Introduction

Care ethics requires a typical care-ethical view on qualitative empirical research in ethics. Amongst many propositions with regard to the relation between ethics and qualitative research, we wish to put forward one unusual but interesting possibility, namely that of what we call ‘N=N case studies’, i.e. inquiries into one or more meaningful exemplary cases, that function as a prism of many ethically relevant issues.

Before outlining our proposition it is necessary to chart the underlying care ethical presuppositions. ‘Presupposition’ here is not meant as an axiom, but as a critical insight gained from earlier research. The question that, in general, preoccupies us as care ethicists is: how can we know what is good care?Footnote 1Footnote 2 On the basis of plausibility gained from past research, we have come to the understanding that what is good in care practices can be found within care practices themselves. What is good not only emerges in practices, but can also be discerned on its own terms, in its actual shape, within care practices. It does not come into existence by applying a normative theory, ethical principles or by indicating pre-given moral values. We argue that it would even be misguided to attempt to introduce the good from outside, either by ethicists, by executives, politicians, or consultants. We are looking for practically proven ‘stepping stones’, political–ethical insights that originate in the practices studied and enable practitioners to deal with newly emerging issues, containing—sometimes probing—moral problems.

A second insight from past research is that to perceive well one needs adequate concepts and theories; what is adequate can only become clear in actually doing empirical research. In our view, digging into care practices requires a specific combination of qualitative empirical research, theory development and ethical reflection that we have come to call ‘empirically grounded ethics of care’. The term ‘grounded’ expresses our aim to develop (political–ethical) concepts, insights and theories that do justice to—are ‘grounded’ in—our (construction and analysis of) empirical material. It refers to the so-called ‘grounded theory approach’ in sociology, actually a family of methodologies to develop (substantive and formal) theories that are ‘grounded in data’ (Glaser and Strauss 1967, p. 33).

In this paper, we discuss and illustrate the specific combination of empirical research and theory development we have been putting in practice. Furthermore, we propose standards to evaluate the rigour and quality of this kind of inquiry and its findings. We do this against the background of existing models of qualitative empirical research and theory development in the behavioural and social sciences. We contrast the nature of our proposal while outlining the current discussions and proposals: plausibility is gained by arguing for and against positions. While we do not wish to pretend that we once for all have defined the appropriate relationship between empirical work and theory development in the study of care practices, we do want to make a mid-way proposal. In short, our view is that extensive descriptions can be condensed in various newly defined terms, terms of their own, and these terms can be far-reaching moral signifiers.

As our aim is to study real-life complexity, on a more detailed level of research our ‘empirical material’ (Alvesson and Kärreman 2007, p. 1265) typically consists of extensive and comprehensive case descriptions. It is for this reason that we characterise our research as ‘case studies’. But what kind of case studies? In the last paragraph, we briefly explain our choice of a particular type of case study, propose a specific name for this category and concisely discuss four quality criteria.

Care ethics

In this section we outline our take on care ethics and the corresponding arguments that are closely related to our view on empirical research. Care ethics is an interdisciplinary approach (1) that understands care as an everyday practice that enables to live together, and therefore is eminently political (Tronto 1993, 2013). (2) Caring practices are socially and institutionally embedded, care not being accessible in a free floating ‘essence’, not ontologically or even metaphysically determinable. (3) Relational interaction in care ethics is seen as both a locus of knowing (i.e. an epistemologically relevant as well as contested locus) and a locus of recognition and attention (Klaver et al. 2014; further references there). (4) Within the explicitly broad view on care good medicine, health care, help, and support are always context- and situation-specific, and can only be validated locally—that is, via the experience of the care receiver (Baart and Vosman 2015b; Klaver et al. 2014). (5) Care ethics as we practice it, studies practices of medicine, health care and social welfare, help, and support, both empirically and theoretically and not just philosophically or via a general political theory only (Baart and Timmerman 2016; Klaver et al. 2014).

