Testability and Ockham’s Razor: How Formal and Statistical Learning Theory Converge in the New Riddle of Induction

Nelson Goodman’s new riddle of induction forcefully illustrates a challenge that must be confronted by any adequate theory of inductive inference: provide some basis for choosing among alternative hypotheses that fit past data but make divergent predictions. One response to this challenge is to distinguish among alternatives by means of some epistemically significant characteristic beyond fit with the data. Statistical learning theory takes this approach by showing how a concept similar to Popper’s notion of degrees of testability is linked to minimizing expected predictive error. In contrast, formal learning theory appeals to Ockham’s razor, which it justifies by reference to the goal of enhancing efficient convergence to the truth. In this essay, I show that, despite their differences, statistical and formal learning theory yield precisely the same result for a class of inductive problems that I call strongly VC ordered, of which Goodman’s riddle is just one example.     

Goodman’s “New Riddle”

First, a brief historical trace of the developments in confirmation theory leading up to Goodman’s infamous “grue” paradox is presented. Then, Goodman’s argument is analyzed from both Hempelian and Bayesian perspectives. A guiding analogy is drawn between certain arguments against classical deductive logic, and Goodman’s “grue” argument against classical inductive logic. The upshot of this analogy is that the “New Riddle” is not as vexing as many commentators have claimed (especially, from a Bayesian inductive-logical point of view). Specifically, the analogy reveals an intimate connection between Goodman’s problem, and the “problem of old evidence”. Several other novel aspects of Goodman’s argument are also discussed (mainly, from a Bayesian perspective).     

No unchallengeable epistemic authority,of any sort,regarding our own conscious experience – Contra Dennett?

Dennett argues that we can be mistaken about our own conscious experience. Despite this, he repeatedly asserts that we can or do have unchallengeable authority of some sort in our reports about that experience. This assertion takes three forms. First, Dennett compares our authority to the authority of an author over his fictional world. Unfortunately, that appears to involve denying that there are actual facts about experience that subjects may be truly or falsely reporting. Second, Dennett sometimes seems to say that even though we may be mistaken about what our conscious experience is, our reports about “what it’s like to be us” must be correct. That view unfortunately requires a nonstandard and unremarked distinction between facts about consciousness and facts about “what it’s like.” Third, Dennett says that reports about experience may be “incorrigible.” However, that claim stands in tension with evidence, highlighted by Dennett himself, that seems to suggest that people can be demonstrably mistaken about their own experience. Dennett needlessly muddies his case against infallibilism with these unsatisfactory compromises.     

Evolution and Two Popular Proposals for the Definition of Function

Summary  In the biological realm, a complete explanation of a trait seems to include an explanation in terms of function. It is natural to ask of some trait, “What is its function?” or “What purpose in the organism does the particular trait serve?” or “What is the goal of its activity?” There are several views concerning the appropriate definition of function for biological matters. Two popular views of function with respect to living things are Cummins’ organizational account and the Griffiths/Godfrey-Smith modern history account. Whereas Cummins argues that a trait functions so as to contribute to the general organization of some organism’s present structure, Griffiths, and Godfrey-Smith argue that a trait functions because of its fitness with respect to the organism’s recent evolutionary history. In this paper, I show how these accounts can be made compatible and compliment one another. Given that structure, organization, operational flexibility, function, and evolutionary history are all factors to be considered in an organism’s makeup, we should expect that the traits of an organism function the way they do because such traits presently contribute to the overall organization of the organism (Cummins) as well as were selected for in the organism’s species’ recent ancestry (Griffiths/Godfrey-Smith).     

Hume’s Solution of the Goodman Paradox and the Reliability Riddle (Mill’s problem)

Many solutions of the Goodman paradox have been proposed but so far no agreement has been reached about which is the correct solution. However, I will not contribute here to the discussion with a new solution. Rather, I will argue that a solution has been in front of us for more than two hundred years because a careful reading of Hume’s account of inductive inferences shows that, contrary to Goodman’s opinion, it embodies a correct solution of the paradox. Moreover, the account even includes a correct answer to Mill’s question of why in some cases a single instance is sufficient for a complete induction, since Hume gives a well-supported explanation of this reliability phenomenon. The discussion also suggests that Bayesian theory by itself cannot explain this phenomenon. Finally, we will see that Hume’s explanation of the reliability phenomenon is surprisingly similar to the explanation given lately by a number of naturalistic philosophers in their discussion of the Goodman paradox.     

