Practices in State Self-Esteem Research: An Analysis of Enacted Ontologies

ABSTRACT

Methodological and empirical questions concerning state self-esteem are contingent upon very specific underlying commitments to “what” state self-esteem and its dynamics actually are. These are questions concerning ontology. These underlying commitments or views about “what actually exists” are not explicit, but enacted through our research actions. It is vital to bring these implicit underlying ontologies to the surface, so that we as researchers can reflect upon them, and on the assumptions that we are communicating and reinforcing with our methodological and empirical practices. In service of a conceptually solid and unambiguous framework of theoretical and methodological approaches to state self-esteem, I aim to lay bare the ontological commitments enacted in current research on state self-esteem. I show that state-self-esteem research forms two different assemblages of practices, which are repertoires of conceptual assumptions, discourse norms, methods of analysis, and operationalizations. One assemblage sketches a narrative of daily self-esteem in mechanistic terms, the other sketches a narrative of daily self-esteem in processual terms. After analyzing how concrete practices enact these ontological commitments, I reflect on how the two research assemblages might converge to benefit research on state self-esteem in the future, emphasizing the need for reflexivity from researchers.

KEYWORDS:

• Daily self-esteem
• dynamic self-esteem
• Self-esteem stability
• ontology
• epistemology

While absolute levels of self-esteem have been the topic of much investigation in modern psychology, it is becoming increasingly pertinent to study the daily fluctuations in self-esteem that occur within individuals, across time and contexts (Alessandri et al., 2016). The stable property of self-esteem (i.e., the absolute level) is commonly referred to as trait self-esteem, whereas the fluctuating property (i.e., daily fluctuations) is referred to as state self-esteem (e.g., Alessandri et al., 2013; De Ruiter et al., 2017; Donnellan et al., 2012; Hank & Baltes-Götz, 2019; Hutteman et al., 2014). The most common definition of state self-esteem, or daily self-esteem, is of “fluctuations in momentary, contextually based self-esteem” (Kernis et al., 1993). These fluctuations are thought to be important, largely because they reveal that self-esteem is quite social, such that individuals demonstrate a sensitivity, and responsivity, to their social surroundings with respect to their self-esteem (Kernis et al., 1993; Leary & Downs, 1995). Furthermore, studies reveal crucial between-individual differences in the variability of self-esteem (Greenier et al., 1999; Kernis, 2005). This research has revealed the importance of understanding the role of state self-esteem fluctuations and the relative stability (or lack thereof) of these fluctuations.

Much of the recent interest in state self-esteem fluctuations (and its stability) has focused on how we might optimally measure and analyze self-esteem fluctuations (Alessandri et al., 2013; Hank & Baltes-Götz, 2019), and which factors predict, and are predicted by, certain characteristics of stability. Regarding the latter, studies in this field focus on the correlates of unstable self-esteem, showing for example, that individuals high in self-esteem variability demonstrate more adverse properties, such as self-consciousness and social anxiety (Oosterwegel et al., 2001), antagonistic narcissism (Geukes et al., 2017) or low levels of emotional stability, agreeableness, and conscientiousness (Zeigler-Hill et al., 2015).

What remains implicit and underappreciated in this line of promising research, however, is a discussion concerning the conceptualization of state self-esteem. To be sure, empirical studies concerning daily self-esteem work from shared definitions (e.g., of “state self-esteem,” “fragile” self-esteem, or “secure” self-esteem; Greenier et al., 1999; Kernis, 2005; Zeigler-Hill, 2011). Definitions, however, can only partially capture the meaning of a concept. For a clear conceptualization, assumptions about the concept’s ontology must also be clarified – the very nature of “what” something is (Morawski, 2022). While there is an exciting surge of empirical work aimed at improving the measurement of state self-esteem and its predictive relationships, there is a striking lack of theoretical work that focuses on clarifying the ontology of state self-esteem and its (in)stability is. The apparent underrepresentation of such theoretical work is not unique to research on state self-esteem (e.g., Hibberd, 2021). In psychology more generally, discussions about the nature of psychology’s objects have been described as “low-level, not mainstage” (Morawski, 2022, p. 168).

Outlining a clearer conceptualization of state self-esteem and its dynamics in terms of underlying philosophical assumptions related to ontology is crucial. First, a coherent scientific investigation requires a clear account of what we are investigating (Maul, 2013). Second, while researchers’ conceptualizations inform their empirical pursuits of a concept (i.e., operationalizations and relevant questions to ask), their empirical pursuits also shape the reality of what a construct actually is and the lived experience of the construct (Hacking, 1995). As such, it is vital that we – as researchers – are cognizant of what we are constructing so that we can reflect on its value for academia and society (Hibberd, 2021).

To address the above knowledge gap, the aim of the current manuscript is to “work out” the different ontologies that researchers assume for state self-esteem and daily self-esteem dynamics. As described above, however, these assumed ontologies are not made explicit by researchers in the field of state self-esteem research. Researchers do, however, communicate particular ontological assumptions via their methodological and empirical practices, as well as their linguistic norms (Levy, 2019). These philosophical commitments to specific ontologies – or “what actually exists” – are thus enacted through our research actions (Derksen & Morawski, 2022). The implied ontological commitments can be brought to the fore via a discursive approach (Edwards, 2004). In particular, the discursive approach that is adopted in this manuscript is the analysis of empirical and linguistic norms in texts (i.e., manuscripts). These are treated as the object of study in themselves, with the aim of discovering repertoires used by researchers. The empirical and linguistic norms adopted in manuscripts about state self-esteem and self-esteem dynamics will be analyzed with an eye for what they do, and more specifically, the commitments to certain ontologies that they communicate. This manuscript builds on and extends the conceptual analyses in chapters four and five of Van Geert and De Ruiter (2022).

Meta-Theoretical frameworks as conceptualization tools: Cartesian-Split mechanistic framework and process-relational framework

To guide the conceptual analysis of underlying philosophical commitments, Overton’s meta-theoretical frameworks entitled Cartesian-split-mechanistic framework of phenomena, and Process-relational framework of phenomena (Overton, 2013a; Overton, 2015) will be used. These frameworks describe the main paradigms – or worldviews – adopted in developmental science more broadly. These frameworks are thus not specific to the study of self-esteem, but are “meta” theoretical in the sense that they can be used to analyze and expose the “epistemological principles (i.e., issues pertaining to the sources and justification of knowledge) and ontological principles (i.e., issues pertaining to fundamental categories of reality)” adopted in a wide array of topics and fields (Overton, 2013b, p. 26).

These worldview frameworks provide a fruitful way of exposing researchers’ worldviews. Previous work has used these two frameworks to examine worldviews implicit in explanatory efforts in developmental science (Witherington & Heying, 2015) and in methodological practices in psychology (Raeff, 2017). Analyses such as these show the ways in which these meta-theories (functioning as paradigms) “ground, constrain, and sustain theoretical and observational concepts” of living systems in general (Overton, 2015, p. 13). In the following sections, I will show the concrete ways in which analytical, methodological, and discourse norms in studies of state self-esteem research fall into these two meta-theoretical frameworks.

Unpacking assemblages of research practices: independent-entities assemblage and coactive-processes assemblage

The sets of practices observed in studies of state self-esteem research form assemblages (i.e., Assemblage Theory; DeLanda, 2016). These are networks of elements (practices, and necessarily, the researchers that do these practices) that are connected, and that demonstrate relative overlap (relative to another assemblage). The assemblages that are analyzed in the current manuscript are the practices of state self-esteem research with regards to how they align with the tenets of the Cartesian-split-mechanistic framework and the Process-relational framework.

Research assemblages are fluid and dynamic, such that there is continuous activity of, and between, elements acting upon each other (the emphasis on activity makes the notion of assemblages similar to Latour’s conceptualization of actor networks; Latour, 1987). Using the concept of “assemblage” is important in this manuscript, because it highlights, firstly, that the identified empirical and linguistic norms (and the researchers that adopt them) are not fixed or static sets of practices or groups of individuals. Secondly, it highlights that – through the actions and interactions between the various elements within the assemblage – the assemblage is constructing something larger than the sum of its parts: the global narrative about what it is that we are studying. Here, I show that two different narratives for the ontology of daily self-esteem are constructed.

In Sections 1 and 2, I describe and the two assemblages of state self-esteem research identified in the current manuscript based on four principles from Overton’s frameworks. In applying these principles, I focus on how they relate to discourse, theories, and methodology adopted by state self-esteem researchers in relation to pivotal concepts for daily self-esteem and their relation to each other: “state self-esteem,” “trait self-esteem,” and “context.”

