Published & Forthcoming Papers
'Ramseyfying' Probabilistic Comparativism | Forthcoming in Philosophy of Science
Abstract: Comparativism is the view that comparative confidences (e.g., being more confident that P than that Q) are more fundamental than degrees of belief (e.g., believing that P with some strength x). In this paper, I outline the basis for a new, non-probabilistic version of comparativism inspired by a suggestion made by Frank Ramsey in `Probability and Partial Belief'. I show how, and to what extent, `Ramseyan comparativism' might be used to weaken the (unrealistically strong) probabilistic coherence conditions that comparativism traditionally relies on.
Betting Against the Zen Monk: On Preferences and Belief | Forthcoming in Synthese
The open-access published version is also available here
Abstract: According to the preference-centric approach to understanding partial belief, the connection between partial beliefs and preferences is key to understanding what partial beliefs are and how they're measured. As Ramsey put it, the 'degree of a belief is a causal property of it, which we can express vaguely as the extent to which we are prepared to act on it'. But this idea is not as popular as it once was. Nowadays, the preference-centric approach is frequently dismissed out-of-hand as behaviouristic, unpalatably anti-realist, and/or prone to devastating counterexamples. Cases like Eriksson and Hajek's (2007) preferenceless Zen monk and Christensen's (2001) other roles argument have suggested to many that any account of partial belief that ties them too closely to preferences is irretrievably flawed. In this paper I provide a defence of preference-centric accounts of partial belief.
Recent Work in Normative Decision Theory | 2019, Analysis Reviews 79 (4): 755-772
The published version is available here
Abstract: A review of some major topics of debate in normative decision theory from circa 2007 to 2019. Topics discussed include the ongoing debate between causal and evidential decision theory, decision instability, risk-weighted expected utility theory, decision-making with incomplete preferences, and decision-making with imprecise credences. (Note: the version uploaded here is slightly longer and more complete than the published version.)
Impossible Worlds and Partial Belief | 2019, Synthese 196 (8): 3433-58
The open-access published version is also available here
Abstract: One response to the problems of logical omniscience is to extend the space of possible worlds to include impossible worlds. It is natural to think that essentially the same strategy can be applied to standard probabilistic models of partial belief, for which parallel problems also arise. In this paper, I note a difficulty with the inclusion of impossible worlds into probabilistic models. With only very weak assumptions about the space of worlds, most of the propositions which can be constructed from possible and impossible worlds are in an important sense inexpressible; whereas the probabilistic model seems committed to saying that agents in general have at least as many attitudes towards inexpressible propositions as they do towards expressible propositions. If it is reasonable to think that our attitudes are generally expressible, then a model with such commitments looks problematic.
Ramsey Without Ethical Neutrality | 2017, Mind 126 (501): 1-51
The published version is available here
Abstract: Frank Ramsey’s ‘Truth and Probability’ sketches a proposal for the empirical measurement of credences, along with a corresponding set of axioms for a (somewhat incomplete) representation theorem intended to characterize the preference conditions under which this measurement process is applicable. There are several features of Ramsey’s formal system which make it attractive and worth developing. However, in specifying his measurement process and his axioms, Ramsey introduces the notion of an ethically neutral proposition, the assumed existence of which plays a key role throughout Ramsey’s system. A number of later representation theorems have also appealed to ethically neutral propositions. The notion of ethical neutrality has often been called into question—in fact, there seem to be good reasons to suppose that no ethically neutral propositions exist. In this paper, I present several new, Ramsey-inspired representation theorems that avoid any appeal to ethical neutrality. These theorems preserve the benefits of Ramsey’s system, without paying the cost of ethical neutrality.
A Representation Theorem for Frequently Irrational Agents | 2017, Journal of Philosophical Logic 46 (5): 467-506
The published version is available here
Abstract: The standard representation theorem for expected utility theory tells us that if a subject's preferences conform to certain axioms, then she can be represented as maximising her expected utility given a particular set of credences and utilities--and, moreover, that having those credences and utilities is the only way that she could be maximising her expected utility (given her preferences). However, the kinds of agents these theorems seem apt to tell us anything about are highly idealised, being (amongst other things) always probabilistically coherent with infinitely precise degrees of belief and full knowledge of all a priori truths. Ordinary subjects do not look very rational when compared to the kinds of agents usually talked about in decision theory. In this paper, I will develop an expected utility representation theorem aimed at the representation of those who are neither probabilistically coherent, logically omniscient, nor expected utility maximisers across the board--that is, agents who are frequently irrational. The agents in question may be deductively fallible, have incoherent credences, limited representational capacities, and fail to maximise expected utility for all but a limited class of gambles.