Contested intuition

Care ethics began with empirical research carried out by the moral psychologist Carol Gilligan and her co-workers. In her famous book In a Different Voice, she mentioned an observation she made halfway through 10 years of research into the way people talk about morality and about themselves (Gilligan 1982). Her argument, based on three empirical studies, was criticised because she was unable to substantiate the claim that women and men reason differently when it comes to morality (e.g. Luria 1986). Her argument however was not about a difference between men and women and not based on statistics. It was about a correction of moral psychological theory and based on a sufficient number of exemplary cases, but comprehensively and extensively described—not on the representativeness of the samples studied. Her argument ‘was interpretive and hinged on the demonstration that the examples presented illustrated a different way of seeing’ (Gilligan 1986, p. 326). Gilligan’s book provoked much response, including many affirmations. According to Margaret Urban Walker, Gilligan offered a new and productive ‘moral vision’ (Walker 1989, p. 127), consisting of the two independent themes Walker discerned in Gilligan’s different voice: a substantive one—‘a care and response orientation’—and a methodological one—‘a contextual-deliberative picture of moral thinking’ (Walker 1989, p. 124). Gilligan’s provocative book is seen by many as the start of what is now called ‘ethics of care’ or ‘care ethics’ (Dingler 2016).

Four formal characteristics of Gilligan’s studies are particularly inspiring for our work: (i) they started with an observation and an intuition, (ii) the argument was based on a sufficient number of exemplary cases, comprehensively and extensively described, (iii) the theoretical concepts she developed were innovative, and (iv) besides criticism and even rejection, they provoked much affirmation. In our own research we have added two more features: we study what participants in practices of care do and refrain from doing, and what they undergo (v); this relates to a non-activist view on action, action rather conceived as part of a practice. Furthermore, we position care-givers as co-researchers, situating them as joint contributors and researchers inquiring into the moral goodness of their own practices, with their own responsibilities (vi); this relates to what research, or rather inquiry with an ethical interest, is about. We prefer to think about our endeavours as ‘inquiries’ rather than as ‘research’, a concept that is associated with subject-object presuppositions. As scholars and participants in a practice, we delve into that particular practice, and look at what gives it its orientation (Baart 2001). The scholar is not without a stance in the practice, is not an ‘atopos’ (Melançon 2014).

Empirically grounded ethics of care

As care ethicists our aim is to contribute to the maintaining, promoting, and strengthening of practices of good care by identifying such practices and by analysing why participants evaluate these practices as good care, and by asking how such practices can be facilitated. We seek to gain an understanding that can help participants in practices of care to come to terms with reality and to act in situations where it is not obvious what the right way to act is. In this paper, we concentrate on care professionals, but we tend to think that ceteris paribus our proposition is valid for caring practices in general. We realise that in late-modern society professional practices take place in highly complex organisations (Vosman et al. 2016). This insight allows for acknowledging the paradoxes in late modernity that reign over practices, instead of singling out individual actors that are deemed to be conscious and responsible of their doings and in command of the circumstances. This also recognizes the entanglement of doings and sayings in complexity, rather than looking at human action as in principal free acting. The qualitative research that we conduct therefore (i) acknowledges the internal morality of practices, (ii) uncovers deeper influences and structures that are not immediately obvious but that nevertheless have a bearing on the care practices in question and index inquiries, and (iii) delivers theoretical concepts and theories that are conducive to good care and are not just descriptive.

Triangle of relationships

Empirically grounded ethics of care is how we call the specific combination of qualitative empirical research, theory development and ethical reflection we endorse. To undergird this proposal we have to methodologically account for the complex, intersecting relationships between empirical research and ethics (flank I of an imaginary triangle), between (social and political) theory and ethics (flank II) and between empirical research and theory development (flank III). We believe that these three flanks have to be considered in their intertwined connectedness. As care ethicists concerned with political ethical theory, we find that this connection lies in a phenomenological inquiry into the lived experience of the persons who act and suffer (care receivers and care givers), and their first-person perspectives. Considerations inspired by care ethics and extensive qualitative empirical research experience cause us to position ourselves, not in the epistemological-egological tradition of phenomenology (Husserl), but in that of political phenomenology (Waldenfels 2011; Vosman 2018; Vosman et al. 2018).

