Concept Creep: Psychology’s Expanding Concepts of Harm and Pathology

TARGET ARTICLE

Concept Creep: Psychology’s Expanding Concepts of Harm and Pathology

Nick Haslam Melbourne School of Psychological Sciences, University of Melbourne, Parkville, Australia

Many of psychology’s concepts have undergone semantic shifts in recent years. These conceptual changes follow a consistent trend. Concepts that refer to the negative aspects of human experience and behavior have expanded their meanings so that they now encompass a much broader range of phenomena than before. This expansion takes “horizontal” and “vertical” forms: concepts extend outward to capture qualitatively new phenomena and downward to capture quantitatively less extreme phenomena. The concepts of abuse, bullying, trauma, mental disorder, addiction, and prejudice are examined to illustrate these historical changes. In each case, the concept’s boundary has stretched and its meaning has dilated. A variety of explanations for this pattern of “concept creep” are considered and its implications are explored. I contend that the expansion primarily reflects an ever- increasing sensitivity to harm, reflecting a liberal moral agenda. Its implications are ambivalent, however. Although conceptual change is inevitable and often well motivated, concept creep runs the risk of pathologizing everyday experience and encouraging a sense of virtuous but impotent victimhood.

Key words: bullying, concepts, moral psychology, prejudice, trauma

Writing in 1993, Daniel Patrick Moynihan, senior senator for New York, alliterated that his country was “defining deviancy down.” Moynihan argued that in response to rising crime and social disorder in the 1970s and 1980s, the public increasingly normalized behavior that would once have been seen as patholog- ical. Sometimes, he proposed, this process was driven by the worthy goal of social inclusion, countering the tendency to stigmatize people on society’s margins. At other times it merely represented a habituation to ongoing social change. Whatever its cause, phenom- ena that had once been seen as deviant were redefined as the new normal.

To Moynihan (1993), the social and political implications of these developments were troubling. By coming to accept crime and family breakdown, he argued, people were “getting used to a lot of behavior that is not good for us” (p. 30). Conservatives, he wrote, are especially opposed to these normalizing redefinitions of deviance because they see them as weakening standards of conduct and loosening moral strictures. Liberals, in contrast, are traditionally wary of the opposite process, in which normal rebellion or alternative ways of living are pathologized.

There is nothing inevitable about the progressive expansion of normality that Moynihan documented. Indeed, I argue that in recent decades the opposite pro- cess has unfolded: The definition of some forms of devi- ance has enlarged and normality has contracted. Psychology has played a significant role in this process, as many of the concepts it employs to make sense of undesirable forms of experience and behavior have extended their meanings, encroaching on phenomena that would once have been seen as unremarkable. More- over, although Moynihan argued that liberals resist attempts to pathologize deviance, psychology’s expan- sionary redefinition of negative phenomena arguably reflects a liberal social agenda. Instead of defining devi- ancy down, psychology has ubiquitized it up.

Conceptual Change Conceptual shifts can be observed in public dis-

course, the focus of Moynihan’s attention. They can also be seen in the discourse of the social and behav- ioral sciences. These fields traffic in what the philoso- pher Ian Hacking (1995) called “human kinds”:

1

Psychological Inquiry, 27: 1–17, 2016 Copyright! Taylor and Francis Group, LLC ISSN: 1047-840X print / 1532-7965 online

DOI: 10.1080/1047840X.2016.1082418

“kinds of people, their behaviour, their condition, kinds of action, kinds of temperament or tendency, kinds of emotion, and kinds of experience” (p. 351–352). According to Hacking, the meanings of human kinds are not fixed. Unlike “natural kinds,” such as biological species or chemical elements, human kinds are moving targets. The changes that they undergo may influence social reality rather than merely mirroring it. Because human kinds form the basis of social judgments and policies, they are sus- ceptible to “looping effects.” How we define, describe, and label a human kind can mould it by influencing how the people so defined come to under- stand themselves. Through analyses of emerging kinds such as “child abuse,” “autism,” and “multiple personality disorder,” Hacking (1991) showed how people come to recognize themselves in professional characterizations, and how they shape their behavior and sense of self in response.

We should therefore expect psychological con- cepts to undergo semantic changes, and for these altered meanings to have looping effects on how peo- ple make sense of themselves personally and collec- tively. The conceptual changes that I explore in this article involve alterations in the semantic “extension” of the relevant concepts, that is, the range of phenom- ena to which they apply. I propose that these altera- tions take two forms. The first, which I dub “vertical expansion,” occurs when a concept’s meaning becomes less stringent, extending to quantitatively milder variants of the phenomenon to which it origi- nally referred. For example, a mental disorder has undergone vertical expansion if its new diagnostic criteria encompass less severe and debilitating clini- cal phenomena than previous criteria. Vertical expan- sion can occur through a lowering of the threshold for identifying a phenomenon or through the relaxation of criteria for defining it. The second form of concep- tual change, which I call “horizontal expansion,” occurs when a concept extends to a qualitatively new class of phenomena or is applied in a new context. For example, the concept of “refugee” has expanded to include people displaced by environmental catas- trophe, whereas it originally referred only to those displaced by conflict.

Overview

The main contention of this article is that in recent decades the meanings of several of psychology’s key concepts have changed in a systematic way. I argue that those changes have targeted particular kinds of concept and moved in a particular direction. Specifi- cally, it is psychology’s negative concepts—those that refer to undesirable, harmful, or pathological aspects of human experience and behavior—that had

meaning changes, and these changes have consis- tently expanded those meanings. The concepts in question continue to refer to the phenomena they denoted at an earlier time, but they now also refer to a horizontally and vertically enlarged range of addi- tional phenomena. This semantic inflation is not widely appreciated by psychologists. When it has been noted it has been discussed in relation to a single concept, and the general pattern has been missed. In the body of the article I illustrate the “concept creep” hypothesis by reviewing changes in six concepts drawn from the provinces of developmental, clinical, and social psychology: abuse, bullying, trauma, men- tal disorder, addiction, and prejudice.

After presenting these six case studies, I examine the causes and implications of the changes they illus- trate. I argue that a good explanation of concept creep must account for why the changes are specific to neg- ative concepts and why they involve expansion rather than contraction. It should also encompass both verti- cal and horizontal expansion and account for the con- sistency of the effect across diverse concepts rather than explaining each change on its own terms. Explanations that invoke technological, social, and cultural developments are entertained, as are some that implicate psychology as a discipline.

I then discuss the wider consequences of concept creep. As Hacking argued, changes in human kind concepts alter social reality, looping back into how people understand themselves and one another and bringing new kinds of people into existence through what he called “dynamic nominalism” (Hacking, 1986). I am at pains not to present concept creep as unambiguously desirable or undesirable, or to write it off as arbitrary or unwarranted. Conceptual revision is to be expected in view of changing scientific and social realities, and it may be appropriately respon- sive to those changes. Although many critics have held psychological concepts responsible for damag- ing cultural trends—such as supposed cultures of fear, therapy, and victimhood—the conceptual shifts I present have some positive implications. Neverthe- less, they also have potentially damaging ramifica- tions for society and for psychology that cannot be ignored.

Case Study 1: Abuse

The concept of abuse has grown in prominence within psychology and related fields, largely through the growing awareness that maltreatment of children and adults, and its implications for mental health, has been underestimated in the past. This underestimation goes back at least as far as Freud’s abandonment of the seduction theory of hysteria. Decades of research have established the disturbing high prevalence of

HASLAM

2

sexual and physical abuse and demonstrated their causal role in a variety of mental disorders.

Hacking (1991) has written at length about the shifting understandings of abuse and the relevance of looping effects to those shifts. He documented the malleability of ideas of child abuse and how these were shaped by cultural trends, legal institutions, and social movements such as feminism and children’s rights activism. However, his historical study primar- ily addresses changes in professional and popular rep- resentations of abuse from the 19th century through to the 1970s and does not focus specifically on psy- chology. My emphasis here is on more recent changes in the definition of abuse within that field.

