Indicative Conditionals and Conditional Probabilities

This chapter introduces and discusses the concepts of necessary and sufficient conditions.

The conventional wisdom about conditionals claims that (1) conditionals that have non-assertive acts in their consequents, such as commands and promises, are not plausibly interpreted as material implications; (2) the most promising hypothesis about these sentences is conditional-assertion theory, which explains a conditional as a conditional speech act, i.e., a performance of a speech act given the assumption of the antecedent. This hypothesis has far-reaching and revisionist consequences, because conditional speech acts are not synonymous with a proposition with truth conditions. (...) This paper argues against this prevalent view in two steps. First, it presents a battery of objections against conditional-assertion theory. Second, it argues that these examples can be convincingly interpreted as categorical assertions of material implications. (shrink)

Triviality results threaten plausible principles governing our credence in epistemic modal claims. This paper develops a new account of modal credence which avoids triviality. On the resulting theory, probabilities are assigned not to sets of worlds, but rather to sets of information state-world pairs. The theory avoids triviality by giving up the principle that rational credence is closed under conditionalization. A rational agent can become irrational by conditionalizing on new evidence. In place of conditionalization, the paper develops a new account (...) of updating: conditionalization with normalization. (shrink)

Four intuitions are recurrent and influential in theories about conditionals: the Ramsey’s test, the Adams’ Thesis, the Equation, and the robustness requirement. For simplicity’s sake, I call these intuitions ‘the big four’. My aim is to show that: (1) the big four are interdependent; (2) they express our inferential dispositions to employ a conditional on a modus ponens; (3) the disposition to employ conditionals on a modus ponens doesn’t have the epistemic significance that is usually attributed to it, since the (...) acceptability or truth conditions of a conditional is not necessarily associated with its employability on a modus ponens. (shrink)

The Equation (TE) states that the probability of A → B is the probability of B given A (Jeffrey, 1964: 702–703). Lewis has shown that the acceptance of TE implies that the probability of A → B is the probability of B, which is implausible: the probability of a conditional cannot plausibly be the same as the probability of its consequent, e.g., the probability that the match will light given that is struck is not intuitively the same as the probability (...) that it will light (Lewis, 1976: 299–300). Here I want to counter Lewis’ claim. My aim is to argue that: (1) (TE) doesn’t track the probability of A → B, but instead our willingness to employ it on a modus ponens; (2) the triviality result doesn’t strike us as implausible if our willingness to employ A → B on a modus ponens implies a similar result; (3) (TE) is still inadequate in this limited role given that some conditionals are only employable on a modus tollens or can’t be employed on a modus ponens; (4) (TE) does not have the logical significance that is usually attributed to it, since inferential disposition is a pragmatic phenomenon. (shrink)

The logic of indicative conditionals remains the topic of deep and intractable philosophical disagreement. I show that two influential epistemic norms -- the Lockean theory of belief and the Ramsey test for conditional belief -- are jointly sufficient to ground a powerful new argument for a particular conception of the logic of indicative conditionals. Specifically, the argument demonstrates, contrary to the received historical narrative, that there is a real sense in which Stalnaker's semantics for the indicative did succeed in capturing (...) the logic of the Ramseyan indicative conditional. (shrink)

This paper argues that two widely accepted principles about the indicative conditional jointly presuppose the falsity of one of the most prominent arguments against epistemological iteration principles. The first principle about the indicative conditional, which has close ties both to the Ramsey test and the “or-to-if” inference, says that knowing a material conditional suffices for knowing the corresponding indicative. The second principle says that conditional contradictions cannot be true when their antecedents are epistemically possible. Taken together, these principles entail that (...) it is impossible to be in a certain kind of epistemic state: namely, a state of ignorance about which of two partially overlapping bodies of knowledge corresponds to one’s actual one. However, some of the more popular “margin for error” style arguments against epistemological iteration principles suggest that such states are not only possible, but commonplace. I argue that the tension between these views runs deep, arising just as much for non-factive attitudes like belief, presupposition, and certainty. I also argue that this is worse news for those who accept the principles about the indicative conditional than it is for those who reject epistemological iteration principles. (shrink)

The Identity principle says that conditionals with the form 'If p, then p' are logical truths. Identity is overwhelmingly plausible, and has rarely been explicitly challenged. But a wide range of conditionals nonetheless invalidate it. I explain the problem, and argue that the culprit is the principle known as Import-Export, which we must thus reject. I then explore how we can reject Import-Export in a way that still makes sense of the intuitions that support it, arguing that the differences between (...) indicative and subjunctive conditionals play a key role in solving this puzzle. (shrink)

Chancy modus ponens is the following inference scheme: ‘probably φ’, ‘if φ, then ψ’, therefore, ‘probably ψ’. I argue that Chancy modus ponens is invalid in general. I further argue that the invalidity of Chancy modus ponens sheds new light on the alleged counterexample to modus ponens presented by McGee. I close by observing that, although Chancy modus ponens is invalid in general, we can recover a restricted sense in which this scheme of inference is valid.

