Attitudes research as a direct source to answer ethical questions: theoretical, methodological and pragmatic considerations

Currently, different positions coexist regarding the interplay of empirical and normative elements in medical ethics.15 ,16 Moral philosophers have warned against ‘ethics by opinion poll’ⁱ which disregards the need for normative analysis.17 Along these lines, it is said that direct inferences from empirical data about people's moral attitudes to the ethical correctness of these attitudes should be avoided. As normative elements play an irreducible role in ethical judgement, they have to be made explicit.12 ,13 However, there are some ethical-theoretical backgrounds which allow for a direct derivation of normative claims from empirical statements. Moral intuition, for example, is considered as the basis of ethical reflection in the different forms of ethical intuitionism.18 ,19 Beginning with such a specific view in moral epistemology, intuition can be regarded as providing a direct access to moral truth. On this basis, research on attitudes may be seen as a means to find a direct answer to normative questions. A similar point can be made for hermeneutic approaches in the field of empirical medical ethics.20 ,21 Here, the implicit moral knowledge of people who are experienced in a certain practice is elicited and serves as a basis to find an answer to ethical challenges. The normative solution here is developed by means of empirical research, for example, by encouraging dialogical processes between different stakeholders.

Despite these approaches, in this article, we will advance a sceptical position against attempts to answer ethical questions directly and exclusively on the basis of empirical research. Our reasons for this scepticism are situated on three different levels.

a) On the theoretical level, which has mostly been discussed so far, several philosophical problems are repeatedly mentioned in the literature, such as the is-ought problem, the naturalistic fallacy and the fact-value distinction.12 ,17 While naturalistic fallacy and fact-value distinction refer to philosophical problems which lie on the metaethical level and do not directly apply to the problem of empirical research in ethics, the is-ought problem is, in fact, a major challenge to attempts to derive normative statements from descriptive facts alone.12

b) Regarding methodology, attitudes research in medical ethics is confronted with the challenge that it does not often become clear which kind of empirical data is actually collected when asking certain groups of people for their moral attitudes.8 ,22 The answers may be expressions of intense moral reflection, as well as products of self-interest, self-deception or historical or cultural accidents. However, it is crucial for medical ethics, as a normative discipline, to know about the nature and origin of moral statements in order to assess if they are reliable and can, therefore, serve as a basis for ethical argument.

c) Finally, there are pragmatic considerations which speak against attempts to answer normative problems in medical ethics on the basis of empirical data on moral attitudes alone. Many highly disputed ethical issues in medicine (eg, treatment decisions for severely ill newborns or patients in a persistent vegetative state) are characterised by the fact that the opinions of relevant stakeholders cannot be explored empirically as these stakeholders do not have the capacity to articulate their views. Furthermore, there is often no consensus among the different parties which are involved in the situation. Whose position should be followed does not emerge from the empirical research itself, but requires further normative inquiry.

These limitations—while they can only be sketched here—suggest that attitudes research in medical ethics may not be sufficient in itself to develop answers to normative questions. What is still needed is a framework of normative justification which clarifies the data's relevance for ethical argument. In the following section, it will be shown how data on people's moral attitudes can be integrated into an ethical framework provided by ethical theories.

Ethical theories as normative frameworks

If no direct inferences are to be made from empirically researched moral attitudes to these attitudes’ ethical justification, additional elements must be introduced in the reasoning process to allow for normative conclusions. What is needed here is a theoretical background clarifying which empirical data should be regarded as ethically relevant and which role they should play in ethical judgement. In the following, we will discuss how far ethical theories are able to provide an adequate background. In addition to the more general remarks on this topic which can be found in the literature,6 ,13 ,23 we substantiate that the theory-dependence of empirical data holds true with respect to three different aspects: (a) The extent of empirical data that is needed, (b) the question of which kind of data is required and (c) the ways in which the empirical data are processed within the framework of an ethical theory.

