1 Introduction

Among the many lessons that the COVID-19 pandemic has forced upon us is the stark reminder that we cannot assume that the world of tomorrow will be like today. Serious changes to our communal lives can happen quickly and with little warning. When that happens, we need to change with it. But, of course, COVID-19 is not alone in driving such developments. Climate change, geopolitical developments, or disruptive technological innovations present serious moral, political, economic, scientific and/or technical challenges. While none of these can be solved by only technological means, technological innovation will undoubtedly be part of confronting such issues.

Unsurprisingly, this translates into responsibilities for those developing these technologies, especially as calls for morally responsible innovation become increasingly pressing (Brundage & Guston, 2019). However, this puts innovators – including engineers – in a difficult position since making design choices implies having sufficient information to make those choices and the accompanying trade-offs. However, in times of rapid and sometimes unpredictable socio-technical change, the necessary knowledge may be lacking, either because the facts of the matter and/or the normative aspects that should guide decision-making are unclear (van de Poel & Robaey, 2017). In such cases of uncertainty, deliberation about what constitutes good engineering design becomes difficult. Furthermore, given the fact that an innovation’s consequences (e.g., in cost-benefit analysis and risk assessment) and norm-based prescriptions (e.g., in engineering codes of conduct) often guide engineering decision-making, it should not surprise us that decision-making becomes more complicated when one or both types of inputs for good engineering decisions are insufficient.

To overcome the deficiencies of consequentialist and deontological approaches to engineering ethics, some have proposed that virtue ethics could help (e.g., Schmidt, 2014) and may be particularly helpful when facing uncertainty (Frigo et al., 2021). This chapter explores these suggestions, strengthening the case for virtue ethics in engineering ethics by showing how it can help deal with different types of uncertainty. We first summarise the case for virtues in engineering ethics (Sect. 9.2), present different types of uncertainty, and compile a list of virtues relevant for engineering (Sect. 9.3). Finally, we present four uncertainty scenarios to analyse the impact of different virtues on engineering decisions (Sect. 9.4).

2 The Case for Virtues in Engineering Ethics

The ancient tradition of virtue ethics experienced a revival in the twentieth century, which eventually saw it positioned next to consequentialism and deontology as one of the three major streams of modern ethical thought (Baril & Hazlett, 2019). It comes as no surprise, then, that virtue ethics has since been proposed as an alternative to consequentialist and deontological approaches for various human activities, not least of which engineering. In this section, we summarise the case made for virtue ethics in engineering, focussing on its alleged advantages vis-à-vis its theoretical rivals.

Generally, virtue ethics helps identify engineering as a normative and purposive practice, in turn facilitating an ethical understanding of its activities. Bowen (2009, 2014) proposes we understand engineering as a practice in the sense presented by MacIntyre, i.e., as a “coherent and complex form of socially established cooperative human activity through which goods internal to that form of activity are realised in the course of trying to achieve those standards of excellence which are appropriate to, and partially derivative of, that form of activity, with the result that human powers to achieve excellence and human conceptions of the ends and goods involved are systematically extended” (1981, p. 187). Viewing engineering through this lens, Bowen presents an explicitly virtue-ethical account of engineering. Following his analysis of its internal goods (e.g., technical excellence, cost-effectiveness, safety and especially, the satisfaction of contributing to the flourishing of others) and its external goods (e.g., prestige, wages, economic benefits for others, but most importantly, technological artefacts), he proposes the following end or goal of engineering: “the promotion of the flourishing of persons in communities through contribution to material well-being” (p. 20). To achieve this normative end by attaining engineering’s internal goods, engineers need to act ethically. However, the question remains: how is virtue ethics better equipped to help engineers act ethically than its consequentialist and deontological counterparts?

The first such advantage of virtue ethics is increased motivation to practise engineering responsibly. The dominant consequentialist and deontological foundations of engineering ethics aim primarily to prevent misconduct, risks, and disasters. This results in a ‘preventive ethics’ of engineering (Harris, 2008). However, while preventive ethics may be partially effective in achieving its goals, it lacks “an internal, motivational, and often idealistic element present in professional life that cannot adequately be accounted for by rules” (p. 155). It is this element that is said to be better accounted for and mobilised by virtue ethics, resulting in an ‘aspirational’ ethics that has a positive rather than a preventive orientation (Bowen, 2009; Harris, 2008; Schmidt, 2014; Steen et al., 2021). This positive orientation also captures some of the zeitgeist we experienced at the beginning of the COVID-19 pandemic. In her essay ‘The Pandemic is a Portal’ writer Arundhaty Roy (2020) invited readers to think about what kind of world we should aspire to after this historically transformative experience, an invitation that virtue ethics would extend to engineers as well.

