Reflections: Real Patterns, Deeper Facts, and Empty Questions
Daniel C. Dennett
From D.C. Dennett, The Intentional Stance (Cambridge, Mass.: MIT Press, 1987)
Several of the themes that receive a brisk treatment in "True Believers" expand into central topics in subsequent chapters. Perhaps the major source of disquiet about my position over the years has been its delicate balancing act on the matter of the observer-relativity of attributions of belief and other intentional states. On my view, is belief attribution (or meaning) in the eye of the beholder? Do I think there are objective truths about what people believe, or do I claim that all attributions are just useful fictions? My discussion of Nozick's objection attempts to place my view firmly on the knife-edge between the intolerable extremes of simple realism and simple relativism, but this has not been recognized as a stable and attractive option by many others in the field, and my critics have persistently tried to show that my position tumbles into one abyss or the other.
My view is, I insist, a sort of realism, since I maintain that the patterns the Martians miss are really, objectively there to be noticed or overlooked. How could the Martians, who "know everything" about the physical events in our world, miss these patterns? What could it mean to say that some patterns, while objectively there, are visible only from one point of view? An elegant two-dimensional microworld provides a clear instance: John Horton Conway's Game of Life (Gardner 1970), a simple yet extraordinarily rich source of insights that should become a part of everyone's imaginative resources as a versatile testbed for thought experiments about the relation of levels in science.
Imagine a huge piece of graph paper. The intersections (not the squares) of this grid are the only places-called cells-in the Life microworld, and at any instant each cell is either ON or OFF. Each cell has eight neighbors: the four adjacent cells north, south, east, and west of it, and the four nearest diagonal cells (northeast, southeast, southwest, northwest). Time in the Life world is also discrete, not continuous; it advances in ticks, and the state of the world changes between each tick according to the following rule:
Each cell, in order to determine what to do in the next instant, counts how many of its eight neighbors is ON at the present instant. If the answer is exactly two, the cell stays in its present state (ON or OFF) in the next instant. If the answer is exactly three, the cell is ON in the next instant whatever its current state. Under all other conditions, the cell is OFF.
The entire, deterministic physics of the Life world is captured in that single unexceptioned law. (While this is the fundamental law of the "physics" of the Life world, it helps at first to conceive this curious physics in biological terms: think of cells going ON as births, cells going OFF as deaths, and succeeding instants as generations. Either overcrowding or isolation causes death; birth occurs only under propitious circumstances.) By the scrupulous application of the law, one can predict with perfect accuracy the next instant of any configuration of ON and OFF cells, and the instant after that, and so forth. So the Life world is a fine Laplacean playpen: a simplified deterministic world in which we finite creatures can adopt the physical stance and predict the future with supreme confidence. Many computer simulations of the Life world exist, in which one can set up configurations on the screen and then watch them evolve according to the rule. In the best simulations, one can change the scale of both time and space, alternating between close-up and bird's-eye view.
One soon discovers that some simple configurations are more interesting than others. There are things that blink back and forth between two configurations, things that grow and then disintegrate, "glider guns" that emit "gliders"-configurations that reproduce themselves translated a few cells over, gradually gliding across the two-dimensional landscape-puffer trains, strafing machines, eaters, antibodies, space rakes. Once one understands the behavior of these configurations, one can adopt the design stance and ask oneself how to design larger assemblages of these objects that will perform more complicated tasks. One of the triumphs of the design stance in the Life world is a universal Turing machine-a configuration whose behavior can be interpreted as the state-switching, symbol-reading, and symbol-writing of a simple computer, which can be "programmed" to compute any computable function. (For more on Turing machines, see "The Abilities of Men and Machines" in Brainstorms, and Dennett 1985a.)
Anyone who hypothesizes that some configuration in the Life world is such a Turing machine can predict its future state with precision, efficiency, and a modicum of risk. Adopt the "Turing machine stance" and, ignoring both the physics of the Life world and the design details of the machine, just calculate the function being computed by the Turing machine; then translate the function's output back into the symbol system of the Life world machine. That configuration of ONs and OFFs will soon appear, you can predict, provided that no stray gliders or other noisy debris collide with the Turing machine and destroy it or cause it to malfunction.
Is the pattern that enables you to make this prediction "real"? So long as it lasts it is, and if the pattern includes "armor" to insulate the machine from noise, the pattern may survive for quite some time. The pattern may owe its existence to the intentions (clear-sighted or confused) of the machine's designer, but its reality in any interesting sense-its longevity or robustness-is strictly independent of historical facts about its origin.
Whether one can see the pattern is another matter. In one sense it would be visible to anyone who watched the unfolding of the particular configurations on the Life plane. One of the delights of the Life world is that nothing is hidden in it; there is no backstage. But it takes a big leap of insight to see this unfolding as the computation of a Turing machine. Ascending to the point of view from which this level of explanation and prediction is readily visible is optional, and might be difficult for many.
I claim that the intentional stance provides a vantage point for discerning similarly useful patterns. These patterns are objectivethey are there to be detected-but from our point of view they are not out there entirely independent of us, since they are patterns composed partly of our own "subjective" reactions to what is out there; they are the patterns made to order for our narcissistic concerns (Akins 1986). It is easy for us, constituted as we are, to perceive the patterns that are visible from the intentional stance-and only from that stance.1
Martians might find it extremely difficult, but they can aspire to know the regularities that are second nature to us just as we can aspire to know the world of the spider or the fish.
So I am a sort of realist. I decline the invitation to join Rorty's (1979, 1982) radical perspectivalism (Dennett 1982a). But I also maintain that when these objective patterns fall short of perfection, as they always must, there will be uninterpretable gaps; it is always possible in principle for rival intentional stance interpretations of those patterns to tie for first place, so that no further fact could settle what the intentional system in question really believed.