Section 1 describes what I will call the Independent-Entities assemblage. In short, this assemblage is characterized by the principles of (1) nature as substance (or entities), (2) split understanding, (3) efficient causal explanation, and (4) objectivism (or realism). Together, these principles sketch a narrative of state self-esteem, trait self-esteem, and context, as (ontologically) separate variables that are independent of each other, where trait self-esteem is a core entity that is both stable and universal. In Section 2, I describe what I will call the Coactive-Processes assemblage. This assemblage is characterized by the principles of (1) nature as process, (2) relational understanding (or holism), (3) circular causality, and (4) constructivism. These principles sketch a narrative of state self-esteem, trait self-esteem, and context as intrinsically and continuously interconnected and coactive, and as fundamentally processual by nature.

The elements of these assemblages (as analyzed in Section 1 and 2 below) are summarized in Table 1, which will be further unpacked in Section 1 and 2. Finally, in Section 3, I discuss the work that both assemblages need to do in order to achieve further conceptual clarity. I also discuss the importance of being aware of how these two assemblages function, and the ways in which they might be encouraged to interact in order to strengthen research concerning state self-esteem in the future.

Table 1. Two assemblages of research on state self-esteem.

Independent-Entities assemblage	Coactive-Processes assemblage
Meta-theory: Cartesian-split-mechanistic framework	Meta-theory: Process-relational framework
Principle 1: Nature as substance (entities)	Principle 1: Nature as processes
Research practices
Discourse: State SE as “ephemeral,” “transitory,” “no carry over,” “independent from each other”; Trait SE as “enduring,” “core,” “essence,” “stable,” and “consistent”; “having SE”	Discourse: State SE as “activity based”; “doing”/’managing’/’constructing’/presenting’ SE; “situated”; demonstrating “intrinsic dynamics”; Trait SE as robust process characterized by “patterns,” “self-structure,” “attractor states”; demonstrating “self-preservation”/’self-maintenance’/ “adaptability” and “activity”
Theories: Sociometer theory	Theories: Constructivism; Dynamic systems
Methodology: Latent State Trait models; Growth Curve models; Standard deviations as indicators of stability levels	Methodology: Discursive analysis; High frequency time-series analysis: temporal structure of state SE
Principle 2: Split understanding	Principle 2: Relational understanding (holism)
Research practices
Discourse: “Isolate” and “quantify” components	Discourse: State SE and trait SE as “levels” of SE as whole; State SE as “enaction” of trait SE; “reciprocal and ongoing relationships”
Theories: Sociometer theory; Self-broadcasting perspective	Theory: Enactment, Dynamic systems
Methodology: Non-concurrent assessments for state SE, trait SE, context; Variable-oriented approach; Inferential statistics; Partitioning variance; Context as primarily proximal	Methodology: SE assessed in situ; Concurrent assessments of SE, trait SE, context; State SE and trait SE as distinct process characteristics of the same process’; Context as both proximal and distal;
Principle 3: Efficient causal explanation	Principle 3: Circular causality
Research practices
Methodology: Context as antecedent; Either/or predictive relationships between state SE and context; Cause-effect relations as temporally distinct	Methodology: Descriptions of constructive processes and constraint; emergence, constraint; “Description” of relational processes
Principle 4: Objectivism (realism)	Principle 4: Constructivism
Research practices
Methodology: Standardized self-report assessments of SE; Focus on “measurement”	Methodology: Various operationalizations of SE: self-report, spoken narrative, behavioral; Focus on “description” of SE

Note: A summary of ways in which concrete research practices in studies of state self-esteem communicate key (ontological and epistemological) principles from the meta-theoretical frameworks Cartesian-split mechanistic framework and Process-relational framework (Overton, 1991, 2013a, 2015). Note that this summary is not intended as an exhaustive list of relevant practices. Rather the summary entails those practices analyzed in the current paper that are illustrative of the two meta-theoretical frameworks. In the table, self-esteem is referred to as “SE.”

Section 1: Independent-Entities assemblage

The Independent-Entities assemblage of studies of state self-esteem research is that which demonstrates key characteristics of the meta-theoretical Cartesian-split-mechanistic framework (see, Table 1). This framework is characterized by a number of principles (see, Overton, 2015), but for simplicity’s sake, the current manuscript will outline how four of the most pivotal principles can be observed in certain practices of studies on state self-esteem: “nature as substance,” “split understanding,” “efficient causal explanation,” and “objectivism.” These form the sub-sections below. These principles demonstrate elements of a certain narrative about what psychological phenomena are and how we can understand and explain them. Each principle stresses a certain characteristic of this narrative. Because these principles, together, sketch a coherent picture of ontology, they will necessarily demonstrate some overlap in their description (Overton, 2015).

The Cartesian-split mechanistic meta-theoretical framework functions as a dominant paradigm in developmental science (e.g., “inheritance, evolution, and organismic – prenatal, cognitive, emotional, motivational, and sociocultural – development”; Overton, 2013b, p. 56) and in psychology (Franz, 2022; Raeff, 2019). The dominance of this meta-theoretical framework is similarly evident in state self-esteem research, as I will show below.

Nature as substance

The dominant theoretical framework for studies on state self-esteem is arguably the Sociometer Theory (Leary et al., 1995). This theory suggests that state self-esteem can be viewed as a “barometer” that fluctuates around one’s self-esteem baseline as a function of current contextual cues. With this, state self-esteem is thus seen as a temporary event, where self-esteem momentarily deviates from an underlying true level of self-esteem, i.e., trait self-esteem (Kernis et al., 1993; Rosenberg, 1986). This fleeting event is commonly assumed to be the result of a social cue, such as inclusion or exclusion (Leary, 1999). Accordingly, the assumption made here is that, without the presence of contextual events, state self-esteem would be equal to the baseline level of trait self-esteem (se also De Ruiter et al., 2015 for a similar analysis).

The above theoretical approach to state and trait self-esteem reflects a substance metaphysical principle that is characteristic of the Cartesian-split-mechanistic framework (see also, Van Geert & De Ruiter, 2022). The word “substance” is derived from the Latin verb substare (to under-stand), or to underlie. From this, a substance metaphysical principle assumes that, first, phenomena can be best understood as stable, permanent, and enduring substances (either in the sense of an “entity,” or “stuff”; Bickhard, 2009). Second, this “entity” or “stuff” is assumed to underlie (or under-stand) fleeting and ephemeral events (i.e. state self-esteem fluctuations).

Aside from the Sociometer Theory, the substance metaphysical principle is explicitly communicated in the statistical approach to state and trait self-esteem within the Independent-Entities assemblage. For example, in their use of the Latent State Trait statistical model, some researchers approach trait self-esteem as “an enduring individual difference” (Donnellan et al., 2012), or a “stable core” that “represent[s] the essence of how people evaluate themselves” (Wagner et al., 2016, p. 531). This stable core is described as “completely stable” and “completely consistent across measurement occasions” (Alessandri et al., 2013, p. 770). These descriptions of trait self-esteem, as “enduring,” “core,” “essence,” “stable,” and “consistent” communicate the first part of the substance metaphysical principle, namely, the existence of an underlying stable phenomenon in the form of trait self-esteem.

Additional self-esteem attributes that are often given “entity”-like properties are “fragile” self-esteem (i.e., unstable and conditional) or “secure” self-esteem (stable and unconditional; Greenier et al., 1999; Kernis, 2005). These attributes of self-esteem are, like trait self-esteem, often characterized as stable between-individual differences, which are approached in terms of an “essential nature” that individuals “possess” (Kernis et al., 1993, pp. 1190–1191). As such, these attributes are treated like entities that underlie (i.e., sub-stare) individuals’ responsivity to changing contexts. This substance metaphysical principle can be further observed when fragile or secure self-esteem are operationalized as stable levels of self-esteem stability, measured with standard deviations of total scores across repeated assessments of state self-esteem (Kernis et al., 1993). Here, a large standard deviation is used to indicate fragile self-esteem, and a small standard deviation is used to indicate secure self-esteem. This practice treats fragile and secure self-esteem as stable and consistent entities, as these researchers are interested in stable levels of within-individual variability that are assumed to endure and remain stable across time and contexts.

In addition to the above assumptions communicated about underlying fixed entities of self-esteem, the second part of the substance metaphysical principle is that an underlying and stable phenomenon (here, trait self-esteem, fragile self-esteem, or stable self-esteem) persists through something. In the Independent-Entities assemblage, the latter is state self-esteem: Trait self-esteem endures through the flux of within-individual state self-esteem variations. Here, state self-esteem is described as being “ephemeral” and “transitory” (Alessandri et al., 2013, p. 770). These descriptions suggest that state self-esteem is not, in itself, part of an iterative process across time (see also, De Ruiter et al. 2015) for a similar argument). This is illustrated by descriptions of state self-esteem as “independent of previous self-esteem–related perceptions” (Wagner et al., 2016, p. 531); or similarly that state self-esteem “does not carry over from one occasion to the next” (Alessandri et al., 2013, p. 771).

