Psychology: Where history,culture,and biology meet

Stability and instability

As noted above, many biological processes appear to be characterized by a millennial stability, otherwise we would not have two eyes, ears, legs, arms, and so forth, over numerous millennia. Similarly, life forms on this planet remain biochemically composed of things like carbon, hydrogen, and oxygen, no matter what we say or do. Unlike the human social universe, we cannot talk these things out of existence. In addition, even though certain life forms, such as viruses, may evolve rapidly, they still appear reliant for their replication on biological processes that may be of considerable duration, such as those involved in lytic or lysogenic cycles.

These processes supporting biological stability are often dynamic. For instance, DNA is not intrinsically stable, since its integrity is continuously maintained by a panoply of proteins that forestall and repair copying mistakes (Keller, 2000). Genomic constancy therefore seems to be the product of extraordinary biological dynamism, such that stability does not represent a simple opposition to instability. Instead, dynamic biological stability is often the outcome of processes that routinely involve elements of instability, such that stable “persistence is at every moment dependent on activity” (Dupré, 2020, p. 99). Nevertheless, given that these processes can remain dynamically stable over evolutionary timescales of several millennia, life scientists have tended to adopt an epistemological approach which presumes that conditional replicability and regularity is possible through experimentation.

Yet the same is far less obviously the case with social processes, and it is therefore not entirely surprising that Gergen (1976) made his “radical” assertion that “social theory is always developed under unique historical circumstances, and the circumstances are in continuous motion” (p. 377). It becomes less likely that observations at one time and place will occur at another because the sociocultural universe can be slippery, mobile, and changeable.

In spite of the historical resistance to this relativizing assertion (Billig, 2008; Ellis et al., 2017), subsequent developments in the philosophy of science have been kind to Gergen’s thesis. For instance, Gergen’s (1973) attention to “enlightenment effects” received strong support from the near identical concept of looping promoted by Hacking (1999). As Gergen stressed, by educating and enlightening the population about social patterns, “the grounds are established for altering these patterns” (Wang, 2016, p. 566) and, as Hacking (1999) later elaborated, the result of this process is that the objects of social science often constitute a “moving target,” reflecting the double hermeneutic of social life (Giddens, 1984; cf. Danziger, 1997). For example, respondents may be more likely to report that stress influences their health, or engage in stress management practices, as a consequence of the plethora of popular media articles and programmes that have reported psychological research related to stress over several decades (Kuokkanen et al., 2020; Newton, 1995). Psychological research does not therefore constitute the independent assessment of a stable reality since people are not just informed by psychological concepts, but also actively respond to them and may change their behaviour, especially where they are heavily mediatized. As noted by the social historian Thomas Dixon (2003, 2012), psychological concepts therefore have a reflexive relationship with our lives, performatively shaping and colouring, as well as explaining, them. Following Gergen, Hacking, and Dixon, social and psychological reality becomes slippery, in motion and inherently less than stable.

Although Gergen’s historicist thesis met with considerable early resistance, some critics did acknowledge that changing historical circumstances may make it difficult to formulate timeless theories. Nevertheless, they still tended to defend the universalism of psychological processes. For example, Manis (1975) responded to Gergen (1973) by suggesting that:

The processes underlying social behavior are probably relatively stable, although they operate on an endless variety of social contents as we vary the time and place of our investigations. For example, the process of fear conditioning may follow a fairly uniform course [emphasis added], even though the particular things we fear depend largely upon changing historical circumstances. (p. 454)

Yet later psychological, historical, and anthropological research questions the universality of emotional processes. Although there may be stable similarities in the neuroendocrinal response to emotional arousal, subsequent research suggests that how we appraise, learn, and feel emotions such as fear varies substantially between cultures: ethnographic anthropology (e.g., Lutz, 1988), historical anthropology (e.g., Reddy, 1997), and experimental (e.g., Gendron et al., 2014) and quantitative colexification analysis (Jackson et al., 2019) all suggest strong cultural variability in the processual conditioning of emotions such as fear, anger, sadness, or love, such that there remains a “wide variation in emotion semantics across 20 of the world’s language families” (Jackson et al., 2019, p. 1522). This research also reinforces the arguments of writers such as Kurt Danziger, Kenneth Gergen, and Gustav Jahoda, who have long criticized the cultural specificity of psychology, particularly the dominance of western conceptions of humanity (Brock, 2006). From this perspective, there is little evidence of the uniform “process of fear conditioning” described by Manis (1975), since the conditioning processes through which people appraise, learn, and label emotions like fear do not appear to be uniform across time and space.