a) Currently, a great variety of ethical theories coexist in the field of applied ethics and, more specifically, medical ethics.24 Authors writing on normative issues in medical ethics refer to a wide spectrum of normative backgrounds which can be delivered by theories stemming from normative philosophical ethics, theological ethics or theories and concepts which have been specially designed in the context of medical ethics. This pluralism of normative theories is relevant for the practice of empirical-ethical studies: There are some theoretical backgrounds that are very much in need of the provision of empirical data, while other theories are less dependent on the results of empirical research. This last characteristic particularly applies to normative approaches which reject or agree with certain practices in medicine for very fundamental reasons. An example would be a theological position towards euthanasia which refuses this practice on the basis of the normative conviction that human life is given and taken by God and should not, therefore, be terminated by man. Ethicists advancing such very fundamental arguments are not so much dependent on the provision of empirical data on people's moral attitudes. Knowledge about the stakeholders’ preferences would not alter their argumentation nor would it make them arrive at different conclusions.

b) In contrast to this, moral attitudes play a major role in ethical evaluation in many other normative approaches which are applied in medical ethics. However, depending on the ethical theory selected, different kinds of data are required. Theories which are based on central ethical concepts, such as ‘autonomy’ or ‘dignity’,25 need empirical information to illustrate the ways in which these concepts are realised in the context given. Other ethical theories imply a special interest in data which are emerging from attitudes research. Preference utilitarianism, for example, is an approach which places special emphasis on people's attitudes and interests.26 The interests of relevant groups of stakeholders regarding a number of controversial issues in medical ethics are not yet well-known, such as patients’ perspectives in medical decision-making. Therefore, attitudes research is very useful in these fields to provide the data which are required for judgements on a preference utilitarian background.

Nonetheless, when an account such as preference utilitarianism is applied in medical ethics, special attention must be directed towards the sort of data which is taken for normative deliberation. The methodological problems raised in the former section of this article again become relevant here: Whose preferences can and should be considered? Who are the relevant stakeholders in ethical decision-making? Which preferences qualify to be used for ethical deliberation? There are some positions in preference utilitarianism which do not distinguish between comprehensible or rational and non-rational preferences, while other positions try to establish criteria for ‘true preferences’ in contrast to merely ‘revealed’ or ‘expressed’ preferences.27 It would be crucial for an empirical-ethical study on the background of preference utilitarianism to determine if stakeholders’ preferences are expected to be qualified in some way, or if they can be directly used as emerging from the empirical research (eg, as indicated in a questionnaire).

c) Ethical theories differ in the ways in which they process empirical data in order to arrive at a normative judgement. In the aforementioned example of a utilitarian theory, the empirical data are taken as a basis for a utilitarian calculus which seeks the maximum accomplishment of people's preferences. People's preferences are weighed and counted against each other to determine which state of the world is in accordance with the greatest possible number of preferences expressed.

Other ethical theories refer to empirical data in different ways. Reflective equilibrium approaches, for example, are very popular in applied ethics28 and have also already been applied in a number of empirical-ethical studies.29 ,30ⁱⁱ Reflective equilibrium approaches represent a form of ethical deliberation which attaches great importance to moral attitudes and intuitions but, at the same time, integrates them into a broader framework of ethical justification. In the case of a wide reflective equilibrium, well-considered moral judgements and principles, and so-called ‘background theories’, which can have a moral as well as a non-moral character, are considered.32 In contrast to utilitarianism, moral attitudes are not counted and weighed, but they undergo a process of further qualification. During that process, considered moral judgments are adjusted and modified in comparison to other elements to arrive at an ethical judgement about the issue at stake.

However, certain limitations have to be considered when a reflective equilibrium approach is applied to an empirical-ethical study. A central challenge in this context is that philosophical models of reflective equilibrium are usually based on the notion of ‘the thinker’ as the (fictional) person who does the reasoning process. In contrast to this, attitudes research in medical ethics is concerned with the moral convictions of a number of people. The reasoning process (the ‘normative analysis’) is often done by the researcher away from the field. Therefore, empirical-ethical research has to provide strategies of how the moral experiences and deliberations of third persons can be considered29 to overcome the status of a doxastic justification (justifying an individual's moral beliefs) and to arrive at a propositional (deindividualised) justification of ethical judgements.33 In summary, it can be said that reflective equilibrium provides a useful background for an integration of attitudinal data into normative reflection. However, the philosophical model has to be developed further to provide an adequate framework for empirical-ethical research.

