EDITORS NOTE: This essay originally appeared on Dan Kaufman’s previous blog, Apophenia. We are reprinting it here, as a preface to a video discussion bewteen Dan Tippens and Dan Kaufman on realism, anti-realism, and our knowledge of the world. That discussion will be posted this Thursday.
If you were to go to the trouble of asking ordinary people about their views on Knowledge and Reality — accosting them, at random, on street corners, perhaps — and succeeded in getting honest answers, you would likely discover that they hold something like the following view: What it is to know something is to possess some body of information — to have a “picture” of things — that squares with or is true to reality. If you were to push further, regarding ‘Reality’, they would likely characterize it along the lines of “everything that actually exists” (the ‘actually’ intended to preclude imaginary and fictional things like unicorns and Sherlock Holmes).
It’s not just ordinary people who think this way. Many philosophers and scientists hold the same view, though they may express it in language that is sufficiently sophisticated to give an impression otherwise. The truth, however, is that many of those philosophers and scientists who identify themselves as “Realists” take a position on Knowledge and Reality that is largely indistinguishable from the ordinary person’s view, just described. Consider the following, written by a well-established astrophysicist, on a discussion-forum, in which I recently participated:
“It is surely for nature to tell us (science) how things are. If it is indeed the case that some low-level quantum events have no cause (are not determined by the prior state of the system) then it is science’s job to report that.”
Now, I know exactly nothing about quantum mechanics and thus, haven’t the faintest idea whether or not quantum events have causes, but the view that reality exists, as it is, and that the job of science is to provide us with descriptions that match it, is stated quite plainly. I would be surprised to discover that there were many scientists, today, who thought otherwise, and as already mentioned, it is a view that is quite common among philosophers, as well.
One of the things that I find most interesting about this is that there are a number of very good reasons for thinking that this “common” view of Knowledge and Reality is false, reasons that have been conceived and expressed by some of our finest thinkers, over the course of the last three hundred years or so. I think it is worth rehearsing a few of these reasons, even if only to raise some eyebrows, with respect to the seemingly uncritical embrace of this brand of Realism today.
The doubts expressed by Descartes, in his Meditations on First Philosophy, are probably the most familiar examples of misgivings about the idea of a “fit” between our picture of the world and the world “as it is.” In Meditation One, Descartes observes that the picture of the world that we obtain by way of perception often does not “fit” the world. Standing on the top of a tall building, I may see tiny people below, who are, in fact, of regular size. Or I may see a person, in a nightclub, bathed in black light, and think that his shirt is purple, when it is actually white.
Of course, such examples are not examples of knowledge — in each case, it turns out that I was wrongabout the quality that I had ascribed — but they point towards more difficult kinds of cases, which Descartes — and John Locke — also recognized. Suppose that you are in an art gallery, making your way through an exhibition of color-field paintings and are struck by a solid, bright orange canvas. Later, you say to your friend, “I saw a solid, bright orange canvas today.” Is this something you know? And if it is, does your knowledge describe “reality as it is”?
If the sources I have consulted are accurate, rabbits can only see two colors: blue and green. (If they are incorrect, then just imagine schmabbits, who can only see blue and green.) Now, I am not sure what a rabbit would see, if you put it in front of the canvas I am speaking of, but let’s assume that it doesn’t see orange, but something else. What does this tell us? Both the rabbit and I have looked at the same canvas, and if Realism is true, there is a way the canvas actually is. And yet, the two of us experience the canvas differently. We have different “pictures” of this portion of the world. Is one of us wrong? Let’s presume that both the rabbit and I are in tip-top shape and that the viewing conditions are ideal. It’s hard to see the basis for claiming that one of us is correct and the other is not — that one of us “knows” the canvas is X, but the other does not know the canvas is Y. (We’re assuming, for the sake of the example, that rabbits can know things.) And yet, it also can’t be the case that we are both correct, because the canvas cannot be both X and not-X.
Our different experiences are explained by the fact that the rabbit and I have different kinds of eyes and different kinds of brains. But this raises a problem for the Realist, who thinks that there is “a way the world is.” Is color part of that way? How could it be, without us being forced to say that some color perceptions are right while others are wrong, which, as we’ve just seen, seems arbitrary. And it doesn’t take much thinking to realize that the same problem can be raised with respect to smell, taste, tactile sensation…indeed, with regard to all of the directly perceptible qualities of the world. After all, there are a lot of animals, with a dizzying variety of perceptual faculties. It is for reasons like this that both Descartes and Locke maintained that these sorts of qualities only exist in the mind of the perceiver; that they are features solely of our picture of the world and not of the world itself. Locke referred to such characteristics as “secondary qualities.” As for the features of things that they reallyhave — that belong to the world “as it is” — Descartes and Locke thought they were limited to what Locke called the “primary qualities” — mass, volume, and motion, the last of which, of course, also implies the reality of causality.
