The Atlas of Reality
Page 51
Holistic Phenomenalism runs into some danger of indeterminacy and relativism, since we may disagree about what counts as the best or simplest theory. It is then incompatible with a scientific realism that takes the statements of science to be true, when interpreted at face value.
13.4.2 Common-sense and Scientific Realism
Both common sense and modern science attribute natures to physical things. These natures, in turn, give rise to powers and dispositions if we embrace either Powerism or Nomism. However, for Phenomenalists physical objects are mere bundles of ideas or mere bundles of possible ideas or mere fictions. Neither bundles nor fictions are the sort of thing to realize powerful natures. Berkeley was quite willing to embrace this conclusion. On his view, only “spirits”, that is God and human souls, are powerful and capable of real action. Therefore, one cost of Phenomenalism is that it puts two attractive views of powers, Powerism and Nomism, beyond the pale. Phenomenalists have to embrace either Strong Hypotheticalism or Neo-Humeism about powers.7
Much of modern science involves the positing of unobservable entities, such as microscopic particles, distant astronomical formations or fields of force like gravity and electric charge. Phenomenalists must be anti-realists about all such unobservable entities, since they do not correspond to actual or even possible sensations. It might be possible for Phenomenalists to suppose that such entities are unobservable only to us but are directly observable by God or by some other actual or merely hypothetical creature. However, there is little or no reason for Phenomenalists to make that assumption, since we have not encountered any creatures with the relevant sort of sensory apparatus. Further, it is doubtful that such an apparatus could actually exist. Given the nature of things like quarks, quasars, and magnetic fields, it is plausible that they are not even possibly directly observable.
Phenomenalists ought, then, take an instrumentalist view of scientific theories. On this view, scientific theories are all literally false, or at least we have no good reason to suppose them to be true. But they are useful fictions. By pretending they are true, we can better make predictions and obtain control every observable phenomenon.
Such an instrumentalist position faces two major challenges. First, there is the no-miracle argument for scientific realism. Defenders of this argument (first clearly proposed by Hilary Putnam in Putnam 1975: 73) claim that the broad instrumental success of scientific theories in a wide range of applications would be an incredible “miracle” unless those theories were true (or at least approximately true). Scientific realism offers a kind of explanation for the instrumental success of science that instrumentalists cannot offer.
Second, scientific realists (such as Quine 1957: 229, 233) have argued that theoretical science is wholly continuous with common sense. We believe in the unobservable entities proposed by science in exactly the same way and with the same justification that we believe in the many observable but not yet observed entities proposed by the exercise of ordinary common sense. Just as the empirical evidence makes it reasonable to believe in an unobserved burglar, so too can empirical evidence make it reasonable to believe in an unobservable field of force. In fact, if the evidence is strong enough, it is unreasonable not to believe in such unobservable entities. Consequently, the fact that Phenomenalism entails the unreasonableness of such beliefs counts against it.
13.4.3 The apparent impossibility of phenomenal reduction
If Phenomenalism is to be plausible, Phenomenalists must provide some sort of account of how facts about perceptions can make true all or nearly all of our ordinary beliefs about physical objects. Carnap attempted just such a reduction of physical theory to truths expressible in a purely phenomenal language (Carnap 1934/1937).
One crucial problem for Phenomenalists, which was noted by Mill, is that there simply aren't enough actual experiences to ground all of the physical truths. As far as we know, there is no one presently perceiving the far side of Pluto, and yet we believe that there are many facts about physical objects in that region of space. Phenomenalists like Mill have attempted to fill this gap by appealing to merely possible or counterfactual perceptions, facts about what observers would perceive if they were located in the right place. Carnap attempted to carry out this project. In doing so, he encountered a problem that he was never fully able to overcome: how to specify the counterfactual situations under which the possible perceptions are supposed to happen. If we say that there would be certain sense perceptions if one of us were looking at Pluto from the other side, we are making reference to a matrix of physical facts, including the possible location of an observer in physical space. In order to carry out the reduction successfully, the antecedents of the relevant conditionals (the if-and-when parts) would have to specify the counterfactual situations in language that is free of all reference to physical objects (including our bodies or sense organs) or physical space. This is a tall order.
In particular, Phenomenalists would have to reduce the spatiotemporal primary qualities of place, distance, and orientation to the intrinsic perceptual states of possible observers. Instead of saying something like ‘If you were on the opposite side of Neptune and looking this way, …’ we would have to say instead something like ‘If you were to have the sensations that would accompany a trip to the opposite side of Neptune, along with the sensation of turning around by 180°, …’ However, it is very far from clear that these two if-clauses would pick out the same possible situations. After all, one might have all of the sensations associated with a trip to Neptune by taking LSD or peyote! What one would then experience would have nothing whatsoever to do with the physical attributes of the farthest planet. The failure of would-be Phenomenalists like Mill and Carnap to solve this sort of problem gives us some reason to think it impossible.
