This paper discusses a scenario borrowed from Williamson (2000) and repurposes it to argue for the possibility of conflict between two prima facie categorical norms of epistemic rationality: the norm to respect one’s evidence and the norm to be coherent. It is argued, pace Williamson, that in the conflict defining the scenario, the evidence norm overrides the coherence norm; that a rational agent with our evidence would lack evidence about some of their own credences; and that for agents (...) whose evidence is limited in this way, incoherence fails to entail irrationality. The above possibility claim has also been defended by Worsnip (2018), albeit on a quite different premise set and in conjunction with a coherence-centered account of epistemic rationality that issues predictions incompatible with those licensed by the evidence-centered account recommended here, as illustrated towards the end of the paper. (shrink)

In this article, a concise argument for computational rationality as a basis for artificial moral agency is advanced. Some ethicists have long argued that rational agents can become artificial moral agents. However, most of their views have come from purely philosophical perspectives, thus making it difficult to transfer their arguments to a scientific and analytical frame of reference. The result has been a disintegrated approach to the conceptualisation and design of artificial moral agents. In this article, (...) I make the argument for computational rationality as an integrative element that effectively combines the philosophical and computational aspects of artificial moral agency. This logically leads to a philosophically coherent and scientifically consistent model for building artificial moral agents. Besides providing a possible answer to the question of how to build artificial moral agents, this model also invites sound debate from multiple disciplines, which should help to advance the field of machine ethics forward. (shrink)

Hans Herzberger's 1973 essay 'Ordinal Preference and Rational Choice' is a classic milestone in the erosion of the idea that rational agents are maximizers of utility. By the time Herzberger wrote, many authors had replaced this claim with the thesis that rational agents are maximizers of preference. That is to say, it was assumed that at the moment of choice a rational agent has a weak ordering representing his or her preferences among the options (...) available to the agent for choice and that the rational agent restricts choice to one of the optimal options. Such an option is an available option judged at least as good as any other.Herzberger explored the prospects of weakening this ordinalist view of maximizing .. (shrink)

Most accounts of immigration ethics implicitly rely upon neoclassical migration theory, which understands migration as the result of poverty and unemployment in sending countries. This paper argues that neoclassical migration theory assumes an account of the human person as solely an autonomous rational agent which then leads to ethics of migration which overemphasize freedom and self-determination. This tendency to assume that migration works as neoclassical migration theory describes is shared by political philosophers, such as Joseph Carens, Michael Walzer, and (...) David Miller. This paper argues that all three philosophers incorrectly frame migration as a contest between the freedom of the migrant and the communal self-determination of the political community. Migration systems theory is presented as a theory that draws upon a relationally embedded understanding of autonomy in order to begin to develop a migration systems ethics. This paper concludes by arguing that the central ethical category for an ethics of migration is not freedom or self-determination, but justice-in-relation. (shrink)

what is now the mainstream view as to the best way forward in the dream of engineering reliable software systems out of autonomous agents. The way of using formal logics to specify, implement and verify distributed systems of interacting units using a guiding analogy of beliefs, desires and intentions. The implicit message behind the book is this: Distributed Artificial Intelligence (DAI) can be a respectable engineering science. It says: we use sound formal systems; can cite established philosophical foundations; and (...) will be able to build reliable and flexible software systems. (shrink)

This paper contends that a Kantian universalizability constraint on theories of practical reason in conjunction with the possibility of collective rational agents entails the surprisingly strong conclusion that no fully agent‐relative theory of practical reason can be sound. The basic point is that a Kantian universalizability constraint, the thesis that all reasons for action are agent‐relative and the possibility of collective rational agents gives rise to a contradiction. This contradiction can be avoided by either rejecting Kantian (...) universalizability, the possibility of collective rational agents, or the tenability of a fully agent‐relative theory of practical reason; we cannot have all three. (shrink)

Andreas Schmidt and Neil Levy have recently defended nudging against the objection that nudges fail to treat nudgees as rational agents. Schmidt rejects two theses that have been taken to support the objection: that nudges harness irrational processes in the nudgee, and that they subvert the nudgee’s rationality. Levy rejects a third thesis that may support the objection: that nudges fail to give reasons. I argue that these defences can be extrapolated from nudges to some nonconsensual neurointerventions; if (...) Schmidt’s and Levy’s defences succeed, then some nonconsensual neurointerventions neither harness irrationality, nor subvert rationality, nor fail to give reasons. This, I claim, poses a challenge both to opponents of nonconsensual neurointerventions, and to defenders of nudging. (shrink)

