Conceptual limitations in emotion regulation self-report scales

Abstract

In this paper, we provide a brief survey of the history of the concepts of emotion and emotion regulation and compare two of the leading scales of emotion regulation—the Emotion Regulation Questionnaire (ERQ) and the Difficulties in Emotion Regulation Scale (DERS). Our purpose is to determine where there might be room for improved conceptual clarity in the study of emotion regulation research today. We conclude by suggesting that the conceptual difficulties within and between these scales are symptomatic of a deeper problem in psychological research, that of reductionism, and suggest a nonreductive read of the concepts in question.

Emotion regulation is well embedded in psychological research as an important measure of mental well-being. Yet, its pervasiveness can shroud the reality that it is a fairly new concept in empirical research, having been introduced in the 1980s by particular researchers using particular scales of measurement (Gross, 1999).

Emotion itself is also a relatively recent concept, one that has so far eluded conceptual consensus among scientists (Campos et al., 2011; Dixon, 2012; Izard, 2010). Layering on “regulation” to the already-contested notion of emotion therefore risks adding to the polyphony. Though it would be too ambitious, and even potentially misguided, to attempt to clarify once and for all what an emotion is, and what it means to regulate an emotion, it is helpful to compare the language employed by the leading theorists of emotion regulation to determine where the conceptual disagreements and confusions are found (Wittgenstein, 1953/1989).

In this paper, we will begin with a brief survey of the history of the concept of emotion regulation. We will then compare two of the leading scales—the Emotion Regulation Questionnaire (ERQ; Gross & John, 2003) and the Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004)—in order to determine where there might be room for improved conceptual clarity in the study of emotion regulation research today. We apply Campos et al.’s (2011) distinction between intrapersonal versus relational approaches to defend and illustrate a nonreductive read of emotion regulation. We conclude by suggesting that the conceptual difficulties within and between these scales are symptomatic of a deeper problem in psychological research, that of reductionism.

A brief conceptual history of emotion regulation

Emotion regulation is a compound term, with both words packing dense meanings. Before addressing the history and conceptualization of emotion regulation, we will briefly review how psychologists arrived at the modern concept of emotion. In the 19^th century, Scottish philosopher Thomas Brown offered up the relatively new folk term “emotion”—a word with French roots meaning something akin to “physical agitation”—to intellectually unify the historical categories of passions and affections (Dixon, 2012). Passions were understood as something like overpowering emotions, usually with overtones of immorality—a dangerous force that needed to be tamed (Oxford English Dictionary, 2024). Affections, by contrast, were understood as the milder, more virtuous counterparts to passions—for example, love and gratitude. Redefining moral and religious categories into one physical category proved to be scientifically useful—“emotion” was quickly taken up by scholars from Darwin (1872) to James (1890). True to its etymology, the focus was on physical rather than moral processes. Brown himself, however, admitted it was difficult to define this new amorphous category. After all, what is left of an essentially moral concept when the morality is removed? That remains unclear some 200 years after Thomas Brown, as there still seems to be no core organizing concept that stands in for its moral antecedents. For example, a recent survey of 34 psychologists and behavioural neuroscientists found that there is still “no generally accepted definition” of emotion other than the concomitant neurological processes (Izard, 2010, p. 369).

That the concept of emotion used to be more commonly found in the territory of sages and clergy but came to be associated with neuroscience and psychology is perhaps emblematic of a larger tr found in the territory of sages and clergy but came to be associated with neuroscience and psychology is perhaps emblematic of a larger trend in modernity, that of isolating the thing from its context in order to purportedly understand it better (Taylor, 1989). This has proven useful in the natural sciences and even in psychology. But when it comes to the human experience, what is removed from its originating context may result in something entirely different. Wittgenstein, for instance, argued that language was only meaningful within a pattern of activity. He went so far as to say that “if a lion could talk, we wouldn’t be able to understand it,” not because the words were unintelligible but because the patterns of activity—the general form of life as he called it, that makes lion language meaningful—would be foreign to us (Wittgenstein, 1953/1989, p. 235). Thus, when Thomas Brown used a new word, emotion (physical agitation), to stand in for value-based appraisals and actions with their concomitant causes and effects in the mind, body, and environment, he effectively redefined and attenuated emotion. The rich value-based, contextual, relational, and moral category of emotions was reduced largely to the physical. A similar point has been made by Campos et al. (2011) where they distinguish between relational versus intrapersonal views on emotion regulation—a discussion we return to later.