On the one hand, this is because virtue ethics may be comparatively better suited to keep engineers pointed towards the normative goal of engineering. This is in part due to the ‘modern’ origins of consequentialist and deontological ethics since, like engineering, they function according to the paradigm of ‘technical rationality’ (Schmidt, 2014), which assumes that the knowledge and skill necessary for a practice can be captured in specific and generalisable rules, prescriptions or instructions. This abstraction of ethics from the actual practice and from the persons involved risks aggravating the disconnect between engineers and the goal of their practice (i.e., the flourishing of people), with them considering values like efficiency and technical ingenuity as ends in themselves instead (Bowen, 2009; Schmidt, 2014). Due to virtue ethics’ focus on the person’s virtues and the connection of professional virtue to the goal of engineering as a practice, it helps minimise this disconnect. On the other hand, virtue could be intrinsically rewarding to virtuous engineers. Indeed, virtue ethics promises to both “identif[y] good behavior and provid[e] the psychological motivation for conforming to that behavior. It may be a more effective prod to achieve good in the world than are less personal calls to maximize utility or conforming to rights and duties, precisely because it is so personal” (Crawford-Brown, 1997 p. 483). This has to do with the nature of virtues as moral dispositions achieved through education and experience. As Aristotle already noted in the Nicomachean Ethics (NE, II.3), both virtue and vice are concerned with pleasure and pain. However, it takes proper education to learn to enjoy the things one ought, to feel delight and pain rightly. It is a sign that a person has achieved virtue that they derive pleasure from exercising it and are characteristically disposed to doing so. This is no less true for professionals than it is for others, even if the virtues demanded by the profession may be specific to it. Indeed, “adhering to professional virtues brings professional satisfaction, just as adhering to personal virtues brings satisfaction to one’s personal life” (Harris, 2008 p. 158). The good engineer receives satisfaction from the virtuous exercise of their profession.

The second critical advantage of virtue ethics is that it helps engineers to perform their profession responsibly in a way not covered by its more act-oriented alternatives: “fulfilling an engineer’s responsibilities to protect public safety, health, and welfare calls as much for settled dispositions, or virtues, as it does for performing this or that specific action.” (Pritchard, 2001 p. 391). This is because those stable dispositions, both personal and professional, are an integral part of professional competence. This makes sense when understanding engineering as a normative practice since virtues are integral to practices that allow practitioners to achieve the practice’s internal goods and end (Harris, 2008). One way virtues help the engineer practise their profession responsibly involves a peculiar aspect of virtue as a stable and deeply entrenched trait of character. Following Hursthouse, Harris sees virtue as ‘multitrack’, involving not simply reason, but “emotions and emotional reactions, choices, values, desires, perceptions, attitudes, interests, expectations and sensibilities.” (Hursthouse, 2006 as cited in Harris, 2008 p.156). To cultivate engineering virtues is thus also to attune one’s perception, values and sensibilities to those salient aspects of the context of engineering that allow the attainment of engineering’s internal goods and goal. As a result, virtuous engineers are more likely to care about and be more sensitive to important aspects of situations and thus more prone to discover them. For example, the rapid development of contact-tracing apps during the COVID-19 pandemic raised difficult trade-offs between privacy, security, public health and a host of other values that led to varied designs. In the Dutch context, for example, a strong emphasis was given to privacy (Verbeek et al., 2020) after consultation with experts. An engineer that is appropriately sensitive to and appreciative of the moral gravity of these trade-offs is also more likely to engineer ethically acceptable applications in these trying times.

Related to these sensibilities comes the idea that virtues, as stable traits of good character, allow the engineer to exercise their powers of discretion and judgement (Harris, 2008), improving decision-making (Sand, 2018), e.g., in selecting the appropriate heuristics in engineering decisions (Schmidt, 2014). Thus, not only does virtue ethics leave more room for the engineer to judge, but being virtuous makes them more capable of doing so. Of course, such virtue has to be cultivated through ample education and experience. However, this focus on education should not be taken as a downside of virtue ethics but rather a call for more disposition-oriented engineering ethics education.

At this point, it is imperative to note that the exercise of virtue both in the engineer’s personal and professional life requires it to be done deliberately, with practical wisdom. Multiple authors have emphasised the importance of practical wisdom or phronesis for the virtuous engineer and the capability for proper engineering judgement (e.g., Frigo et al., 2021; Harris, 2008; Schmidt 2014; Steen et al., 2021). Aristotle defined practical judgement as “a reasoned and true state of capacity- to act concerning human goods” (NE VI.5). For engineers, this complex virtue would allow them to know what aspects of a situation are relevant, what virtues are appropriate in a given situation, how they fit together, and how they should be exemplified in practice (Athanassoulis & Ross, 2010; Steen et al., 2021). As such, the virtue of practical wisdom may be of particular use when the consequences of an action are not entirely known, or the appropriate norms are not given, i.e., it may help engineers when significant aspects of the case are uncertain (Frigo et al., 2021).

In light of the above, the case for virtue ethics in engineering -possibly supplementary to consequentialist and deontological approaches- is extensive and prima facie convincing. As such, our goal is not to make this case anew. Rather, we aim to strengthen the case for virtue in engineering ethics by further highlighting its compatibility with current changes in engineering – specifically the increased appreciation of uncertainty in engineering work, which might be exactly the type of attitude needed for engineering during and after the pandemic.