We define and prove a formal semantics divided into two complementary interacting components: the strictly linguistic semantics, we call linguistic agent, and the strictly logical and referential semantics, we call rational agent. This Linguistic \ Rational Agents’ Semantics applies to Deep Dependency trees or more generally, to discourses, i.e. sequences of DD-trees, and interprets them by functional structures we call Meaning Representation Structures, similar to the DRT, but interpreted very differently. LRA semantics incrementally interprets the discourses by (...) minimal finite models, called proto-models, in a monotonic logic of the LA and checks the proto-models with respect to the classical models of the RA. The proto-model is considered as the linguistic sense of the discourse. We define in full detail the LA which, as we believe, must be universal. On the other hand, we don’t propose a particular RA. We only define the scheme of interaction between the two agents and the stimuli of the RA used by the LA. After all, every discourse has in LRA semantics the single meaning and the single sense for every Rational Agent used to interact with the Linguistic Agent. (shrink)

Answering the question, ‘Who is a rational agent?’ is of utmost importance for all moral theories which conceive of the rational agent as their basic moral unit. Surprisingly enough, these theories do not pay much attention to this question, and assume, without offering detailed discussions, certain characterizations of the rational agent. In this paper, I examine what kind of attribute ‘rational’ is. In light of this examination I claim that the rational moral theories are based (...) on a mistaken characterization of the rational agent. (shrink)

Andreas Schmidt and Neil Levy have recently defended nudging against the objection that nudges fail to treat nudgees as rational agents. Schmidt rejects two theses that have been taken to support the objection: that nudges harness irrational processes in the nudgee, and that they subvert the nudgee’s rationality. Levy rejects a third thesis that may support the objection: that nudges fail to give reasons. I argue that these defences can be extrapolated from nudges to some nonconsensual neurointerventions; if (...) Schmidt’s and Levy’s defences succeed, then some nonconsensual neurointerventions neither harness irrationality, nor subvert rationality, nor fail to give reasons. This, I claim, poses a challenge both to opponents of nonconsensual neurointerventions, and to defenders of nudging. (shrink)

This book is an archetypal product of the Belief-Desire-Intention (BDI) school of multi-agent systems. It presents what is now the mainstream view as to the best way forward in the dream of engineering reliable software systems out of autonomous agents. The way of using formal logics to specify, implement and verify distributed systems of interacting units using a guiding analogy of beliefs, desires and intentions. The implicit message behind the book is this: Distributed Artificial Intelligence (DAI) can be a (...) respectable engineering science. It says: we use sound formal systems; can cite established philosophical foundations; and will be able to build reliable and flexible software systems. (shrink)

Christian List and Philip Pettit argue that some groups qualify as rational agents over and above their members. Examples include churches, commercial corporations, and political parties. According to the theory developed by List and Pettit, these groups qualify as agents because they have beliefs and desires and the capacity to process them and to act on their basis. Moreover, the alleged group agents are said to be rational to a high degree and even to be (...) fit to be held morally responsible. And the group agents under consideration are autonomous, according to the List-Pettit Theory, because their beliefs and desires cannot easily be reduced to the beliefs and desires of the groups’ members. I want to show that we should not accept the List-Pettit Theory, because it implies the absurd claim that instrument-user units, like car-driver units, are rational agents over and above their user-parts, say drivers. The focus of my argument is on whether instrument-user units are autonomous in relation to their user-parts on the List-Pettit Theory. (shrink)

In this paper, we introduce a new four-valued logic which may be viewed as a variation on the theme of Kubyshkina and Zaitsev's Logic of Rational Agent textbf{LRA} cite{LRA}. We call our logic $ bf LIRA$. In contrast to textbf{LRA}, it has three designated values instead of one and a different interpretation of truth values, the same as in Zaitsev and Shramko's bi-facial truth logic cite{ZS}. This logic may be useful in a situation when according to an agent's point (...) of view her/his reasoning is rational, while from the external one it might be not the case. One may use textbf{LIRA}, if one wants to reconstruct an agent's way of thinking, compare it with respect to the real state of affairs, and understand why an agent thought in this or that way. Moreover, we discuss Kubyshkina and Zaitsev's necessity and possibility operators for textbf{LRA} definable by means of four-valued Kripke-style semantics and show that, due to two negations of textbf{LRA}, two more possibility operators for textbf{LRA} can be defined. Then we slightly modify all these modalities to be appropriate for $bf LIRA$. Finally, we formalize all the truth-functional $ n $-ary extensions of the negation fragment of $bf LIRA$ as well as their basic modal extension via linear-type natural deduction systems. (shrink)