In sum, following Gergen (1973, 1976), the epistemological ability to generalize across time and space, which is tendentiously applicable to biological processes, cannot be so easily utilized in sociocultural psychology. This epistemological distinction also received support from the philosophical argument of Hacking (1999), particularly his distinction between the indifferent kinds studied within natural sciences and the interactive kinds witnessed in social sciences (cf. Collingwood, 1960; Connelly & Costall, 2000). As noted, Hacking (1999) saw social science as the study of “a moving target” because our studies interact with the social objects we are trying to observe. However, he suggested that the same is not true of natural kinds, such as quarks: “Quarks . . . do not form an interactive kind; the idea of the quark does not interact with quarks. Quarks are not aware that they are quarks and are not altered simply by being classified as quarks” (Hacking, 1999, p. 32). In this fashion, “the targets of the natural sciences are stationary” (Hacking, 1999, p. 108). In contrast, social objects are interactive, on the move, and, in consequence, slippery and unstable: they remain continuously open to change as people interact with them.

Yet, as Collins and Bunn (2016) note, drawing on Hacking:

It would seem that psychology has for much of its history assumed that it is in the business of discovering and examining indifferent kinds when it is more appropriate to describe it as contributing to the construction of interactive kinds [emphasis added]. (p. 94)

In asserting that natural kinds such as quarks are “indifferent,” Hacking (1999) was cognizant of debates within science and technology studies and the sociology of scientific knowledge. Yet, in spite of the constructionist logic of much of science and technology studies and the sociology of scientific knowledge, Hacking argued that natural objects are often indifferent to our musings and adventures. As Danziger (1997) put it, unlike “psychological objects,” “natural objects . . . have no capacity for self-reference” (p. 190).

Nevertheless, it might be argued, following varieties of antidualist argument (e.g., actor–network, more-than-human, performative materiality, etc.), that the human epistemological representation of material objects forms part of the ontological constitution of the objects themselves. Yet, as I have argued elsewhere (Newton, 2007), there remains a need to question totalizing tendencies which suggest that the technolinguistic chattering of one (human) animal fundamentally alters the nature of natural kinds. Take the famous wave–particle duality of natural light (Barad, 2007): before it could decide whether to behave like a wave or a particle at the quantum level, did natural light have to wait for billions of years for a biped animal to measure it? Was it the arrival of humanity that caused natural light to vary between the two? In other words, is it rather colonial of social scientists to seek to imperially capture everything—the natural as well as the social—within a constructionist framework? As Barad (2007) argues, we need to avoid the “reification of . . . categories of the ‘social’ and the privileging of them as explanatory factors over the ‘natural’” (p. 86).

Biology and “nature”: Across the great divide?

In spite of the overall cogency of Hacking’s proposition, I am not persuaded by all aspects of his conjecture. As noted, Hacking’s (1999) argument that “the targets of the natural sciences are stationary” and “the targets of the social sciences are on the move” (p. 108) suggests that the former are stable whereas the latter are mobile, and always potentially unstable. However, it is difficult to maintain this binary distinction between natural processes characterized by stability over time, and sociocultural processes associated with flux and instability. First, as noted above, biological stability reflects dynamic processes, as illustrated by the continuous processual repair of copying errors that maintains DNA stability. Second, not all natural processes operate like Hacking’s quarks, unaffected by human knowledge of them. For instance, microbial viruses are aware of human knowledge to the extent that they may quickly evolve to evade human attempts to control them through vaccines (as may be the case with COVID-19 variants). Microbial viruses can rapidly respond to, and interact with, our knowledge about them, and therefore also constitute a moving target. Contra Gergen and Hacking, such natural kinds can be both interactive and unstable.

