By Owen Morawitz
There’s a scene in the dystopian science fiction film The Matrix (1999), where the character Morpheus (Laurence Fishburne) introduces the film’s protagonist Neo (Keanu Reeves) to the “Desert of the Real”. The scene takes place within “the construct”: a “loading program” made to resemble the Matrix of the film’s namesake, a “neural-interactive simulation” created by sentient machines to enslave humanity inside a virtual reality. At this shocking revelation, Neo dares to ask: “This… this isn’t real?”—a query to which Morpheus replies with the following: “What is real? How do you define real? If you’re talking about what you can feel, what you can smell, what you can taste and see, then real is simply electrical signals interpreted by your brain.” And while The Matrixis certainly not alone in using a bleak sci-fi premise to explore fundamental questions surrounding the nature of reality, Morpheus’ query does wrestle with one of the most heavily-debated issues within the Philosophy of Science: How do we know that scientific descriptions of the real world are in fact, real?
In this essay, I focus on two of the central positions in this debate: the ‘no miracles’ argument, which posits that scientific realism is the only philosophical position that does not treat the success of science throughout human history as a miraculous achievement; and conversely, the argument for pessimistic induction, which states that the dramatic changes in the history of science do not provide a sufficient level of confidence for belief in current scientific theories to provide accurate descriptions of the real world. I explore the concept of realism in a general sense, along with a more-specific formulation in terms of scientific realism, and how this formulation helps to categorise each argument as being a realist or an anti-realist position (along with their many variations). After analysing the flaws of each argument and considering case examples from the history of science, I then argue that neither argument presents a sufficiently convincing position for the support or rejection of scientific realism, and that a different position—one embracing the merits and flaws of both arguments—is required. To support this claim, I concentrate on the philosophical work of Richard Rorty and Hilary Putnam, who advocate for a pragmatic blend of fallibilism and anti-scepticism, where truth is a normative notion that cannot be naturalised or relativised to any specific culture or scientific practice.
To begin, let us consider the ‘no miracles’ argument (henceforth referred to as NMA) and the affirmative position for scientific realism. American philosopher and mathematician Hilary Putnam is often credited with providing the formalised version of the NMA in his work Mathematics, Matter and Method, which states that realism “is the only philosophy that doesn’t make the success of science a miracle.” As Pfeifer and Sarkar note in The Philosophy of Science: An Encyclopedia, “Realism takes the explanatory and predictive success of theories to warrant an ontological commitment to the existence of the entities they posit.” In layperson’s terms, things or external objects perceived by our minds through our senses are real, and they exist independent of our mind’s perceiving them: that is, “Theoretical posits, like electrons or genes, are not just useful ideas but real entities.” Realists argue that science has been extremely successful in providing a “detailed account of how theory and reality are connected,” through empirical predictions and explanations within different fields of scientific inquiry. If we take our best scientific theories to be “true (or approximately true, or [to] correctly describe a mind-independent world of entities, laws, etc.),” then considering the opposite scenario, where our best scientific theories are instead false, means their success in describing the real world as it is would be nothing short of miraculous. As Putnam states:
If those [external] objects don’t really exist at all, then it is a miracle that a theory which speaks of gravitational action at a distance successfully predicts phenomena; it is a miracle that a theory which speaks of curved space-time successfully predicts phenomena; (…) That science succeeds in making many true predictions, devising better ways of controlling nature, etc., is an undoubted empirical fact. If realism is an explanation of this fact, realism must itself be an over-arching scientific hypothesis. 