THERE would appear to be no philosophical consensus about the nature of human action, even though discussion of that ancient topic has intensified in the last two decades. I shall nevertheless ask the reader to suppose that the question has at last been settled in its main lines, and settled in a way I shall describe in a moment. The supposition I have in mind is no light matter. The universe it envisions is radically different from what it would be (...) if some other supposition were in its place. A detailed justification of it would require a thoroughgoing reassessment of many of the notions about causality that have prevailed in this century. Though I have offered something of that kind in earlier writings, the most I can hope to do here is to state the supposition in as brief, vivid, and concrete a way as I can manage. After that I shall ask what view of the rational agent it leads us to. No matter how action itself is interpreted, common sense attributes action to an agent. If the actions are those that take place in an atmosphere of rational discourse, it is natural to think of the agent as a rational one. Do we have any right to make this attribution in any way more exacting than that of common sense? Precisely what ontological status does the rational agent have? What, in particular, are we to make of its self-identity, seeing that some actions seem to modify the self-identity in which in a prima facie sense they originate? (shrink)

In the context of strategic interactions, individuals sometimes find themselves better off when they have fewer options. This mechanism is known under the name of ‘strategic commitment’, as it is usually the individuals themselves who ‘commit’ to following a certain course of action by restricting their options; but that is not necessary. I explain how a paternalistic intervention may be conceived where it is a third party who paternalistically restricts rational individuals’ choices to improve their welfare. This kind of (...) intervention, which I call ‘strategic paternalism’, contradicts the narrative according to which welfare economics is incompatible with paternalism because it assumes individual rationality, which would make paternalism irrelevant. To prove this point, I show why and in what sense this ‘strategic paternalism’ deserves to be called that way. (shrink)

Explanations of cognitive processes provided by traditional artificial intelligence were based on the notion of the knowledge level. This perspective has been challenged by new AI that proposes an approach based on embodied systems that interact with the real‐world. We demonstrate that these two views can be unified. Our argument is based on the assumption that knowledge level explanations can be defined in the context of Bayesian theory while the goals of new AI are captured by using a well established (...) robot based model of learning and problem solving, called Distributed Adaptive Control (DAC). In our analysis we consider random foraging and we prove that minor modifications of the DAC architecture renders a model that is equivalent to a Bayesian analysis of this task. Subsequently, we compare this enhanced, “rational,” model to its “non‐rational” predecessor and a further control condition using both simulated and real robots, in a variety of environments. Our results show that the changes made to the DAC architecture, in order to unify the perspectives of old and new AI, also lead to a significant improvement in random foraging. (shrink)

Explanations of cognitive processes provided by traditional artificial intelligence were based on the notion of the knowledge level. This perspective has been challenged by new AI that proposes an approach based on embodied systems that interact with the real‐world. We demonstrate that these two views can be unified. Our argument is based on the assumption that knowledge level explanations can be defined in the context of Bayesian theory while the goals of new AI are captured by using a well established (...) robot based model of learning and problem solving, called Distributed Adaptive Control (DAC). In our analysis we consider random foraging and we prove that minor modifications of the DAC architecture renders a model that is equivalent to a Bayesian analysis of this task. Subsequently, we compare this enhanced, “rational,” model to its “non‐rational” predecessor and a further control condition using both simulated and real robots, in a variety of environments. Our results show that the changes made to the DAC architecture, in order to unify the perspectives of old and new AI, also lead to a significant improvement in random foraging. (shrink)

Neoclassical economics assumes that individuals have stable and context-independent preferences, and uses preference satisfaction as a normative criterion. By calling this assumption into question, behavioural findings cause fundamental problems for normative economics. A common response to these problems is to treat deviations from conventional rational choice theory as mistakes, and to try to reconstruct the preferences that individuals would have acted on, had they reasoned correctly. We argue that this preference purification approach implicitly uses a dualistic model of the (...) human being, in which an inner rational agent is trapped in an outer psychological shell. This model is psychologically and philosophically problematic. (shrink)

Aristotle’s account of virtues follows the same line of argument found in his account of the good. He relies on his metaphysical theory of essence as form and function, and on his psychological theory of human function as rational agency. He uses these theories to organise, explain, defend, and modify common beliefs. In doing so, he shows that his ethical theory is not purely dialectical, but also strong dialectic.

