There are fundamental differences between the explanation of scientific change and the explanation of technological change. The differences arise from fundamental differences between scientific and technological knowledge and basic disanalogies between technological advance and scientific progress. Given the influence of economic markets and industrial and institutional structures on the development of technology, it is more plausible to regard technological change as a continuous and incremental process, rather than as a process of Kuhnian crises and revolutions.

There are fundamental differences between the explanation of scientific change and the explanation of technological change. The differences arise from fundamental differences between scientific and technological knowledge and basic disanalogies between technological advance and scientific progress. Given the influence of economic markets and industrial and institutional structures on the development of technology, it is more plausible to regard technological change as a continuous and incremental process, rather than as a process of Kuhnian crises and revolutions.

Applications of artificial intelligence (AI) for cybersecurity tasks are attracting greater attention from the private and the public sectors. Estimates indicate that the market for AI in cybersecurity will grow from US$1 billion in 2016 to a US$34.8 billion net worth by 2025. The latest national cybersecurity and defence strategies of several governments explicitly mention AI capabilities. At the same time, initiatives to define new standards and certification procedures to elicit users’ trust in AI are emerging on a global scale. (...) However, trust in AI (both machine learning and neural networks) to deliver cybersecurity tasks is a double edged sword: it can improve substantially cybersecurity practices, but can also facilitate new forms of attacks to the AI applications themselves, which may pose severe security threats. We argue that trust in AI for cybersecurity is unwarranted and that, to reduce security risks, some form of control to ensure the deployment of ‘reliable AI’ for cybersecurity is necessary. To this end, we offer three recommendations focusing on the design, development and deployment of AI for cybersecurity. (shrink)

This article argues that an ethical framework will help to harness the potential of AI while keeping humans in control.

This article presents the first, systematic analysis of the ethical challenges posed by recommender systems through a literature review. The article identifies six areas of concern, and maps them onto a proposed taxonomy of different kinds of ethical impact. The analysis uncovers a gap in the literature: currently user-centred approaches do not consider the interests of a variety of other stakeholders—as opposed to just the receivers of a recommendation—in assessing the ethical impacts of a recommender system.

Online service providers —such as AOL, Facebook, Google, Microsoft, and Twitter—significantly shape the informational environment and influence users’ experiences and interactions within it. There is a general agreement on the centrality of OSPs in information societies, but little consensus about what principles should shape their moral responsibilities and practices. In this article, we analyse the main contributions to the debate on the moral responsibilities of OSPs. By endorsing the method of the levels of abstract, we first analyse the moral responsibilities (...) of OSPs in the web. These concern the management of online information, which includes information filtering, Internet censorship, the circulation of harmful content, and the implementation and fostering of human rights. We then consider the moral responsibilities ascribed to OSPs on the web and focus on the existing legal regulation of access to users’ data. The overall analysis provides an overview of the current state of the debate and highlights two main results. First, topics related to OSPs’ public role—especially their gatekeeping function, their corporate social responsibilities, and their role in implementing and fostering human rights—have acquired increasing relevance in the specialised literature. Second, there is a lack of an ethical framework that can define OSPs’ responsibilities, and provide the fundamental sharable principles necessary to guide OSPs’ conduct within the multicultural and international context in which they operate. This article contributes to the ethical framework necessary to deal with and by endorsing a LoA enabling the definition of the responsibilities of OSPs with respect to the well-being of the infosphere and of the entities inhabiting it. (shrink)

This article presents the first thematic review of the literature on the ethical issues concerning digital well-being. The term ‘digital well-being’ is used to refer to the impact of digital technologies on what it means to live a life that is good for a human being. The review explores the existing literature on the ethics of digital well-being, with the goal of mapping the current debate and identifying open questions for future research. The review identifies major issues related to several (...) key social domains: healthcare, education, governance and social development, and media and entertainment. It also highlights three broader themes: positive computing, personalised human–computer interaction, and autonomy and self-determination. The review argues that three themes will be central to ongoing discussions and research by showing how they can be used to identify open questions related to the ethics of digital well-being. (shrink)

This article presents the first thematic review of the literature on the ethical issues concerning digital well-being. The term ‘digital well-being’ is used to refer to the impact of digital technologies on what it means to live a life that isgood fora human being. The review explores the existing literature on the ethics of digital well-being, with the goal of mapping the current debate and identifying open questions for future research. The review identifies major issues related to several key social (...) domains: healthcare, education, governance and social development, and media and entertainment. It also highlights three broader themes: positive computing, personalised human–computer interaction, and autonomy and self-determination. The review argues that three themes will be central to ongoing discussions and research by showing how they can be used to identify open questions related to the ethics of digital well-being. (shrink)

This paper argues that there is an urgent need for an international doctrine for cyberspace skirmishes before they escalate into conventional warfare.