This observation reinforces the need to explain what is particular to human sociality that makes our social world interactive and unstable. As discussed next, by focusing on the relation between human language and social change, the work of Elias presents one answer to this question.

Language

In the case of human societies and human languages the malleability, the ability to change, has become so great that extensive changes of social life . . . can occur within the biological lifetime of the same species. (Elias, 1991, p. 29)

My principal interest in human language, and the work of Elias, derives from the fact that our exceptional linguistic flair enables a far more rapid sociocultural change than that observed with other species. This is not to suggest that language and technology are solely agents of instability, since they constitute central processes through which we both make and remake our sociotechnical universe. Nevertheless, these key human facets enable us to (inter)generationally refashion our sociocultural world in a manner that is unavailable to other species, as reflected in the relatively rapid transitions in the West from tribal to feudal, mercantile, industrial, financial, and global capitalist society.

In asserting that human linguistic exceptionality is central to this sociocultural fluidity, I do not mean to imply that we do not have most things in common with our animal cousins. Yet other animals lack our particular linguistic skills and the ability to learn them. Although our close primates, such as chimpanzees, have some technological skills and have been taught to understand a limited vocabulary, it remains the case that we are the only primate characterized by complex language and technology. As Horigan (1988) notes, “ape language experiments exhibit a marked degree of anthropocentrism” (p. 100), and sometimes resemble a desperate attempt to impose a symbolic system on our closest primates (Chomsky, 2007/2008) because they are “at best limited-vocal learners” (Petkov & Wilson, 2012, p. 2077). Quite apart from the resistance of other primates to such anthropocentric ambition, it remains the case that no other primate uses symbolic and technological systems to enable the kind of (inter)generational sociocultural fluidity witnessed with humanity.

This is important to note because it is not necessarily the dominant emphasis in much of the existing work on the evolution of language—whether it focuses, for example, on language analysis (e.g., Jackson et al., 2022), semiotics and signification (Lazzarato, 2014), ethology (e.g., Hauser et al., 2002), extinct hominids such as Neanderthals and Denisovans (e.g., Conde-Valverde et al., 2021), or genomics such as FOXP2 (e.g., Atkinson et al., 2018)—or the particularity of human cortico-motor neurons (Fitch, 2018). What I find surprising about this research tradition is that comparatively less attention has been paid to the key sociocultural significance of language—namely, its centrality to relatively rapid sociocultural change—even though cultural psychologists have been directly concerned with the effect of language on sociocultural behaviour (e.g., Imai et al., 2016). For instance, the 522-page Routledge Handbook of Language and Culture (Sharifian, 2015) provides detailed analyses of the interrelation between linguistic change and culture but fails to consider the core significance of human language for the plasticity of our species. On the one hand, it has not gone unnoticed that human language affords “a unique kind of open-ended adaptability” (Kirby et al., 2008, p. 10681) that allows for cultural innovation (Maynard Smith & Szathmáry, 1995; cf. Hauser et al., 2002). On the other hand, comparatively less emphasis has been placed on the fact that, compared to other species, our linguistic skill has facilitated rapid sociocultural change. Without language and other forms of symbolization (e.g., mathematics), people could not communicate arguments, ideas, or techniques to others either within generations or, critically, between generations. For instance, the historical evolution of universities (de Ridder-Symoens, 1991) can be seen as an important means for the intergenerational symbolic communication of linguistic and technical repositories of knowledge, thereby enabling sociocultural change and development at a much faster pace than the relatively slow biological evolution observed with many species. These symbolic repositories enable subsequent generations to learn, innovate, and change in a species-atypical manner. As Elias (1991) notes, it is because of language that “in the case of human societies a great social change such as that from tribe to empire can occur without any biological change [emphasis added]” (pp. 31–32).