The concept of constraining the passions is ancient, with both Stoic and Christian roots (Sorabji, 2000). For example, the famous Stoic Marcus Aurelius (2002) wrote on the importance of cognitively reframing situations (to use modern psychological jargon) in order to tame the passions and stay in rational control. Emotion regulation, however, came to prominence in psychology research only in the 1980s. Gross (1999), in his review of the history of emotion regulation research in psychology, argues that Freud’s anxiety-reduction theory and Hans Selye’s coping theory were the precursors to modern emotion regulation theory.

As is well known, for Freud (1926/1959), the reduction of anxiety was a time-intensive process that involved addressing one’s unconscious inner conflicts. It is interesting to note that as the putative great-grandfather of emotion regulation, Freud believed it required hundreds of hours to decrease one’s anxiety—psychoanalysis typically requires three to five sessions per week for months or years (Zusman et al., 2007). From the anachronistic standpoint of a history of emotion regulation, this sounds like an absurdly long time to regulate negative emotion. And yet, Freud’s theory of anxiety was developed in reference to morality, not cognitive appraisals or mere physiological overarousal. To learn to balance between the id (often associated with vice) and the superego (the strict taskmaster of learned standards of virtue) meant resolving moral conflicts learned implicitly through one’s life experiences—a task that transcends mere skill or strategy (Freud, 1926/1959).

In Gross’ (1999) account, next in historical line comes Selye writing only a few decades after Freud. Coping, for Selye (1950), was primarily about how the body’s homeostasis systems responded to external demands. Stress mounts when external demands become too great for the body’s ability to “cope”—that is, maintain homeostasis—leading to biological problems. Where Freud’s proto-vision of emotion regulation was one of time-intensive psychological excavation, Selye’s was essentially biological. Since he was not a psychologist, Selye did not offer a complete theory of emotion regulation. In his limited comments on how to consciously down-regulate stress, he deferred to Freud and offered simple solutions such as smiling and practicing gratitude (Selye, 1976).

After acknowledging Freud and Selye’s pioneering work, Gross (1999) himself defines this new field of emotion regulation as the “heterogeneous set of processes by which emotions are themselves regulated” (p. 557). In his view, emotions are experiential, behavioural, and physiological systems adapted for solving problems. However, these adapted problem-solving systems may not always be beneficial, and thus require up-or-down regulation of either positive or negative emotions depending on the situation (Gross, 1999). There are two noteworthy developments that emerge with Gross’ theory. First, compared to Freud’s therapeutic process, Gross’ emotion regulation is more moment-to-moment. Gross’ (1999) examples of regulating one’s emotions include “deciding to change an upsetting topic” or “lighting a cigarette when anxious,” both of which bear similarities to Freud’s concept of defence mechanisms (pp. 557, 558). They also likely reflect a general shift away from intensive long-term psychodynamic approaches toward shorter-term CBT-style interventions. Second, in contrast to Selye, Gross’ emotion regulation focuses on behaviours one does to modify or control emotion, rather than automatic biological processes that the body is adapted to perform.

Gross’ (1999) survey is a useful starting point for understanding the modern conception of emotion regulation and how it developed from earlier theories. But more can be learned about its conceptual foundations by comparing two of the most widely used self-report scales for measuring emotion regulation: Gross’ own ERQ (Gross & John, 2003), and the DERS (Gratz & Roemer, 2004). We chose these two scales because they are the two most frequently cited scales for measuring adult emotion regulation and have broad applicability across clinical and social psychological domains. Collectively, these scales have been cited well over 20,000 times. They can therefore be considered representative for how emotion regulation is conceptualized and measured in adult populations. By analyzing the language used in these measures, and the theoretical frameworks offered by their creators, we hope to better understand what each scale is trying to accomplish, and whether there are any hidden conceptual complexities or presuppositions contained in them (Bennett & Hacker, 2022b).