Two experiments investigate two versions of Aristotle’s Thesis for the first time. Aristotle’s Thesis is a negated conditional, which consists of one propositional variable with a negation either in the antecedent or in the consequent. This task allows us to infer if people interpret indicative conditionals as material conditionals or as conditional events. In the first experiment I investigate between-participants the two versions of Aristotle’s Thesis crossed with abstract versus concrete task material. The modal response for all four groups is (...) consistent with the conditional event and inconsistentwith the material conditional interpretation. This observation is replicated in the second experiment. Moreover, the second experiment rules out scope ambiguities of the negation of conditionals. Both experiments provide new evidence against thematerial conditional interpretation of conditionals and support the conditional event interpretation. Finally, I discuss implications formodeling indicative conditionals and the relevance of this work for experimental philosophy. (shrink)

The surprise exam paradox is an apparently sound argument to the apparently absurd conclusion that a surprise exam cannot be given within a finite exam period. A closer look at the logic of the paradox shows the argument breaking down immediately. So why do the beginning stages of the argument appear sound in the first place? This paper presents an account of the paradox on which its allure is rooted in a common probabilistic mistake: the base rate fallacy. The account (...) predicts that the paradoxical argument should get less and less convincing as it goes along—a prediction I take to be welcome. On a bleaker note, the account suggests that the base rate fallacy may be more widespread than previously thought. (shrink)

Reasoning with conditionals is central to everyday life, yet there is long-standing disagreement about the meaning of the conditional. One example is the puzzle of so-called missing-link conditionals such as "if raccoons have no wings, they cannot breathe under water." Their oddity may be taken to show that conditionals require a connection between antecedent ("raccoons have no wings") and consequent ("they cannot breathe under water"), yet most accounts of conditionals attribute the oddity to natural language pragmatics. We present an experimental (...) study disentangling the pragmatic requirement of discourse coherence from a stronger notion of connection: probabilistic relevance. Results indicate that mere discourse coherence is not enough to make conditionals assertable. (shrink)

A spectre haunts the semantics of natural language — the spectre of Triviality. Semanticists (in particular Rothschild 2013; Khoo and Mandelkern 2018a,b) have entered into a holy alliance to exorcise this spectre. None, I will argue, have yet succeeded.

In this paper, new evidence is presented for the assumption that the reason-relation reading of indicative conditionals ('if A, then C') reflects a conventional implicature. In four experiments, it is investigated whether relevance effects found for the probability assessment of indicative conditionals (Skovgaard-Olsen, Singmann, and Klauer, 2016a) can be classified as being produced by a) a conversational implicature, b) a (probabilistic) presupposition failure, or c) a conventional implicature. After considering several alternative hypotheses and the accumulating evidence from other studies as (...) well, we conclude that the evidence is most consistent with the Relevance Effect being the outcome of a conventional implicature. This finding indicates that the reason-relation reading is part of the semantic content of indicative conditionals, albeit not part of their primary truth-conditional content. (shrink)

Adams' Thesis states that the probability of a conditional is the probability of the consequent conditional on the antecedent. S. Kaufmann introduced a rival method, the so-called ``local interpretation'', for calculating the probability of a conditional that, according to a purported majority, squares better with intuition in some circumstances. He also gives an example purporting to show that this new method sometimes corresponds to rational action. We challenge the intuitions and expose a mathematical error in the example. We also offer (...) a model for the local interpretivist semantics. This model puts theoretical local interpretivists on ground as solid as that of Thesis abiders for whom conditionals have truth conditions. (shrink)

Many recent theories of epistemic discourse exploit an informational notion of consequence, i.e. a notion that defines entailment as preservation of support by an information state. This paper investigates how informational consequence fits with probabilistic reasoning. I raise two problems. First, all informational inferences that are not also classical inferences are, intuitively, probabilistically invalid. Second, all these inferences can be exploited, in a systematic way, to generate triviality results. The informational theorist is left with two options, both of them radical: (...) they can either deny that epistemic modal claims have probability at all, or they can move to a nonstandard probability theory. (shrink)

KK is the thesis that if you can know p, you can know that you can know p. Though it’s unpopular, a flurry of considerations has recently emerged in its favour. Here we add fuel to the fire: standard resources allow us to show that any failure of KK will lead to the knowability and assertability of abominable indicative conditionals of the form ‘If I don’t know it, p’. Such conditionals are manifestly not assertable—a fact that KK defenders can easily (...) explain. I survey a variety of KK-denying responses and find them wanting. Those who object to the knowability of such conditionals must either deny the possibility of harmony between knowledge and belief, or deny well-supported connections between conditional and unconditional attitudes. Meanwhile, those who grant knowability owe us an explanation of such conditionals’ unassertability—yet no successful explanations are on offer. Upshot: we have new evidence for KK. (shrink)

Bradley offers a quick and convincing argument that no Boolean semantic theory for conditionals can validate a very natural principle concerning the relationship between credences and conditionals. We argue that Bradley’s principle, Preservation, is, in fact, invalid; its appeal arises from the validity of a nearby, but distinct, principle, which we call Local Preservation, and which Boolean semantic theories can non-trivially validate.

