|
|
|
|
 Naturalism and Pragmatism
[NOTE: The kind of pragmatic naturalism defended in this essay is best categorized into the Perspectival Liberal Pluralism variety of naturalism, as described by "Naturalism and Science"]
Pragmatic naturalists, despite their explicit affirmation of the reality of natural entities whose existence is not dependent on mind or experience, nevertheless receive censure for failing to be genuine realists. Self-appointed protectors of realism abound, and their high standards demand additional affirmations upon the nature of, and knowledge about, natural entities. Pragmatists have consistently run afoul of such standards, due to their theories of naturalism and scientific method. One version of pragmatic naturalism, based primarily on strands drawn from the philosophies of Charles Peirce and John Dewey, will be defended and examined for its ability to satisfy a variety of realist standards. Pragmatic naturalism developed from an abiding esteem for science and its achievements that was not expressed by most idealisms. But reactions against idealism took a wide variety of forms, and thus realism was a splintered movement from the beginning. While pragmatism was arguing that scientific warrant is measured by a theory’s utility in human efforts to reliably control natural processes, realisms were seeking a way to safeguard science from any entanglement with mind-dependence.
The reality of natural entities
The early years of the 20th century found naive realists repelling idealism with the principle that natural entities (to use a neutral term to cover “object” or “event” or “regularity,” etc.) have only a contingent relationship with experience. When present to experience, the entity itself is in experience (denying dualistic representationalism), and when the entity departs from experience, it retains the same qualities that were observed while in experience. Unless the entity’s qualities were as contingently related to experience as the entity itself, the naive realists argued, idealism would prevail by asking the troubling question of how an entity stripped of its observed qualities could possibly be conceived. As the 20th century progressed, an alternative to naive realism and idealism arose in the form of a revived scientific realism, which uses scientific conceptions of entities to give them some experience-independent characteristics. There are two basic types of scientific realism. Moderate scientific realism requires that most or all observed qualities of entities are only in experience, because their existence is dependent on the involvement of a living observer. “Extreme” scientific realism locates all observed qualities, even those obtained through scientific instruments, only within experience. The conceptions of the “real” characteristics of entities, and thus the conceptions of the really existing entities themselves, are composed solely of abstract traits that cannot even in principle be observed by any means. Extreme scientific realism invites the sort of reductionism that supports a privileged status for subatomic physics, but that reductionism is positively irrelevant to the other sciences. No geologist would wait for the natural laws of geo-physics and plate tectonics to be expressed solely in terms of sub-atomic forces, and the natural entities sought in confirmation of a geologist’s theories are eminently identifiable in experience. The physicalist conclusion that color, for example, may only be the surface spectral reflectance of an object, is of little interest to a chemist looking for the distinctive color of a fluid’s chemical reaction. The practical acceptance of anomalous monism’s respect for the laws of each scientific field is matched in scientific practice by a respect for the experienceable qualities of the entities pursued by the respective sciences. The respect due to experienceable qualities of natural entities should not be diminished by their necessarily perspectival constitution. Bertrand Russell’s famous complaint that no perspective has superior epistemic status, leaving the defining characters of an entity mysterious and thus unobservable, is no longer generally taken to be a good proof that perspectives completely fail to supply qualities of the natural entity itself. Indeed, perspectives are the best explanation of our competence for discriminating between natural entities in space/time and entities that exist only within us, since such discrimination is predicated on our motile ability to get various perspectives upon the changing qualities of natural entities. The visual aura symptomatic of