The conceptualization of emotion regulation in the Emotion Regulation Questionnaire

In their scale—arguably the first validated self-report scale of emotion regulation—Gross and John (2003) conceptualize emotion and its regulation in the following way:

an emotion begins with an evaluation of emotion cues. When attended to and evaluated in certain ways, emotion cues trigger a coordinated set of response tendencies that involve experiential, behavioral, and physiological systems. Once these response tendencies arise, they may be modulated in various ways. (p. 348)

According to their model, emotion can be modulated through anticipatory regulation (e.g., selecting the situations you choose to put yourself in, or appraising the meaning of a situation before the emotion fully sets in) or response-focused regulation (e.g., suppressing your behavioural response once you are already experiencing an emotion). The scale was intended to be measured against outcomes of “affect, well-being, and social relationships” in an adult population (Gross & John, 2003, p. 348). Thus, it can be surmised that it is meant to apply to any area of adult psychology where these variables are studied.

Gross and John (2003) do not explicitly define “regulation” or explain why they have used this particular word to describe the phenomenon they are trying to measure, but their implicit definition can be inferred from the language used in the scale items. For instance, various items include phrases such as “I control my emotions,” “I change the way I’m thinking,” “I make myself think,” and “I make sure not to express” (Gross & John, 2003, p. 351). These words bring to mind the earliest English uses of the word regulate, which referred to controlling mechanical devices such as watches. More commonly today, regulate usually means to control or direct according to some external standard, as in “the state regulates the mining industry” or as in biology where genes are regulated (that is, turned off or on by environmental inputs), or the processes by which mechanistic systems operate to achieve some set goal (Oxford English Dictionary, 2023). One example of mechanistic regulation would be the way in which a cybernetic control circuit (e.g., a thermostat) “senses” temperature input and “acts” accordingly to regulate the temperature according to some set point. In using the phrase “emotion regulation,” then, it seems that Gross and John are conveying that persons are able to bring themselves under control according to some rule or standard.

Gross and John (2003) argue that expressive suppression is a maladaptive regulation strategy, in contrast to the more adaptive strategy of cognitive reappraisal, for several reasons. One, expressive suppression is employed late in the process of emotion, once the emotion is already being experienced, making it a responsive rather than anticipatory regulation strategy. Continually suppressing an ongoing emotion is cognitively taxing, leading to psychological problems. Two, they assert that one cannot suppress the experience of the emotion, only the expression of it, which allows negative emotions to “linger” and “accumulate unresolved” (Gross & John, 2003, p. 349). This disjunction between the inner and outer reality of the person creates a sense of incongruence and inauthenticity, leading to alienation from self and others (Gross & John, 2003). Cognitive reappraisal is believed to circumvent these problems by altering one’s appraisal of a situation, thus changing the emotion before it is experienced. It is worth considering three possible conceptual concerns present in this process model of emotion regulation and in the particular scale that operationalizes it.

First, there seems be an arbitrarily narrow selection of emotion regulation strategies drawn from the fuller process model, along with certain assumptions made about each stage. In their process model, an emotion could be regulated in any of the following ways: by selecting the situation one is in, modifying the situation, deploying one’s attention differently within that situation, cognitive reappraising one’s interpretation of the situation, and finally, suppressing one’s response to the situation. Yet Gross and John (2003) only select two strategies to measure: cognitive reappraisal and suppression (which happen to be, temporally, the two latest strategies one can use in a given situation). Theoretically, situation selection would be just as legitimate as reappraisal and suppression (and temporally the first strategy one could use), yet is not measured in their scale. To use an illustrative example from Bennett and Hacker (2022a), a fear of heights might primarily manifest in one’s avoidance of skyscrapers and ladders. The emotion of fear is present, but it is successfully “regulated” by avoiding fear-eliciting situations (and, one might add, this fear-motivated avoidance means the bodily perturbations of sweaty palms and a pounding heart would likely be successfully avoided too). Each stage of the process model could thus have an accompanying emotion regulation strategy of equal legitimacy, but Gross and John (2003) only select two stages from the model. As a result, this influential scale (named the Emotion Regulation Questionnaire, rather than the more accurate Reappraisal and Suppression Questionnaire) has cascaded into thousands of research studies on emotion regulation that only address reappraisal and suppression, implicitly narrowing the scope of how emotion regulation more broadly is understood for many researchers.

