The limits of language as the limits of psychological explanation

The common sense limits of hypothesis testing

The present journey begins with what was an obscure but profound challenge to the experimental tradition in psychology. In 1978, Norwegian psychologist Jan Smedslund published an article in the Scandinavian Journal of Psychology proposing that Bandura’s (1977) highly acclaimed theory of self-efficacy essentially duplicates common sense suppositions. Most cogently, he argued that the extensive experimental support for the theory was essentially pointless, as failures to support its hypotheses would violate common sense. In effect, they would be unintelligible. To expand, Bandura’s theory was concerned with people’s coping behavior, how long it can be sustained, and whether they would press on in spite of obstacles. The determinants of such activity, Bandura reasoned, are primarily cognitive in nature, with self-expectations playing a critical role. Smedslund subjected the major explanatory propositions to careful conceptual analysis, demonstrating, one by one, that not only are the suppositions derived from common sense, but they are empirically untestable.

To illustrate, Bandura proposes that a “person’s convictions of their own effectiveness is likely to affect whether they will even try to cope with given situations” (1977, p. 193). Smedslund then converts this proposal into a formal theorem: “If P wants to do T in S and if P believes with complete certainty that he can do T in S, and no other circumstances intervene, then P will try to do T in S” (1978, p. 3). He then goes on to demonstrate the unintelligibility of this not being the case: “The alternative to P trying to do T in S is P not trying to do T in S. But P not trying to do T in S is not acceptably explained by P’s wanting to do T in S and by P’s certainty that he can do T in S. Hence some additional circumstances must be involved” (p. 3). Simply put, if people are convinced they can do something they want to do, and no other circumstances prevail, then it will make no sense (unless qualifications are inserted) that they will not try to do it. To introduce evidence that they will try to do what they want, knowing they can, is pointless. The contrary would be cultural nonsense.

Such analyses are as intriguing as they are profound. Are all our mental explanations lurking within the rules of ordinary language? Can we never uncover the psychological origins of our actions, never find illumination through careful and systematic research? Or more generally, has more than a century of empirical work in psychology simply served as a chorus for common sense understandings? Indeed, Smedslund (1972) surmised that all psychological theory is derived from a conceptual network embedded in the ordinary language of the culture. Yet, such a proposal would surely meet with resistance in the discipline of psychology. Its theories often seem to cut against the grain of common sense. For decades, psychoanalytic theory was viewed by outsiders with suspicion, Jungian theory was regarded as a fairy tale, and Skinnerian theory was repugnant—while simultaneously embraced by cadres of serious and well-trained professionals. The most attractive feature of Festinger’s (1957) cognitive dissonance theory was its capacity to make counter-intuitive predictions. Clearly, there is more to be said on the issue of linguistic determination, and this issue cuts to the core of psychology.

Tautological explanation and the logic of originary resemblance

One of the most compelling questions triggered by Smedslund’s work concerns the origins of logical necessity. Why, in the preceding example, does it make no sense to say that if a person wants to do something, and he can do it, that he won’t do it? In this case and others, Smedslund (1978) argues that we are compelled by common sense understandings within the culture. Yet, how did these understandings come into being? How indeed did we discover that people have mental processes and these processes influence their behavior? In later publications, Smedslund (2004) draws from the work of Anna Wierzbicka (1996) in proposing that our concept of human action rests on a series of “semantic primitives,” amounting to a universal and “complete archetypical conception of human being” (Smedslund, 2004, p. 81). As proposed, for example, it is simply axiomatic that people can “know, think, want, and feel…” (p. 82).

Although the idea of universally shared conceptions of the person is a fascinating possibility, it is not immediately compelling. There are simply too many historical and anthropological accounts of variations in people’s constructions of the person to warrant such a sweeping generalization. Where, for example, in the semantic primitives would one place the soul, nirvana, repression, instinct, or a fractional anticipatory goal response? Simultaneously, there are numerous instances in which psychological science seems to have affected common sense understandings. (Consider the proliferation of “mental illness” discourse into Western culture.) And as pointed out, counter-intuitive explanations are not uncommon in psychology.