3 Virtue, Uncertainty and Engineering

The latter of the abovementioned advantages, the usefulness of practical wisdom or phronesis, is fitting for a practice like engineering that is both a) inescapably normative and b) supposed to provide practical and acceptable solutions in ever-changing circumstances. These characteristics parallel Aristotle’s characterisation of practical wisdom as not only concerned with human goods (NE VI.5) but also not “concerned with universals only – it must also recognise the particulars; for it is practical, and practise is concerned with particulars” (VI.7). This positions practical wisdom as a virtue of careful deliberation about the good that is paramount in exercising virtue more generally (VI.13). Nevertheless, to deliberate well about proper action in a world of particulars is not straightforward because practical wisdom will run up against practical and epistemic limits, i.e., deliberation about practical action is often made more difficult by uncertainty, something the COVID-19 pandemic has proven time and time again. That is, many rapid technological developments were considered while knowledge of their effects or the situation into which they were to be implemented was incomplete, e.g., contact tracing apps, new vaccine technologies, testing kits, and new anti-viral drugs. The complex and ongoing deliberations that these prompted are emblematic of the fact that as the consequences of our technological interventions (and/or our appreciation of them) have come to extend farther in time, space and our lives, uncertainties have likewise grown despite significant efforts to map them better. Unsurprisingly, then, virtue ethicists have explicitly argued for the need to account for uncertainty in virtue ethics as well as the need for virtue ethics in light of uncertainties brought about by rapid (socio-)technological change. In “Seven Traits for the Future” (MacIntyre, 1979), MacIntyre grapples with the question of what traits would be desirable to promote in our society going forward. Interestingly, the virtue that tops his list is the “Ability to Live with Uncertainty”. In line with the above, MacIntyre points out that “there are necessary limits to our predictions about the future of technology […] The answer is clear: we will have to design people with all those traits [...] necessary for living in an unpredictable environment, people with an ability to live with a large lack of certainty about their future” (p. 5). A more developed account of the need for virtues in light of the uncertainty in an increasingly technological world is provided by Shannon Vallor (2016), who makes the case that the unpredictability of technological development and its consequences brings about a state of ‘acute technosocial opacity’, which in turn requires that we develop the proper ‘technomoral virtues’ if we are to deal with that opacity well. From her list of twelve technomoral virtues, some stand out as specifically conducive to dealing with the uncertainty involved in our technological future, i.e., those of humility, courage, flexibility, perspective and, of course, technomoral wisdom (the latter being structurally similar to phronesis in more general virtue ethics). The central conclusion from these accounts is clear: deliberating about proper action when we are uncertain about future consequences and normative changes requires appropriately virtuous and wise deliberators. The extremely polarised and uncharitable nature of many societal discussions about COVID-related technological interventions would also seem to indicate a need for more of such virtuous deliberators.

Suppose this is true for society at large. In that case, it should be at least as important for those who have the power to steer the very technological developments that bring about a significant share of the uncertainties we are discussing. As such, it most certainly applies to engineers. Some of those who have argued for a role for virtue ethics in engineering (discussed in the previous section) have also made the link with uncertainty explicit. Sand (2018), while discussing the virtues of innovators more broadly, recognises that even responsible innovation processes have significant, unpredictable effects on society (dealing in so-called ‘wicked problems). Countering the critique that a virtue-ethical approach would be unable to provide an answer to the question of what innovators should do in such a context, he points out that a) that is equally a problem for deontological and consequentialist approaches in those circumstances, and b) virtues could nevertheless provide “guidance and orientation for becoming a more creditable person and avoiding making moral mistakes. Certain dispositions and capacities help to assess risks properly and, thereby, enhance good decision making” (p. 84). Sand’s focus on risk is not coincidental. Innovation generally, and engineering projects specifically, are characterised by the impossibility “to predict absolutely accurately what their consequences will be” (Ross & Athanassoulis, 2010 p. 148). This leads to situations of risk, which “involve, of necessity, uncertainty; therefore, the outcomes of one’s actions will be uncertain” (Ross & Athanassoulis, 2012 p. 838). This central feature of engineering practice puts a “peculiar ethical burden” on engineers: “the assessment, management, and communication of risk—the very real possibility that engineered projects and products could detract from the material well-being of some people, rather than enhancing the material well-being of all.” (Schmidt, 2014 p. 998). As such, we should expect engineers to be well-equipped to deal with such situations, which, according to Ross and Athanassoulis (2012), is enhanced by having virtuous dispositions, including phronesis. As such, virtue ethics’ focus on the engineer’s moral character and their responsiveness to relevant contextual features can help them deal with uncertainty in the form of risks (Athanassoulis & Ross, 2010). However, as Sand’s reference to ‘wicked problems’ already indicates, technical risks are only one instantiation of uncertainty engineers may be confronted in their work. In what follows, we present different types of uncertainty, including normative ones, that engineers may encounter when designing the very technologies that will shape our future and ideally help others flourish. In so doing, we hope to show that uncertainty in engineering is more multifaceted than a singular focus on risk would disclose.