Classic psychological investigations of abuse rec- ognized two forms, physical and sexual. Physical abuse involved the intentional infliction of bodily harm, whereas sexual abuse involved inappropriate sexual contact, including penetrative sex or nonpene- trative molestation. Childhood exposure to these forms of abuse was found to increase vulnerability to adult psychopathology, relationship difficulties, and physical ill health.

Three changes to the conceptualization of abuse that have occurred within the psychological literature over recent decades represent clear cases of horizon- tal expansion. First, “emotional abuse” (Thompson & Kaplan, 1996)—sometimes labeled “psychological abuse”—was introduced as a new abuse subtype. It refers to forms of maltreatment that need not involve bodily contact, unlike physical and sexual abuse, but includes verbal aggression and other behavior that is domineering, intimidating, threatening, rejecting, degrading, possessive, inconsistent, or emotionally unresponsive. This form of abuse was commonly studied within intimate domestic relationships. This new focus on behavior exchanged between adults represents a second horizontal extension of the abuse concept from its traditional focus on the behavior of adults toward children.

A third horizontal extension of the abuse concept is its incorporation of neglect. Neglect implies a lack of appropriate care and concern, as when negligent parents fail to tend to their children’s basic needs for food, shelter, clothing, physical contact, and affec- tion. In the early literature on child maltreatment, neglect and abuse were traditionally considered sepa- rately—the field’s flagship journal, which com- menced publication in 1976, was entitled Child Abuse and Neglect—but increasingly neglect has been understood as a form of abuse. Cicchetti and Barnett’s (1991) taxonomy of child abuse, for exam- ple, considers physical neglect as one of its subtypes. Similarly, Goldsmith and Freyd (2005) considered emotional neglect, or “emotional unavailability,” to be a form of emotional abuse.

Emotional abuse and neglect-as-abuse are ideas that represent horizontal extensions of the abuse con- cept. The former extends abuse into the realm of non- physical harm, where damage is done indirectly through language or social interaction. The latter extends the abuse concept by including acts of omis- sion. Whereas physical and sexual forms of abuse represent the commission of undesirable acts toward a victim, neglect involves the failure to commit desir- able acts. Neglect, like physical or sexual abuse, can be an act in the sense of being deliberate, but it differs from these prototypes of abuse by referring to inaction.

The inclusion of emotional abuse and neglect within a broadened concept of abuse may also repre- sent a vertical expansion of that concept. Emotional abuse encompasses some forms of interpersonal mal- treatment that are more diffuse and ambiguous than those that fall within the realms of physical and sexual abuse, which, because they require bodily con- tact, are intrinsically more tangible. Determining what counts as emotional abuse may have a larger element of subjectivity. Whether a particular interac- tion represents humiliation or teasing, possessiveness or protectiveness, and aggressiveness or assertiveness may be uncertain and the parties involved may have very different perceptions. If deciding whether emo- tional abuse has occurred depends on the self-identi- fied victim’s perception, abuse can be invoked as a description that might seem innocuous from an inde- pendent observer’s standpoint. This reliance on highly subjective impressions is a feature of some methods of assessing abuse, as in the following item from a popular self-report measure: “As a child, did you feel unwanted or emotionally neglected?”

A similar vertical expansion of the abuse concept can result when it incorporates neglect. Because crite- ria for judging omissions (i.e., what was not done that should have been) tend to be less concrete than those for judging commissions (i.e., what was done that should not have been), the boundary of neglect is indistinct. As a consequence, the concept of neglect can become overinclusive, identifying behavior as negligent that is substantially milder or more subtle than other forms of abuse. This is not to deny that some forms of neglect are profoundly damaging, merely to argue that the concept’s boundaries are suf- ficiently vague and elastic to encompass forms that are not severe.

This brief discussion of abuse reveals that the con- cept’s meaning has undergone significant inflation, horizontal and vertical. Its message is well captured by Furedi (2006), who noted a “continuous expansion of the range of human experiences which can be labelled as abusive,” such that “neglect and unin- tended insult become equated with physical violence

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

3

and incorporated into an all-purpose generic category ” (p. 86).

Case Study 2: Bullying

Abuse originally referred to objectionable and damaging behavior directed toward children by adults. Bullying also refers to destructive behavior, but it has been primarily examined in the contexts where children are both victims and perpetrators. Understood as a form of proactive aggression, bully- ing has been the focus of an explosion of research since the concept emerged in Scandinavia in the 1970s. As Olweus (2013) has observed, citations of bullying research increased 100-fold from 1990 to 2010.

Olweus, the father of bullying research, proposed three core elements that define the phenomenon: aggres- sive or otherwise negative actions that are directed toward a child by one or more other people, where that behavior is intentional, repetitive, and carried out in the context of a power imbalance. The victim has less power—whether in numbers, size, strength, age, status, or authority—than the bully. Bullying is therefore con- ceptually distinct from peer aggression, where the aggression may not be repeated and the combatants may be of more or less equal power.

The behavior that constitutes bullying takes a vari- ety of forms. Prototypically it involves direct physical harassment, including hitting, kicking, coercion and intimidation, and verbal harassment, including racial and sexual comments, unfriendly teasing, name-call- ing, and threats. Bullying also incorporates indirect or relational behaviors that involve third parties, such as spreading rumors, manipulating friendship net- works, and deliberately excluding or isolating the bul- lied child from joint activities.

The primary conceptual expansion of “bullying” has been horizontal. One extension that has attracted great media attention is “cyber-bullying,” understood as bullying behavior conducted using the Internet or mobile technologies (e.g., making threats, spreading rumors, posting offensive images). Although some of the online behaviors that qualify as cyber-bullying closely resemble traditional verbal bullying, con- ducted through a new medium, others are distinctive. Although surveys indicate that cyber-bullying may not be as common as “traditional” bullying, and that relatively few children who have been cyber-bullied have not also been bullied traditionally, it never- theless represents a new class of behaviors that stretch the concept of bullying beyond its earlier meaning.

An even more striking horizontal extension of the bullying concept is its growing application to adult workplaces (e.g., Salin, 2003). Bullying is now

researched and studied in organizational contexts almost as much as in schoolyards. According to the Web of Science database, 12.7% of the articles men- tioning bullying published in the 1990s were in devel- opmental psychology journals, holding relatively constant at 14.0% in the 2000s, and 13.3% in the 2010s. Across these decades the share of articles pub- lished in occupational and organizational psychology journals rose from an insignificant 1.3% to 8.8% and then 10.8%.

Bullying in workplaces resembles school bullying at least superficially, and the same elements of physi- cal, verbal, and indirect or relational bullying may occur, although physical intimidation is probably less frequent and verbal harassment more subtle than playground name-calling. However, even if analo- gous behavior occurs in the two settings, a horizontal concept extension has occurred because phenomena that would once have been conceptualized as some- thing other than bullying (e.g., repeated public denun- ciations of employees by supervisors) now fall under its umbrella.