In empirical bioethics, an intensive debate is ongoing about its methodology and standards (Christen et al. 2014; Davies et al. 2015; Ives et al. 2018; Wangmo et al. 2018). This debate however is almost exclusively focused on the relationship between empirical research and ethics, ignoring the other two flanks and their connectedness. We have recently discussed our stance regarding this relationship of empirical research and ethics in its connectedness with the other two flanks (Baart and Timmerman 2016; Vosman at al. 2018). Essential aspects of our position are (a) the assumption that reality (and not just care) is morally laden, not neutral, (cf. Taylor 1989) and (b) a kind of ethical reflection that helps to articulate this moral ladenness.

The relationship between empirical research and theory development, in its connectedness with the other two flanks, is what concerns us here. This interconnectedness is charted by the German sociologist Gesa Lindemann. She has pointed out that data collection and analysis are not free of theoretical assumptions. According to her, these assumptions cannot be falsified by empirical data and cannot thus be further developed. Drawing on Georg Simmel, she has proposed a distinction between three levels of sociological theory development (Lindemann 2008, 2009). The first level encompasses ‘social theories’, containing the theoretical and methodological assumptions about what a social phenomenon is and what concepts are crucial to studying it. The second level consists of ‘limited range theories’, theories about specific social phenomena including ‘grounded (substantial and formal) theories’ and ‘middle-range theories’. The third level consists of ‘theories of society’, i.e. ‘theories about large historical formations such as modern society, capitalist society, or functionally differentiated society’ (Lindemann 2009, p. 6).Footnote 3 Lindemann has further shown that the relationship between theory and empirical research has to be conceptualised in different terms on each of these levels. Thinking in terms of verification and falsification is adequate only on the second level, that of limited range theories. She has proposed that the empirical basis of theories about society (level three) should be investigated with the help of the difference between plausible and implausible Gestalt extrapolation (Lindemann 2008, p. 124; 2009, p. 10), and that of first-level theories with the help of the ‘precise/troubled distinction’ (Lindemann 2008, p. 124, 2009, p. 14). Social theories can be understood and pursued as heuristic devices: they do not define nor interpret what is perceived but allow for disclosure of what presents itself. The researcher’s troubling experience (Irritation) that they cannot capture reality precisely enough by using these devices, can become a motivation to develop them further (cf. Baart and Timmerman 2016; Vosman and Niemeijer 2017). Our inquiries deliver on the second level, but also and substantially on the first level, that of concepts and theories as heuristic devices to perceive well.

The empirically grounded ethics of care that we are proposing also involves an intertwining of empirical research on the one hand, and concept and theory development on the other. Theoretical and conceptual reflections are used to recognise details that are particularly relevant and are saturated by complexity, to perceive more closely what has been perceived, to perceive what was not perceived before, and to give a plausible account of what has been discerned. Empirical research is carried out to operationalise and differentiate abstract concepts (relationality, vulnerability, corporeality) and to repair and enlarge deficient concepts (autonomy, self-reliance, co-creation, shared decision making). Empirical research makes it possible to ‘talk back’ to theory and to ask questions that stimulate further, relevant theory development. This sometimes also implies showing how the use of a particular theory, e.g. on vulnerability or power, has become shallow or uninformative. Going beyond mere descriptions of experiences, our aim is to develop concepts and theories grounded in our empirical research. We propose an ‘oscillating’ movement between empirical work and theory development, committing both to each other. ‘Oscillating’ means swinging in an intended, methodological way, detecting what questions the observed practice raises, as well as what questions the theoretical notions used raise, rigorously inquiring into these questions, and updating the preliminary findings.

In this oscillating movement between empirical work and theory development, research by other scholars in ethics and social science is important—both for empirical work and for theory development. The empirical and theoretical research findings of these ‘fellow travellers’ can broaden our knowledge, raise questions, or offer concepts that support and maintain the oscillating movement. Researchers can thus gradually gain a ‘deeper’ understanding of the situation or the practice, develop ‘clearer’ theoretical concepts as it were, and avoid both empirical naivety and fact-free reasoning. The inquiry is fuelled by observations of what seems to be athwart, what doesn’t make sense, and prepares the way for a different perspective, possibly also a different theoretical take on the practice, if the theory used does not fit or ignores the fine fabric of the care practice. The focus is always on a practice (Nicolini 2012; Schmidt 2012), and primarily on acting and undergoing (or to put it in phenomenological terms: the passibilité) of living beings (Vosman et al. 2016), and on the knowledge and practical wisdom of the persons who offer care and help, here professionals. At the same time, theoretical understanding is developed further, from concepts to grounded theories to more comprehensive theories that point beyond the concrete situation or practice. This reciprocal approach focuses primarily on theory development at the first two levels that Lindemann has distinguished (‘social theories’ and ‘limited range theories’). Our proposal has two prerequisites: (a) the assumption that every perception and experience is theory laden (cf. Hirschauer 2008) and (b) a procedure that helps to articulate the theoretical implications of the practice instead of distorting them or even replacing them with inappropriate theoretical concepts.