3.1 Broadening the Scope of Engineering Uncertainty

In a technical sense, uncertainty represents the lack of probabilistic knowledge for a given event (Doorn & Hansson, 2011). Engineering’s preoccupation with risk rather straightforwardly fits this ‘technical’ sense of uncertainty. However, scholarship in the philosophy/ethics of technology and engineering has gone further in defining types of uncertainties that engineers may encounter and capture the complexity of what is unknown. In this section, we present four types of uncertainty from recent publications from these fields. In each type, uncertainty represents a situation that may hamper engineering decision-making, i.e. design decisions in which the potential risks of new technologies play a defining role.

A first type of uncertainty identified in the engineering ethics literature is scenario uncertainty (van de Poel & Robaey, 2017). This type of uncertainty captures the lack of full knowledge about a situation, where different potential ways forward (scenarios) can be imagined based on the available information. Still, we lack the knowledge to reasonably predict how likely these scenarios are to unfold. We often lack complete knowledge for many innovations, but there are some benchmark or a history of use that provide reasonable expectations of a way forward. In scenario uncertainty, this is not the case. Moreover, it is not only that we do not know which scenarios will be likely to happen but -because the form of our knowledge or the lack thereof has normative consequences- also which ones would be more or less desirable. Thus, this type of uncertainty will have normative implications and relate to epistemic normative uncertainty (Taebi et al., 2020).

Another type of uncertainty, also related to epistemic normative uncertainty, is ignorance: a situation where there is simply no knowledge of some potential consequences of a technological intervention (van de Poel & Robaey, 2017). Here, we don’t even know what scenarios we don’t know.

There are two more types of uncertainties we include in our analysis. In these, the normative aspects of engineering applications take centre stage. The third type of uncertainty identified in the literature mentioned above is indeterminacy, the situation in which causal chains are uncertain, and different actions of different agents could lead to different outcomes that might be unforeseen (van de Poel & Robaey, 2017). This type of uncertainty also evokes evolutionary normative uncertainty where, as technology and moral norms co-evolve (Taebi et al., 2020), it is unclear how to normatively assess a technological innovation because of unpredictable moral change. A fourth and final type of uncertainty is normative ambiguity representing a disagreement about values and norms (van de Poel & Robaey, 2017). Normative ambiguity can be further specified as theoretical normative uncertainty, where different ethical theories will justify different ways of dealing with a problem (Taebi et al., 2020), or conceptual normative uncertainty where the norms themselves allow for different interpretations or prioritizations (Taebi et al., 2020).

Even from this condensed summary, it is clear that uncertainty has many forms and, importantly, includes normative unknowns. In light of this, if virtues are supposedly effective at helping engineers deal with uncertainty, and the exercise of virtue aims at the practical and the good, it stands to reason that virtue may be helpful for both epistemic and normative uncertainties. However, whether and to what extent this is the case remains to be seen. As such, this chapter aims to evaluate the usefulness of engineering virtues when faced with various forms of uncertainty in engineering. Before it can do so, however, we must first consider what engineering virtues might be applicable in the first place.

3.2 What Are the Engineering Virtues?

Above, we said that this chapter does not aim to make a case for virtue ethics in engineering anew. Likewise, it does not aim to develop a new list of most important engineering virtues, nor is this necessary given the impressive array of engineering virtues presented by others. Table 9.1 offers an extensive overview of engineering virtues as proposed by those making a case for virtue ethics in engineering; virtues that we take as inspiration in analysing the uncertainty scenarios below.

Table 9.1 Overview of previously recognised virtues relevant for engineering
Full size table

This table structures the overview of engineering virtues differently from how one would find them in the sources from which they were extracted. The presentation of these virtues, their character, and their connection to other virtues and to engineering varied widely across those sources. Thus, the virtues in Table 9.1 have been brought together under a new structure. However, because of these varied foundations, and because we are not taking a strong position on the nature of virtue, Table 9.1’s virtue categories should be read heuristically.

For example, it follows the general distinction between moral and intellectual virtues. However, it also has an ‘in-between’ category that contains virtues that do not neatly fall in either of those two. For example, whether humility is a moral or an intellectual virtue may depend on whether we have outcomes – or motivations based understanding of virtue (Wilson, 2017) and/or the context in which it is acted upon (Bommarito, 2018). Likewise, the virtue of anticipation has been linked to both moral responsibility as well as intellectual virtue in engineering (Steen et al., 2021; Stone et al., 2020). It has even been argued that some virtues can be hybrid, simultaneously moral and intellectual (e.g., hermeneutic justice, see Fricker, 2007). Answering the question of which of these possibilities fits which virtue in the category is beyond the ambitions of this chapter. However, it indicates that exercising these virtues may exhibit both moral and intellectual excellence.

Next, Table 9.1 distinguishes between ‘fundamental human virtues relevant for engineering’ and ‘specifically engineering virtues’. This is simply to indicate that the former would be expected of virtuous persons generally but are also important for the virtuous engineer, while the latter are specific to the practice of engineering. This specificity also prompted a redesignation to ‘social’ and ‘technical’ virtues, indicating the community-orientedness of the former and the partially practical rather than solely intellectual nature of the latter. Such a division is to be expected for engineers since their being persons and being engineers is not distinct. It is important for engineers to find “continuity and coherence in both professional and personal life. [They] are human persons always and only sometimes engineers” (Bowen, 2014 p. 25).