In this paper we design a new logical system to explicitly model the different deductive reasoning steps of a boundedly rational agent. We present an adequate system in line with experimental findings about an agent’s reasoning limitations and the cognitive effort that is involved. Inspired by Dynamic Epistemic Logic, we work with dynamic operators denoting explicit applications of inference rules in our logical language. Our models are supplemented by (a) impossible worlds (not closed under logical consequence), suitably structured according (...) to the effect of inference rules, and (b) quantitative components capturing the agent’s cognitive capacity and the cognitive costs of rules with respect to certain resources (e.g. memory, time). These ingredients allow us to avoid problematic logical closure principles, while at the same time deductive reasoning is reflected in our dynamic truth clauses. We further show that our models can be reduced to awareness-like plausibility structures that validate the same formulas and we give a sound and complete axiomatization with respect to them. This approach to the agent’s internal deductive reasoning is finally combined with actions of external information. (shrink)

Aristotle has applied his dialectical method to the structure and design of whole communities. He claims that the discipline undertaken in Ethics is a political science, and continuous the same inquiry in Politics. Aristotle’s methods and doctrines in both works are examined together to provide a clearer view of the structure and connexion of the arguments and conclusions.

In this article I am concerned with whether it could be morally significant to distinguish between doing something 'in order to bring about an effect' as opposed to 'doing something because we will bring about an effect'. For example, the Doctrine of Double Effect tells us that we should not act in order to bring about evil, but even if this is true is it perhaps permissible to act only because an evil will thus occur? I discuss these questions in (...) connection with a version of the so-called Trolley Problem known as the Loop Case. I also consider how these questions may bear on whether a rational agent must aim at an event which he believes is causally necessary to achieve an end he pursues. (shrink)

Agents know some but not all logical consequences of what they know. Agents seem to be neither logically omniscient nor logically incompetent. Yet finding an intermediate standard of minimal rationality has proven difficult. In this paper, I take suggestions found in the literature (Lewis, 1988; Hawke, Özgün and Berto, 2020; Plebani and Spolaore, 2021) and join the forces of subject matter and impossible worlds approaches to devise a new solution to this quandary. I do so by combining a (...) space of FDE worlds (Berto and Jago, 2019) with a Lewisian (1988) understanding of subject matters as partitions. By doing so, I show how subject matters impose some order in the anarchic space of FDE worlds, while the worlds allow for distinctions between contents which would not otherwise be available. Combining the two approaches, then, brings us closer to the desired closure principles for knowledge of minimally rational agents. (shrink)

In this article I am concerned with whether it could be morally significant to distinguish between doing something 'in order to bring about an effect' as opposed to 'doing something because we will bring about an effect'. For example, the Doctrine of Double Effect tells us that we should not act in order to bring about evil, but even if this is true is it perhaps permissible to act only because an evil will thus occur? I discuss these questions in (...) connection with a version of the so-called Trolley Problem known as the Loop Case. I also consider how these questions may bear on whether a rational agent must aim at an event which he believes is causally necessary to achieve an end he pursues. (shrink)

In this article I am concerned with whether it could be morally significant to distinguish between doing something 'in order to bring about an effect' as opposed to 'doing something because we will bring about an effect'. For example, the Doctrine of Double Effect tells us that we should not act in order to bring about evil, but even if this is true is it perhaps permissible to act only because an evil will thus occur? I discuss these questions in (...) connection with a version of the so-called Trolley Problem known as the Loop Case. I also consider how these questions may bear on whether a rational agent must aim at an event which he believes is causally necessary to achieve an end he pursues. (shrink)

Research in psychology and behavioral economics shows that individuals’ choices often depend on “irrelevant” contextual factors. This presents problems for normative economics, which has traditionally used preference-satisfaction as its criterion. A common response is to claim that individuals have context-independent latent preferences which are “distorted” by psychological factors, and that latent preferences should be respected. This response implicitly uses a model of human action in which each human being has an “inner rational agent.” I argue that this model is (...) psychologically ungrounded. Although references to latent preferences appear in psychologically based explanations of context-dependent choice, latent preferences serve no explanatory purpose. (shrink)

Agents know some but not all logical consequences of what they know. Agents seem to be neither logically omniscient nor logically incompetent. Yet finding an intermediate standard of minimal rationality has proven difficult. In this paper, I take suggestions found in the literature (Lewis, 1988; Hawke, Özgün and Berto, 2020; Plebani and Spolaore, 2021) and join the forces of subject matter and impossible worlds approaches to devise a new solution to this quandary. I do so by combining a (...) space of FDE worlds (Berto and Jago, 2019) with a Lewisian (1988) understanding of subject matters as partitions. By doing so, I show how subject matters impose some order in the anarchic space of FDE worlds, while the worlds allow for distinctions between contents which would not otherwise be available. Combining the two approaches, then, brings us closer to the desired closure principles for knowledge of minimally rational agents. (shrink)

