Charles R. Pigden

In the philosophical debate over the epistemic status of conspiracy theories, the view that each theory ought to be judged on its own merits, ‘particularism’, has the upper hand. But challenges to this view continue to be put forth; this paper summarizes that debate and reaffirms the particularist perspective. In this paper, we address how different conceptions of what counts as a ‘conspiracy theory’ impact how one might evaluate particularism, with specific emphasis on (1) a ‘simple definition’ of a conspiracy theory as a theory that posits a conspiracy, and (2) ‘contrarian conspiracy theories’ (those that contrast with an official account).

In Milgram's experiments, subjects were induced to inflict what they believed to be electric shocks in obedience to a man in a white coat. This suggests that many can be persuaded to torture, and perhaps kill, another person simply on the say‐so of an authority figure. But the experiments have been attacked on methodological, moral and methodologico‐moral grounds. Patten argues that the subjects probably were not taken in by the charade; Bok argues that lies should not be used in research; and Patten insists that any excuse for Milgram's conduct can be adapted on behalf of his subjects. (Either he was wrong to conduct the experiments or they do not establish the phenomenon of immoral obedience). We argue that the subjects were indeed taken in, that lies (though usually wrong) were in this case legitimate, and that there were excuses available to Milgram which were not available to his subjects. So far from ‘disrespecting’his subjects, Milgram enhanced their autonomy as rational agents. We concede however that it might be right to prohibit what it was right to do. [1]

A great deal of conspiracy theory research presupposes a falsehood – that conspiracy theories as such are irrational to believe – and that conspiracy theorists as such suffer from a range of cognitive defects. But since people frequently conspire, many people believe in a wide range of conspiracy theories because they themselves are historically and politically literate. Thus, research questions like ‘Why Do People Believe in Conspiracy Theories?’ (with the presupposition that there is something wrong with them if they do) are misconceived, since people believe in conspiracy theories (like many other theories) for a wide range of reasons, some good and some bad. But there are legitimate research questions here such as ‘Why do so many people believe in irrational conspiracy theories?’ But you can’t answer such questions systematically without a robust set of criteria for determining which kinds of conspiracy theories are irrational to believe and which are not. Furthermore, you need to keep the whole question in mind in conducting your research. If this were done, then it might very well turn out that people believe in irrational conspiracy theories for much the same reasons that they believe in irrational theories of other kinds (for example, religious theories).

Hume contends that you can’t get an ought from an is. Searle professed to prove otherwise, deriving a conclusion about obligations from premises about promises. Since can’t derive a substantive ought from an is by logic alone, Searle is best construed as claiming that there are analytic bridge principles linking premises about promises to conclusions about obligations. But we can no more derive a moral obligation to pay up from the fact that a promise has been made than we can derive a duty to fight a duel from the fact that a challenge has been issued—just conclusions about what we ought to do according to the rules of the relevant games. Hume suggests bridge principles that would take us from the rules of the games to conclusions about duties, but these are distinctly dubious. The argument features an obtreperous earl, an anarchist philosopher and a dueling Prime Minister.

I argue that ‘conspiracy theory’ and ‘conspiracy theorist’ as commonly employed are ‘tonkish’ terms (as defined by Arthur Prior and Michael Dummett), licensing inferences from truths to falsehoods; indeed, that they are mega-tonkish terms, since their use is governed by different and competing sets of introduction and elimination rules, delivering different and inconsistent results. Thus ‘conspiracy theory’ and ‘conspiracy theorist’ do not have determinate extensions, which means that generalizations about conspiracy theories or conspiracy theorists do not have determinate truth-values. Hence conspiracy theory theory – psychological or social scientific research into conspiracy theorists and what is wrong with them – is often about as intellectually respectable as an enquiry into bastards and what makes them so mean.

