Norbert Paulo

Wie sollten selbstfahrende Autos für Dilemma-Situationen programmiert werden? Soll eine ältere Person geopfert werden, wenn dadurch ein Kind gerettet werden kann, oder eine Person, wenn dadurch zwei gerettet werden können? Dies ist weder eine Horrorgeschichte noch Science-Fiction. Das Thema dieses Buches wird sehr bald jeden direkt betreffen, ohne Ausnahme. Letztendlich wird die breite Öffentlichkeit diese Entscheidung treffen müssen, und das Ziel dieses Buches ist es, sie darauf vorzubereiten, diese Entscheidung nicht spontan oder in Panik zu treffen, sondern erst nach sorgfältiger Abwägung der relevanten Argumente. Dieses Buch soll sie in die Lage versetzen, eine fundierte Entscheidung über einige der schwierigsten moralischen Fragen zu treffen, mit denen Gesellschaften weltweit in den kommenden Jahren konfrontiert sein werden, nämlich jene über Dilemmata im Zusammenhang mit selbstfahrenden Autos.

Dieses Kapitel konfrontiert die Leserschaft mit einer unbequemen, aber unvermeidlichen Tatsache: Selbstfahrende Autos werden in Unfälle verwickelt sein. Dies unterstreicht die Notwendigkeit einer Regulierung, ist aber mit einer schweren Bürde verbunden. Menschenleben können gerettet werden, aber in einigen Fällen müssen bewusste Entscheidungen darüber getroffen werden, wem Priorität eingeräumt wird. In diesem Kapitel wird argumentiert, dass es an der Gesellschaft liegt, zu entscheiden, wie selbstfahrende Autos für moralische Dilemmata programmiert werden sollen. Es ist das erste Mal in der Geschichte der Menschheit, dass derartige Fragen entschieden werden müssen. Die Entscheidung ist nicht einfach, aber keine Entscheidung zu treffen, ist auch keine Option, da die Technologie es ermöglicht, selbstfahrende Autos für diese Situationen vorab zu programmieren. Die Entscheidung, keine Entscheidung zu treffen, ist also auch eine Entscheidung.

The controversy over moral bioenhancement has fallen into a stalemate between advocates and critics. We wish to overcome this stalemate by addressing some of the key challenges any moral enhancement project has to meet. In particular, we shall argue that current proposals are unpersuasive as they, first, fail to diagnose the often complex causes of contemporary moral maladies and, second, are premised on methodological individualism. Focusing on brains and minds neglects social and environmental factors. Solving the mega-problems of today very likely requires more than transforming individual brains, it requires structural and higher-level changes. By itself, moral bioenhancement is thus insufficient for solving these problems. Moreover, we outline some of the yet unresolved problems of legitimacy a mandatory state-run bioenhancement project faces and show why they cannot be defused through analogies with moral education. Finally, normatively less worrisome means of enhancing morality, such as psychotherapies affording self-exploration, are already available. Moral bioenhancement may thus not even be necessary for solving today’s mega problems. The overemphasis on speculative future technological fixes may cloud the view on possibilities to tackle pressing problems instantly.

Brauchen wir selbstfahrende Autos, und wie sollen sie sich in Gefahrensituationen verhalten? In diesem anschaulich geschriebenen Buch gehen Norbert Paulo und Lando Kirchmair diesen faszinierenden Fragen auf den Grund. Expert: innen gehen davon aus, dass selbstfahrende Autos wesentlich sicherer sein werden als herkömmliche Autos. Dennoch wird es weiterhin Unfälle geben, auch tödliche. Die Innovation des selbstfahrenden Autos gibt uns die neue, großartige und zugleich beunruhigende Möglichkeit, im Voraus zu entscheiden, was in welcher Unfallsituation geschehen soll. Sie wird uns zwingen zu entscheiden, wer sterben muss, um andere in Notsituationen zu retten. Dieses Buch bereitet die Leser: innen darauf vor, eine fundierte Position zu den schwierigen moralischen Dilemmata im Zusammenhang mit selbstfahrenden Autos einzunehmen. Paulo und Kirchmair zeigen, dass die ethischen und rechtlichen Herausforderungen, die selbstfahrende Autos mit sich bringen, real, neu und unvermeidlich sind - und dass sie einer Lösung bedürfen. Sie diskutieren auch die erstaunlichen Diskrepanzen zwischen den bestehenden Regeln für Kollisionen mit selbstfahrenden Autos in einigen Rechtsordnungen und empirischen Erkenntnissen darüber, wie die Menschen darüber denken, wie sie reguliert werden sollten.