Interestingly, one of Smedslund’s (1978) demonstrations of Bandura’s common sense assumptions provides a clue to an alternative. In this formalization of a Bandura proposal, “If P believes he is capable of handling S and if no other circumstances intervene, then P will behave assuredly” (p. 4), Smedslund’s proof of common sense is that “it follows directly from the meaning of theorems involved. To ‘behave assuredly’ means to ‘behave believing one is capable of handling the situation’” (1978, p. 4). To describe a person as behaving confidently is not fundamentally different from saying the person is confident. In effect, the explanation is tautological. Believing oneself to be capable is not a cause of behaving assuredly; it is essentially a definition of what it is to be assured. The relation between cause and effect is circular. With further effort, it also proves possible to convert virtually all of Smedslund’s common sense theorems to tautology. For example, in the earlier illustration, “belief” that one can succeed is part of what it means to “try” to succeed; trying to succeed is premised on the assumption that success is possible. With this move in place, the door is open to considering the more general possibility that most (if not all) psychological explanations are tautological.

Why should one suspect this might be the case? Here it is useful to consider the challenge of using psychology to explain behavior, whether professional psychologists or in the population more generally. The challenge rests on two conceptual assumptions: mind/behavior dualism and mechanistic causation. In brief, mental events are the underlying determinants of physical action. Now, the researcher does have human action available for observation (the explanandum), but wishes to explain this action by virtue of a non-observable psychological state, condition, or mechanism (the explanans). Tracing the relation between cause and effect would thus require (a) ontological specification (i.e., what exists within the mental world) and (b) mind–behavior linkages (i.e., a specification of the relationship between the mental existents and manifest in action). Yet, the observed behavior in itself furnishes no principled guidance in specifying either the nature of the psychological cause or how it is related to behavior. For example, we scarcely doubt the social psychological proposal that attitudes toward political candidates affect one’s voting. But we cannot derive this conclusion from the fact of voting. Rather, it appears we must presume a priori that (a) attitudes are constituents of the mental world and (b) they are realized in the act of voting.

By what means, then, can we establish the plausibility of psychological explanations? Here let us consider the possibility of a culturally shared logic of originary resemblance, to wit, with no other information available, attempts to explain the causal source of an action will bear a definitional similarity to this action. Because there are no observables on which to rely, one reasons that the psychological origin of an action must resemble in some form the action to be explained. Relevant to psychology, the iconic example of the principle is furnished by Aristotle’s answer to what animates the otherwise lifeless body. In De Anima, he posits the psyche as the “form of the living thing,” that which itself has the capacity for movement and growth. In effect, movement of the body and that which gives rise to the movement are almost identical. Likewise, in Descartes, there is virtually no distinction drawn between the way in which he describes his reasoning process and the written language that serves as its expression. And today, in the psychology of language, it is typically presumed that we use language to express thoughts or concepts. Yet, concepts are typically defined as mental categories, virtually indistinguishable from the activity of using nouns. The plausibility of an explanation thus appears to derive from the fact that the explanans bears definitional resemblance to the explanandum. The dependency on tautology is further clarified in two additional ways. First, consider explanations that are non-tautological. Without further qualification it would make little sense to say, for example, “he was so sleepy that he went out to vote,” or “her passion for antiques caused her to vote.” Remove the common definition and the explanatory sense dissolves. Second, in cases where both explanans and explanandum are observable, we are not limited to circularity. To explain the growth of crops in terms of temperature and rainfall, for example, does not depend on tautology. It is the absence of a commonly observable explanans that invites reliance on tautology. In sum, we confront the radical possibility that: all propositions concerning the relationship between mental states and human behavior owe their intelligibility to the degree to which they share definitional space.

In a certain sense the question here is empirical. If we survey the vast range of psychological explanations that appear in the research literature, to what extent do they rely on this definitional process? Here the most extensive work to date has been carried out by Lisa and Michael Wallach (L. Wallach & M. A. Wallach, 1994, 2001; M. A. Wallach & L. Wallach, 1998). As their research suggests, the vast bulk of explanation in social psychology research relies on near-tautologies. The attendant research is unfalsifiable. And while controversial (Schaller, Crandall, Stangor, & Neuberg, 1995), their conclusions remain robust and challenging. Yet, further questions remain. How can the present proposal be squared with the many instances of counter-intuitive explanations in psychology? Especially important are cases in which actions do not appear to be in the individual’s best interest. Further, we find many instances in psychological research in which the results are either unexpected or the reverse of the prediction. In effect, they violate the demands for circularity. We gain clarity on these and related issues by exploring the definitional process more fully.