This statement helps to affirm Putnam’s formulation of the NMA and a position for scientific realism. However, some philosophers, like Peter Godfrey-Smith, argue that this “line of argument has been unimpressive as a defence of realism.” A proper analysis of NMA requires a more rigorous definition, especially when considering how the Philosophy of Science interprets realism. We can define scientific realism along three useful dimensions: “a metaphysical (or ontological) dimension; a semantic dimension; and an epistemological dimension.” The metaphysical (or ontological) aspect relates to the definition of philosophical realism mentioned above, where external objects perceived by the mind are taken to have a mind-independent existence. The semantic dimension commits scientific realism to a “face value” or literal interpretation of theoretical statements, where “claims about scientific objects, events, processes, properties, and relations, whether they be observable or unobservable, should be construed literally as having truth values, whether true or false.” The final—and perhaps most important—dimension frames scientific realism as an epistemic philosophy that yields knowledge of the world as it is; “a position concerning the actual epistemic status of theories (or some components thereof).” Combining these three dimensions, we get a more concrete definition of the scientific realism advocated by the NMA: the external world of perceived objects is real and mind-independent, and our best scientific theories provide true (or approximately true) and accurate descriptions of both the observable and unobservable components of the real world. However, as Anjan Chakravartty notes: “any epistemology of science that does not accept one or more of the three dimensions of realism—commitment to a mind-independent world, literal semantics, and epistemic access to unobservables—will thereby present a putative reason for resisting the miracle argument.” For this reason, scientific realism is not immune to critique and counter-argument, and the anti-realist positions which represent these critiques are numerous. Proponents of idealism are directly opposed to the metaphysical (or ontological) dimension of realism, rejecting mind-independence and insisting that “there is no world external.” Instrumentalism (or Bas van Fraassen’s more-intricate formulation of constructive empiricism) asserts that scientific realism cannot make literal or meaningful claims about unobservable objects; science “aims to give us theories which are empirically adequate,” and nothing is to be gained by asking if a scientific theory provides a true description of the world (either observable or unobservable). Points of contention also arise with the formulation of the NMA when the jump is made from the accuracy and success of “our best scientific theories” to the implied notion of “approximate truth.” As Michel Ghins explains in his critique of Putnam’s NMA, a theory’s supposed truth value is not required to make it scientifically valid in relation to real world phenomena.
The main source of the philosopher’s uneasiness with a scientific role for truth lies in the:
peculiar nature of [saying] the predicate ‘is true’. Truth (like falsehood) applies to statements and is thus a semantic property of some of our language components. That electrons do possess the property of having a negative charge may well cause certain facts, but nothing scientifically explanatory is added by saying that it is true that electrons have that property. 
As for our best scientific theories, Godfrey-Smith notes that “accurate representation of the world” is not the only justification for the success of science, and theories “can contain errors that compensate for each other.” One such example from the history of science is the Carnot cycle from the early nineteenth-century, which explains the operation of heat engines using the now-obsolete caloric theory (where heat consisted of a fluid called ‘caloric’), yet still includes certain elements which would later form the “basic ideas of thermodynamics.” Therefore, as in this example and many others, it does not seem compelling for realists to say that the success of science points “directly or unambiguously toward the truth of theories.” I submit that this is the fatal flaw regarding the optimistic view of scientific realism and the NMA. While scientific theories can be empirically successful in predicting aspects of the real world, it does not seem philosophically prudent, or convincing, to describe the only aim or success of science as an epistemic pursuit towards underlying truth-value claims of the real world, nor to invoke miracles (or their absence) as proof of this pursuit.
Returning briefly to The Matrix and the Baudrillard-name-dropping dialogue between Morpheus and Neo, we can see that Morpheus’ cryptic description of reality underlies a larger empirical dilemma. If all Neo knows of the real world is perceived through his senses, as electrical signals processed by his brain, what then is the fundamental difference (if any) between a simulated, virtual reality like the Matrix and real reality? For Neo, this revelation in perception is deeply troubling. Likewise, for the Philosophy of Science, this line of questioning leads to a cornerstone of scientific anti-realism: the argument for pessimistic induction (also referred to as pessimistic meta-induction; henceforth referred to here as PI). At its core, PI is an empirical argument against realism concerning “the ‘ground level’ inductive inferences that generate scientific theories and law statements.” To simplify, inductive inferences refer to cases where “one takes the past as grounds for beliefs about the future, or the observed as grounds for beliefs about the unobserved, or in general an inference that is ampliative—[that is] having more content in the conclusion than in the premises.” Proponents of PI also refer to specific uses of enumerative induction, where “one simply generalises from instances to all cases or to the next case.” As a critique against scientific realism, American philosopher and epistemologist Larry Laudan’s formulation of PI is the most well-known and comprehensive. However, I first focus on Putnam’s own version of PI, also known as ‘disastrous induction’, and reconstructed as follows by Peter J. Lewis as a reductio:
Putnam’s argument for PI
(P1) Assume most current scientific theories are true.