Ultimately, this view is a refinement of the common conception of Knowledge and Reality, rather than a rejection of it. While it is true that our immediate picture of the world may not “fit” with “reality, as it is,” once that picture is revisited and we screen out those characteristics that are solely features of the picture and not the world, we end up with a revised picture that does fit reality. (This is precisely the process of discovery that Descartes describes in Meditation Two, by way of an experiment with a ball of wax.) But working through these problems and refinements has introduced an idea that is essential to what will be a reason to reject Realism, in its common form: namely, that a part of our picture of the world comes from us and not from “reality as it is.”
David Hume represents the next big step forward in this idea. Giving greater scrutiny to the characteristics that Descartes and Locke were comfortable to assign to “reality, as it is,” Hume demonstrated that to a great degree, they too come from us. Take the idea that the world consists ofobjects that exist independently of anyone’s perception; that have “distinct and continued existence,” as Hume put it. This would seem an essential characteristic of “reality as it is,” as the Realist conceives it. And yet Hume can see no rational reason why anyone should think that there are such objects. Why, for example, should we believe that a chair in the room continues to exist, after we’ve exited and can no longer see it? Why should we not think, instead, that the chair exists when perceived and ceases to exist when not perceived? Hume searches in vain for some evidence — any evidence — that objects actually have such distinct and continued existence, but cannot find any. All that we have evidence of is the chair existing from t1 to t4, not existing from t-5 to t-10, and existing again from t-11 to the present.
The idea that the chair continues to exist when I am not experiencing it, Hume says, does not come from the chair, but instead, is the result of a number of psychological tendencies and habits, including the tendency to connect things that are close together, whether in space, time, or in terms of resemblance, and the habit of “filling in the gaps,” when sequences, like the one I’ve described with the chair, happen over and over again. Like the so-called secondary qualities, then, the distinct and continued existence of objects is also a part of our picture of the world that would seem to come fromus and not from “reality, as it is.”
Hume makes essentially the same argument regarding our idea of causality. When we say that a billiard ball striking another billiard ball caused that second billiard ball to move, we are saying that one series of events made another series of events happen. Hume describes this notion of one thing “making” something else happen as an idea of “necessary connection” and maintains (rightly) that it is at the heart of our common notion of causality. But like “distinct and continued existence,” Hume can find no rational grounds for the belief in “necessary connection.” In terms of what we actually have evidence of — what we observe, strictly speaking — we just see one event followed by another. We may, of course, see these two events, one following the other, over and over again, but never, in the course of these observations, do we observe the “making” relation; the necessary connection. As in the case of our idea of distinct and continued existence, Hume suggests that though our idea of one event following another comes from the world, “as it is,” the idea of necessary connection — of one eventmaking another event happen — comes from us. Specifically, it is due to the fact that (a) when we see one event, followed by another, over and over again, upon seeing the first event, we anticipate or expect the second event; (b) we unconsciously project this anticipation and expectation into the events themselves, mistaking them for a necessary connection.
By the time we get to Kant, every significant, substantive quality that we commonly attribute to “the world as it is” seems to come, instead, from us (Kant adds spatiality and temporality to those qualities supplied by the human mind). Finding it impossible to accept that there is no independent reality at all, Kant describes reality “as it is” as noumenal and, in principle, unknowable. Our knowledge, consequently, can only be of phenomena — of things as they are experienced — but this means that the idea of knowledge consisting of sentences or pictures that “fit” an “as it is” reality becomes impossible to sustain.
In the 20th century, new challenges to common Realism arose, but rather than the focus being on perception and the degree to which our own minds contribute to the picture of the world that results from it, as it was in the 17th and 18th centuries, the new concern was with the role played by language and in particular, by linguistic frameworks, conceptual schemes, and frames of reference.
One of the most well-known of these new challenges came from the philosopher, Willard Van Orman Quine, and specifically from his work on translation. In the second chapter of Word and Object, Quine describes what he calls a “radical translation” scenario, in which we imagine a linguist trying to translate the utterances of natives, whose language has never been translated before and for which there are no bilinguals. The linguist is working on one particular word, ‘gavagai’, which, after many days of observation, he determines is uttered when and only when a rabbit is present. As Quine points out, however, this is not sufficient to show that ‘gavagai’ means rabbit or that the speaker intends torefer to rabbits. After all, every time a rabbit is present, a set of un-detached rabbit parts is present, as is a time-slice-of rabbit — call it a “rabbity moment.” No further amount of evidence will enable us to make these further distinctions, since everything that is evidence that a rabbit is present is also evidence that un-detached rabbit parts and rabbity moments are present. Neither will further acts of ostension — i.e. pointing — succeed in more precisely identifying the speaker’s intended referent, as every act of pointing at a rabbit is an act of pointing at un-detached rabbit parts and at rabbity moments.
