INTRODUCTION

Systemic¹ couple and family therapy (CFT) and family science is a multidisciplinary field that emerged from a rich myriad of ideas inspired by cybernetics, physics, theater, film, and literature, in addition to a variety of psychotherapy approaches (Lebow, 2017). Besides social workers, psychologists, and psychiatrists, input has also come from sociologists, anthropologists, engineers, biologists, and communication scientists. We argue that in contrast with mono-disciplinary academic fields such as psychology and medicine (psychiatry), systemic CFT and family science have yet to be considered an academic discipline in their own right. This is reflected in the myriad professional backgrounds of those who identify with and practice systemic CFT. Thus, our discipline has not yet obtained a traditional “academic home”, nor does it have a comprehensive program of research with systemic or relationship-focused underpinnings. Accordingly, there is no plethora of “scientific ambassadors” to efficiently promote the legitimacy of the broad and comprehensive field of systemic CFT. Although systemic research programmes such as the scientific paradigm of family psychology (Pinsof & Lebow, 2005) have been proposed previously, our impression is that these suggestions have been heeded by only a few. Over the past couple of decades, great advances have been made in research methods, including a wide range of qualitative and quantitative research methods and mixed methods approaches (Friedlander et al., 2021). We thus find ourselves better able to examine complex relational processes than ever before. Given these advances, we believe the scientific paradigm of family psychology should be updated. In this second article of two published in Family Process on the ideological and philosophical preferences relating to and conducting research in the field of systemic CFT, we present the rationale, metaphysical assumptions, and implications for this update. Observation of this updated iteration of the paradigm would enable the application of relevant systemic research methods that coincide with the goals of a multitude of organizations that promote the interest of systemic therapy around the world (e.g., International Family Therapy Organization [IFTA], American Family Therapy Association [AAFT], European Family Therapy Association [EFTA]) (Lyness, 2020; Northey, 2009). All family therapy organizations share a commitment that emphasizes scientific and rigorous training standards, and in some countries, the job of the family therapist is an established profession, while in others efforts are ongoing to obtain recognition and statutory licensing (Borcsa et al., 2013; Józefik et al., 2013; Lee et al., 2013; Moloney, 2013; Northey, 2009; Relvas et al., 2013; Wagenaar & Baars, 2012; Welter-Enderlin, 2005). EFTA formulates these as efforts “to achieve recognition for family and systemic therapy as a distinct, scientifically-based form of psychotherapy practice and to ensure rigorous standards of training and professional practice throughout Europe” (EFTA Training Standards, Section 2.1). Achieving these EFTA standards would, firstly, imply building on the common principles, acknowledged manuals, common factors, and regulations of psychotherapy research.

Furthermore, Alan Carr (2013, p. 196) suggests that the first step in promoting family therapy as a profession is:

to develop a research infrastructure for generating and using scientific evidence to inform family therapy practice and convince public health service funders of its value. The next steps are to create a career structure and to establish a context to support the introduction of statutory registration.

Our own interest in promoting a new paradigm of systemic research is two-fold: We have different professional backgrounds (as social workers and clinical psychologists) but we have a shared conviction of the need to promote a systemic science practitioner or local clinical scientist approach to our field (Stricker & Trierweiler, 1995). Consequently, the local clinical scientist engages in the science-based pursuit of emerging data that is considered relevant to her own or her colleagues' clinical practice to support and inform clinical decisions (Lebow, 2014). On the basis of our experience of quantitative, qualitative, and mixed methods, we find it appropriate to advocate methodological multiplicity, although some of us have particular knowledge of quantitative methods (KJW and TT) while others (JS and LL) have greater competency in qualitative methods. Furthermore, the first authors (KJW and JS) have experienced the importance of self-reflexivity; the alienation they felt in their scientific inquiries as students at the MA and PhD levels was an epiphany that mobilized them to overcome the methodological restraints they faced in the CFT field. For instance, the identified need for a mixed methods approach led one of them to become self-taught in quantitative methods on a master's level program in family therapy that at that time focused mostly on qualitative methods. Constraints like this on the use of heterogeneous methodologies illustrate our basic concern for our field, which emphasizes mainly qualitative student projects. This experience also included self-reflexivity, which led to their realization of and perplexity over the apparent limited interest and know-how in terms of discovering the relational underpinnings of how change occurs in systemic CFT. Thus, motivated by our own experience and concerns, and aligning ourselves with the goals of the family therapy organizations previously mentioned, in this article, we aim to promote the idea of a new systemic research paradigm. This idea was first presented at international workshops (EFTA in 2022 and IFTA in 2023), where it was much welcomed. In this second of two articles, we will in particular promote multiplicity, including the use of quantitative methods and the application of transmethodological reflexivity, to promote ethical transparency and reflection on the positioning of the researcher.