(P2) Then most past scientific theories are false, since they differ from the current theories in significant ways.
From the above, we see that the pessimistic aspect of PI arrives with Putnam’s conclusion (C1). If reasoning through PI indicates that our current scientific theories are likely to be false, then what level of confidence should we have in their ability to predict and reflect accurate descriptions of the real world in the present or the future? In many ways, this line of reasoning is an extension of the Carnot cycle example mentioned above in critique of the NMA. However, PI applies more-broadly to all scientific theories—that is, all scientific knowledge of the real world. As Godfrey-Smith suggests: “Some hold that fundamental ideas have changed so often within science—especially within physics—that we should always expect our current views to turn out to be wrong.” Against this charge of extreme pessimism, realists typically respond with various qualifications and ‘looser’ definitions in favour of keeping the optimistic position of scientific realism intact. While “often described as the greatest single challenge facing scientific realism,” the argument for PI and anti-realism itself is not entirely impenetrable. Critiques of PI from within the Philosophy of Science vary in scope and strength, however I focus here on the response from Lewis, which critiques both Putnam’s and Laudan’s general formulations of PI for employ fallacious reasoning. In his first premise (P1), Putnam assumes a truth-value claim regarding scientific theories and for the second premise (P2), his justification implicitly refers to our current standards of science, where “most past theories contain central terms that we now regard as non-referring, and presumably any such theory must be false.” Lewis takes issue with the inductive step Putnam makes from (P2) to (C1), in requiring “that current theories do not differ from past theories in any respect relevant to truth, [which] is precisely what a convergent realist would deny.” Convergent realism is a variation or qualification of scientific realism which focuses on the development of science over a historical timescale, where scientific “theories are [considered to be] converging on (‘moving in the direction of’, ‘getting closer to’) the truth.” As Lewis notes, for a convergent realist, this forward momentum towards approximate truth provides an explanatory distinction between the predictive successor failure of scientific theories: “Given this distinction, a realist need not accept that the falsity of past theories provides any inductive evidence for the falsity of current theories; the fact that unsuccessful (past) theories are false does nothing by itself to cast doubt on the truth of successful (current) theories.” While Putnam’s fallacious reasoning undermines his formulation of PI, Laudan’s 1981 paper “Confutation of Convergent Realism” presents a stronger rebuff to convergent realism’s appeal to success, formulating PI with “explicit appeal to the falsity of past theories in an attempt to undermine our justification for thinking that current theories are true.” Lauden’s argument for PI is reconstructed here by Lewis as a reductio:
Laudan’s argument for PI
(P1) Assume that the success of a theory is a reliable test for its truth.
(P2) Most current scientific theories are successful.
(P3) So most current scientific theories are true.
(P4) Then most past scientific theories are false, since they differ from current theories in significant ways.
(P5) Many of these false past theories were successful.
The fifth premise (P5) is Laudan’s major historical claim, as evidenced by a long list of “successful but false theories in the history of science, for example the Carnot cycle. However, once again, the inductive step between final premise and conclusion evidences fallacious reasoning in Laudan’s formulation of PI. For Lewis, the fact “that many false past theories were successful does not warrant the assertion that success is not a reliable test for truth.” While Lewis’ overall refutation of Laudan’s formulation of PI is rigorous and extensive, it can be most succinctly explained using the notion of relatability and an example of the false positives paradox:
The realist maintains that success can be used as a test of the truth of theories, since we can directly observe the success of a theory, but not its truth. For such a test, a false positive would be a case in which a theory is false but successful, and a false negative would be a case in which a theory is true but unsuccessful. In affirming that success provides a reliable test for truth, the realist is claiming that the rates of false positives and false negatives are low. If this is indeed the case, then if most current scientific theories are successful, it follows deductively that most current theories are true, as required by the realist. 
