Empirical Ethics: When Ethics meets Data*

 * This text is retrieved from my dissertation Empirical Essays on Artificial Intelligence Ethics

The relation between normative and empirical fields is not optimal. Pragmatic reasons related to the different nature and aims of these disciplines are part of the problem. Ethics is a prescriptive discipline focused on moral evaluation, conceptual clarification, and normative justification, whereas empirical sciences is a descriptive discipline focused on description, reconstruction, and analysis [1]. Historically, Ethics is modeled after the field of Philosophy to produce a rational and de-contextualized discourse. Accordingly, it tends to reject the empirical sciences amid the fear that sociological contextualization may lead to ethical relativism [1]. 

Important meta-ethical fallacies (is-ought problem; naturalistic fallacy; and fact-value distinction) [The classical formulation of the is-ought problem is found in a passage of A Treatise of Human Nature by David Hume and relates to the fallacy of drawing an ought-conclusion from a set of is-premises [2]. The naturalistic fallacy by George Edward Moore (Principia Ethica) relates to the fallacy of identifying the predicate good with any natural or metaphysical predicate [2]. The fact-value distinction refers to different meta-ethical views about facts and values: (i) no statement or concept is irreducibly both evaluative or factual; (ii) evaluative discourse fails to have certain characteristics essential to paradigmatic realistic discourses like the scientific discourse; and (iii) scientific facts do not presuppose values (science is value free) [2].]  have provided theoretical ground for such rejection and are often used to dismiss the empirical work in the normative domain [3]. Despite the strained relation between normative and empirical sciences, it is not uncommon for ethics to undergo an empirical turn in order to improve its realism and context specificity [1, 4, 5]. Such a turn was evident in the field of medical ethics. As a response to the critiques of foundationalism and as part of the attempt to operationalize its key principles, the focus of medical ethics shifted from purely normative to an approach that is more grounded in practical reality [1].

Broad conceptions of empirical ethics include empirical and philosophy-driven research. In Empirically driven empirical ethics, empirical claims describe or explain the world either to identify normative issues that arise in a particular domain of practice or to make sense of the relevant experiences, understandings, judgments, or intuitions of individuals concerning those issues. In Philosophically driven empirical ethics, normative principles and guidelines are operationalized with the aim of practical implementation. Both conceptions require a prescriptive element where some course of action is recommended. Still, this element is more relevant in the empirically driven empirical ethics, as it is distinctive of the social scientific endeavor of describing normative dimensions of practice [6].

Other more nuanced combinations of empirical and normative research have also been described in the literature [3, 6–8]. These combinations tend to feature empirical and normative research as parallel, symbiotic, or integrated. A parallel combination means that the disciplines have each their own specific task which cannot be reduced to the other [7, 8]. A symbiotic combination entails that both disciplines influence each other in various ways. Empirical research may be used solely to ensure the application of the prescriptive function of moral theories or to improve moral theories, or it can also be part of a more dynamic process in which the critical function of Ethics is applied to empirical data and simultaneously the moral theory is criticized by empirical data [3]. A integrated combination may be said to be a new discipline with its premises, theories, topics, and methods to integrate empirical and normative research [7, 8] in which ethicists and descriptive scientists try to integrate moral theory and empirical data in order to reach a normative conclusion concerning a particular social practice [3].

[1] P. BORRY, P. SCHOTSMANS, and K. DIERICKX, The birth of the empirical turn in bioethics, Bioethics 19, 49 (2005), https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1467-8519.2005.00424.x .

[2] R. De Vries and B. Gordijn, Empirical ethics and its alleged meta-ethical fallacies, Bioethics 23, 193 (2009), https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1467-8519.2009.020.x .

[3] B. Molewijk, A. M. Stiggelbout, W. Otten, H. M. Dupuis, and J. Kievit, Scientific contribution. empirical data and moral theory. a plea for integrated empirical ethics, Medicine, Health Care and Philosophy 7, 55 (2004).

[4] A.W.Musschenga, Empirical Ethics, Context-Sensitivity, and Contextualism, The Journal of Medicine and Philosophy: A Forumfor Bioethics and Philosophy of Medicine 30, 467 (2005), https://academic.oup.com/jmp/article-pdf/30/5/467/2729155/30-5- 467.pdf .

[5] M. PARKER, Two concepts of empirical ethics, Bioethics 23, 202 (2009), https://onlinelibrary.wiley.com/doi/pdf/10.1111/j.1467-8519.2009.018.x .

[6] M. Dunn,M. Sheehan, T. Hope, and M. Parker, Toward methodological innovation in empirical ethics research, Cambridge Q. Healthcare Ethics 21, 466 (2012).

[7] G. R.Weaver and L. K. Trevino, Normative and empirical business ethics: Separation, marriage of convenience, or marriage of necessity? Business Ethics Quarterly , 129 (1994).

[8] L. van der Scheer and G.Widdershoven, Integrated empirical ethics: Loss of normativity? Medicine, Health Care and Philosophy 7, 71 (2004).

[9] E. J. Topol, High-performance medicine: the convergence of human and artificial intelligence, Nature medicine 25, 44 (2019).

[10] M.-C. Laï,M. Brian, and M.-F.Mamzer, Perceptions of artificial intelligence in healthcare: findings from a qualitative survey study among actors in france, Journal of Translational Medicine 18, 14 (2020).

[11] C. Blease, T. J. Kaptchuk, M. H. Bernste