Semiotic slippage: Tautology goes on holiday

The preceding arguments suggest that the limits to psychological explanations of an action are furnished by the definitional attributes of the action itself. Thus, for example, we can say that “she gave a donation to charity because she wanted to help, because giving to charity is an act of helping, and wanting to help is part of what it means to help. (One cannot sensibly say, “she wanted to help, and therefore didn’t help.”) But how are we to understand the nature of definition and its confinements? The word “definition” is commonly defined as a statement that circumscribes the “meaning of a word.” This definition may seem acceptable enough, until we ask how we are to understand the word, “meaning.” Here our analysis takes a new turn. One of the most philosophically relevant contributions emerging from post-structural theory is Jacques Derrida’s (1982) concept of différance. Simply put, the concept points to the fact that the meaning of a word is never given in itself. Its definition must be sought in other words. In effect, we must defer our understanding of a word until we locate a word (or words) that differ from the original. (The concept of différance represents an elision of the joint requirements of differing and deferring.)

In this light, we must re-examine what has so far been characterized as a “constrained determinism” in psychological explanation, that is, one cannot escape tautology. However, the fact that terms for mental events or entities (e.g., thought, emotion, motivation) are defined by other mental terms, suggests an extended or dispersed definitional network from which intelligible explanations may be derived. In short, we confront a condition of extended semiotic slippage, resulting in significant plasticity in explanation. To explore, for analytic purposes let us consider that for any description of behavior, there is a range of defining attributes. These may be said to form an intelligibility nucleus (Gergen, 1984). Consider, for example, the various terms or phrases that define what it is to call someone’s actions aggressive. At the center of the nucleus we might place the common definition, that aggression is “hostile or violent behavior toward another.” It is also “an attack” or “an assault.” Following the preceding arguments, we thus establish a range of explanations that are either nonsense, congenial, or contrary to our understanding. In terms of nonsense, one cannot intelligibly explain that, “his green mind caused his violence,” and it would be puzzling to explain that “she didn’t feel at all hostile so she attacked him.” Yet, to explain that “people with hostile personality dispositions are often aggressive,” would be unremarkable. One does not escape the definitional nucleus.

However, the case rapidly gains complexity when we begin to define the range of terms within the nucleus. Or, one might say, that every term within a definitional nucleus is also featured in at least one other nucleus. For analytic purposes we can call these second order nuclei. For example, “hostility”—a component of the aggression nucleus—is also the center of a second order nucleus, and within this nucleus it is often defined as an “emotional state.” Most importantly, we find that by extending the definitional network in this way, we open new potentials for what “makes sense.” It becomes plausible to say, for example, that “emotional arousal can be channeled into aggression”—essentially an extended tautology.

To press further, the term “emotional state” will also occupy the center of yet another nucleus, and as its definition is unpacked, a further range of potentially plausible utterances is generated. For example, an “emotional state” can be defined as the “absence of reason.” By definitional extension, this makes it possible to say that “losing all reason, he struck her,” or by further extension, “He lost his wits and lashed out.” One may also expand such analyses to include for example, fourth and fifth order nuclei. Because war is defined as combat, and armies engage in combat, for instance, we can intelligibly conjecture that “hostile people are more likely to volunteer for the armed services.” The reverse of this proposal would be suspicious. Such extensions also permit us to conjecture that, owing to our defining education as acquiring skills in reasoning, “the less educated are more hawkish in their attitudes toward war.” Importantly, such conjectures do not depend on observation to acquire their intelligibility. They constitute the background from which we make sense of the world. One might say that tautologies—whether transparent or extended—are essential ingredients for sustaining our cultural capacities for making sense.

As we find, the potential for semiotic slippage can vastly expand the range of what may be felicitously said about the underlying causes of behavior. However, with sufficient slippage we also approach the borders of implausibility. For example, love may be defined as intense attraction, which may be defined as intense desire, then defined as a state of desperate need, subsequently defined as a helpless dependency, which can be defined as a state of slavery, defined as morally reprehensible, defined as something to be eradicated. With extended slippage, we might thus conclude that we wish to destroy those we love, and love should be abolished. Wittgenstein (1953) used the metaphor of language “going on holiday,” by which he meant that the meaning of a term is not confined to its usage in a given context. The term may be lifted from one context and used metaphorically in others. Meaning is set loose from its moorings. As we find in the present case, through semiotic slippage, tautology is indeed invited on holiday.

