Ronald De Sousa

The Philosophical search for Natural Kinds is motivated by the hope of finding ontological categories that are independent of our interests. Other requirements, of varying importance, are commonly made of kinds that claim to be natural. But no such categories are to be found. Virtually any kind can be termed 'natural' relative to some set of interests and epistemic priorities. Science determines those priorities at any particular stage of its progress, and what kinds are most 'natural' in that sense is always a real and lively scientific question. The general philosophical problem of scientific realism is also real; but between the scientific and the metaphysical, the Problem of Natural Kinds' sits otiose.

Some wag reported the following story: Scholars have recently established that the Iliad and the Odyssey were not, after all, written by Homer. They were actually written by another author, of the same name.The majority of current theories of naming and reference, including ones as divergent in other respects as those of Russell and Searle, would rule this story impossible. They would do so on roughly these grounds: the sense and reference of the name ‘Homer’ is determined, given the absence of other reliable testimony, by the sole description: ‘Author of the Iliad and the Odyssey’. ‘Homer is the author of the Iliad and the Odyssey’ is therefore known a priori, hence necessarily true. There could not be another author of that name and claim to fame.In lectures delivered at Princeton in the Winter of 1970, Saul Kripke offered a lucid alternative to such theories.

Ainslie’s Picoeconomics presents an ingenious theory, based on a remarkably simple basic law about the rate of discounting the value of future prospects, which explains a vast number of psychological phenomena. Hyperbolic discount rates result in changes in the ranking of interests as they get closer in time. Thus quasi-homuncular “interests” situated at different times compete within the person. In this paper I first defend the generality of scope of Ainslie’s model, which ranges over several personal and subpersonal levels of psychological analysis. I raise a problem which results from the temporal relativity of assessments of value, and affects the possibility of objective values. Finally, I offer one example of a form of time-related irrationality on which Ainslie’s scheme does not seem to have a grip, namely one which relates not to a situation’s relative position in time, but to its temporal (‘progressive’ or ‘perfect’) temporal aspect.

Ira Brevis furor, said the Latins: anger is a brief bout of madness. There is a long tradition that views all emotions as threats to rationality. The crime passionnel belongs to that tradition: in law it is a kind of “brief-insanity defence.” We still say that “passion blinds us;” and in common parlance to be philosophical about life's trials is to be decently unemotional about them. Indeed many philosophers have espoused this view, demanding that Reason conquer Passion. Others — from Hume to the Emotivists — have appeared to reverse this hierarchy.” But those philosophers who refuse to join in the general denigration of emotion as irrational usually share the presupposition that the role of rationality is limited to the calculation of means. In so far as emotions are concerned with the determination of ends, they remain, on this view, beyond the pale of rationality. Modern decision theorists have worked out schemes to assess the rationality of desires, as well as actions, against the background of beliefs and other desires.1 But these schemes leave no room at all for emotions, except, by implication, as disrupters of the rational process.

Nothing seems to follow strictly from 'X believes that p'. But if we reinterpret it to mean: 'X can consistently be described as consistently believing p'--which roughly renders, I think, Hintikka's notion of "defensibility"--we can get on with the subject, freed from the inhibitions of descriptive adequacy. But defensibility is neither necessary nor sufficient for truth: it tells us little, therefore, about the concept of belief on which it is based. It cannot, in particular, specify necessary conditions for the consistent ascription of belief--as opposed to rational belief. If there are no such conditions, all belief ascriptions must be treated as atomic: which is implausible. If there are some, they must be settled on before an account of consistency can be complete. The reason is simple: it is that we have beliefs about our own beliefs. A set of first order beliefs is consistent if they can all be true together; but it is a lesson of Moore's paradox that consistency of second-order beliefs requires additional constraints. If someone both believes that p and that he does not believe that p, the propositions he believes might all be true together, yet he is inconsistent. And the characterization of second order consistency will remain incomplete, so long as nothing is said about the consistency of first order belief ascriptions. Suppose someone says and believes: "I am inconsistent: I believe both p and ~p." This is "indefensible": but is it insight, or nonsense? Breast-beating of this sort is guaranteed success: something he believes is bound to be false. But that remains diagnostically frustrating: is he inconsistent because what he said is true, or on the contrary because it is inconsistent?

Abstract The notion of representation lies at the crossroads of questions about the nature of belief and knowledge, meaning, and intentionality. But there is some hope that it might be simpler than all those. If we could understand it clearly, it might then help to explicate those more difficult notions. In this paper, my central aim is to find a principled criterion, along lines that make biological sense, for deciding just when it becomes theoretically plausible to ascribe to some process or state a representational role. I shall be especially concerned with some differences, in this regard, between classical and connectionist models. The relation between ?standard? artificial intelligence and connectionism turns out to illustrate a ?first in, last out? principle: What we most easily understand (and so can program) is what we have most recently invented; tasks we ourselves perform best, by contrast, are a lot harder to understand. Classical AI has modelled the former; connectionism tries to tackle the latter. I end with some speculations about the possible implications of these considerations for our understanding of understanding

Emotions can be the subject of moral judgments; they can also constitute the basis for moral judgments. The apparent circularity which arises if we accept both of these claims is the central topic of this paper: how can emotions be both judge and party in the moral court? The answer I offer regards all emotions as potentially relevant to ethics, rather than singling out a privileged set of moral emotions. It relies on taking a moderate position both on the question of the naturalness of emotions and on that of their objectivity as revealers of value: emotions are neither simply natural nor socially constructed, and they apprehend objective values, but those values are multidimensional and relative to human realities. The axiological position I defend jettisons the usual foundations for ethical judgments, and grounds these judgments instead on a rationally informed reflective equilibrium of comprehensive emotional attitudes, tempered with a dose of irony.