A blueprint for achieving this—as suggested in the first of the two articles—is the principles of evidence-based practice (EBP) (American Psychological Association, 2006). Adherence to these principles would ensure that the field of systemic CFT is building on the professional and scientific ground that it has in common with the rest of the psychotherapy community. Second, adhering to international standards (IFTA/EFTA) would mean identifying, developing, and applying research approaches tailored to the greater complexity and comprehensiveness of systemic CFT. Such methodological approaches would not be relevant exclusively to our field, as they would also be applicable to other branches of psychotherapy research. For instance, as our object of interest is primarily family relationships, the recommended quantitative research approach is statistical analyses at the dyadic level rather than exclusively at the individual level (Friedlander et al., 2021). Since psychotherapy is, by definition, relational, such methods should also be of great relevance for those who examine other relationships in psychotherapy, such as the client–therapist alliance or the development of relations between members attending group therapy (Kivlighan, 2021; Kivlighan et al., 2016). However, we still lack international standards for a scientific program integrating science and practice into systemic CFT training similar to the scientist–practitioner model of graduate education in psychology (Boulder Model) (Raimy, 1950). Unless more is done within training programs in respect of standards and the development of competent scientist-practitioners, there is a risk that systemic CFT research will be conducted by other professionals with limited knowledge about the field of systemic CFT.

A systemic perspective on CFT focuses on relations between people and the contexts they interact within to a greater extent than on the individuals' intrapsychic reality (which is the norm in traditional psychotherapy research). This demonstrates that systemic research questions (Carr, 2010) and, accordingly, which methods are suitable, may differ from individual-oriented psychotherapy research questions. For this reason, we see it as crucial that systemic CFT practitioners be involved in research, partly to address relevant research questions and partly to make sure the results from such studies are understandable and useful for clinical implementation. Furthermore, in order to encourage systemic CFT practitioners to take an interest in conducting research themselves, there should be a viable educational program available for training and certification. We believe that the more tailored a scientific training program is to the systemic clinician's daily practice, the better the chances those recruited to it will become systemic-oriented researchers. Perhaps, the most accessible starting point for a clinician to enter research is through the use of a practice-oriented research design (POR) (Castonguay & Muran, 2015) that implements research as a natural part of clinical practice. Here systematic feedback, such as routine outcome monitoring (ROM) (Tilden & Wampold, 2017), is applied, in part for clinical purposes and in part to collect quantitative research data. Due to the greater comprehensiveness of studying systems (dyads, groups, families, and organizations), such a research program needs to build on multiple theoretical frameworks (e.g., family systems theory, resilience theory, attachment theory, developmental psychology) to improve our understanding of complex relational processes. Furthermore, as many human processes are intrapsychic as well as systemic, such systems need to be part of the understanding of a comprehensive systems theory in line with General Systems Theory (von Bertalanffy, 1968) and current family systems theory (FST) (Priest, 2021). As Priest (2021) argues, not only does a family systems theory require the support of research, this theory should provide the basis for hypotheses and research questions for future research. For instance, scientifically investigating how systemic processes interact will, accordingly, contribute to theory development as a crucial part of systemic research and research-practice integration. Thus, research is needed as a component in the development of the theoretical infrastructure of systemic CFT to capture a holistic ethos by recognizing intrapsychic and interpersonal systems as mutually interdependent (Priest, 2021; von Bertalanffy, 1968). However, in an individual-oriented society, it is a challenge to translate intrapsychic concepts and reveal them as relational processes, such as by conveying these phenomena in a relational language and developing validated assessments of these relational or biopsychosocial processes (e.g., from resilience in individuals to family resilience) (Walsh, 2016). Wampler and Patterson (2020) suggest that these relational processes should (a) evoke interpersonal interaction instead of intrapsychic conditions, (b) focus on the interaction between individuals in family relationships (i.e., family members broadly defined), and (c) be applicable to the family as a whole and to different subsystems, family situations, cultures, and stages of the life cycle, as well as interactions between therapists and clients. Many of the constructs used in systemic CFT are descriptive of dyadic interaction. For example, interparental conflict evokes the dyadic relationship between parents. Explicitly using the construct “interparental family conflict” could help broaden our thinking about the construct (Stokkebekk, 2022). Similarly, the “secure base” construct brings to mind the dyadic relationship between parent and child. However, the application of the construct “family as a secure base” could help expand our thinking (Wampler & Patterson, 2020). As this discussion easily addresses the schism between individually oriented professionals and relationally oriented professionals, it is our objective here to avoid taking an “either-or” position. We would rather encourage the extension of our notions traditionally considered individually oriented so that they can also be understood from a systemic perspective, and vice versa, which may mean all systems on all levels (Priest, 2021; von Bertalanffy, 1968).