Our inquiries: acknowledging complex, real-life practices of care

So much for the theory. In practice, we have been carrying out qualitative inquiries which bring to light what participants in practices of care do and refrain from doing, and what they undergo (Vosman et al. 2016). Care-ethical ‘critical insights’ guide us in this search for good care. Our interest is a critical phenomenological one: what phenomena become apparent in care practices, and what do people who partake in the practice perceive? What do they reflect upon? How do they act and how do they evaluate their actions? What things do they undergo and how do they call these things? Those who are involved in the practice studied, especially care givers, are positioned as co-researchers. We are also interested in the development of theory. We use ‘grounded theory’ procedures and criteria, such as the researcher’s theoretical sensitivity, the inscription of experience and observation in diary entries, field notes, interview transcripts, et cetera, the coding along various dimensions (initial, focused, axial, and theoretical coding), constant comparison, memo writing and sorting, saturation, and theoretical sampling.

We discuss three examples of our care-ethical empirical research: (i) a case study of urban neighbourhood pastoral care (by Baart), (ii) a case study of how hospital staff deals with a patient’s relatives (by Baart, Vosman and others), and (iii) a case study of end-of-life care by general practitioners (by Baart and Timmerman 2016).Footnote 4 Before presenting these in more detail, we discuss a number of general features that occurred in all three cases, and that exemplify what we are proposing. Each of these three studies started with a strong intuition, i.e. that sound and thorough qualitative research would produce an insight into the practice in question that would be substantially different from, and more conducive to good care than, current ideas about the practice. In each of these studies, we studied a limited number of cases, which were nonetheless comprehensive and extensive. A central focus in these cases was what participants in the practice in question perceive and reflect upon, and how they assess, weigh, act and evaluate—i.e. not their beliefs or opinions, or their responses to prepared vignettes. The three studies that we discuss are based on analyses of two, one, and six cases respectively.Footnote 5 Our inquiry involved various kinds of extensive qualitative analysis. In addition, we carried out a considerable amount of conceptual and theoretical work so as to make sound observations and to properly articulate what we observed. We subsequently presented a reconstruction of the outcome of our analysis to the practitioners who participated in the study, and they supplemented, corrected, and ultimately generally endorsed this.

Urban neighbourhood pastoral care

Baart’s study about urban neighbourhood pastoral care in the Netherlands, a form of outreach pastoral care in disadvantaged, resource-poor urban neighbourhoods, was initiated by the interest in what these workers actually do and in the vitality of these neighbourhoods (Baart 2001; Baart and Vosman 2011; Timmerman and Baart 2016b). Although the most comprehensive of the three, we discuss it only briefly. During 9 years, Baart followed two workers closely, interviewing them monthly, and analysing case descriptions, reflection reports, diary excerpts et cetera. More than 100 times, the workers were asked for feedback on preliminary findings. The empirical work was accompanied by conceptual and theoretical studies about distantiation/engagement, professionalism, and poverty. Eventually, the researcher characterised what the workers do as the practising of ‘presence’, basically understood as a qualified, efficacious way of being for the other by being with the other. The core of the practice of presence is relational work: connecting with and attuning to the lifeworld, life course, longings, concerns and turning points of the other(s). The theoretical outcome of this longitudinal empirical study consisted of ‘a theory of presence’, an empirically grounded, care-ethical theory explaining why the practising of presence works and is valuable, and specifying which particular understanding of common humanity it implies. This theory can be located at Lindemann’s second level of theories, that of the limited range theories.