ophthalmic migraines, for example, remain stubbornly unaffected by blinking, looking around, moving one’s head, etc. The rationalistic prejudice towards defining the entity’s reality solely in terms of its unchangingly permanent characters has gradually faded, and no longer provokes many philosophers into locating perspectival qualities exclusively within the observer at the perspective’s locus. Just as anomalous monism is still naturalistic, so too is a perspectivalism that locates observed qualities in nature instead of within a subjective realm of mind or brain. The decline of rationalism does not imply that observed qualities must be naively located in the object, either. Scientific realism could, and likely should, prefer the naturalistic perspectivalism promoted by such philosophers as John Dewey, A. N. Whitehead, George Mead, and Justus Buchler. To use the example of colors again, this naturalism agrees with science that color does not exist exclusively in the light source or the photons (after all, a photon is not seen to be colored upon close scrutiny) or on the object or in the observer, but rather in their interaction together in a particular situation. While some philosophers remain who would prefer that colors exist only where one of these three things are, they cannot explain why making changes to either of the other two things alters the observed color. Instead, in their fallacious quest for a concrete location for colors, these philosophers at best promote an additional definition of “color,” and at worst, reductively suppose that their preferred “colors” really exist instead of the color actually observed. Pragmatic naturalism is a type of perspectival naturalism which has no complaint against science’s novel concepts (“red” light as that radiation of 4.3 x 1014 Hz), but protests any substitution of such scientific entities for the observed color. Pragmatic naturalism endorses a situational contextualism for colors and other traditional secondary qualities to replace a naturalism committed to the fallacy of concrete location. To be properly aligned with science and its results, it is quite unnecessary for a naturalism to obey the outdated rationalistic prejudice towards objectification and its concomitant premise that only the intrinsic non-relational characters of entities are its real characters. Relational qualities are just as natural as intrinsic qualities. Reflected colors exist where they are experienced: in situations that include a source of light, the light, a reflective object, and a sensitive organism. While it is the situation which is colored, it remains convenient to say instead that the reflective object is colored since the reflection occurs on the object’s surface and so the color seems to be where the object is. In the language of emergent naturalism, the color is an emergent property (not a supervenient property) of the interacting portions of the situation. And because it is the situation which is experienced, the color is thus a feature of experience. This is a naturalistic understanding of experience, of course. No ontological gap between nature and experience opens up, because experience is directly and immediately of its objects, and should not be confused with some sort of subjectively internal intermediary. The kind of pragmatism defended in this essay endorses naturalistic empiricism, which in turn is compatible with a moderate scientific realism that respects perspectives. Some realists argue that perspectival naturalism is unable to be truly realistic because relational qualities are by definition not independent of human experience. But relational qualities do not exhaust reality; perspectivalism is not driven towards sense-data empiricism or positivism. We experience the natural world first and foremost as a world of objects. We learn fairly quickly in life that most natural objects endure through time and tend to continue to exist while they are not being experienced. Pragmatists repudiate a positivist skepticism about the existence of things beyond experience. Our experience in due course brings us to the conception of objects and processes as retaining some characteristics (such as weight, size, motion, growth) after they leave situations in which humans experience them. Against the sensationalist empiricisms of Hume and Mill, Peirce and Dewey found that experience displays stable regularities that make some universal judgments adequate to natural processes. Peirce argued that the observation of a series of events can induce us to firmly believe in real regularities, such as the falling of dropped stones.