Interestingly, it would seem that the specific virtue characterisation of the practice of engineering has generally focused on the ‘technical’ rather than the ‘social’ virtues, with the moral virtues for engineers being left less specific to the practice. While this is likely partly due to the cognitive nature of engineering work, its practical and normative orientation could be a reason to explore this imbalance further.

Lastly, the central position of practical wisdom in the table is no coincidence. Not only is it one of the virtues most often cited, but when it is, it is given a central, regulative role in the exercise of the engineering virtues as well as in appropriate, context-sensitive engineering judgement (Frigo et al., 2021; Schmidt, 2014; Steen et al., 2021). This, of course, runs parallel to the role that practical wisdom plays in virtue ethics more generally.

Armed with the different types of uncertainty and a list of engineering virtues, the latter’s usefulness in dealing with the former can now be investigated. In the next section, we do so by sketching several scenarios based on real situations and technological innovations in which hypothetical engineering professionals face difficult decisions under different conditions of uncertainty.

4 Applying the Virtues to Cases of Uncertainty in Engineering

In this section, we present real events and persons to provide a context for discussion. Our subsequent analysis develops hypothetical agent-centred considerations grounded in those real events. This allows us to discuss the role of virtues in decision-making by engineers in situations burdened by different types of uncertainty. Each case exemplifies a particular uncertainty situation. However, few real-life cases are likely to be so ideal-typical as to present only one type of uncertainty. Although that does not make their application in our analysis less salient for the purposes of the chapter, we recognise that they do not tell the whole story. Another observation about the cases we present is that they all capture potentially undesirable situations. We acknowledge that uncertainties need not always be about undesirable situations, and in this analysis, we will see that all cases can present desirable and undesirable uncertainties.

Moreover, the cases presented below are all from the life sciences, where the application of recent scientific findings allow translating them to useful engineering applications. The technologies presented thereby add a layer of uncertainty because of their relation to new knowledge not always resulting from traditional science but also of techno-science (Bensaude-Vincent et al., 2011). This allows examining situations where uncertainty is not only an epistemic endeavour but also a moral one. Finally, the cases presented are engineering applications that result from various forms of bio-engineering that relate directly to the pandemic, like mRNA vaccines, or that could come play a role in mitigating effects of the pandemic, like mechanical wombs, and synthetic milk.

4.1 Situations of Scenario Uncertainty and Ignorance

4.1.1 The mRNA COVID Vaccines

Recently developed COVID-19 vaccines have raised many questions on their potential side effects. Receiving emergency approval by local regulatory bodies, recommendations on their use for specific age groups changed as they were rolled out. The main discourse concerning administering vaccines under emergency approval has been that benefits outweigh the risks of the disease itself. This consequentialist claim was made from a general public health perspective. However, for vulnerable populations, like pregnant people, children, or people with certain existing conditions, the risks and benefits of the vaccines had not been researched when they were first rolled out.

One event that represents scenario uncertainty, in particular, is the June 2021 citizen petition on the assessment of mRNA vaccines, led by Dr. Linda Wastila, professor and Parke-Davis Chair in Geriatric Pharmacotherapy at the University of Maryland. Leading the Coalition Advocating for Adequately Licensed Medicines (CAALM) comprised of scientists, clinicians, and patients advocates, their citizen petition to the Food and Drug Administration (FDA) asked for more caution in the full approval of mRNA COVID vaccines (BMJ Opinion, 8 June 2021b). The petition raises eight points of consideration to the Food and Drug Administration. We highlight three of these here: CAALM asks the FDA to provide evidence that the new vaccines will actually benefit vulnerable groups, to research biodistribution of mRNA vaccines, and to further investigate all severe reactions following vaccination. Following the 23 August 2021, FDA decision to grant full approval to the mRNA COVID-19 vaccine by Pfizer, another member of CAALM, Peter Doshi, senior editor of the BMJ, reiterated the need for the 2-year requirement in phase 3 clinical trials in order to “have the science right” (BMJ Opinion 23 August 2021a).

This citizen petition essentially demands less scenario uncertainty by defining specific areas of concern that need further investigation. Reducing scenario uncertainty, in turn, allows reducing epistemic normative uncertainty. To make the right choices about vaccine rollout and considering further measures like mandatory vaccination, we need to know the likelihood of different scenarios.

For certain groups of the population, like pregnant people, this goes even further than a situation of scenario uncertainty but rather of ignorance. Pregnant people are excluded from medical clinical trials (Smith et al., 2020), understandably so, given the potentially devastating side effects on their future child. In this sense, they are protected, but in the face of the urgency of the COVID-19 pandemic, pregnant people are also at specifically high risks (Wastnedge et al., 2021). Until governments emitted recommendations on vaccination during pregnancy,Footnote 1pregnant people could, in certain cases, elect to receive vaccines. While not all vaccines are safe to administer during pregnancy, like those containing live viral material, mRNA vaccines have the advantage of not containing live virus material and thus presented the option of being administered to pregnant people. One can only imagine the heightened sense of epistemic normative uncertainty for this particular population group, with it possibly impacting future generations.