In summary, the use that can be made of empirical data on moral attitudes is heavily dependent on the ethical-theoretical background. Furthermore, philosophical theories are often restricted to certain parts of ethical evaluation, while a more comprehensive notion of ‘ethical judgement’ is required in applied ethics. In the following section, we will discuss further functions which attitudinal data can fulfil and which lie beyond the ‘traditional’ focus of ethical theories.

Further functions of attitudes research in medical ethics: explorative research and moral pragmatics

Ethical theories usually concentrate on the justification of ethical judgements. However, we consider it crucial in applied ethics research to overcome the scope of ethical theories and also take into account aspects of practice-oriented ethics which are often not captured in philosophical theory. In applied ethics, further steps have to be taken into consideration, which precede and succeed the ethical judgement proper. These further aspects also represent additional fields in which attitudes research can be successfully employed. Two of the additional functions which attitudes research can have in this context will be discussed in the following pages, namely the ideas of (a) ‘explorative empirical research’ and (b) ‘moral pragmatics’. The first concept concerns the preconditions of ethical judgement and, therefore, has a logically upstream character to judgement itself. The second additional function addresses considerations which have to follow after the moral judgement has been made.

a) At the beginning of every ethical analysis of a concrete issue (such as a new medical technology or a societal development) is the question of how to identify and characterise the ethical problems existing in the respective field. In the process of identifying ethical challenges and relevant features of the situation, important points may be overlooked without an intimate knowledge of the stakeholders’ views and experiences. Empirical research on moral attitudes is also able to detect formerly unknown ethical problems or aspects which suggest an alternative judgement about certain issues.11 ,12 Furthermore, research into stakeholders’ perceptions and views can reveal answers and practical solutions to ethical problems which have been developed and successfully applied in practice. These different functions of attitudes research can be categorised under the label of ‘explorative empirical research’, as they deliver new information which, however, has to undergo further investigation regarding its soundness and consistency.ⁱⁱⁱ Along these lines, explorative empirical research is expected to bring about new information which is relevant for an ethical evaluation and may have slipped the attention of the ‘armchair philosopher’ so far. One example would be a qualitative study on young adults who had been conceived by in vitro fertilisation (IVF).34 Asked for their personal attitudes regarding this medical technology, one pro argument which was brought forward was the knowledge of being a deeply wanted child. This was a new and unexpected point for the researchers which is relevant for the ethical discussion on IVF.

Nevertheless, the empirical information revealed by explorative research should undergo a critical evaluation regarding its soundness and consistency and its coherence with other relevant features of the situation. Coming back to the study on IVF, what normative weight should be attributed to the study participants’ experience of feeling ‘wanted’ must be discussed in comparison to other factors which may play a role in an ethical evaluation of IVF practices. If an argumentation rejects IVF for fundamental reasons, for example, because of its ‘artificiality’, the stakeholders’ views will play more of a minor role than if other ethical accounts are taken as an argumentative basis. Therefore, explorative research cannot be described as being fully independent of underlying normative assumptions and ethical background theories.

b) Attitudes research has a role to play before the beginning of, and subsequent to an ethical evaluation: It helps to answer the question of what should be done after an ethical judgement has been formulated. When the normative judgement is fixed, exploring and analysing the pragmatic conditions of bringing the theoretical position into practice must be seen as the next step. This practical, for example political or educational, implementation of ethical norms can be described as ‘moral pragmatics’, an expression coined by the German philosopher Dieter Birnbacher.^iv35 In order to address the pragmatic circumstances, ethics needs empirical information about people's actual thinking to implement measures which are constructed in reference to this social reality. This is of special importance, as ethical guidelines or interventions have to be in line, at least partially, with people's already existing attitudes and behaviours to gain the motivational power which is necessary for these interventions to become action-guiding. A study, for example, which shows that a major section of the physicians in a country are opposed to a legalisation of physician-assisted suicide, may provide important information for the implementation of a new legal regulation on this issue.36 It informs the decision-makers that developing guidelines for how to deal with physicians who have objections against assisting patients with their suicides will be a crucial point.