The flux of daily self-esteem is therefore assumed to be equal to the deviations from a baseline (where the baseline is the trait self-esteem level, in line with the Sociometer theory), where these deviations are “independent” from each other. An exception to this may be Hutteman et al. (2015), who suggest that “the environment evokes changes in trait self-esteem via consistent changes in state self-esteem” (p. 769). With this, the authors bring attention to the bottom-up causal role that state self-esteem may have over time (while not necessarily describing state self-esteem events in terms of an iterative process).

Split understanding

The Cartesian-split mechanistic framework describes split understanding (or splitting) as the principle that the whole can be split into independent parts, and that it is possible to isolate these parts from each other (Overton, 2013a, 2015) – a principle that stems from Descartes (Overton, 2015; Westerman & Steen, 2007). This splitting occurs between body and mind, between what is thought to be the foundation or core of a phenomenon and what we can observe of the phenomenon, and between a person and their context.

The assumption that parts of a whole are fundamentally independent from each other is widely communicated in psychological research more globally by the use of inferential statistics, where “partitioning the variance and discerning main effects are based on the prior theoretical assumption that one is dealing with phenomena that function separately and contribute independently to some aspect of behavior” (Raeff, 2019, p. 318). From this, the assumption within the Cartesian-split-mechanistic framework is that one can reassemble and understand the whole in additive terms, and that an understanding of the whole will be reached when we understand all of the parts.

One way in which splitting is assumed in the Independent-Entities assemblage is in the approach to trait self-esteem and state self-esteem, which builds on the “nature as substance” principle described above. Splitting assumes that trait self-esteem exists independently from state self-esteem, as trait self-esteem persists and endures through the fleeting experiences of state self-esteem (described above). Evidence of this principle is the popular Latent State Trait models adopted, which illustrate the notion that these two can be isolated from each other (for example, Alessandri et al., 2013; Donnellan et al., 2012; Wagner et al., 2016). These models “isolate and quantify the completely stable components of constructs like self-esteem from the more ephemeral components”(Alessandri et al., 2013, p. 770). These models thus illustrate the Cartesian-split-mechanistic tendency of isolating phenomena (trait self-esteem from state self-esteem) and examining their workings in an additive fashion, as if they can be split from each other.

The second, and perhaps most characteristic, form of splitting that occurs in the Independent-Entities assemblage is that of splitting between state self-esteem and the context. Experiences of the self in the moment (i.e. state self-esteem) and experiences of a certain context are extracted and partitioned from each other in order to map their association. This occurs in relation to contextual experiences such as relationship quality (Mund et al., 2015), perceived (lack of) social approval (Geukes et al., 2017), perceived social inclusion (Hutteman et al., 2015), daily events (in domains such as academics and relationships; Greenier et al., 1999; Hayes et al., 2004) and perceived belongingness (Mazereel et al., 2021).

The statistical approaches adopted in these studies approach contextual experiences primarily as antecedents to certain state self-esteem consequences. With this, the cause (context) and effect (state self-esteem changes) are viewed as ontologically isolable, where one exists outside of the other. It is thus in the empirical focus of these studies that splitting between context and state self-esteem can often be seen in this assemblage. As I will describe in the following section, this form of causal explanation is itself another (highly interconnected) principle of the Cartesian-split-mechanistic framework.

Another way that splitting can be observed in the Independent-Entities assemblage is in the procedures for measuring state self-esteem and contextual experiences. Most studies in this assemblage (for example, those listed above) do not synchronize assessments of state self-esteem with the relevant (social) context or experience. This suggests that, while the authors obviously expect there to be a causal association between the two, state self-esteem and social experiences themselves are not necessarily assumed to be inherently intertwined experiences that must be investigated as such. They are treated as if they are separate, split, experiences. Exceptions to this are those that use Experience Sampling Methods to study the real-time experiences of state self-esteem and the (social or intra-personal) context concurrently, for example, Clasen et al. (2015), Crowe et al. (2019), Hank and Baltes-Götz (2019), and Mazereel et al. (2021).

Finally, splitting can be observed in the Independent-Entities assemblage in studies that isolate state self-esteem from other intra-personal phenomena, such as one’s own body (its condition, attractiveness, and strength; (Minelli et al., 2003)), mood and affect (Oosterwegel et al., 2001), suicidality and tiredness (Crowe et al., 2019), physical activity (Mazereel et al., 2021), and personality (Zeigler-Hill et al., 2015). Here, intra-personal elements are examined as an internal context. The same additive approach that is described above – with regards to state self-esteem and external contexts, occurs in these studies.

Efficient causal explanation

Another pivotal tenet of the Cartesian-split-mechanistic framework that is of relevance here is that of deterministic causality via efficient causation – which is that of an antecedent condition or event producing a consequence (Overton, 2013a, 2015; Witherington & Heying, 2015). From this principle, it follows that dynamics or movement of an object are the result of a “push” from one or multiple independent parts (Overton, 2013).

Common antecedent-consequent relations of interest in studies of state self-esteem are “whether self-views either affect or are affected by others” views of an individual’ (Srivastava & Beer, 2005, p. 967). Here, the antecedent-consequence relations correspond with the Sociometer Theory – where social experiences are thought to influence subsequent changes in state self-esteem (Leary et al., 1995), and the self-broadcasting perspective – where broadcasting one’s self-evaluations is thought to subsequently influence social experiences (Srivastava & Beer, 2005). Both of these theories, and any attempts to investigate the underlying cause-effect predictions, demonstrates efficient causal explanation. What is yet more characteristic of the Cartesian-split-mechanistic framework, is when an either/or approach is taken when investigating these causal relationships, such as communicated in the quote above by Srivastava and Beer (2005). The question is asked “whether” self-views later affect others’ views or vice versa, thus communicating that they cannot both be true simultaneously.

Some researchers have attempted to go past an either/or approach, and have investigated the transactional antecedent-consequent relations between self-esteem and context (e.g., Hutteman et al., 2015; Mund et al., 2015). Hutteman et al. (2015) suggest, for example, that “in line with a dynamic-transactional point of view on personality development, which suggests reciprocal transactions between individuals and their environment, both processes (i.e., sociometer and self-broadcasting) can go hand in hand and mutually influence each other” (p. 769). With the assumption that bi-directional causal relationships occur, these studies move more to the periphery of the Cartesian-split-mechanistic framework (and closer to the Relational-process framework, which I will describe in Section 2). Nevertheless, these Growth Curve Models demonstrate efforts to map the transactional cause-and-effect relations between state self-esteem and context as separate effects in time, which therefore still shows traces of the Cartesian-split-mechanistic framework. This is because these efforts split these two causal relations apart and study them as temporally distinct causes, taking place one after another rather than as co-acting processes. As described in the previous sub-section, splitting such causal relationships is a key tenet of the Cartesian-split-mechanistic framework.

Objectivism

Finally, a central principle of the Cartesian-split-mechanistic worldview is that of objectivism (or epistemological realism; Holden & Lynch, 2004), where the nature of phenomena is assumed to be fixed and universal, existing independently of our observing it (Overton, 1991). Evidence of this principle lies in the practice of assessing state self-esteem with the standardized state self-esteem self-report method. This is a practice that can be observed in all studies within the Independent-Entities assemblage, but which is not sufficient on its own to characterize a study as belonging to this assemblage (this will be described further in the section Coactive-Processes assemblage).

The standard procedure for assessing state self-esteem within this assemblage is the use of a modified version of a traditional self-esteem scale, where individuals are asked to give the response that best reflects how they feel at the moment they complete the measure (e.g., Alessandri et al., 2013; Hank & Baltes-Götz, 2019). When studies on state self-esteem utilize these standardized scales for state self-esteem, the implied assumption is that state self-esteem is experienced as a universal, that is, in the same qualitative way across individuals. Here, between-individual differences in state self-esteem are limited to the level of state self-esteem or its variability.

Within this assemblage, within-individual variation is necessarily also approached in terms of quantity of self-esteem; for example, with the question “does an individual have the same amount of self-esteem from one day to the next” (Alessandri et al., 2013, p. 770). The assumption that is made is that the experience of state self-esteem must then also be the same across time and contexts for an individual.

Crucially then, both between- and within-individual variability of state self-esteem are not approached in terms of quality. Instead, state self-esteem is approached as a quantifiable immaterial substance (i.e., the “stuff” of self-esteem, to refer back to the substance metaphysical principle described earlier) that can be measured with some margin of error (Maul, 2013; Scheff, 2015). In this sense, this assemblage measures state self-esteem and its variability much like we might measure someone’s weight – as an objective attribute that exists universally and can be quantified (Cigman, 2004). Specifically, the Independent-Entities assemblage focuses on measuring how much “worth” (Rosenberg, 1986), “likability” or “competence” (Tafarodi & Swann, 1995) one experiences. Here, it is not necessarily the quantitative approach that is characteristic of this assemblage, but the emphasis on “measuring” state self-esteem and its variability. As we will see in the Coactive-Processes assemblage, quantitative approaches may also be adopted in order to achieve the goal of describing a process.