4.1.2 What Virtues to Consider for Scenario Uncertainty and Ignorance?

Normally, in drug development, stage-gate models are used to “fail early and fail often”, to ensure that whatever drug is produced at the end has higher safety standards (Hjorth et al., 2017). Now, we enter a hypothetical scenario of a vaccine developer: typically, this would be a scientist working on applying knowledge from biochemistry, pharmacology, molecular life science, immunology and so on to the end of making an effective vaccine. For this discussion, our hypothetical scientist engineers a new vaccine and is acutely aware of the concerns raised in the CAALM citizen petition to the FDA. We could even imagine this scientist being pregnant, thus embodying the two types of uncertainties and experiencing the urgency of developing vaccines to fight the pandemic and protect vulnerable groups.

Considering the list of virtues presented above, here are some we could consider relevant and why we think so. Here, the virtue of justice seems particularly relevant when making design choices in terms of access to the new vaccines. Here, as MacIntyre points out, we face the challenge of multiple meanings of justice, where he suggests considering the issue of desert (MacIntyre, 1981, p.249). One interpretation we can offer here is that everyone deserves access to health, so this could mean designing vaccines to benefit all groups of the population. This might come at the cost of other internal engineering goods, like efficiency, or run into other challenges like clinical trials regulations. Typically, engineering solutions do not, at first, focus on justice though they can certainly contribute to it. A just engineer would aim to realise the normative goal of engineering; whether it is in her ability and power to do something about it is another question.

Here, general virtues of integrity, honesty and perseverance can support the just engineer. Aiming to act as a just engineer might prove frustrating, and especially in an urgent and business context where outcomes are needed fast. So having virtues of integrity, honesty and perseverance can help her remain just in such a high-pressure environment. Finally, virtues of open-mindedness, originality, and thoroughness can help realise the goal of a just engineer, for they help engineer safe and accessible vaccines for all.

Looking at specific engineering virtues, inclusion and responsiveness are a specification of the virtue of justice, accompanied by virtue of anticipation, and the virtues of sensitivity to ‘tight coupling’ and ‘complex interaction’. We present these as specifications as it is not clear what the exercise of these engineering virtues would amount to without the virtue of justice. For instance, one could excel at anticipating and yet not mobilise it towards justice or human flourishing.

So far, we’ve discussed scenario uncertainty and ignorance as the same thing. Does ignorance call for different dispositions than scenario uncertainty? With this specific technological intervention, being a just engineer will likely be paramount in either case. Lacking information, conceptions of justice and moral dispositions become increasingly important. As an analogy, consider Rawl’s veil of ignorance. A lack of knowledge prompts an increased need for moral virtues of compassion, empathy and care (and, arguably, in the case of technology development, imaginativeness and creativity).

In this first analysis of relevant virtues in cases of scenario uncertainty and ignorance, it seems that an engineer apt to deal with these situations is a just engineer. The other virtues come to support the exercise of justice as a virtue but could be different, depending on the context and type of scenario uncertainty.

4.2 Situations of Indeterminacy

4.2.1 The Mechanical Womb

In March 2021, a New York Times article heralded the success of mechanical womb research giving birth to thousands of mice embryos (Kolata, 2021). While this article underlines the scientific advantages of studying the development of mice through a mechanical womb, e.g., by pausing development, it also points to potential future applications to human embryos. Needless to say, one can imagine many applications of a mechanical womb for people who struggle with infertility (Berglund, 2021), new avenues to replace problematic issues of surrogacy (Abecassis, 2016), and for increasing chances of survival of preterm babies (Werner and Mercurio, 2021). At the moment, laws prohibit any research on embryos older than 14 days.

In the New York Times article, one of the scientists interviewed on the matter, Dr. Tesar, not involved in the development of the mechanical womb, is quoted saying: “[e]ven assuming they could [grow human embryos], whether that is appropriate is a question for ethicists, regulators and society.” In the enthusiasm of this techno-scientific achievement, it seems that scientists interviewed in this New York Times piece defer moral judgement and potential future use to other agents in society.

Questions of indeterminacy arise at several levels. In this case, we can expect an entanglement of causal events, different decisions of different stakeholders, and changing norms and values concerning this. Here, we list just a few of the indeterminate issues in relation to the mechanical womb (cf. Verbeek, 2008):

  • How would society decide to use this technology: to support pre-term babies in neonatal intensive care units or to implant an embryo until birth to remedy infertility and replace surrogacy?

  • How would parents and doctors decide when to use the mechanical womb in either case? Who would set the guidelines and based on what norms and values?

  • How will parents experience relating to their babies born from a mechanical womb? Would this threaten the integrity of the baby’s future? Past and recent controversial interventions with babies have brought international media attention to the cases of Louise Brown, the first baby born from in vitro fertilisation, or Lulu and Nana, the first gene-edited human babies. While today, in-vitro fertilisation is a common procedure supporting many families in their reproductive journey, gene-editing is forbidden, while it could also further support families in bearing viable children in some cases.