Structuring Ends

There is disagreement among contemporary theorists regarding human well-being. On one hand there are “substantive good” views, according to which the most important elements of a person’s well-being result from her nature as a human, rational, and/or sentient being. On the other hand there are “agent-constituted” views, which contend that a person’s well-being is constituted by her particular aims, desires, and/or preferences. Each approach captures important features of human well-being, but neither can provide a complete account: agent-constituted theories have difficulty accounting for the normativity of their claims, and substantive good theories have difficulty accounting for how a person’s actually adopted aims shape what is good for her and hence what she has reason to do. I articulate and defend a hybrid view that equals these approaches in systematicity and completeness of explanation yet seeks to surpass them in coherence with our ordinary judgments about what human well-being consists in. This hybrid view maintains, with agent-constituted theories, that a person’s well-being is (1) significantly constituted by her actually adopted aims; (2) deeply contingent; (3) agent-relative; (4) significantly dependent on spatially and temporally remote events; and (5) significantly independent of her experiences. The hybrid view also maintains, with substantive good theories, that a person’s well-being is (6) in part determined by facts independent of her aims, desires, and preferences; (7) such that all her aims are subject to critical evaluation and revision; and (8) constituted by her aims only if these aims are choiceworthy.     

A New Form of Agent-Based Virtue Ethics

In Morals From Motives, Michael Slote defends an agent-based theory of right action according to which right acts are those that express virtuous motives like benevolence or care. Critics have claimed that Slote’s view— and agent-based views more generally— cannot account for several basic tenets of commonsense morality. In particular, the critics maintain that agent-based theories: (i) violate the deontic axiom that “ought” implies “can”, (ii) cannot allow for a person’s doing the right thing for the wrong reason, and (iii) do not yield clear verdicts in a number of cases involving “conflicting motives” and “motivational over-determination”. In this paper I develop a new agent-based theory of right action designed to avoid the problems presented for Slote’s view. This view makes morally right action a matter of expressing an optimal balance of virtue over vice and commands agents in each situation to improve their degree of excellence to the greatest extent possible.     

Was Brzozowski a “constructionist”? A contemporary reading of Brzozowski’s “philosophy of labour”

Brzozowski’s ‘philosophy of labour’—to which he devoted a number of writings starting in 1902—presents problems of interpretation. A conceptual approach to his conception shows it to be a sometimes uneasy mix of realist and anti-realist notions. Brzozowski appears to have thought that labour is not first of all about the things it supposedly transforms, but rather about itself. I suggest that Brzozowski can be read in the spirit of Nelson Goodman’s nominalist constructionalism (“worldmaking”). On this account, labour in Brzozowski’s idiom turns out to be the constitution of forms of symbolizing sufficient unto themselves and the needs they satisfy. However, that Brzozowski was not entirely consistent in the views I impute to him—he forever sought for some ‘external’ measure of the rightness of labour/symbolizing—can be explained at least in part by his ‘humanism’, that is, his commitment to the task he assigns humankind, that of creating the one meaningful world attesting to virtually unrestricted human power.     

Reductive levels and multi-scale structure

I discuss arguments about the relationship between different “levels” of explanation in the light of examples involving multi-scale analysis. I focus on arguments about causal competition between properties at different levels, such as Jaegwon Kim’s “supervenience argument.” A central feature of Kim’s argument is that higher-level properties can in general be identified with “micro-based” properties. I argue that explanations from multi-scale analysis give examples of explanations that are problematic for accounts such as Kim’s. I argue that these difficulties suggest that some standard assumptions about causal competition need to be revised.     

Ordinary experience and the epoché: Husserl and Heidegger versus Rosen (and Cavell)

In various publications, Stanley Cavell and Stanley Rosen have emphasized the philosophical importance of what they both call “the ordinary.” They both contrast their recovery of “the ordinary” with traditional philosophy, including the phenomenological philosophy of Edmund Husserl. In this paper, I address Rosen’s claims in particular. I argue that Rosen turns the real situation on its head. Contra Rosen, it is not the case that the employment of Husserl’s epoché distorts the authentic voice of “the” ordinary—a voice that is clearly audible only from within everyday life. For (pace both Cavell and Rosen) there is no single “voice” of the ordinary: There are many such “voices,” not all of which are to be relied upon. Therefore, if we want to achieve an adequate grasp of ordinary experience, and Rosen does want this, we precisely need the epoché to curtail the misleading messages of certain other “voices of the ordinary.” Moreover, and somewhat surprisingly, this positive evaluation of the Husserlian epoché finds support in Heidegger’s writings from the twenties. I argue that Heidegger, too, believed that the epoché was an indispensable tool for the philosophical attempt to capture ordinary experience.     