With reference to the principles of EBP (American Psychological Association, 2006), the methodological approaches need to respond to the research questions and assess which methods are most suitable for answering the questions. Therefore, the researcher's initial preference for specific research designs and methodologies should not influence this process. Rather, it is important to be open-minded and have an overview of the existing methods and research designs in order to plan and optimize a research study. Often, a single approach is not sufficient and it may be necessary to use a combination of different methods at different stages of a research study in order to address different aspects of the research question (For an example of such a series of studies, the reader is referred to the author's [KJW] PhD project, which examined the processes and outcomes of patients with a history of childhood trauma who were receiving systemic CFT; Whittaker et al., 2021; Whittaker, Johnson, et al., 2022; Whittaker, Stänicke, et al., 2022). Hence, in multidimensional systemic research, there is a need for self-reflexivity, meaning that the researcher must be critically conscious of their own positioning and prejudices and of how these may influence the phenomenon being studied (Davies & Harré, 1990; Stokkebekk, 2022; Stokkebekk et al., 2021).

RESEARCH AS A PARADIGM AND A PROGRAM

A paradigm for systemic family research should be grounded in a philosophy of science that embraces the need for multiplicity to discover and explore the complexity and relational underpinnings of the phenomena of interest. By this, we mean that the research designs and methods intended for implementation in research projects are supported by the researcher's underlying assumptions about the target phenomena, such as the assumed interdependence of observations (e.g., how family members' responses are contingent upon one another) (Kenny et al., 2020) when the researcher is selecting an appropriate statistical method.

The field of systemic CFT has been full of epistemological debate about the philosophical assumptions that are relevant to research and clinical practice (Hoffman, 1985; Lorås & Sundelin, 2018). Our contribution to such debate seeks to address the need for a program of systemic research. The central focus of these paradigm debates has been on general systems theory and systemic ideas. Differences in opinion relate to whether one should view a system as an entity or a metaphor and how to apply cybernetic theory to practice and research (Wampler & McWey, 2020).

The underlying premise of general systems theory (GST) is that a system is a unified whole formed by interrelated parts, where a change in one part of the system affects the rest of the system (von Bertalanffy, 1968). An updated version of GST with particular relevance for our field is Priest's iteration of family systems theory (FST) (2021). FST encourages a focus on the relationship between the parts of a system as well as on each of the parts that comprise the system. The addition of cybernetics has provided an important companion to GST, as it endeavors to discover how systems use feedback loops. Earlier research in family therapy followed the assumptions of first-order cybernetic thinking, which perhaps reflects a (post-) positivistic research paradigm, according to which systems and information are observable. This was followed by a second-order shift in the view of cybernetics, where the emphasis was on the impossibility of conceptualizing the researcher or therapist as independent of the system subject of their study or treatment (Howe & Von Foerster, 1974). Thus, the focus was not on patterns but on systems of meaning and how meaning was constructed. This made constructivism and social construction relevant as a research paradigm that emphasized the researcher's reflexivity above all else. As Hoffman (1993) famously put it, “We do not ‘discover’ the world-out-there but, on the contrary, ‘invent’ it” (p. 390). Broadly, the shift parallels a move in the social sciences towards social constructionism in general and away from a mechanistic version of systems theory. Although the observation of patterns was still seen as an important starting point, the emphasis shifted to exploring how these patterns were shaped by meaning-making and the beliefs and stories of family members. Dallos and Urry (1999) suggest a third-order cybernetic shift, which in our opinion resembles a research paradigm from critical realism (Pocock, 2015; Tilden et al., 2022) with its critical examination of society's impact on families. Arguably, third-order cybernetics presents a “neutral view of problems” in that it acknowledges the pernicious influence of factors outside of the family's control. In effect, pathology is seen as inevitable where, for example, ethnic minorities experience racial abuse and discrimination, or poverty and deprivation are seen to strip people of their self-respect and foster a sense of helplessness. Families are viewed as microcosms that reproduce rather than cause these inequalities and oppressions (Dallos & Urry, 1999, p. 173).