  • What will become of entire bodies of professionals such as midwives, or doulas, were mechanical wombs to become the norm in reproductive health?

These questions capture potential evolutionary normative uncertainty. Of course, we might change our moral views on many of these issues, but we just don’t know what kind of possibilities the mechanical womb will afford, even within a pandemic situation like ours. For example, pregnant people with COVID-19 are more at risk of pre-term births than their healthy counterparts (Villar et al., 2021). Recent news reports in the Netherlands indicated an alarming number of unvaccinated pregnant people needing emergency C-sections as early as 24 weeks of gestation (NOS, 2021). Despite the fact that the mechanical womb is still under development, one can readily imagine a use case for it in similar circumstances, with all the normative uncertainties that brings.

4.2.2 What Virtues to Consider for Indeterminacy?

Let us now imagine our bioengineer, with a background in developmental biology, or training in obstetrics, designing a carefully balanced environment meant to support the development of a human being.

Bioengineers’ choices impact the integrity of the life of children, but also on the relationships to their parents, and potentially on the future of labour for pregnancy care. Decisions on the use of technology will impact technical choices and vice versa. It will require understanding the impact that technical limitations might have on the fundamental questions listed above. This is typically beyond the purview of a designer, as the focus is on optimising technology to help grow healthy babies.

In order to capture the range of effects on different agents, an important set of virtues here are compassion, empathy and care. Indeed, designing the mechanical womb is not just a matter of optimisation, but rather a matter of what should this technology afford for our human identities and integrity. With this comes a requirement for intellectual carefulness, in order to verify assumptions on these various relationships and inquire on how such a development changes things for a complex set of people, also in the future. Some engineering virtues can further specify how to be careful: for instance, cooperativeness with healthcare providers, patient organisations, surrogacy advocates, and various types of prospective parents. Furthermore, in order to be able to be cooperative, virtues of inclusion and responsiveness, as well as the virtues of perspective, will help give depth to cooperativeness by inviting speaking to a broad range of stakeholders and engaging with relevant moral issues at stake. Another possible engineering virtue that would help support this endeavour is seeing the ‘big picture’ as well as the details of smaller domains.

Therefore, a compassionate, empathetic, and caring engineer could make design choices that accompany indeterminate situations, where norms are bound to evolve.

4.3 Situations of Normative Ambiguity

4.3.1 Synthetic Maternal Milk

There is certainly a lot of debate on the ‘right’ way to nourish a newborn with global health recommendations focussing on maternal milk, and in the cases where breastfeeding is not an option due to health or socio-economic factors, formula from cow milk is presented as the next best alternative. A new development in synthetic maternal milk might significantly change these discussions, creating new opportunities for personalized nutrition. Just like the mechanical womb, the case of synthetic maternal milk could also serve as an helpful alternative for mothers infected with COVID-19 to provide superior nutrition to their children while lowering chances of infection.

An in-depth portrait of Dr. Leila Strickland (Kleeman, 2020), founder of Biomilq, a synthetically produced maternal milk, questions how such a development in precision fermentation could disrupt what we understand as a good way to feed the next generations of newborns and even further prevent breastfeeding in public which is a taboo in many western countries (Hauck et al., 2021). If ‘breast is best’, how will synthetic breast milk change how people perceive the need for breastmilk, or what kind of added price tag will this put on parents who already pay a premium for formula milk derived from cow milk?

Here we can see various layers of normative ambiguity that can be captured at an individual but also societal level. From a public health perspective, an innovation like Biomilq presents several advantages: it is more environmentally friendly as it reduces reliance on dairy farming and reflects parents’ preferences in diets as it would be customizable. In the beginning, it would probably be a luxury product using a biopsy from the feeding parent. Still, its production could become available at an attractive price point for consumers in the long run. This is a consequentialist perspective on the matter, often preferred in public health recommendations (Markmann et al., 2015). However, synthetic maternal milk could steer parents away from either formula, or breastfeeding. This can hide a host of problems concerning health and justice. For instance, the designers of Biomilq recognise that synthetic milk will never be equivalent to breast milk due to the adaptive nature of breastmilk: becoming more diluted on warmer days, containing antibodies when a child is sick. Another issue is justice. If synthetic maternal milk is superior to formula, how will parents without means have access to this better option? These two issues point to a rights-based approach to health, where the right to health of a child and the right to access healthy food might be put at risk or might create new demands on the health system. Which of these theories will help us deal with these developments? It is hard to say and rather likely that these different reasoning will create conflicts on the issue of feeding children. This is what theoretical normative uncertainty captures. Within this, there will likely be conceptual normative uncertainty as to how parents and doctors, hospitals, or even international organisations like the World Health Organization see what as healthy or just.

4.3.2 What Virtues to Consider for Normative Ambiguity?

Let us now consider the role of a bioengineer, active in precision fermentation and developing synthetic maternal milk. Looking at the personal story of Dr. Strickland and the account of experiencing emotional stress that comes from not being able to breastfeed, an act linked to being a good parent, providing the best nutrition for a child’s health, is a laudable motivation. At the same time, the social disruption potential of such an innovation can change many norms of what is desirable for a child’s health, from a public and individual perspective, and what is acceptable in public places in terms of breastfeeding.

