Behavioral Pragmatism: Making A Place for Reality and Truth

Abstract

In “Behavioral Pragmatism: No Place for Reality and Truth,” Barnes-Holmes (2000) proposed a behavior-analytic version of philosophical pragmatism he called behavioral pragmatism (BP), a perspective which challenges two bedrock tenets of Western culture: (1) metaphysical realism, the view that an external, physical reality exists which is mind-independent and (2) the correspondence theory of truth (CTT), a theory which maintains that true statements are those which correspond to mind-independent reality. Many (perhaps most) behavior analysts accept both of these tenets (though they typically name and describe these tenets using different terms). By contrast, in lieu of the first, BP offers, as a replacement tenet, nonrealism, in place of the second, the pragmatic truth criterion. The account of reality and truth of BP has gained increasing prominence within behavior analysis because of its inclusion within relational frame theory, a perspective with a growing number of adherents. In this paper, I first argue that the realism/pragmatism dispute needs to be resolved because it threatens the coherence of radical behaviorism as a philosophy of science. Next, I present a detailed account of the differing conceptions of reality and truth as articulated within: (1) metaphysical realism, (2) behavioral pragmatism, and (3) Richard Rorty’s version of pragmatism (Rortian pragmatism). Finally, using the insights of Rortian pragmatism (RP), I offer three proposals for modifying the core tenets of behavioral pragmatism. If adopted, these proposals would help narrow the realism/pragmatism divide, thereby reducing the threat to radical behaviorism’s coherence.

Concisely put, metaphysics is “the study of ultimate reality” (van Inwagen 2009, p. 1). Among Western philosophers, a prominent, popular, and enduring approach to metaphysics has been metaphysical realism (a term popularized by Putnam 1978, 1981) which asserts that the ultimate reality is a mind-independent reality (Chakravartty 2011; Psillos 1999; Putnam 1981). In this context, the term mind is not being used to denote an immaterial substance (more about this monistic usage in due course). For metaphysical realists (realists, for short), mind-independent reality is an external, physical reality which exists whether or not anyone is present to perceive, think, or talk about it or otherwise interact with it (Devitt 2010). Furthermore, realists assert that mind-independent reality has a determinate structure—a structure which it possesses independent of our human sensory/perceptual systems, taxonomic practices, and conceptual schemes (Rorty 1999). Finally, as a crucial corollary to realism, its proponents have typically endorsed the correspondence theory of truth which maintains that truth bearers (i.e., true statements or beliefs) correspond to mind-independent reality (Devitt 2010).

Was Skinner an adherent of realism and the correspondence theory of truth? A thorough answer to this question is beyond the scope of this paper. However, examining some of the evidence relevant to answering this question helps set the stage for my later discussion of the realism vs. pragmatism debate within behavior analysis. Consider first the issue of realism. Employing Reese’s (1993) exegesis of the metaphysical assumptions of mainstream behavior analysis, Barnes and Roche (1994) concluded that “most behavior analysts assume that there exists a real, physical, and ordered universe” (p. 165; emphasis added). In effect, they concluded that most behavior analysts are realists. If accurate, this should not be surprising, given that Skinner—in his explanations of behavior—often differentiated between the behaving organism on the one hand and a behavior-independent, external world on the other. For example, Skinner (1953) stated that “our ‘perception’ of the world—our ‘knowledge’ of it—is our behavior with respect to the world. It is not to be confused with the world itself” (p. 140; emphasis in original). Similarly, according to Skinner (1974), our behavior is determined by “the environment, past or present, which … lies outside the behaving person” (p. 144). These and other passages support the claim that Skinner assumed the existence of a reality independent of us.

Barnes and Roche (1994) acknowledged that Skinner, in some of his writings, “clearly suggests that the world exists in parts ‘about us’ (i.e., independently of us)” (p. 165). However, they also reported that in other passages, he appears to contradict this realist stance. For example, while Skinner’s apparent realist sentiments were on full display when he explained that the task of science is to discover “the laws which govern a part of the world about us” (Skinner 1953), he then contradicted himself (according to Barnes and Roche) by suggesting that “scientific laws … are not obeyed by nature but by men who deal effectively with nature” (p. 166)—a position not consonant with realism. In offering an additional argument against portraying Skinner as a realist, Barnes and Roche began by noting that Skinner characterized stimulus and response classes as inseparable, co-defining classes. They then concluded that, as a result of this inseparability, the objects/events serving as stimuli “are known or defined in terms of behavioral function, rather than as physical things that exist independently of behavior” (Barnes and Roche 1997, p. 545). Further elaborating,

if we talk of a real, physical universe, we are saying that stimuli have some form of existence beyond our behavior; this clearly contradicts behavior-analytic epistemology in which there can be no stimuli (i.e., a physical universe) if there is no organism to provide responses that define those stimuli (Barnes and Roche 1994, p. 166; emphasis in original).

By so arguing, Barnes and Roche easily called into question any portrayal of Skinner as a strict adherent of realism (for a critique and a counter-critique of Barnes and Roche’s view, see Tonneau 2005 and Barnes-Holmes 2005).

Next, consider Skinner’s views on the topic of truth. Zuriff (1980) identified two competing interpretations. Under one interpretation, many passages can be culled from Skinner’s writings which “amount to a modern version of the correspondence theory of truth” (Zuriff 1980, p. 343). For instance, Skinner (1957) explained that we call a verbal response true “when the correspondence with a stimulating situation is sharply maintained” (p. 147). Obversely, we say that a verbal response is false—e.g., in perjured testimony—when there is a “lack of customary correspondence between a verbal response and certain factual circumstances” (p. 339). Under a second interpretation, Skinner’s writings provide other passages in which he appears instead to endorse “a behaviorist version of the pragmatic theory of truth” in which “a statement is true if it ‘works’” (Zuriff 1980, p. 344). For example, Skinner (1974) stated that “a proposition is ‘true’ to the extent that with its help the listener responds effectively to the situation” (p. 235). Similarly, with respect to any empirical claim, “there is a special sense in which it could be ‘true’ if it yields the most effective action possible” (Skinner 1974, p. 235). However, despite the evidentiary support for both interpretations, Zuriff opted for interpreting Skinner as endorsing a version of the pragmatic theory of truth because “in one form or another this pragmatic theme is woven throughout Skinner’s comments on knowledge,” thereby comprising the theory “more prominent in his work” (p. 344).

As previously noted herein, while Barnes and Roche raised doubts about Skinner’s supposed allegiance to realism, they nonetheless believed that most behavior analysts were realist in orientation. By contrast, Barnes-Holmes (nee Barnes) and his colleagues have opted for a different behavior-analytic account of reality dubbed behavioral pragmatism (Barnes-Holmes 2000). For the behavioral pragmatists—and for other pragmatist behavior analysts as well—the objects which populate the world “are defined or known as behavioral functions, instead of physical things that exist independently of behavior” (Barnes-Holmes 2000, p. 197). Behavioral pragmatists neither accept nor reject the existence of mind-independent reality, opting instead for a position dubbed nonrealism (Barnes-Holmes 2003). Regarding the topic of truth, behavioral pragmatists avoid the apparent inconsistency in Skinner’s treatment of truth by rejecting all versions of correspondence-based truth, opting instead for the pragmatic truth criterion in which “successful working” serves as their criterion of truth (Barnes and Roche 1997, p. 555; Barnes-Holmes 2000, p. 198).