A third form of horizontal creep of the bullying concept involves types of behavior rather than medium or setting. Although early definitions emphasized direct physical and verbal forms of school bullying, current definitions include behav- iors that are not directed at the bullied child but operate by manipulating relationships with other people. These behaviors include ignoring and excluding children. As Mishna (2012) noted, “It is only fairly recently that indirect and social exclu- sionary forms of peer victimization were labelled as bullying” (p. 41). For instance, one early bullying scale (Neary & Joseph, 1994) omitted indirect forms entirely, referring only to physical (hit and pushed, picked on, and bullied) and verbal (teasing, name- calling, being laughed at) actions. Olweus’s (1989) description of bullying in the original Olweus Bully- ing Questionnaire mentioned a single exclusionary behavior (“a kid is being bullied . . . when no one ever talks to them”) among a large number of physi- cal and verbal behaviors. The recent extension of the bullying concept to give greater attention to exclusion is also evident in definitions of workplace bullying, one checklist giving as an example times when someone “gave you the silent treatment” (Fox & Stallworth, 2005)

The horizontal creep of the bullying concept into electronic media and adult workplaces, and into indirect and exclusionary actions, is striking. How- ever, the concept has also undergone some vertical expansion such that milder or less extreme phenom- ena have come to count as bullying. One form of vertical creep is a relaxation of the repetitiveness criterion in the definition of bullying. To some extent this loosening has been prompted by the

4

HASLAM

advent of cyber-bullying, in which posting a single offensive picture or message may be considered to constitute bullying. However the same relaxation has occurred in the traditional bullying domain as well. The original Olweus Bullying Questionnaire, for example, specified in its instructions to children that “when we talk about bullying, these things hap- pen repeatedly,” but the revised version stated that “these things may happen repeatedly” or “are usually repeated” [emphasis added].

A similar loosening of the definition of bullying can be observed with the power imbalance criterion. Olweus has insisted that this defining element is vital for distinguishing bullying from general peer aggres- sion, but it has been difficult to nail it down. Tradi- tionally power imbalance was understood primarily in terms of size, age, or number, as when one child was victimized by a group, all of these factors making it difficult for victims to defend themselves. More recent understandings consider power imbalance in terms of differential peer-group status, popularity, and even self-confidence (Olweus, 2013), none of which transparently implicate “power” as it is usually understood.

The horizontal creep of bullying into cyberspace and workplaces makes the retention of a restricted view of power imbalance even more difficult. Smith, del Barrio, and Tokunaga (2012) note that in cyber- bullying, where the bully may be anonymous and therefore of unknown relative to the victim, power imbalance can be understood in terms of differences in “technological know-how between perpetrator and victim, relative anonymity, social status, number of friends, or marginalized group position” (p. 36). Other difficulties in restricting the meaning of power imbalance arise in organizations, where power differ- entials based on rank are officially legitimated, unlike schoolyards. The person who is accused of bullying for repeatedly criticizing a subordinate’s work may have formal responsibility for the management of that employee’s performance so that the power imbal- ance is intrinsic to the relationship and to the behavior itself. In addition, definitions of workplace bullying allow for bullies being same-rank coworkers.

A third form of vertical creep can be seen in the relaxation of the intentionality criterion in workplace bullying research. As Salin (2003) observed, “intent is typically not part of the definition, but instead the subjective perception of the victim is stressed” (p. 1215–1216). Thus bullying can be said to occur even if the identified bully had no intent to harm the identi- fied victim. This broadens the traditional concept of bullying by including behavior that might be inadvertent.

This opening of the definition of bullying to the subjectivity of victims arguably represents a fourth form of vertical creep and is also observed in school

bullying scholarship. Olweus (2013), for example, proposed that “the ultimate” power of definition “must reside with the targeted student” (p. 757) as to when a power imbalance occurs. Similarly, Mishna (2012) argued forcefully that victims’ judgments of whether they have been bullied should take prece- dence over those of perpetrators and adult observers, such as parents and teachers. (This principle sits uneasily with Mishna’s acknowledgment that chil- dren may not identify their mistreatment as bullying and may have to be educated into accepting the label.) Thus, if a child perceives social exclusion to have been deliberate, repeated, and hurtful, or “jokes” to have been said with malice rather than jest, then bullying has occurred. In the workplace, similarly, it is victims’ “perceived” power imbalance that is taken as relevant to the definition of bullying and their per- spective may be privileged in deciding what counts as bullying. One measure of experiences of “general bullying behavior” lists a variety of ambiguous behaviors: a person “limited your ability to express an opinion,” “gave excessively harsh criticism of your performance,” “made unreasonable work demands,” and “applied rules and punishments incon- sistently” (Fox & Stallworth, 2005). The inclusion of behaviors such as these within the concept of work- place bullying is pungently criticized by Furedi (2006). Considering an organization’s report on its members’ concerns about the phenomenon, he wrote: “It became clear that what MSF categorizes as bully- ing in the workplace is what used to be called office politics” (p. 87).

To summarize, the concept of bullying has spread from its original meaning to encompass a wider range of phenomena. It has expanded horizontally into online behavior, into adult workplaces, and into forms of social exclusion that do not directly target the victim with hurtful actions, as distinct from hurt- ful omissions. It has also expanded vertically so that behavior that is less extreme than prototypical bully- ing now falls within its bounds, primarily by loosen- ing defining criteria. In some circumstances bullying behavior need not be repeated or intentional, and it need not occur in the context of a power imbalance as traditionally conceived. Greater weight in determin- ing when bullying has occurred is now given to the subjective perceptions of the victim. As a result, “bullying” can now refer to a much greater variety of actions than it did originally.

As Cascardi, Brown, Iannarone, and Cardona (2014) recently observed, the broadening definition of bullying has significant practical consequences. They argued that the relaxation of the original requirement that behavior must occur in a power imbalance and be repeated has eroded distinctions between bullying, on one hand, and harassment and peer aggression, on the other. Cascardi et al. noted

5

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

that the expanded definition of bullying is now inscribed in U.S. state antibullying statutes and can have troubling implications for free speech rights and for schools that could “be required to report and investigate every aggressive transgression, from play- ground teasing and roughhousing to aggravated assault” (p. 255). Equally, they argued, blurred boundaries between bullying, harassment, and peer aggression can lead to inappropriate interventions, as these forms of aggression typically require different therapeutic and legal responses.

Case Study 3: Trauma

Trauma, from the Greek for “wound,” originally referred to a morbid condition of the body produced by a physical insult. Its cause was an external event, and its effect was an organic disturbance that might manifest in psychological symptoms. This meaning was operative in the first edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-I; American Psychiatric Association [APA], 1952), which described a class of “chronic brain disorders associated with trauma,” including disorders caused by “brain trauma, gross force” and “electrical brain trauma.” Related classes of chronic brain disorders were associated with infections, poisoning, and con- genital conditions. A trauma was therefore seen within mid-20th-century psychiatry as a physical agent causing organic brain pathology.

Trauma now refers to a much broader set of phe- nomena, although the earlier meaning persists within it, as in “traumatic brain injury.” DSM-III (APA, 1980) was a turning point, recognizing “post-trau- matic stress disorder” (PTSD) as a mental disorder for the first time. According to the manual, PTSD was a distinctive cluster of symptoms linked causally to a traumatic event. However, in contrast to the DSM-I understanding of trauma, these symptoms were not understood to spring from an organic injury to the brain but from a psychological injury to the mind, caused not by a physical insult but by a dis- tressing experience. This is a classic example of hori- zontal creep.

Although trauma can now refer not only to an event that causes a wound (physical or psychic) but also to the psychological symptoms that result, I focus on the former meaning. Defining what counts as a traumatic event has been an enduring source of controversy in trauma studies. Disagreement is almost inevitable because “there is a continuum of stressor severity and there are no crisp boundaries demarcating ordinary stressors from traumatic stressors” (Weathers & Keane, 2007, p. 108), and because people’s differing percep- tions of an event’s seriousness may determine whether it traumatizes them.