Peirce therefore could not agree with Hume and Mill that belief in observed regularities was just a result of mental habit that had nothing to do with rational acceptance. Such beliefs are instead the foundation of all rationality. However, natural regularities need not be, and should not be confused with, universally invariant laws of nature. There will never be evidence sufficient to prove the existence of such fixed laws. Dewey agreed with Peirce completely on these points:
That there are operations in nature sufficiently stable to be depended upon is testified to by every art and science. But that they are absolutely invariant is both unproved and unprovable, since it is a matter of circumstantial fact where we must be content with an order of probability.2
While it is the function of scientific laws to attempt to match natural regularities, there will never be any perfect match. This is so, not because scientific laws need further refinement, but because natural regularities are more like statistical generalities than exact equations. As Dewey explains,
Universals, definitions, laws, are subject to revision. They are not merely hypothetical in linguistic form (if-then propositions) but what is more important they are, as prescribing operations, working hypotheses, and hence are subject to modification through the consequences to which they give rise. The revision of conceptions and principles is a constant phenomenon of scientific practice.3
Most philosophers of science have gradually come around to the idea that scientific realism should not include a commitment to the existence of exact laws of nature. Laws are useful as idealizations that can be used in logical inferences to propose experimental predictions. But as Ronald Giere points out, expressing a widespread conclusion, “they are neither universal nor necessary – they are not even true.”4 But the denial that exact laws are real is not a denial of the regular processes of nature. Following Giere, laws should be understood as important components of models of natural processes. Models “need only be similar to particular real-world systems in specified respects and to limited degrees of accuracy.”5 This conception of models has long been a feature of pragmatic philosophies of science, going back to Peirce. In the next section, an expanded role for models will be developed that builds on this foundation. Science is properly concerned with understanding the regular processes of objects and events that we observe in nature. Science uses their independent qualities to explain the behavior of these enduring entities as they interact together regardless of whether they are being experienced as well. Science therefore legitimately investigates the conditions of the existence of both perspectival qualities and independent qualities. It is an open question whether it is best to conceive of the enduring characteristics as themselves relational, or to prefer to conceive of some as intrinsic qualities. Some naturalists hold that enduring independence requires intrinsic qualities, but others have suggested that all qualities are ultimately relational. This a pragmatic question best left to the judgment of future scientific development. Our discussion of the scientific investigation of independent entities (again dropping “objects” as too limiting because science investigates other things like events, patterns of events, processes, etc.) should now recognize that there are five basic kinds of independent entities. These kinds are differentiated by the means used in identifying the identity. Three questions must be asked: is the identifying quality perspectival or independent, is the identifying quality experienced or not, and is the identifying quality discerned by a human sense by the aid of an instrument? Because all perspectival qualities are experienceable by definition, the variant combinations of answers to these questions produce five kinds of entities. The kinds are:
Examples of DOPQ: a chemist identifying a mineral by its color, a ornithologist identifying a bird by its song, and a geologist identifying a rock by its texture. Examples of IOPQ: an astronomer identifying a red giant star by its flickering color through a telescope, and submarine sonar operator identifying a surface vessel by its amplified propeller noise. Examples of DOIQ: a paleontologist identifying a fossil bone by its shape, and an oceanographer identifying a tide by the water height. Examples of IOIQ: an official of the bureau of weights and measures using a standard gallon container to identify a full gallon of gasoline from a station pump, and an engineer using a calipers to measure the size of a machine part. Examples of IDIQ: a physicist identifying a metal by calculating its density from its measured volume and weight, and a geologist identifying an iron ore by measuring its magnetic attraction. These are all respectable examples of the role of entity identification in the pragmatic method of hypothesis confirmation, and illustrate pragmatism’s moderate scientific realism. In actual scientific practice, identifications usually proceed by using a combination of two or more of these means. Extreme scientific realism is irrelevant to most of the sciences’ methodologies because they prescribe evidence collection and experimental confirmation in terms of some sort of observable qualities. Unless we were to follow those dualistic philosophers who holds that experience does not include the entities of scientific inquiry but only private inner ideas or sensations or qualia, moderate scientific realism as defined above can suffice to account for most scientific inquiry. Pragmatic naturalism sanctions moderate scientific realism. There is one kind of natural entity which must receive additional scrutiny in this essay: that entity which, due to the scientific conception of that entity, can only be identified by one or more IDIQs, and thus which cannot be identified by any direct or instrumental observation. Examples of such non-observable entities are black holes, quarks, the force of gravity, and the curvature of space-time. Evidence for such entities must always consist of the detection of their effects on scientific instruments. Extreme scientific realism is strictly focused on this kind of entity, and characteristically claims that in order to be a realist, a philosopher must confess that current theories about non-observable entities are completely or at least approximately true. Pragmatism, as we shall see, cannot endorse this extreme scientific realism. We shall directly deal with non-observable entities after some further issues regarding the nature of scientific theories are disentangled. What has been established in this first section is that there is a robust version of pragmatic naturalism which is fully compatible with, and supportive of, all five modes of scientific identification.
The realism of scientific theories
The best place to start is a provisional definition of a pragmatically realistic philosophy of science. The proposed definition is as follows: The proper object of scientific knowledge is the technologically created natural object in human experience. This definition has three components.