Summary

With the various practices described above, the Independent-Entities assemblage sketches a specific picture of the ontology of state self-esteem (i.e., of “what” it is and how it can be explained) that reflects the meta-theoretical paradigm described by the Cartesian-split-mechanistic framework: 1) A substance (i.e., in the sense that there is a fixed core or essence “under-lying” that which is fleeting), which 2) can be split into independent parts that – in an additive way – make up the whole of daily self-esteem dynamics, where 3) one part (e.g., context) deterministically causes movement in other part (e.g., state self-esteem variability) or vice versa, and which 4) can be studied objectively in terms of quantities. This assemblage thus demonstrates significant overlap with the larger assemblage in psychology that, as Morawski (2022) described, “sees stable, singular, and determined objects” (p. 167).

Section 2: Coactive-Processes assemblage

The Coactive-Processes assemblage of state self-esteem research is that which demonstrates key characteristics of the meta-theoretical Process-relational framework (see, Table 1). This meta-theoretical framework functions as an alternative to the Cartesian-split-mechanistic framework. It holds that “all facets of the individual and the context exist in mutually influential relations” (Overton, 2013b), and functions as a paradigm for meta-theories such as social constructivism, sociocultural theory, and systems theories (Raeff, 2019).

Like the Cartesian-split-mechanistic framework, this framework is characterized by a number of overlapping principles that – together – sketch a certain narrative about what it is that we study and how this should be done (see, Overton, 2015). The four principles that will be used here are: “nature as process,” “relational understanding,” “circular causality,” and “constructivism.”

Nature as process

A unifying stance adopted by researchers within the Coactive-Processes assemblage is the explicit rejection of a substance metaphysical principle that views (trait, fragile, or secure) self-esteem as a “thing-like” property that individuals “have” or “possess” (e.g., Cigman, 2004; De Ruiter et al., 2017; Levy, 2019; Pomagalska, 2005; Sabat et al., 1999; Strandell, 2017). Rather, the Coactive-Processes assemblage communicates a commitment to a process metaphysical principle or process ontology. This metaphysical principle is pivotal in the broader Process-relational paradigm, where any phenomenon that may seemingly resemble a “thing” or “entity” is actually in a constant state of action that occurs and changes across time – making it a process (Overton, 2013; Overton, 2015; or similarly, the Interactivist model; Bickhard, 2009).

From this metaphysical stance, researchers within this broader paradigm often refer to “becoming” rather than “being” (Malafouris, 2019; Marshall, 2016; Overton, 2015; Di Paolo & De Jaegher, 2021; Rescher, 1996; Van Geert & De Ruiter, 2022). With this, the assumption is that phenomena cannot be reduced to any foundational entity, because any phenomenon is “process (relations) all the way up and all the way down” (Bickhard, 2011, p. 13). For self-esteem, this translates to the notion that trait self-esteem (and other attributes like fragile self-esteem or secure self-esteem) and state self-esteem are inherently processual, and specifically, processes of “activity” that individuals “do” (Raeff, 2010), rather than “have.”

A mid-range meta-theory, which is positioned within the larger Process-relational framework, that has been used to study state self-esteem is the dynamic systems theory (see, Van Geert & De Ruiter, 2022). With this theory, researchers within the Coactive-Processes assemblage have conducted research to conceptualize and study the precise nature of self-esteem as processes.

Specifically, state self-esteem is thought to be an iterative process, where each state self-esteem experience is (in part) the product of the previous state self-esteem experience. This assumption was explicitly investigated by De Ruiter et al. (2015), Delignières et al. (2004), and Wong et al. (2014), where the temporal structure of state self-esteem was analyzed (rather than the bandwidth, as is the case in the Independent-Entities assemblage). These studies used high frequency time-series as well as methods such as Auto-Regressive-Integration-Moving-Average (ARIMA) (i.e., Delignières et al., 2004) and Detrended Fluctuation Analysis (DFA) (i.e., (De Ruiter et al., 2015; Wong et al., 2014) – both of which examine the nature and evolution of the auto-correlations between state self-esteem moments across time. The underlying assumption in studies like those by Delignières et al. (2004), De Ruiter et al. (2015), and Wong et al. (2014) contrasts the assumption made in the Independent-Entities assemblage that – as described in the section Nature as substance – state self-esteem “does not carry over from one occasion to the next” (Alessandri et al., 2013, p. 771).

Methodologically, these time-series studies conducted within a complex dynamic systems approach can be differentiated from those that obtain repeated measures of state self-esteem in order to analyze the growth curves of state self-esteem (e.g., Hutteman et al., 2015). Those used growth curve analyses used successive assessments that are relatively widely spaced – making possible dependence between subsequent values indiscernible (Fortes et al., 2004, p. 11). In contrast, the time-series studies within the Coactive-Processes assemblage acquired high-frequency data with small intervals of time between them, making it more feasible to assess iterativity across the assessments. For example, the studies that gathered the data with the smallest intervals of time between assessments were those that examined the second-to-second changes in state self-esteem. Wong et al. (2014, 2016) gathered real-time narrations of individuals’ self-concepts that were later mapped in terms of their moment-to-moment changes in valence, and De Ruiter et al. (2015,2018, 2019) studied parents interacting with their adolescent children and captured the moment-to-moment affective and behavioral actions that demonstrate positive or negative experiences of the self. The noteworthy pluralism in operationalizations of state self-esteem is another characteristic of the Coactive-Processes assemblage, which will be discussed in the section Constructivism.

Aside from time-series approaches, qualitative approaches are also adopted within the Coactive-Process assemblage when studying state self-esteem as a process. For these researchers, the focus is less on the temporal structure of state self-esteem (such that “time” is not explicitly taken into account) and more on the ways that people discursively “do” state self-esteem. For discursively-oriented studies, for example, individual’s “do” self-esteem through acts of “managing” and “presenting” self-evaluations from moment-to-moment. For example, Sabat et al. (1999) suggested that “high self-esteem is the favorable way I present myself” (p. 11), denying that it is a “hidden causal attribute of people” or something that is “in me” (p. 11). Similarly, Strandell (2017) stresses the importance of studying the “process by which self, context and goals interact in performative actions” (p. 80).

To illustrate in more depth how these researchers study ways of “doing” state self-esteem, Fearon (2004) used self-report questionnaires – not as a way of “measuring” self-esteem – but as an invitation for participants to discuss their self-worth in relation to social relationships. In this study, conversation-analytic methods were used to study the ways that individuals engaged in performative emotional and behavioral acts toward the researcher, moving through a sequence of 1) evaluating issues with regard to the social self, 2) expressing shame regarding a perceived issue, and 3) performatively repairing a perceived issue (together, called a bond threat sequence). In another example, Sabat et al. (1999) analyzed the real-life behavior and discourse of sufferers of Alzheimer’s disease within the context of their homes or an adult day-care center. Here the authors showed how individuals used their language or behavior to maintain positive daily self-regard. Specifically, these individuals would distance themselves (physically or discursively) from the “negative” attributes related to having Alzheimer’s (and other individuals who share these attributes), while they attempted to make attributes from their past (i.e., prior to being afflicted with Alzheimer’s) more salient and to align themselves (again, physically and discursively) with non-sufferers of Alzheimer’s. Individuals thus managed their daily self-esteem by means of social positioning.

Discursive studies like those by Fearon (2004) and Sabat et al. (1999) show daily self-esteem as a process of successful and unsuccessful self-protective and performative acts that are intrinsically situated within contexts, such that individuals actively negotiate the meaning and experience of a given context, and where these negotiations occur over time. These researchers rejected the study of daily self-esteem as the level of self-esteem in the moment, and instead, they examined how people engage in self constructing activities (see also, Raeff, 2010). Importantly, they did not view these activities as separate from – and influencing – daily self-esteem (which would imply “splitting,” see the previous section).

Aside from the time-series or discursive approaches to state self-esteem (or daily self-esteem) as a process, this assemblage also views trait self-esteem as a process. Importantly, researchers within this assemblage reject the view of trait self-esteem as “an entity that makes people do things” (Strandell, 2017, p. 85). Instead, trait self-esteem has been referred to as “an abstract concept denoting an outcome of processes” (Strandell, 2017, p. 75) – where the “outcome” is not fixed but dynamic. For discursive approaches within this assemblage, the outcome may be the ever-evolving establishment of identity in terms of its valence (i.e., being positive or negative). This may include an individual’s “sense of pride, burden, failure” – and importantly – how these are intrinsically part of a person’s past and present life (Sabat et al., 1999, p. 16). This conceptualization stresses that these positive and negative characteristics should thus be seen as an ongoing process, of past and present identity, and the tendencies that emerge.