The working definition of “traumatic event” is embodied in Criterion A of the DSM’s diagnostic rules for PTSD (Long & Elhai, 2009). In DSM-III (APA, 1980), Criterion A required that a traumatic event “would evoke significant symptoms of distress in almost everyone” (p. 238) and be “outside the range of usual human experience” (p. 238). It stated that “such common experiences as simple bereave- ment, chronic illness, business losses, or marital con- flict” (p. 247) generally fail to meet this requirement, and listed rape, assault, military combat, natural dis- asters, car accidents, and torture as events that gener- ally succeeded. DSM-III-R (APA, 1987) retained the key elements of this criterion but specified the nature of the distress evoked by the traumatic stressor (“intense fear, terror, and helplessness”; p. 250) and of the threat that the stressor represented. In particu- lar, it noted that this threat could be to one’s kin or friends rather than oneself, that it could involve learn- ing about an event that had afflicted them, and that witnessing serious injury or death in another person could also count as a traumatic event. DSM-III-R therefore expanded the definition of trauma to include indirect exposures. DSM-IV (APA, 1994) continued to include indirect exposures as potential traumatic events—it lists as one example learning about the diagnosis of a life-threatening illness in one’s child— and increased the emphasis on the subjectivity of the traumatic stressor by introducing a new criterion con- cerning the distress experienced in response to the stressor.

Although the revisions of Criterion A from DSM- III to DSM-IV seem relatively subtle, they have been criticized for broadening the definition of trauma and described as “conceptual bracket creep” by one critic (McNally, 2004). In particular, the opening up of indirect experiences as traumas represents an enlarge- ment of the original concept. A study by Breslau and Kessler (2001) found that only 14 of 19 experiences that would qualify as potentially traumatic by DSM-IV’s Criterion A1 would have met Criterion A in DSM-III-R, all of the remainder representing indi- rect exposures. The later version of Criterion A led to a 22% increase in the number of traumatic events to which their sample had been exposed. Weathers and Keane (2007) also argued that DSM-IV’s listing of “developmentally inappropriate sexual experiences” as potential traumatic events represents a break with the earlier understanding that traumas must involve threats of serious injury or death.

This trend toward including indirect and noncata- strophic events within the definition of trauma—which has been partially arrested by DSM-5 (Zoellner, Bedard-Gilligan, Jun, Marks, & Garcia, 2013)—exem- plifies vertical expansion. It has led to an enlargement of the range of events that are recognized as potential triggers of PTSD. In recent years, trauma theorists and

6

HASLAM

practitioners have proposed including childbirth, sexual harassment, infidelity, and emotional losses such as abandonment by a spouse or loss or a sudden move or loss of home within that range. These extensions are sometimes justified empirically by research showing that these events can precipitate PTSD symptoms (e.g., Carlson, Smith, & Dalenberg, 2013). Nevertheless, they represent a lowering of the threshold of severity for traumatic events.

A recent definition of trauma produced by the U.S. Government’s Substance Abuse and Mental Health Services Administration exemplifies this lowering:

Individual trauma results from an event, series of events, or set of circumstances that is experienced by an individual as physically or emotionally harmful or threatening and that has lasting adverse effects on the individual’s functioning and physical, social, emotional, or spiritual well-being.

This definition abandons most of the restrictive ele- ments of DSM’s Criterion A. A traumatic event need not be a discrete event, need not involve serious threats to life or limb, need not be outside normal experience, need not be likely to create marked distress in almost everyone, and need not even produce marked distress in the traumatized person, who must merely experi- ence it as “harmful.” Under this definition the concept of trauma is rendered much broader and more subjec- tive than it was even three decades ago.

Case Study 4: Mental Disorder

The proliferation of mental disorders in successive editions of the DSM is well known. The DSM’s pre- cursor, War Department Technical Bulletin “Medical 203,” published in 1943, listed 47 conditions, a num- ber that more than doubled to 106 diagnoses in DSM- I (APA, 1952), leapt to 182 in DSM-II (APA, 1968), jumped to 265 in DSM-III (APA, 1980), and then hopped to more than 300 in the revised edition of the DSM-IV (DSM-IV-TR; APA, 2000). The swelling population of mental disorders has led critics of psy- chiatry to accuse DSM of disease-mongering.

DSM’s expanding cast of mental disorders need not reveal an expansion in the concept of “mental dis- order.” DSM-IV-TR might map the same psychopath- ological territory as DSM-I but in a more fine- grained, high-resolution manner. Later editions of DSM might merely split conditions recognized in ear- lier conditions into narrower variants rather than expanding into new territory as horizontal creep would require. However, a comparison of the earliest and more recent editions of DSM demonstrates that successive DSMs not only subdivide existing disor- ders but also open up new psychiatric terrain.

DSM-I contained seven groupings of mental disor- ders: acute and chronic brain disorders, mental defi- ciency, psychotic disorders, psychophysiologic disorders, psychoneurotic disorders, personality dis- orders (which included addiction), and transient situa- tional personality disorders. DSM-II expanded the range of psychiatric conditions in three ways. First, it introduced a new “special symptoms” grouping that included problems with sleep and eating, domains that were not covered in the earlier edition. Second, it extended the range of conditions afflicting young people beyond DSM-I’s “mental deficiency” cate- gory, recognizing a new grouping of behavior disor- ders of childhood and adolescence. Third, it added sexual deviations to DSM-I’s list of personality disorders.

DSM-III divided up several of DSM-II’s disorder groupings: It carved off substance-related disorders and sexual disorders from personality disorders, split the special symptoms grouping into separate eating and sleep disorder categories, and cleaved psycho- neurotic conditions into separate anxiety and mood disorder groups. However, in addition to these divi- sions of existing disorder groupings, DSM-III also pushed back the psychiatric frontier by recognizing new kinds of disorder in a clear demonstration of hor- izontal creep. New groupings of factitious, impulse- control, and dissociative disorders were defined, none of their conditions corresponding in a straightforward way to those described in previous DSM editions. Further horizontal expansion can be seen in DSM- III’s acquisition of new conditions within groupings recognized in DSM-II. Disorders involving cognitive difficulties were added to the disorders first diagnosed in childhood and adolescence, sexual disorders were expanded to include gender identity disorder (a con- dition of gender, not sexuality), anxiety disorders incorporated social fears and extreme shyness for the first time (“social phobia”; see Lane, 2008), and sub- stance-related disorders expanded to include prob- lematic usage (substance abuse) that fell short of addiction or dependence.

As a result of this consistent pattern of diagnostic spread, many people whose clinical presentation would not have warranted a DSM-I diagnosis—alco- hol abusers, insomniacs, bulimics, Touretters, gender dysphorics, anorgasmic women, dyslexic children, and shy adults—would have received a DSM-III diag- nosis. DSM-IV and DSM-5 have introduced further horizontal creep, which is not reviewed here. The key point is that successive editions of the manual have progressively dilated the ostensive definition of men- tal disorder. Although the prototypical psychiatric conditions are continuously represented in some fash- ion from DSM-I through DSM-5, new domains of psychopathology have been added. Phenomena that might previously have been understood as moral

7

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

failings (e.g., substance abuse) bad habits (e.g., eating problems), personal weaknesses (e.g., sexual dys- functions), medical problems (e.g., sleep disturban- ces), character foibles (e.g., shyness), or ordinary vicissitudes of childhood now find shelter under the umbrella concept of mental disorder.

The expanding register of mental disorders indi- cates horizontal creep, but the concept of mental dis- order has also undergone vertical creep. Recent editions of DSM sometimes loosen the criteria for determining where normality ends and mental disor- der begins. This quantitative easing allows milder, less disabling psychological phenomena to qualify as disordered. Sometimes this relaxation of criteria takes the form of recognizing less severe “spectrum” condi- tions, as with cyclothymia, a less impairing variant of bipolar disorder, and Asperger’s syndrome, a less impairing variant of autistic disorder, which has recently been reincorporated in the latter diagnosis, thereby vertically expanding it.

Especially powerful cases for such a lowering of diagnostic thresholds have been made by Horwitz and Wakefield (2007, 2012) in their historical studies of depression and anxiety disorders. They argued that recent ways of diagnosing these conditions systemati- cally misdiagnose normal affective responses as forms of psychopathology. For example, symptom- based diagnosis of depression conflates contextually justified sadness with melancholia, the more restric- tive traditional understanding of depression as “sadness without cause,” resulting in a recent explo- sion of diagnosed depression (Shorter, 2013). By mis- representing normal sadness, worry, and fear as mental disorders, the mental health professions over- medicate, exaggerate the population prevalence of disorder, and deflect resources away from more severe conditions.