This definition holds that the object of scientific knowledge is real, not just ideal. Thus, a pragmatist is a realist about the objects of scientific knowledge and hence should be classified as a scientific realist. From pragmatism’s standpoint, realism is lost if any of pragmatism’s components are denied. Realism can be put in jeopardy by any one of these three claims:
Each of these claims mounts a challenge to pragmatism. However, each claim is an obstacle to a genuinely realistic philosophy of science. An examination of each claim will expose its threat to realism.
Option One: Scientific knowledge aims to describe the world as it really is completely beyond the experience of human knowers. This view can be termed “transcendent realism.” The typical justification for option one attests that the experience of a human knower is limited to the private contents of one’s conscious states, and that knowledge is a matter of correspondence between human ideas and independent reality.
Option Two: Scientific knowledge is of theories only, and not reality. This view can be termed “scientific idealism.” The typical justification for option two argues that genuine scientific knowledge consists largely of propositions with the sort of logical form forbidding them from having a semantic status or referential use.
Option Three: Science aims to understand the world as it proceeds in the absence of any human observation, intervention, or participation. This view can be termed “objectivism.” The typical justification for option three affirms that scientific knowledge aims at understanding the world’s processes or natural laws, which cannot be created, abrogated, or altered by human powers.
Self-proclaimed scientific realists nearly unanimously accept options one and three, “transcendent realism” and “objectivism,” following in the tradition of modern philosophy. This tradition makes a sharp ontological distinction between the “real” features of reality and those features which exist only within human experience. It has been plausibly argued that this philosophical distinction was given new life with the advent of the atomistic materialism of the 1700s. I will not pursue that theme here. But by the late 1800s the proliferation of successful scientific models of unobservable entities, in physics, chemistry, and biology especially, led some philosophers such as Ernst Mach, William James, and Pierre Duhem to question whether science could still aim at correctly discerning the ultimately real features and processes of nature. This questioning took two different forms. Perhaps science’s rightful attempt to correctly describe reality cannot succeed to any significant degree. Alternatively, perhaps science should not attempt to correctly describe reality. Scientific realism thus confronted two rivals: “skepticism” and “antirealism.” The scientific realist not only holds that there really exists an experience-transcendent reality, but also holds that science legitimately aims at, and possesses, some knowledge of it. The skeptic is a transcendent realist who does not question experience-transcendent reality nor science’s efforts to gain knowledge of it, but only doubts whether science actually has gained any knowledge. The antirealist rejects as incoherent the idea that science should attempt to know anything of an experience-transcendent reality. A pragmatic philosophy of science should take the antirealist stance as described here, since the pragmatist holds that transcendent entities are not appropriate objects of scientific knowledge. Pragmatism takes this antirealist stance not because the notion that nature extends farther than experience is incoherent. It is not incoherent because we experience a world of enduring objects and relatively stable processes. Indeed, our experience of nature is of a nature that continually outruns our experiences of it. Nature’s depth and breadth seems all the more inexhaustible as we explore its secrets, both microscopically and astronomically, with an ever-expanding array of instruments and probes. An antirealist philosophy of science should not be confused with metaphysical antirealism’s claim that no experience-transcendent reality exists. An antirealistic philosophy of science is more of a stance taken on the nature of science than on the nature of reality. That means that it is antirealism, and not skepticism, which can possibly gain support from the actual procedures and empirical results of scientific investigation. Skepticism can be countered by scientific realism’s claim that the future progress of science will erode the grounds for skepticism, particularly in the areas of theory unification and theoretical refinement. This is so, because the skeptic is a potential realist about specific entities who is waiting for the evidence to rise to some level pre-set by the