Inevitably, digitization and the increasing use of intelligent programs and machines have fundamentally changed the world of work. Moreover, it is to be expected that trends will continue in the near future and that other far-reaching changes will occur. Work is such an essential part in the lives of most members of society. It is not only the primary source of income but also crucial for one’s self-fulfillment, identification, and the achievement of social recognition. Therefore, from a societal, legal, and political perspective, there is a need to understand the imminent changes, to critically reflect on them, and to draft alternative proposals to shape the future in order to provide normative guidance.

The idea of using biomedical means to make people more likely to behave morally may have a certain appeal. However, it is very hard to find two persons – let alone two moral philosophers – who agree on what it means to be moral or to act morally. After discussing some of the proposals for moral enhancements that all ethicists could agree on, I engage more closely with the recent idea of “procedural moral enhancement” that aims at improving deliberative processes instead of particular moral views, motivations, or dispositions. I argue that it is better understood as a contribution to moral epistemology and should thus be labeled “moral-epistemic enhancement”. I then defend perspective-taking as a moral epistemic capacity which can be enhanced by both traditional and non-traditional biomedical means; a capacity which almost always contributes to the epistemic value of moral decision-making. Perspective-taking seems to be an uncontroversial non-trivial capacity for moral decision-making reasonably widely shared by proponents of ethical beliefs within the academic community. The enhancement of this capacity is thus a good candidate for an uncontroversial non-trivial moral enhancement.

What should we do if climate change or global injustice require radical policy changes not supported by the majority of citizens? And what if science shows that the lacking support is largely due to shortcomings in citizens’ individual psychology such as cognitive biases that lead to temporal and geographical parochialism? Could then a plausible case for enhancing the morality of the electorate—even against their will –be made? But can a democratic government manipulate the will of the people without losing democratic legitimacy? This paper explores the problems that governmental manipulation of voters pose for democratic legitimacy and the tensions between non-manipulated input and morally acceptable output. These venerable issues of political theory resurface in light of recent suggestions to tackle today’s global mega-problems by Ingmar Persson and Julian Savulescu. They suggest that to avert the looming catastrophe, governments should alter psychological traits of the citizenry through biomedical means, from pharmaceuticals to genetics. However, we argue that a government cannot rule with democratic legitimacy if elected by a will of the people it manipulated before. Normatively, conferring power from the governed onto governors is a one-directional relation that is incompatible with manipulation. But while it is tempting to rebut suggestions to morally enhance the people as antithetical to essential ideas of democracy, swift rebuttals tend to overlook the deeper challenge: Majoritarian decision-making may lead to inacceptable outcomes. The dilemma between input and output runs through major works in political theory. Rather than wishfully ignoring the dangers of democracy, democratic theory has to provide answers.

The development of autonomous vehicles has stimulated research into preferences in sacrificial moral dilemmas. Studies such as the Moral Machine Experiment, which has attracted enormous attention in both philosophy and psychology, have been criticized for an important methodological flaw: forcing participants to choose between two options, like killing a man or a woman. It has been shown that many people actually prefer a third option, namely treating people “equally”. While this is an important improvement, we argue that “equal” treatment can be understood in different ways. Instead of “equality”, we propose to use the concept of impartiality and argue that impartiality needs to be complemented with action-guiding decision rules. To support this conceptual point, we conducted a vignette study. Our results suggest that impartiality is indeed the main preference and that the most attractive decision rule to complement impartiality is not random choice, as many papers suggest, but inaction.