In order to be able to think about others, whom a disruptive innovation rather do disservice, our bioengineer might have to exercise beneficence next to justice. For example, is it a good invention if it’s only good for a portion of parents? This could be further specified as cultivating appropriate ambition for an invention and exercising the virtue of humility in order to evaluate design choices beyond their own goal. This also calls for the intellectual virtue of carefulness. Here this could be understood as being careful to make synthetic milk that is good for children but also being careful about wider issues. A common idea of potentially disruptive innovation is ‘break things and move fast’ where exactly in these cases, carefulness would be advised.

These fundamental virtues can be further specified in terms of engineering virtues. For instance, to think of beneficence and appropriate ambitions, engineers can develop techno-social sensitivity and practise inclusion and responsiveness to recalibrate these ambitions and notions of the good for others. This also requires the exercise of anticipation of effects or seeing the big picture as well as the details of smaller domains. Our engineer needs to be open to criticism and correction for this recalibration of ambitions and notions of the good.

Therefore, a beneficent engineer would be equipped to deal with theoretical and conceptual normative ambiguity by virtue of thinking of the good beyond her own experience.

4.4 The Regulative Role of Phronesis in Situations of Uncertainty

The cases described above highlight the interaction of different virtues that support the ultimate exercise of fundamental moral virtues. Depending on how authors classify them, these various types of virtues are human virtues, more specific to engineering, moral/social, intellectual/technical, or somewhere in between. This is the main finding from analysing these cases: types of uncertainties do not necessarily require more knowledge of the truth, but more ways of relating to various issues at stake: for instance, the health of unborn children, parents who struggle with infertility, and parents who want to feed their children with the best possible options. Our interpretative analysis did not make the role of practical wisdom explicit, so we would like to return to it here. In order to navigate between different virtues and be able to know which ones could help support the exercise of other virtues: this demands practical wisdom to be regulative of the various types of virtues (see summary Table 9.2).

Table 9.2 Summary of relevant virtues for life science engineering in different situations of uncertainty
Full size table

In the cases presented, we suggest that some of these virtues are the most relevant virtues and that these are supported by other fundamental human virtues that can then be further specified into engineering virtues. This is consistent with Aristotle’s account where moral virtues and practical wisdom depend on each other in order to achieve human goods (NE VI.13). This analysis from relevant virtues to virtue specification is evaluative and interpretative and might not yield the same conclusions in every case of either type of uncertainty. Rather it underlines the necessity of practical wisdom in recognising which virtues to practise when and which virtues to acquire.

Therefore, the deliberate exercise of practical wisdom is primordial to making design choices, as these choices will require the exercise of various virtues, as our cases illustrate. It is not sufficient to be inclusive if the exercise of justice does not accompany this or to be open to criticism if not for the exercise of beneficence. This confirms the important role of practical wisdom in situations of uncertainty (Frigo et al., 2021).

5 Conclusion: Virtues for Designing under Uncertainty

To conclude this chapter, we find that moral virtues are paramount to making good design decisions in situations of uncertainty, like many presented by the COVID-19 pandemic. For the cases we evaluate, we conclude that: situations of scenario uncertainty and ignorance call for a just engineer. Situations of indeterminacy call for a compassionate, empathetic and caring engineer. Finally, situations of normative ambiguity call for a beneficent engineer. Because of the nature of the situations described, we found that the most relevant virtues were, in fact, moral virtues. In addition, we find that these will require the exercise of many other virtues of different kinds. Our analysis also indicates that without a central moral virtue and without practical wisdom, there is perhaps no guarantee of goodness in the exercise of other, more technical virtues.

Our analysis reinforces existing scholarship on virtue ethics and engineering and highlights the need for virtues-based approaches therein, with a specific focus on engineering ethics education. New approaches are needed to complement the idea that uncertainties can either be avoided or eliminated through more knowledge. Engineering education could also embrace different situations of uncertainties in order to give space for the cultivation of other virtues; or, as MacIntyre writes, we should have the ability to live with uncertainty.

There are further avenues for research that we have not touched upon in this chapter. First, we have not discussed vices corresponding to the virtues we present. Doing so could bring to light virtue conflicts and present a more complete account of designing under uncertainty. Second, we have not explored poietic virtues (Poznic & Fisher, 2021) as it would require a more sustained interaction with empirical material to add to this discussion than this chapter would allow. Nevertheless, doing so would present additional opportunities to deepen philosophers’ and ethicists’ empirical engagement with design situations under uncertainty. Third, empirical research could broaden the array of design cases and possibly relevant virtues for engineers, which could, in turn, broaden and/or confirm our idea that some virtues lend themselves better to specific types of uncertainties.

We would like to end this chapter by once more stressing the importance of experience and education for shaping responsible engineers. That is, if we want virtuous engineers ready to tackle an uncertain future, we need to think about how to cultivate the necessary virtues through engineering education in pandemic times and beyond.