Probabilism, Representation Theorems, and Whether Deliberation Crowds out Prediction | 2017, Erkenntnis 82 (2): 379-99
The published version is available here
Abstract: Decision-theoretic representation theorems have been developed and appealed to in the service of two important philosophical projects: (i) in attempts to characterise credences in terms of preferences, and (ii) in arguments for probabilism. Theorems developed within the formal framework that Savage developed have played an especially prominent role here. I argue that the use of these 'Savagean' theorems create significant difficulties for both projects, but particularly the latter. The origin of the problem directly relates to the question of whether we can have credences regarding acts currently under consideration and the consequences which depend on those acts; I argue that such credences are possible. Furthermore, I argue that attempts to use Jeffrey's non-Savagean theorem (and similar theorems) in the service of these two projects may not fare much better.
Epistemic Two-Dimensionalism and Arguments from Epistemic Misclassification | 2013, Australasian Journal of Philosophy 91 (2): 375-89 (with Kelvin McQueen and Clas Weber).
The published version is available here
Abstract: Epistemic Two-Dimensional Semantics (E2D) claims that expressions have a counterfactual intension and an epistemic intension. Epistemic intensions reflect cognitive significance such that sentences with necessary epistemic intensions are a priori. We defend E2D against an influential line of criticism: arguments from epistemic misclassification. We focus in particular on the arguments of Speaks [2010] and Schroeter [2005]. Such arguments conclude that E2D is mistaken from (i) the claim that E2D is committed to classifying certain sentences as a priori and (ii) the claim that such sentences are a posteriori. We aim to show that these arguments are unsuccessful as (i) and (ii) undercut each other. One must distinguish the general framework of E2D from a specific implementation of it. The framework is flexible enough to avoid commitment to the apriority of any particular sentence; only specific implementations are so committed. Arguments from epistemic misclassification are therefore better understood as arguments for favouring one implementation of E2D over another, rather than as refutations of E2D.
Abstract: Comparativism is the view that comparative confidences (e.g., being more confident that P than that Q) are more fundamental than degrees of belief (e.g., believing that P with some strength x). In this paper, I outline the basis for a new, non-probabilistic version of comparativism inspired by a suggestion made by Frank Ramsey in `Probability and Partial Belief'. I show how, and to what extent, `Ramseyan comparativism' might be used to weaken the (unrealistically strong) probabilistic coherence conditions that comparativism traditionally relies on.
Betting Against the Zen Monk: On Preferences and Belief | Forthcoming in Synthese
The open-access published version is also available here
Abstract: According to the preference-centric approach to understanding partial belief, the connection between partial beliefs and preferences is key to understanding what partial beliefs are and how they're measured. As Ramsey put it, the 'degree of a belief is a causal property of it, which we can express vaguely as the extent to which we are prepared to act on it'. But this idea is not as popular as it once was. Nowadays, the preference-centric approach is frequently dismissed out-of-hand as behaviouristic, unpalatably anti-realist, and/or prone to devastating counterexamples. Cases like Eriksson and Hajek's (2007) preferenceless Zen monk and Christensen's (2001) other roles argument have suggested to many that any account of partial belief that ties them too closely to preferences is irretrievably flawed. In this paper I provide a defence of preference-centric accounts of partial belief.
Recent Work in Normative Decision Theory | 2019, Analysis Reviews 79 (4): 755-772
The published version is available here
Abstract: A review of some major topics of debate in normative decision theory from circa 2007 to 2019. Topics discussed include the ongoing debate between causal and evidential decision theory, decision instability, risk-weighted expected utility theory, decision-making with incomplete preferences, and decision-making with imprecise credences. (Note: the version uploaded here is slightly longer and more complete than the published version.)