This chapter starts from Owen Flanagan’s Principle of Minimal Psychological Realism for first-order moral theory. The author argues that a similar principle of psychological realism also applies to (second-order) moral epistemic decision procedures which are used to determine the proper contents of first-order morality. He calls it the Principle of Minimal Psychological Realism for Moral Epistemic Decision Procedures (PRDP): “Make sure when constructing a moral epistemic decision procedure that the character and decision processing prescribed are possible, or are perceived to be possible, for creatures like us”. Just as consequentialists implicitly endorse PMPR in first-order morality, so do proponents of Rawlsian moral epistemology implicitly endorse PRDP. Applying PRDP to different elements of Rawls’ moral epistemic decision procedure, the author shows that his requirements concerning considered judgements do not violate the PRDP. Things are more complicated when it comes to the epistemic requirements concerning competent judges though. Some of them - particularly openness for belief revision and bias avoidance - are very demanding and might even violate the ought-implies-can principle. However, they do not violate the more flexible PRDP. There is one requirement in Rawls that does violate PRDP though, namely the requirement to imagine a social deliberation among a group of competent judges. PRDP does not apply to epistemic ideals that are not meant to be actually carried out. If proponents of such moral epistemic views were to develop decision procedures, PRDP would apply to these, too, the author argues.

In diesem Kapitel werden konkrete Regeln für die Programmierung selbstfahrender Autos vorgeschlagen. Diese Regeln sind das Ergebnis dessen, was in den vorangegangenen Kapiteln diskutiert wurde. Sie basieren auf der CREP-Methode, die sowohl die Präferenzen der Bevölkerung als auch traditionelle Moraltheorien und die grundlegenden Menschenrechte berücksichtigt. Sie sind notwendig, da es weiterhin Autounfälle geben wird und moralische Dilemmata unvermeidlich sind. Beispielsweise haben die meisten Menschen die moralische Präferenz, möglichst viele Menschenleben zu retten, und einige Moraltheorien stützen diese Ansicht. Das aktuell in Deutschland geltende Gesetz zum autonomen Fahren verbietet aber eine entsprechende Programmierung selbstfahrender Autos. Die Autoren argumentieren, dass das geändert werden sollte.

In diesem Kapitel werden einige der wichtigsten Argumente zur Regulierung moralischer Dilemmata bei selbstfahrenden Autos unter Berücksichtigung der Empfehlungen hochrangiger Ethik-Kommissionen und anderer Expertinnen erörtert. Zwar scheint es klar zu sein, welche Bedeutung der Risikominderung und der Priorisierung von Menschen gegenüber bloßen Gegenständen und Tieren zukommt, doch gibt es kontroverse Debatten über andere Fragen. Sollten wir, wie einige empirische Studien nahelegen, persönliche Merkmale wie den sozialen Status, das Geschlecht oder das Alter berücksichtigen, wenn wir entscheiden, wem Priorität eingeräumt werden soll? In diesem Kapitel werden einflussreiche Argumente für gegensätzliche Positionen – z. B. immer so viele Menschen wie möglich retten versus niemals Menschenleben aufrechnen – präsentiert und eingeordnet. So wird ein besseres Verständnis der Vor- und Nachteile möglicher Lösungen erreicht.

What should we do if climate change or global injustice require radical policy changes not supported by the majority of citizens? And what if science shows that the lacking support is largely due to shortcomings in citizens’ individual psychology such as cognitive biases that lead to temporal and geographical parochialism? Could then a plausible case for enhancing the morality of the electorate—even against their will –be made? But can a democratic government manipulate the will of the people without losing democratic legitimacy? This paper explores the problems that governmental manipulation of voters pose for democratic legitimacy and the tensions between non-manipulated input and morally acceptable output. These venerable issues of political theory resurface in light of recent suggestions to tackle today’s global mega-problems by Ingmar Persson and Julian Savulescu. They suggest that to avert the looming catastrophe, governments should alter psychological traits of the citizenry through biomedical means, from pharmaceuticals to genetics. However, we argue that a government cannot rule with democratic legitimacy if elected by a will of the people it manipulated before. Normatively, conferring power from the governed onto governors is a one-directional relation that is incompatible with manipulation. But while it is tempting to rebut suggestions to morally enhance the people as antithetical to essential ideas of democracy, swift rebuttals tend to overlook the deeper challenge: Majoritarian decision-making may lead to inacceptable outcomes. The dilemma between input and output runs through major works in political theory. Rather than wishfully ignoring the dangers of democracy, democratic theory has to provide answers.