skeptic. Scientific realists and skeptics thus get into prolonged debates over how much evidence is needed to show that a sufficient degree of epistemic correspondence has been achieved. An antirealist philosophy of science should not be confused with skepticism, despite the fact that many antirealists start out as skeptics and find themselves taking an antirealist stance as a result. But the reasons for being skeptical about science are not the reasons for being an antirealist. For example, the problem of the underdetermination of theories by experience is only a debatable issue between scientific realists and skeptics because they all agree upon transcendent realism. The antirealist does not construe scientific theories as attempting to correctly describe transcendent reality. Some antirealists take refuge in scientific idealism, dropping the notion that science aims at knowledge of any natural objects. The debate between scientific realism and idealism proceeds over whether science can legitimately seek causal explanations in which something non-observed is understood as causing observed effects. Scientific realists understandably complain that without non-observed causes, patterns of observed events are simply monstrous coincidences. Scientific realists sometimes make this argument against skepticism but it fails, since the skeptic has no difficulty accepting that something non-observed does cause observes effects; the skeptic simply finds insufficient reason to credit any theory with correctly describing the non-observed cause. The skeptic rightly complains in response that the understandable need for some non-observed cause hardly justifies assenting to any specific theory now under consideration. Thus the “inference to the best explanation” argument supporting scientific realism is at best a useful move against idealism but it has no effect on skepticism. Does the “monstrous coincidence” argument have any force against pragmatic antirealism? To answer this question, the pragmatist should make use of an all-important distinction between four types of non-observed things postulated by scientific theories.6 Some things postulated by science are directly observable by human beings using their own natural senses but no one has yet gone to the trouble of observing them. Some things postulated by science are non-observed because while current technology permits their observation through instrumentation assisting the natural senses, such observations have not yet been carried out. Other postulated things have not yet been observed because the technology that could permit their instrumental observation has not yet been invented. Finally, some things postulated by science could never ever be even instrumentally observed even using infinitely powerful technology because they are not the sorts of things which possess any detectable features at all. Let us give these labels to these four sorts of things: “directly observable,” “instrumentally observable,” “hypothetically observable,” and “never observable.” The first three shall be mentioned in the rest of this paper as “observable” in contrast to the never-observable. Pragmatism’s rejection of transcendent realism entails that science should not attempt to correctly describe postulated things that are never observable. Pragmatism has no objection to the first three kinds of observable things, for they all are potentially within human experience, broadly construed. We will discuss shortly why pragmatism should have such an accommodating notion of experience. If scientific realism’s “monstrous coincidence” argument has any relevance against pragmatism, then this move must be formulated as arguing that sometimes a scientific theory must postulate a never-observable thing to adequately explain observable patterns of events. Pragmatism’s antirealism must counter-argue that no scientific theory is ever driven by necessity to undertake that sort of explanation. The counter-argument can only be briefly sketched here. Let us take an example of a scientific theory postulating a never-observable, and let us call such a theory a transcendental theory. First, an analysis of the propositions used by that theory to define the nature of the never-observable will allow the exposure of their logical forms. Second, it is shown that these propositions actually have the same form, that of the “universal” form, which cannot carry any existential meaning. For example, when a physics text declares that electrons have a charge of –1, it does not mean to say that all of the electrons observed so far have turned out to have a charge of –1. Rather, the text is explaining what current physics means by the term “electron”: “If a thing is an electron, then it has a charge of –1.” The meaning of an universal proposition is conserved even if no electrons were to exist.