It seems that the issues with these formulations of PI leave us at something of a philosophical stalemate, where the success or failure of a scientific theory is not a question that can be easily “settled solely by the historical track record.” As was the previous case for the NMA, I argue that the fallacious flaws in both Putnam and Laudan’s formulations of PI—as illustrated by Lewis—provide sufficient evidence to disregard the main charge of the argument for PI as a “serious threat to scientific realism.” Ultimately, the level of confidence we have in any given scientific theory appears to be largely contextual, where it’s possible for the Philosophy of Science to have “different kinds of confidence, in different domains of science.”
By now, the attentive reader has likely noticed a curious thread running through this essay. References to Putnam’s philosophy have occurred twice now: first, in his formulation of NMA as an argument for scientific realism; and second, for his formulation of PI as a form of empirical argument and a position of anti-realism. This juxtaposition is entirely intentional, and the reasoning behind this contradiction will become clearer to the reader shortly. For Godfrey-Smith, pragmatism is defined as: “A family of unorthodox empiricist philosophical views that emphasise the relation between thought and action. For pragmatists, the chief purpose of thought and language is practical problem-solving. (…) Pragmatists reject the correspondence theory of truth.” This empirical approach acknowledges that our theoretical picture of the external world is derived solely through sensory observation and experience; namely, thought and action. However, we are never able to judge or prove with absolute certainty whether these beliefs about the external world are objectively ‘true’ or representative of real reality. One of the strongest advocates and defenders of pragmatism was American philosopher Richard Rorty, who outlined three important characterisations of pragmatism in his essay, Pragmatism, Relativism, Irrationalism. First, pragmatism “is simply anti-essentialism applied to notions like ‘truth,’ ‘knowledge,’ ‘language,’ ‘morality,’ and similar objects of philosophical theorizing.” Second, for the pragmatist, “there is no epistemological difference between truth about what ought to be and truth about what is, nor any metaphysical difference between morality and science.” Third, pragmatism “is the doctrine that there is no constraints on inquiry save conversational ones—no wholesale constraints derived from the nature of objects, or of the mind, or of language, but only those retail constraints provided by the remarks of our fellow-inquirers”. Rorty did admit that pragmatism was itself a “vague, ambiguous, and overworked word,” and while this might appear to further obfuscate the definition of pragmatism, we can boil down Rorty’s pragmatic philosophy to a single ethos: “There is no method for knowing when one has reached the truth, or when one is closer to it than before.” Along with Rorty and other neo-pragmatists such as American logician Willard Van Orman Quine, Putnam rejected the traditional conception of philosophy as a ‘Master Science’ that supplied “all other sciences with ontological and epistemological foundations.” As it turns out, Putnam was nothing if not comfortable with changing his mind and embracing the views of his peers when it came to philosophical positions, which I would argue is precisely what the practice of philosophy entails. While Putnam was careful not to “identify himself as a pragmatist,” his views on philosophy—particularly the philosophy of science—were influenced by two key commitments: “pragmatism and realism.” Like Rorty’s three characterisations of pragmatism, Putnam’s pragmatic view was defined by four central theses:
(1) anti-scepticism: pragmatists hold that doubt requires justification just as much as belief (…)
(2) fallibilism: pragmatists hold that there is never a metaphysical guarantee to be had that such-and-such a belief will never need revision (…)
(3) the thesis that there is no fundamental distinction between ‘facts’ and ‘values’; and
Pragmatists fundamentally reject the correspondence theory of truth: “that there is only one true and complete scientific theory, one true and complete set of beliefs about the world.” In this fallibilistic view, no belief or scientific theory is immune to revision. However, this does not require that we doubt all our beliefs or theories at once; instead, we are only required to doubt our beliefs or theories when we have sufficient reasons to doubt them. Such a rejection then typically yields a common-sense response from the view of both optimistic realists and pessimistic anti-realists: namely, how should we understand the concept of truth? On such a difficult (and clearly controversial) question, even the pragmatists are split. Putnam maintained the traditional pragmatic view, where “truth is a normative notion,” however he was adamant that such “normativity [could not] be naturalised or relativised to any particular culture or practice.” Putnam’s pragmatic view of the Philosophy of Science—and philosophy in general—demands “a continuous review of our scientific and philosophical beliefs and theories, guided by a notion of ‘goodness’ appropriate to those beliefs and theories.” In this view, pragmatism implies several practical obligations regarding both science and philosophy. The