In line with the third-order shift of cybernetics, Pinsof and Lebow (2005) proposed the scientific paradigm of family psychology as an alternative to the prevailing paradigm of constructivism/social constructionism, summarizing their position as follows:

There are several pragmatic implications to the interactive constructivist position and the notion of progressive, but partial or incomplete, scientific knowledge. Quantification does not necessarily imply objectivity. It is a tool to help us understand and manipulate reality, but it does not confer an inherent objectivity on the knowledge it generates. As a consequence, this paradigm incorporates both quantitative and qualitative methods. They are not incompatible. On the contrary, within this paradigm, they exist in a circular, reiterative relationship. They generate somewhat different, yet compatible types of knowledge that are complementary and mutually enriching. (p. 8)

As previously mentioned, we advocate revisiting the tenets that Pinsof and Lebow (2005) refer to as the scientific paradigm of family psychology, as a solid fundament for further development. Since their initial proposal, there have been significant developments in research methodology that are relevant to systemic CFT, such as advances in qualitative, quantitative, and mixed methodologies. Two important approaches that we will advocate for here are trans-methodological reflexivity (from qualitative methodology) and interdependence of observations (also known as nonindependence of observations in quantitative methodology) (Kenny et al., 2020). First, the applicability of mixed methods research approaches makes self-reflexivity a trans-methodological research skill that is relevant in both quantitative and qualitative approaches. Reflexivity is the process of reflecting critically on oneself as a researcher (Lincoln et al., 2018). Although reflexivity is mostly applied in qualitative methodology, it has also been proposed as a general research skill (Finlay, 1998) and referred to as vital in quantitative methodology (Jamieson et al., 2023). Furthermore, positional reflexivity is also referred to as an important principle in systemic practice in general (Lini & Bertrando, 2022; Stokkebekk et al., 2022). We argue that reflexivity forces researchers to come to terms not only with their choice of research problem (e.g., their various adherence biases) but also with those whom they engage in the research process. Positioning oneself as a systemic (and multimethodological) researcher also calls for transparency in one's choice of methodology and transparency in the choices one makes during the research process. In our view, future systemic research should, as a norm, be a multi-methodological research endeavor. Systemic research is thus referred to as a “complexity science” where “the systemic approach is a meta-theoretical or paradigmatic framework for multi-methods research” (Schiepek, 2020, p. 12).

Second, as our main objective is the study of the reciprocal world of human relationships, we find it necessary to explicitly adopt the assumption of interdependence of observations. In practical terms, interdependence (Kenny et al., 2020) implies that people who find themselves in the same circumstance, such as members of the same family seeking therapy together, will be contingent upon one another as they undergo therapeutic processes. Conversely, strangers who attend individual therapy are likely to undergo therapeutic processes independently of one another. As argued by Kenny et al. (2020), the study of interdependence of observations is the focus of any research that wants to examine two or more people who are related. To our knowledge, this approach to relational research has not been systematically applied in any systemic couple and family therapy research program, which we consider a grave oversight. Interdependence of observations needs to be understood by all researchers in the field irrespective of methodological preference and further applied statistically by those that use quantitative methods. Regarding the latter, statistical methods that allow for the testing of assumed correlations (i.e., a test of interdependence of observations) should be applied when analyzing quantitative data collected from systemic CFT settings (Kenny et al., 2020). We will illustrate this with one of the most commonly applied outcome instruments in couple therapy, the Revised Dyadic Adjustment Scale (RDAS) (Busby et al., 1995). This questionnaire is usually completed by both members of a dyad and addresses their subjective satisfaction in their relationship. Although we can assume that the degree of correlation between spouses' assessments on such an instrument is hardly controversial, the potential reciprocity of scores cannot be handled by traditional multivariate analyses (Kenny et al., 2020). Such analyses, when applied to psychotherapy, are based on assumptions rooted in what one might call an individual-oriented paradigm of psychology (i.e., the assumption of the independence of observations). Consequently, even if there were consensus on which measures of outcome should be used in couple and family research and practice, the results of such evaluations would be of limited value if founded on the assumption of the independence of observations. At best, research guided by such an assumption would apply methods to analyze outcomes that could show only how individuals change from one point to another (and usually just at two time points) in accordance with the nomothetic approach (i.e., with the objective of making general predictions about the population) (Beltz et al., 2016). Such methods cannot divulge how family systems change across time and are therefore of limited interest beyond illustrating the general efficacy of treatments.

THE PARADIGM OF SYSTEMIC FAMILY RESEARCH

An iteration of the scientific paradigm of family psychology (Pinsof & Lebow, 2005) was presented as the five pillars by Pinsof et al. (2018) in Integrative Systemic Therapy: Metaframeworks for problem solving with individuals, couples, and families. We have found this iteration of the scientific paradigm of family psychology to be suitable as an epistemological and ontological foundation for what we have dubbed the paradigm of systemic family research.