Impossible Worlds and Partial Belief | 2019, Synthese 196 (8): 3433-58
The open-access published version is also available here
Abstract: One response to the problems of logical omniscience is to extend the space of possible worlds to include impossible worlds. It is natural to think that essentially the same strategy can be applied to standard probabilistic models of partial belief, for which parallel problems also arise. In this paper, I note a difficulty with the inclusion of impossible worlds into probabilistic models. With only very weak assumptions about the space of worlds, most of the propositions which can be constructed from possible and impossible worlds are in an important sense inexpressible; whereas the probabilistic model seems committed to saying that agents in general have at least as many attitudes towards inexpressible propositions as they do towards expressible propositions. If it is reasonable to think that our attitudes are generally expressible, then a model with such commitments looks problematic.
Ramsey Without Ethical Neutrality | 2017, Mind 126 (501): 1-51
The published version is available here
Abstract: Frank Ramsey’s ‘Truth and Probability’ sketches a proposal for the empirical measurement of credences, along with a corresponding set of axioms for a (somewhat incomplete) representation theorem intended to characterize the preference conditions under which this measurement process is applicable. There are several features of Ramsey’s formal system which make it attractive and worth developing. However, in specifying his measurement process and his axioms, Ramsey introduces the notion of an ethically neutral proposition, the assumed existence of which plays a key role throughout Ramsey’s system. A number of later representation theorems have also appealed to ethically neutral propositions. The notion of ethical neutrality has often been called into question—in fact, there seem to be good reasons to suppose that no ethically neutral propositions exist. In this paper, I present several new, Ramsey-inspired representation theorems that avoid any appeal to ethical neutrality. These theorems preserve the benefits of Ramsey’s system, without paying the cost of ethical neutrality.
A Representation Theorem for Frequently Irrational Agents | 2017, Journal of Philosophical Logic 46 (5): 467-506
The published version is available here
Abstract: The standard representation theorem for expected utility theory tells us that if a subject's preferences conform to certain axioms, then she can be represented as maximising her expected utility given a particular set of credences and utilities--and, moreover, that having those credences and utilities is the only way that she could be maximising her expected utility (given her preferences). However, the kinds of agents these theorems seem apt to tell us anything about are highly idealised, being (amongst other things) always probabilistically coherent with infinitely precise degrees of belief and full knowledge of all a priori truths. Ordinary subjects do not look very rational when compared to the kinds of agents usually talked about in decision theory. In this paper, I will develop an expected utility representation theorem aimed at the representation of those who are neither probabilistically coherent, logically omniscient, nor expected utility maximisers across the board--that is, agents who are frequently irrational. The agents in question may be deductively fallible, have incoherent credences, limited representational capacities, and fail to maximise expected utility for all but a limited class of gambles.
Probabilism, Representation Theorems, and Whether Deliberation Crowds out Prediction | 2017, Erkenntnis 82 (2): 379-99
The published version is available here
Abstract: Decision-theoretic representation theorems have been developed and appealed to in the service of two important philosophical projects: (i) in attempts to characterise credences in terms of preferences, and (ii) in arguments for probabilism. Theorems developed within the formal framework that Savage developed have played an especially prominent role here. I argue that the use of these 'Savagean' theorems create significant difficulties for both projects, but particularly the latter. The origin of the problem directly relates to the question of whether we can have credences regarding acts currently under consideration and the consequences which depend on those acts; I argue that such credences are possible. Furthermore, I argue that attempts to use Jeffrey's non-Savagean theorem (and similar theorems) in the service of these two projects may not fare much better.
Epistemic Two-Dimensionalism and Arguments from Epistemic Misclassification | 2013, Australasian Journal of Philosophy 91 (2): 375-89 (with Kelvin McQueen and Clas Weber).
The published version is available here
Abstract: Epistemic Two-Dimensional Semantics (E2D) claims that expressions have a counterfactual intension and an epistemic intension. Epistemic intensions reflect cognitive significance such that sentences with necessary epistemic intensions are a priori. We defend E2D against an influential line of criticism: arguments from epistemic misclassification. We focus in particular on the arguments of Speaks [2010] and Schroeter [2005]. Such arguments conclude that E2D is mistaken from (i) the claim that E2D is committed to classifying certain sentences as a priori and (ii) the claim that such sentences are a posteriori. We aim to show that these arguments are unsuccessful as (i) and (ii) undercut each other. One must distinguish the general framework of E2D from a specific implementation of it. The framework is flexible enough to avoid commitment to the apriority of any particular sentence; only specific implementations are so committed. Arguments from epistemic misclassification are therefore better understood as arguments for favouring one implementation of E2D over another, rather than as refutations of E2D.
