Micha H. Werner

This article discusses the significance of “Large Language Models” and similar AI systems for the process of cultural reproduction. It argues that such systems pose a risk of cultural exploitation. Human language use resembles the cultivation of crops: based on our own purposes and values, (agri-)cultural heritage is utilized and further developed in a critical and selective manner. This selective and creative appropriation and transmission contains an element of intergenerational fairness. However, AI systems cannot evaluate and select cultural data based on their own purposes or values. They lack access to an inner world of their own, a lived practice, or a social community. The technologized appropriation and reproduction of cultural traditions therefore cannot perform critical selection tasks on its own. It shifts this burden onto users or social institutions, for whom dealing with AI-generated “slop” becomes a challenge that is difficult to manage.

Since Barry Stroud's classic paper in 1968, the general discussion on transcendental arguments tends to focus on examples from theoretical philosophy. It also tends to be pessimistic, or at least extremely reluctant, about the potential of this kind of arguments. Nevertheless, transcendental reasoning continues to play a prominent role in some recent approaches to moral philosophy. Moreover, some authors argue that transcendental arguments may be more promising in moral philosophy than they are in theoretical contexts. Against this background, the current volume focuses on transcendental arguments in practical philosophy. Experts from different countries and branches of philosophy share their views about whether there are actually differences between "theoretical" and "practical" uses of transcendental arguments. They examine and compare different versions of transcendental arguments in moral philosophy, explain their structure, and assess their respective problems and promises. This book offers all those interested in ethics, meta-ethics, or epistemology a more comprehensive understanding of transcendental arguments. It also provides them with new insights into uses of transcendental reasoning in moral philosophy.

Hatte Jeremy Bentham den von ihm geprägten Begriff deontology in seinem gleichnamigen Werk noch im etymologisch buchstäblichen Sinne (von griech. to deon: das Schickliche, die Pflicht) als »Lehre vom Sollen« verstanden und mit der »Wissenschaft von der Moral« gleichgesetzt (Bentham 1834), so dient das Adjektiv ›deontologisch‹ heutzutage ausschließlich der Kategorisierung einer spezifischen Teilklasse allgemeiner Konzeptionen normativer Ethik. ›Deontologische‹ Theorien normativer Ethik werden dann in der Regel ›teleologischen‹ Theorien (so erstmals Muirhead 1932), und/oder ›konsequentialistischen‹ Theorien (so etwa Alexander/Moore 2008) gegenübergestellt.

In der Technikethik spielt das Konzept der Verantwortung eine zentrale Rolle. ›Verantwortung‹ ist ein Basiskonzept, das, ähnlich wie die Konzepte ›Pflicht‹ oder ›Schuld‹, in vielfältigen Kontexten gebraucht wird. Dennoch lassen sich einige allgemeine Aussagen über seine Bedeutung treffen. Gerade Autoren, die im Bereich der Technik- und Wissenschaftsethik aktiv sind, haben sich nachdrücklich um eine Klärung des allgemeinen Verantwortungsbegriffs und seiner verschiedenen Aspekte und Gebrauchsweisen bemüht (z. B. Grunwald 1999; Hubig/Reidel 2003; Lenk/Maring 1991; Ott 1997, 252–255; Ropohl 1994).

The book introduces a conception of discourse ethics, an intersubjectivist version of Kantian ethics. Analysing contributions by Jürgen Habermas, Karl-Otto Apel, Wolfgang Kuhlmann, Albrecht Wellmer, Robert Alexy, Klaus Günther, Rainer Forst, Marcel Niquet and others, it reconstructs critical discussions of the justification of the principle of morality (Part I) and the various proposals for its application (Part II). It defends an alternative model of how discourse ethics can provide guidance in non-ideal circumstances and avoid both arbitrariness and rigorism.

The ongoing digital transformation of almost all areas of human action and agency calls for a readjustment of the norms that regulate these practices. For example, the digitisation of communicative practices poses new challenges to their functioning. This paper explains some of these challenges and argues that they cannot be met by a normative framework that focuses mainly on defensive (free speech and property) rights. In the context of mediated digital communication, the application of such a framework may even have paradoxical consequences. Accordingly, this paper argues for a broader system of communicative rights, understood as rights to meaningful participation in well-functioning communicative practices.