In this article, we focus on the attribution of moral responsibility for the actions of autonomous weapons systems (AWS). To do so, we suggest that the responsibility gap can be closed if human agents can take meaningful moral responsibility for the actions of AWS. This is a moral responsibility attributed to individuals in a justified and fair way and which is accepted by individuals as an assessment of their own moral character. We argue that, given the unpredictability of AWS, meaningful (...) moral responsibly can only be discharged by human agents who are willing to take a moral gambit: they decide to design/develop/deploy AWS despite the uncertainty about the effects an AWS may produce, hoping that unintended and unwanted or unforeseen outcomes may never occurs, but also accepting to be held responsible if such outcomes will occur. We argue that, while a moral gambit is permissible for the use of non-lethal AWS, this is not the case for the actions of lethal autonomous weapon systems. (shrink)

Accountability is a cornerstone of the governance of artificial intelligence (AI). However, it is often defined too imprecisely because its multifaceted nature and the sociotechnical structure of AI systems imply a variety of values, practices, and measures to which accountability in AI can refer. We address this lack of clarity by defining accountability in terms of answerability, identifying three conditions of possibility (authority recognition, interrogation, and limitation of power), and an architecture of seven features (context, range, agent, forum, standards, process, (...) and implications). We analyse this architecture through four accountability goals (compliance, report, oversight, and enforcement). We argue that these goals are often complementary and that policy-makers emphasise or prioritise some over others depending on the proactive or reactive use of accountability and the missions of AI governance. (shrink)

This article argues that personal medical data should be made available for scientific research, by enabling and encouraging individuals to donate their medical records once deceased, similar to the way in which they can already donate organs or bodies. This research is part of a project on posthumous medical data donation developed by the Digital Ethics Lab at the Oxford Internet Institute at the University of Oxford. Ten arguments are provided to support the need to foster posthumous medical data donation. (...) Two major risks are also identified—harm to others, and lack of control over the use of data—which could follow from unregulated donation of medical data. The argument that record-based medical research should proceed without the need to secure informed consent is rejected, and instead a voluntary and participatory approach to using personal medical data should be followed. The analysis concludes by stressing the need to develop an ethical code for data donation to minimise the risks, and offers five foundational principles for ethical medical data donation suggested as a draft code. (shrink)

An Evaluation of the 2008 Loebner Contest.

In this report we focus on the definition of autonomous weapons systems (AWS). We provide a comparative analysis of existing official definitions of AWS as provided by States and international organisations, like ICRC and NATO. The analysis highlights that the definitions draw focus on different aspects of AWS and hence lead to different approaches to address the ethical and legal problems of these weapons systems. This approach is detrimental both in terms of fostering an understanding of AWS and in facilitating (...) agreement around conditions of deployment and regulations of their use and, indeed, whether AWS are to be used at all. We draw from the comparative analysis to identify essential aspects of AWS and then offer a definition that provides a value-neutral ground to address the relevant ethical and legal problems. In particular, we identify four key aspects—autonomy; adapting capabilities of AWS; human control; and purpose of use—as the essential factors to define AWS and which are key when considering the related ethical and legal implications. (shrink)

Machine learning platforms have emerged as a new promissory technology that some argue will revolutionize work practices across a broad range of professions, including medical care. During the past few years, IBM has been testing its Watson for Oncology platform at several oncology departments around the world. Published reports, news stories, as well as our own empirical research show that in some cases, the levels of concordance over recommended treatment protocols between the platform and human oncologists have been quite low. (...) Other studies supported by IBM claim concordance rates as high as 96%. We use the Watson for Oncology case to examine the practice of using concordance levels between tumor boards and a machine learning decision-support system as a form of evidence. We address a challenge related to the epistemic authority between oncologists on tumor boards and the Watson Oncology platform by arguing that the use of concordance levels as a form of evidence of quality or trustworthiness is problematic. Although the platform provides links to the literature from which it draws its conclusion, it obfuscates the scoring criteria that it uses to value some studies over others. In other words, the platform “black boxes” the values that are coded into its scoring system. (shrink)