Second, they argue that cognitive reappraisal is an “antecedent-focused” regulation strategy, that is, changing one’s appraisal of a situation before the emotion arises (Gross & John, 2003, p. 349). Yet this seems contradictory, as it presupposes that one has already participated in an emotion-eliciting situation, attended to an emotion-eliciting aspect, and appraised the situation in a presumably unsatisfactory or upsetting way, causing one to then reappraise the situation in a less upsetting way. Phenomenologically it is also hard to imagine how someone would be motivated to revise their appraisals without having experienced any negative emotion. Therefore, on conceptual grounds, reappraisal must be a response-focused, rather than a truly antecedent, strategy.

What might be the consequences of this type of conceptual confusion? If reappraisal is a response-focused strategy rather than an anticipatory strategy, it may turn out to have similar difficulties as suppression. For instance, suppression is cited as being maladaptive because, for one, it is inauthentic, creating an incongruency between what one is feeling (inner) versus expressing (outer). Yet, changing one’s mind solely to adapt is no less inauthentic. Presumably, a person’s revised appraisal is not what they truly think, it is only an adaptive set of beliefs, much like a teenager who changes their clothing style and music taste to fit in better with a more popular friend group, and who is then labeled “inauthentic” or “phony” by their previous friends.

The third conceptual difficulty with this scale is the degree to which the authors believe emotion regulation is a conscious activity at all. They are ambivalent about the term emotion regulation “strategy,” “because it might be taken to imply that these emotion regulation processes are executed consciously” when in fact they can be executed automatically, “without much conscious awareness” (Gross & John, 2003, p. 348). It is important to note that Gross and John (2003) are not saying that emotion regulation is probably a conscious activity but there is a chance they can be executed unconsciously, but rather, they make the stronger claim that they do not even want to “imply” that emotion regulation strategies are conscious activities (p. 348). Some of the items in the scale, particularly in the suppression subscale, require participants to self-report on a process that Gross and John theorize may be automatic. For instance, item eight reads: “When I am feeling negative emotions, I make sure not to express them” (Gross & John, 2003, p. 351). “Making sure” seems a conscious deliberate activity, as when someone tells you to “make sure to turn off the lights.” As such, it does not make sense to ask a person to “make sure” to do something of which they may not be consciously aware, as in “make sure to have a dream about a horse tonight.” One may or may not be able to control a dream and may not remember even if one did have the dream. If so, it does not make sense to expect an accurate self-report of the degree to which one “makes sure” to carry out a process which is not fully available to conscious awareness.

By analogy, the body under normal circumstances automatically adjusts its heartrate to facilitate equilibrium throughout its many interconnected systems. But some individuals also possess the ability to deliberately modulate their heartrate to some degree. Would it make sense to ask such a person on a Likert scale if they “make sure” to lower their heart rate when it is too high? Certainly, their body normally does that automatically. On occasion, they might also consciously adjust their heartrate through breathing or some other technique. In either case, the same outcome is achieved, yet it is ambiguous the degree to which the individual “made sure” to regulate their heartrate. Similarly, if an ERQ respondent often suppresses their emotions automatically (that is, without deliberate or conscious action), but occasionally “makes sure” to suppress them (that is, deliberately), what should they answer on a self-report scale? In either case they have successfully suppressed their emotions, but to what degree have they “made sure” to suppress them?

Perhaps by using the language of “make sure,” the authors are intentionally limiting the scope of the question to only the processes that the person deliberately chooses. But this introduces a related difficulty: total scores on the suppression subscale would not conceptually refer to a person’s total levels of emotion suppression, but only the levels of emotion suppression they are consciously aware of and deliberately enact. The total levels of actual suppression would remain unknown, since the proportion of emotion suppression that is automatic would be hidden from their own view. Thus, it is not clear what correlations and regressions relating emotion suppression to other scales would mean—such calculations may be entirely misleading.