Given that the main interest of systemic researchers is the interdependence of the observed objective (e.g., the interaction between family members), this implies explicitly including this understanding regardless of the choice of methodological approaches. Accordingly, a statistical test of interdependence of observations should be applied when relevant, and/or interdependence in family systems should be explored with, for instance, attention being paid to self and other positions of the social agents in the scope of a qualitative methodological enquiry. Hence, both qualitative and quantitative data should be collected from all participants deemed to be directly involved or affected by the phenomena studied. This would ensure that both of our objectives (i.e., trans-methodological reflexivity and interdependence of observations) are attended to. We will therefore argue that these objectives require explicit emphasis in the paradigm of systemic family research.

As a further addendum to the paradigm of family psychology, we have chosen to replace the term “psychology” with “systemic.” The former term connotes that the paradigm has a proclivity toward the individual's intrapsychic reality, while the latter emphasizes the link between research and systemic concepts. This does not, however, preclude systemic research from studying individuals but rather emphasizes that it should include a systemic perspective in line with GST and FST (Priest, 2021; von Bertalanffy, 1968). We also emphasize that systemic research is equivalent to a complex enquiry requiring a multi-methodological focus. The researcher is the nexus of research in many aspects of the enquiry, where methodological awareness, applicability of relevant methodology, and self-reflexivity are crucial. We have thus coined the term trans-methodological reflexivity to refer to the constant reflexive awareness of the positionality and choices of the researcher (Finlay, 2003). Furthermore, all research methods are subject to power- and positionality-laden interpretations. Using different methods is one way of highlighting this issue and also of paving the way for an alternative position for the problematic (and un-reflexive) “claims of objectivity and neutrality made by the vast majority of researchers working with quantitative methods” (Nightingale, 2003, p. 79).

The paradigm of systemic family research proposes methods that are applicable to examining relationships of any degree of closeness, including collegial and peer relationships. However, the use of the term family is preferred in view of the topics and phenomena with which couple and family therapists are concerned. An example of such a topic is intrafamilial abuse and how it affects the close romantic relationships of survivors of such trauma as adults. By exploring such topics, hypotheses informed by systemic theories may be tested. As such, the paradigm of systemic family research may potentially confirm, disconfirm, or further develop family systems theory (Priest, 2021) as a relevant framework for understanding relational phenomena. In the following section, we shall present the assumptions underlying the paradigm of systemic family research in greater detail.

THE FIVE PILLARS

The core theories (i.e., epistemology and ontology) proposed as the five pillars by Pinsof et al. (2018 in Integrative Systemic Therapy: Metaframeworks for problem solving with individuals, couples, and families can be considered integrating FST (Priest, 2021) and perspectives of the postmodernist movement. Within this framework, the differences between competing views have been reconciled, with some of the more radical social constructionist standpoints often associated with postmodernism being moderated and FST being aligned with the values of present-day society. The five conceptual pillars expound on why certain theories or treatment strategies are prioritized and provide the foundation for the multiplicity of frameworks that make up integrative systemic therapy (Pinsof et al., 2018). We suggest extending these five metaphysical assumptions with the principles of interdependence of observations and transmethodological reflexivity to support the proposed paradigm of systemic family research.

The epistemological pillar pertains to how knowledge about reality is obtained (i.e., its epistemological stance): the recognition that an objective reality exists but is only partially accessible to any given individual (Pinsof et al., 2018). Thus, we recognize the existence of objective reality, but the world as we know it is always perceived and understood through the lens of subjectivity. Objective reality, although ultimately unknowable, may thus be understood as a continuous intersubjective process that results in cognitive approximations of the environment that allow the organism to interact with and physically manipulate it (Pinsof & Lebow, 2005). As indicated, this epistemological stance, with its emphasis on the social aspect of the knowledge of reality, has commonalities with the postmodernist perspective. However, it moderates it by acknowledging the existence of an objective reality that may be at least partially knowable. Consequently, from a purely postmodernist point of view, one would not, as a therapist, be interested in underlying psychological structures (e.g., structural theory) (Marcus, 1999). As a scientist, one would not research latent variables such as anxiety or depression; one would be interested only in the meaning and structure of the stories told by clients and informants. From the perspective of postmodernist critique, language shapes the experience that we call reality. This scope of interest thus gives preference to language as the medium of relating meaning and structuring reality; hence language has been the target of research (Priest, 2021). According to the epistemological pillar, one would not be bound by the kind of limitations that ensue from adhering to a strictly postmodernist perspective. Through the lens of the epistemological pillar, one can be interested as a therapist and a scientist in both the narratives and the underlying psychological structures without sacrificing one out of preference for the other. By extension, the argument may be made that the epistemological pillar encourages integration by acknowledging the benefits of multiple perspectives. As such, the epistemological pillar is similar to the epistemological stance underlying the paradigms of pragmatism (Maarouf, 2019) and critical realism (Patomäki & Wight, 2000), as these allow for the application of a range of methods to collect and analyze data. Therefore, research and practice based on the epistemological pillar are not an “either-or” venture; when it comes to the selection of methods of enquiry, the epistemological pillar embraces multimodality (Pinsof et al., 2018). However, we argue that the epistemological pillar (i.e., acknowledgment of reality as contingent on the interpersonal) also supports the concept of transmethodological reflexivity (Finlay, 2003), and we underscore its importance. Hence, this addresses the necessity of being transparent about the that researcher's contribution to the enquiry and the production of knowledge. We thus argue for the need of transmethodological reflexivity regarding the methodological choices of the researcher, self-reflexive accounts regarding adherence and biases to the phenomena of inquiry are vital to ensure research validity.