The “struggle between liberties and authorities”, as described by Mill, refers to the tension between individual rights and the rules restricting them that are imposed by public authorities exerting their power over civil society. In this paper I argue that contemporary information societies are experiencing a new form of such a struggle, which now involves liberties and authorities in the cyber-sphere and, more specifically, refers to the tension between cyber-security measures and individual liberties. Ethicists, political philosophers and political scientists have (...) long debated how to strike an ethically sound balance between security measures and individual rights. I argue that such a balance can only be reached once individual rights are clearly defined, and that such a definition cannot prescind from an analysis of individual well-being in the information age. Hence, I propose an analysis of individual well-being which rests on the capability approach, and I then identify a set of rights that individuals should claim for themselves. Finally, I consider a criterion for balancing the proposed set of individual rights with cyber-security measures in the information age. (shrink)

Business scholars have recently proposed that the virtue of personal wisdom may predict leadership behaviors and the quality of leader–follower relationships. This study investigated relationships among leaders’ personal wisdom—defined as the integration of advanced cognitive, reflective, and affective personality characteristics (Ardelt, Hum Dev 47:257–285, 2004)—transformational leadership behaviors, and leader–member exchange (LMX) quality. It was hypothesized that leaders’ personal wisdom positively predicts LMX quality and that intellectual stimulation and individualized consideration, two dimensions of transformational leadership, mediate this relationship. Data came from (...) 75 religious leaders and 1–3 employees of each leader (N = 158). Results showed that leaders’ personal wisdom had a positive indirect effect on follower ratings of LMX quality through individualized consideration, even after controlling for Big Five personality traits, emotional intelligence, and narcissism. In contrast, intellectual stimulation and the other two dimensions of transformational leadership (idealized influence and inspirational motivation) did not mediate the positive relationship between leaders’ personal wisdom and LMX quality. Implications for future research on personal wisdom and leadership are discussed, and some tentative suggestions for leadership development are outlined. (shrink)

In this article, I analyse deterrence theory and argue that its applicability to cyberspace is limited and that these limits are not trivial. They are the consequence of fundamental differences between deterrence theory and the nature of cyber conflicts and cyberspace. The goals of this analysis are to identify the limits of deterrence theory in cyberspace, clear the ground of inadequate approaches to cyber deterrence, and define the conceptual space for a domain-specific theory of cyber deterrence, still to be developed.

In this article, we focus on the scholarly and policy debate on autonomous weapon systems and particularly on the objections to the use of these weapons which rest on jus ad bellum principles of proportionality and last resort. Both objections rest on the idea that AWS may increase the incidence of war by reducing the costs for going to war or by providing a propagandistic value. We argue that whilst these objections offer pressing concerns in their own right, they suffer (...) from important limitations: they overlook the difficulties of calculating ad bellum proportionality; confuse the concept of proportionality of effects with the precision of weapon systems; disregard the ever-changing nature of war and of its ethical implications; mistake the moral obligation imposed by the principle of last resort with the impact that AWS may have on political decision to resort to war. Our analysis does not entail that AWS are acceptable or justifiable, but it shows that ad bellum principles are not the best set of ethical principles for tackling the ethical problems raised by AWS; and that developing adequate understanding of the transformations that the use of AWS poses to the nature of war itself is a necessary, preliminary requirement to any ethical analysis of the use of these weapons. (shrink)

Ethics Without Intention tackles the questions raised by difficult moral dilemmas by providing a critical analysis of double effect and its most common ethical and political applications. The book discusses the philosophical distinction between intended harm and foreseen but unintended harm. This distinction, which, according to the doctrine of double effect, makes a difference to the moral justification of actions, is widely applied to some of the most controversial ethical and political questions of our time: collateral damages in wars and (...) acts of terrorism; palliative care, euthanasia, abortion, and embryo research; self-defence, suicide, and self-sacrifice. It is also crucial to the now notorious theoretical cases of the trolley problem and the knobe effect. -/- Di Nucci approaches the doctrine of double effect from four key directions: its historical origins, which can be traced further back than the classic attribution to Aquinas; its theoretical coherence, which is the subject of a lively contemporary debate in philosophy; its moral intuitiveness, which has always been taken for granted but has recently begun to be questioned; and finally its relevance to the difficult moral and political decisions of our time. An engaging and comprehensive introduction to the doctrine of double effect. (shrink)