The article examines the extent to which the scope for social design of responsibility regimes is limited by pre-positive boundaries. A distinction is made between purely performative, purely constitutive and normative attributions of responsibility. It is argued that statements about pre-positive responsibilities in a broad area are only possible depending on specific normative-ethical theories. Nevertheless, the article attempts to defend both an outer limit of possible attributions of responsibility and a minimum content of any plausible ethics of responsibility. The outer limit is characterised by the principle 'ultra posse nemo obligatur', the minimum content results from a reflection on the general normative preconditions of each specific attribution of responsibility. This reflection also allows for some tentative considerations regarding the interpretation of processes of transnationalisation.

Advances in science and technology have added to our insights into the vulnerabilities of human agency as well as to the methods of exploiting them. This has raised the stakes for efforts to clarify the concept and ethics of manipulation. Among these efforts, Robert Noggle’s influencer-centered account of manipulation has been most significant. He defines manipulative acts as those whereby an agent intentionally influences a recipient’s attitudes so that they do not conform as closely as they otherwise would to the pertinent norms and ideals endorsed by the influencer. This provides a relatively simple and in many ways clear definition of manipulation. It sidesteps thorny debates about autonomy, freedom, or practical rationality. It also promises to reveal a conceptual parallel between manipulating and lying, and thus to explain why manipulation is pro tanto wrong. In one respect, however, the account remains ambiguous: It remains unclear whether, and to what extent, it requires that influencers’ beliefs about what is ideal for their recipients to be grounded in some effort on the part of the influencer to identify with or take on the role of her recipient. This paper explains this ambiguity. It argues that influencer-centrism cannot remain indifferent to the validity of an agent’s beliefs about the ideal state of the recipient and provide an identification requirement that would render the whole account plausible and sufficiently determinate.

Recent contributions to the philosophy of manipulation have challenged assumptions explicitly understood as “Kantian”; especially the assumption that the concept and the negative value of manipulation could be explained by regarding it as a subversion of rational agency. This paper examines Robert Noggle’s concerns about Kantian accounts of manipulation and confronts them with Kant’s considerations about the “moral illusion”. It argues that, while the original framework of transcendental idealism makes it hard to understand the value and vulnerability of rational agency, an appropriately revised Kantian perspective on rational agency and manipulation may still be promising. It would also overcome the limitations of Noggle’s own account of manipulation that result from a kind of methodological subjectivism. The considerations suggest that a multidimensional model is needed to fully account for the normative status and value(s) of (the different facets of) autonomy

Hatte Jeremy Bentham den von ihm geprägten Begriff deontology in seinem gleichnamigen Werk noch im etymologisch buchstäblichen Sinne als »Lehre vom Sollen« verstanden und mit der »Wissenschaft von der Moral« gleichgesetzt, so dient das Adjektiv ›deontologisch‹ heutzutage ausschließlich der Kategorisierung einer spezifischen Teilklasse allgemeiner Konzeptionen normativer Ethik. ›Deontologische‹ Theorien normativer Ethik werden dann in der Regel ›teleologischen‹ Theorien, und/oder ›konsequentialistischen‹ Theorien gegenübergestellt.

In der Technikethik spielt das Konzept der Verantwortung eine zentrale Rolle. ›Verantwortung‹ ist ein Basiskonzept, das, ähnlich wie die Konzepte ›Pflicht‹ oder ›Schuld‹, in vielfältigen Kontexten gebraucht wird. Dennoch lassen sich einige allgemeine Aussagen über seine Bedeutung treffen. Gerade Autoren, die im Bereich der Technik- und Wissenschaftsethik aktiv sind, haben sich nachdrücklich um eine Klärung des allgemeinen Verantwortungsbegriffs und seiner verschiedenen Aspekte und Gebrauchsweisen bemüht.

The article examines philosophical conceptions of responsibility found in the contributions of Martin Buber, Hans Jonas and Emmanuel Levinas. It argues that, despite the significant differences of these contributions, they all share important goals, significant structural features, and corresponding challenges. All three thinkers try to overcome the solipsistic limitations of Edmund Husserl’s phenomenology as well as the egocentrism of Heidegger’s concept of "solicitude" or "self-care." All three try to overcome the Kantian subject-object dichotomy. All three understand responsibility as a bipolar relation only. In consequence, all of them face similar challenges resulting from new forms of dichotomies and the exclusion of “thirdness” that are the result of their conceptual decisions.