Critiques of opaque machine learning models, used to guide consequential decisions, are gaining traction in moral philosophy. They are said to undermine informed self-advocacy, violate duties of consideration to decision-subjects, or reinforce existing power structures, among other things. Even though motivated by different normative frameworks, the received view is that the appropriate amelioration strategy is to make opaque machine learning models explainable. We challenge the received view. Focusing on Vredenburgh’s account of the ‘right to explanation’, we argue that the goals of informed self-advocacy cannot plausibly be satisfied by a combination of causal and rule-based normative explanations. Here we make three arguments in particular: (i) we highlight problems concerning the feasibility of causal explanations of the machine learning output for the purpose of social prediction; (ii) the relevant explanations are likely to be intellectually over-demanding since their correct interpretation will require a combination of domain-knowledge and statistical proficiency that cannot be presumed for laypersons; and (iii) shifting the burden to detect inadequacies to laypeople makes scrutinizing explanations very costly for them. Drawing on the empirical and philosophical literature on informed consent, we argue that it is more appropriate to revise the goal of the right to explanation from an interest in informed self-advocacy to enabling public trust: The public should be justified in believing that it can count on an institution using machine learning models competently, and that the goals for which they are used are morally legitimate.

Die Philosophie ist eine merkwürdige Disziplin. Sie hat keinen klar bestimmten Gegenstandsbereich und keine einheitliche Methodik. Es gibt aber eine philosophische Methode, die in fast allen Epochen und in fast allen Bereichen der Philosophie intensiv genutzt wurde und weiter genutzt wird: das Gedankenexperiment. Was genau Gedankenexperimente sind, wie sie funktionieren und welchen wissenschaftlichen Wert sie haben, wird in der Philosophie seit den 1990er Jahren intensiv diskutiert. Diese Diskussion ist jedoch stark auf Gedankenexperimente in den Naturwissenschaften und in der theoretischen Philosophie fokussiert, obwohl sie auch in der praktischen Philosophie allgegenwärtig sind. Diesem Phänomen ist der Schwerpunkt „Gedankenexperimente in der praktischen Philosophie“ gewidmet. Diese Einleitung stellt die einzelnen Beiträge des Schwerpunkts vor und diskutiert deren jeweiligen Kontext.

In 2021, Germany passed the first law worldwide that regulates dilemma situations with autonomous cars. Against this background, this article investigates the permissibility of trade-offs between human lives in the context of self-driving cars. It does so by drawing on the debate about the traditional trolley problem. In contrast to most authors in the relevant literature, it argues that the debate about the trolley problem is both directly and indirectly relevant for the ethics of crashes with self-driving cars. Drawing on its direct normative relevance, the article shows that trade-offs are permissible in situations with self-driving cars that are similar to paradigmatic trolley cases. In scenarios that are unlike paradigmatic trolley cases, the debate about the trolley problem can have indirect normative relevance because it provides reasons against the use of moral theories and principles that cannot account for the trolley problem.

Empirically minded philosophers and scientists have recently challenged the traditional view that legal and moral decision making are guided by reason rather than emotion. The rationalistic ideal no longer seems to be an appropriate picture of normative decision making. This paper uses the work of Joshua Greene, a philosophically trained psychologist, to exemplarily introduce some of the challenges for the rationalistic ideal from the point of view of empirical ethics. An outline of Greene’s empirical research is followed by a detailed examination of the arguments for and against the normative implications of this research. It is argued that legal scholars as well as ethicists should seriously engage with the recent advances in empirical research concerning normative decision making. Prompting this engagement is the underlying aim of this article.