This idea is not new. It is a quite direct extension of Quine's (1960) thesis of the indeterminacy of radical translation, applied to the "translation" of not only the patterns in subjects' dispositions to engage in external behavior (Quine's "stimulus meanings"), but also the further patterns in dispositions to "behave" internally. As Quine says, "radical translation begins at home" (1969, p. 46), and the implications of his thesis extend beyond his peripheralism or behaviorism.
The metaphor of the black box, often so useful, can be misleading here. The problem is not one of hidden facts, such as might be uncovered by learning more about the brain physiology of thought processes. To expect a distinctive physical mechanism behind every genuinely distinct mental state is one thing; to expect a distinctive mechanism for every purported distinction that can be phrased in traditional mentalistic language is another. The question whether ... the foreigner really believes A or believes rather B, is a question whose very significance I would put in doubt. This is what I am getting at in arguing the indeterminacy of translation. (Quine 1970, pp. 180-81)
My argument in "Brain Writing and Mind Reading" (1975) enlarged on this theme, exposing the error of those who had hoped to find something in the head to settle the cases Quine's peripheralism left indeterminate: exactly the same considerations applied to the translation of any "language of thought" one might discover once one abandoned behaviorism for cognitive science. Another Quinian2 who has defended this position with regard to belief is Davidson (1974a):
Indeterminacy of meaning or translation does not represent a failure to capture significant distinctions; it marks the fact that certain apparent distinctions are not significant. If there is indeterminacy, it is because when all the evidence is in, alternative ways of stating the facts remain open. (p. 322)
More recently another application of the idea of indeterminacy has been ably defended by Parfit (1984), who has argued that the principles we (should) rely on to determine questions of personal identity will also inevitably leave open the possibility in principle of puzzling indeterminate cases. This is extraordinarily hard for many people to accept.
Most of us are inclined to believe that, in any conceivable case, the question "Am I about to die?" must have an answer. And we are inclined to believe that this answer must be either, and quite simply, Yes or No. Any future person must either be me, or someone else. These beliefs I call the view that our identity must be determinate. (p. 214)
We are just as strongly attracted to the view that the contents of our thoughts or beliefs must be determinate and resist the suggestion that the question "Do I believe there is milk in the fridge?" could fail to have a determinate answer, yes or no. Parfit shows that there are other cases in which we rest content without an answer to such questions.
Suppose that a certain club exists for several years, holding regular meetings. The meetings then cease. Some years later, some of the members of this club form a club with the same name, and the same rules. We ask: "Have these people reconvened the very same club? Or have they merely started up another club, which is exactly similar?" There might be an answer to this question. The original club might have had a rule explaining how, after such a period of non-existence, it could be reconvened. Or it might have had a rule preventing this. But suppose that there is no such rule, and no legal facts, supporting either answer to our question. And suppose that the people involved, if they asked our question, would not give it an answer. There would then be no answer to our question. The claim "This is the same club" would be neither true nor false.
Though there is no answer to our question, there may be nothing that we do not know.... When this is true of some question, I call this question empty. (p. 213)
We recognize that in the case of the club's existence, and in similar cases, there is no "deeper fact" that would settle the matter, but in the case of personal identity, the supposition that there is-must be-just such a deeper fact dies hard. Not surprisingly, this is one of the brute convictions that Nagel (1986) cannot abandon:
Why isn't it enough to identify myself as a person in the weaker sense in which this is the subject of mental predicates but not a separately existing thing-more like a nation than a Cartesian ego?
I don't really have an answer to this, except the question-begging answer ... (p. 45)
I make the analogous case for what we might call belief identity or belief determinacy. In chapter 8, "Evolution, Error, and Intentionality," I show how Searle, Fodor, Dretske, Kripke, and Burge (among others) all are tempted to search for this deeper fact, and how their search is forlorn. I claim, in other words, that some of the most vigorously disputed questions about belief attribution are, in Parfit's sense, empty questions.
How is this persistent illusion created? In the reflections on "Making Sense of Ourselves," I describe a false contrast (between our beliefs and those of lower animals) that creates the mistaken conviction that our own beliefs and other mental states must have determinate content.
Another leitmotif that gets its first play in "True Believers" is the comparison between Putnam's Twin Earth thought experiment and various simpler environmental dislocations or exchanges-of thermostats, in this instance. The theme is developed more fully in chapter 5, "Beyond Belief," and in the example of the "two-bitser" in chapter 8.
Finally, the attenuated and conditional endorsement of the idea of a language of thought at the end of "True Believers" is further elaborated in chapter 5 and in chapter 6, "Styles of Mental Representation," and the reflections following it.
1. Anscombe spoke darkly of "an order which is there whenever actions are done with intentions" (1957, p. 80) but did not say where in the world this order was to be discerned. In the brain? In behavior? For years I could not make sense of this, but now I see what she may have intended and why she was so coy in her description (and location) of the order. It is as hard to say where the intentional order is as it is to say where the intentional patterns are in the Life world. If you "look at" the world in the right way, the patterns are obvious. If you look at (or describe) the world in any other way, they are, in general, invisible.
2. Some non-Quinians have maintained versions of this idea. Wheeler (1986) insightfully shows that Derrida can be seen to "provide important, if dangerous, supplementary arguments and considerations" to those that have been advanced by Davidson and other Quinians. As Wheeler notes:
For Quineans, of course, it is obvious already that speech and thought are brainwriting, some kind of tokenings which are as much subject to interpretation as any other.... Still, there seems in non-Quinean circles to be a covert belief that somehow inner speech is directly expressive of thought. (p. 492)
This covert belief is exposed to attack in chapter 8.