Escaping linguistic determinism?

Given an indeterminate network of extended definitions, we may reconsider the limits of what can intelligibly be said about the psychological sources of a given action. Given the liquidity of extended definition, is it possible that: any psychological state may satisfactorily explain any human action? Or conversely, given someone’s behavior, can it be explained in terms of virtually any psychological state? There are good reasons to resist such a conclusion. Mental testing, assessment, and diagnostic analyses all have a long history of development. Claims are made to scientific objectivity, based on logical and systematic practices of test design. With multiple, carefully screened, and inter-related test items, psychologists inform the world about an individual’s intelligence, personality, prejudices, abilities, cognitive proclivities, mental illness, and so on. Does the methodological rigor with which these tests are constructed not resist the possibility for infinite re-explanation?

To explore this question, research was conducted on the potentials of a standardized personality test to limit the range of explanations (Gergen, Hepburn, & Comer, 1986). Selected for inquiry was Rotter’s (1966) internal–external (I–E) control scale, used to assess the degree to which individuals see themselves as responsible for their actions and their consequences, as opposed to external world conditions. Agreement with an item such as “There is a direct connection between how hard I study and the grades I get,” would be indicative of an internal locus of control. To agree that, “Unfortunately, an individual’s worth often passes unrecognized no matter how hard he tries” would reveal a tendency to see one’s outcomes as controlled by external circumstances. Scores on the test measure have been used to predict a wide variety of behaviors (Lefcourt, 2014).

The first step in the exploration was to determine the extent to which one could explain responses on individual items to virtually any psychological disposition. Two dozen undergraduate students were thus enlisted in a series of “interpretation puzzles.” As one of their challenges, they were asked if they could show how it would make sense for someone who had a given disposition to agree with a given item from the Rotter scale. The trait dispositions were randomly drawn from a list of some 500 common traits.

It was found that the participants rarely encountered difficulty. For example, it was explained that a lonely person would say that who gets to be boss is a matter of luck (external) because, “A lonely person lacks self-confidence and thus believes his actions will make no difference in the outcome.” A person who is impulsive would agree that an individual’s worth passes unrecognized no matter how hard he tries, because “he might very well need to justify his feelings of staying too short a time with one project or another by believing that no matter how persevering he remains, he won’t be acknowledged anyway.” An independent group of research participants reviewed the offered explanations, and on a 5-point scale judged almost all of them as “plausible” to “highly plausible.”

The participants were then asked if they could take the same trait and show how it could be expressed in two opposing items from the I–E scale. Again, the students were undaunted. For example, they could easily see why a broad-minded person would agree that “for well-prepared students there are no unfair tests” (an Internal indicator), because having a broad mind means understanding the views of others. They could also see why a broad-minded person would agree that we are “victims of forces out of our control” (External), because the “broad-minded person would not try to blame world events on a particular politician or groups.” The research also revealed that the various rationales employed by participants seldom duplicated each other. In other words, there were multiple and intelligible ways in which a given response on the Rotter scale could be explained in terms of a given psychological trait. And further, participants had little difficulty in relating any given psychological trait to a group of multiple and contradictory items on the Rotter scale.

This pattern of findings suggests that the common claims that psychological testing (including psycho-diagnostics) successfully illuminates the psychological sources of behavior are groundless. With little effort, any measure of any underlying source could also be explained as a measure of virtually any other source. Nor would attempts to justify inferences through predictive validity studies unseat such a conclusion. The behaviors selected for prediction in such studies stand in the same relationship to the “underlying” disposition as do the items on the trait measures. The explanatory possibilities are without clear limits.