Are conspiracy theorists epistemically vicious? That is the conventional wisdom. It has distinguished supporters, including Quassim Cassam, Cass Sunstein and Adrian Vermeule. For me, a trait is an epistemic virtue if leads to the discovery of salient truths and the avoidance of pernicious falsehoods, and an epistemic vice the contrary. As such epistemic virtues and vices are role‐relative, context‐relative and end‐relative. I argue that that it is not necessarily or even usually vicious to be a conspiracy theorist, even if we restrict the conspiracies in question to conspiracies on the part of Western government agencies, the reason being that many such theories are now known to be true. Finally I contend that the policy suggested by some polemicists of systematic skepticism towards conspiracy theories would be intellectually suicidal and hence vicious.

I have two aims in this paper. In §§2-4 I contend that Moore has two arguments (not one) for the view that that ‘good’ denotes a non-natural property not to be identified with the naturalistic properties of science and common sense (or, for that matter, the more exotic properties posited by metaphysicians and theologians). The first argument, the Barren Tautology Argument (or the BTA), is derived, via Sidgwick, from a long tradition of anti-naturalist polemic. But the second argument, the Open Question Argument proper (or the OQA), seems to have been Moore’s own invention and was probably devised to deal with naturalistic theories, such as Russell’s, which are immune to the Barren Tautology Argument. The OQA is valid and not (as Frankena (1939) has alleged) question-begging. Moreover, if its premises were true, it would have disposed of the desire-to-desire theory. But as I explain in §5, from 1970 onwards, two key premises of the OQA were successively called into question, the one because philosophers came to believe in synthetic identities between properties and the other because it led to the Paradox of Analysis. By 1989 a philosopher like Lewis could put forward precisely the kind of theory that Moore professed to have refuted with a clean intellectual conscience. However, in §§6-8 I shall argue that all is not lost for the OQA. I first press an objection to the desire-to-desire theory derived from Kripke’s famous epistemic argument. On reflection this argument looks uncannily like the OQA. But the premise on which it relies is weaker than the one that betrayed Moore by leading to the Paradox of Analysis. This suggests three conclusions: 1) that the desire-to-desire theory is false; 2) that the OQA can be revived, albeit in a modified form; and 3) that the revived OQA poses a serious threat to what might be called semantic naturalism.

In 1913 Russell gave up on the Moorean good. But since naturalism was not an option, that left two alternatives: the error theory and non-cognitivism. Despite a brief flirtation with the error theory Russell preferred the non-cognitivist option, developing a form of emotivism according to which to say that something is good is to express the desire that everyone should desire it. But why emotivism rather than the error theory? Because emotivism sorts better with Russell’s Fundamental Principle that the “sentences we can understand must be composed of words with whose meaning we are acquainted.” I construct an argument for emotivism featuring the Fundamental Principle that closely parallels Ayer’s verificationist argument in Language, Truth, and Logic. I contend that Russell’s argument, like Ayer’s, is vulnerable to a Moorean critique. This suggests an important moral: revisionist theories of meaning such as verificationism and the Fundamental Principle are prima facie false. Any modus ponens from such a principle to a surprising semantic conclusion (such as emotivism) is trumped by a Moorean modus tollens from the negation of the surprising semantics to the negation of the revisionist principle.

Patrick Stokes has argued that although many conspiracy theories are true, we should reject the policy of particularism (that is, the policy of investigating conspiracy theories if they are plausible and believing them if that is what the evidence suggests) and should instead adopt a policy of principled skepticism, subjecting conspiracy theories – or at least the kinds of theories that are generally derided as such – to much higher epistemic standards than their non-conspiratorial rivals, and believing them only if they are proven up to the hilt. The reason is that although some conspiracy theories are true (or otherwise epistemically kosher) there is a widespread practice of conspiracy theorizing which leads to the social acceptance of conspiracy theories which (much like Lord Byron) are mad, bad and dangerous to ‘know’, leading to radically false beliefs about the natural and political worlds (for instance to Trump’s claim that Global Warming is a ‘Chinese Hoax’). Thus Stokes is a defender of the epistemic status quo in which the punditocracy routinely dismisses and sneers at ‘conspiracy theories’, simply because they are conspiracy theories, whilst admitting sotto voce that once in a blue moon there are some conspiracy theories that are true. It is just the he would prefer the epistemic policy of the punditocracy to be the policy of the populace at large. Stokes seems to be suggesting that although particularism looks like a sensible epistemic policy (and might indeed be the correct policy in an ideal world), if it were put into practice at the level of public debate it would give a specious air of plausibility to many false and dangerous conspiracy theories with pernicious social effects. I reply that the actual moral costs of conspiracy ‘skepticism’ (as it is currently practiced) exceed the likely costs of particularism, especially if it is informed by some sensible heuristics, ruling out some conspiracy theories as crazy and ruling in others as relatively sane. For the conspiracy ‘skepticism’ as currently practiced is highly selective, pouring scorn and contempt on conspiracy theories which threaten current elites whilst giving the conspiratorial concoctions of those elites something close to a free pass.