Much of the recent controversy surrounding DSM-5 (2013) concerns this vertical creep and the attendant risks of overdiagnosis and inappropriate treatment. It prompted the architect of DSM-IV, Allen Frances, to launch a crusade to “save normal- ity” from the new edition (Frances, 2013). Several of the targets of his campaign were clear examples of relaxed diagnostic rules or new diagnoses with less stringent criteria than those they replaced. The removal of the bereavement exclusion for major depression, which allowed grieving people to receive a diagnosis of depression within 2 months of the death when in the previous edition they could not, is one example (Wakefield, Schmitz, First, & Horwitz, 2007). Others are the relaxation of rules for diagnosing attention deficit hyperactiv- ity disorder among adults; the listing of a new “somatic symptom disorder” for people who had unusually strong worries about physical symptoms that fell well short of previous diagnostic criteria

for hypochondriasis; and the advent of “mild neuro- cognitive disorder,” a sort of “dementia lite.”

In sum, the evolving concept of mental disorder has not only proliferated conditions but also expanded sideways into new forms of psychopathol- ogy and downward into milder forms. As a result, the proportion of humanity warranting a diagnosis has risen and the proportion of human experience and behavior that counts as disordered has swelled.

Case Study 5: Addiction

Addiction can be considered under the rubric of mental disorder, but it warrants its own discussion, in part because the concept does not feature in many classifications or disorder. The meaning of that con- cept within psychology, psychiatry, and general med- icine in the first half of the 20th century involved physiological dependence on an ingested psychoac- tive substance. The pharmacological properties of the substance lead the addicted person to require progres- sively more of it to attain the desired state (tolerance) and cause the person to experience an unpleasant physiological state when deprived of it (withdrawal). This dependency creates an increasingly joyless pat- tern of compulsive consumption. The classic addic- tive drugs—alcohol, cocaine, heroin, nicotine—all have properties that interact with the human body’s motivational apparatus to generate these effects.

In recent decades the concept of addiction has been enlarged by the identification of addictions that do not involve substances. So-called “behavioral” or “process” addictions to the Internet and gambling have been proposed in the mental health literatures (e.g., Potenza, 2006). For the first time compulsive gambling (“gambling disorder”) has been officially recognized within a class of “substance-related and addictive disorders” in DSM-5 (APA, 2013), having previously resided in a DSM-IV’s grouping of “impulse control disorders.” DSM-5’s developers entertained the idea of including a wider assortment of behavioral addictions, and popular and academic writers have provided many to choose from, propos- ing behavioral addictions to sex, pornography, shop- ping, online gaming, food, chocolate, exercise, social media, TV, work, and tanning, among other things.

The recognition of behavioral addictions reflects in part a growing acknowledgment among scientists that certain compulsive behaviors overlap substantially with substance addictions in their phenomenology, neurobiol- ogy, natural history, personality correlates, and response to treatment (Grant, Potenza, Weinstein, & Gorelick, 2010). Like substance addictions, behavioral addictions involve recurrent failure to resist urges to engage in a particular activity that is harmful to the person, generally with a subjective experience of compulsion and

8

HASLAM

powerlessness. They may involve withdrawal—and tol- erance-like experiences and repeated unsuccessful attempts to quit. In view of these similarities, it has been argued that what matters for addiction is not the mani- fest substance or behavior but the underlying process whereby an activity that can provide gratification and escape from discomfort becomes something over which the person loses control and continues despite negative consequences (Goodman, 1991). If the addictive process is internal, it makes little sense to restrict the concept of addiction to dependence on substances.

In addition to expanding horizontally to include behavioral addictions alongside the earlier substance addictions, recent developments in psychology have also opened the concept to milder, less extreme forms of compulsive behavior. This vertical expansion is illustrated well by the concept of “soft” addictions (Wright, 2006), which represent persistent activities that carry some cost in money, time, energy, or inti- macy. Soft addictions lack the sense of powerlessness, dependency, and compulsion that is typical of standard addictions and the harm they cause is relatively innoc- uous. By recognizing them as addictions that concept creeps downward into the realm of bad habits and other repetitive pleasurable behaviors where there is a risk of inappropriate diagnosis. As Petry (2006) warned, “One must be cautious of where to draw the line between simply an excessive behavior pattern and a bona fide psychiatric disorder” (p. 157).

The concept of addiction has clearly undergone a substantial semantic enlargement in the last few deca- des, both horizontal and vertical. It can be argued that this expansion represents something of a return to an earlier understanding of addiction as being “given over” some activity. Alexander and Schweighofer (1988) argued that this broad understanding of addic- tion, which could include being intensely involved with desirable activities such as the reading of books, prevailed until the temperance movement installed a restrictive, disease-based definition in the mid-19th century. Even if it is true that the meaning of addic- tion constricted before it swelled, it is the expansion that has taken place in recent decades that concerns this article. Some elements of that expansion have a solid scientific basis, grounded in the many common- alities of substance and behavioral addictions, but the key point for my argument is the fact of the expan- sion, not whether it is well justified.

Case Study 6: Prejudice

In examining concept creep in the domain of prej- udice, we move from the domain of clinical psychol- ogy to social psychology. Prejudice is one of the most well-researched topics within the field, repre- senting a form of intergroup animosity that social

psychologists have been eager to study, theorize, and address since Allport’s (1954) The Nature of Preju- dice, the seminal work on the subject. As Dixon, Lev- ine, Reicher, and Durrheim (2012) documented, Allport understood prejudice to involve intergroup antipathy: The prejudiced person holds hostile atti- tudes toward members of an outgroup. This definition of prejudice as negative evaluation of outgroups has persisted, featuring in many textbook definitions. However, Dixon and colleagues argued that the idea of prejudice-as-antipathy may not be sufficient, and social psychological accounts of prejudice in the past three decades have begun to broaden it.

Early social psychological researchers began with an understanding of prejudice as blatant bigotry, examining endorsement of hostile and derogatory statements about African Americans, Jews, and others. However, as rates of endorsement of these statements began to wane later in the 20th century, the understanding of prejudice was broadened. McConaghy (1986) drew a distinction between “old- fashioned” racism, exemplified by endorsement of explicit bigotry, and a subtler and more prevalent “modern” racism. Modern racists, like so-called “symbolic” racists (Sears, Henry, & Kosterman, 2000), do not endorse direct hostility to traditional targets of prejudice but instead denied the continuing existence of racism and expressed opposition to affir- mative action policies. It was possible to score high on a questionnaire measure of modern racism, and later sexism, without agreeing with any derogatory evaluations of the target group. Nevertheless, such scores were taken to indicate prejudice because they were conceptualized as revealing tacit negative evalu- ations and were associated with other indicators of prejudice, such as discriminatory behavior.

The ideas of modern and symbolic racism extended the concept of prejudice from direct, expressed antipathy to a group to inferred antipathy: Modern and symbolic racists held hostile attitudes toward racial outgroups but suppressed them, know- ing better than to reveal socially prohibited senti- ments on questionnaires. Two new ideas extended the concept of prejudice even further. The concept of aversive prejudice (Dovidio & Gaertner, 2004) applies to liberally minded people who deny personal prejudice but hold aversions, sometimes unconscious, to other-race people. These aversions are not based on hostile antipathy by on fear, unease, or discomfort. The related concept of implicit prejudice (Dovidio, Kawakami, & Gaertner, 2002) referred to people who unconsciously linked negative concepts with racial minority groups more strongly than with majority groups, as demonstrated by tasks such as the Implicit Association Test. The ideas of implicit and aversive prejudice can be prejudiced not only if they suppress their negative racial sentiments but even if they are

9

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

unaware of having them. Just as some critics had rejected the claim that modern or symbolic racism is true prejudice, arguing that it could merely indicate sincerely held conservative opinions (Sniderman & Tetlock, 1986), later critics challenged the view that automatic associations between racial outgroups and negative stimuli necessarily implied prejudice as dis- tinct from knowledge of widely held stereotypes (Arkes & Tetlock, 2004).