Some Metaethical Desiderata and the Conceptual Resources of Theism

In this paper, I argue that theists are extremely well-situated with respect to developing metaethical accounts that qualify as ‘robust’ versions of moral realism. In the first part of the essay, a number of metaethical desiderata are identified. In the second part, theistic strategies for accommodating those desiderata are explained and defended. The upshot is that, contrary to the received philosophical wisdom, there are good theoretical reasons for theistic philosophers to seek to develop metaethical accounts that ground moral facts in facts about God.     

Bealer and the autonomy of philosophy

George Bealer has provided an elaborate defense of the practice of appealing to intuition in philosophy. In the present paper, I argue that his defense fails. First, I argue that Bealer’s theory of determinate concept possession, even if true, would not establish the “autonomy” of philosophy. That is, even if he is correct about what determinate concept possession consists in, it would not follow that it is possible to answer the central questions of philosophy by critical reflection on our intuitions. Furthermore, I argue that Bealer’s account of determinate concept possession in fact faces serious problems. Accordingly, I conclude that Bealer does not succeed in vindicating the appeal to intuition in philosophy.     

To Understand Understanding: How Intercultural Communication is Possible in Daily Life

I propose a few epistemological and methodological reflexions to account for intercultural daily communication. These reflexions emerged during a sociological research in Mendoza, Argentina, with Huarpes Indigenous students at the University of Cuyo. I observed that Indigenous people became quasi “ethnographers” of diverse environments. To make intelligible their classmates’ behavior, and to account for their own behavior, Huarpes follow, in diverse environments and interactions, public rules of meaning. The objective of this paper is twofold: (a) to stress the methodological scope of ordinary communication and ordinary reasoning in order to study understanding between people from different groups and categories, and (b) to contest a kind of “pessimist” standpoint in social sciences and philosophy according to which the use of ordinary language reduces possibilities for understanding. Interviews, participant observation in natural situations, and a review of literature about language and understanding are the basis of this paper.     

The Anomalous Foundations of Dream Telling: Objective Solipsism and the Problem of Meaning

Little sociological attention is directed to dreams and dreaming, and none at all is directed to how people tell one another about dreams. Ordinary settings in which dreams are told mimic the conditions of “breaching” experiments and should produce anomie, but dream telling proceeds without trouble. Foundational orientations of ordinary dream talk assimilate into professional dream studies, where dream narratives are “data” and the analysis of narratives is “dream analysis.” That such practices proceed without trouble poses some interesting problems for sociology in terms of how anyone experiences “constraint” in the telling and hearing of dreams.     

“Cannot” Implies “Not Ought”

I argue for a version of “ought” implies “can”. In particular, I argue that it is necessarily true that if an agent, S, ultima facie ought to do A at T’, then there is a time T* such that S can at T* do A at T’. In support of this principle, I have argued that without it, we cannot explain how it is that, in cases where agents cannot do the best thing, they often ought to do some alternative action – such as get help or do the promised action later; nor can we explain the phenomenon of necessary enablers or the phenomenon of more stringent prima facie obligations overriding less stringent ones in cases where the agent cannot fulfill both.     

Are there a posteriori conceptual necessities?

I critically assess Stephen Yablo’s claim that “cassinis are ovals” is an a posteriori conceptual necessity. One does not know it simply by mastering the relevant concepts but by substantial empirical scrutiny. Yablo represents narrow content by “would have turned out”-conditionals. An epistemic reading of such conditionals does not bear Yablo’s claim. Two metaphysically laden readings are considered. In one reading, Yablo’s conditionals test under what circumstances concepts remain the same while their extensions diverge. As an alternative, I develop a more literal metaphysical interpretation: Yablo’s conditionals draw on scenarios which are qualitatively identical to some original situation. None of these interpretations sustains Yablo’s core thesis.     

In Defense of Imperative Inference

“Surrender; therefore, surrender or fight” is apparently an argument corresponding to an inference from an imperative to an imperative. Several philosophers, however (Williams 1963; Wedeking 1970; Harrison 1991; Hansen 2008), have denied that imperative inferences exist, arguing that (1) no such inferences occur in everyday life, (2) imperatives cannot be premises or conclusions of inferences because it makes no sense to say, for example, “since surrender” or “it follows that surrender or fight”, and (3) distinct imperatives have conflicting permissive presuppositions (“surrender or fight” permits you to fight without surrendering, but “surrender” does not), so issuing distinct imperatives amounts to changing one’s mind and thus cannot be construed as making an inference. In response I argue inter alia that, on a reasonable understanding of ‘inference’, some everyday-life inferences do have imperatives as premises and conclusions, and that issuing imperatives with conflicting permissive presuppositions does not amount to changing one’s mind.