In addition to the processual emphasis of trait self-esteem, in the sense of ongoing (i.e., past and present) self-evaluations, Strandell (2017) also stressed the importance of placing action in the center of this conceptualization, suggesting that “self-esteem cannot be separated from self-performative actions, so any mechanism of self-esteem must be understood as a means by which people try to establish an identity” (p. 85). This highlights the centrality of viewing trait self-esteem as both part of an ever-evolving narrative and as the (performative) actions that help to construct such a narrative. Here, we can see that trait self-esteem is thus not conceptualized as a fixed outcome, but in terms of on an ongoing process of identity “construction” (Sabat et al., 1999).

The above narrative of trait self-esteem as an ongoing constructive process can also be found in time-series studies within the Coactive-Processes assemblage. In these studies, trait self-esteem is conceptualized as the emergence of ever-evolving patterns of self-evaluation. The patterns consist of similar dynamic and active elements as those studied in the discursive studies described above, such as “personal shortcomings” or “memorable accomplishments and personal values” (Vallacher et al., 2002, p. 371). Importantly, these elements “are not static in valence or passive, but rather influence each other in order to achieve a common evaluation” (Vallacher et al., 2002, p. 371). This “common evaluation” is thus conceptualized as being trait self-esteem. Vallacher et al. (2002) suggested that elements of self-evaluation thus interact over time and emerge as a dynamic “self-structure” (or pattern). Here, emphasis is placed on the ongoing “intrinsic dynamics” of trait self-esteem, and thus also on trait self-esteem as a process.

Vallacher et al. and others (De Ruiter et al., 2017; Wong et al., 2016) emphasize the constructive nature of trait self-esteem (as a structure or pattern of self-evaluative elements) by suggesting that it functions as attractor states. These are dynamic patterns that demonstrate self-maintenance over time while, at the same time, constantly evolving and adapting to new experiences. This conceptualization thus encompasses the relative stability that is generally characteristic of trait self-esteem (i.e., “self-maintenance”) and the dynamic and constructive nature of trait self-esteem (i.e., “adaptability”). Empirically illustrating trait self-esteem in terms of attractor states, or similar patterns of self-maintenance and adaptation, has been accomplished with time-series studies (De Ruiter et al., 2018; Fortes et al., 2004; Wong et al., 2016).

In summary, we can see that – while methodologically distinct – the empirical endeavors in both discursive and time-series approaches to daily self-esteem communicate (often explicitly) a commitment to a process metaphysical principle. Both methodological traditions approach state self-esteem and trait self-esteem as active processes involving the construction, presentation, and maintenance of the self-evaluative aspect of identities or personal narratives. By avoiding “splitting” between activities on the one hand (e.g., constructing, presenting, maintaining) and a resulting “thing” or “stuff” on the other (i.e., state self-esteem), studies within the Coactive-process assemblage of daily self-esteem research fall within the self-as-process tradition, promoted by Ryan and Brown (2003). Ryan and Brown described that researchers within this tradition “view the self not merely as a concept, or as an object of self-evaluation, but as the very process of assimilation and integration” where “the self is both an inherent tendency and a dynamic, synthetic process” (p. 71). The Coactive-process assemblage approaches daily self-esteem in a similar way. As such, a working definition of state self-esteem that is specific to the Coactive-process assemblage could be contextually-based and fluctuating self-constructing activity that is experienced with a certain valence (i.e., positive and negative experiences).

The studies within this assemblage represent fruitful first attempts to study daily self-esteem – not as quantifiable “stuff” that fluctuates – but as active processes of self-constructing activity that individuals do. More research within this assemblage is necessary to further flesh out the concrete ways in which state self-esteem can be studied as a situated process, and how time-series and discursive approaches may work together to generate new knowledge and an alternative conceptualization of daily self-esteem.

Relational understanding

A crucial facet of the general Process-relational model is holism (Overton, 2007), meaning that there is reciprocal determination between parts of the whole. With holism, the assumption is that the parts cannot be separated from each other without disturbing and unraveling the “whole” which they comprise. “Parts” of the whole include internal parts (such as state and trait self-esteem) and external parts (such as context).

A commitment to holism necessitates a relational understanding in order to understand the whole, where one seeks to understand how specific parts (e.g., state and trait self-esteem; state self-esteem and context) influence and are influenced by each other as coactive processes (or interpenetration, as Overton (2015) referred to it). As such, a relational understanding (as a necessary way of understanding holism) stands in contrast to split understanding (i.e., which is characteristic of the Independent-Entities assemblage described earlier). A relational-understanding approach responds to “increasing calls for conceptualizing human functioning holistically and integratively” (Raeff, 2019, p. 321).

It is important to note that a commitment to relational understanding does not prohibit researchers from focusing on specific parts of a whole. Indeed, researchers are not expected to study all parts of a whole. Researchers can thus pursue relational understanding while focusing on certain parts of the whole by analyzing how the parts function in the of the whole (Overton, 2013b). For example, Bloom (2016) described that “looking at the parts is necessary. However, a recursive process should involve zooming in to the depths of the parts, then zooming out to the whole, then back to the parts and so on” (p. 29).

With regards to internal parts of self-esteem, relational understanding can be observed in the Coactive-processes assemblage in the way that state and trait self-esteem are analyzed. In contrast to the variable-oriented approach adopted in the Independent-Entities assemblage, studies within the Coactive-Processes assemblage tend to analyze state and trait self-esteem as distinct characteristics of a larger process. De Ruiter et al. (2017), for example, referred to trait self-esteem as a process that occurs on the “macro-level” of self-esteem development, and to state self-esteem as a process that occurs on the “meso-level” of self-esteem development. Together (also with micro-level processes of situated self-experiences), these processes are mutually connected in one larger process of circular causality (which will be described in more detail in the next section). Due to circular causality, the “whole” of self-esteem consists of parts that are involved in mutually-occurring bottom-up processes and top-down processes. The bottom-up processes denote the “constructive” characteristic of self-esteem and the top-down process denotes the “self-maintaining” characteristic, described in the previous section.

The above conceptualization of state and trait self-esteem as co-occurring processes that permeate each other at all times is common in the time-series studies of this assemblage. State self-esteem and trait self-esteem cannot be understood, or even assessed, without each other. This is because trait self-esteem emerges out of the iterative processes of state self-esteem, while state self-esteem variability is – at the same time – constrained by the emerging trait self-esteem. Trait self-esteem is thus assessed as the patterning (in terms of “attractor states” or “self-maintaining intrinsic dynamics”) of state self-esteem iterations (De Ruiter et al., 2015; De Ruiter et al., 2018; Delignières et al., 2004; Fortes et al., 2004; Ninot et al., 2005).

In a similar way, discursive approaches to daily self-esteem approach state and trait self-esteem as parts of the same process. For these studies, there is a striking lack of explicit reference to “state self-esteem” and “trait self-esteem.” Rather, these studies simply refer to “self-esteem” and the notion that it is always “enacted” in daily concrete context (e.g., Miller & Cho, 2017). Here, “enacted” means that any self-esteem experience is necessarily a concrete moment of “performing,” and at the same time, “constructing” trait self-esteem. In this way, differentiating between state self-esteem and trait self-esteem is impossible as they are two ends of the same process, where the former is “done,” “performed,” or “constructed” via the latter.

Next, the Coactive-Process assemblage places great emphasis on viewing self-esteem and the context as a whole that can only be understood through relational understanding. Within this assemblage, proximal contexts (i.e., interpersonal and situational contexts (De Ruiter, 2019; Fearon, 2004; Sabat et al., 1999) and distal contexts (i.e., culture; (Göncü, 2020; Miller & Cho, 2017; Pomagalska, 2005) are explicitly conceptualized as an intrinsic part of the constructive and performative process of self-esteem. These studies approach the two as having a reciprocal and ongoing relationship that makes them inseparable (De Ruiter, 2019). With this approach, researchers do not assume that a given context “causes” daily self-esteem to change in a certain way, but instead, the context is assumed to co-construct daily self-esteem together with other intrinsic dynamics (and vice versa; Cigman, 2004). Sabat et al. (1999), for example, describe how presenting or displaying self-esteem “require[s] a listener, another person who will honor the expressions of the speaker, and who will, by that cooperation, participate in the construction of [the way in which self-esteem is presented]” (p. 15).