Although modern, symbolic, aversive, and implicit prejudices are less blatant and hostile than the old-fashioned variant that dominated early prejudice research, they retain the view that prejudice involves negative group evaluations. Further extensions of the concept of prejudice relaxed this requirement. The concept of benevolent sexism (Glick & Fiske, 1996) extended prejudice to include group evaluations that were at least superficially warm and positive. Benev- olent sexists idealize women as pure creatures who are too delicate and morally superior to inhabit the hurly-burly public world of men.

All of the forms of prejudice just reviewed are usu- ally understood from the standpoint of the perpetrator of prejudice. Particular social actors are prejudiced, and their attitudes are objective elements of their psy- chology. However, some research implies that preju- dice exists at least in part in the eyes of the target. Research on microaggressions (Sue et al., 2007), for example, takes the target’s perceptions of prejudice as clear evidence of its existence: If a target perceives a slight as evidence of prejudice, then it is taken as such, even if the slight is ambiguous and its author denies it. Of course, many prejudiced acts are unam- biguous, target perceptions may tend to be accurate, and denials of prejudice are frequently not credible. Nevertheless, to count perceived discrimination and ambiguous microaggressions as unqualified instances of prejudice is to subjectivize the concept. In addition to this subjectivity, the concept of microaggression extends the concept of prejudice by encompassing acts of omission and phenomena that reflect anxiety rather than hostility. Proposed examples of microag- gression include the faltering speech, trembling voice, and mispronunciation of words by anxious White therapists discussing racial issues with minor- ity clients, and “the sheer exclusion of decorations or literature that represents various racial groups” in environments that they inhabit (Sue et al., 2007, p. 274).

These more recent understandings of prejudice extend its meaning far beyond the blatant antipa- thy that was the original phenomenon of interest to Allport and other early social psychologists. Prejudice is no longer exclusively blatant, but can be subtle and nonconscious. It is not necessarily hostile, but can be anxiously avoidant or patronis- ingly positive. It may not even be inherent in the

acts or attitudes of a prejudiced person, existing instead in another person’s perception. This expan- sion of the meaning of prejudice reflects a process of vertical concept creep in which the concept’s elastic boundaries stretch to include increasingly mild and subtle phenomena.

Examples of horizontal creep also exist in the domain of prejudice, as new forms are recognized that early psychologists would not have dreamed of. Those psychologists primarily studied varieties of racism, including anti-Semitism, whereas researchers now also study prejudices based on sexual orientation, gender identity, religion, physical appearance and stature, marital status, and even species. (Some of these preju- dices—homophobia, transphobia, Islamophobia—also illustrate the horizontal creep of the concept of phobia from irrational fear to attitudinal aversion.) The expan- sion of the concept of prejudice has been both horizon- tal and vertical.

It is important to reiterate here that by document- ing the expanding meaning of prejudice in recent social psychology I am not questioning the validity of this expansion or advocating a return to a narrower understanding of the concept. Each extension of the concept of prejudice was arguably well justified. The idea of modern prejudice was justified by changing social conditions: public disapproval of racial bigotry drove blatant prejudice underground, where it contin- ued to promote discriminatory behavior. The notion of implicit prejudice was justified by emerging evi- dence of the influence of nonconscious processes on social behavior. The failure of social integration in the absence of intergroup antipathy gave a warrant to the idea of aversive prejudice. “Benevolent sexism” was justified by the realization that patronizing atti- tudes protect women from full participation in the public sphere rather than from their own fragility. Ideas of perceiver-defined prejudice and microag- gression can be justified by evidence that displays of prejudice can have harmful effects on their targets while remaining subtle and deniable. It is possible to define “prejudice” in a coherent way that encom- passes all of these variants. My point is simply that the concept now refers to much more than it did several decades ago.

Overview of the Case Studies

These six conceptual case studies replicate a pattern of semantic enlargement and point to a few consistent ways in which it has taken place. First, most of the concepts have stretched to include milder, subtler, or less extreme phenomena than those to which they referred at an earlier time. This stretching is evident in definitions of abuse that can count angry arguments as instances of emotional abuse, definitions of bullying

10

HASLAM

that can include once-off displays of office tyranny, relaxed diagnostic criteria for mental disorders such as depression, the recognition of nonconscious forms of prejudice, and definitions of trauma that allow vicari- ous experiences to be counted as traumatic.

Second, some of the concepts that initially referred to the commission of undesirable acts have stretched to include acts of omission and avoidance. This pat- tern is illustrated by the inclusion of neglect within the concept of abuse, the growing recognition of exclusionary forms of bullying, and the proposal of aversive forms of prejudice that involve avoidance of others rather than hostile attack.

Finally, several concepts have acquired a more subjective aspect. Emotional abuse may be claimed if one party feels abused rather than by a set of objec- tive abusive behaviors, bullying if employees per- ceives that their work has been criticized too harshly, prejudice if its target perceives it despite the sincere protestations of the perpetrator, and trauma if its vic- tim experiences classic posttraumatic symptoms even if the triggering event would not qualify as traumatic on classical criteria.

In sum, then, conceptual creep has occurred across a diverse assortment of concepts and has commonly involved an increased sensitivity to negative experi- ence and behavior, an increased focus on harmful forms of inaction, and an increased acceptance of sub- jective criteria for deciding when the concepts apply.

Explaining Concept Creep

The preceding sections illustrate how the mean- ings of some of psychology’s concepts have dilated in recent decades. The six case studies reveal similar patterns of horizontal and vertical expansion in con- cepts drawn from developmental, social, and clinical psychology. The consistency of this pattern across diverse concepts calls for a generalized explanation rather than separate explanations for each. For exam- ple, although the expansion of mental disorder could be attributed to the rise of “therapism” (Sommers & Satel, 2005) or medicalization (Frances, 2013), the expansion of abuse and trauma to an emerging “culture of fear” that is preoccupied with risk (Furedi, 2006), and the expansion of prejudice to “political correctness,” the fact that similar expansions of meaning occur across these disparate concepts sug- gests that a more general phenomenon may be implicated.

A strong and parsimonious explanation of concept creep should therefore be capable of accounting for the generality of the phenomenon. It should also pro- vide satisfactory answers to three key questions. First, it should be able to account for why semantic expan- sion rather than contraction occurs. If human kinds

are intrinsically fluid, as Hacking maintained, why have their meanings spread rather than receded? Sec- ond, a good explanation for concept creep should account for why the conceptual expansion is asym- metrical, evident only for negative concepts. It is dif- ficult to find examples of semantic inflation among psychology’s positive concepts, with the arguable exception of intelligence. Third, an adequate explana- tion of concept creep should be capable of accounting for both horizontal and vertical expansion, the quali- tative and quantitative aspects of the phenomenon.

One possible explanation of concept creep is that the concepts in question are human kinds in Hacking’s sense rather than natural kinds. None of them refer to timeless categories that are independent of social convention and practice. Similarly, none of them have sharp, discoverable category boundaries that psychology’s definitions might try to approxi- mate. For example, most mental disorders—not to mention the concept of mental disorder itself—fall on a continuum with normality and with other disorders: The psychiatric domain is a spectral blur rather than a collection of discrete conditions (Haslam, Holland, & Kuppens, 2012). Definitions of human kinds are therefore bound to be in flux.

