In 2015, Davies and colleagues summarized thirty-two empirical bioethics integrative methodologies that combine normative analysis and empirical data obtained using social-science research. Following this, scholars have discussed the integration of the normative and life sciences research (Mertz and Schildmann 2018), using critical realism in empirical bioethics (McKeown 2017), and integrating experimental philosophical bioethics and normative ethics (Earp, et al. 2020; Mihailov, et al. 2021). In line with the systematic review of empirical bioethics methodologies’ two broad categories of dialogical and consultative processes of integration (Davies, et al. 2015), our participants indicated two familiar approaches. The first one is based on a reflective equilibrium–type process, and the other, an interdisciplinary collaboration between and among different stakeholders.

In addition, several participants suggested integration was inherent with the normative and empirical intertwined within the overall research process. Our participants’ accounts of inherent integration shared some similarities with, for example, moral case analysis (Dunn, et al. 2012), integrated empirical ethics (Molewijk, et al. 2003), and dialogical empirical ethics (Landeweer, et al. 2017; Widdershoven, et al. 2009). The shared similarities were in the sense that there were no separate normative and empirical parts to be distinguished in a project and that the project itself was normatively oriented. However, we should be critical of this view. The mere fact the empirical and the normative is inseparably intertwined throughout a research process does not mean (1) that these claims cannot be conceptually separated and (2) that such a method is free of methodological concerns. For instance, there would still be the need to specify what moral principles demand in a particular situation, decide which ethical theory to use, or make normative judgements with the help of empirical data (Frith 2012; Salloch, et al. 2015). Apart from that, several of these “inherently integrated” methods lacked a clear normative side and the enterprises described seemed purely empirical. Upon closer analysis, one could interpret some of the accounts of “integration was always inherently present” as a way of avoiding looking into the black box.

Furthermore, within these “inherently integrated” approaches, a few scholars described their descriptive research on ethical issues as empirical bioethics. Based on the available definition of empirical bioethics (Ives, Dunn, and Cribb 2017; Mertz, et al. 2014) and the standards offered by Ives and colleagues (2018), the works of these participants would thus not count as empirical bioethics. This is because there was no evidence of any integration happening. In our opinion, this mismatch between the practice of some scholars and what is “agreed” to in the literature as empirical bioethics may be pointing to the fact that empirical work in bioethics is in essence heterogeneous (Ives, Dunn, and Cribb 2017; Mertz, et al. 2014). For one, it is possible that scholars look at their projects as fitting an empirical bioethics because they start from research questions relating to the normative and because their projects, even with purely descriptive parts (and papers), are aimed to eventually lead to normative conclusions. But also in that case, we need to be clear about the nature of such particular (sub)projects and about the absence of integration efforts in these parts. Second, it is possible that scholars have different perspectives on the matter than the one expressed in the standards paper (Ives, et al. 2018). In that case as well, these must be brought out in the open. Third, some scholars may simply be mistaken when they consider their projects to be empirical bioethics. Their mistaken belief might be based on the idea that the empirical findings were at some point integrated in normative reasoning, which results in a normative claim. This simply might not be the case. This then, more than anything, would point to the need for transparency about and agreement on the use of methods. A heterogeneity of approaches in the field should be applauded. However, for all of them, we need to be able to identify where and how the integration happens. In the remaining part of this discussion, we focus on the overall vague manner in which our participants talked about their methods and what that implies for the field of empirical bioethics.

Reflective equilibrium, broadly construed, is a deliberative process that seeks coherence between attitudes, beliefs, and competing ethical principles (Daniels 2020). A standard objection against reflective equilibrium methodology is that it is insufficiently determinate in practical contexts to be action-guiding or to help decide between conflicting views (Arras 2009; Paulo 2020; Raz 1982). The iterative process of going back-and-forth between the normative and the empirical to come to a coherent account, similarly, is fraught with indeterminate indications. The way study participants relayed their approaches and explained their practices underscored the vagueness they felt. It further showed the difficulties even scholars with expertise in using these methods had in illustrating the “how” in an exact manner.

Such vagueness was also evident in collaboration methods of integration reported by our study participants. This collaboration involves an iterative and deliberative process of sharing information and engaging with different perspectives (Rehmann-Sutter, et al. 2012). It requires ongoing dialogue between social scientists and bioethicists. Their practical know-how guides the conclusion about the normative significance of empirical data. Even though the experience and implicit know-how of the experts can be rich in content and varied, how the communication process is done and who decides the outcome often remains indeterminate. This was noted in the voices of our participants.

The difficulty in clearly explaining the “how” of the integration process is something that researchers who have carried out an integration or wished to do so are likely to be familiar with. Several scholars have pointed to this unclear process as well (Ives and Draper 2009; Mertz and Schildmann 2018; Strong, et al. 2010). One explanation for this finding may be that, given the numerous tailored versions of the reflective equilibrium methodology for empirical bioethics (de Vries and van Leeuwen 2010; Ives 2014, Ives and Draper 2009; Van Thiel and Van Delden 2010; Savulescu, et al. 2021), there may be confusion surrounding how to make a choice and how to implement it in practice. As noted earlier, there are many available empirical bioethics methodologies (Davies, et al. 2015), and it has been suggested that each researcher could be using his or her own version (Wangmo and Provoost 2017). This situation, to us, points in two directions. First, it may convey a general need to remain flexible and open to creativeness, key components of the normative reasoning that is central to the integration method. We may thus have to stop looking for a method that is akin to empirical standards, especially those of quantitative methods, and recognize that the empirical and normative integration is in many ways a normative enterprise, which does not follow an exact method. Second, the wide variation of approaches makes it even clearer that we need to seek more methodological clarity on the overarching level. This is where the debate on standards (Ives, et al. 2018), for instance, has been an added value. It allows for heterogeneity while at the same time striving to create more clarity. In fact, we point out that the integration methods are inherently indeterminate and that this is a good thing. That said, an acceptance of the indeterminate character of this integration does not absolve us from the need to identify the foundations of what we are doing in a theoretical-methodological way.

