as inferential devices). Finally, (c) a stance involves certain propositional attitudes, such as beliefs, desires, or hopes as well as commitment, detachment, pursuit, or suspension. These attitudes, however, typically are not identified with a stance, in the sense that normally they are not invoked to individuate it. As an illustration, the following example of a stance may be helpful. Consider constructivism in mathematics. Conceptualized as a stance, constructivism involves the three components just discussed: (a) The mode of engagement is the critical attitude toward the introduction of mathematical objects without a corresponding method of construction. The constructivist insists that, in the absence of a proper method, no such introduction is acceptable. (b) The style of reasoning involves arguing that proofs by reductio ad absurdum are not legitimate devices for positing mathematical objects. More generally, non-constructive methods of inference are systematically questioned (for example, cases in which the axiom of choice is used to establish the existence of certain mathematical objects are typically contested). Only constructive inferences are accepted. Finally, (c) among the propositional attitudes, due to the mode of engagement and the style of reasoning that are adopted, the disbelief in certain results from classical mathematics emerges. In just the same way, it is possible to consider realism about science also as a particular stance. The three relevant components are satisfied: (a) The mode of engagement, not surprisingly, is the search for true (or approximately true) scientific theories. This is a realist mode of approaching the investigation of the world. (b) The style of reasoning involves accepting inference to the best explanation as an inferential device within science. Although strictly speaking not necessary for realism in science, such inference is fairly common among

158 Bueno realists. Finally, (c) among the propositional attitudes we find, for instance, the belief in the existence of unobservable features of the world, which is another common trait among realists. However, as noted, such a belief need not be used to individuate a stance. In examining contemporary science, Pyrrhonists could be realists in the sense of invoking a realist stance. The mode of engagement they invoke is one in which the search for truth is crucial (in either of the two senses discussed above). It s just that, as noted, Pyrrhonists haven t found it yet. In good Pyrrhonian style, they keep searching. So this key aspect of realism is really not in conflict with Pyrrhonism. Moreover, their style of reasoning may involve the adoption of inference to the best explanation as a device of hypotheses generation. Once a hypothesis is identified as the best explanation of a certain phenomenon (given the available alternatives), it is considered as a candidate for the truth. Pyrrhonists will then critically examine, as good realists do, how strong is the evidence in support of that hypothesis, and whether some alternative hypotheses that haven t been considered yet could be better supported. What this means, however, is that the investigation will then continue, by uncovering alternative possibilities and assessing them. The commitment to establish the truth is not a requirement for a realist view. This would be an unreasonably high demand for a philosophical proposal at any rate. The central component is the systematic search for it and that s precisely what Pyrrhonists do. As part of this process of investigation, Pyrrhonists will entertain, as realists do, a number of propositional attitudes; in particular, the belief that there are unobservable features of the world. They can do that, and have done so, dialectically. They will consider these unobservable features and contrast rival accounts of them together with proposals that emphasize entirely different unobservable traits not to mention conceptions that deny the existence of these features altogether. Pyrrhonists will examine critically, as realists do, the weaknesses and strengths of the various possibilities, and they can do that non-dogmatically, that is without having to be committed to any particular proposal. In this way, even this component of realism the belief in the existence of an unobservable reality beyond the appearances is something available to Pyrrhonists, as long as it is understood as something they explore dialectically, as part of their investigation. The result is a non-dogmatic form of realism, in which at issue is the search for the truth, rather than its establishment; the search for hypotheses that better explain the phenomena, rather than the commitment to one unique explanation; and the search for unobservable traits of the world, rather than the advocacy of a single one proposal among many possibilities. In this way,

Realism and Anti-Realism about Science 159 understood as a stance, realism becomes a form of investigation rather than a particular philosophical doctrine. It is a kind of investigation in which the emphasis lies on the search for the goal rather than on its guaranteed establishment. As is well known, Pyrrhonists are those who keep searching. And as long as the mode of engagement involves a critical attitude and a lack of commitment, they can adopt a realist stance. Understood in this way, realism provides a form of investigation that is not in conflict with Pyrrhonism. To sum up: it seems that Pyrrhonists can be realists in the sense that they explore a mode of engagement and a style of reasoning that are crucial to a realist stance, precisely because this excludes the sorts of commitments that one may fear enter into the propositional attitudes that are part of realism and which would seem on their face to conflict with Pyrrhonism. However, once one sees how these propositional attitudes function in their dialectical process, even the Pyrrhonist can non-dogmatically embrace them. Typically most realists adopt the sorts of beliefs referred to in the third component of a realist stance. In fact, one might think that this is constitutive of realism. The suggestion here is that, on the stance conception, such commitments need not be constitutive of realism. One can envision other attitudes towards such propositions being compatible with realism, including the dialectical attitude characteristic of the Pyrrhonist. The result is that, as long as realism is conceived of as a stance, there need not be any conflict with Pyrrhonism. 4 Pyrrhonism and Anti-Realism As in the case of realism, there are many anti-realist views about science. I will consider two: (i) Instrumentalism (Nagel 1950): According to the instrumentalist, scientific theories are not truth-apt; they are simply tools of inference that are used to derive predictions about the observable world and, in this way, are helpful to organize experience. These theories can also be employed to connect empirical laws and observations. As a result, they help to establish the relevance that some such observations have to one another, while also guiding empirical research. For the instrumentalist, theories are unable to represent anything beyond experience.7 7 For a helpful survey of different kinds of instrumentalism, see Psillos (1999: 15 37).