The fact that human kinds are not natural kinds explains why psychology’s concepts may have elas- tic meanings, but it does not explain concept creep adequately. Having elastic conceptual boundaries does not dictate that those concepts should expand. In principle, concepts with intrinsically flexible definitions should be just as likely to contract their meanings. The unnaturalness of human kinds also fails to explain why concept creep is most evident for negative concepts or why it takes horizontal and vertical forms. Finally, it does not offer a uni- fying explanation for the common pattern of expan- sion across concepts. The fact that psychology’s concepts are human kinds enables creep but does not explain it.

Another class of explanations invokes technologi- cal and societal change. By these accounts, concept creep is a response to altered social realities. Cyber-bullying and Internet addiction, for example, could not exist before the emergence of the Internet. The horizontal expansion of the concepts of bullying and addiction to include them simply reflects the emergence of a new online medium in which intimi- dation and dependency can occur. Similarly, the hori- zontal creep of the concept of prejudice could be attributed to the rising prominence of minority groups, such as transgendered people and Muslims, in societies where they had not previously been salient. In these instances, concept creep is a side effect of large-scale societal changes, in which new forms of an existing phenomenon are added by accretion.

11

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

Explanations of this kind may have some merit, but they also have limitations. It is difficult to identify a single technological or other societal development that could drive the generalized expansion of con- cepts as disparate as bullying, mental disorder, and prejudice. Although accounts that invoke such devel- opments can explain why concepts tend to broaden rather than narrow, as with cyber-bullying, they fail to account for why that broadening should be clearer for negative concepts. Unless technological and soci- etal change has uniformly negative implications, it should also broaden positive concepts. In addition, such change can explain horizontal expansion—the addition of new forms of experience or behavior enabled by new technologies or social conditions— but it is unclear how it could account for vertical expansion. Why technological or societal change should lower the threshold for detecting established forms of behavior or experience is unclear.

A third kind of explanation for concept creep implicates psychology as a field of knowledge rather than societal or technological change. It could be argued that just as successful species increase their territory, invading and adapting to new habitats, suc- cessful concepts and disciplines also expand their range into new semantic niches. Concepts that suc- cessfully attract the attention of researchers and prac- titioners are more likely to be applied in new ways and new contexts than those that do not. For example, the success of the concept of bullying in the develop- mental psychology literature of the 1970s and 1980s may have made it an appealing concept to apply to analogous behavior observed in workplaces for schol- ars working in the 1990s. This tendency for success- ful concepts to colonize new semantic territory has several dimensions. Concepts that are particularly illuminating may be more readily extended by anal- ogy, and those that receive more academic attention may be more readily extended as researchers strive to make novel contributions in a crowded marketplace of ideas. The former possibility implicates the intrin- sic properties of successful concepts, whereas the lat- ter implicates the dynamics of popular research topics.

Figure 1 shows how many of the creeping con- cepts from the case studies have indeed become more prominent in recent decades. Panels (a) to (e) present data on the relative frequency with which particular words appear in Google’s Ngram database of approx- imately 5 million books (Michel et al., 2011), reported separately for books published at 10 time- points from 1960 to 2005. These relative frequencies are scaled so that 100 represents the highest relative frequency for each word across the 10 time-points (i.e., the highest relative frequency of the word over the period in question). In every case, the relative fre- quency of words associated with abuse, bullying,

Figure 1. Rising usage of words denoting (a) abuse, (b) bullying, (c) trauma, (d) addiction, and (e) racism from 1960 to 2005, as a propor-

tion of all words in the Google NGram corpus, and static usage of (f)

psychology. Values are scaled so that 100 represents the highest pro-

portion obtained for the term at any of the 10 time-points.

12

HASLAM

trauma, addiction, and prejudice has risen steeply, especially in the period 1970–2000, reaching peak salience in the most recent time-points. If increases in the salience of particular concepts can drive the expansion of concept meanings, these graphs demon- strate the plausibility of this mechanism of concept creep.

The “Darwinian concepts” account, according to which more successful concepts tend to expand their semantic range, has significant promise as an explana- tion of concept creep. Psychological concepts have had unquestionable cultural success in recent decades, with writers pointing to the rising “psychologization” of experience in postwar society (De Vos, 2010). The discipline has grown steeply in public influence, research output, professional numbers, and undergrad- uate enrolments, and as Horwitz and Wakefield (2007) observe, “all professions strive to broaden the realm of phenomena subject to their control” (p. 213). The psy- chologization account can explain why several nega- tive concepts have expanded rather than contracted their semantic range. It can account for what ties these conceptual changes together as a general phenomenon: all of the concepts in question are prominent psycho- logical ideas. It may also account for the asymmetrical tendency for negative rather than concepts to expand, in view of the discipline’s disproportionate emphasis on negative phenomena (Gable & Haidt, 2005). Argu- ably the psychologization account can also explain why both vertical and horizontal forms of expansion have taken place. It proposes that successful and popu- lar concepts have a tendency to inflate, and that ten- dency could be expressed equally well by bringing new phenomena under the concept’s semantic umbrella or by extending the concept downward to more subtle phenomena.

Although an explanation of concept creep that emphasizes factors internal to the discipline of psy- chology is promising, there is evidence that the psy- chologization account is insufficient. Figure 1’s panel (f), for example, shows that psychology itself, at least as indicated by the relative frequency of use of the word “psychology” and “psychological,” has not become more salient during the period when the creeping concepts have crept most vigorously, and has in fact declined to some degree. I propose that a broader cultural shift may be implicated in concept creep. That shift can be approached through the anal- ysis of historical trends in violent behavior offered by Pinker (2011). Pinker documents a relentless decline in all forms of violence over several time-scales. Over the recent decades that are the focus of my anal- ysis, Pinker identifies the “rights revolutions” as the main driver of this reduction. Movements for the rights of women and minorities have conducted what he describes as a “civilizing offensive” targeting forms of aggression and inequity that had previously

been accepted. Their campaigns “are propelled by an escalating sensitivity to new forms of harm” (p. 460). Pinker argues that although the prescriptions of these campaigns can seem excessive—as an example he gives the prohibition of schoolyard dodgeball—they simply reflect an overshooting of that well-motivated sensitivity to harm.

Pinker’s analysis offers a way to comprehend con- cept creep in terms of wider cultural trends rather than those that originate within psychology. It can explain why conceptual change should be expansion- ary and why it should apply asymmetrically to nega- tive (i.e., harm-related) concepts. Its claim that the driving force in the phenomenon is a rising sensitivity to “new forms of harm” (horizontal expansion) and to “the slightest trace of a mindset that might lead to it” (p. 469) (vertical expansion) indicates that it can explain both dimensions of creep. The only limitation of Pinker’s analysis as an account of concept creep is that it is specifically addressed to the topic of violence rather than negativity in general. The idea that the rights revolutions were based on “a rising abhorrence to violence” (p. 469) can help to explain the expan- sion of concepts directly related to violence such as abuse and bullying, and it could be stretched to explain the enlargement of less directly related con- cepts such as prejudice and trauma. However a vio- lence-based analysis cannot account for the expansion of unrelated concepts such as mental disor- der and addiction. Nevertheless, an extension of Pinker’s argument that sees the “civilizing offensive” as involving not only a desire to eradicate violence but also a general sensitization to harm could go a long way toward accounting for concept creep.

I would contend that concept creep represents a combination of intellectual forces at work within psy- chology and cultural trends as work within society at large. These forces and trends overlap. Psychology has played a role in the liberal agenda of sensitivity to harm and responsiveness to the harmed, and the growth of the field and its increased focus on negative phenomena—social and personal harms such as abuse, addiction, bullying, mental disorder, prejudice, and trauma—has been symptomatic of the success of that social agenda.