In noting this sociocultural significance of language, my interest is meso/macroanalytical, rather than the microanalytical focus that is more apparent within linguistic research. In keeping with this orientation, I do not wish to suggest that language represents a simple tool that people use to straightforwardly effect change. Instead, following Elias (1991), language is seen as representing a collective experience through which people haphazardly communicate and act (with collective consequences that may be different, or contrary, to those intended by any one individual or group). This interpretation is in keeping with Elias’s fundamentally networked conceptualization of change and agency. Predating approaches such as actor–network theory, Elias continuously questioned the supposed sovereign independence of the individual actor on the grounds that their actions are frequently conditioned by the wider networks in which they operate—as the actor–network theorist John Law (1994) noted, “the explanatory attitude of [actor–network] writers is not so different from that of Norbert Elias” (p. 113).

In consequence, rather than portraying language as a purposive instrumentalist tool, Elias (1991) described it as a key social development that allows for the possibility of collective sociocultural change distinct from other species, which, in the absence of complex communication skills, tend to be more reliant on biological, rather than sociocultural, evolution. A species that has language has this possibility of a built-in plasticity because languages enable collective sociocultural change without change in the human genome. Liquid language remains an inherently fluid phenomenon, both changing our social universe and continuously open to change itself.

Elias’s work on language speaks to this human social plasticity as well as the limitations in Hacking’s (1999) analysis. Although Hacking’s account is persuasive, it can appear foreshortened. It fails to fully explain what it is about human subjectivity that particularly differentiates it from that of a quark, another primate, or other life forms. On the one hand, Hacking (2007) presents a detailed consideration of systems of knowledge through his post-Foucauldian attention to “engines of discovery,” extending what “we learned from Michel Foucault [about] the capital role of knowledge” (p. 297). On the other hand, although this argument further elucidates the interrelation between systems of knowledge and “moving targets,” it does not, of itself, address the question of how, and why, human social interactivity, fluidity, and plasticity differs from other life forms. Yet, by drawing on Elias, we can see that human language is an essential ingredient of the instability of Hacking’s “interactive kinds,” differentiating us from other current species. It is the capacity for language and symbolization that also allows human beings to immediately reflect on, and react to, the looping of social science research concepts and categorizations, so that the latter may come to constitute “enlightenment effects” and “moving targets” (Gergen, 1973; Hacking, 1999).

Technolinguistics

The preceding discussion is not meant to suggest that Elias’s (1991) argument is without limitations. In particular, his overall thesis is circumscribed to the extent that language is not the only predominant source of sociocultural changeability, as human beings have also transformed their worlds through a remarkable range of technology. Although Elias attends to the history of technology, he does not place a strong emphasis on it as a possible source of social change (Burkitt, 1999). Yet human flexibility also derives from the fact that we are biologically equipped to develop increasingly complex technologies (Newton, 2007). In making this statement, I do not wish to invoke technological determinism, or imply that tools are something that individuals use to straightforwardly effect change, devoid of any sociotechnical network (Latour, 1999). Instead, as with the discussion of language above, I merely wish to posit technological networks as key social processes that have collectively allowed humans a unique technological competence. Although our primate cousins evince some sophistication in tool use, this does not match the complex technologies that were already apparent among Neolithic humanity (Robb, 2013), even though this skill may have once been shared with extinct hominids such as Neanderthals (Hardy et al., 2020) and Denisovans (Gibbons, 2011).

At the same time, the combination of human language and technology allows for a far greater degree of social reasoning, reflection, and intergenerational communication than exists amongst other living primates. Indeed, there is a long-standing argument that tools and signs should not be seen as separate things (Ilyenkov, 1977), as well as a related tradition that stresses how language and technology may have developed in tandem (Washburn, 1960). As Elias (1991) argued:

could human knowledge have been extended, as it was for example from the making of hunting-axes to that of computers, from the perception of the sun as . . . the vehicle of a God to that as a kind of helium burning furnace, without . . . languages? (p. 104)

There are also archaeological indications that the development of language may have been critical to greater human technological refinement and finesse (Morgan et al., 2015). As Mithen (2019) argues, language needed to evolve alongside human technological development, such as with the Neolithic transition from hunter-gatherer to farming societies, where there was a need for new concepts of property and ownership, as well as a managerial lexicon relating to issues such as the watering, weeding, and storing of crops (cf. Ferretti et al., 2018).