These two opposing views on reality and truth are currently amply represented within the behavior-analytic community and constitute a realism/pragmatism divide. In this paper, I first argue why overcoming the realism/pragmatism divide should matter to behavior analysts. Next, I present a detailed account of the differing conceptions of reality and truth as articulated within: (1) metaphysical realism, (2) behavioral pragmatism, and (3) Richard Rorty’s version of pragmatism (Rortian pragmatism). Finally, in the “Conclusion” section, using the insights of Rortian pragmatism, I offer three proposals for modifying the core tenets of behavioral pragmatism. I argue that the adoption of these proposals by behavioral pragmatists (and by other pragmatist behavior analysts as well) would help settle the realism/pragmatism schism within behavior analysis.

Why It Matters

The discipline of behavior analysis consists of three major branches: (1) basic science, (2) applied behavior analysis, and (3) the theoretical/philosophical/conceptual (TPC) underpinnings of the science. Marr (2013) observed that, of these three branches, discussions targeting the TPC precepts “evince, by far, the greatest internal dissension” (p. 195) within behavior analysis. A prominent example of this dissension has been the realism vs. pragmatism debate among behavior analysts. As Hackenberg (2009) reported, one of the “internecine struggles” has been “the debate over whether behavior analysis is a form of realism or pragmatism” (p. 401). As an exemplar of the pragmatist side of the debate, behavioral pragmatism deserves critical examination because such an exegesis may help clarify—and assist in resolving—this intramural conflict. Here, the eventual goal can be the achievement of a broad consensus among behavior analysts on the nature of reality and truth.

As further evidence of the realism vs. pragmatism debate within behavior analysis, consider the conflicting views of four prominent behavior analysts: J. E. R. Staddon (1993), J. Burgos (2004), W. Baum (1994), and S. C. Hayes (Hayes et al. 2001). The first two—Staddon and Burgos—assume the existence of an independent, physical reality. For instance, Staddon (1993) maintained that “the evolution of science has always been sustained by a faith—yes, faith, it cannot be proved—that there is a single, unchanging, underlying reality, imperfect as our apprehension of it may be” (p. 246; emphasis in original). Similarly, Burgos (2004) advocated a “realism about behavior” (p. 72). For Burgos, behavior “exists objectively, where ‘objectively’ typically means ‘independently of the mind’” (p. 73).

On the other hand, Baum and Hayes largely share the behavioral pragmatist’s view on reality. For instance, Baum (1994) professed “no commitment to any idea of real behavior,” denying the realist view that “there is some real behavior that goes on in the real world” (p. 26). Similarly, in his role as a principal architect of relational frame theory (RFT), S. C. Hayes (Hayes et al. 2001) proposed a theoretical account of language and cognition which allegedly precludes “ontological assumptions” (p. 34). For instance, within RFT, operants are not assumed to be real, but rather “merely useful constructions” (p. 23). RFT also shares behavioral pragmatism’s endorsement of successful working, rather than correspondence with reality, as its truth criterion (Hayes and Long 2013). RFT has been successful not only in winning over an increasing number of adherents (Wilson, Whiteman, and Bordieri 2013) but also in producing a burgeoning research program (Dymond, May, Munnelly, and Hoon 2010). Thus, RFT represents what is perhaps the most prominent and influential exemplar of the pragmatist side of the debate within behavior analysis.

What challenges does this realism vs. pragmatism schism present to the philosophy of behavior analysis, radical behaviorism? I argue that chief among the challenges it presents is that of threatening radical behaviorism’s coherence as a philosophy of science. Chiesa (1994) characterized radical behaviorism as “perhaps the most coherent philosophy of science in psychology today” (p. 7). Others (e.g., Buskist and Critchfield 1994; Hillix and Marx 1974) have offered similar appraisals of radical behaviorism. By contrast, Chiesa noted that some disciplines within psychology demonstrate their lack of coherence by failing to achieve consensus on what constitutes the very subject matter of their respective disciplines. Using Chiesa’s yardstick, the ongoing realism vs. pragmatism divide suggests that radical behaviorism currently faces challenges to its coherence because it too has failed to achieve consensus on whether or not its subject matter includes a reality which exists independently of us. Are operant behaviors real (as Burgos contends) or merely useful constructions (as Hayes contends)? Critically examining behavioral pragmatism—and, in particular, its refusal to posit an independent reality—can help clarify the realism/pragmatism divide, thereby helping to eventually bridge the schism. Importantly, helping bridge that schism reduces the threat which it presents to the coherence of radical behaviorism.

Metaphysical Realism

Reality

Most Westerners share a firmly held belief in metaphysical realism, a belief which van Inwagen (2009) dubbed the “Common Western Metaphysic” (p. 25). According to this preeminent metaphysical belief, an external material world exists—a world populated by real objects which “exist independently” of a person’s “beliefs and anything else present in her mind” (p. 24). In describing what he alternatively called the “Standard Metaphysical Picture,” Ebbs (1997, p. 203) offered essentially the same explication of this well-entrenched Western view. Traditional developmental psychology has chronicled the normal occurrence of this belief in its account of the emergence of object permanence among typically developing toddlers (Shaffer 1989). Identified as a developmental milestone, object permanence is reportedly achieved by a person when an object “is conceived to exist independently of a person’s perception of it” (Ginsburg and Opper 1988, p. 41; for a behavior-analytic account, see Schlinger 1995, pp. 142–148). The doctrine of metaphysical realism provides a formal, theoretical account of this mainstream view. Specifically, metaphysical realism (realism, for short) makes two major claims about mind-independent reality: (1) an existence claim and (2) an independence claim (Devitt 1991, 2010; Brock and Mares 2007). I shall discuss each claim in turn.

The Existence Claim

The existence claim asserts that an external, physical reality exists—a reality populated by “entities that our best science and common sense say exist” (Devitt 2010, p. 50). For example, this claim commits the realist to the existence of commonplace objects like rocks and trees, as well as entities (e.g., atoms, molecules, muons) which science, at its best, says exist (Devitt 1991). For the realist, “it is not just that our experiences are as if there are cats, there are cats. It is not just that the observable world is as if there are atoms, there are atoms” (Devitt 2010, p. 33; emphasis in original).

The Independence Claim

This claim asserts that the existing objects of reality are mind-independent. Given that this claim employs the term mind, a clarification is needed. Within contemporary Western philosophical discourse, the expression mind-independent does not usually connote mind/matter dualism. For most present-day Western philosophers, the mind does not typically possess “any ontological reality as an entity or substance” (Angeles 1992, p. 187). Rather, “you have a mind if you think, perceive, or feel” (Morton 2005, p. 603). Nevertheless, to avoid any lingering connotations of dualism, in place of the expression mind-independent reality, hereafter in this paper I shall instead use the expression absolute reality.

For realists, absolute reality is independent in two senses of the term. In the first sense—which I dub autonomous independence—absolute reality is independent in that it exists whether or not anybody is present to perceive (or otherwise interact) with it. In the second sense—which I dub intrinsic independence—absolute reality is independent in the sense that it is “as it is independently of how humans take it to be” (Khlentzos 2011, p. 1; emphasis added). As Hilary Putnam explained, realists consider reality as having intrinsic independence because it is “as it is in itself, independent of perspective” (Putnam 1992, p. x) and “determinate … independent of conceptual scheme” (Putnam 1983, p . viii). More specifically, absolute reality has intrinsic independence in the sense that it contains objects traditionally identified as natural kind objects. Bird and Tobin (2012) explained that “to say that a kind is natural is to say that it corresponds to a grouping or ordering that does not depend on humans” (p. 1; emphasis in original).