first is to remove the need entirely for objective or metaphysical claims about the nature of reality. The second, to know that our theoretical picture of the world is dependent on our response to observations of the world, as we inhabit and work within the world. Thirdly, it obliges one to adopt theories or beliefs because they are useful, practical and intellectually productive. The fourth obligation is to acknowledge that decisions regarding which theories or beliefs to adopt are not related to truth-value claims. Finally, it also obliges one to recognise that all decisions regarding theory or belief adoption “must always reflect present knowledge and the rich cultural, social and historical context in which scientists and philosophers act.” I argue that these practical elements provide ample justification for supporting a pragmatic philosophical position in this debate, over strict adherence to positions of realism or anti-realism respectively. Pragmatism resolves the apparent tension between the three dimensions of scientific realism (not contingent on mind-independence of external objects; no need for truth-value claims, or a convergent path to approximate truth; science is still able to provide accurate descriptions of the world, both observable and unobservable, through epistemic knowledge). Likewise, a pragmatic view dissolves the need for reasoning through pessimistic induction regarding the history and progression of scientific theories. In the later decades of his life, Putnam’s work in the field of neo-pragmatism left an indelible mark on the ceaseless and seemingly-inexhaustible debate surrounding scientific realism, as evidenced by his self-confessed “manifesto”:
Where are we then? (…) If there is no fact of the matter about what cannot be tested by deriving predictions, then there is no fact of the matter about any philosophical statement, including that one. On the other hand, any conception of rationality broad enough to embrace philosophy—not to mention linguistics, mentalistic psychology, history, clinical psychology, and so on—must embrace much that is vague, ill-defined, no more capable of being ‘scientized’ than was the knowledge of our ancestors. The horror of what cannot be ‘methodised’ is nothing but method fetishism. It is time we got over it. Getting over it would reduce our intellectual hubris. We might even recover our sense of mystery; who knows? 
Pragmatism allows us to disregard the semantic word games, the endless shuffling of vague qualifications, and directs the Philosophy of Science back to the task at hand: that is, doing real science. As Putnam’s philosophical peer (and wife), Ruth Anna Putnam, states quite persuasively: “To take pragmatism seriously is to take oneself to be living in a world that one shares with others (…) It is to see oneself not as a spectator of but as an agent in the world.”
In looking to summarise the positions discussed in this essay concerning the debate within the philosophy of science, it will be customary to make one final reference (I promise) to The Matrix. Prior to Neo’s awakening and knowledge of humanity’s inevitably dire fate at the hands (?) of the machines, his introduction to Morpheus and the Matrix itself takes the form of a dialogue questioning the nature of truth. A question which Morpheus, ever cryptically, answers with a choice—blue pill or red pill:
You are a slave, Neo. Like everyone else you were born into bondage, born into a prison that you cannot smell or taste or touch. A prison for your mind. (…) You take the blue pill, the story ends, you wake up in your bed and believe whatever you want to believe. You take the red pill, you stay in Wonderland, and I show you how deep the rabbit-hole goes. 
After much thought and deliberation, I have concluded that the debate between realism and anti-realism presents a similarly troubling choice: agree with the NMA (the blue pill), embrace the optimistic view of scientific realism, our best scientific theories, the aims and success of science as an epistemic pursuit, and notions of approximate truth; or, acknowledge the stiff and extreme conclusion of PI (the red pill), that the dramatic changes in the history of science do not provide a sufficient level of confidence for belief in current scientific theories to provide accurate descriptions of the real world. I contend that both arguments are thoroughly uncompelling as complete justifications and philosophically-rigorous positions in favour of scientific realism and anti-realism respectively, and—as argued in this essay—I find such a choice between them to be inherently difficult to swallow. However, if we extend the analogy just a little, and imagine placing Putnam in Neo’s reality-hacking shoes, then the answer becomes somewhat clearer: combine these blue and red pills, take the best from both (or the resulting mess), and arrive at something resembling what I wish to call the ‘purple pill of pragmatism’. Now, to this solution the reader might respond: ‘Well, isn’t this merely an example of philosophical fence-sitting on your part?’ Perhaps. However, when faced with such a difficult choice – a spot on the fence of this debate or a tumble down the philosophical rabbit hole—my preferred option is obvious. As Rorty deftly articulates: “In the end the pragmatists tell us, what matters is our loyalty to other human beings clinging together against the dark, not our hope of getting things right.”