The fourth conceptual difficulty with this scale is that the process model that supports the scale items assumes emotion always emerges in the context of appraisals of specific situations, while the scale items themselves are abstract and context-independent (Campos et al., 2011). For instance, item one, measuring reappraisal, states: “I control my emotions by changing the way I think about the situation I’m in” (Gross & John, 2003, p. 351). Unfortunately, aggregating all the participant’s past situations into an overarching abstract statement may leave us with relatively little insight. Perhaps in certain types of situations, depending on one’s mood, current bodily state, and quality of attention, one tends to regulate emotions through reappraisal, and in an entirely different permutation of factors one tends to suppress. If there can be no coherent meaning of the concept of an emotion apart from how it operates in specific situations, in what sense can we ask participants about their emotions in the abstract? The authors themselves acknowledge this limitation and recommend diary and experience-sampling methods as a supplement in emotion regulation research (Gross & John, 2003). However, if these are conceptual limitations, they cannot be solved solely on empirical grounds—that is, by adding more and better data (Maraun, 1998). Campos et al. (2011) have helpfully framed this particular conceptual limitation in terms of intrapersonal versus relational emotion regulation. Gross and John (2003) clearly depict emotion regulation as an event within the self that a solitary individual can make avowals about. And yet, this view misses that emotion regulation is also a relational process, located in a broader interpersonal and sociocultural context (Campos et al., 2011).

To summarize, the ERQ has four main conceptual difficulties: (a) its artificially narrow scope, (b) its misplaced categorization of reappraisal as an antecedent emotion regulation strategy, (c) its equivocation of the automaticity of emotion regulation strategies, and (d) the context-independent intrapersonal nature of its items.

The conceptualization of emotion regulation in the Difficulties in Emotion Regulation Scale

The second emotion regulation self-report scale under consideration is the DERS (Gratz & Roemer, 2004). Rather than addressing discrete emotion regulation strategies, it contains six subscales, each focusing on what the authors consider a maladaptive aspect of emotion regulation. When totalled together, the authors assert that this represents the participant’s total level of difficulty with regulating their emotions. The six subscales are: (a) nonacceptance of emotional responses, (b) difficulties engaging in goal-directed behaviour, (c) impulse control difficulty, (d) lack of emotional awareness, (e) limited access to emotion regulation strategies, and (f) lack of emotional clarity (Gratz & Roemer, 2004). The DERS was published in the Journal of Psychopathology and Behavioral Assessment, and in contrast to the ERQ, measures its outcomes against more specifically clinical outcomes, such as domestic abuse and self-harm.

Similar to the ERQ, the authors of the DERS assert that emotion suppression is maladaptive. Indeed, they conceptually summarize their whole scale by arguing that a total DERS score—that is, total levels of difficulty in regulating emotion—should positively correlate with experiential avoidance of emotion, and negatively correlate with emotional expressivity (Gratz & Roemer, 2004). In other words, adaptive emotion regulation is synonymous with experiencing and expressing one’s emotions. Their reasoning rests on empirical studies, such as that of Gross and Levenson (1997) which shows that avoiding one’s “unwanted thoughts and feelings” is associated with increased physiological arousal, leading to greater emotion dysregulation (Gratz & Roemer, 2004, p. 42). They do cite other theoretical and empirical work supporting the idea that suppression is associated with “psychological disorders,” but these citations and the concepts they may contain are not explained (Gratz & Roemer, 2004, p. 42). Thus, despite the warnings of Maraun (1998) and Bennett and Hacker (2022a), there are no additional conceptual arguments provided for why suppression would theoretically be associated with psychological disorders or emotion dysregulation.

The DERS shares some conceptual difficulties with the ERQ, namely the context-independent nature of its questions, and the contradictory nature of asking participants about emotion regulation processes that the authors also avow can be unconscious—though the authors acknowledge these both as limitations (Gratz & Roemer, 2004). However, acknowledging these as limitations is not tantamount to recognizing them as metatheoretical commitments (Campos et al., 2011). In addition, the DERS contains its own unique conceptual challenges.