Natural kind objects are typically characterized as having intrinsic features (Bird and Tobin 2012). In this context, having intrinsic features means that any natural kind object (e.g., gold, bananas, tigers) has features which “make it what it is, regardless of its surroundings or its relation to other things” (Blackburn 2008, p. 189). Thus, absolute reality has intrinsic independence in the sense that each of its natural kind objects has specific properties which differentiates it from the others—properties which have existed prior to, and independent of, the human sensory/perceptual systems and taxonomic/conceptual schemes. In short, the defining attributes of natural kind objects are “set” by nature, not by us. Of course, natural kind objects are also said to have extrinsic features. For example, an intrinsic feature of a piece of gold is reportedly its atomic number 79, while its extrinsic features include its weight, shape, size, place of origin, and present location (the intrinsic/extrinsic distinction is not without controversy; for a review, see Weatherson and Marshall 2013).

Truth

I previously reported that a critical corollary to realism has typically been the correspondence theory of truth (CTT) which holds that statements or beliefs are true when they correspond to absolute reality. For proponents of CTT, such statements or beliefs provide us with absolute truth. CTT is an exemplar of what I dub truth correspondism, the broadly defined view that truth bearers (i.e., true statements/beliefs) correspond to reality (the term reality being variously defined, depending on the version of truth correspondism). Truth correspondism has dominated within Western culture, offering a mainstream explanation of how truth is achieved. For instance, in van Inwagen’s (2009) account of the Common Western Metaphysic, truth bearers “correspond to external reality, just as it is up to a map to correspond to the territory” (p. 24). The Standard Metaphysical Picture (Ebbs 1997) offers essentially the same account of this prevailing approach to truth. Notwithstanding the various versions of truth correspondism, when realist philosophers affirm their commitment to truth-as-correspondence, CTT is usually what they have in mind.

Bertrand Russell (1912) described CTT as the “commonest” view on truth among philosophers. More recently, CTT was called the “most venerable” (Kirkham 1992, p. 119) and “best known” (Lowe 1995, p. 881) theory of truth. Similarly, Devitt (2010) reported that “the most popular theory of truth has probably been the correspondence theory” (p. 155). While CTT has been challenged numerous times by its many critics (e.g., Kuhn 1962; Putnam 1981; Rorty 1979, 1999), modern versions of this doctrine continue to be proposed by well-known contemporary philosophers (e.g., Davidson 2001; Devitt 2010). Just as the doctrine of truth correspondism encompasses a number of different versions, so too CTT is not a single theory but, rather, a family of theories. Each member of the CTT family typically offers its own explicit, formal, theoretical account of the layperson’s commonsense view of truth-as-correspondence. Therefore, I begin my discussion of CTT by first examining the layperson’s account of truth. This prototypical version of truth correspondism I dub folk correspondence (a slightly altered version of an expression used by Prado 1987). As Prado explained folk correspondence, “everyone understands that sentences are true if they accurately capture ‘the facts’ and take truth-as-correspondence to be unquestionable” (1987, p. 8). Moreover, on this view, these facts are facts about absolute reality. Folk correspondence exemplifies folk theory, the latter being defined as “a loose network of largely tacit principles, platitudes, and paradigms” (Stich 1983, p. 1) which—in the case of folk correspondence—determine how terms like truth are used by the lay public.

Like folk correspondence, CTT begins with the “simple observation” that determining whether a statement about the world is true “surely must depend on how the world is” (Rundle 1995, p. 166). However, unlike folk correspondence, CTT typically provides a specific account of the nature of the “special relatedness” (Prado 1987, p. 9) of correspondence which holds between absolute reality and truth bearers, or as Putnam (1990) explained it, between “items independent of us and items in language” (p. 172). Kirkham (1992) distinguished between two principal, competing approaches taken by CTT proponents when explaining the nature of the special relatedness of correspondence: (1) correspondence-as-congruence (e.g., Russell 1912) and (2) correspondence-as-correlation (e.g., Austin 1970). The correspondence-as-congruence approach argues for a “structural isomorphism” (Kirkham 1992, p. 119) between a truth bearer and the fact to which it corresponds. On this view, the structure of a truth bearer “mirrors or pictures the structure of facts much in the way in which a map mirrors the structure of that portion of the world of which it is a map” (p. 119; emphasis added).

On the other hand, in its portrayal of the correspondence relation, proponents of correspondence-as-correlation deny that “the truth bearer mirrors, pictures, or is in any sense structurally isomorphic with the state of affairs to which it is correlated” (Kirkham 1992, p. 119). Instead, according to correspondence-as-correlation, “a truth bearer as a whole is correlated to a state of affairs as a whole” (p. 119; emphasis in original). Historically, the correspondence-as-congruence approach has been the dominant version—the version of CTT most often subjected to criticism and counter-theorizing by pragmatists (e.g., Rorty 1979) who take a different view. Therefore, given its dominance as a target of pragmatist critiques, I will further explicate correspondence-as-congruence. During the modern era in philosophy, Russell (1912) offered one of the earliest and best known versions of this approach. Consider the statement the cat is on the mat. Using Russell’s (1912) analysis, this statement primarily consists of a three-term relation: two object terms (cat and mat) and an object-relation term (on). In addition to these three terms, the statement has (speaking metaphorically) an “order” or “direction” which is indicated “by means of the order of words in the sentence” (p. 126). Changing the order—e.g., the mat is on the cat—yields a different statement which, if true, has a different corresponding fact.

Behavioral Pragmatism

Reality

Earlier I reported that realism makes two major claims about absolute reality: an existence claim and an independence claim. The first claim asserts that an external, physical reality exists, while the second claim asserts that this reality has both autonomous and intrinsic independence. Behavioral pragmatism (BP) may also be explicated by examining any claims it makes (or does not to make) regarding the existence and independence of an absolute reality. With respect to the existence claim, the proponents of BP are noncommittal, refusing to either accept or reject this claim. They are “silent” (p. 68) on the issue because they consider it “irrelevant” (p. 72) to their pursuits (Barnes-Holmes 2005). Consequently, BP also remains silent on the issue of whether or not absolute reality is independent. Consider the treatment of BP of the existence and independence claims in more detail.

The Existence Claim

BP assumes nothing “fundamental, final, or absolute” about “the nature or substance” of absolute reality (Barnes-Holmes 2005, p. 68). Simply stated, the existence of absolute reality is “neither affirmed nor denied” (p. 70). However, as the proponents of BP have explained, this does not mean that BP is antirealist. Rather, behavioral pragmatists profess a neutral position with respect to the existence or nonexistence of absolute reality—or, as they prefer to call it—“ontological reality” (p. 68). According to the proponents of BP, their position is neither realist nor antirealist. Put another way, “the technical terms of behavior analysis are simply empty with respect to ontological reality, and thus neither realism nor antirealism is implied” (p. 74). Rather, behavioral pragmatists consider themselves to be adherents of nonrealism (Barnes-Holmes 2003). To further clarify the agnosticism of BP with respect to the possible existence of absolute reality, I now turn to the issue of independence.

The Independence Claim

As a result of their agnosticism about the existence of an absolute reality, behavioral pragmatists must (to avoid contradiction) also maintain an agnosticism about the possible independence of absolute reality. Since they neither affirm nor deny the existence of absolute reality, they can neither affirm nor deny its independence (in either its autonomous or intrinsic sense). For instance, Barnes-Holmes (2005) stated that “the a-ontological position of behavioral pragmatism argues neither for nor against an independent reality” (p. 72). Consider, for example, an apple. “In common sense terms, the apple is a physical thing that exists independently of behavior. For the behavioral pragmatist, however, the apple is defined only in terms of its behavioral functions” (Barnes-Holmes 2000, p. 197).