Defence agencies across the globe identify artificial intelligence as a key technology to maintain an edge over adversaries. As a result, efforts to develop or acquire AI capabilities for defence are growing on a global scale. Unfortunately, they remain unmatched by efforts to define ethical frameworks to guide the use of AI in the defence domain. This article provides one such framework. It identifies five principles—justified and overridable uses, just and transparent systems and processes, human moral responsibility, meaningful human control (...) and reliable AI systems—and related recommendations to foster ethically sound uses of AI for national defence purposes. (shrink)

Idealism is the view that reality is fundamentally mental. Idealism has been influential historically, but it has been neglected in contemporary metaphysical debate. This volume of 17 essays by leading philosophers rectifies the situation.

This paper provides a new analysis of e - trust , trust occurring in digital contexts, among the artificial agents of a distributed artificial system. The analysis endorses a non-psychological approach and rests on a Kantian regulative ideal of a rational agent, able to choose the best option for itself, given a specific scenario and a goal to achieve. The paper first introduces e-trust describing its relevance for the contemporary society and then presents a new theoretical analysis of this phenomenon. (...) The analysis first focuses on an agent’s trustworthiness , this one is presented as the necessary requirement for e-trust to occur. Then, a new definition of e-trust as a second-order-property of first-order relations is presented. It is shown that the second-order-property of e-trust has the effect of minimising an agent’s effort and commitment in the achievement of a given goal. On this basis, a method is provided for the objective assessment of the levels of e-trust occurring among the artificial agents of a distributed artificial system. (shrink)

I analyse an argument according to which medical artificial intelligence represents a threat to patient autonomy—recently put forward by Rosalind McDougall in the Journal of Medical Ethics. The argument takes the case of IBM Watson for Oncology to argue that such technologies risk disregarding the individual values and wishes of patients. I find three problems with this argument: it confuses AI with machine learning; it misses machine learning’s potential for personalised medicine through big data; it fails to distinguish between evidence-based (...) advice and decision-making within healthcare. I conclude that how much and which tasks we should delegate to machine learning and other technologies within healthcare and beyond is indeed a crucial question of our time, but in order to answer it, we must be careful in analysing and properly distinguish between the different systems and different delegated tasks. (shrink)

This theme issue has the founding ambition of landscaping Data Ethics as a new branch of ethics that studies and evaluates moral problems related to data (including generation, recording, curation, processing, dissemination, sharing, and use), algorithms (including AI, artificial agents, machine learning, and robots), and corresponding practices (including responsible innovation, programming, hacking, and professional codes), in order to formulate and support morally good solutions (e.g. right conducts or right values). Data Ethics builds on the foundation provided by Computer and Information (...) Ethics but, at the same time, it refines the approach endorsed so far in this research field, by shifting the Level of Abstraction of ethical enquiries, from being information-centric to being data-centric. This shift brings into focus the different moral dimensions of all kinds of data, even the data that never translate directly into information but can be used to support actions or generate behaviours, for example. It highlights the need for ethical analyses to concentrate on the content and nature of computational operations — the interactions among hardware, software, and data — rather than on the variety of digital technologies that enables them. And it emphasises the complexity of the ethical challenges posed by Data Science. Because of such complexity, Data Ethics should be developed from the start as a macroethics, that is, as an overall framework that avoids narrow, ad hoc approaches and addresses the ethical impact and implications of Data Science and its applications within a consistent, holistic, and inclusive framework. Only as a macroethics Data Ethics will provide the solutions that can maximise the value of Data Science for our societies, for all of us, and for our environments. (shrink)

Thinking is overrated: golfers perform best when distracted and under pressure; firefighters make the right calls without a clue as to why; and you are yourself ill advised to look at your steps as you go down the stairs, or to try and remember your pin number before typing it in. Just do it, mindlessly. Both empirical psychologists and the common man have long worked out that thinking is often a bad idea, but philosophers still hang on to an intellectualist (...) picture of human action. This book challenges that picture and calls on philosophers to wake up to the power of mindlessness: it is our habits, skills and conventions that help us cope with a world way too diverse for us to hope to always reinterpret it. The book presents the empirical evidence that has been accumulating over the last few decades and offers a philosophical analysis of mindless phenomena such as habits, skilled activity, automatic actions, emotional and spontaneous reactions and social conventions, arguing that traditional philosophical theories of action should be revised to do justice to this forgotten but important part of our lives: when we act mindlessly, we are free and fully rational even though we neither deliberate nor are aware of what we are doing. (shrink)