The principle of relational understanding is thus often an explicit part of the conceptual frameworks adopted within the Coactive-Processes assemblage. This conceptual importance is mirrored by empirical choices. Empirical studies of state self-esteem within this assemblage do not attempt to isolate measures of state self-esteem from context in a way that is often (but not necessarily) done in the Independent-Entities assemblage. Instead, the two parts are assessed as the concurrent experiences of a context and one’s self in this context.

Noticeably, studies within the Coactive-Processes assemblage approach “context” in a broader way than is typically done in the Independent-Entities assemblage (where it is primarily the proximal social context that is investigated). In the Coactive-Processes assemblage, context is often examined from a cultural-historical framework, a tradition that has its roots in the Vygotskian emphasis on how children’s development is contingent upon their culture (Hedegaard et al., 2008). The cultural-historical framework thus focuses on how larger societal and cultural norms and values become personalized in individual’s own narratives and behavioral norms. For this assemblage, daily self-esteem is thus often conceived of as being situated in these larger contexts of societal and cultural norms and values. For example, in Miller and Cho (2017)’s comprehensive ethnographic study of how self-esteem is enacted and personalized in domestic spaces in America, they show how cultural ideals inform (and are reinforced by) every-day parenting (and management of children’s self-esteem) and daily experiences of self-esteem. They found that self-esteem was managed in daily practices such as praise, and that the kind of praise offered (and other practices) reflected the Western belief in the importance of high self-esteem in child rearing, and in self-esteem as predominantly related to individual strengths. Similarly, Gulbas (2013) showed how women in Venezuela often internalized race-based oppression and western beauty norms, such that these norms informed their conceptualization of what high self-esteem is.

In short, the Coactive-Processes assemblage views state self-esteem and context as parts of one whole, such that they are ontologically inseparable. This ontological assumption is often described explicitly in these studies, and is also enacted through the methodological choice to describe daily self-esteem in situ, where the contexts studied vary from the interpersonal to the wider cultural and historical contexts.

Causal explanation

Whereas the Independent-Entities assemblage often seeks to explain daily self-esteem dynamics by means of the antecedent responsible for pushing state self-esteem (i.e. efficient causal explanation, described earlier), the causal explanation adopted in the Coactive-Processes assemblage is grounded in the principle of relational understanding (described above). From this explanatory aim, the whole of self-esteem can only be understood by means of examining how parts relate to each other and how parts relate to the whole (Overton, 2015). This calls for a form of explanation that focuses on, what is called, circular causality (this was briefly touched upon in the section Relational understanding; Overton, 2015; Witherington, 2011). Circular causality involves the ongoing and mutual processes of bottom-up construction (or emergence as it is typically called) and the top-down self-maintenance (or loose constraint; Van Geert & De Ruiter, 2022; Witherington, 2011).

Causal explanation focusing on circular causality places an emphasis on the intrinsic dynamic of state self-esteem, and the notion that these intrinsic dynamics partly cause trait self-esteem (De Ruiter et al., 2017). In addition, this form of causal explanation places emphasis on processes themselves as explanatory (Chatham-Carpenter & DeFrancisco, 1998; De Ruiter et al., 2015; Fortes et al., 2004; Strandell, 2017; Vallacher et al., 2002).

The concrete research practices that communicate this form of causal explanation are the same as those that communicate the principle of “relational understanding” and “nature as process,” described above. Thus, researchers from this assemblage demonstrate a focus on circular causality when they aim to describe a particular sub-process within the larger circular process (acknowledging that it is part of the whole explanation). Studies have does this by, one, describing the ways in which intrinsic dynamics of state self-esteem emerge into patterns that we can call trait self-esteem (De Ruiter et al., 2015; Delignières et al., 2004; Fortes et al., 2004; Ninot et al., 2005; Vallacher et al., 2002; Wong et al., 2014, 2016) – via time-series analyses, two, describing how patterns of self-esteem (i.e., trait self-esteem) constrain the moment-to-moment-experiences of state self-esteem (De Ruiter et al., 2018) – via time-series analyses, and three, how contexts constrain and become personalized in individual’s experiences of self-evaluations (De Ruiter, 2019; Gulbas, 2013; Miller & Cho, 2017) – via time-series and discursive analyses.

What should be noted here, is that these studies share a common aim of “describing” dynamic processes in order to explain how parts relate to each other or to the whole (Overton, 2015; Raeff, 2017; Van Geert & De Ruiter, 2022). Here, researchers demonstrate “description as a meaningful and legitimate means of understanding” (Bloom, 2016, p. 27). This stands in contrast with the aim of explanation via “prediction” that is observed in the Independent-Entities assemblage.

Constructivism

An important principle within the larger Process-Relational framework is that of constructivism. With constructivism, an “organism must construct its own reality by actively structuring and transforming the world, a process whereby the world and the organism structure one another, with the ‘world as known’ – or knowledge – being a truly emergent product of this continual interplay rather than a discovery of a preexisting reservoir of inherently meaningful information” (Witherington & Margett, 2011, p. 287). This principle contrasts that of “objectivism” and “realism” (demonstrated in the Independent-Entities assemblage), where phenomena are assumed to be fixed and universal in nature, existing independently of our observing them (Overton, 1991). Thus, whereas the Independent-Entities assemblage necessarily strives to “measure” the quantitative between-individual differences in the universal “stuff” that is self-esteem (see the section Objectivism), the Coactive-Process assemblage works from the premise that the very nature of an individual’s daily self-esteem is actively constructed in relation to the context in which these experiences occur. From this perspective, psychological phenomena such as state self-esteem are not simply internal representations that are communicated to others, but customized and molded to fit the current interaction (McGann & de Jaegher, 2009). Individuals therefore necessarily “do” (i.e., maintain, construct, perform) state self-esteem in different ways depending on the context in which it is situated.

Given the assumption that each individual’s state self-esteem experience (and the quality thereof) will be uniquely constructed within the current context, the Coactive-Processes assemblage demonstrates a commitment to assessing daily self-esteem in a time-, context-, and person-specific way, rather than with one universal form of assessment. An assessment of daily self-esteem is, firstly, developed based on the specific context in which daily self-esteem occurs in the relevant study, resulting in a wide variety of operationalizations of daily self-esteem. Within the context of discussing a self-report (Fearon, 2004) or a self-descriptive narrative (Wong et al., 2014, 2016) researchers have assessed daily self-esteem as self-reflective thoughts. In contrast, when daily self-esteem is assessed within interactions with others, researchers have assessed it by means of the relevant behaviors that occur in that context (De Ruiter et al., 2015, 2018; Sabat et al., 1999). Researchers have thus designed each particular assessment of daily self-esteem based on the kind of self-experiences that a particular context affords.

In addition to a context- (or study)-specific approach to the assessment of daily self-esteem, the qualitative approach to data collection that is adopted in each of these studies allows for the assessment of daily self-esteem to evolve qualitatively across time and to vary between individuals. As such, these studies do not only operationalize state self-esteem variability in terms of fluctuations in the level of state self-esteem, but in terms of the quality that state self-esteem takes on from moment to moment.

The emphasis placed on person-specific experiences of state self-esteem highlights why – for this assemblage – it is crucial to “describe” daily self-esteem, rather than to “measure” it (see the previous section). Each operationalization of state self-esteem must therefore be justified based on the affordances provided by a research context, and the description of daily self-esteem must explicitly take this into account by describing the situated processes of daily self-esteem.

The plurality of approaches to daily self-esteem within the assemblage is supported by phenomenological accounts of self-esteem, which show that it is important to look at how individuals actually experience and conceive of their own self-esteem. The traditional conceptualization of self-esteem assumes that individuals place value on characteristics like worth, competence and likability, and often in comparison with others (e.g., “I am able to do things as well as most other people”; Rosenberg, 1965). However, lived experiences of self-esteem may be greatly different, not just across contexts, but across individuals. As Chatham-Carpenter and DeFrancisco (1998) argued, while self-esteem may represent the evaluative component of self-concept for all individuals, the standards for these evaluations will likely vary across individuals, and they may or may not be reflected in the standard self-esteem scales. They argue that the common self-esteem scales likely reflect androcentric standards (such as independence and assertiveness) that may not be suitable for women, for example. Similarly, others have argued that standard self-esteem instruments may also make assumptions about what individuals (should) value in themselves based on western ideals and norms (Markus & Kitayama, 1991). Explicitly describing the ways in which individuals experience self-esteem across cultures, contexts, and time is central in this assemblage.

Discussion: toward a synthesis between the two assemblages

The above sections are intended to make explicit the ways in which research on state self-esteem communicate diverging commitments to what state self-esteem is (i.e., ontological commitments) and how it can be understood (i.e., epistemological commitments) – summarized in Table 1. These commitments are communicated either explicitly (often through discourse norms or reference to meta-theories that take an ontological stance) or communicated implicitly through practices such as analytical methods and assessment methods. With this, a narrative for the what and how of state self-esteem emerges across the studies that demonstrate these commitments.