The first conceptual problem, related to asking direct questions about unconscious processes, is that various DERS items require participants to report on the degree to which they have clarity and awareness of the emotions they are experiencing, while simultaneously assuming emotional clarity on the remaining questions. For instance, item five states: “I have no idea how I am feeling” (Gratz & Roemer, 2004, p. 48). Supposing a participant answers this question with the highest available score (a score of five indicating they “almost always” have “no idea” how they are feeling), it is uncertain how such a participant might understand 27 of the 36 scale items which all begin with the statement, “When I’m upset . . .” which of course presupposes knowing that one is upset (Gratz & Roemer, 2004, p. 44). If a participant has very high scores on the subscales lack of emotional awareness and lack of emotional clarity, to what degree can we be confident of such participant’s total scores correlating with experiential avoidance and emotional expression (the author’s stated criteria of validity)? If we must take seriously a participant’s avowals that they are not aware of their emotions, then how do their remaining avowals make sense—of their emotional experience, emotional expression, and so on?

The second conceptual difficulty with the DERS is that it is theoretically opposed to emotion suppression, yet many of its items could be negatively correlated with high scores of emotion suppression, contrary to the stated intentions of its authors. Unlike the authors of the ERQ, Gratz and Roemer (2004) do not restrict suppression to outward expressiveness. On the contrary, they take for granted that one can indeed attenuate the intensity and duration of the experience of emotional “arousal” itself (Gratz & Roemer, 2004, p. 42). They state that adaptive emotion regulation involves attenuating unhelpful arousal in order to maintain control over one’s impulses (e.g., “When I’m upset, I feel like I can remain in control of my behaviors” reverse-coded) and to achieve one’s goals (Gratz & Roemer, 2004, p. 48). The goals outlined in the scale include things like “concentrating,” “focusing,” and “getting work done” (Gratz & Roemer, 2004, p. 48). One could note how similar this conceptualization of emotion “regulation” is to that of a cybernetic control circuit such as a thermostat. Recall that a thermostat has feedback inputs of temperature and then regulates itself to enact heating or cooling processes toward instrumental goals of attaining a set temperature. Likewise, in this conception of emotion regulation, one has “arousal” inputs, and in place of heating and cooling processes, one has emotion regulation strategies in order to achieve highly instrumental—even neoliberal—goals of productivity, focus, and getting work done.

The adaptive value of controlling one’s emotions in order to accomplish one’s goals seems obvious. However, it is less obvious how this relates to their overall conception of healthy emotion regulation as low emotional avoidance and high emotional expressivity. Surely a participant who is emotionally avoidant could still excel at meeting their goals, especially instrumental, work-focused goals (indeed, every item in the goals subscale is related to productivity, focus, and work). For example, Theodore Roosevelt was admired for his ability to suppress grief over the death of his father and excel in his school exams (Morris, 1979/2001). Thus, DERS participants could score very low on each of the items in the difficulties engaging in goal-directed behaviour subscale—no difficulty with productivity when upset—which would lower their overall DERS score (meaning more adaptive emotion regulation). In other words, some participants could suppress their emotions and still be productive, very much like Roosevelt. A related conceptual problem is that emotion suppression is often employed precisely to meet one’s goals, whether this is adaptive in the long term or not. For example, a male adolescent might suppress expressions of sadness during a movie in order to meet the goal of not being teased by his friends. Indeed, attachment theory conceptualizes emotion regulation in exactly these terms: one has learned certain emotion regulation styles in order to adapt to one’s early environment. For instance, a child may learn to suppress their emotions because they have learned that expressing their distress leads to greater disappointment and rejection (Mikulincer & Shaver, 2019). Such examples show that suppressing emotion and meeting one’s goals are not mutually exclusive.

Having goals consisting of concentrating, focusing, and getting work done also appears to be an arbitrary narrowing of the broad range of goals that constitute the human experience. Yet, this productivity-focused operationalization of “goals” is fitting with the mechanical definition and use of “regulate” mentioned above. It is normally mechanisms, systems, parts of systems, and formal groups amenable to legislative control that are said to be “regulated” in everyday use of the word. Concentration and productivity surely represent only a portion of the broad range of human goals that involve our emotional experience, and perhaps not even the most important ones when one considers the lifelong goals of attaining satisfying relationships and achieving a sense of purpose. Perhaps this is yet one more expected outcome of reducing a previously moral concept—the passions—into a scientific one (Dixon, 2003). Without a sense of value-based appraisal or moral meaning, emotion as a concept becomes merely instrumental.