Owen Morawitz is a freelance writer and thirty-something human male. He enjoys carving out a meaningless existence in the abyssal void, venturing beyond the bounds of the Southern Hemisphere, and listening to music that’s at times poignant, abrasive and restless—except when hungover. Alongside his B.A. (Philosophy and English Literature), Owen is currently completing an Honours thesis at The University of Queensland, focusing on manifestations of emptiness as object-oriented ontology in author Cormac McCarthy’s magnum opus, Blood Meridian.
 Anjan Chakravartty, “Scientific Realism,” The Stanford Encyclopedia of Philosophy, last modified June 12, 2017, https://plato.stanford.edu/archives/sum2017/entries/scientific-realism/.
 Michel Ghins, “Putnam’s No-Miracle Argument: A Critique,” Australasian Studies in History and Philosophy of Science, vol. 17 (2002): 121-138, Accessed June 3, 2018, doi:10.1007/978-94-017-2862-1_6 (emphasis added).
Chakravartty, Anjan. “Scientific Realism,” The Stanford Encyclopedia of Philosophy. Last modified June 12, 2017. https://plato.stanford.edu/archives/sum2017/entries/scientific-realism/.
Conant, James, and Ursula Zeglen. Hilary Putnam: Pragmatism and Realism. Edited by James Conant and Ursula Zeglen. London: Routledge, 2002.
van Fraassen, Bas C. The Scientific Image. Oxford: Oxford University Press, 1980.
Ghins, Michel. “Putnam’s No-Miracle Argument: A Critique.” Australasian Studies in History and Philosophy of Science, vol. 17 (2002): 121-138. Accessed June 3, 2018. doi:10.1007/978-94-017-2862-1_6.
Godfrey-Smith, Peter. Theory and Reality: An Introduction to The Philosophy Of Science. Chicago: University of Chicago Press, 2003.
Lewis, Peter J. “Why The Pessimistic Induction Is A Fallacy.” Synthese, vol. 129, no. 3 (2001): 371-380. Accessed June 3, 2018. doi.org/10.1023/A:1013139410613.
“Problem of Induction.” In The Philosophy of Science: An Encyclopedia, 379. Edited by Jessica Pfeifer and Sahotra Sarkar. New York: Routledge, 2006.
Putnam, Hilary. Mathematics, Matter and Method. Cambridge, UK: Cambridge University Press, 1975.
Putnam, Hilary. Meaning and the Moral Sciences. London: Routledge, 1978.
Putnam, Hilary. Realism with a Human Face. Edited by James Conant. Cambridge, MA: Harvard University Press, 1990.
Putnam, Hilary. Words and Life. Edited by James Conant. Cambridge, MA: Harvard University Press, 1994.
Putnam, Hilary. “Pragmatism and Non-scientific Knowledge.” In Hilary Putnam: Pragmatism and Realism, 14, 23-24. Edited by James Conant and Ursula Zeglen. London: Routledge, 2002.
Putnam, Ruth Anna. “Taking Pragmatism Seriously.” In Hilary Putnam: Pragmatism and Realism, 10. Edited by James Conant and Ursula Zeglen. London: Routledge, 2002.
“Realism.” In The Philosophy of Science: An Encyclopedia, 686. Edited by Jessica Pfeifer and Sahotra Sarkar. New York: Routledge, 2006.
Rorty, Richard. Consequences of Pragmatism (Essays: 1972-1980). Minneapolis: University of Minnesota Press, 1982.
Wachowskis, The, dir. The Matrix. 1999; Burbank, CA: Warner Bros. Pictures, 2008, DVD.
Zeglen, Ursula. “Introduction to Part I.” In Hilary Putnam: Pragmatism and Realism, 3-5. Edited by James Conant and Ursula Zeglen. London: Routledge, 2002.
Featured image by BMan_au via flickr