A further scanning of the explanations was undertaken to determine the extent to which they relied on definitionally extended tautologies. While unpacking these definitional sequences was sometimes challenging, for the most part the linkages were evident. Yet, there was one commonly used definitional adjustment that was unanticipated. We may term this definitional maneuver self-repairing, and its implications for the freedom of explanation are substantial. Simply put, self-repairing explanations result from the fact that any trait, desire, or other psychological state commonly judged as bad or unsatisfactory carries with it the understanding that repair is desired. Simply put, that which is defined as undesirable presumes the desire to be otherwise. To illustrate, consider how one might explain why a lazy person would take up rock climbing. The simple tautology is cultural nonsense. However, one can explain that the lazy person is distressed with his lazy way of life, and thus takes up an active sport. Such an explanation draws from the unstated assumption that one does not wish to be lazy, or more broadly, that negative states carry with them the desire for positive repair. The tautological form remains, as the desire to be active is simultaneously a desire to avoid being inactive.

This capacity for self-repair enabled research participants to inject plausibility into numerous cases of otherwise nonsensical connections between trait and behavior. However, it also sheds light on the means by which otherwise counter-intuitive explanations in psychology achieve plausibility. Freud’s (1933) mechanisms of defense are among the most interesting. In the case of reaction formation, for example, one expresses love for that which is unconsciously hated because the latter feeling is an undesirable threat to the ego. In defining the “undesirable,” the condition is provided for seeking the desirable, in this case a defense mechanism. The defense mechanism projection involves attributing to others that which one desires because the desire would threaten the ego. The dangerous sexual desire is the definitional equivalent of saying that it should be resisted; resistance is realized in the embrace of its opposite.

Proofs and provocations

This inquiry into the relationship of language to psychological explanation began by recounting Smedslund’s (1978) ground-breaking demonstrations of the common sense underpinnings of hypothesis testing. As he proposed, experiments do not test hypotheses about the relationship between psychological process and behavior because any failure to verify would defy cultural understanding. Upon further examination, however, it was advanced that the intelligibility of relationships between psychological explanans and behavioral explananda does not rest so much on cultural understandings as on tautology. Or to put it otherwise, explanations in both psychology and the culture at large draw from the same linguistic practice. As further reasoned, the reliance on tautology is born of the fact that ostensive definitions of the explanans in psychology are impossible. Thus, one makes psychological sense by explaining a given behavior in terms of a “miniaturized” form of itself. This default practice was captured in the concept of originary resemblance—with the behavioral outcome resembling its psychological origin.

Further exploration of the definitional process brought to attention the potential for unbridled diffusion of definition. This possibility emerges from the recognition that each definition of a psychological state or condition relies on further definition for its meaning. In this light it was possible to account for explanations far removed from simple tautology. By drawing on extended definitional sequences, it was found that a given behavior (or its negation) could be explained by virtually any randomly drawn motive or trait. Because definitional dispersion also allows definition by self-repair (e.g., every negative characteristic is defined as that which should be resisted or repaired), it was possible to account for counter-intuitive or paradoxical explanations. With Wittgenstein, our language may furnish the limits of psychological explanation, but within those limits lies an infinitude of plausibility.

Explaining mind–world relations

Our spoken and written language is unlike mathematics in that its logics are not fixed. In Wittgenstein’s (1953) terms, the system acquires its utility from its use, and its use may shift with time and circumstance. Thus, the proposals put forward here cannot be proved as mathematical theorems. There may be exceptions, and indeed there are logics not revealed by the present analysis (Tellings, 2017). At the same time, the present analysis suggests new and significant directions of inquiry. For one, the analysis is limited to psychological explanations of behavior. Yet, can the same logics not be applied to the relationships between the “stimulus world” and psychological states? An opening to this possibility is provided by Rorty’s (1981) proposal that the philosophical problem of epistemology is essentially a linguistic one: How to relate two domains of discourse—a discourse of material reality on the one side and mental reality on the other. Austin’s (1962) analysis of the linguistic play between sense and sensibilia expands on this possibility. Given the logic of originary resemblance, one might thus inquire into the reliance on extended tautology in explaining the relation between the “external” world and the world as “experienced.” To what extent are descriptions of one’s experiences of the world drawing from the same definitional source as our descriptions of the world itself. If we commonly agree that “there is a rabbit,” we cannot intelligibly explain, without qualification, that we thus “experience a horse.” Unless we disqualify the observer, we must conclude that the experience bears some resemblance to the thing in itself.