Specifically, the previous sections show the ways in which two distinct narratives emerge, each within an “assemblage” (i.e. dynamic collection) of researchers and practices; which I have called the Independent-Entities assemblage and the Coactive-Processes assemblage. Each of these demonstrates adherence to a certain metatheoretical framework – which functions as a worldview, or paradigm – about what exists in the world. The former assemblage communicates principles from the larger Cartesian-split-mechanistic framework and the latter communicates principles from the Process-relational framework (Overton, 2013, 2015).

To summarize the characteristics of these two assemblages: For the Independent-Entities assemblage, state self-esteem is approached as an ephemeral event through which an essence or core endures, which is trait self-esteem. For this assemblage, state self-esteem can be isolated (i.e., “split”) from core self-esteem attributes (including trait self-esteem, fragile self-esteem, and secure self-esteem) in order to understand how daily self-esteem works. Given this “splitting,” the form of explanation that is generally pursued in this assemblage is that of antecedent-consequence or cause-effect relations (i.e. “efficient causal explanations”). In line with this emphasis on predictive relationships, the focus is on quantitative effects, and thus “measuring” state self-esteem as a universal and objective variable that differs across people in terms of its quantity (i.e., objectivism).

The general ontological narrative that emerges from the Independent-Entities assemblage is that of daily self-esteem as a “machine” (a metaphor explicitly used to describe the underlying assumptions of the larger Cartesian-split-mechanistic framework, e.g., Overton, 2015). A machine is fundamentally stable (i.e., a stable entity), where activity and variability (i.e., ephemeral events) can be explained and reduced to movement in one or multiple parts of the machine or outside of the machine (i.e. antecedents or causes). Each activity of the whole machine, and each individual movement of parts, can thus be isolated (i.e., “split”). Within the Independent-Entities assemblage, we can compare the machine to trait self-esteem and the moment-to-moment variability in its activity to state self-esteem. Variability in daily self-esteem is then thought to be caused by a push, or antecedent that deterministically brings about the expected behavior, in the same way that a machine moves when something specific happens to it.

In contrast, the Coactive-Processes assemblage does not assume that trait self-esteem is a core or essence that endures through fleeting events of state self-esteem. Rather, trait self-esteem and state self-esteem are both thought to be, inherently, processes, which cannot be reduced to any stable entity (i.e., process ontology). With this conceptualization of the nature of self-esteem and its elements, trait and state self-esteem are both approached in terms of activity, and the notion of “doing” self-esteem (rather than “having” self-esteem). Yet, state and trait self-esteem are distinct in the following way: State self-esteem is viewed as an active process that is primarily variable from moment-to-moment. Within time-series approaches, this activity is studied by characterizing the precise iterative nature of this and the nature of the intrinsic dynamics that emerge over time. Within discursive traditions, this activity is studied in terms of the quality and function of these activities as situated experiences, emphasizing activities like performing, constructing, and managing self-esteem. With the above methodological approached, trait self-esteem is viewed – not as a separate process that occurs alongside this daily process – but as a description of what comes about across time; dynamic patterns or ever-evolving narratives of self-evaluation. To capture the processual and activity-based focus of these two approaches, I proposed the following working definition of state self-esteem from within the Coactive-Processes assemblage: contextually-based and fluctuating self-constructing activity that is experienced with a certain valence (i.e., positive and negative experiences).

The relationships between these two – the daily activity of self-esteem (i.e., state self-esteem) and the evolution of patterns that emerge out of this daily process (i.e., trait self-esteem) – are explained and understood based on their ongoing and mutual relationship with each other (i.e., relational understanding). Thus, daily self-esteem emerges into what can be called trait self-esteem, and trait self-esteem loosely constrains or invites certain self-evaluations in the moment. Relational understanding goes hand in hand with the causal explanation that is generally adopted in this assemblage, which is that of circular causality, where both elements (as well as context) are always co-causing each other by the very fact that they are coactive and mutually constituting each other. Finally, within this assemblage, state self-esteem is operationalized and assessed in a multitude of different ways, including verbal reflections, self-report reflections, behaviors, and affect.

Both quantitative (i.e., time-series) and qualitative (i.e., discursive) approaches are adopted to study daily experiences of self-esteem. This reflects an underlying assumption that state self-esteem is always a constructed experience that is unique across time, contexts, and individuals (i.e., constructivism), and is in line with the denial of a reducible entity that exists independently from our studying it. The narrative that emerges from this assemblage, of what state self-esteem is, is that of an “act in context” (Overton, 2015). This conceptualization stresses the notion that it is active, moving, and always situated in an ongoing and moving context. As such, an attempt to stop or isolate any parts of this “act in context” will be, at best, incapable of studying that which is of interest, or at worse, a disturbance to (or unraveling of) the act in context.

Given the differences in underlying philosophical commitments enacted in the two assemblages summarized above, it may seem unclear whether a synthesis between the two is possible. They are rooted, after all, in different “worldviews.” In practice, there is currently next to no scholarly interaction between the two assemblages – demonstrating that they function as separate assemblages. Below, I argue that, firstly, a large practical separation between the two assemblages is not inevitable, but rather, a synthesis between them is possible (in terms of communication and interaction between practices and researchers). Secondly, I propose how we as a field might achieve a successful synthesis. I will argue that the field of daily self-esteem requires open reflexivity from researchers regarding their own philosophical commitments about daily self-esteem and those communicated by their research practices (as demonstrated in the above sections) – as well as reflection on whether these align or not.

Synthesis is possible because … ontology is not an empirical question

Importantly, the two divergent conceptualizations of what state self-esteem is in relation to trait self-esteem or context, or how it is experienced, need not be in direct competition. As Franz (2022) described, the nature of psychological phenomena (for example, as being fundamentally quantitative versus qualitative) is not an empirical question. It is, in other words, not possible to “prove” whether state self-esteem is a phenomenologically varying phenomenon or a phenomenologically universal and stable one. Similarly, it is not possible to “prove” whether state self-esteem can be isolated from a context or not. This is because the very operationalization of “state self-esteem” and the “context” will inevitably constrain and direct our answer to this question. Even the answer to the question of whether or not the temporal dynamics of state self-esteem are stable or not is contingent upon the mode and frequency of state self-esteem assessments.

The answers to these types of questions are thus not directly comparable across assemblages, as it is impossible to disentangle the methodological approach from the specific knowledge that is produced (Van Geert & De Ruiter, 2022). Self-esteem is indeed always going to be a construct (Levy, 2019), and that construct will vary depending on the philosophical position that motivates the specific pursuit. This is good news, as it means that results produced by both approaches to dynamic self-esteem do not resemble opposing hypotheses, such that one must be false.

Synthesis is possible because … divergent research aims demonstrate unique merits

It is vital that we recognize that the two assemblages of daily-self-esteem research have largely different scientific aims. The Independent-Entities assemblage is focused on establishing between-individual predictive relationships between state self-esteem (i.e., its fluctuations or levels of variability) and other constructs of interest, while the Coactive-Processes assemblage is focused on understanding the idiosyncratic and situated nature or function of daily self-esteem. In recognizing these different aims, we can work toward more interaction between the two assemblages without friction – i.e., a plurality of approaches.

Both assemblages contribute successfully to an understanding of daily self-esteem, both in their own unique ways: One outlining large-scale relations between variables, and the other offering rich descriptions and explanations of processes. As Chang (2019) stated, “the plurality of aims within science gives plurality to the meaning of ‘success,’ for different scientists and even for an individual scientist” (p. 175). Both of these assemblages are successful in elucidating something about daily self-esteem based on different definitions of what it means for a study to be successful. Chang also outlined that, with such a plurality of success (stemming from different aims), a similar plurality of methodology logically follows. The two assembles offer different kinds of knowledge production, necessitating different methodological practices.

In order to fully understand daily self-esteem, we need practices adopted in both assemblages. We need the knowledge produced by the Independent-Entities assemblage, because we need to know which dispositions tend to correspond with stable and unstable self-esteem, which contexts tend to correspond with an increase in daily dynamics, and we need to know how levels of self-esteem tend to correspond with developmental outcomes. In short, it is useful to understand the association between state self-esteem (and the variability thereof) and internal, contextual, and developmental factors of interest.

We also need the knowledge produced by the Coactive-Processes assemblage. We need to understand what daily self-esteem means to individuals as a lived and situated experience. We need to understand how people make sense of these experiences through their interactions with interactional partners and as contingent upon larger societal discourses. We need to understand how these experiences emerge into entrenched ways of experiencing one’s self over time, and thus how daily self-esteem forms robust patterns through processes of intrinsic dynamics. This knowledge production provides a rich understanding of what state self-esteem is in terms of activity, and how it functions as a coactive process in relation to ongoing contextual changes (De Ruiter & Gmelin, 2021). Importantly, because this information is gained within the context in which self-esteem dynamics naturally occur, it promises to be conceptually close to that which we are interested in, and robust against threats to validity such as the ecological fallacy and Simpson’s paradox (Simpson, 1951).