Moral pragmatics is an important task of medical ethics which is highly dependent on empirical information. Knowledge about people's moral attitudes can be used here without being integrated into the framework of a specific ethical theory. Nevertheless, some account of normative justification has to be applied in advance to arrive at the ethical judgement which forms the basis of the following pragmatic considerations. Therefore, this function of attitudes research cannot be regarded as being completely independent of a theoretical background. However, the task of moral pragmatics is an important aspect in the practice of medical ethics which is often not captured in philosophical theory.

The two functions of ‘explorative empirical research’ and ‘moral pragmatics’ and their relationship to a moral evaluation on an ethical-theoretical basis are illustrated in table 1.

Further methodological perspectives

There are some methodological and practical questions regarding attitudes research in medical ethics which can only be sketched here and should be the subject of future research. A more detailed discussion, for example, would be worthwhile about the different functions of quantitative and qualitative research on moral attitudes for normative reflection in medical ethics. Some arguments have been put forward in the literature which could lead to the perception of qualitative research being the better alternative38 ,39 as it provides the researcher with a better opportunity to explore the research subjects’ motives. However, there may be other contexts where quantitative research is more useful, for example, in political decision-making on bioethical issues which relies heavily on majority opinions.

A second point for further discussion is the interaction between the sociotheoretical background of the empirical methods which are applied in attitudes research and the ethical theories which are used in the respective research projects. Ethical theories often include empirical or anthropological premises, while social theories are frequently based on certain normative assumptions which are not made explicit. Therefore, a mutually critical look from both sides could be helpful to make the respective background assumptions explicit. Different ways of combining the diverging theoretical backgrounds could be chosen for empirical-ethical studies to achieve such a constructive mutual critique.40

Furthermore, there is still a need for the development of quality criteria for empirical-ethical studies. High-quality empirical research alone must be seen as a necessary but insufficient condition for empirical-ethical studies, as it does not cover the normative dimension of the topic. Hence, empirical-ethical studies have to meet the basic scientific standards of empirical and normative methods.41 ,42 Furthermore, there are also criteria pertaining to the communication between the two sides which are not yet well developed.

Outlook for research practice

In the preceding sections, we have tried to provide a systematic outline of the conditions under which empirical data on people's moral attitudes can be used for normative deliberation. Based on what has been illustrated so far, we suggest the following concrete points to be considered by researchers who are planning and conducting attitudinal studies in medical ethics.

• Researchers should reflect on how far the planned study is supposed to contribute to the normative debate, or whether it is meant to remain on the empirical level. If the study is supposed to enrich the normative debate, a normative background (eg, ethical theory, normative concepts, ethical principles, etc) should be selected and be made explicit to clarify the normative dimension of the study.

• As we have shown before, the selection of a specific ethical-theoretical background has a decisive impact on the question of which kind of empirical data is needed. Therefore, the choice for a normative basis should take place along with the development of the empirical research design. Given the theory-dependence of the empirical data's ethical relevance, a strategy of gathering data before any reflection on the way of analysing them normatively would not be advisable.

• Attitudinal studies which are supposed to deliver a contribution to normative discussion are very much dependent on an open and close interdisciplinary cooperation between representatives of the empirical sciences and ethicists with a philosophical background. A basic mutual understanding of each other's methodology is thereby required for fruitful cooperation.

• The empirical methodology and the normative background of an attitudinal study in medical ethics should be made explicit in the publication of the data. Furthermore, the intercourse between the empirical data basis and normative claims should be made transparent.

Therefore, diverse issues on the border between socioempirical research and ethical theory have to be considered and further developed to facilitate high-quality work in medical ethics which is, at the same time, empirically informed and theoretically sound. Being reflective on these issues, empirical research on stakeholders’ moral attitudes will inform the ethical debate, and will be an indispensable element of normative work in medical ethics.