In sum, technology and language appear closely interrelated, and it seems reasonable to characterize Homo sapiens as a quintessentially technolinguistic primate. Furthermore, in terms of temporality, I contend that it is this combined technolinguistic plasticity that gave humans the collective ability to detect, and use, stable processes in the natural world, such as those surrounding the magnetic compass, paper, printing, and gunpowder (Needham, 1956). Humans learned that these technologies would produce the same result again and again due to the dynamic processual stabilities that they incorporated, such as the explosive gases produced by gunpowder or the curative effects of penicillin. Similarly, our current concern to control COVID-19 is based on stimulating dynamic biological stabilities within our immune system. In this sense, the history of science and technology addresses the relation between differing temporalities, as the human technolinguistic capacity for change, instability, and adaptation became interwoven with the dynamic processual stability of many biological and physical processes. Furthermore, this exceptional technolinguistic plasticity to control nature’s dynamic stabilities has come to define an anthropocentric era, where the key question for other life forms is whether they can survive in a “human universe” (Monastersky, 2014), especially given climate change and habitat depletion.

The biosocial: The example of cultural neuroscience

As argued above, the consequence of human technolinguistic flair is a sociocultural plasticity that may result in differences in the epistemological assumptions that can be reliably applied to the sociocultural and biological provinces of psychology. At the same time, many fields of psychology implicitly traverse both these provinces, such as health, clinical, cognitive, forensic, sport, developmental, comparative, rehabilitation, addiction, psychodynamic, environmental, and psychophysiology—an observation that suggests a need to address epistemological pluralism.

The field of cultural neuroscience is symbolic of this need since it reveals how even seemingly more biological fields are also inevitably sociocultural. For instance, cultural neuroscientists must address not only the likelihood that neuroplasticity may vary between cultures (e.g., Chiao et al., 2013; Murata et al., 2013), but also the possibility that it is variable across time, since cultures may change relatively rapidly (Greenfield, 2017; Rosa, 2013), as illustrated by Chen et al.’s (2005) longitudinal study of rapid sociocultural change in Chinese culture (consequent on its transition towards a market economy). On the one hand, critical psychologists may remain understandably doubtful of cultural neuroscience because of its reductionism (Martínez Mateo et al., 2012) and neuro-essentialism (Racine et al., 2005; cf. Haslam et al., 2022; Miller, 2010), and so forth. On the other hand, the field remains interesting since it symbolizes the way that all fields of psychology, including those that may appear more biologically defined, need to account not only for sociocultural processes, but also for the epistemological implications of this requirement. Most significantly, the requirement to add cultural to neuroscience not only challenges the Cartesian divorce between cultural and neural learning, but also questions whether the same epistemological assumption can be applied to the study of neural and cultural processes. For instance, it is not only that neuro-reductionism cannot easily be applied to sociocultural processes (e.g., west brain–east brain formulas; Denkhaus & Bös, 2012), but also that natural science assumptions of longue durée stability become far more questionable in the sociocultural arena.

In sum, the field crystalizes the contrast between the millennial stability of neural processes and the changeable fluidity of the sociocultural. In particular, when studying cultural change, we are likely to be looking at Hacking’s “moving targets,” and such instability may weaken or invalidate validity and measurement at a later date, with consequent difficulties in the replication of earlier studies (see below). As a result, the kind of epistemological assumptions of replicability that can be more readily entertained with neural functioning may be called into question when it comes to studying cultural and historical influences on such functioning.

Using replication to determine biological and sociocultural stability?

In spite of the foregoing argument, it might be suggested that replication studies could provide a vehicle to demonstrate the stability of sociocultural psychological processes, as Gergen (1973) suggested in his seminal article. As a consequence, if replication studies evidenced sociocultural stability across time and space, similar assumptions about stability might then be entertained to those that apply with the biological. In so doing, researchers might employ a comparable epistemology and research design across social and biological arenas of psychology, thereby helping to justify disciplinary unity, and warranting comparison with natural science. In this manner, as the “cornerstone of science” (Simons, 2014, p. 76), replication studies might rescue psychology from questions about its scientific legitimacy by validating stable, and consequently replicable, psychological processes.