Some implications for emotion regulation research

Both the ERQ and DERS purport to measure the same construct: emotion regulation. That these scales both share some conceptual confusions or arbitrary categories has been established already, but there arises an even deeper problem when one attempts to compare their contrasting conceptualizations of emotion regulation. The ERQ operationalizes emotion regulation as the presence of adaptive emotion regulation strategies (namely, cognitive reappraisal). One who regulates their emotions well in this scheme focuses on reappraising their valuations of specific situations, rather than suppressing their reactions to upsetting valuations of situations. The DERS operationalizes emotion regulation as something closer to emotional skills that are entirely response focused: those who can regulate their emotions well are those who can accept their negative emotions, stay productive despite negative emotions, control their impulses when upset, become aware of their current emotions, and have confidence they can find a way to calm down.

That the DERS focuses entirely on response-focused emotion regulation skills may seem like an innocuous difference from the ERQ, yet according to Gross and John (2003), the difference between antecedent and response-focused strategies is “central to [their] theory” (Gross & John, 2003, p. 349). Why? Because any response-focused strategy (such as suppression) requires tremendous amounts of energy to modulate an emotion that has already fully activated all the emotion-related bodily and cognitive systems, leading to maladaptive outcomes (Gross & John, 2003). Suppression is simply one such strategy, though presumably, based on Gross and John’s comments, all response-focused strategies would share the same dilemma. Given that no item on the DERS addresses any antecedent-focused strategies, it would seem that, based on Gross and John’s conceptualization, no DERS score could tell us anything about the positive emotion regulation capabilities of a person. Thus, while the authors of the ERQ and DERS agree that emotion suppression is maladaptive, their two measures are conceptually incompatible with one another on nearly every other level.

We hope to have shown that psychological research on emotion regulation has been influenced by scientific reductionism at several stops along the way, beginning with Brown’s redefinition of emotion from a thick moral concept into a bodily perturbation, to the abstract and context-free nature of self-report questionnaires, to the implicitly mechanistic vision of the person offered in those same questionnaires. While isolation and focus has largely been a boon to the natural sciences and some aspects of psychology, it seems to be detracting from conceptual clarity in the psychological study of emotion regulation. A risk is that while one might know, for example, exactly what it means for a circuit to be regulated, and even to construct a circuit diagram to represent its regulation, it is less clear what it means for an emotion to be regulated—and unclear what additional insight we gain into emotional experience by describing it in terms of regulation.

If we can temporarily bracket the reductionistic and scientific vision of the emotions that has developed since Thomas Brown in the 19^th century, what kind of conception of the emotions might we have? Bennett and Hacker (2022a), for their part, offer an interpretation: emotions are valuations of our appraisal of situations, which then motivates subsequent behaviour and can only be understood in the context of our whole person and the life we find ourselves in. An example they provide is that of Edmond Dantès in The Count of Monte Cristo, whose emotion of hatred is manifest mainly in a long-term motivation to carefully scheme revenge on those who betrayed him. He knows no abstract hatred, but only the particular cruelty of the Château d’If where he is imprisoned. What might it mean then for Dantès, in this broader contextual scheme, to regulate his emotion of hatred? His hatred is resulting from a value-based judgment of a particular situation, that of his friends betraying him. The object of his hatred is thus his friends, and what motivates him is presumably the belief that their actions were wrong, and that justice demands he make them suffer. These beliefs are motivating his actions in crafting a long-term plan for revenge. To regulate this complex web of belief and behaviour would require many things, perhaps involving forgiveness, or a revised belief (à la Stoicism) that it does not matter what his friends did since it is no longer in his control. Given the historical language used for “constraining” and “taming” the passions, clearly there is historical precedence for the notion of altering one’s emotions. It just may not be possible to do so apart from the ancient humanistic roots of meaning, value, and virtue. For Dantès to regulate his “emotions” in this fuller sense would involve becoming a more virtuous person at a deeper level, one who is not hateful. This is no easy task, and certainly one that is not as straightforward as regulating a mechanical system.