My first paper on the Is/Ought issue. The young Arthur Prior endorsed the Autonomy of Ethics, in the form of Hume’s No-Ought-From-Is (NOFI) but the later Prior developed a seemingly devastating counter-argument. I defend Prior's earlier logical thesis (albeit in a modified form) against his later self. However it is important to distinguish between three versions of the Autonomy of Ethics: Ontological, Semantic and Ontological. Ontological Autonomy is the thesis that moral judgments, to be true, must answer to a realm of sui generis non-natural PROPERTIES. Semantic autonomy insists on a realm of sui generis non-natural PREDICATES which do not mean the same as any natural counterparts. Logical Autonomy maintains that moral conclusions cannot be derived from non-moral premises.-moral premises with the aid of logic alone. Logical Autonomy does not entail Semantic Autonomy and Semantic Autonomy does not entail Ontological Autonomy. But, given some plausible assumptions Ontological Autonomy entails Semantic Autonomy and given the conservativeness of logic – the idea that in a valid argument you don’t get out what you haven’t put in – Semantic Autonomy entails Logical Autonomy. So if Logical Autonomy is false – as Prior appears to prove – then Semantic and Ontological Autonomy would appear to be false too! I develop a version of Logical Autonomy (or NOFI) and vindicate it against Prior’s counterexamples, which are also counterexamples to the conservativeness of logic as traditionally conceived. The key concept here is an idea derived in part from Quine - that of INFERENCE-RELATIVE VACUITY. I prove that you cannot derive conclusions in which the moral terms appear non-vacuously from premises from which they are absent. But this is because you cannot derive conclusions in which ANY (non-logical) terms appear non-vacuously from premises from which they are absent Thus NOFI or Logical Autonomy comes out as an instance of the conservativeness of logic. This means that the reverse entailment that I have suggested turns out to be a mistake. The falsehood of Logical Autonomy would not entail either the falsehood Semantic Autonomy or the falsehood of Ontological Autonomy, since Semantic Autonomy only entails Logical Autonomy with the aid of the conservativeness of logic of which Logical Autonomy is simply an instance. Thus NOFI or Logical Autonomy is vindicated, but it turns out to be a less world-shattering thesis than some have supposed. It provides no support for either non-cognitivism or non-naturalism.

This includes a methodological meditation (in blank verse) on the history of philosophy as a contribution to philosophy (rather than as a contribution to history) plus a conspectus of the issues surrounding Hume, the Motivation Argument and the Slavery of Reason Thesis. However I am posting it here mainly because it contains a novel restatement of the Argument from Queerness. Big Thesis: the Slavery of Reason Thesis (via the Motivation Argument) provides no support for non-cognitivism or emotivism, but there is a plausible version of the Slavery of Reason Thesis that provides some support for the Error Theory. As in other papers I stress the importance of DTADs, dispositions to acquire desires as well as desires conceived as propositional attitudes.

