Reply to Bermúdez

I have some disagreements with Bermúdez position, but I think what he takes to be the main difference between our positions on the uses of contrariety may be a simple misunderstanding that we can just clear up. I welcome this opportunity to say, and think, a bit more about contrariety than I did in Varieties. First, however, the disagreements. I attempted, in Varieties, to base my discussion of the development of representational systems in animals very closely on experimental evidence concerning what various kinds of animals are actually known to be capable of. Varieties was not an armchair project. What I am missing in Bermúdez’s discussion is a any evidence that the kinds of practical inferences he credits animals with performing actually occur in any species other than man. I suspect they do not, based on various kinds of evidence cited in Varieties. (Modes tollens is pretty hard for freshmen, let alone gazelles.) But let us leave the actual world aside for a moment, as philosophers are (much too much) prone to do, and pursue only theoretical possibilities (should there actually be any such things). In Chapter 7, “Intensionality,” I was at some pains to explain why the truth or satisfaction conditions of a representation do not determine either its articulation, or what concepts, if any, are required to make use of it. (This claim is a naturalist relative and generalization of the view that the meanings of sentences are not merely functions from possible worlds to truth values (c.f., truth conditions) but structured propositions.) This view is supported by a comprehensive theory of intentional representation (Varieties, Chapters 5-6) and a detailed theory of empirical concepts (Millikan, 1984, 2000). It is neatly illustrated with the (indispensable) example of bee dances, whose truth conditions are most easily expressed in English with sentences that make explicit reference to the home hive of the bees, to the sun, to nectar, to relations among these involving angles and to distances measured in some given metric. But the dance itself has variable places only for angle and distance, expressed, of course, without reference to units of measure, and it is also highly improbable that the bees themselves have concepts of most of the above-named things. So I am unprepared to follow Bermúdez when he claims, for example, that an animal capable or representing that a is not F, supposing this kind of negation to express indefinite contrariety, would have to think that “there is some property G such that a is G and G is a contrary of F,”1 and to have concepts of properties, of contraries, modal concepts and so forth. In simple cases, the satisfaction condition for a representation is merely what it must correspond to by rule if it is to guide its consumer(s), its user(s) (for example, another bee, another part of the brain) in performing their tasks in an (historically) normal way. The idea that logical inference in humans requires meta-representation is likewise insupportable on the view I have presented. All that is required is an ability to move from representations having appropriate forms to further representations of appropriate form (abilities not to be confused with mere dispositions (Millikan, 2000, Chapter 4)). (Computers, too, are guided by the forms of input that they receive without knowing anything about form or about truth.) I have questioned whether animals prior to humans possess anything like subject-predicate representations at all. They oviously do have practical concepts, abilities to reidentify various kinds of things and to learn how to deal with them for various practical purposes. But, as argued in Chapter 18, I could find no reason, given what what we know of animal behavior, to suppose that animals recognize the fundamental ontological distinctions that so obsess us philosophers between kinds, features or properties, and individuals (see also Millikan, 2000, Chapters 2-5). It doesn’t follow that they are are incapable of inference. Sticking to theoretical possibilities, why might not two inner maps, A and B, of different sections of an animal’s environment be combined to produce a third map that shows relations not shown on A or on B taken separately? But Bermúdez is concerned with the way an animal might use a practical grasp of contrariety in performing inferences. Let me try then to be more constructive, for I think there is ample evidence that animals do have to have a practical grasp of certain kinds of contrariety, even if not for exactly the purposes Bermúdez proposes. Still, we should begin by acknowledging a bit more of a gap between contrariety and negation than Bermúdez’s discussion suggests, for he mentions only pairs of contraries that are exhaustive, indeed, that could as easily be defined using simple negation. To be absent is, indeed, not to be present and to be undetected is, indeed, not to be detected. He tells us, for example, that “[t]here is no distinction between acting upon the absence of the gazelle and acting upon it’s not being the case that the gazelle is present.” But if we replace “absent” and “present” with a non-exhaustive contrary pair, such as “at the pond” and “in the upper meadow,” we get a different result. Consider the lion, for whom there would surely be a large gap between acting on the Gazelle’s being at the pond