This paper discusses “impartiality thought experiments”, i.e., thought experiments that attempt to generate intuitions which are unaffected by personal characteristics such as age, gender or race. We focus on the most prominent impartiality thought experiment, the Veil of Ignorance (VOI), and show that both in its original Rawlsian version and in a more generic version, empirical investigations can be normatively relevant in two ways: First, on the assumption that the VOI is effective and robust, if subjects dominantly favor a certain normative judgment behind the VOI this provides evidence in favor of that judgment; if, on the other hand, they do not dominantly favor a judgment this reduces our justification for it. Second, empirical investigations can also contribute to assessing the effectiveness and robustness of the VOI in the first place, thereby supporting or undermining its applications across the board.

Reflective equilibrium has been criticized for various reasons ever since the publication of Rawls’ A Theory of Justice. Recent empirical research into moral decision-making poses new challenges to RE because it questions the reliability of moral intuitions. This research might discredit moral intuitionism in general and RE in particular insofar as it ascribes epistemic value to moral intuitions. These findings suggest, for instance, that moral intuitions vary with cultural background, gender or framing. If it could be shown that all or certain kinds of moral intuitions are unreliable, this would diminish the plausibility of the claim that they provide good reasons to believe in the truth of their content, especially given the growing evidence that professional philosophers are just as biased and susceptible to distorting factors as lay people. In this paper I critically discuss three recent defenses of RE against this unreliable intuitions objection. Two of these defenses argue that the UIO misses its target, because RE does not depend on intuitions, or because the relevant considered judgments in RE are not intuitions. The third defense grants that RE depends on intuitions and that some Rawlsian considered judgments are intuitions. But it argues that the notion of considered judgments can be revised, and that wide reflective equilibrium has the means to adopt the criticism because it can use information about unreliable intuitions and exclude those from further consideration. I argue that the first two defenses do not succeed. The third defense is the most promising one. However, I argue that this defense also fails. In this defense, RE is not a distinctive moral epistemic theory: different versions of RE would be conceivable, all of which are deeply problematic. I conclude with a discussion of the broader philosophical implications of the UIO, specifically whether or not accepting the UIO would warrant skepticism regarding moral knowledge.

Many contemporary ethicists use case-based reasoning to reach consistent beliefs about ethical matters. The idea is that particular cases elicit moral intuitions, which provide defeasible reasons to believe in their content. However, most proponents of case-based moral reasoning are not very explicit about how they resolve inconsistencies and how they abstract principles from judgments about particular cases. The aim of this article is to outline a methodology—called Consistency Reasoning Casuistry—for case-based reasoning in ethics. This methodology draws on Richmond Campbell and Victor Kumar’s naturalistic model for the resolution of inconsistencies between the content of intuitions about particular cases. I argue that reasons similar to those that motivate their model also support a more abstract form of moral reasoning that goes beyond mere resolutions of inconsistencies between case judgments and demands the formulation of more abstract moral norms. Consistency Reasoning Casuistry, it is argued, is a good candidate for a methodology for case-based moral reasoning that is in harmony with paradigms of contemporary moral psychology and that can accommodate the methodology implicit in the work of many contemporary ethicists.

The moral psychologist Joshua Greene has proposed a number of arguments for the normative significance of empirical research and for the unreliability of deontological intuitions. For these arguments, much hinges on the combination of various components of Greene's research – namely the dual-process theory of moral judgement, ‘personalness’ as a factor in moral decision-making, and his functional understanding of deontology and consequentialism. Incorporating these components, I reconstruct three distinct arguments and show that the Personalness Argument for the claim that empirical research can advance normative ethics and the Combined Argument against deontology are both sound and interesting in themselves. They do not, however, cast doubt on traditional deontology or reserve a specific role for neuroscience. The Indirect Route argument overcomes some of the other arguments’ limitations. It is, however, invalid. I conclude by pointing out the broader philosophical relevance of Greene's arguments as shedding light on second-order morality.