I take it there is general acknowledgment that a radical difference is found between universal and particular propositions. The latter alone are existential in import, the former being hypothetical or of the “if-then” type.7
Third, pragmatism explains that when a thing is defined only by universal propositions of a theory, that transcendental theory cannot be understood as taking any existential stance towards the thing. This explains its status in the theory as a never-observable. Fourth, pragmatism accounts for the use of universal propositions by pointing out how they are used in scientific inferences to control the processes by which observable things are manipulated. Fifth, pragmatism explains that universal propositions are evaluated and revised as they are tested against the observable predicted events in experimentation. In essence, pragmatism holds that the nature of explanation is a matter of locating those things which in the right circumstances will interact to produce another thing. This is a statement of pragmatism’s productionism. As Dewey explains, a paradigmatic illustration of the functional role of universal propositions in science
is provided by hypothetical universals contrary to fact such as are constantly employed in science, as for example the proposition “If bodies interact without friction, then . . .” or “If a body moves upon impact of one body only without being affected by other bodies, then. . . .” The value of such propositions is proved by their constant use in scientific calculations. Upon any other theory than that of the ultimate connection of hypothetical universals with conduct of observational experimental operations in inquiry, the proved utility of propositions contrary to fact presents an insoluble paradox. The attempt has been made to resolve the paradox by saying that while the propositions in question do not affirm anything of existence, they “ascribe to reality a character which is the ground of the connection stated in the hypothetical judgment.” Regarding this mode of interpretation, it has been pertinently asked “How can there be the ground in the real universal of something which nevertheless does not exist?” The seeming paradox completely disappears when it is seen that such propositions do not intend or purport to have reference to existence but to be relevant to inquiry into existence – a very different matter. There is indeed something of the nature of contrary-to-factness in all definitions. For they are ideal as well as ideational. Like ideals, they are not intended to be themselves realized but are meant to direct our course to realization of potentialities in existent conditions – potentialities which would escape notice were it not for the guidance which an ideal, or a definition, provides.8
Scientific realism could also have this productionist model of explanation, but in practice most scientific realists instead are objectivists who hold that the object of scientific knowledge is unalterable by human control. If productionism were adopted instead, it would be seen that a transcendental theory does not explain because its transcendent propositions pick out those never-observables which produce observables, but instead it explains because its universal propositions are used to predict how some observables can be used to produce another observable. Universal propositions have only an operational meaning, not a referential meaning, and where a theory defines a term only with universal propositions, that theory is not using the term existentially. No proposition can meaningfully refer to transcendent reality. Empiricism traditionally refuses to grant to theories in the natural sciences much ability to make existential reference. Van Fraasen’s constructive empiricism, for example, finds a proliferation of non-existential propositions in chemistry and physics. Empiricism’s antirealism is indeed congenial to pragmatism. However, any empiricism must clearly decide which types of non-observables should be assigned non-existential status. Van Fraassen notoriously refuses to grant existential reference not only to transcendent propositions but also to propositions putatively about things observable through instruments. His constructive empiricism therefore takes an idealistic stance towards all theories excepts those concerned solely with direct observables. The most effective rebuttal to this extremely restrictive empiricism points out that the human sense organs are natural instruments that focus and amplify our interactions with the environment. Selecting out the basic equipment nature provides to humans as the epistemic paradigm of maximum verity only endorses a narrow parochialism that cannot even explain why many people benefit from wearing glasses. Pragmatism understands experience not as something exclusively pertaining to events occurring within passive observers, but rather as a natural process that occurs wherever human beings actively interact with their environment. Technological instruments surely transform the results of our interactions, but only pragmatic considerations should privilege some transformations over others. Dualism was inspired by an excessive privileging of the scientific measurement of primary qualities. However, dualism is not necessarily avoided by the converse privileging of unmediated contact with nature. Naturalistic empiricism, by locating experience in the situational context of organisms interacting with the environment, reconstructs the theory-observation distinction. Since there are no propositions capable of referring to transcendent reality, theories only make reference to observables. Since observables exist within situations of human interaction with nature, a specification of any particular observable is a complex recipe articulating what an observer must do to control a certain kind of environment for the production of the observable. Because theories specify observables and the methods of producing them, the evidence of the observable’s actual production is precisely the evidence relevant for justifying the theory. In other words, a theory receives evidential justification to the degree that its specification of the observable is capable of directing the production of that observable. Theories putatively about never-observables can never receive any evidential support. Unfortunately, discussion of van Fraassen’s constructive empiricism has perpetuated a spurious distinction between epistemic and pragmatic justifications for believing scientific theories. The gratuitous use of “epistemic” in this discussion, apparently a code-word for “truth-relevant,” gathers together anything “pragmatic” as having nothing to do with a theory’s evidential support. The very notion of pragmatic reasons for believing a theory to be true is ruled out as a contradiction in terms. But from the standpoint of pragmatism, there cannot be any reasons, epistemic or pragmatic, to believe that theory about never-observables is correct. And for theories about observables, the epistemic-pragmatic distinction collapses. Pragmatically using a theory to produce the postulated observable is the same act as experimentally finding the best kind of evidence to rationally justify acceptance of that theory.