As noted, a more contextual rather than intrapsychic and acontextual construal of emotional regulation is central to Campos et al.’s (2011) metatheoretical critique of scales such as the ERQ and DERS. Campos et al. rightfully point out that a problem with these measures is that they conceptualize emotion regulation as an internal event that can be studied simply by asking participants to self-report on their internal states. One should ask how the conceptual complexity of everyday emotional states such as Dantès’ and countless others could be reduced to a simple intrapersonal calculus. To acknowledge this as a mere limitation of such scales, as both sets of authors do (Gratz & Roemer, 2004; Gross & John, 2003), appears to also skate over the conceptual nexus of these issues (Bennett & Hacker, 2022a; Maraun, 1998). As Campos et al. (2011) note, the issue that this raises is epistemological and “centers on the implicit and widespread belief that studying emotion regulation in simpler, nonsocial contexts will yield findings that will eventually elucidate emotion regulation in more complex interactional settings” (p. 34). For the reasons that we have reviewed in the present article and others that we do not have the space to review, we believe that Campos et al. (2011) are correct to “doubt this assumption” (p. 34). As such, we believe that these authors are correct to urge a relational approach to both conceptualizing and operationalizing emotion regulation and invite interested researchers to consult Campos et al. (2011, pp. 31–34) for suggestions on how they might do so.

Conclusion

It seems impossible for psychologists to scientifically study human behaviour without using everyday folk language. Brown himself did this in taking the relatively new folk term “emotion” and applying it to the study of the passions and affections (Dixon, 2012). As reflexive subject-objects, it also would be impossible for psychologists to discover and study some aspect of human behaviour that was unknown to humanity, and thus without any known linguistic referent—such is the paradox of psychological science (Morawski, 2014). Yet the multifarious nature of our folk concepts may elude scientific operationalization by virtue of its diffuse embeddedness in the human form of life (Maraun, 1998). For instance, to speak of “measuring” anxiety connotes accuracy. Yet, the dizzying complexity and subtlety of a particular person’s anxiety from moment to moment, or day to day (not to mention their conscious awareness and recollection of their complex anxiety to be relayed on a self-report questionnaire), makes psychological measurement a monumental task (Maraun, 1998). The most expedient recourse for psychologists then becomes operationalization by reduction—“emotion” becomes a matter of context-independent items on a questionnaire.

Emotion regulation has become a concept of central importance, with many psychologists arguing it is a key transdiagnostic factor of psychopathology (e.g., Fernandez et al., 2016). Given that the DERS has clinical populations explicitly in mind, the importance of clarifying the conceptual nature of emotion regulation is of ever-increasing importance, as clinical diagnoses and therapeutic-medical decisions are made on the basis of clients’ total scores on emotion regulation questionnaires.

Yet naming the conceptual problems that arise from these scales does not undermine that these scales have real goods to offer. They both offer empirical evidence that their measures strongly correlate with important outcomes in the realms of self-harm, interpersonal functioning, well-being, and so on. Clearly there is something that differs between those with high versus low scores on these scales. However, if two of the leading emotion regulation self-report scales have conceptual limitations within them, and even deeper, fundamental conceptual incompatibilities between them, it may be that we do not yet know what that something is which the empirical evidence nonetheless assures us is there. In addition, as Bennett and Hacker (2022a) remind us, it is a fallacy to mistake the part for the whole. So too, one’s conscious responses to particular items on a scale may indeed represent one important part of our emotional life, but it would be a mistake to confuse it for the whole.

Gross and John (2003) recognize that emotion always arises from our valuations of specific situations, which is to be commended. Gratz and Roemer’s (2004) conception takes a more situation-agnostic, skills-based approach. However, it may be that both scales practically fail to pay sufficient attention to the way emotion is embedded in the “hurly burly” of life, with its near-infinite permutation of differing situations and valuations (Wittgenstein, 1980/1998, p. 108). Although researchers should heed Gross and John’s (2003) own recommendations to pursue diary, experience-sampling, and related methodologies for measuring emotion regulation, the work of Campos et al. (2011) serves as a reminder that one can ground these research strategies in either an intrapersonal or relational frame. As such, it is also important that researchers pay more careful attention to how individuals experience and respond to their emotions in the everyday stuff of life—an endeavour that requires conceptual precision.