Works in Progress
Comments Always welcome
Unawareness and Implicit Belief: Possible Worlds Models of Informational Content | Conditionally accepted
Abstract: Possible worlds models of belief have difficulties accounting for unawareness, the inability an agent may have to entertain (and hence believe) certain propositions. Accommodating the possibility of unawareness is important for adequately modelling epistemic states, and representing the informational content to which agents have access given their explicit beliefs. In this paper, I use neighbourhood structures to develop an original multi-agent model of explicit belief, awareness, and informational content, along with an associated sound and complete axiom system. I also defend the model against the seminal impossibility result of Dekel et al. (1998), according to which three intuitive conditions preclude non-trivial unawareness on any 'standard' model of knowledge or belief.
Comparativism and the Measurement of Partial Belief | Under revise & resubmit
Abstract: According to comparativism, degrees of belief are reducible to a system of purely ordinal comparisons of relative confidence. (For example, being more confident that P than that Q, or being equally confident that P and that Q.) In this paper, I raise several general challenges for comparativism, relating to (i) its capacity to illuminate apparently meaningful claims regarding intervals and ratios of strengths of belief, (ii) its capacity to draw enough intuitively meaningful and theoretically relevant distinctions between doxastic states, and (iii) its capacity to handle common instances of irrationality.
Representation Theorems and Radical Interpretation | Under review
Abstract: David Lewis' theory of radical interpretation for mental content is founded on two key ideas: that beliefs and desires can be understood in terms of their causal-functional roles within folk psychology, and that folk psychology is more or less Bayesian in outline. This paper concerns a puzzle for Lewis' theory. On the one hand, Lewis argued that the facts about an agent's sensory evidence and choice dispositions will always underdetermine the facts about her beliefs and desires. On the other hand, we have various representation theorems (e.g., in Ramsey 1931; Savage 1954) that are widely taken to show that if an agent's choice dispositions satisfy certain structural conditions, then those dispositions \textit{alone} suffice to determine her beliefs and desires. Here, I will argue that Lewis' conclusion is correct---any tension with representation theorems is merely apparent, and relates primarily to the difference between how 'choice dispositions' are treated in Lewis' theory and the problematic way they're understood in the context of the representation theorems. Indeed, there's no plausible sense in which theorems like Ramsey's or Savage's show that beliefs and desires can be determined by choice dispositions, even in principle---ultimately, they're of limited relevance to functionalism and to the project of radical interpretation.
What is 'Real' in Interpersonal Comparisons of Confidence | Under review
Abstract: According to comparativism, comparative confidence is strictly more fundamental than absolute confidence; indeed, comparativists think that comparative confidence ought to be treated as one of the fundamental theoretical concepts in decision theory and epistemology. In a pair of recent papers in this journal, Stefansson (2016; 2018) has defended comparativism, in particular against objections raised by Meacham and Weisberg (2011). One of those objections concerns whether comparativism is capable of adequately explaining interpersonal comparisons of confidence---for instance, whether it has the resources to make sense of one agent's having more confidence in a given proposition than another agent does. In this paper, I will argue that (i) Stefansson's proposed explanation of interpersonal confidence comparisons is inadequate; (ii) we have good general reasons to think that comparativism cannot handle intepersonal comparisons; and (iii) that the best explanation of interpersonal comparisons requires thinking about confidence in a fundamentally different way than that which comparativists propose.
Don't Stop Believing: On the Rationality of Inquiry into the Moral Error Theory (with J. Isserow) | Under review; email me for a draft
Abstract: It's natural to think that there's value to improving the accuracy of one's beliefs. If beliefs are a map by which we steer our efforts to bring the world in line with our preferences, then, all else being equal, we should want a more accurate map. However, the world could be structured so as to punish learning with respect to certain topics; i.e., by coming to learn certain things, your situation as a whole may end up worse than it would have been otherwise. In this paper, we investigate whether the world is structured so as to punish learning specifically with respect to the moral error theory. We ask: If an ordinary human agent, with ordinary human preferences, had the option to learn whether the error theory is true, should she take it? We will argue that she should not.