As Quine points out, in a subsequent essay, “Ontological Relativity,” the only thing that distinguishes rabbits from sets of un-detached rabbit parts and rabbity moments is their principle of individuation: “If you take the total scattered portion of the spatiotemporal world that is made up of rabbits, and that which is made up of un-detached rabbit parts, and that which is made up of rabbit stages,” Quine wrote, “you come out with the same scattered portion of the world each of the three times. The only difference is in how you slice it.” Does reality “as it is,” then, consist of rabbits, sets of un-detached rabbit parts, or rabbity moments? Given what Quine has just told us, to ask this is to ask a somewhat ill-formed question. Specifically, it is to ask what Rudolph Carnap described as an “external question,” by which he means a question that is asked independently of any linguistic framework or frame of reference. Within such a framework — that is, given a set of rules for counting, distinguishing one thing from another, etc. — I can ask questions like “Do rabbits exist?” or “How many rabbity moments are there?”, questions that Carnap characterizes as “internal,” but such questions are ill-formed when asked outside of a frame of reference, which is what we do when we say, “Yes, but what really exists, rabbits or rabbity moments?” As Nelson Goodman, the most sophisticated of the twentieth-century’s anti-Realists put it, “If I ask about the world, you can offer to tell me how it 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.”
The Realist may want to protest that when he asks the external question, “What really exists, rabbits or rabbity moments?” he is really asking, “Which linguistic framework/conceptual scheme/frame of reference” describes reality, as it is?” but this, too, is to ask a question that is somewhat ill-formed. Questions of reality are questions that we ask within frames of references, as it is only in the context of a frame of reference that we have the linguistic and logical tools to make ontological commitments. “To be real…means to be an element of the system,” Carnap wrote, “hence this concept cannot be meaningfully applied to the system itself.” To the extent that we are asked to choose between conceptual schemes or frames of reference that are at odds with one another, in terms of what they say exists or how many things exist, we can only do so on pragmatic grounds. Take a “micro-universe,” consisting of X1, X2, and X3. If I ask how many objects exist in this universe, you might say “three,” but of course, this presupposes both a conception of what counts as an object and of how one counts objects. As Hilary Putnam pointed out, if one were to ask a “Polish logician,” who treats mereological sums as objects, then in the universe we have described there are actually seven objects: X1, X2, X3, (X1+X2), (X2+X3), (X1+X3), and (X1+X2+X3). If one were to ask, “Well, which way of counting is the one that matches the way reality actually is,” we would not be able to answer. The choice between counting the first way or counting the second way will depend on the relative usefulness of each respective way of counting, relative to one’s aims, not on some notion that there is a privileged way of counting that reflects reality “as it is.”
At this point, a person who is unwilling to let go of his Realism may want to say: “Okay, there isn’t a way the world is, in itself, in the sense that it exists pre-conceptualized, independently of us, but there is a kind of generic reality — a world of “stuff” out there, which, depending on which linguistic frameworks/conceptual schemes/frames of reference we employ, becomes fully fleshed out as a picture of this, that, or another world.” In many ways, such a move is reminiscent of Kant’s characterization of the world-in-itself, as “noumenal,” and would seem to be motivated by similar intuitions and concerns. Alas, this will not do, for to say that something is to be organized, classified, arranged, etc., presupposes that it already has determinate characteristics — parts and pieces, so to speak. As Donald Davidson described it:
“We cannot attach a clear meaning to the notion of organizing a single object (the world, nature etc.) unless that object is understood to contain or consist in other objects. Someone who sets out to organize a closet arranges the things in it. If you are told not to organize the shoes and shirts, but the closet itself, you would be bewildered.”
The problem with this move, then, is not in the relationship between “scheme and content,” per se, but rather, in the idea that there can be scheme-less content; that we can separate frameworks from the things they are frameworks for. But this is just another way of saying that the problem lies in the idea, inherent to the common view of Knowledge and Reality that we have been talking about: that there is a world, separate from us, which our various descriptions and pictures (and the conceptual schemes they operate within) try to “get right.”
In closing, I am tempted to speculate as to why there are so many common Realists among professional philosophers and working scientists, today, but I won’t. The reasons are likely varied and many and guesses as to what they might be strike me as being of little value. What is important to understand is that it is not because the problems that I have described and the questions I have asked have been solved or answered. They have not. Like the many attempts to provide answers to modern Skepticism, solutions to the problems raised here continue to elude us.
Fortunately, our failure to solve the problems raised need have no effect whatsoever on scientific practice or even on our ordinary, common efforts to navigate our world. The Realism/Anti-Realism schism is empirically neutral, and one’s choice of a position has no scientific or other practical impact. This may seem to suggest that the questions we’ve raised are unimportant, and if what one means by “important” is that something is consequential for our practices, then I would agree that the Realism/Anti-Realism debate is unimportant. I do not define ‘important’ in this way, however, and for me, the question of how we understand our fundamental relationship to the world is of tremendous interest and something that I never tire thinking, reading, or talking about.
Rene Descartes, Meditations on First Philosophy (1641)
John Locke, An Essay Concerning Human Understanding (1689)
David Hume, A Treatise of Human Nature (1738)
Immanuel Kant, Critique of Pure Reason (1781)
Rudolph Carnap, “Empiricism, Semantics, and Ontology” (1950)
W.V.O. Quine, Word and Object (1960)
W.V.O. Quine, “Ontological Relativity” (1968)
Donald Davidson, “On the Very Idea of a Conceptual Scheme” (1973)
Nelson Goodman, Ways of Worldmaking (1978)
Hilary Putnam, The Many Faces of Realism (1987)