Hume is widely regarded as the grandfather of emotivism and indeed of non-cognitivism in general. For the chief argument for emotivism - the Argument from Motivation - is derived from him. In my opinion Hume was not an emotivist or proto-emotivist but a moral realist in the modern ‘response-dependent’ style. But my interest in this paper is not the historical Hume but the Hume of legend since the legendary Hume is one of the most influential philosophers of the present age. According to Michael Smith ‘the Moral Problem’ – the central issue in meta-ethics - is that the premises of Hume’s argument appear to be true though the non-cognitivist conclusion appears to be false. Since the argument seems to be valid, something has got to give. Smith struggles to solve the problem by holding on to something like the premises of the argument whilst trying to fend off the conclusion. In my view this is a wasted effort. Hume was not arguing for non-cognitivsm in the first place, and the arguments for non-cognitivism that can be extracted from his writings are no good. Either the premises are false or the inferences are invalid. And this is despite the fact that Hume was substantially right about reason and the passions. Thus ‘the Moral Problem’ is not a problem, and the legendary Hume does not deserve his influence. An important theme in this paper is the concept of a DTAD or a dispositions to acquire desires. These play an important role in motivation but unlike desires (with which they are sometimes confused ) they are NOT propositional attitudes

This paper deals with what I take to be one woman’s literary response to a philosophical problem. The woman is Jane Austen, the problem is the rationality of Hume’s ‘sensible knave’, and Austen’s response is to deepen the problem. Despite his enthusiasm for virtue, Hume reluctantly concedes in the EPM that injustice can be a rational strategy for ‘sensible knaves’, intelligent but selfish agents who feel no aversion towards thoughts of villainy or baseness. Austen agrees, but adds that ABSENT CONSIDERATIONS OF A FUTURE STATE, other vices besides injustice can be rationally indulged with tolerable prospects of worldly happiness. Austen’s creation Mr Elliot in Persuasion is just such an agent – sensible and knavish but not technically ‘unjust’. Despite and partly because of his vices – ingratitude, avarice and duplicity – he manages to be both successful and reasonably happy. There are plenty of other reasonably happy knaves in Jane Austen, some of whom are not particularly sensible. This is not to say that either Austen or Hume is in favor of knavery It is just that they both think that only those with the right sensibility can be argued out of it.

Milgram’s experiments, subjects were induced to inflict what they believed to be electric shocks in obedience to a man in a white coat. This suggests that many of us can be persuaded to torture, and perhaps kill, another person simply on the say-so of an authority figure. But the experiments have been attacked on methodological, moral and methodologico-moral grounds. Patten argues that the subjects probably were not taken in by the charade; Bok argues that lies should not be used in research; and Patten insists that any excuse for Milgram’s conduct can be adapted on behalf of his subjects. (Either he was wrong to conduct the experiments or they do not establish the phenomenon of immoral obedience). We argue a) that the subjects were indeed taken in b) that there are good historical reasons for regarding the experiments as ecologically valid, c) that lies (though usually wrong) were in this case legitimate, d) that there were excuses available to Milgram which were not available to his subjects and e) that even if he was wrong to conduct the experiments this does not mean that he failed to establish immoral obedience. So far from ‘disrespecting’ his subjects, Milgram enhanced their autonomy as rational agents. We concede however that it might (now) be right to prohibit what it was (then) right to do.

The Analytic Tradition in Philosophy is an excellent successor to an excellent book : It is a fine an example of the necromantic style in the history of philosophy where the object of the exercise is to resurrect the mighty dead in order to get into an argument with them, either because we think them importantly right or instructively wrong. However what was a pardonable a simplification and a reasonable omission in the earlier book has now metamorphosed into a sin of omission and an oversimplification in its successor—the book lacks is an adequate discussion of the Ramified Theory of Types and Russell’s reasons for adopting it. I conclude with a brief meditation on Open Question Arguments and the threats that they pose to some of the analyses proposed by Russell and Moore themselves