and acting on its not being the case that the Gazelle is in the upper meadow. My suggestion, in the chapter to which Bermúdez refers, was that simple negation involves an assertion of indefinite contrariety. To know what A is not, is not usually to know what A is. A main conclusion from the evidence on animal behavior cited in Part IV of Varieties is that the behavior of non-human animals is always guided by current perception (current perception of what’s inside or outside the animal), or by an extension of current perception, in light of the animal’s practical history or the practical history of its species. Giving just one example, there is no evidence that I know of that animals pay any attention to correlations or causal connections of kinds that have so far had no practical consequences either for themselves or for their species. And although perception often yields (fallible) knowledge of the future (Chapter 13 ff.) it never directly yields an indefinite negative, but only, perhaps, a positive contrary. So a question arises what the use of a simple (that is, indefinite) negative representation would be. I suggested that its use was in theoretical judgment, paradigmatically, judgment about things neither we nor our ancestors have so far learned to use for practical purposes. Animals, I suggested, do not play the game of theoretical judgment, which involves the identification of subjects of judgment and of properties of these that have, so far as guidance by immediate perception and its extensions is concerned, no immediate significance for the guidance of action. The story I then told about our use of theoretical judgment was an involved one that cannot be repeated here. But it is the ability to recognize contraries of theoretical properties (qua properties, that is, no longer as undifferentiated from real kinds and from objects) and the ability to recognize theoretical subjects for judgment that I said must be developed in parallel. I apologize that I was probably the cause of misunderstanding in two ways. First, I used the term “theoretical judgment” in a way that fit my own theory rather than, say, more familiar usage (such as Kant’s). Second, I carefully pointed out that animal communication systems used with conspecifics not only fail ever to express negation but cannot generally even express contraries, which is true, but probably not relevant, hence thoroughly misleading in this context. For if we look more deeply into the way inner representational systems work, in both animals and humans, Bermúdez is absolutely right, I think, in pointing to the importance of the recognition of certain kinds of contraries. A fundamental principle involved in the operation of the nervous system is that there exist ubiquitous inhibitory relations among various neurons preventing certain combinations of them from firing together. One result that is popularly known involves the necker cube, which will switch back and forth, appearing first to have a corner in front and then to have the same corner behind, but that never presents these contraries at the same time. Contrast the waterfall illusion, whereby after staring at a waterfall for a minute or so, other things looked at then appear both to be moving rapidly upward and to be staying in exactly the same place. In the necker cube case, contrariety is directly recognized by the perceptual systems, as expressed in their refusal to represent certain contraries together in connection with the same part of an object, whereas in the waterfall case, contrariety is not recognized at this early perceptual stage. Extrapolate this so as to connect more closely with Bermúdez’s concerns. We can speculate that the animal who is afraid of lions but not of gazelles perceives these as contraries. That is, the visual system of such an animal will not return the verdict that what is moving across the visual field over there is at once a lion and a gazelle, at once to be run from and approached. It will recognize these as contraries, and if necessary, prompt the animal to make further observations before acting.2 Similarly, I’m quite sure that my cat’s visual system treats perceptions of our dog, who loves him, and of the neighbor’s dog, who chases him, as strong contraries. This can easily be interpreted as a precursor to recognition of contradiction in thought, which is most immediately expressed in the inhibition a contradiction induces in one’s confidence with regard to the perceptions or judgments that gave rise to it

Articulation (sociology · Aside · Epistemology · Linguistics · Naturalism · Political science · Politics · Representation (politics · Simple (philosophy · Sociology · Suspect · Utterance · Law · Linguistics and Discourse Analysis · Philosophy · Philosophy and Theoretical Science · Syntax, Semantics, Linguistic Variation

• Doxastic Revision in Non-Human Animals

  Open Access•Laura Danón, Daniel Kalpokas•Review of Philosophy and Psychology•2024

  Add to list

• Attributing Psychological Predicates to Non-human Animals

  Open Access•Andrés Crelier•Review of Philosophy and Psychology•2023

  Add to list

• Language, Thought, and Other Biological Categories

  Ruth Garrett Millikan•Language, Thought, and Other…•1984

  Add to list

• On Clear and Confused Ideas

  Open Access•Ruth Garrett Millikan•On Clear and Confused Ideas•2000

  Add to list