Since the publication of the research report, the theoretical elaboration and justification of the practice of presence have been received very positively, also far outside the realm of pastoral care. The vocabulary of presence has come to be appreciated by social workers, nurses, youth workers, workers in addiction care, general practitioners et cetera, because it helps them to articulate their own motivation and aims, and their frustrations with the way they are increasingly being trained, assessed, and managed.

Hospital staff dealings with a difficult family

During 2009–2014, Baart and Vosman carried out a major transition and research project in a general hospital. In this framework they conducted a short, but comprehensive and complex study of the last 3 months of a patient and her family, in the clinic’s daily routine perceived as difficult, dissatisfied and inappropriately responsive. Baart and Vosman were interested in what this family actually experienced, how the care givers actually responded and how these were related (Baart and Vosman 2015a). Using phenomenological, content, thematic and discourse analyses of interviews with relatives, nurses and doctors, of patient records, and of the daughter’s diary, these last 3 months were reconstructed and interpreted from very different perspectives. As part of this case study, a 5-h interview with the widower and the daughter of the deceased was analysed, with a focus on the change in the course of the emotions expressed. To understand how the emotions of the family vary over the episodes, we had to see that this variety of emotions is not, as is often assumed, divided into phases, but is always present in layers. During almost the whole period, sadness is continuously present as a kind of basso continuo, except in the last two episodes where it is completely absent. Often there is a thin layer of desperateness. In turn, thick layers of anger and, most importantly, fear can be seen, which can sometimes swell and even come to the surface. When that happens, the top layer of positivity disappears. This episodic course of emotions could be linked to the timeline, which lists the incidents in the care for the patient and her family. Then it became obvious that the incidents were directly related to the disappearing of the upper layer and the coming to the surface of fear and anger, provoked by the declining attitudes of the carers and their systematic underestimating of the (relevance of the) family knowledge. At such moments, a pinprick was enough to make the family’s behaviour escalate in the eyes of nurses and doctors. This theory about layered emotions is equivalent to Lindemann’s first level of theories, that of the theoretical and methodological assumptions about what a phenomenon is and what concepts are crucial to studying it.

The results of the analyses were presented to the nurses, doctors and managers involved in this case. The nurses generally endorsed the reconstruction of the case. The idea of layered instead of phased emotions was experienced as helpful.

End-of-life care by general practitioners

Our research into end-of-life (EoL) care by general practitioners (GPs) in the Netherlands was inspired by our amazement at the way one specific GP was treated, and how his actions in a specific case were judged (Timmerman and Baart 2016a). This GP gave an overdose of morphine to a long-time patient whose condition had deteriorated sharply, after which the patient died to the relief of the patient’s spouse. Before the GP had the opportunity to explain and justify his actions, he was subjected to an accusative public debate and got accused by the National Health Inspection of improper handling. He committed suicide. We were surprised at how little informed society is about what GPs in general perceive and reflect upon, and how they assess, weigh, act, and evaluate at the bedsides of their sick and dying patients, this despite much quantitative and qualitative research. Our intuition was that it would be worthwhile and relevant to study this lack of understanding, and that heuristic use of the concept of ‘practical wisdom’ would be helpful.

The empirical material was constructed by conducting 11 in-depth, semi-structured interviews and five focus group interviews with six experienced GPs. This resulted in extensive and comprehensive information about six cases, one of each of the six participating GPs, and fragmentary information about circa 80 more cases. We reconstructed the EoL care process in such a way as to retain rather than reduce and unduly simplify its complex, dynamic, and emergent character. To discern what EoL care by GPs is about, we had to reconstruct several, complementary ways for GPs to determine pain and suffering. To reconstruct the whole process, we had to develop our own vocabulary, grammar, and syntax. To preserve the emergent dynamics of the process, we had to introduce the idea that the current situation and process is continuously interpreted through images and stories, references to lived reality, and an imagined synthesis (‘what will the final situation be?’). The different concepts and theories we developed or elaborated upon in this study can be located at Lindemann’s first and second level of theories.