Truth

While realists typically favor some version of the correspondence theory of truth (CTT), behavioral pragmatists do not. In arguing against CTT, Barnes-Holmes (2000) utilized a version of a famous argument offered by philosopher Hilary Putnam (1981). In his “God’s eye point of view” argument, Putnam observed that realists—by embracing the correspondence theory of truth (CTT)—demonstrate that their “favorite point of view is a God’s eye point of view” (p. 49). Taking the God’s eye point of view means that we, as truth seekers, have to “‘stand outside’ and compare our thought and language with the world” (Putnam 1994, p. 297). In other words, CTT requires that we access absolute reality in a manner free of the limitations of our human sensory/perceptual systems and taxonomic/conceptual schemes. Putnam protested that we are not able to do this. With CTT, “to single out a correspondence between two domains one needs some independent access to both domains” (Putnam 1987, p. 43; emphasis added). While we have access to the domain of our language “we have no access to ‘unconceptualized reality’”(Putnam 1994, p. 297). Whenever we try to access absolute reality, we necessarily employ our anthropocentric perspectives. Therefore, Putnam concluded, CTT fails as an account of truth.

In offering their own version of Putnam’s argument, Barnes-Holmes and colleagues asserted that the goal of scientific inquiry cannot be the attainment of an increasingly accurate description of absolute reality. Why? Because “no special point of vantage is available” (Barnes-Holmes 2000, p. 198) to us, so we can never determine the accuracy of our descriptions of absolute reality by comparing them with absolute reality itself. Hence, like Putnam, BP rejects CTT. Barnes and Roche (1994) argued that “if we assume that direct access to reality is impossible, then we have no objective reality with which we can establish the truth of our inferences” (p. 166; emphasis in original). Therefore, “it is contradictory to argue that our inferences are … inferences about the external world” (p. 166). BP resolved this contradiction by simply not assuming the existence of absolute reality. Without this assumption, the traditional CTT cannot be employed.

For example, consider the case of a behavioral researcher who states that a particular output from a cumulative record of a pigeon’s behavior displays a “scallop” pattern. According to BP, the output pattern does not represent “what the … pigeon ‘really’ did” (Barnes-Holmes 2000, p.198). In other words, the output pattern—as well as any statement ascribing that pattern to the pigeon’s behavior—does not (and cannot) correspond to absolute reality. “Instead, the pattern may be defined as a discriminative stimulus for a particular ‘scientific’ response, such as ‘scallop’ or’break-and-run,’ that has been differentially reinforced in the presence of that pattern” (p. 198). As traditionally defined, absolute truth, is “immutable, absolute, and final” (p. 198). According to BP, the researcher’s verbal response cannot provide us with the absolute truth because of the contingent nature of her response. The researcher’s response is the product of her behavioral history—“a different or more extended history may have produced a different truth” (p. 198)— and therefore can never be “immutable, absolute, and final” as absolute truth requires.

For the behavioral pragmatist, the failure of a researcher (or anyone else) to achieve absolute truth is not a problem. In place of CTT, the behavioral pragmatist instead offers the pragmatic truth criterion in which “successful working”—rather than correspondence with absolute reality—serves as the criterion of truth (Barnes and Roche 1997, p. 555; Barnes-Holmes 2000, p. 198). “The behavioral pragmatist always appeals to utility, and never correspondence, as a truth criterion” (Barnes-Holmes 2000, p. 202). Put another way, for the behavioral pragmatist, the truth of a statement is “defined in terms of its usefulness in achieving particular goals” (Barnes and Roche 1997, p. 543; emphasis added). Providing additional clarification of the position of BP, Barnes-Holmes (2000) stated that “If a scientific statement is useful in helping the behavioral pragmatist to achieve the goals of prediction and control with some degree of scope and precision, then the statement is considered true” (p. 198). As a consequence, “the issue of correspondence is, therefore, simply irrelevant” (Barnes and Roche 1997, pp. 548–549).

Rortian Pragmatism

Introduction

Having originated during the end of the nineteenth century, pragmatism lost much of its currency by the middle of the twentieth. The subsequent publication of Richard Rorty’s (1979) impactful Philosophy and the Mirror of Nature has been frequently credited (e.g., Kogler 2005) with launching pragmatism’s eventual resurgence. In his book, as well as numerous other publications spanning decades, Rorty championed a pragmatist account (albeit his own version) of reality and truth. Indeed, in large measure because of his pragmatist views, Rorty has drawn the attention of many within the behavior-analytic community (e.g., Lamal 1983; Leigland 1999, 2003; Malone 2004; Schoneberger 2002, 2003, 2006). Following Rorty’s death in 2007, Richard Posner eulogized him as the one who “single-handedly revived pragmatism” (quoted in Metcalf 2007, p.1), an appraisal shared by others (e.g., Margolis 2002).

Of course, Rorty has also had his critics. For example, he has been described by some as “the bad boy of American philosophy” (Ree 1998, p. 7) because—by becoming a pragmatist—he abandoned his prior, decades long commitment to analytic philosophy. With roots in ancient Greek philosophy, analytic philosophy emerged at the beginning of the twentieth century. While there is no consensus on how best to define it, nearly all commentators would agree that analytic philosophers share the common goal of “articulating the meaning of certain concepts, such as ‘knowledge,’ ‘belief,’ ‘truth,’ and ‘justification’” (Stroll 2000, pp. 7–8) by utilizing the analytical tools specific to their philosophical school (e.g., by analyzing the “logical geography” of concepts; see Ryle 1949). As a result of becoming a pragmatist, Rorty renounced his allegiance to analytic philosophy, earning the aforementioned reputation as American philosophy’s “bad boy.”

While not initially trained as an analytic philosopher, in the early1960s, Rorty began practicing philosophy in that tradition. As he later explained in an interview, “analytic philosophy was taking over” so “I retooled myself so as to become an analytic philosopher” (Rorty et al. 2002, p. 52). He subsequently made a number of important contributions (e.g., Rorty 1965, 1970a, b) to the philosophy of mind from the perspective of analytic philosophy. However, with the 1979 publication of his Philosophy and the Mirror of Nature, Rorty publicly signaled his break with analytic philosophy. Specifically, in that book, Rorty (1979) critiqued analytic philosophy’s representationalist theories of knowledge—theories in which “knowledge is conceived of as accurate representing … as the Mirror of Nature” (p. 170). Rorty’s observed that “analytic philosophy … is marked principally by thinking of representation as linguistic … and of philosophy of language … as the discipline which exhibits the ‘foundations of knowledge’” (Rorty 1979, p. 8). This view conceives of the language/reality relation as a “relation between a medium of representation and what is purportedly represented” (Rorty 1992, p. 371). In his book, Rorty (1979) provided a sustained set of arguments against representationalism. Broadly speaking, his strategy was to question the “frame of reference” (1979, p. 7) which resulted in these representationalist accounts of knowledge. More ambitiously, his arguments were meant to “put us in a position to criticize the very notion of ‘analytic philosophy’” (p. 8).