Toward a synthesis of assemblages

The above sections describe why, even though the underlying ontological commitments communicated (either implicitly or explicitly) manifested in the two assemblages may be competitive, the assemblages themselves need not be. The two assemblages demonstrate different ontological views (not necessarily explicitly adopted) that cannot be empirically proven or falsified, in part because they cannot be disentangled from the research practices adopted in order to study these ontologies. Moreover, these two ontological perspectives are connected to different research aims that are not directly combative.

With this, we can begin to sketch ways in which the differences stemming from the two assemblages may be utilized to collectively advance our understanding of daily self-esteem.

Assemblages as open, dynamic, and practice-based

For a synthesis between assemblages to take place, it is important to stress that, while the norms, aims, and underlying ontological commitments observed in the two assemblages are indeed divergent, this does not mean that the assemblages represent separate or fixed entities themselves. Similar to Raskin’s (2020) analysis of critical and indigenous psychologists’ epistemologies, we should refrain from treating divergent characteristics as “either intellectual commitments from which we are not allowed to stray or baked-in psychological qualities” (p. 397). Accordingly, the two assemblages – despite appearances – are not fixed “types” of research or researchers; nor do they study psychological phenomena with true ontological differences. The assemblages are simply interacting components (practices, methods, researchers) that are dynamic and open to change.

Similarly, we should remember that researchers that are part of these assemblages do not have distinct essences that are fixed and mutually exclusive. As Warren (1992) suggests, researchers are not “subjectivists” versus “objectivists” (or any other categories stemming from the two assemblages described in this article). Instead, researchers are simply researchers who do “subjecting” or “objecting,” respectively (Raskin, 2020; Warren, 1992). This conceptualization emphasizes the fact that there may be instances in which a single researcher adopts philosophical tenets or methods from the “other” assemblage, or engages in components (i.e., philosophical commitments or practices) from both assemblages simultaneously. This de-essentializes these two assemblages (and their norms and researchers), and allows researchers to contribute to the research aims from both assemblages in a flexible way. Indeed, “if we conceive of [objecting and subjecting] as things we do rather than states we achieve, we can alternate between them, thus taking advantage of their respective strengths and weaknesses” (Raskin, 2020, p. 397).

A similar suggestion was made by De Ruiter and Albert (2017) in their analysis of how Conversation Analysis and experimental designs might be able to interact. Drawing from this, perhaps inferential-statistical studies of state self-esteem and descriptive ones can join one empirical cycle, where the context of discovery and the context of justification join as one cyclical research endeavor. For example, descriptive studies can be used to discover ways in which daily self-esteem is managed in daily life, or what the intrinsic dynamics of state self-esteem might be, and inferential-statistical methods can then be used to test the extent to which these findings can be generalized to larger groups of individuals.

Flexible movement between the assemblages is exemplified in Hank and Baltes-Götz’s (2019) study of state and trait self-esteem processes across the span of two days. While these authors used many practices characteristic of the Independent-Entities assemblage (e.g., methods of analysis and assessment of state self-esteem), the authors drew many of their theoretical assumptions and their research aims from studies from the Coactive-Processes assemblage (e.g., intrinsic dynamics of daily and situated self-esteem, and circular causality; De Ruiter et al., 2015; Fortes et al., 2004; Ninot et al., 2005). These studies provided motivation for Hank and Baltes- Götz’s focus on iterativity and context dependency. Moreover, the authors’ results were largely interpreted through the theoretical propositions of the Self-Organizing Self-Esteem (SOSE) model (De Ruiter et al., 2017), which aligns with the broader Process-Relational model and is characteristic of the Coactive-Processes assemblage. Furthermore, Hank and Baltes-Götz (2019) described their study as providing support for the theoretical propositions put forward in the SOSE model. This example illustrates how researchers can take part in both assemblages by drawing from practices (methodological or philosophical) from both. As a result, the two assemblages of self-esteem can inform, and indeed support, each other. Note that this example illustrates one way in which the two assemblages can support each other, such that there are other ways in which this can be done. Exploring these alternative forms of synthesis forms a fruitful direction for future research on daily self-esteem.

Reflexivity

For the above scenario of synthesis to be achieved, I suggest that reflexivity must be incorporated into the communication that takes place (via manuscripts) for all research on daily self-esteem. Reflexivity is the reflection on the relationship between knowledge production (i.e., what we know) and the epistemological, ontological assumptions, and larger historical and social contexts in which our research is embedded. This involves a bi-directionality between these two, where we reflect on how our assumptions and contexts inform our research, and how our research reinforces or privileges certain assumptions and contexts (Morawski, 2014; Pilgrim, 2020).

An important part of reflexivity is the understanding that the researcher and the knowledge produced by the researcher cannot be separated (Morawski, 2014), as discussed throughout this manuscript. This understanding, however, is not always shared in areas of psychology where more positivist approaches are adopted, in which there is a search for an objective truth (Pilgrim, 2020). As such, reflexivity is not a common component of empirical work in psychology in general (Morawski, 2005), nor in the Independent-Entities assemblage more specifically.

With an eye on encouraging successful knowledge production in studies on daily self-esteem, reflexivity is necessary in order to ensure that any attempts to combine practices from both assemblages are, in fact, fruitful. This is because combining disparate methodologies that stem from and embrace different underlying assumptions can create “muddle” as research output (Potter, 2003, p. 787).

When drawing from practices across assemblages, muddling is potentially avoided with open reflexivity, where researchers can make explicit a) which philosophical assumptions they – as researchers of daily self-esteem – hold, b) whether the philosophical assumptions communicated by their practices align with those that they hold, and c) if a divergence is present, they can reflect on why this divergence exists. Indeed, it is not uncommon that researchers may actually be trying to express something theoretical that simply cannot yet be done methodologically (Overton & Lerner, 2014). This is because methodological developments often lag behind theoretical developments. For example, the intention of studying trait self-esteem, state self-esteem, and context as coactive processes that demonstrate circular causality may be realized as closely as possible via transactional effects – such as those found by Hutteman et al. (2015). With open reflexivity, researchers can bring these assumptions to the fore and develop an “open-minded and generous” attitude to divergent practices (necessary for pluralism; Chang, 2019, p. 182) that is based on an understanding of where these differences come from and what their epistemological consequences are.

A synthesis between assemblages may therefore not require a full adherence to methods and practices that communicate the ontological claims one wishes to communicate. The aim, instead, should be that of a “sufficiently” coherent set of practices, even if they stem from diverging ontological commitments. This stems from Chang (2019)’s coherence theory of reality, where researcher practices need only “fit” together in order to work toward an explanation of reality that makes sense. I argue that opening communicating our reflexivity can be the means for achieving sufficient fit.

Currently, reflexivity of underlying ontological commitments manifested in research practices takes place within the Coactive-Processes assemblage (e.g., Cigman, 2004; De Ruiter et al., 2017; Levy, 2019; Miller & Cho, 2017; Sabat et al., 1999; Van Geert & De Ruiter, 2022), but not yet in – or with – the Independent-Entities assemblage. It is crucial that researchers from both assemblages openly and publicly reflect on the philosophical assumptions underlying their chosen methodologies and research questions and engage with each other’s accounts of reflexivity. This work may take the form of conceptual studies (such as this one), but it can also occur within studies that are primarily empirical. Only then can we work toward collective reflexivity as a larger field of daily self-esteem research, which is necessarily built upon communication and interaction between those that represent different epistemologies or contexts (Teo, 2017).

Taken together, at minimum, it should be possible for the two assemblages to fit together and form one larger assemblage of daily-self-esteem research practices, where practices are adopted in accordance with the particular aim of the researcher. Ideally though, through open reflexivity, we may additionally find a practical synthesis between practices and ontological commitments. With reflexivity taking place across assemblages, the entire community of daily-self-esteem researchers can engage in open discussions regarding the specific ontological commitments (outlined in Table 1) that we wish to conserve and those that we wish to reject, and we can assess which differences can co-exist with sufficient “fit” (Chang, 2019). This is crucial for the development of a solid conceptual foundation that we can collectively work with.

With this article, deeply rooted ontological commitments are brought to the fore, and in particular, the ways in which these commitments are communicated and reinforced by the practices we adopt. As such, I hope to have provided the grounds and conceptual work necessary for a reflective process that is communal and that furthers this exciting field of research.

Disclosure statement

No potential conflict of interest was reported by the author(s).