160 Bueno Concerns have been raised regarding the viability of taking scientific theories in such instrumentalist terms. A significant form of revisionism regarding scientific practice is required to get such a view off the ground, since it demands that scientific theories not be taken literally. But there are anti-realist alternatives that manage to take scientific theories literally, such as the following one. (ii) Constructive empiricism (van Fraassen 1980, 1989, 1991, 2008): For constructive empiricists, scientific theories need not be true to be good, as long as they are empirically adequate (and informative); that is, roughly, as long as what they say about the observable world is correct. However, constructive empiricists recommend agnosticism regarding the unobservable, since according to them it is unclear how one can establish what is really going on beyond the observable level, given particular instances of underdetermination the same phenomena are compatible with conflicting unobservable descriptions. Consider, for instance, different interpretations of non-relativist quantum mechanics, such as the Copenhagen interpretation and the many-worlds view: although they agree with the empirical predictions generated by the theory, what each of them has to say about the unobservable world is significantly different. In the end, it is unclear how we can be in a position to know what is really going on beyond the appearances. Moreover, the constructive empiricist insists, it is possible to make sense of significant features of scientific practice without the commitment to the existence of unobservable entities, their properties, and relations. Are Pyrrhonists anti-realists? In one respect, clearly they are not. Anti-realist views about science are dogmatic proposals about the nature of the scientific enterprise. Despite the different commitments they recommend vis-à-vis their realist alternatives, they still advance definite views about the issues under consideration. Instrumentalism is committed to a particular conception of the nature of scientific activity, highlighting the central role that predictions play in this enterprise, and they deny that truth is the proper criterion for evaluation of a scientific theory. Constructive empiricists defend particular claims about the aim of science (the search for empirically adequate theories), the nature of the observable (which is restricted only to what can be seen with the naked eye), and the limitations of our epistemic access to unobservable features of the world (given particular underdetermination arguments). Pyrrhonists, not surprisingly, will suspend judgment about all of these claims, invoking if needed realist arguments to counterbalance the considerations provided by anti-realists. In another respect, however, Pyrrhonists can be thought of as being antirealists. Consider anti-realism as a philosophical stance. (a) As its mode of engagement, we have a critical attitude toward the idea that scientific theories

Realism and Anti-Realism about Science 161 need to be true to be good. Pyrrhonists will challenge, along instrumentalist or constructive empiricist lines, the adequacy of truth as a proper aim for science. They will note the difficulty of adopting, as an aim for science, something that one cannot know that has been reached even if it has indeed been reached (Laudan 1984). They will raise objections to the introduction of a concept (of truth) that, when taken as an aim of science and applied to scientific theories, commits one to all kinds of objects that are invoked in such theories: from whatever mathematical objects that are referred to, to all sorts of unobservable entities, processes and structures that may be in the world. In order to prevent such consequences, one needs to develop some nominalization strategy for mathematics (for a critical survey, see Bueno 2013) or some way of restricting the scope of the truth predicate when applied to scientific theories (for a defense, see Psillos 1999). A critical mode of engagement, which is common to a number of antirealist views, is central to anti-realism since this stance is typically adopted as a reaction to what is perceived as the excesses of realist alternatives. For the anti-realist, the various commitments that are embraced in realism are often questionable. (Realist conceptions can, of course, also adopt this critical mode of engagement, but it is not typical of what makes them realist.) Precisely this sort of critical engagement is similarly crucial to Pyrrhonism, which also emerges in response to the excesses of dogmatic philosophies in their attempt at establishing the truth (or approximate truth) about the relevant domains. From the perspective of a stance, anti-realism and Pyrrhonism share the same mode of engagement. Some anti-realist views, however, go beyond the critical mode of engagement, and attempt to establish the negative claim that nothing in a given domain (e.g., about unobservable objects) can be known. But to establish a claim of this sort can be just as difficult as to establish the corresponding positive claim (namely, to the effect that knowledge of unobservable objects is possible). Given the stance understanding of anti-realism, the critical mode of engagement will keep the anti-realist searching for whether it is indeed impossible to establish the relevant knowledge claim. By being critical about what it takes to establish such a negative claim, the anti-realist will pursue alternative arguments and will keep searching for better ways of trying to settle the matter. The critical mode of engagement is, thus, crucial to the process of making it clear that more is required to establish claims of this negative sort.8 Not 8 Perhaps Pyrrho himself was a metaphysical anti-realist, according to whom we must suspend judgment because no beliefs are true or false (for a thorough discussion of Pyrrho, see Bett 2000).