A self-described “syncretist” (Rorty 1998, p. 10), Rorty employed a diverse set of American and European sources—performing what Malachowski (2002) called “acts of appropriation” (p. 67)—as he abandoned analytic philosophy and honed his own version of pragmatism. Among Rorty’s American influences were the leading classical philosophical pragmatists (i.e., Dewey, James, and to a lesser degree, Peirce) as well as these more recent philosophers: Brandom, Davidson, Goodman, Kuhn, Putnam, Quine, and Sellars. His European influences included Darwin, Nietzsche, Wittgenstein, Sartre, Gadamer, Heidegger, Derrida, and Foucault. Out of these myriad influences emerged Rortian pragmatism (RP). In an effort to distinguish his pragmatist views on reality from the realist’s position, Rorty drew a distinction between two accounts of reality: absolute vs. contingent reality (Rorty himself did not use these terms to make this distinction). As I earlier reported, realists champion absolute reality. By contrast, as part of his version of pragmatism, Rorty championed contingent reality (why contingent shall be explained in due course). With respect to truth, Rorty distinguished between two accounts: absolute vs. contingent truth (again, these are not Rorty’s terms). Realists have typically endorsed absolute truth, while Rorty endorsed contingent truth. Rorty’s concepts of contingent reality and contingent truth shall now be discussed in detail.

Reality

“I am always trying to kick the habit of the realism issue. There is something a little shameful about spending one’s time defending something so apparently humdrum as the independent existence of the familiar world. But the provocations are so great, and my flesh is weak” (Devitt 1991, p. vii).

Though Rorty disagreed with Devitt about the nature of reality, like Devitt, he nonetheless assumed that an external, physical reality exists. Specifically, he assumed the existence of contingent reality. Like realism’s conception of absolute reality, the contingent reality of RP can also be characterized as making two claims: (a) an existence claim and (b) an independence claim. I shall examine both claims, beginning with contingent reality’s existence claim (why I characterize the conception of reality of RP as contingent shall be explained when I discuss the independence claim).

The Existence Claim

Like realism’s concept of absolute reality, the contingent reality of RP posits the material existence of both commonplace objects (e.g., trees; Rorty 2007, p. 106) and scientific objects (e.g., neutrinos; Rorty 1998, p. 87). Within RP, the term reality serves as “a name for the aggregate of all such things” (Rorty 2007, p. 106; emphasis in original). Rorty (1991a) observed that the existence of material reality often manifests itself as “brute physical resistance” (p. 81). For example, Rorty cited the physical resistance which Dr. Johnson’s boot famously encountered when kicking a rock in an effort to refute Bishop Berkeley’s immaterialism. According to Rorty, kicking a rock puts you in direct contact with reality—but so does engaging in verbal behavior about reality. Both kicking rocks and talking about reality—as well as countless other ways of interacting with reality—are “as direct as contact with reality can get” (Rorty 1991a, pp. 145–146). The foregoing examples (trees, rocks, neutrinos) of Rorty’s assumption of an extant external, material reality are not anomalies, but rather represent a central tenet of RP. As Putnam (2000) observed, “virtually all of Rorty’s writings contain passages intended to reassure us that he is not denying that there is a world” (p. 81). However, as I will discuss next, while RP agrees with realism that an external, physical world exists, RP soon parts company with realism with respect to the latter’s conception of the independence of reality. I intend to show that the differing conception of reality’s independence of RP explains why I have termed the conception of reality of RP contingent.

The Independence Claim

To review, realism asserts that absolute reality is independent in two senses of the term. In the first, autonomous sense, being independent means that reality exists whether or not anyone is present to perceive (or otherwise interact with) it. In the second, intrinsic sense, being independent means that natural kind objects populate reality—objects which are distinguishable from each other by virtue of their intrinsic properties. In claiming that absolute reality has an intrinsic independence, realists assume a fundamental intrinsic/extrinsic distinction. As Rorty described this distinction, for the realist, the intrinsic features of objects are “absolute, non-description relative” features (Rorty 1998, p. 2), while the extrinsic features of objects are features they have “merely in relation to, e.g., human desires and interests” (Rorty 1991b, p. 130). According to realism, we successfully identify natural kind objects when our descriptions (e.g., our taxonomies and conceptual schemes) carve nature at its intrinsic joints.

RP agrees with realism that an external, physical reality is independent in the first, autonomous sense. As Rorty put it, “most of the world is as it is whatever we think about it” (1982, p. xxvi) and “would be as it is had we never existed” (Rorty 2000b, p. 264). For example, “one of the obvious truths about mountains is that they were here before we talked about them” (1998, p. 72). In a like manner, “if there had been no human beings there would still have been giraffes” (1999, p. xxvi). However, unlike realism, RP rejects independence in the second, intrinsic sense. For Rorty, “there is no such thing as an intrinsic, nonrelational property” (1999, p. 135). Within RP, all of an object’s features are considered “description-relative features” (Rorty 1998, p. 85). Thus, Rorty (1998) advocated that we “discard” (p. 85) the distinction between intrinsic vs. extrinsic properties.

Consider now Rorty’s primary argument against intrinsicality, couched in terms of a response to John Searle, a philosopher who has been a prominent, contemporary proponent of the intrinsic/extrinsic distinction. In his explication of the distinction, Searle (1992) asserted that “the expressions ‘mass, ‘gravitational attraction,’ and ‘molecule’ name features of the world that are intrinsic” (p. 211). On the other hand, “expressions such as ‘nice day for a picnic’” identify features of the world that have been “assigned”—i.e., extrinsic “observer relative” features (p. 211) of the world. In response, Rorty argued that no such distinction obtains, that a molecule is just as observer relative as “the suitability of a day for a picnic” (Rorty 1998, p. 73). For Rorty, to name some feature of the world gravitational attraction is to engage in the descriptive practices of observers trained in the physical sciences. According to Rorty, rather than distinguishing between the intrinsic vs. extrinsic features of the world, Searle’s examples instead demonstrate “an arbitrary preference for the human purposes served by physicists over those served by picnickers” (Rorty 1999, p. 69). Arguing that objects only have relational properties, Rorty maintained that

everything that can serve as the term of a relation can be dissolved into another set of relations, and so on forever. There are, so to speak, relations all the way down, all the way up, and all the way in every direction: you can never reach something which is not just one more nexus of relations… there are no terms of relations which are not simply clusters of further relations. (Rorty 1999, pp. 53–54)

By rejecting realism’s claim that reality has an intrinsic nature, RP conceptualizes the nature of reality as contingent, not absolute. Within RP, the identity of any object is contingent on our human sensory/perceptual systems, and on the taxonomic/conceptual verbal practices we have learned to employ when describing reality. In short, within RP, “thinghood” is not intrinsic, but rather “description-relative” (Rorty 1991b, p. 4). As children (and continuing into adulthood), we become increasingly successful in accurately distinguishing between the objects of reality. Given such success, our accurately discriminating between reality’s objects cannot have entailed carving reality at its own joints because such joints (if they exist) are unknowable to us. Instead, we carve our own joints into reality. Any descriptive vocabulary we employ generally carries with it the standards of usage which determine when we have carved contingent reality correctly. Making what he characterized as a “banal point,” Rorty noted that when we are educated within a culture, we find out “a lot about the description of the world offered by our culture (e.g., by learning the results of the natural sciences)” (1979, p. 365). During language acquisition, we proceed through “stages of implicit, and then explicit and self-conscious, conformity to the norms of the discourses going on around us” (p. 365; emphasis added). In this manner, we gain knowledge of the objects comprising contingent reality.