This perspective on research has something in common with movements in the field of psychotherapy, which is heading perceptibly forward toward integration (Norcross et al., 2013; Oddli & Kjøs, 2021) and thus beyond decades of debate over which therapy models are most efficacious. We would argue that the epistemological pillar and its related epistemological stances (e.g., pragmatism [Maarouf, 2019] and critical realism [Patomäki & Wight, 2000]) are in fact developing into the cornerstone of all contemporary relational and systemic therapies. Consequently, as regards scientific endeavors, we would propose that the perspectives described suggest that no single theoretical framework can capture all the nuances of relational systems. By extension, we suggest that mixed-method designs, although contingent upon the stated research question, are more advantageous than single-method designs. Besides multiple methods being implemented within one study (i.e., merged), methods may also be implemented sequentially (e.g., a quantitative study building on a qualitative study). For instance, some research projects are based on an initial explorative design (e.g., action research and ethnographic or qualitative enquiry) that may lead to the generation of hypotheses that can be tested statistically according to an explanatory design (e.g., multilevel modeling; Bolger & Laurenceau, 2013). Conversely, other research projects are often based on an initial explanatory design (e.g., statistical hypothesis-testing design) that can result in the need for further explorative enquiry into the target phenomena (e.g., deviant case analysis; Seawright & Gerring, 2008). In systemic research, the main phenomena of interest (i.e., the intersubjective) are often of such complexity that several methods (i.e., several perspectives) deployed in tandem are better suited to examining such intricacies (i.e., multiple subjective perspectives).

The ontological pillar of IST connects the paradigm of systemic family research to GST (von Bertalanffy, 1950, 1968) and cybernetics (Wiener & von Neumann, 1949), which serve to describe the interaction between people. Priest's (2021) empirical documentation of the research supporting these theories, as suggested in his iteration of FST, supports the rationale of the ontological pillar. From this perspective, every aspect of human life, whether relating to the individual or the family or the society as a whole, may be conceptualized as a system made up of subsystems. More than one system or subsystem may have an influence on another system or subsystem. “The whole is always more than the sum of its parts,” an axiom sometimes credited to Aristotle, rings true when describing this ontological stance. Systems are organized in such a manner that boundaries and power matter, and behavior may be understood only within the context of systems. Feedback from systems (e.g., ROM [Tilden & Wampold, 2017] and verbal and/or nonverbal communication) provides the information needed to understand the relationship between subsystems (e.g., a family and its constituent members). Within systems, a driving force toward morphogenesis (i.e., change) and a pull toward homeostasis may both exist. The ontological pillar orients research conducted within the paradigm of systemic family research towards the examination of such driving forces: how people who are tightly knitted together influence one another within their social environments (e.g., the influence of parents' relationship satisfaction on their child's externalizing behavior). However, it is not only family members who influence each other; the researchers also interact with and are part of the system that is being observed. Hence, this calls for paying reflexive attention to how we as researchers are influenced by and influence the family system we study. Scotland (2012, in Bager-Charleson & McBeath, 2020) makes a key point when he states that “it is impossible to engage in any form of research without committing (often implicitly) to ontological and epistemological positions” (p. 10). For instance, if you assume a critical realism position, which we as authors encourage, this involves a “stratified ontology” that allows both quantitative and qualitative research approaches to co-exist. However, the ontology of critical realism also involves a critical reflexive viewpoint and ethical responsiveness to people and their containment of resources from their social reality (Price & Martin, 2018). The self-positioning of the researcher as “a neutral observer” is always in danger of othering people with labels and concepts that alienate them from restraints inherent in the context of their social reality. This underscores the need for self-reflexivity and a reflexive account of the viewpoints and labels used and their applicability to capturing people in the context of their social reality.