Yet, rather than confirming the scientific stability of sociocultural research findings, replication studies have appeared as a primary vehicle in the problematization of sociocultural stability, given the much-reported difficulties in effecting replication (Open Science Collaboration, 2012, 2015). As Gergen (1976) noted, “early studies may not be replicated because the historical climate . . . has shifted” (p. 378). Furthermore, the possible reasons for such historical change are so varied that we need to have a clear idea of the differences in historical, demographic, educational, technological, and cultural context that might affect results across test populations (Irvine, 2021).

For some of those who argued that “there is no such thing as an exact replication” (Schmidt, 2009, p. 92), conceptual replication offered an approach that promised to accommodate sociocultural change. Yet there may be a conceptual problem with the notion of conceptual replication. First, it can be difficult to reliably specify an updated experimental procedure because the theoretical premises of the original study are often constructed too loosely (Earp & Trafimow, 2015). Second, it is hard to be certain whether any failure to replicate is a consequence of the updated procedure being unable to capture the processes assessed in the original study, such that any failure may be a consequence of differences in method rather than weakness of the original effect (Doyen et al., 2014). Third, there is still room for doubt even when a successful replication looks apparent, since it is possible that the updated procedures may have introduced a new dynamic. In effect, the conceptual replication may involve a subtle blurring of experimental method so that the successful replication is actually assessing a different construct or process.

Furthermore, underlying all these replication constraints are difficulties that are more likely to be observed in social science. On the one hand, no science can be described as remotely pure in the sense that it is not (partly) socially constructed, free from questionable customs and research practices, or commercial, normative, and political pressures. On the other hand, natural science epistemologies and methods suit some terrains more than others. For example, as Cesario (2014) observes, “we know that physics generates objective knowledge precisely because we will both get the same answer if you measure the rate of the ball dropped from the tower and I measure it as well” (p. 41). Although such Galilean regularities are not unconditional (Cartwright, 1999), it is still the case that mechanics is a powerful tool for prediction. Yet this relatively straightforward replication is often impossible in social science due to the likelihood that the experimental environment will “vary in myriad . . . ways, however small” (Wiggins & Chrisopherson, 2019, p. 213), such as occasioned by variance in the experimental lighting deployed (e.g., de Kort & Veitch, 2014). Furthermore, the question remains as to how we disentangle these possible effects from other known issues in conceptual replication, including concerns over confirmation and publication bias (Pashler & Harris, 2012), and the possibility of “concept creep” through the redefinition of psychological concepts (Haslam et al., 2020). In consequence, following the Heracleitean aphorism, it could be that sociocultural experiments can never “enter the same river twice” (Morawski, 2019), and in many areas of social science, replications may rarely resemble those that Galileo is supposed to have observed with falling bodies.

These difficulties appear to be compounded if we follow Sullivan’s (2020) argument that psychology has favoured a Galilean approach to science, “bringing social psychology in line with the worldview of contemporary natural science” (p. 80). Although this may be appropriate in some natural sciences, Sullivan suggests that it is inappropriate to psychology as it is not governed by “Galilean regularities,” whereby experimental configuration X habitually leads to outcome Y. This argument appears yet more convincing if we augment it by applying Argyris and Schön’s (1974) well-known distinction: although psychology may in practice resemble Lewin’s (1931) critique of Aristotelian science as being implicitly value-laden, and so forth, the espoused ideal of the mainstream discipline appears to be predominantly that of a Galilean science. Furthermore, as Bruner (1990) and Sullivan (2020) suggest, this espoused Galilean project may have promoted a “dual distrust of culture and history” (Sullivan, 2020, p. 80), encouraged by Lewin’s (1931) “Galilean” desire of “getting rid of the historical bent” (p. 165).