Judith Jarvis Thomson has recently proposed a new argument for the thesis that killing the one in the Trolley Problem is not permissible. Her argument relies on the introduction of a new scenario, in which the bystander may also sacrifice herself to save the five. Thomson argues that those not willing to sacrifice themselves if they could may not kill the one to save the five. Bryce Huebner and Marc Hauser have recently put Thomson's argument to empirical test by asking (...) people what they should do in the new trilemma case, in which they may also sacrifice themselves. They found that the majority judge that they should either kill the one or sacrifice themselves; Huebner and Hauser argue that those numbers speak against Thomson's argument. But Thomson's argument was about the dialectical effect of the new trilemma on the traditional dilemma, rather than about the trilemma itself. Here I present the results of a study in which I asked subjects first what they should do in the trilemma and then what they should do in the traditional Trolley Problem. I found that, if asked first about the trilemma, subjects then have the intuition that killing the one in the traditional Bystander at the Switch is not permissible?exactly what Thomson's argument had predicted. (shrink)

According to the Simple View of intentional action famously refuted by Bratman , φ-ing is intentional only if the agent intended to φ. In this paper I show that none of five different objections to Bratman's counter-example – McCann's , Garcia's , Sverdlik's , Stout's , and Adams's – works. Therefore Bratman's contention that SV is false still stands.

I argue against Appel's recent proposal – in this JOURNAL – that there is a fundamental human right to sexual pleasure, and that therefore the sexual pleasure of severely disabled people should be publicly funded – by thereby partially legalizing prostitution. I propose an alternative that does not need to pose a new positive human right; does not need public funding; does not need the legalization of prostitution; and that would offer a better experience to the severely disabled: charitable non-profit (...) organizations whose members would voluntarily and freely provide sexual pleasure to the severely disabled. (shrink)

This article reviews eight proposed strategies for solving the Symbol Grounding Problem (SGP), which was given its classic formulation in Harnad (1990). After a concise introduction, we provide an analysis of the requirement that must be satisfied by any hypothesis seeking to solve the SGP, the zero semantical commitment condition. We then use it to assess the eight strategies, which are organised into three main approaches: representationalism, semi-representationalism and non-representationalism. The conclusion is that all the strategies are semantically committed and (...) hence that none of them provides a valid solution to the SGP, which remains an open problem. (shrink)

According to the Simple View of intentional action famously refuted by Bratman, φ-ing is intentional only if the agent intended to φ. In this paper I show that none of five different objections to Bratman's counter-example – McCann's, Garcia's, Sverdlik's, Stout's, and Adams's – works. Therefore Bratman's contention that SV is false still stands.

In this paper I refute an apparently obvious objection to Frankfurt-type counterexamples to the Principle of Alternate Possibilities according to which if in the counterfactual scenario the agent does not act, then the agent could have avoided acting in the actual scenario. And because what happens in the counterfactual scenario cannot count as the relevant agent's actions given the sort of external control that agent is under, then we can ground responsibility on that agent having been able to avoid acting. (...) I illustrate how this objection to Frankfurt's famous counterexample is motivated by Frankfurt's own 'guidance' view of agency. My argument consists in showing that even if we concede that the agent does not act in the counterfactual scenario, that does not show that the agent could have avoided acting in the actual scenario. This depends on the crucial distinction between 'not φ-ing' and 'avoiding φ-ing'. (shrink)

In this paper I argue that there is an important anomaly to the causalist/compatibilist paradigm in the philosophy of action and free will. This anomaly, which to my knowledge has gone unnoticed so far, can be found in the philosophy of Harry Frankfurt. Two of his most important contributions to the field – his influential counterexample to the Principle of Alternate Possibilities and his ‘guidance’ view of action – are incompatible. The importance of this inconsistency goes far beyond the issue (...) of coherence within Frankfurt’s philosophy. I shall argue that this inconsistency represents an important anomaly within the causalist/compatibilist framework; so that we should start to seriously consider having to move on from the established paradigm. (shrink)