The sequence pillar breaks down the relationship between subsystems and systems into sequences and thus situates phenomena of interest (e.g., patterns of reciprocity) within the dimension of time (Pinsof et al., 2018). By defining behavior as sequences, we can identify and target recursive patterns of thought, emotion, and behavior. The phenomena we seek to understand when we engage in research are embedded in such sequences. For example, when we attempt to explore how mental distress influences couple satisfaction, a possible line of enquiry is how depressive thoughts, emotions, and behaviors may be embedded in a couple's interactions with each other (i.e., sequences). Furthermore, we may be motivated to study how therapeutic interventions undertake to change or modify such sequences so as to ameliorate both individual and relational distress.

The constraint pillar is informed by Bateson's (1972) concept of cybernetic explanation, which proposes that an occurrence is the result of the cessation of a constraint (e.g., the influence of event A constrains the occurrence of event B; Pinsof et al., 2018). This contrasts with casual explanations that propose that event B occurs as a consequence of event A occurring. As regards therapy, constraints may be conceptualized as “something” that prevents a client system (e.g., a family) from solving the problem it presents in therapy (i.e., a constraint to change). The constraint pillar prioritizes cybernetic explanations over causal explanations because they promote constraint questions (e.g., what constraints this couple from having a satisfying marriage?). Constraint questions inherently facilitate circular and systemic understandings of change (e.g., the couple's marriage satisfaction is constrained by the negative aftereffects of past trauma).

This theory of change is also known as the theory of constraints (Breunlin, 1999). The constraint pillar informs the hypothesizing process innate to both therapy and research with regard to what the constraints are composed of and where such constraints are located (e.g., within the individual, within the couple dyad, the parent–child dyad or triad, or at the level of the local community and/or society, etc.). Constraints may manifest anywhere within a system and can range from a genetic disorder constraining an individual's capacity to break down essential amino acids and resulting in severely impeded cognitive development (the individual level) to an undisclosed affair creating emotional distance in a couple's relationship (the dyad level) or the detrimental effect on local communities of the lack of affordable healthcare (the societal level). In all of these examples, one may postulate that removing or altering the specific constraint would either directly solve the presented problem or allow for the implementation of strategies that target it (e.g., restricting phenylalanine in the diet, disclosing the affair, and changing government policy).

The constraint pillar is of immense value when designing a research project, particularly if one seeks to identify particularities of families, such as the factors associated with a lack of response to therapy, as it prompts the formulation of research questions as cybernetic explanations. The constraint pillar also directly promotes engagement with informants to identify constraints. This intentional stance, in conjunction with the epistemological and ontological pillars, means that research is a collaborative project between researcher and informant. This may be most evident in qualitative methods; however, it is also relevant when quantitative methods are used. This is especially true in what has been called practice-oriented research (POR; Castonguay & Muran, 2015) when ROM (Tilden & Wampold, 2017) is being applied. In POR, the roles of researcher and clinician often overlap and interactions between researcher and client/informant are frequent. Subsequently, research guided by the theory of constraint (Breunlin, 1999) may have practical implications: it may facilitate the development of strategies to improve the quality of life for people whose responses to therapy and/or community interventions have been less than optimal by identifying constraints to therapy and thereby improving outcomes.

The causality pillar emphasizes the recursive influence that subsystems and systems have on each other (i.e., their assumed interdependence). Basically, systems theory builds on the notion of systemic, mutual influences rather than linear causality. For this reason, systemic research represents a departure from positivistic research, which often reduces and isolates the phenomena of interest to enable the analysis of one-way predictive or causal associations. The systemic notion that “every part of a system has a mutually influence on every associated part of the system” could nevertheless be considered a simplification. From a research point of view, such a theoretical statement can be considered an empirical question addressing the question “Is it necessarily so?” An alternative theoretical understanding of causality supported by the ontological pillar claims that subsystems and systems do not necessarily all have an equal influence on other subsystems and systems but may rather show systems and subsystems as having an unequal influence on each other (i.e., power matters; Minuchin, 1982; Pinsof & Lebow, 2005). Within our field, systemic CTF, such differential causality (Pinsof, 1995) has long been rightly assumed but only within the last decade has it become the target of systematic enquiry (Friedlander et al., 2021; Whittaker, Johnson, et al., 2022). The causality pillar reminds us that we are interested in investigating whether and how subsystems and systems influence one another. By founding our research on the causality pillar and thereby, by extension, assessing for the assumed interdependence of observations (i.e., a test of nonindependence in quantitative methodology; Kenny et al., 2020), we are in a position to empirically test hypotheses generated by family systems theory (Priest, 2021) relating to how systems change across time. Similarly, one could supplement quantitative tests of non-independence with in-depth interviews to method triangulate (Fielding, 2012) and hence explore links between statistical associations (e.g., actor and partner effects; Kenny et al., 2020) and meaning-making from different subject positions on relational processes (e.g., self/another positioning, positioning theory; Stokkebekk, 2022; Stokkebekk et al., 2019, 2021, 2022).