Lewis De-Humanized: Resolving the Problem of Radical Indeterminacy | It'll be done when it's done
Abstract: Over the course of nearly a dozen works, David Lewis developed a unique---and uniquely sophisticated---functionalist theory of beliefs and desires, centred around an essentially Bayesian picture of typical human psychology. There is a lot to like about his theory, but it's not perfect. In this paper, I will do two things. First, I will describe Lewis' theory in detail, taking care to correct some common misunderstandings. Second, I will describe some unfortunate consequences of Lewis' preferred response to radical indeterminacy, and show how that response might be improved.
Credal Sets and Strength of Belief | It'll be done when it's done
Abstract: I distinguish between several common ways to interpreting credal sets (i.e., sets of probability functions), which are often conflated--the supervaluational, interval, and intersectional interpretations. I discuss each in relation to the idea that the strength of a belief is a psychological quantity. I show that on the intersectional interpretation, but not on the supervaluational or interval interpretations, a number of very intuitive claims about strengths of belief become difficult to maintain, and the intersectional interpretation requires a more radical picture of the mind than is often realised by its proponents. Finally, I suggest the need for generalising the credal sets model, along with a mixed interval-supervaluational interpretation of the resulting model.
Abstract: Possible worlds models of belief have difficulties accounting for unawareness, the inability an agent may have to entertain (and hence believe) certain propositions. Accommodating the possibility of unawareness is important for adequately modelling epistemic states, and representing the informational content to which agents have access given their explicit beliefs. In this paper, I use neighbourhood structures to develop an original multi-agent model of explicit belief, awareness, and informational content, along with an associated sound and complete axiom system. I also defend the model against the seminal impossibility result of Dekel et al. (1998), according to which three intuitive conditions preclude non-trivial unawareness on any 'standard' model of knowledge or belief.
Comparativism and the Measurement of Partial Belief | Under revise & resubmit
Abstract: According to comparativism, degrees of belief are reducible to a system of purely ordinal comparisons of relative confidence. (For example, being more confident that P than that Q, or being equally confident that P and that Q.) In this paper, I raise several general challenges for comparativism, relating to (i) its capacity to illuminate apparently meaningful claims regarding intervals and ratios of strengths of belief, (ii) its capacity to draw enough intuitively meaningful and theoretically relevant distinctions between doxastic states, and (iii) its capacity to handle common instances of irrationality.
Representation Theorems and Radical Interpretation | Under review
Abstract: David Lewis' theory of radical interpretation for mental content is founded on two key ideas: that beliefs and desires can be understood in terms of their causal-functional roles within folk psychology, and that folk psychology is more or less Bayesian in outline. This paper concerns a puzzle for Lewis' theory. On the one hand, Lewis argued that the facts about an agent's sensory evidence and choice dispositions will always underdetermine the facts about her beliefs and desires. On the other hand, we have various representation theorems (e.g., in Ramsey 1931; Savage 1954) that are widely taken to show that if an agent's choice dispositions satisfy certain structural conditions, then those dispositions \textit{alone} suffice to determine her beliefs and desires. Here, I will argue that Lewis' conclusion is correct---any tension with representation theorems is merely apparent, and relates primarily to the difference between how 'choice dispositions' are treated in Lewis' theory and the problematic way they're understood in the context of the representation theorems. Indeed, there's no plausible sense in which theorems like Ramsey's or Savage's show that beliefs and desires can be determined by choice dispositions, even in principle---ultimately, they're of limited relevance to functionalism and to the project of radical interpretation.
What is 'Real' in Interpersonal Comparisons of Confidence | Under review
Abstract: According to comparativism, comparative confidence is strictly more fundamental than absolute confidence; indeed, comparativists think that comparative confidence ought to be treated as one of the fundamental theoretical concepts in decision theory and epistemology. In a pair of recent papers in this journal, Stefansson (2016; 2018) has defended comparativism, in particular against objections raised by Meacham and Weisberg (2011). One of those objections concerns whether comparativism is capable of adequately explaining interpersonal comparisons of confidence---for instance, whether it has the resources to make sense of one agent's having more confidence in a given proposition than another agent does. In this paper, I will argue that (i) Stefansson's proposed explanation of interpersonal confidence comparisons is inadequate; (ii) we have good general reasons to think that comparativism cannot handle intepersonal comparisons; and (iii) that the best explanation of interpersonal comparisons requires thinking about confidence in a fundamentally different way than that which comparativists propose.