According to the Simple View of intentional action, I have intentionally switched on the light only if I intended to switch on the light. The idea that intending to is necessary for intentionally -ing has been challenged by Bratman (1984, 1987) with a counter-example in which a videogame player is trying to hit either of two targets while knowing that she cannot hit both targets. When a target is hit, the game finishes. And if both targets are about to be (...) hit simultaneously, the game shuts down. The player knows that she cannot hit both targets, but still she concludes that, given her skills, the best strategy is to have a go at each target at the same time. Suppose she hits target 1. It seems obvious that she has hit target 1 intentionally. But, Bratman argues, she could not have intended to hit target 1. Since the scenario is perfectly symmetrical, had the player intended to hit target 1, she would have also had to intend to hit target 2. But the player knows that she cannot hit both targets. (shrink)

This article analyses the ethical aspects of multistakeholder recommendation systems (RSs). Following the most common approach in the literature, we assume a consequentialist framework to introduce the main concepts of multistakeholder recommendation. We then consider three research questions: who are the stakeholders in a RS? How are their interests taken into account when formulating a recommendation? And, what is the scientific paradigm underlying RSs? Our main finding is that multistakeholder RSs (MRSs) are designed and theorised, methodologically, according to neoclassical welfare (...) economics. We consider and reply to some methodological objections to MRSs on this basis, concluding that the multistakeholder approach offers the resources to understand the normative social dimension of RSs. (shrink)

This paper aims to problematize pregnancy and support the development of a safe alternative method of gestation. Our arguments engage with the health risks of gestation and childbirth, the value assigned to pregnancy, as well as social and medical attitudes toward women’s pain, especially in labor. We claim that the harm caused by pregnancy and childbirth provides a prima facie case in favor of prioritizing research on a method of extra corporeal gestation.

We think less than we think. My thesis moves from this suspicion to show that standard accounts of intentional action can't explain the whole of agency. Causalist accounts such as Davidson's and Bratman's, according to which an action can be intentional only if it is caused by a particular mental state of the agent, don't work for every kind of action. So-called automatic actions, effortless performances over which the agent doesn't deliberate, and to which she doesn't need to pay attention, (...) constitute exceptions to the causalist framework, or so I argue in this thesis. Not all actions are the result of a mental struggle, painful hesitation, or the weighting of evidence. Through practice, many performances become second nature. Think of familiar cases such as one's morning routines and habits: turning on the radio, brushing your teeth. Think of the highly skilled performances involved in sport and music: Jarrett's improvised piano playing, the footballer's touch. Think of agents' spontaneous reactions to their environment: ducking a blow, smiling. Psychological research has long acknowledged the distinctiveness and importance of automatic actions, while philosophy has so far explained them together with the rest of agency. Intuition tells us that automatic actions are intentional actions of ours all the same (I have run a survey which shows that this intuition is widely shared): not only our own autonomous deeds for which we are held responsible, but also necessary components in the execution and satisfaction of our general plans and goals. But do standard causal accounts deliver on the intentionality of automatic actions? I think not. Because, in automatic cases, standard appeals to intentions, beliefs, desires, and psychological states in general ring hollow. We just act: we don't think, either consciously or unconsciously. On the reductive side, Davidson's view can't but appeal to, at best, unconscious psychological states, the presence and causal role of which is, I argue, inferred from the needs of a theory, rather than from evidence in the world. On the non-reductive side, Bratman agrees, with his refutation of the Simple View, that we can't just attach an intention to every action that we want to explain. But Bratman’s own Single Phenomenon View, appealing to the mysterious notion of 'motivational potential', merely acknowledges the need for refinement without actually providing one. So I propose my own account of intentional action, the 'guidance view', according to which automatic actions are intentional: differently from Davidson and Bratman, who only offer necessary conditions in order to avoid the problem of causal deviance, I offer a full-blown account: E's phi-ing is intentional if and only If phi-ing is under E's guidance. This account resembles one developed by Frankfurt, with the crucial difference that Frankfurt – taking 'acting with an intention' and 'acting intentionally' to be synonymous – thinks that guidance is sufficient only for some movement being an action, but not for some movement being an intentional action. I argue that, on the other hand, Frankfurt's concept of guidance can be developed so that it is sufficient for intentional action too. In Chapter One I present and defend my definition of ‘automatic action’. In Chapter Two I show that such understanding of automatic actions finds confirmation in empirical psychology. In Chapter Three I show that Davidson's reductive account of intentional action does not work for automatic actions. In Chapter Four I show that the two most influential non-reductive accounts of intentional action, the Simple View and Bratman's Single Phenomenon View, don't work either. And in Chapter Five I put forward and defend my positive thesis, the 'guidance view'. Also, in the Appendix I present the findings of my survey on the intentionality of automatic actions. (shrink)