The study findings confirm the image of an indeterminate process. As research on this topic is developing, it is ever more clear that the scholars involved come from a wide variation of disciplines. This is another argument as to why this indeterminate character is indispensable. The findings thus substantiate what has already been written about the indeterminate status of the methods used in empirical bioethics (Arras 2009; Davies, et al. 2015; Dunn, et al. 2008; Huxtable and Ives 2019), despite efforts to delimit and standardize empirical bioethics work (Mertz, et al. 2014; Ives, et al. 2018). One way of reading the vagueness we encountered is the scholars’ struggle to explain their own integration process, and perhaps even a lack of full comprehension of that process. Another interpretation is one that is in line with the wish for creativity and flexibility, and a level of indeterminacy in the methods we look for, namely an expression of leaving things open. Creativity can be a medicine against the belief that precise and transparent standards can account for such a “maze of interactions” (Feyerabend 2010) between experts with fertile know-hows. Too much standardization misses how particular research situations inspire novel ways of seeing the ethical relevance of empirical data. We should nevertheless be aware that the indeterminate nature of any integrative methodology makes it subject to risks of post-hoc rationalizations and motivated reasoning (Ives and Dunn 2010; Mihailov 2016). In the end, demands for creativity—however valid—should go hand in hand with demands for a thorough theoretical foundation as well as practical understanding of the method at hand.

Reflective equilibrium is a deliberation method that helps us come to a conclusion about what we ought to do (Daniels 1996; Rawls 1951, 1971). If we describe the integration process only in terms of going back-and-forth between data and theory, or in terms of collaboration between different experts, we risk obscuring the normative nature of using empirical data to help elaborate ethical prescriptions, which is the goal of doing such an integration (Ives and Draper 2009; Mertz, et al. 2014). Researchers often talk about integration as if it is a process half empirical and half normative or something that just needs normative reasoning alongside empirical data. But the very act of integration is normative in nature. While facts are essential for addressing bioethical issues, the task of integration ultimately depends on normative assumptions about the normative weight of moral intuitions.

Our data show that many of our participants rely on a reflective equilibrium characterized in their explanations mostly by moving back-and-forth between empirical results about moral attitudes and intuitions. Although the cyclical thinking is an important part of reflective equilibrium, there is more to it. Often, however, our participants did not move beyond this aspect. Ideas of coherence between moral intuitions and moral principles, and the fundamental willingness to adjust moral principles in light of what we discover were rarely touched upon. Perhaps what we see here is that several study participants embarked on an intuitive account of a—sometimes simplified—reflective equilibrium inspired methodology. At least in the interviews, it was not shown that they were fully aware of theoretical commitments to coherence, giving normative weight to moral intuitions, and screening them for bias.

The need to clarify the essential normative nature of integration appeals to normatively trained bioethicists, who may be in a better position to debate and assess how empirical input should be integrated into normative recommendations. We are not claiming that bioethics should be the arena of philosophers. Empirical research in bioethics is widespread (Borry, et al. 2006; Wangmo, et al. 2018), and scholarly perceptions about who belongs in the field are no longer exclusivist. There is thus a need to look at empirical bioethics projects in a broader way, including studies where empirical data are gathered but not used directly as part of a normative argumentation. Such empirical data may thus contribute to a larger body of work aimed at reaching normative conclusions. They can include, for example, empirical studies that explore stakeholders’ views relating to bioethical matters and explain how people arrive at certain reasoning patterns or studies that reveal the lived experience of stakeholders and explore how moral questions are experienced in practice (Mihailov, et al. 2022). To our view, despite the central role of normative know-how to integration, this does not mean that integration efforts need to be exclusively the work of ethicists or that empirical researchers will be unable to engage in it.

Our findings are, first and foremost, not generalizable, as they are based on an exploratory qualitative study design. The data come from a small non-representative sample of researchers. Other scholars, with different or greater experience in using particular (interdisciplinary) integration methods may have different opinions. They could perhaps have provided us with more concrete information about the way they carried out such integration. Also, only one of our participants described him/herself as a normative researcher. It would have been interesting to have more participants who were normatively oriented to include their views on how empirical data can be of use to the adaptation or formation of normative recommendations. Second, we asked scholars to tell us the process they use in integrating the normative and the empirical. This is a challenge task in and of itself. Not only did the scholars have limited time for the interview, but also it is generally difficult to explain how exactly this process pans out post-hoc. We thus acknowledge that we presented the participants with questions which were in no way easy for them to address in a single conversation. Because we wanted to focus on the scholars’ own reports, we did not confront them with approaches adopted by others in as systematic way. We did not also engage in a critical assessment of the reported method at the time of the interview. It would be interesting for further research to include such an approach and, for instance, study this using focus group methods. Using confrontation with other approaches or other views could offer the opportunity for a more critical reflection. For this paper, however, we opted to enrich the ongoing debate first and foremost with the accounts of the scholars. Third, we underline that a minority of our participants had already published methodological papers related to empirical bioethics as evident from the EBE sample. We did not ask the scholars to discuss the method that they have written about or most liked, nor did we ask them to discuss the paper that led to their identification for this study. During the interviews, however, we sought to address acquiescence and social desirability by using Socratic questioning and probing, to provide time for participants to explain their method of integration.