Pragmatism and technology
Pragmatism does not want to discard the notion of scientific theories as models of reality, but what should be discarded is the notion that theories model simply because they represent. There are two primary meanings of the word “model.” An existing structure can be modeled, and a good model, in the first sense, will copy the original. But before a structure is actually built, a model is designed to prepare for the building. The structure is produced from the model in the second sense that the builders use the model to guide their construction. The test of a model in the first sense is correspondence, because a theory aims at copying its object just as it exists prior to being known. The test of a model in the second sense is productivity. Pragmatism recommends that scientific theories should be understood as models in this second sense, and thus the test of a scientific theory is its ability to reliably and efficiently guide the production of the theory’s objective. Pragmatism thus relinquishes the scientific realist’s objectivism, instead proposing that a theory’s purpose is to direct the application of technology towards producing its object where it had not existed before.
The office of physical science is to discover those properties and relations of things in virtue of which they are capable of being used as instrumentalities; physical science makes claim to disclose not the inner nature of things but only those connections of things with one another that determine outcomes and hence can be used as means.9
Pragmatism owes a detailed account of the relationships between the progress of scientific inquiry and the nature of instrumentally observable and hypothetically observable objects of scientific knowledge. This account will explain why the dividing lines between kinds of objects of knowledge are permeable, since theoretical innovation can bring never-observables into the realm of the hypothetically observable, and technological invention can bring hypothetically observables into the realm of the instrumentally observable. For example, DNA was once the object of a disreputable theory of “germ plasm” responsible for reproductive transmission in the late 1800s that aroused critical scorn for its notorious undetectability. After the germ plasm theory gained respectability and microscope technology improved, the possibility of observation arose and biologists began to theorize upon its chemical structures and features. DNA is now instrumentally observable with advanced electron microscopes. The same sort of scientific development has occurred in atomic theory and cosmology, and in many other sciences. Scientific realists try to use such developments, despite their rarity in the history of science, to perform a meta-induction supporting the existence of the many transcendent entities postulated by present-day science. This meta-induction fails, because the history of science is littered with discarded transcendent theories that far outnumber those few transcendent entities that have been successfully transformed into observables. There can be no meta-abduction across the history of science to establish realism. Nor can scientific realism embarrass pragmatism by those few examples of successful transformation, since the pragmatist, unlike the metaphysical antirealist, affirms the notion of as-yet discovered entities and the legitimacy of science’s search for them. A somewhat different type of “meta-abduction” supporting scientific realism has been offered recently by Jarrett Leplin. The best explanation of the predictions of novel phenomena by successful scientific theories is that their descriptions of non-observable entities is, at least to some degree, representionally accurate. Instrumentalist empiricism (positivism) can only say that successful novel predictions are coincidences. “We thus expect a successful theory to continue to be successful when pressed beyond the phenomena involved in its development, provided that we interpret the theory realistically.”10 Unfortunately for Leplin, he does not consider the possibility that there is a weaker abductive conclusion to be drawn: that successful theories have only managed to capture some aspect of the regular processes of natures. There is no need to interpret theories as attempting to accurately represent non-observable entities. Pragmatism is already committed to the reality of natural regularities, so pragmatism can also account for theoretical success. Pragmatic naturalism, while hostile to realistic interpretations of non-observables, does not descend to the denial of real processes in nature. Pragmatism also owes a detailed account of the nature of observable objects. Pragmatism’s allegiance to a process-oriented ontology has been one of its durable insights, in light of 20th century developments in relativistic quantum mechanics. As Dewey explains in Experience and Nature, “the ultimate objects of science are guided processes of change.”11 Consistent with its rejection of objectivism, pragmatism not only relieves science of attempting to trace natural