162 Bueno surprisingly, the compatibility of anti-realism and Pyrrhonism with regard to the mode of engagement emerges once again. (b) As its style of reasoning we have the questioning of inference to the best explanation as an acceptable inferential device within science or with which to interpret scientific activity (or both). Anti-realists will question the reliability of using this inferential mechanism given its inability to guarantee the truth of the relevant conclusion even if one starts from true premises. Moreover, one could be dealing with a bad lot of potential explanations, and thus one would infer the best explanation from the lot as the true one, despite the fact that the true explanation is not there (van Fraassen 1989). Once again, the critical considerations that are raised against inference to the best explanation should not be understood as conclusively deciding the issue. Additional arguments in support of the rival perspective can always be advanced, and the anti-realist will keep investigating the matter. Finally, (c) among its propositional attitudes we find agnosticism about the existence of unobservable entities. Given the critical mode of engagement and the difficulties that were raised to inference to the best explanation, it is expected that, as part of an anti-realist stance, a non-committal attitude be in place. But why should agnosticism be embraced rather than the (admittedly stronger) denial of the existence of unobservable entities? An anti-realist need not be committed to rejecting the unobservable. After all, to establish its nonexistence is just as difficult as to establish the opposite. For all we know, perhaps unobservable entities do exist. But, the anti-realist will argue, nothing requires such commitment one way or the other. An agnostic attitude then emerges. For dialectical purposes, however, when mounting a critique of standard commitments found in realism, the anti-realist can, and often does, engage with negative claims to the effect that unobservable entities do not exist. But this does not amount to any kind of commitment on their part, provided that the points are only made dialectically. In this way, understood as a stance, anti-realism becomes a form of investigation rather than a particular philosophical doctrine. As long as the mode of engagement involves a critical attitude and a lack of commitment, Pyrrhonists can adopt an anti-realist stance. Moreover, while engaging with anti-realist issues, Pyrrhonism would explore them dialectically, contrasting them, as needed, with realist views. To conclude: it seems that Pyrrhonists can be anti-realists in the sense that they explore a mode of engagement and a style of reasoning that are crucial to an anti-realist stance, precisely because this excludes the sorts of commitments that one may fear enter into the propositional attitudes that are part of

Realism and Anti-Realism about Science 163 anti-realism (such as the belief that unobservable entities don t exist) and which would seem on their face to conflict with Pyrrhonism. However, once one sees how these propositional attitudes function in their dialectical process, even the Pyrrhonist can non-dogmatically embrace them. Certain anti-realist beliefs (e.g., about the nonexistence of unobservable entities) may be thought of as constitutive of anti-realism. The suggestion here is that, on the stance conception, such beliefs need not be understood in these terms. One can envision other attitudes towards the relevant propositions being compatible with anti-realism, including the dialectical attitude and the agnostic way characteristic of the Pyrrhonist. The result is that, as long as anti-realism is conceived of as a stance, there need not be any conflict with Pyrrhonism. 5 Pyrrhonism, Scientific Practice, and the Realism/Anti-Realism Debate By conceptualizing realism and anti-realism as stances to be taken rather than doctrines to be believed, the Pyrrhonist is able to avoid taking a stand on the debate between realists and anti-realists about science. Conceptualized in these terms, no dogmatic commitment to particular beliefs is involved (such as those concerning the existence of unobservable entities): just different attitudes toward scientific research are at issue. As we saw, realists typically take the scientific enterprise to involve the search for underlying features of the phenomena, whereas anti-realists tend to question whether the search for any such traits is in fact required, noting that if the phenomena are saved, in informative ways, this is enough to make sense of scientific research. These are, of course, different ways of implementing an investigation of the world. But to make them metaphysically substantive, additional commitments are needed, e.g., regarding the existence of unobservable objects, the search for true (or approximately true) scientific theories, the assignment of a central role to inference to the best explanation as a guide to truth (or approximate truth) of the relevant theories or, alternatively, the denial of each of these proposals. Commitments of this kind are typically invoked in order to argue that the underlying traits of the phenomena have been identified and possess the properties that have thereby been uncovered (following a version of realism), or that no such traits can ever be detected or that it s impossible for one to be in a position to settle such issues (following some forms of anti-realism). But these features involve particular interpretations of scientific activity and its achievements, and thus go beyond the simple specification of the relevant