In this article I am concerned with whether it could be morally significant to distinguish between doing something 'in order to bring about an effect' as opposed to 'doing something because we will bring about an effect'. For example, the Doctrine of Double Effect tells us that we should not act in order to bring about evil, but even if this is true is it perhaps permissible to act only because an evil will thus occur? I discuss these questions in (...) connection with a version of the so-called Trolley Problem known as the Loop Case. I also consider how these questions may bear on whether a rational agent must aim at an event which he believes is causally necessary to achieve an end he pursues. (shrink)

Frances Kamm sets out to draw and make plausible distinctions that would show how and why it is, in some circumstances, permissible to kill some to save many more, but is not so in others. To do so she draws on a famous, and famously artificial, example of Judith Thomson, which illustrates the fact that people intutitively reject some instances of such killings but not others. The irrationality, implausibility and in many cases the self-defeating nature of such distinctions I had (...) attempted to expose in my 'The Survival Lottery' over 25 years ago. I still think these distinctions irrational and implausible and I will try, in this response, to show why this remains the case and why doctrines of additional effects, to however many powers they are taken, remain unhelpful in ethics. (shrink)

Can we coherently conceive of an agent whose practical rationality is limited to merely instrumental reasoning? I argue that we cannot. Existing arguments to this effect have focused on what is required in order to have reasons to take means to our ends-or on what is required in order to be bound by the so-called ‘instrumental principle’. By contrast, I argue that consideration of the special kind of concept-use characteristic of instrumental reasoning reveals that a merely instrumentally rational agent (...) would not be so much as able to identify means to ends in the first place. I then elucidate this line of thought by testing it against three different conceptions of merely instrumentally rational agency found in the literature. I conclude by highlighting a risk associated with remaining agnostic concerning the possibility of merely instrumentally rational agents: namely, that we arrive at a false conception of our own capacity to identify means to ends as somehow isolated from our conceptions of the good. (shrink)

The advent of functional brain imaging has revolutionized the ability to understand the biological mechanisms underlying decision-making. Although it has been amply demonstrated that assumptions of rationality often break down in experimental games, there has not been an overarching theory of why this happens. I describe recent advances in functional brain imaging and suggest a framework for considering the function of the human reward system as a discrete agent.

Frances Kamm sets out to draw and make plausible distinctions that would show how and why it is, in some circumstances, permissible to kill some to save many more, but is not so in others. To do so she draws on a famous, and famously artificial, example of Judith Thomson, which illustrates the fact that people intutitively reject some instances of such killings but not others. The irrationality, implausibility and in many cases the self-defeating nature of such distinctions I had (...) attempted to expose in my 'The Survival Lottery' over 25 years ago. I still think these distinctions irrational and implausible and I will try, in this response, to show why this remains the case and why doctrines of additional effects, to however many powers they are taken, remain unhelpful in ethics. (shrink)

ABSTRACT In this paper, I argue that a number of influential Millian responses to Frege’s puzzle, which consist in denying that Frege’s data apply to natural languages, are not viable if logic is to play its role in legitimizing the logical appraisal of rational subjects. A notion of validity which does justice to the normativity of logic must make room for a distinction between valid inferences and enthymemes. I discuss the prospects of formal, relevant and manifest validity as candidates (...) for a notion which complies with this desideratum. Their success, or failure is argued to hang on the viability of a semantical account of de jure co-reference, which is in tension with standard Millian tenets. I conclude that these Millian theories face the following dilemma: either accept that there is no notion of logical validity which makes logic normative for reasoning, thus jeopardizing our well entrenched practices of rational appraisal; or accept that de jure co-reference is a real semantical relation. (shrink)

This paper uses the exchanges between the lovers Dido and Aeneas in Aeneid IV to undercut the pretensions of Stoic philosophers to lead a dispassionate, imperturbable life under the sole guidance of “reason.” It takes Aeneas as an example of Stoicism’s lawyer-like, falsified rationality—“I will say just a few words in regard to this matter [pro re]” (IV 336)—and Dido as an example of someone who, though under the sway of furor, nevertheless makes honest, reasoned arguments that are continuous with (...) the feelings she is experiencing. The point is not that one is more at fault than the other but the rather more radical thesis that with his Aeneas character Virgil is showing that Stoicism’s ataraxia and apatheia are inevitably dissimulation, inevitably fake. (shrink)