In addition, a Galilean aspiration may partly explain why there has been a lengthy and continuing denial of differences between psychology and natural science, seemingly motivated by a desire to maintain psychology’s status as a proper science (e.g., Schlenker, 1974; Van Bavel et al., 2016; Zwaan et al., 2018). For instance, Van Bavel et al. (2016) defend an epistemological similarity between psychology and natural science by referring to the difficulties experienced in replicating Isaac Newton’s light experiments, consequent on his contemporaries’ use of bad prisms. However, as these authors immediately go on to note, such replication constraints were soon resolved by using better prisms. Yet, contrary to Van Bavel et al. (2016), I am not convinced that such straightforward resolution is available within psychology since measurement cannot be precisely crafted in the manner available with prisms and because of the considerable discrepancy between human behaviour and the constancy of the speed of light.

A similar problem arises with Zwaan et al.’s (2018) detailed defence of replication, as they also pay insufficient attention to differences in replication between psychology and natural science. As Petty notes in an invited commentary on Zwaan et al. (2018), “in the physical sciences . . . when mixing hydrogen and oxygen to create water, the choice of operation to represent the hydrogen and oxygen concepts is constrained because there is a tight link between concepts and operations” (p. 35). Yet, as a consequence of human technolinguistic flair and its affordances for sociocultural change, there is simply no such parallel in many fields of psychology. The problem is not only that “the choice of operations to represent concepts is vast” (Zwaan et al., 2018, p. 35), but also that sociocultural change in the intervening period may mean that the constructs employed are no longer valid—a difficulty accentuated by the fact that construct validity receives neither sufficient attention (Eronen & Bringmann, 2021; Muthukrishna & Henrich, 2019; Oberauer & Lewandowsky, 2019) nor epistemological scrutiny (Uher, 2021). Similarly, I would suggest that Meehl’s (1978) original complaint that theories “never die, they just slowly fade away” (p. 807) needs to be seen in the context of sociocultural change. In other words, because human technolinguistic flair makes such change likely, it is also probable that many theories may suffer “a period of disillusionment as the negative data come in, a growing bafflement about inconsistent and unreplicable empirical results, multiple resort to ad hoc excuses” (Meehl, 1978, p. 807). They may no longer fit a different time or place.

Lastly, as Zwaan et al. (2018) themselves note, there remains the issue of whether the underlying theories are impoverished, so that psychology rarely approximates the theoretical strength of natural science. To maximize the probability of an exact replication, we need a theory that is plausible and well worked out (Irvine, 2021). Yet we rarely have a causal model like that which Galileo considered when analysing the acceleration of falling bodies, where there is a strong theoretical relationship between time and distance, and, in consequence, a likelihood of invariant replicability across experiments (Meehl, 1978). Although Trafimow and Earp (2016) attempt to qualify this relationship by noting that Galileo’s original studies were not based on formal theory, their relatively straightforward replicability occurs because there is a strong underlying theory to be constructed (cf. Klein, 2016). I am not persuaded that there is such a parallel in psychology if only because “psychologists do not actually propose ‘strong’ testable theories” (Zwaan et al., 2018, p. 15). Furthermore, as Irvine (2021) observes, the wealth of possible factors that might influence psychometric invariance is vast, particularly in sociocultural psychology. One might extend Irvine’s argument to suggest that it is the triangulation between theory, experimental design, and measurement that is critical to effective replication. Although this was something that was relatively easy to achieve with respect to Galileo’s falling bodies, or hydrogen and oxygen, it remains far more difficult in sociocultural psychology due to the wide range of possible variables (Meehl, 1978, 1990), as well as questions over the validity of the measurement procedure (Irvine, 2021).

In sum, replication studies have constituted a key means through which the assumption of sociocultural stability has become problematized, rather than validated. Notwithstanding questions about their reliability, they frequently point to the likelihood of change, rather than stability, over time and space.

Conclusion

In questioning whether psychology studies situations that are “stable, unvarying, and singular” (Morawski, 2022, p. 159), much of the preceding logic lends support to the critical-historical approach to psychology, especially discursive psychology’s attention to language and historical/cultural psychology’s focus on variation across time and space. Yet it also departs from some of its argument. First, although the salience of human language remains paramount, there is a need to avoid the kind of linguistic determinism that reduces the natural world to language and social constructionism alone (Nightingale & Cromby, 1999; Soper, 1995), whilst also acknowledging the centrality of biology to human existence.