This doctoral dissertation investigates what influence indicative conditionals have on belief updating and how learning from conditionals may be modelled in a probabilistic framework. Because the problem is related to the interpretation of conditionals, we first assess different semantics of indicative conditionals. We propose that conditionals should be taken as primary concepts. This allows us to defend a claim that learning a conditional is equivalent to learning that the relevant conditional probability is 1. This implies that learning a conditional can (...) be modelled as learning a material conditional. We then turn to a few competing approaches to the problem of learning from a conditional -- an approach that models these situations with regards to changes in the explanatory status of the antecedent, as minimising probability distances between the prior and posterior probability distributions, and, particularly, an approach based on epistemic entrenchment. It is argued that although all approaches are important for their pioneering contributions, none of them ultimately provides a fully satisfactory solution. We then formulate a proposal according to which learning a conditional may be modelled as learning a material conditional in conjunction with other contextually inferred or observed information. We show that the view has an intuitive appeal on the grounds of pragmatics and empirical research in psychology of reasoning. We also show that the proposed approach successfully resolves all standard cases from the literature. The proposal is then extended to left-nested conditionals. To do so, we first inspect some preliminary empirical data about actual reasoning with nested conditionals and then show that the proposed approach also successfully models these cases. We then investigate how the approach may be extended to the so-called imperative conditionals. We introduce a tripartite typology of imperative conditionals and argue that although imperatives cannot be probabilistically modelled in any straight-forward way, at least some general insights of the proposed method help us understand the influence that the three different types of imperative conditionals have on listeners’ beliefs. We conclude with a summary and some open questions that should be addressed in future work (e.g. the influence of counterfactuals on belief updating). (shrink)

Adams famously suggested that the acceptability of any indicative conditional whose antecedent and consequent are both factive sentences amounts to the subjective conditional probability of the consequent given the antecedent. The received view has it that this thesis offers an adequate partial explication of Ramsey’s test, which characterizes graded acceptability for conditionals in terms of hypothetical updates on the antecedent. Some results in van Fraassen may raise hope that this explicatory approach to Ramsey’s test is extendible to left-nested conditionals, that (...) is, conditionals whose antecedent is itself conditional in form. We argue that this interpretation of van Fraassen’s results is to be rejected. Specifically, we provide an argument from material inadequacy against a generalization of Adams’ thesis for left-nested conditionals. (shrink)

This collection introduces the reader to some of the most interesting current work on conditionals. Particular attention is paid to possible world semantics for conditionals, the role of conditional probability in helping us to understand conditionals, implicature and the material conditional, and subjunctive versus indicative conditionals. Contributors include V.H. Dudman, Dorothy Edgington, Nelson Goodman, H.P. Grice, David Lewis, and Robert Stalnaker.

Some left-nested indicative conditionals are hard to interpret while others seem fine. Some proponents of the view that indicative conditionals have No Truth Values (NTV) use their view to explain why some left-nestings are hard to interpret: the embedded conditional does not express the truth conditions needed by the embedding conditional. Left-nestings that seem fine are then explained away as cases of ad hoc, pragmatic interpretation. We challenge this explanation. The standard reasons for NTV about indicative conditionals (triviality results, Gibbardian (...) standoffs, etc.) extend naturally to NTV about biconditionals. So NTVers about conditionals should also be NTVers about biconditionals. But biconditionals embed much more freely than conditionals. If NTV explains why some left-nested conditionals are hard to interpret, why do biconditionals embed successfully in the very contexts where conditionals do not embed? (shrink)

Famous results by David Lewis show that plausible-sounding constraints on the probabilities of conditionals or evaluative claims lead to unacceptable results, by standard probabilistic reasoning. Existing presentations of these results rely on stronger assumptions than they really need. When we strip these arguments down to a minimal core, we can see both how certain replies miss the mark, and also how to devise parallel arguments for other domains, including epistemic “might,” probability claims, claims about comparative value, and so on. A (...) popular reply to Lewis's results is to claim that conditional claims, or claims about subjective value, lack truth conditions. For this strategy to have a chance of success, it needs to give up basic structural principles about how epistemic states can be updated—in a way that is strikingly parallel to the commitments of the project of dynamic semantics. (shrink)