Don't Stop Believing: On the Rationality of Inquiry into the Moral Error Theory (with J. Isserow) | Under review; email me for a draft
Abstract: It's natural to think that there's value to improving the accuracy of one's beliefs. If beliefs are a map by which we steer our efforts to bring the world in line with our preferences, then, all else being equal, we should want a more accurate map. However, the world could be structured so as to punish learning with respect to certain topics; i.e., by coming to learn certain things, your situation as a whole may end up worse than it would have been otherwise. In this paper, we investigate whether the world is structured so as to punish learning specifically with respect to the moral error theory. We ask: If an ordinary human agent, with ordinary human preferences, had the option to learn whether the error theory is true, should she take it? We will argue that she should not.
Lewis De-Humanized: Resolving the Problem of Radical Indeterminacy | It'll be done when it's done
Abstract: Over the course of nearly a dozen works, David Lewis developed a unique---and uniquely sophisticated---functionalist theory of beliefs and desires, centred around an essentially Bayesian picture of typical human psychology. There is a lot to like about his theory, but it's not perfect. In this paper, I will do two things. First, I will describe Lewis' theory in detail, taking care to correct some common misunderstandings. Second, I will describe some unfortunate consequences of Lewis' preferred response to radical indeterminacy, and show how that response might be improved.
Credal Sets and Strength of Belief | It'll be done when it's done
Abstract: I distinguish between several common ways to interpreting credal sets (i.e., sets of probability functions), which are often conflated--the supervaluational, interval, and intersectional interpretations. I discuss each in relation to the idea that the strength of a belief is a psychological quantity. I show that on the intersectional interpretation, but not on the supervaluational or interval interpretations, a number of very intuitive claims about strengths of belief become difficult to maintain, and the intersectional interpretation requires a more radical picture of the mind than is often realised by its proponents. Finally, I suggest the need for generalising the credal sets model, along with a mixed interval-supervaluational interpretation of the resulting model.
PhD Thesis |
Representation Theorems and the Grounds of Intentionality
This work evaluates and defends the idea that decision-theoretic representation theorems can play an important role in showing how credences and utilities can be characterised, at least in large part, in terms of their connection with preferences (i.e., characterisational representationism). Roughly, a decision-theoretic representation theorem tells us that if an agent’s preferences satisfy constraints, then that agent can be represented as maximizing her expected utility under a unique set of credences (modelled by a credence function) and utilities (modelled by a utility function). Such theorems have been thought by many to not only show how credences and utilities can be understood via their relation to preferences, but also to show how credences and utilities can be naturalised—that is, characterised in wholly non-mental, non-intentional, and non-normative terms.
There are two broad questions that are addressed. The first (and more specific) question is whether any version of characterisational representationism, based on one of the representation theorems that are currently available to us, will be of much use in directly advancing the long-standing project of showing how representational mental states can exist within the natural world. I argue that there is no current representation theorem which lends itself to a naturalistic interpretation suitable for the goal of reducing facts about credences and utilities to a naturalistic base. A naturalistic variety of characterisational representationism will have to await a new kind of representation theorem, quite different from any which have yet been developed.
The second question is whether characterisational representationism in any form (naturalistic or otherwise) is a viable position—whether, in particular, there is any value to developing representation theorems with the goal of characterizing what it is to have credences and utilities. Of this I am less sceptical. In particular, I defend a weak version of characterisational representationism against a number of philosophical critiques. With that in mind, I also argue that there are serious drawbacks with the particular theorems that decision theorists have developed thus far; particularly those which have been developed within the four basic formal frameworks developed by Savage, Anscombe and Aumann, Jeffrey, and Ramsey.