The participating GPs responded affirmatively to the final presentation of our findings. The gist of their comments was that the vocabulary we developed was particularly helpful to think about what they do, talk with colleagues about what they do, and give a public account of what they do. This was confirmed by other GPs’ responses to presentations about the outcome of our research.Footnote 6

These three exemplary case studies resulted in the reconstruction of (i) what has become known as the practice of presence, (ii) shifts in the layered emotions of hospital patients’ relatives, (iii) the practical wisdom of general practitioners. An important criterion for the quality of our analyses is the nature and extent of the approval these reconstructions receive from the participants in the practice researched. Many participants acknowledged and approved our work in these three studies, often expressing recognition: ‘Yes, that’s how we do that! Yes, that’s how it is for us!’ There was a substantial degree of recognition, and this recognition was critical of received ideas about the object of study. Our research has shown respectively that, contrary to what is often believed, (i) assisting people in dire social conditions is not about solving problems but about relentlessly staying with them and their troubles; (ii) in hospital emotions of relatives are not experienced in distinct phases, rather in layers that flare up, and (iii) GP’s have a nuanced practical savoir faire at EoL care, rather than operating in a ‘rational choice’ mode. In neither of these studies, the number of cases was determined by considerations of (theoretical) saturation on the level of cases. The cases studied, however, were considered to be exemplary: rich in information, meaningful, exemplifying the practice studied, featuring practically wise practitioners, suitable for clarifying other cases, flexible, etc.

Relevant theoretical frameworks in social science research methodology

In behavioural and social science research methodology, there is an ongoing debate on qualitative research and theory development. We will discuss some of the methodologies proposed in order to account for our own methodology, methods of inquiry, and quality criteria: (1) grounded theory, (2) abductive analysis, (3) exemplar methodology, (4) mystery as method and (5) intuitive inquiry. From each of them we adopt elements in our own care-ethical way.

(1) Grounded theory

The first methodology we discuss is grounded theory, actually a family of methodologies to develop (sociological) theories on the basis of the construction and analysis of empirical material; such theories are thus ‘grounded’ in empirical material. In its most radical form, grounded theory methodology has been defined by Barney Glaser and Anselm Strauss as an inductive method (Glaser and Strauss 1967). The quality criteria of a grounded theory are: the theory must fit the research field (fit). It must address the basic processes and the most important problems in the field (relevance). It must interpret and explain the phenomenon researched (work). Grounded theory methodology is currently nearly 50 years old, and it has evolved from Glaser and Strauss’s post-positivism, through Anselm Strauss and Juliet Corbin’s symbolic interactionism and pragmatism, to Kathy Charmaz’ constructivism (Ralph et al. 2015). Since Glaser and Strauss (1967), proponents of grounded theory methodology have acknowledged that what Glaser and Strauss called ‘inductive’ was actually ‘abductive’, a word derived from Charles Sanders Peirce (see Reichertz 2007). The influential grounded theory methodologist Kathy Charmaz has added four more criteria for a grounded theory study to the criteria of quality mentioned above: credibility, originality, resonance, and usefulness (Charmaz 2006). According to her, assessing the quality of a grounded theory study necessarily involves evaluating its resonance by asking the researcher: ‘Does your analysis offer [your participants or people who share their circumstances] deeper insights about their lives and worlds?’ (Charmaz 2006: 183). We would like to adopt from this the idea of abduction as a way of accounting for the role that theoretical concepts fulfil in the process of analysis, and the idea of resonance as a quality criterion.

(2) Abductive analysis

Consistent with this, Stefan Timmermans and Iddo Tavory have contended that grounded theory methodology should be transcended by a methodology of ‘abductive analysis’:

‘Abductive analysis is a qualitative data analysis approach aimed at generating creative and novel theoretical insights through a dialectic of cultivated theoretical sensitivity and methodological heuristics. Abductive analysis emphasizes that rather than setting all preconceived theoretical ideas aside during the research project, researchers should enter the field with the deepest and broadest theoretical base possible and develop their theoretical repertoires throughout the research process’ (Timmermans and Tavory 2016, p. 180).