Second, there remains a need to explore “research methods enabling us to discern the relative durability [emphasis added] of social phenomena” (Gergen, 1973, p. 318). In other words, although human technolinguistic skill increases the likelihood of ongoing, and accelerating, sociocultural change (Rosa, 2013), change processes are nevertheless likely to remain variable and uneven (Newton & Keenan, 1991). In certain contexts and demographics, sociocultural stability may well pertain for a period, even though it should not be assumed a priori.

Third, the problems represented by the replication “crisis” cannot easily be resolved. As Muthukrishna and Henrich (2019) argue, multiple replication is not a feasible solution across large numbers of studies. Yet their proposed solution of a unifying theoretical framework also remains problematic, given the limited number of strong theories (see above). In addition, even with conceptual replication, it is difficult to determine whether the modifications to the experimental design adequately model the social, cultural, and technological changes that are supposed to have occurred. There is always a reasonable likelihood that we are addressing an ambiguous moving target, so that “concept creep” is less likely to represent the refinement of a concept as its continual, and problematic, adjustment to changing technical and sociocultural circumstances (Haslam et al., 2020).

Fourth, these issues are unlikely to be adequately resolved by the traditional, if infrequently applied, call for longitudinal design since similar constraints to those observed with conceptual replication arise. In particular, the concepts we operationalize at one point in time may no longer be the same at a later date, or it may be difficult to formulate an adequate assessment of them.

Fifth, the traditional anxieties about unification and universalism in psychology are likely to be as problematic to resolve as in any other science (Brock, 2016; Galison & Stump, 1996; Gieryn, 1999). That both unification and universalism remain chimeric derives from the fact that our biological selves can exhibit stability, but our social selves may not. This contextual variability is likely to continue to frustrate desires for universalism and epistemological unification as the epistemological assumptions of dynamic stability that we employ with biological processes may not apply to sociocultural processes, due to the exceptional technolinguistic plasticity of our species.

Sixth, we need to re-examine Kim’s (1999) claim that the cause of psychology’s crises derives from the inappropriate emulation of natural science. Within the biological processes of human psychology, the epistemological assumptions of natural sciences can appear more appropriate. It is when we look at sociocultural processes that natural science epistemologies become more tendentious, relying as they implicitly do on an assumption of process stability across time and space. This means that, as Sullivan (2020) suggests, we require much greater attention to questions of epistemology. However, there is also a need to address the fact that psychology operates across a biosocial terrain, and consequently we need plural epistemologies, since the assumptions we make with respect to the biological aspects of human existence do not necessarily hold with human sociality.

Seventh, to adequately address these issues, there is a need for both interdisciplinary and multimethod endeavours. With the former, exploration of the philosophy and sociology of science is of particular interest, as is the work of sociologists, historians, and anthropologists examining psychological issues. With the latter, ethnographic research may prove particularly fruitful, triangulating between observational, interview, and documentary data. Ethnographic research may also be used to address changes in processes and concept creep over time and space. For instance, Kuokkanen et al. (2020) use documentary data to explore changing perceptions of stress, well-being, and mental health in Finland between 1995 and 2014. In particular, they argue that a gradual shift toward a neoliberal economy occasioned an individualization of mental health discourse that increasingly placed responsibility for the management of stress and well-being with the individual employee rather than the corporation (cf. Hutmacher, 2019; Newton, 1995). Although this historical research necessarily lacked triangulation with other forms of ethnographic data, it does provide an example of research attempting to explore “the interrelation of events over extended periods of time” (Gergen, 1973, p. 319).

To sum up, ontological and epistemological tensions may continue to characterize psychology. Many of the questions relating to the scientific status of psychology, and its legitimacy, derive from the different kinds of temporality that apply to biological and social processes. The problem for the discipline is that it cannot be governed by a single epistemology and method because the different temporalities of biological and social processes require multiple ways of investigating what it is to be human. Yet, rather than constituting grounds for “crisis anxiety,” embracing pluralism represents a potentially fruitful and creative terrain on which to advance knowledge (Wegerhoff et al., 2022).