For example, as a result of conforming to the norms of discourse of the English-speaking verbal community, English speakers/listeners exhibit nearly unanimous agreement about the existence of the commonplace objects (e.g., trees, rocks, birds) of everyday life. Of course, norms of discourse vary across verbal communities. What counts as an accurate description of reality within one verbal community may not be considered accurate by another employing a different vocabulary. As members of smaller verbal communities (e.g., physicists, philosophers, behavior analysts), we employ specialized vocabularies. We may describe reality “as atoms and the void, or sense data and awareness of them, or ‘stimuli’ of a certain sort brought to bear upon organs of a certain sort” (Rorty 1982, p. 14). For Rorty, while any given vocabulary may have utility in achieving specific purposes (e.g., prediction and control), no vocabulary can provide the one true description of reality, for there is no such perspective-free description attainable by us. According to the concept of contingent reality of RP, “the question of what ‘X’ refers to is… a question of how best to make sense of a community’s linguistic behavior” (Rorty 1982, p. xxiv).

Rorty (1999) recognized that his conception of contingent reality can be mistaken for a type of linguistic idealism. As Rorty defined it, linguistic idealism is the doctrine that “there was really no thing there to be talked about before peopled began talking—that objects are artifacts of language” (1999, p. 58). Similarly, Hacking (1999) defined it as “the doctrine that only what is talked about exists; nothing has reality until it is spoken of, or written about” (p. 24). To help distinguish RP from linguistic idealism, first consider what is arguably a prominent exemplar of linguistic idealism: Goodman’s (1978) Ways of Worldmaking. Goodman proposed that there are a “multiplicity of worlds” (1978, p. 1). Furthermore, he maintained that these worlds are “made from nothing by use of symbols” (p. 1). For instance, “we make a star as we make a constellation, by putting its parts together and marking off its boundaries” (p. 213). He explained,

we do not make stars as we make bricks; not all making is a matter of moulding mud. The worldmaking mainly in question here is making not with hands but with minds, or rather with languages or other symbol systems. Yet when I say that worlds are made, I mean it literally. (Goodman 1980, p. 213)

As this passage and the previous quotes indicate, Goodman offered a version of linguistic idealism.

To be sure, RP shares an important feature with linguistic idealism. Namely, like that doctrine, RP denies that we can “pick out objects” (Rorty 1999, p. 58) in an external world by discerning their alleged intrinsic properties. However, despite this shared feature, RP is not a type of linguistic idealism. Unlike linguistic idealism, the contingent reality of RP posits the existence of an external, physical reality which has autonomous independence; in short, it exists whether or not humans do. I previously reported that Rorty provided specific examples of objects (mountains and giraffes) that would have existed in the world even “if there had been no human beings” (1999, p. xxvi). That said, to avoid a serious misinterpretation, Rorty’s discussion of such objects requires additional clarification.

When Rorty affirmed that mountains and giraffes existed before we talked about them, he was not making the claim that the terms mountains and giraffes carve nature at its joints. Such a claim would clearly contradict a central premise of RP; to wit, refusing to assume that reality has an intrinsic nature. Consider again the example of giraffes. Rorty stated that “a giraffe is an object in the natural world… if there had been no human beings there would still have been giraffes … But this … does not mean that giraffes are what they are apart from human needs and interests” (1999, p. xxvi). Further elaborating,

We speak a language which includes the word ‘giraffe’ because it suits our purposes to do so … All descriptions we give of things are descriptions suited to our purposes. No sense can be made, we pragmatists argue, of the claim that some of these descriptions pick out ‘natural kinds’—that they cut nature at the joints. The line between a giraffe and the surrounding air is clear enough if you are a human being interested in hunting for meat. If you are a language-using ant or amoeba, or a space voyager observing us from above, that line is not so clear … More generally, it is not clear that any of the millions of ways of describing the piece of space time occupied by what we call a giraffe is any closer to the way things are in themselves than any of the others. (Rorty 1999, p. xxvi; all emphases added)

This passage indicates that RP denies that our concept of giraffe carves nature at its joints. Put another way, the term giraffe does not pick out a natural kind. In Rorty’s view, in our carving of nature we have drawn a line between (1) a “piece of space time” which we ended up calling giraffe and (2) the “surrounding air.” Why have we drawn it there? According to Rorty, not because this line reflects reality’s putative intrinsic nature. As previously noted, Rorty forcibly argued against the doctrine of intrinsicality. According to RP, reality cannot supply to us its own criteria for our identification of objects like giraffes and mountains. Instead, we have drawn the lines where we have because, in large measure, it has suited our purposes to draw them there. The lines can be drawn differently by us to serve different purposes. The lines could also be drawn differently by other sentient organisms. Employing a thought experiment, Rorty imagined that different lines of demarcation might be drawn by fanciful nonhuman, language using terrestrials or extraterrestrials—creatures with presumably different perceptual systems and perhaps different purposes served by their taxonomies and conceptual schemes. Furthermore, to avoid yet another possible misinterpretation, note that Rorty’s use of the expressions piece of space time and surrounding air (as used above) do not carve reality at its joints either. Like giraffe and mountain, these expressions also impose a contingent human conceptual/classificatory scheme on reality.

As the foregoing illustrates, in developing the account of contingent reality of RP, Rorty navigated a course somewhere between the Scylla and Charybdis of linguistic idealism and realism. As a consequence, RP exhibits similarities to both doctrines. Like linguistic idealism, RP denies that we can identify objects in an external world by discerning their alleged intrinsic properties. However, like realism, RP posits an external reality which exists whether or not humans are present. Moreover, like realism, RP maintains that reality does not spring into existence ex nihilo when we use our words. On the contrary, instead of our vocabularies somehow causing the world (as linguistic idealism maintains), within in RP, it is the world that “causes us to acquire the vocabularies we employ” (Rorty 1991a, p. 56).

Lastly, consider the treatment of RP of an enduring problem posed within philosophy—the problem customarily referred to as “the problem of the external world” (Bonjour 2011, p. 1). According to Hookway (2005), the problem of the external world consists of “showing how our subjective data provide us with reason for believing that there are external things” (p. 841). This problem raises the question of “whether and how beliefs about physical objects and about the physical world generally can be justified or warranted on the basis of sensory or perceptual experience” (Bonjour 2011, p. 1). By agreeing with realism that there is a world “out there,” RP potentially falls victim to this problem. More specifically, RP claims that we have genuine knowledge of the world and that at least some of our assertions about the world are true. How does Rorty justify these claims to true knowledge of an external world? To address that question, I turn now to the account of RP of truth.

Truth

“Questions such as “Does truth exist?” or “Do you believe in truth?” seem fatuous and pointless. Everybody knows that the difference between true and false beliefs is as important as that between nourishing and poisonous foods… the ability to wield the concept of “true belief” is a necessary condition for being a user of language, and thus for being a rational agent” (Rorty 2007, p. 89).

As this passage indicates, Rorty did not deny the importance of the concept of truth in the conduct of our daily lives (an approach to truth which I discuss in due course). Like the behavioral pragmatists, Rorty rejected the correspondence theory of truth (CTT). Within RP, absolute reality is considered inaccessible to us; lacking such access, we can never tell whether or not our statements correspond to it. In other words, we can never tell whether or not our statements are absolutely true. In arguing against CTT, Rorty observed that we cannot “penetrate beneath appearances and see nature ‘in its own terms’” (Rorty 1982, p. 192). Because we lack such access to “reality plain” (p. 154), “there is no way to hold the world in one hand and our descriptions of it in the other and compare the two” (p. 179) as CTT requires. Many other philosophers of stature (e.g., Blackburn 2005, 2008; Goodman 1978; Kuhn 1962; Putnam 1978, 1981, 1994, 1999) have offered similar arguments.