l. There is an antinomy in Hare's thought between Ought-Implies-Can and No-Indicatives-from-Imperatives. It cannot be resolved by drawing a distinction between implication and entailment. 2. Luther resolved this antinomy in the l6th century, but to understand his solution, we need to understand his problem. He thought the necessity of Divine foreknowledge removed contingency from human acts, thus making it impossible for sinners to do otherwise than sin. 3. Erasmus objected (on behalf of Free Will) that this violates Ought-Implies-Can which he supported with Hare-style ordinary language arguments. 4. Luther a) pointed out the antinomy and b) resolved it by undermining the prescriptivist arguments for Ought-Implies-Can. 5. We can reinforce Luther's argument with an example due to David Lewis. 6. Whatever its merits as a moral principle, Ought-Implies-Can is not a logical truth and should not be included in deontic logics. Most deontic logics, and maybe the discipline itself, should therefore be abandoned. 7. Could it be that Ought-Conversationally-Implies-Can? Yes - in some contexts. But a) even if these contexts are central to the evolution of Ought, the implication is not built into the semantics of the word; b) nor is the parallel implication built into the semantics of orders; and c) in some cases Ought conversationally implies Can, only because Ought-Implies-Can is a background moral belief. d) Points a) and b) suggest a criticism of prescriptivism - that Oughts do not entail imperatives but that the relation is one of conversational implicature. 8. If Ought-Implies-Can is treated as a moral principle, Erasmus' argument for Free Will can be revived (given his Christian assumptions). But it does not 'prove' Pelagianism as Luther supposed. A semi-Pelagian alternative is available.

The history of philosophy can be seen either as a contribution to history or a contribution to philosophy or perhaps as a bit of both. Hutchinson fail on both counts. The book is bad: bad in itself (since it quite definitely ought not to be) and bad as a companion to Principia (since it sets students a bad example of slapdash, lazy and pretentious philosophizing and would tend to put them off reading Moore). As a conscientious reviewer I ploughed through every page and I have to say that I resented every minute of my life that I wasted on the book. Don’t waste any of yours.

The paper reconstructs Moore's Open Question Argument (OQA) and discusses its rise and fall. There are three basic objections to the OQA: Geach's point, that Moore presupposes that ?good? is a predicative adjective (whereas it is in fact attributive); Lewy's point, that it leads straight to the Paradox of Analysis; and Durrant's point that even if 'good' is not synonymous with any naturalistic predicate, goodness might be synthetically identical with a naturalistic property. As against Geach, I argue that 'good' has both predicative and attributive uses and that in moral contexts it is difficult to give a naturalistic account of the attributive 'good'. To deal with Lewy, I reformulate the OQA. But the bulk of the paper is devoted to Durrant's objection. I argue that the post-Moorean programme of looking for synthetic identities between moral and naturalistic properties is either redundant or impossible. For it can be carried through only if 'good' expresses an empirical concept, in which case it is redundant since naturalism is true. But 'good' does not express an empirical concept (a point proved by the reformulated OQA). Hence synthetic naturalism is impossible. I discuss direct reference as a possible way out for the synthetic naturalist and conclude that it will not work. The OQA may be a bit battered but it works after a fashion.

In Part 1, I contend (using Coriolanus as my mouthpiece) that Keeley and Clarke have failed to show that there is anything intellectually suspect about conspiracy theories per se. Conspiracy theorists need not commit the ‘fundamental attribution error’ there is no reason to suppose that all or most conspiracy theories constitute the cores of degenerating research programs, nor does situationism - a dubious doctrine in itself - lend any support to a systematic skepticism about conspiracy theories. In Part 2. I argue (in propria persona) that the idea that there is something suspect about conspiracy theories is one of the most dangerous and idiotic superstitions to disgrace our political culture.

Abstract Conspiracy theories should be neither believed nor investigated - that is the conventional wisdom. I argue that it is sometimes permissible both to investigate and to believe. Hence this is a dispute in the ethics of belief. I defend epistemic “oughts” that apply in the first instance to belief-forming strategies that are partly under our control. But the beliefforming strategy of not believing conspiracy theories would be a political disaster and the epistemic equivalent of selfmutilation. I discuss several variations of this strategy, interpreting “conspiracy theory” in different ways but conclude that on all these readings, the conventional wisdom is deeply unwise