According to Timmermans and Tavory, abductive analysis arises from the researcher’s social and intellectual position. Is must be further aided by the methodological steps of revisiting the same observation again and again, de-familiarising the known world, and applying alternative ‘casings’ to the observations. Abductive analysis can be evaluated using three criteria: the theoretical claims have to be supported by the empirical material (fit). The theoretical claims have to be stronger than competing theories (plausibility). The theorisations matter to the ‘community of inquiry’ (relevance). We would like to adopt from this as criteria of quality the acknowledgement of the researcher’s position, the assessment of the plausibility of different theoretical conceptualisations, and the idea of relevance to a community of inquiry.

(3) Exemplar methodology

An important concept in grounded theory is theoretical saturation. The researcher continues constructing and analysing empirical material until no additional issues or insights emerge from its analysis and all relevant concepts have been constructed, differentiated and explored. As such, theoretical saturation functions as a guiding principle to assess the adequacy of the empirical material (Hennink et al. 2017). An alternative adequacy criterion, one we endorse, is exemplariness. This criterion is used in the third methodology we discuss, one used in developmental studies, whereby exemplars are studied that demonstrate the developmental construct of interest in its most consistent and intense form:

‘The exemplar methodology is a sample selection technique that involves the intentional selection of individuals, groups, or entities that exemplify the construct of interest in a highly developed manner’ (Bronk 2012, p. 1).

Important steps in sampling are the careful designation of nomination criteria, representing the standards for qualifying exemplars, and the thoughtful selection of nominators, who are supposed to use the criteria to select appropriate candidates. Nominators can be experts, practitioners or lay people. We would like to adopt from this the idea of viewing constructs of interest in its most intense form.

(4) Mystery as method

Mats Alvesson and Dan Kärreman have developed a research methodology which examines and sometimes induces ‘breakdowns’ in understanding – deviations from what is expected in empirical contexts on the basis of established wisdom. These breakdowns could be of interest to theory, leading to novel theoretical ideas. Alvesson and Kärreman call these theoretically interesting breakdowns ‘mysteries’. ‘Empirical material’ fulfils the role of inspiring the problematisation of existing theoretical ideas and vocabularies. Empirical material is conceived not as a judge or a mirror, but as a critical dialogue partner. Alvesson and Kärreman have drawn attention to friction rather than harmony in the interplay between theory, researcher subjectivity, and empirical material. Theories are instruments that provide illumination, insight, and understanding. Rather than emphasising procedures and techniques for collecting and analysing data, they are interested in the researcher’s reflexivity in dealing with the empirical material (Alvesson and Kärreman 2007, 2011). We would like to adopt from this the interplay between theory, researcher subjectivity, and empirical material, and the relevance of friction, resembling Lindemann’s ‘troubledness’ (Irritation).

(5) Intuitive inquiry

Yet another method that is relevant to our thinking about methodology is ‘intuitive inquiry’, a method developed by the psychologist Rosemarie Anderson:

‘Intuitive inquiry is a hermeneutical research method that joins intuition to intellectual precision. Intuitive researchers explore topics that claim their enthusiasm and invite the inquiry to transform both their understanding of the topic and their lives. As a method, intuitive inquiry seeks to both describe what is and to envision new possibilities for the future through an in-depth, reflection process of interpretation’ (Anderson 2004, p. 307).

Anderson describes the process of intuitive inquiry with the concept of interpretative lenses. The researcher first develops preliminary interpretative lenses, transforms and revises them in interaction with the empirical research and confronts their final versions and their theoretical implications with the literature. Anderson has proposed two new grounds for determining external validity: ‘resonance validity’, i.e. the capacity to produce sympathetic resonance in its readers, and ‘efficacy validity’, i.e. the capacity to add value to its readers’ lives. We would like to adopt from this the acknowledgement of the role of the researcher’s subjectivity and intuitions, the idea of heuristic—rather than ‘interpretative’ (Vosman and Niemeijer 2017)—lenses that can be revised during the research process, and resonance and efficacy validity as criteria of quality.

Each of these five alternative methodologies from the field of behavioural and social sciences give rise to questions that remain unanswered. They nonetheless delineate in a relevant way a field of discussion about the role of qualitative research, abductive reasoning, exemplary cases, existing theories, researcher subjectivity, intuition and reflexivity, and the community of inquiry in theory development. They also offer different accounts of how qualitative research can be evaluated and what criteria can be used to do this. Some of these criteria apply in one way or another to our inquiries, particularly: fit, relevance (to the field and to the community of inquiry), exemplariness, illuminative power, and resonance validity. Returning to the field of ethics, we now present a partial alternative.