13.5 Arguments against Solipsism
Besides the Veil of Perception, arguments for Idealism and Solipsism depend on an appeal to Ockham's Razor. This appeal can succeed only if Idealism or Solipsism provide the simplest overall account of the world, given our evidence. There is a battery of arguments in the history of philosophy that challenge this assumption. These arguments suggest that mental phenomena are dependent on or interdependent with a world of extra-mental physical objects. These arguments can be taken either as providing additional support for Perceptual Realism (and against the Veil of Perception) or as a defense of Inferred Anti-Idealism:
13.1A.1 Inferred Anti-Idealism. Non-mental (physical) things exist but cannot be perceived (either directly or indirectly).
13.5.1 Appeals to common sense and to perception
To believe in both physical objects and other minds is a matter of common sense. It is also an inseparable presupposition of the methods of modern science, from particle physics to sociology. Many of our ethical, legal, and political practices are grounded on the supposition of the reality of the external world. We can therefore justify the rationality of belief in the external world by appealing both to common sense and to our rational practices.
In addition, since the arguments for the Veil of Perception were unsuccessful, we can justify the appeal to physical objects on the grounds that we seem to perceive them via the senses. The Reliable Perception Presumption (PEpist 4) supports the idea that it is reasonable to assume that things are as they appear to us to be, unless we have strong grounds to think otherwise.
Complex thought depends on language, and language is a social phenomenon. Wittgenstein (1953) argued that, without a real society as a source of linguistic norms, we have no criterion for the correct use of language. Whatever seems right will be right, especially if we are “speaking” an idiolect of the present moment. We can't know what we are thinking if we have no criterion for the correct use of our language. Thus, Solipsism is self-refuting. The argument goes something like this (following the so-called “communitarian” interpretation):8
I would not know what I was thinking unless there were a society that provided me with linguistic norms.
I do know what I am
thinking.
Therefore, there exists a society that provides me with linguistic norms.
The main weakness in Wittgenstein's argument is his assumption that thought depends upon language. In particular, Wittgenstein assumes (on the this communitarian interpretation of his argument) that the norms of thought (the correct application of our concepts) always depend on the social norms of language. Wittgenstein spends much of The Philosophical Investigations (1953) trying to substantiate this claim, but he doesn't demonstrate that a concept cannot somehow carry within itself some principle of application that determines which uses are correct and incorrect, apart from the judgments of others. In addition, Wittgenstein's argument seems to assume that the social norms embedded in language are necessarily reliable, and so are the final court of appeal for the correctness and incorrectness of our judgments. This seems wrong. It is natural to think that everyone in a language community could err systematically in the application of some concept. For example, it seems possible that everyone wrongly believes that a particular tree is older than every other tree.
13.5.2 The Paradigm Case Argument against ubiquitous illusion
Solipsists argue that all of our perceptions are in fact illusions, insofar as they present to us a world of things beyond our minds. However, as J.L. Austin argued (1962), the concept of illusion makes sense only if we can also grasp its contrary, real (veridical) perception. In order to know that some of our representations are illusory we must be also be able to know that they are not genuine perceptions of physical things. However, it is possible to grasp a concept in only two ways: by being able to define it in terms of simpler concepts or by knowing paradigm cases of the concept and its complement. The concepts perception and physical are not definable. Any attempt to define them would make use of a circle of synonymous concepts, such as knowledge, seeing, and hearing in the case of perception, and material, external, colored, and located in the case of physical. On Solipsism, we have no knowledge of any paradigms of any of these concepts. Hence, we cannot grasp the concept illusory, and Solipsism is self-refuting.
Austinians are appealing here to something like the following principle of concept possession:
PEpist 5 Conceptual Acquaintance. It is possible to know the truth of a proposition that involves the use of a concept C only if either (i) one knows a definition of C in terms of other concepts or (ii) one knows a non-empty class of instances of C and of instances of non-C.
Definitional Foundationalism. A system of definitions known by a single person at a time forms a finite, acyclic directed graph (that is, there are no definitional circles or infinite regresses—all definitional “trees” terminate in undefinable “leaves”).
The combination of Conceptual Acquaintance and Definitional Foundationalism entails that all of our fundamental, undefinable concepts are concepts we know to apply in certain concrete cases, the paradigm cases. Solipsists believe that all concepts of physical things are uninstantiated, that there are in fact no instances of these concepts. This would contradict the combination of Conceptual Acquaintance and Definitional Foundationalism. Solipsists can't have the concepts that they claim are uninstantiated. Similarly, Solipsists believe that they have knowledge of only a single thing, and that thing is mental. This would also contradict the combination of Conceptual Acquaintance and Definitional Foundationalism, since we can't grasp the concept being mental without knowing some non-mental things.