As a self-identified syncretist, Rorty borrowed some of those arguments from his fellow philosophers. Of those, arguably the most prominent has been his invocation (e.g., Rorty 1999, p. 38) of Putnam’s (1981, p. 49) “God’s eye point of view” argument. Recall that the behavioral pragmatists also employed a version of Putnam’s argument in making their case against CTT. Rorty agreed with Putnam that CTT fails as a theory of truth because it requires that truth seekers have access to what is patently inaccessible, namely, “unconceptualized reality”(Putnam 1994, p. 297). As Goodman (1978, p. 3)—another influence that shaped RP—put it,

you can offer to tell me how it < i.e., absolute reality > is under one or more frames of reference; but if I insist that you tell me how it is apart from all frames, what can you say? We are confined to ways of describing whatever is described.

In rejecting CTT, RP also rejects, as the putative goal of scientific inquiry, the development of increasingly accurate descriptions of absolute reality. Instead, RP opts for utility as the goal. Concisely put, “inquiry aims at utility” (Rorty 1999, p. xxvi). Like the behavioral pragmatists, Rorty viewed the perennial goal of science as the search for a “more useful description of the world” (p. 48) rather than absolute truth. However, despite this shared conviction, the behavioral pragmatists disagree with Rorty over the role that utility plays in their differing conceptions of truth. The behavioral pragmatists define truth as utility; RP does not. What then is the account of RP of truth?

To begin to answer that question, consider again the general doctrine of truth correspondism, the view that true statements are true because they correspond to reality (where reality is variously defined, depending on the version of truth correspondism). Rorty (1982, 1998) differentiated between two principal types of truth correspondism: nontrivial and trivial. CTT is an example of the nontrivial type. Across its various versions, nontrivial truth correspondism provides a specific account of the “special relatedness” of correspondence which holds between truth bearers and the intrinsic features absolute reality. Recall that, in Russell’s (1912) version, the correspondence relation consists of a “structural isomorphism” (p. 119)—a one-to-one correspondence—between parts of a true statement and parts of absolute reality. On the other hand, the trivial type of truth correspondism asserts that truth bearers correspond to contingent (not absolute) reality. In other words, truth bearers correspond to an external, physical reality in which “thinghood” is determined, not by intrinsic properties, but rather by the conceptual schemes and taxonomies which we employ. RP champions a version of the trivial type of truth correspondism. I dub the truth criterion of RP mundane correspondence, and the truths about contingent reality, contingent truth (Rorty himself did not use these labels).

By labeling this type of correspondism trivial, Rorty did not intend to disparage its importance as an account of truth. Rather, he considered it trivial in the sense that it offers no technical account of the correspondence relation. According to Rorty (1998), from the perspective of mundane correspondence, one can provide an account of “word-world relations for particular words used in a particular way by particular people” (p. 90) but no general account of correspondence when broadly conceived. For mundane correspondence, truth bearers correspond to the world only in the sense that “it is the world that determines truth” (Rorty 1982, p. 14). In espousing mundane correspondence, RP explicitly endorsed anti-representationalism. The anti-representationalism of RP maintains that no vocabularies—not even the vocabulary of particle physics—provides “representations of how things really are” (Rorty 1982, p. xlvi). The “principal argument” of RP for its anti-representationalism asserts “the relativity of descriptions to purposes… the view that inquiry aims at utility for us rather than an accurate account of how things are in themselves” (Rorty, 1999m p. xxvi). That said, the external world is, for Rorty, nonetheless the cause of our true beliefs. There are “relations of causation” between our true knowledge and “other items in the universe, but no relations of representation” (Rorty 1991a, p. 5; emphasis in original).

Unlike its cousin CTT, mundane correspondence attempts no analysis of the difficult-to-fathom correspondence relation. By opting to not enter the fray, Rorty escapes the problems associated with attempting to explicate that relation. Indeed, the history of philosophical discourse on this subject demonstrates that “many centuries of attempts to explain what ‘correspondence’ is have failed” (Rorty 1982, p. xxvi). By espousing mundane correspondence, Rorty (1982) embraced “the simple Aristotelian notion of truth as correspondence with reality with a clear conscience—for it will now appear as the uncontroversial triviality that it is” (p. 15). In effect, mundane correspondence restates Aristotle’s oft-cited stance on truth; namely, “to say of what is that it is and of what is not that it is not, is true” (quoted in David 2013, p. 2; emphasis added). Despite abjuring CTT’s assertion of a special correspondence relation, mundane correspondence nonetheless maintains that “there is no argument about the point that it is the world that determines truth” (Rorty 1982, p. 14). For example, on this view, contingent truths such as “the cat is on the mat” and “this rock is hard to move” are true because they correspond to—in the sense of being determined by—the world.

In offering the account of RP of mundane correspondence and contingent truth, Rorty insisted that he was not offering “a new theory of truth” (Rorty 1998, p. 11; emphasis added). Rather, he offered “an account of how the marks and noises made by certain organisms … can be fitted into our overall account of the interaction between these organisms and their environment” (Rorty 1991a, p. 10). In Rorty’s view, when we assert a proposition like “the cat is on the mat”—thus making the implicit claim that it corresponds to reality—our primary burden is that of providing sufficient justification in support of our claim. In his words, “the only criterion we have for applying the word ‘true’ is justification” (Rorty 1998, p. 4). Furthermore, justification is “always relative to an audience” (p. 4)—i.e., relative to the normative practices of a particular linguistic community. Therefore, Rorty argued that the goal of inquiry cannot be to accurate represent reality’s alleged intrinsic nature, but rather “to justify our belief to as many and as large audiences as possible” (1998, p. 39).

Different linguistic communities employ different criteria of justification. For example, “scientists have been programmed so as to respond to certain retinal patterns with ‘there goes a neutrino’” (Rorty 1991a, p. 56). A scientist’s description of those retinal patterns (i.e., a description of what she saw) constitutes justification for the assertion. On the other hand, a truth claim can, for example, fail to be justified, e.g., “if the litmus paper turns blue” (Rorty 1991a, p. 80). Within American jurisprudence, the juries in criminal trials must find the defendants guilty beyond a reasonable doubt to justify the assertion of a guilty verdict, while in civil trials, juries typically need only a preponderance of the evidence to justify an assertion of guilt. Of course, within RP, any justificatory claim—a claim that specific criteria of justification have been met—must correspond to contingent (not absolute) reality. According to Rorty, a community’s criteria of justification are typically established because of their demonstrated success over time. While our truth claims, and the justification we offer, are intimately linked, Rorty acknowledged that justification and truth are nonetheless distinct. Specifically, an assertion may be described as “fully justified, but perhaps not true” (Rorty 1998, p. 21). By acknowledging that an assertion may be fully justified but not be true, Rorty was acknowledging that our present justification for making that assertion may be successfully challenged in the future—for example, by newly acquired data. In this sense, characterizing an assertion as “fully justified, but perhaps not true” serves to caution us against a dogmatic commitment to that assertion. Concisely put, the difference between justification and truth is the difference between “present and future justifiability” (Rorty 2000a, p. 5).

Conclusion

Recall that one of the “internecine struggles” (Hackenberg 2009, p. 401) within behavior analysis has been the dispute over whether its underlying philosophy (radical behaviorism) is realist or pragmatist. Earlier, I argued that this dispute should matter to behavior analysts because, left unresolved, it threatens the coherence of that underlying philosophy. To help clarify the nature of the dispute, I provided detailed accounts of (1) realism, (2) behavioral pragmatism (BP), and (3) Rortian pragmatism (RP). Finally, using the insights of RP, I now offer three proposals for modifying the core tenets of behavioral pragmatism. These proposals, if accepted by behavioral pragmatists (and by other like-minded pragmatist behavior analysts), would go a long way in resolving the realism/pragmatism schism within behavior analysis.