Let us apply this reasoning to psychological experimentation. For example, classical studies on attitude change demonstrated that characteristics of a communicator will affect attitude change. Thus, if a communicator is an expert, is attractive, or trusted, for example, one is more likely to agree with him or her. Yet, in terms of definition, this is to say little more than we agree with what is agreeable, or are attracted to what attracts us. The empirical findings are superfluous. Or, in the case of Wertheimer’s (1912) classic Gestalt research on apparent motion, it was shown that a string of lights, rapidly illuminated in succession, is experienced by the observer as motion. Yet, we might also conclude that the same event is simply defined by the researchers as “a rapid succession of lights” and by the experimental subject as “motion.” They could have both used the same terms, as they are mutually defining. Is it possible, then, that the entire S-O-R paradigm on which the edifice of experimental psychology is built depends for its intelligibility on extended tautology (Gergen, 1987)?

Of further concern, we may inquire into the extent to which some form of tautology affects even those hypotheses for which ostensive definition of both explanans and explananda are available. Why, for example, is it unremarkable for psychologists to explain crime, drug addiction, and suicide in terms of deficient child-rearing, poor education, poverty, trauma, and the like? Is it impossible that all such behavior could be the result of excellent parenting, a good education, sufficient income, and a trauma-free life? Does a culturally devalued effect always have a devalued cause? How could we make sense of the contrary?

Such issues invite continued attention. In the meantime, we confront a variety of provocations that emerge from the present analysis.

Hypothesis testing and beyond

Research relating mind to world has been a mainstay in psychology for over a century. Inquiry is launched into the effects on behavior of cognitive processes, emotions, motivation, prejudice, values, personality, and so on. Yet, as the present line of argument suggests, regardless of the rigor of the research, the psychological explanation for an action will be limited to a partial definition of the action itself. Further, because of the process of definitional dispersion, an indeterminate array of alternative explanations of the research results can be generated, while a vast array of alternative explanations (non-tautological) can be discarded as nonsense. All such explanations can be generated (or discarded) without the benefit of the empirical research itself. In effect, the research cannot itself furnish an account of human action that is not already available; nor can it demonstrate that one explanation is objectively superior to another. An indeterminate number of explanations can be provided for any behavioral pattern. Buttressing and extending Smedslund’s (1978) early surmise, the value of testing hypotheses about the relationship of psychological process to behavior is unclear.

Alternatives to this tradition demand renewed attention. At the outset, one could abandon either one or both of its grounding suppositions: mind/world dualism and/or mechanistic causation. Skinnerian behaviorists along with actuarially oriented researchers have followed the first path. Without positing mental causes, potentially useful predictions of drug dependency, suicide, school success, and the like can result from multivariate statistical designs. At the same time, such research findings beg for explanation. Why, for example, would suicide rates spike in both the teen years and in old age? Answering the question of “why” will almost inevitably invite the language of psychological process. In this case, for example, issues of stress and loneliness might well become focal. To abandon psychological explanation would be to abandon cultural life.

Many humanists eschew mechanist causation in favor of intentional or purposive explanations (Schneider & Krug, 2010; Taylor, 1964). They also question the research goal of prediction and control, and the power relations that are implied. The central challenge of research becomes that of enhancing human understanding. Again, however, that which is to be understood is psychological in nature (e.g., experience, emotion, desires). Drawing from the present analysis, the number of intelligible “understandings” of another’s subjectivity are limited only by one’s linguistic skills. One does not acquire “understanding” of the mental world through empirical research—humanist or otherwise—but through linguistic interchange.

A third orientation to psychological explanation is supplied by discursively oriented psychologists (Edwards & Potter, 1992; Tileaga, 2012). In this case, one abandons concern with the mind as a reality to be explored and attention turns to the use of mental discourse in social life. Jettisoning as well the assumption of causation, the focus is on social life as a continuous process of negotiating meaning. In this case, semiotic slippage serves a pivotal role in the pragmatics of everyday life. The question of whether, for example, the actions of a nation’s president are the result of his/her mental illness or of clever reasoning, is not an empirical question. Conclusions will largely depend on whether such negotiations will lead to describing presidential actions as irrational or tactical. From this perspective, the conception of research as a report on contemporary affairs is delimited. It would be akin to taking a snapshot of an ocean wave. Attention is thus directed to research as a future-forming process (Gergen, 2015), where the concern shifts from reporting on current conditions to building more promising futures.