Solipsists' claim that there appears to be a physical world in space but there is in fact no such world couldn't be true, since they cannot even conceive of the possibility of such a physical world without knowing some actual paradigms of their physical concepts.
There are other variations on Conceptual Acquaintance that would lead to the same conclusion. For example, consider a second version:
PEpist 5.1 Conceptual Acquaintance (Second Version). It is possible to know the truth of a proposition that involves the use of a concept C only if either (i) one knows a definition of C in terms of other concepts or (ii) one's idea of C is connected causally with at least one actual instance of C and one actual instance of non-C.
These Conceptual Acquaintance principles are far from obviously true, however. It would seem to be possible, for example, to acquire the concept green entirely from colorful hallucinations without ever seeing anything green or non-green. Suppose, for example, that we stimulate the optical nerve of a blind person in such a way as to produce a variety of colorful experiences. We could tell the person, ‘This is red’, ‘Now this is green’, and so on. It would seem that the blind person might acquire the relevant concepts in advance of any veridical experience of colored surfaces. Also, imagine someone born with color-altering lenses who learns to use color-words correctly (as terms for physical surfaces) but who mislearns what each of the colors really looks like. Suppose, for example, that orange things look red to such a person. He would have the concept red, even though he calls the color ‘orange’ (matching the behavior of normally-sighted people) and even though what he takes to be paradigms of that color and of its complements are not really so.
In response, anti-Solipsists could weaken Conceptual Acquaintance so that it only requires that, in order to have some concept C, either oneself or some other member of one's species must have been acquainted with both instances and counter-instances of C. Even if one has never encountered anything really red but base one's concept red on hallucinatory experience, it still must be the case that other members of the human species (including many of one's ancestors) must have used their perceptual capacities to distinguish red things from non-red things How else could our concepts and sensory appearances be connected with real properties?
This argument thus depends ultimately on a kind of causal theory of conceptual content, the theory that concepts have the content they do because of causal interactions between the concept-holder or his ancestors and the relevant properties. Support for such a theory would seem to rely heavily on empirical evidence, especially from cognitive psychology and evolutionary biology, evidence whose validity Solipsists would reject. However, do Solipsists have an equally plausible theory of conceptual content to offer?
Solipsists might attempt to appeal to the Realist-Aristotelian account of intentionality sketched briefly in Section 7.3. On this view, we can think about properties simply by incorporating the corresponding universals into our thoughts. A universal like SQUARENESS would be literally a part of each of our thoughts or apparent perceptions of square things. Solipsists would still owe us some account of why those universals are present in our thoughts and perceptual states. What is responsible for putting them there? Aristotle thought (in De Anima Book III) that it was the fact that these universals are instantiated in our natural environment, together with the fact that we are naturally equipped to perceive them by means of our senses, that are jointly responsible for the presence of the universals in our minds. Obviously, Solipsists would not have access to such an account.
13.6 Conclusion and Preview
Solipsism and Idealism run up against the obvious fact that we perceive many things beyond our own minds, things that co-exist with us in a world of space and time. The burden of proof is on Solipsists and Idealists, and historically they have sought to carry this burden by appeal to the Veil of Perception that is supposed to lie between our perceptual capacities and the world beyond our minds. We've found plenty of reason to doubt the existence of this Veil. In addition, critics of Solipsism have marshaled a battery of arguments in defense of the common-sense conviction that we ordinarily perceive physical things and other people.
In light of these conclusions, we can now turn to those metaphysical questions that arise from the world as we find it, questions about physical parts and wholes, space and time, possibility and necessity, and cause and effect.
Notes
1. Even a no-subject Idealist like David Hume will recognize a difference between mental actions (like believing or intending) and mental objects (like color
s and smells). For Hume, a mind is a bundle of mental events and acts. His conception of the self will fit our definition of ‘mind’ so long as some bundles include actions essentially.
2. This overlooks one possible complication. It could be that there are mental entities (like sense data) that are perceived by more than one subject at the same time, even if one subject's act of perceiving is not the object of any other subject's awareness. Nonetheless, Solipsism remains a very economical theory if one has no reliable evidence for the existence or perceptual activity of other subjects.
3. The reader may have noticed that the Phenomenal Exportation Principle is formally very similar to the Barcan formula, a formula that licenses the exportation of the existential quantifier through the possibility operator. The Phenomenal Exportation Principle licenses the same exportation of the existential quantifier through the ‘It appears to S that’ operator.
4. Jackson later recanted this objection to Intentionalism.
5. For details about the processing of color information in the brain, see Hunt 1982, Hardin 1988/1993, and de Valois and de Valois 1993.
6. Thanks to Richard Lawton Davis for this argument.
7. It might be possible for Berkeleyan Phenomenalists to embrace Nomism. The laws of nature could connect bundles of divine-ideas of various kinds. However, such a position does not seem to be available for Holistic Phenomenalists, since there are no actual entities to play the role of physical objects on those views.