Reality

• Proposal no. 1: Behavioral pragmatists should jettison nonrealism and adopt, in its place, the Rortian pragmatist’s existence assumption; to wit, the assumption that an external, physical, reality exists.

• Proposal no. 2: Behavioral pragmatists should adopt the Rortian pragmatist’s assumption that reality has an autonomous (but not intrinsic) independence; to wit, the assumption that reality exists whether or not humans are present to interact with it—a reality which is not assumed to have an intrinsic nature.

• Rationale: By acting in accordance with these two proposals, behavioral pragmatists would be (in effect) espousing the Rortian pragmatist’s concept of contingent reality. Put another way, these proponents of pragmatism would thereby be implicitly championing two of realism’s principal assumptions; namely, that (1) an external, physical reality exists and (2) this reality is independent in the sense that it exists regardless of whether or not we are present to perceive (or otherwise interact) with it—a reality not assumed to possess intrinsic features. Because such a change in the core tenets of BP would narrow the divide between pragmatist vs. realist behavior analysts, it could likely prove helpful in the overall effort to overcome the realist/pragmatist impasse in behavior analysis. Of course, helping bring both sides closer together may not be a goal which behavioral pragmatists are willing to entertain. Recall that behavioral pragmatists consider the issue of whether or not an external, physical reality exists as “irrelevant” (Barnes and Roche 1994, p. 70). Given their stated dismissive insouciance regarding this issue, behavioral pragmatists may be similarly uninterested in helping settle the pragmatism vs. realism debate. However, that said, by espousing the Rortian pragmatist’s position, BP would likely benefit in manner which could indeed serve as a motivating factor for making this change in their tenets.

To see how BP could benefit, first consider Barnes and Roche’s (1994) acknowledgement that “most behavior analysts assume that there exists a real, physical, and ordered universe” (p. 165). By adopting the Rortian pragmatist’s existence and independence assumptions, BP would benefit because such a change in its core tenets would make BP more attractive to those behavior analysts who espouse realism. What is more, because most members of Western mainstream culture reportedly share these two assumptions (Ebbs 1997; van Inwagen 2009), opting for the Rortian positions would also make BP more palatable to the culture at large. As Barnes and Roche (1994) acknowledged, behavior analysts “are more likely to be taken seriously by the modern, Western verbal community if we believe in an independent reality” (p. 166). The central goal of BP is “successful working” (Barnes-Holmes 2000, p. 198), the obtaining of “demonstrable effects in the domain of practical affairs” (p. 194). By embracing the Rortian pragmatist’s existence and independence assumptions, BP would achieve, as one of its “demonstrable effects,” an increased likelihood of gaining wider acceptance by other behavior analysts as well as mainstream culture. This, in turn, could also result in recruiting new members to the ranks of BP. Finally, these proposed modifications in the tenets BP would help narrow the divide between realists and pragmatists and, as a consequence, lessen the threat which that divide poses to the coherence of radical behaviorism.

Truth

• Proposal no. 3: Behavioral pragmatists should adopt the Rortian pragmatist’s truth criterion—mundane correspondence—in which truth bearers (statements, beliefs) are true because they correspond to contingent reality; to wit, they correspond to an extant, external, physical reality in which “thinghood” is determined, not by intrinsic properties, but rather by our human sensory/perceptual systems, taxonomic practices, and conceptual schemes.

• Rationale: Like it is nonrealism, the pragmatic truth criterion of BP contradicts the traditional view on the nature of truth held by most behavior analysts and, more broadly, by Western culture. Indeed, proponents of relational frame theory—who agree with the positions of BP on reality and truth—acknowledged that the pragmatic truth criterion “lies at odds with both commonsense notions of truth and, importantly, truth as it is understood in most mainstream science and philosophy” (Wilson, Whiteman, and Bordieri 2013, p. 27). With the behavioral pragmatist’s out-of-the-mainstream position in mind, I now offer two principal reasons for recommending that the behavioral pragmatists replace the pragmatic truth criterion with the Rortian proposal of mundane correspondence as their criterion of truth.

My first reason is that, by adopting this proposal, behavioral pragmatists would move closer to the mainstream view on truth. Just as adopting the Rortian approach to reality would bring the behavioral pragmatists closer to the traditional view held by most behavior analysts and the greater culture, so too would adopting the Rortian approach to truth. Thus, making this change would help rectify what is at issue between realists and pragmatists within the behavior-analytic community. Crucially, behavioral pragmatists could make this change without having to abandon successful working as the central goal of scientific inquiry. While Rortian pragmatists (unlike behavioral pragmatists) do not equate truth with successful working, they do agree with behavioral pragmatists that the pursuit of successful working—instead of absolute truth—serves as the principal goal of science. What could motivate behavioral pragmatists to make this change? Again, as I earlier argued in support of my two proposals that BP adopt the Rortian view on reality, embracing the Rortian position on truth would increase the likelihood that BP would be viewed more favorably by others, perhaps gaining new adherents to BP as well.

My second reason for offering this proposal is that, by adopting the Rortian approach to truth, behavioral pragmatists would remove a major deficiency in their explication of scientific verbal responses. Recall that, for the behavioral pragmatist, a cumulative record’s particular output pattern—and, importantly, any scientific verbal response asserting the presence of such a pattern—cannot represent “what the rat or pigeon ‘really’ did” (Barnes-Holmes 2000, p.198). Rather, the pattern is a discriminative stimulus for a specific scientific response (e.g., scallop, break-and-run, etc.) “that has been differentially reinforced in the presence of that pattern” (p. 198). In my view, this alternative account by BP is incomplete because it fails to tell us what criteria must be met before such reinforcement is provided (or withheld) by others. Presumably, reinforcement is provided when the specific scientific response corresponds—in some sense of the term—to the pattern displayed by the cumulative record and withheld when such correspondence is absent. Furthermore, BP also fails to tell us what criteria must be met when someone correctly asserts that in some particular endeavor, successful working has been achieved. Again, such assertions are presumably reinforced by others when they correspond (in some sense) to instances of successful working (e.g., in the lab).

As we have seen, within BP, the criterion cannot be correspondence to the intrinsic features of absolute reality; it cannot be correspondence to “what the rat or pigeon actually did.” Why? Because the researcher (like the rest of us) lacks access to absolute reality. The Rortian pragmatist agrees. However, the behavioral pragmatist goes further by ruling out any invocation of correspondence in explaining what makes a statement true. For example, Barnes-Holmes (2000) reported that “in defining truth behaviorally, the behavioral pragmatist always appeals to utility, and never correspondence, as a truth criterion” (202). Unlike the behavioral pragmatist, the Rortian pragmatist does not conclude that any use of the term correspondence is off-limits when providing an account of truth. If behavioral pragmatists were to adopt the Rortian concept of truth as mundane correspondence, they would thereby be providing the criterion which must be met before a scientific response is reinforced. In addition, they would also be providing what criterion must be met when someone correctly asserts that successful working has been achieved in some particular instance. By adopting this third proposal, as well as the aforementioned first and second proposals, behavioral pragmatists (and other pragmatist behavior analysts) would be narrowing the realism/pragmatism divide, thereby helping reduce the threat to the coherence of radical behaviorism as a philosophy of science.