Mental testing and diagnosis as world construction

Mental testing is now lodged deeply within our cultural institutions. Psycho-diagnostics are pivotal to the therapeutic profession, and psychological tests are widely employed in education, counseling, and business. As the present analysis suggests, the relationship between any diagnostic indicator or test score and an underlying psychological condition is indeterminate. There are no compelling grounds for claims of diagnostic accuracy or testing certitude. One might well ask whether diagnostic tests are in any sense measures of mind, as opposed to the conversational preferences of their creators. Drawing from the preceding analysis, we see that the choice of behavioral descriptions contains the implications for what may be said about their mental source. If a child’s behavior is described as “attention deficit,” the mental cause will take the form of pathology; this would not necessarily follow if the same behavior were described as boredom or irrepressible curiosity. Thus, as authoritative claims are made to mental causes, one simultaneously creates the reality of those causes. In claims to diagnosis or measurement, one becomes an author of mental ontology. It is here that the present analysis offers significant support for critical analyses of the societal consequences of psycho-diagnostics and mental testing (Whitaker, 2010). Removing the veil of “objective assessment,” significant questions must be raised as to the cultural impact of classifying people as introverted, neurotic, deficient in happiness, mentally disordered, and in other invalidating ways.

Therapeutic practice

From the late 1800s to the present, most psychotherapeutic practice has been oriented around mental causes. Whether repression, archetypes, self-regard, construals, emotional traumas, or cognitive dispositions, the attempt is to illuminate the psychological causes of one’s anguish or inabilities. As the present analysis suggests, such searches for psychological well-springs can only yield definitional simulacra of the way in which the symptom is described. Despite hours of therapeutic exploration, beneath a problematic symptom will inevitably be located a problematic condition of mind. Further, with semiotic slippage in motion, the relationship between the particular cause and the symptom is indeterminate; the explanation for any effect (e.g., depression, bi-polarity), may be traced to an indeterminate number of mental causes (e.g., repression, cognitive mechanisms, emotional state); and therapists of any school may provide an explanation for virtually all symptoms.

One might thus conclude that therapeutic attempts at “mental plumbing” are futile, mere linguistic exercises. If therapy were indeed a matter of “locating the truth about one’s mental life” as a route to healing, this might be reasonable. However, as commonly agreed (Muran & Barber, 2010), the quality of the relationship between therapist and client is far more important to the therapeutic outcome than the content of the conversation. On these grounds, because one’s inner life in Western culture is a privileged sanctum, conversations about its character may be a critical element in forming a therapeutic alliance.

At the same time, we must consider the way in which objectification of mental causes also lends weight to the psychologizing of social problems. To inquire into the depths of a client’s depression, for example, removes from attention the possibility of workplace oppression and exploitation. To presume that people’s problems result from mental illness fails to address the potential role of social injustice in our institutional settings. From a more discursive standpoint, the challenge of therapy might profitably be shifted from mental exploration to enhancing client skills in the social negotiation of personal descriptions and explanations. Understanding when, where, and how to describe one’s self as “depressed,” for example, may be far more useful than exploring the “depths” of one’s depression.

My attempt in the present undertaking has been to bring together a range of related inquiries in order to further discussion on the utility of psychological explanation. My specific concern is with the way in which language is insinuated into understanding of the mental world. To complete the discussion, however, it is useful to reflect on the impact of language practices in theoretical psychology itself. The field has yielded a rich array of innovations in psychological theory, along with incisive conceptual analyses of existing formulations. Further, scholarship has successfully illuminated ideological and moral dimensions of various theoretical standpoints. Yet, the present analysis suggests that our tools of representation—in this case language—influence our understandings of mental life. In the same way, we may ask whether these same tools may not affect our efforts at conceptual analysis and critique. Or more broadly, to what degree does our dependency on spoken and written language limit or otherwise influence our capacities for effectively grappling with the issues of the day? We inherit, for example, a language of nouns and pronouns. While unremarkable in itself, reliance on these tools will necessarily create a world of separate units—of persons, places, and things. We presume such an ontology in our analyses not because “this is the way the world is.” The language is there as a fore structure that virtually requires such a world. Thus, worlds of configuration, harmonies, co-constitutions, and continuous process—all irreducible to independent units—become territories beyond elementary articulation. It is here that theoretical psychology might usefully open a dialogue with the arts.