IMPLICATIONS FOR SYSTEMIC RESEARCH

Part 1 and Part 2 of “Researching What We Practice” have addressed several unique aspects of research in the field of systemic family therapy. We will now discuss their implications for our future research efforts.

As suggested in Part 1, systemic practitioners and researchers should adhere to the principles of EPB (American Psychological Association, 2006). This could bridge the scientist–practitioner gap by improving the dialog between clinicians and researchers. Such an improvement in dialogue could be facilitated by incorporating frequent client assessments via ROM, thus establishing a culture of POR. Furthermore, participation in research should also be incorporated into student training so that clinical work and research are perceived as inseparable, thereby optimizing the translation of research findings into clinical practice. Consequently, systemic research will join the unified standards of psychotherapy research. Additionally, in the first article, we addressed specific challenges in the systemic field that relate to voices claiming preference for specific research methodologies (e.g., social constructionism) that eschew other schools of thought. Our response is partly to acknowledge that such differences in ideology and the philosophy of science represent the diversity in our field, but the challenge remains to deal with such diversity without creating further polarization in the field, such as between those preferring qualitative or quantitative research designs to the exclusion of others. In this regard, we suggest that the application of the principles of EBP could ensure equal acknowledgment of all research methodologies and sources of knowledge. EBP further implies adherence to established methodological standards common to the entire psychotherapy field, such as the choice of research design being dependent on what can best answer the research question rather than the researcher's ideological preference. EBP can thus include what is considered unique in systemic research. Accordingly, adhering to aspects of EBP such as common guidelines for systemic practitioners and researchers would also help bridge the professional gaps in our field.

Given our critique that some prominent systemic researchers and therapists appear to be selective in their choice of design and source of knowledge, we advocate greater awareness of underlying assumptions and how these relate to the selection of research methods. For this reason, we find it timely to suggest a program for systemic research that involves guidance and training for systemic practitioners and researchers. Such a program should build upon the valuable contributions of Pinsof and Lebow's (2005) Scientific Paradigm of Family Psychology and Pinsof et al.'s (2018) Integrative Systemic Therapy: Metaframeworks for problem solving with individuals, couples, and families. Predicated on these works we conclude that the five pillars are the preferred philosophy of science. On the basis of our experience as clinicians and researchers, we find this integrative systemic approach very useful as it promotes the applicability of the five pillars as a philosophy of science and serves to integrate competing views and methods into a coherent scientific approach to CFT research. In addition to the impressive contributions it makes, we suggest that it is also necessary for systemic researchers to show (a) transmethodological reflexivity and (b) ways of capturing interdependence of observations as unique phenomena of interest in systemic therapy.

CONCLUDING COMMENTS

In summary, our two articles address central tenets in the ontology and epistemology of systemic family therapy that affects both clinical practice and research. We are motivated by the concern that parts of the systemic field relate to only certain ideologically preferred elements of research design and knowledge sources. We consider this perspective and practice as unbalanced and deviating from the principles of EBT that should guide psychotherapy practice and research in general. We believe that such deviations could harm the entire field of systemic family therapy. For this reason, we suggest that principles of EBP should be implemented as an overarching compass, regardless of the preferred paradigm. Furthermore, we suggest a paradigm and programme of systemic research built upon the contributions of Pinsof and Lebow (2005) and Pinsof et al. (2018) and supplemented with the unique methodology for capturing the systemic objectives of interdependence of observations (Kenny et al., 2020) and our own suggested approach of transmethodological reflexivity. Our proposed paradigm and program are suitable for use as a theoretical foundation and as a practical guide for conducting systemic family therapy research. They build upon decades of achievement in the field and are up to date and in line with overall scientific advances, whether one prefers a qualitative or quantitative approach or a mixture of these. This inclusiveness with regard to methods and ideas can incentivize collaboration across ideological boundaries and traditions and can thus counteract the existing dichotomy in our field through the acknowledgment of differences under a single umbrella of commonly acknowledged and accepted principles.