In this paper, we claim that the problem of conditionals should be dealt with by carefully distinguishing between thinking conditional propositions and conditional thinking, i.e. thinking on the basis of some supposition. This distinction deserves further investigation, if we are to make sense of some old and new experimental data concerning the understanding and the assertion of conditional sentences. Here we will argue that some of these data seem to refute the mental models theory of conditional reasoning, setting the ground (...) for a different approach to the cognitive study of conditionals. (shrink)

I’ll describe a range of systems for nonmonotonic conditionals that behave like conditional probabilities above a threshold. The rules that govern each system are probabilistically sound in that each rule holds when the conditionals are interpreted as conditional probabilities above a threshold level specific to that system. The well-known preferential and rational consequence relations turn out to be special cases in which the threshold level is 1. I’ll describe systems that employ weaker rules appropriate to thresholds lower than 1, and (...) compare them to these two standard systems. (shrink)

In two experiments, we investigated how people interpret and reason with realistic conditionals in the form of inducements (i.e., promises and threats) and advice (i.e., tips and warnings). We found that inducements and advice differed with respect to the degree to which the speaker was perceived to have (a) control over the consequent, (b) a stake in the outcome, and (c) an obligation to ensure that the outcome occurs. Inducements and advice also differed with respect to perceived sufficiency and necessity, (...) as well as the degree to which these statements were perceived to be effective in changing the behaviour described in the antecedent of the conditional. Multiple regression analyses indicated that perceived control over the consequent, necessity, and sufficiency emerged as the best predictors of (a) the degree to which statements were perceived to be effective in changing the behaviour of the addressee, and (b) inference patterns on a conditional arguments task. (shrink)

Ernest Adams has claimed that a probabilistic account of validity gives the best account of our intuitive judgements about the validity of arguments. In particular, he claims, it has the best hope of accounting for our judgements about many arguments involving conditionals. Most of the examples in the literature on this topic have been arguments framed in the language of propositional logic. I show that once we consider arguments involving predicates and involving identity, Adams’s strategy is less successful.

This book shows how these developments have led researchers to view people's conditional reasoning behaviour more as succesful probabilistic reasoning rather ...

To qualify the truth of a proposition probabilistically is to place it within the scope of a special type of alethic modality. We expect that, as in other modal contexts, the merely probabilistic truth of an assumption in a valid inference must carry over to whatever conclusions are derived from the assumption. That expectation, however, is not always fulfilled in ordinary reasoning about conditional probabilities. There are simpler ways of illustrating what I shall call the paradoxes of conditional probabilistic reasoning (...) in ordinary language, but the following argument is a colorful example. Consider an apparently deductively valid inference, by an imaginary inmate of a penal institution supervised by a bitterly hated warden who is surrounded by dangerous criminals, including the argument’s author. (shrink)

Two experiments were conducted to investigate the roles of covariation and of causality in people's readiness to believe a conditional. The experiments used a probabilistic truth-table task (Oberauer & Wilhelm, 2003) in which people estimated the probability of a conditional given information about the frequency distribution of truth-table cases. For one group of people, belief in the conditional was determined by the conditional probability of the consequent, given the antecedent, whereas for another group it depended on the probability of the (...) conjunction of antecedent and consequent. There was little evidence that covariation, expressed as the probabilistic contrast or as the pCI rule (White, 2003), influences belief in the conditional. The explicit presence of a causal link between antecedent and consequent in a context story had a weak positive effect on belief in a conditional when the frequency distribution of relevant cases was held constant. (shrink)

The formulation of the conditional probability in classical tasks does not guarantee the effective transmission of the independence of the hit rate from the base rate. In these kinds of tasks, data are all available, but subjects are able to understand them in the specific meanings proper to a specialized language only if these are adequately transmitted. From this perspective, the partitive formulation should not be considered a facilitation, but rather, a way of effectively transmitting the conditional probability.Consider the following (...) two phrases:1 The death-rate among men is twice that for women.2 In the deaths registered last month there were twice as many men as women.Are these two different ways of saying the same or are these different events? In fact, they are different events. (Lindley 1985, p. 44). (shrink)

From a point of view like de Finetti's, what is the judgmental reality underlying the objectivistic claim that a physical magnitude X determines the objective probability that a hypothesis H is true? When you have definite conditional judgmental probabilities for H given the various unknown values of X, a plausible answer is sufficiency, i.e., invariance of those conditional probabilities as your probability distribution over the values of X varies. A different answer, in terms of conditional exchangeability, is offered for use (...) when such definite conditional probabilities are absent. (shrink)

The conditional probability of h given e is commonly claimed to be equal to the probability that h would have if e were learned. Here I contend that this general claim about conditional probabilities is false. I present a counter-example that involves probabilities of probabilities, a second that involves probabilities of possible future actions, and a third that involves probabilities of indicative conditionals. In addition, I briefly defend these counter-examples against charges that the probabilities they involve are illegitimate.