Our proposal: N=N case study

As our aim is to study real-life complexity of inevitably morally imprinted care processes, our empirical material typically consists of extensive and comprehensive case descriptions. We therefore characterise our inquiries as ‘case studies’. Because of the goal of our inquiries, the number of cases included (N) is not decisive for their quality, as long as the analysis of these cases provides sufficiently profound theoretical insight into the practice studied. It is neither the requirements of (statistical) generalisation nor (data, code, meaning or theoretical) saturation (Hennink et al. 2017) that determine the number—although saturation can have a function in the analysis. Essential is the information richness or ‘information power’ (Malterud et al. 2016) of the case descriptions. Because (a) the number of cases is not determined by considerations of generalisation or saturation and (b) we do have the goal to develop theory, we propose to call the design of this kind of inquiry: ‘N=N case study’. A more accurate, but longer definition would be: ‘phenomenological, theory-oriented N=N case study in empirically grounded ethics of care’. By calling it N=N case studies, we distinguish our method from (descriptive) single-case studies (N=1) and multiple-case studies (N=n)—although also in a N=N case study as we propose to understand it, only one case can be studied. The validity, trustworthiness and transferability (instead of generalisability) of the findings of a N=N case study are based on the exemplariness and extensiveness of the case descriptions, the rigour of the analysis, the innovative power of the empirically grounded theoretical concepts that are developed, and the approval the findings receive from the participants in the practice.

We clarify each of these four criteria (although they deserve detailed discussion). The first criterion concerns the choice of cases. These should be exemplary, i.e. derived mostly from the practice of practitioners who are recognised as good practitioners by, for example, colleagues or patients. These cases should be reconstructed both benevolently and critically, in all their relevant dimensions. The case descriptions should be extensive and comprehensive, written after in-depth, semi-structured interviews of the participants—possibly preceded by or embedded in some kind of observation, document study or group interaction. By describing exemplary cases extensively and comprehensively, using newly developed theoretical concepts, we also construct exemplars that are suitable to clarify other cases and more easily can be transferred to other contexts and practices.

The second criterion concerns the rigour of the analysis, mainly based on construct and process validity, many criteria for which can be found in the methodological literature. Which of these are applicable depends on the research aim and the empirical material that is at hand or can be constructed. The third criterion concerns the concepts, metaphors, and theories that are developed. These should be innovative, i.e. should contribute in new and productive ways to our perceiving, understanding, and acting. The judgement of innovativeness is to, first, the researchers and, second, the practitioners who can judge the usability of these concepts. In our search for good care, we are especially interested in theoretical concepts that can help professionals to act in cases where there is no obviously right way to act, to account for their acting, and to keep anything at bay that serves goals other than contributing to good care. The last criterion is resonance validity; the approval the findings receive from participants in the practice that is studied. This is not the same as the extent to which participants in the research believe the findings represent themselves. Resonance validity involves a judgement about the response that participants give to the findings, the way in which, and the extent to which they understand, confirm, and criticise it and explain how it helps them to act. These four criteria are not like electrical switches, either on or off, but are like the strands of a rope—they all contribute together to the overall strength of the argument.

Conclusion

For a political-ethical take on care ethics that wants to do justice to actual care practices, their internal morality and the practical wisdom of the participants in those practices, and to offer ‘stepping stones’ for these participants, there is an urgent need for a care-ethical version of a further developed grounded theory design. We offer a theoretically substantiated and practically tested proposal: the so-called N=N case study. N=N case studies are phenomenological, theory-oriented care-ethical inquiries into one or more extensively and comprehensively described exemplary cases; cases that are rich in information, meaningful, exemplifying the practice studied, featuring practically wise participants, and suitable for clarifying other cases. They are guided by the wish to engage burning issues of the participants in a practice and by an intuition about where to look. We propose four standards for this kind of inquiry: (1) the exemplariness and extensiveness of the case descriptions analysed, (2) the rigour of the analysis, (3) the innovative power of the empirically grounded theoretical concepts that are developed, and (4) resonance validity: the approval the findings receive from participants in the practice.