In the final part of the work, however, I develop a new representation theorem, which I argue goes some of the way towards resolving the most troubling issues associated with earlier theorems. I first show how to construct a theorem which is ontologically similar to Jeffrey’s, but formally more similar to Ramsey’s—but which does not suffer from the infamous problems associated with Ramsey’s notion of ethical neutrality, and which has stronger uniqueness results than Jeffrey’s theorem. Furthermore, it is argued that the new theorem’s preference conditions are descriptively reasonable, even for ordinary agents, and that the credence and utility functions associated with this theorem are capable of representing a wide range of non-ideal agents—including those who: (i) might have credences and utilities only towards non-specific propositions, (ii) are probabilistically incoherent, (iii) are deductively fallible, and (iv) have distinct credences and utilities towards logically equivalent propositions.
This work evaluates and defends the idea that decision-theoretic representation theorems can play an important role in showing how credences and utilities can be characterised, at least in large part, in terms of their connection with preferences (i.e., characterisational representationism). Roughly, a decision-theoretic representation theorem tells us that if an agent’s preferences satisfy constraints, then that agent can be represented as maximizing her expected utility under a unique set of credences (modelled by a credence function) and utilities (modelled by a utility function). Such theorems have been thought by many to not only show how credences and utilities can be understood via their relation to preferences, but also to show how credences and utilities can be naturalised—that is, characterised in wholly non-mental, non-intentional, and non-normative terms.
There are two broad questions that are addressed. The first (and more specific) question is whether any version of characterisational representationism, based on one of the representation theorems that are currently available to us, will be of much use in directly advancing the long-standing project of showing how representational mental states can exist within the natural world. I argue that there is no current representation theorem which lends itself to a naturalistic interpretation suitable for the goal of reducing facts about credences and utilities to a naturalistic base. A naturalistic variety of characterisational representationism will have to await a new kind of representation theorem, quite different from any which have yet been developed.
The second question is whether characterisational representationism in any form (naturalistic or otherwise) is a viable position—whether, in particular, there is any value to developing representation theorems with the goal of characterizing what it is to have credences and utilities. Of this I am less sceptical. In particular, I defend a weak version of characterisational representationism against a number of philosophical critiques. With that in mind, I also argue that there are serious drawbacks with the particular theorems that decision theorists have developed thus far; particularly those which have been developed within the four basic formal frameworks developed by Savage, Anscombe and Aumann, Jeffrey, and Ramsey.
In the final part of the work, however, I develop a new representation theorem, which I argue goes some of the way towards resolving the most troubling issues associated with earlier theorems. I first show how to construct a theorem which is ontologically similar to Jeffrey’s, but formally more similar to Ramsey’s—but which does not suffer from the infamous problems associated with Ramsey’s notion of ethical neutrality, and which has stronger uniqueness results than Jeffrey’s theorem. Furthermore, it is argued that the new theorem’s preference conditions are descriptively reasonable, even for ordinary agents, and that the credence and utility functions associated with this theorem are capable of representing a wide range of non-ideal agents—including those who: (i) might have credences and utilities only towards non-specific propositions, (ii) are probabilistically incoherent, (iii) are deductively fallible, and (iv) have distinct credences and utilities towards logically equivalent propositions.
Miscellaneous bits and pieces
1. Frank P. Ramsey: Partial Belief and Subjective Probability | Stanford Encyclopedia co-authored entry
2. 'Ramseyfying' Probabilistic Comparativism | Presented at F.E.W. 2018 and the 2019 APA Eastern Division Meeting.
3. Ramsey and the Ethically Neutral Proposition | Short unpublished exposition piece for the Leeds NatRep group.
4. The Instability of Savage's Foundations: The Constant Acts Problem | Short unpublished exposition piece for the Leeds NatRep group.
5. What is Lewisean Interpretivism | Presentation at Manchester 2019 Lewis conference.
2. 'Ramseyfying' Probabilistic Comparativism | Presented at F.E.W. 2018 and the 2019 APA Eastern Division Meeting.
3. Ramsey and the Ethically Neutral Proposition | Short unpublished exposition piece for the Leeds NatRep group.
4. The Instability of Savage's Foundations: The Constant Acts Problem | Short unpublished exposition piece for the Leeds NatRep group.
5. What is Lewisean Interpretivism | Presentation at Manchester 2019 Lewis conference.