processes as they happen in the absence of human involvement. Pragmatism also comprehends natural processes as processes happening in the presence of human involvement. Human efforts and natural events are organic wholes. Our conceptions of natural processes are our conceptions of what we creatively effect when we interact with selected portions of the environment. The objects of science are the events that we can to some degree control. Quantum physics, at least in its present state of development, has empirically confirmed that a surrender of objectivism as well as transcendent realism is necessary. The international community of physicists have become increasingly satisfied with experimental results confirming that a relativistic quantum system cannot be described by precise values of physical quantities in the absence of observation upon that system. Rather, any theoretical description of a relativistic quantum system is essentially a prediction of what measurable events would occur under certain experimental conditions. But this validation of pragmatism is not also an endorsement of positivistic instrumentalism. Positivism, like its recent cousin constructive empiricism, goes to far in the regressive direction of limiting experimental evidence to what is directly observable in some vain verificationist quest for infallible data. High-energy physics experiments do indeed result in mountains of data. To find one genuine observation of a subatomic particle, computers must sorts through millions of measured events. But such measurements are hardly the stuff of which epistemological certainty is made, since only one piece of data in a million, if confirmed by many repeated experiments, may actually count as an observation. This implies that the notion of “observation” must be expanded beyond the observer’s own eyes to include the observer’s technological manipulation of incredibly sophisticated machinery and the observer’s theoretical stance on why this machinery has focused on the desired events to be produced. This also implies that the “observer” is actually the wide community of inquirers that assist in the selection of relevant and reliable observations. But this inevitable entanglement of theory and observation does not entail any relativistic conclusion that genuine testing of a theory by observation is impossible. To draw that conclusion, it must further be assumed that the purpose of a theory is to model real processes as they occur in the absence of observation. If objectivism is assumed from the outset, then indeed it becomes hard to understand how a theory that creates its own supporting evidence could claim justification. But when the objectivism that lies at the heart of the realism-relativism debate is exposed, questioned, and replaced with productionism, then the genuine theory-observation relationship can be discovered. The sole purpose of a theory is to direct the production of desired events, so it is no longer a mystery or paradox that theory should produce its own evidence. Indeed, the very best way for a theory to gain rational justification is for it to reliably and repeatedly produce its predicted evidence. Furthermore, theories of greater scope, possessing superior reliability across a wider variety of purposes for the community of inquirers and technology users, are the legitimate aim of progressive scientific inquiry. In summary, pragmatism offers an empirically naturalistic and moderately realistic philosophy of science. Pragmatism’s expansive concept of interactive experience harmoniously complements a naturalism wary of the transcendent. Philosophy of science should embrace pragmatism’s view that the proper object of scientific knowledge is the technologically created natural object in human experience. Pragmatism offers the most realistic interpretation of the growth of scientific knowledge that is consistent with both actual scientific practice and experimentally confirmed results.
Notes
1. The Collected Papers of Charles Peirce, eds. Charles Hartshorne, Paul Weiss, and Arthur W. Burks (Cambridge: Harvard University Press, 1931-1958), vol. 5, paragraphs 94, 100, 101. 2. The Later Works of John Dewey, vol. 11, ed. Jo Ann Boydston (Carbondale: Southern Illinois University Press, 1987), p. 112. Henceforth references to The Later Works will be cited by LW followed by volume and page numbers. 3. Ibid., pp. 110-111. 4. Ronald Giere, Science Without Laws (Chicago: University of Chicago Press, 1999), p. 90. 5. Ibid., p. 93. 6. This distinction is a version of Rom Harré’s account of scientific investigation in Varieties of Realism: A Rationale for the Natural Sciences (Oxford: Blackwell, 1987). 7. LW 5: 197. 8. Logic: The Theory of Inquiry, LW 12: 302-303. 9. Experience and Nature, LW 1: 6. 10. Jarrett Leplin, A Novel Defense of Scientific Realism (Oxford: Oxford University Press, 1997), pp. 104-105. 11. Experience and Nature, LW 1: 128.
First published as "A Pragmatically Realistic Philosophy of Science" in Pragmatic Naturalism and Realism (Amherst, N.Y.: Prometheus Books, 2003). Copyright 2003 by John R. Shook |