Implications of Conceptual Creep

If concept creep is real, how are its implications to be evaluated? What ‘looping effects’ might the altered extensions of psychology’s negative concepts have on the self-understanding of people to whom they apply and on the views of the wider public? Those drawn to a pessimistic assessment of these changes might argue that the expanding meaning of concepts such as abuse, bullying, and mental disorder

13

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

is creating a culture of weakness, fragility, and excuse-making, in which everyone is a victim and no one is responsible for their predicament. Those drawn to a more optimistic assessment might applaud the growing sensitivity to suffering and maltreatment. A balanced evaluation of concept creep would be more ambivalent, falling somewhere between conservative reaction and liberal celebration.

The nature of this ambivalence is well captured by the work of moral psychologists. Concept creep can be seen as a form of expansion of the ‘moral circle’ (Laham, 2009; Singer, 1981), representing an enlarge- ment of the sphere of people whose experiences or behavior are recognized by psychologists as deserving of moral concern. I have argued that the basis of this expansion is a rising sensitivity to harm, implicating the harm/care moral foundation (Graham, Nosek, Haidt, Iyer, Koleva, & Ditto, 2011). This foundation is strongly associated with political liberalism (Graham, Haidt, & Nosek, 2009), as well as with empathy and compassion (Graham et al., 2011). In essence, the con- cept creep phenomenon broadens moral concern in a way that aligns with a liberal social agenda by defining new kinds of experience as harming and new classes of people as harmed, and it identifies these people as need- ful of care and protection.

As an expansion of the moral circle into new and milder forms of harm, concept creep might appear to be an entirely beneficial sign of moral progress. It defines previously tolerated forms of abusive, domineering, and discriminatory behavior as problematic, and extends pro- fessional care to people who experience adversity and suffering that would once have been ignored. However, by increasing the range of people who are defined as moral patients—people worthy of moral concern, based on their perceived capacity to suffer and be harmed (H. M. Gray, Gray, & Wegner, 2007)—it risks reducing the range of people who see themselves as capable of moral agency. According to research on ‘moral typecasting’ (K. Gray & Wegner, 2009), there is an inverse relation- ship between moral patiency and agency, such that peo- ple tend to be typecast either as victims who suffer harm but lack responsibility and the capacity to act inten- tionally, or as perpetrators who are blameworthy but lack the capacity to suffer. A possible adverse looping effect of concept creep is therefore a tendency for more and more people to see themselves as victims who are defined by their suffering, vulnerability, and innocence, and who have diminished agency to overcome their plight. The flip-side of this expanding sense of victim- hood would be a typecast assortment of moral villains: abusers, bullies, bigots, and traumatizers.

Moral typecasting theory helps to understand some of the mixed blessings of concept creep. Consistent with Pinker’s analysis of violence, the expanding reach of psychology’s negative concepts is likely to have civi- lizing effects, sensitizing people to harm and suffering.

Identifying intimidation at work as bullying rather than tolerating it as office politics is an important step towards creating more nurturing workplaces. Similarly, recognizing an extended range of problems as mental disorders, and a wider variety of life events as traumas, promotes treatment over neglect and sympathy over blame, a common problem for people whose troubles are often moralized as signs of personal weakness.

However, these extensions of harm-based moral concern also have a dark side. Expanding the concept of mental disorder can pathologize normal experien- ces, generate over-diagnosis and over-treatment, and engender a sense of diminished agency. There is evi- dence that people understand their psychological problems as psychiatric diseases tend to be more pes- simistic about recovery and less confident of their capacity to exert control over their difficulties (Has- lam & Kvaale, 2015; Lebowitz, 2014). Similarly, identifying a self-destructive behavior as an addiction encourages people to see themselves as powerless in the face of it, a perception that provides an exculpa- tory moral benefit but also a cost in likelihood of self- change. Although gambling is the only behavioral addiction to be ratified by DSM, popular discourse is riddled with supposed addictions to love, sex, mobile phones, and video games. Expansion of negative con- cepts such as addiction within psychology is likely to trickle down to the lay public, shaping the experience and self-understanding of many more people than the psychology profession can influence directly.

Concept creep may have additional adverse side- effects beyond those attributable to moral typecasting. First, by applying concepts of abuse, bullying, and trauma to less severe and clearly defined actions and events, and by increasingly including subjective ele- ments into them, concept creep may release a flood of unjustified accusations and litigation, as well as exces- sive and disproportionate enforcement regimes (Cascardi et al., 2014). Second, concept creep can produce a kind of semantic dilution. If a concept expands to encompass less extreme phenomena than it did previously, then its prototypical meaning is likely to shift in that direction. If trauma, for example, ceases to refer exclusively to terri- fying events that are outside normal human experience, and is applied to less severe and more prevalent stresses, then it will come to be seen in a more benign light. As Weathers andKeane (2007) remark:

it is essential to set a threshold of stressor severity as part of the diagnostic criteria for PTSD. Doing other- wise would result in a substantial departure from the original conceptualization of PTSD and risk trivializ- ing the suffering of those exposed to catastrophic life events (p. 114)

Similarly, people whose experience fits the more stringent early definition of other negative

14

HASLAM

concepts—who have suffered sexual abuse, melan- cholic depression, hostile bigotry—may find that experience downplayed or trivialized when it is equated to the less severe experiences that fit under the new, expanded definition. In extreme situations, the meaning of negative concepts can be almost completely debased, as when people describe their ordinary, transient sadness as ‘depression.’

Conceptual expansion also carries risks for psy- chological research and practice. The enlarged mean- ings of the concepts discussed in this paper have produced semantic overlaps, and increasingly these concepts can be applied to the same phenomena. An episode in which one person uses an ethnic slur towards another can now count as abuse, bullying, trauma, and prejudice by some definitions. The verti- cal expansion of trauma to include relatively mild maltreatments and of bullying to encompass single incidents, and the horizontal expansion of abuse to include events causing emotional harm, mean that their meanings have become somewhat redundant. This redundancy can lead to conceptual confusion and parallel literatures that address similar phenom- ena using different terminology.

A final risk of conceptual creep concerns the pub- lic’s view of psychology itself. The more the field is seen to traffic in concepts that emphasize the undesir- able and pathological in human life, and the more those concepts are seen as encompassing normal expe- rience, psychology will be identified with negativity and a view of people as dysfunctional victims. The positive psychology movement has laid out this cri- tique based on what it sees as the disproportionate focus of the field on negative phenomena, and I sug- gest that the expansion of negative concepts makes this issue even more salient. Concept creep runs the risk of creating a public impression of psychology as a field that exaggerates misery, inflates mental disorder, excuses misbehavior, and is oversensitive to perceived bias and discrimination. None of these impressions may be justified, but a serious appraisal of conceptual creep must reckon with these potential downsides.

Conclusions

The expansion of psychology’s negative concepts represents a historical development that has signifi- cant implications for the field and for the wider soci- ety that it influences. There is a consistent trend for these concepts to encompass an increasingly broad range of human behavior and experience, and for their meanings to spread and change as a result. Understanding what drives this trend and evaluating its costs and benefits are important goals for people who care about psychology’s place in our cultures. Equally important is the task of deciding whether the

trend should be encouraged, ignored, or resisted. Ulti- mately this question depends on whether we would be content for most interpersonal frictions to be ascribed to abuse and bullying, for everyday stresses to be described as traumas and habits as addictions, for mental disorder to be more common than its absence, and for prejudice to be seen as a constant undercur- rent in social life.

Note

Address correspondence to Nick Haslam, Mel- bourne School of Psychological Sciences, University of Melbourne, Parkville VIC 3010, Australia. E-mail: nhaslam@unimelb.edu.au

References

Alexander, B. K., & Schweighofer, A. R. F. (1988). Defining “addiction.” Canadian Psychology, 29, 151–162.

Allport, G. (1954). The nature of prejudice. New York, NY: Addi-

son-Wesley. American Psychiatric Association. (1952). Diagnostic and statisti-

cal manual of mental disorders. Washington, DC: Author.

American Psychiatric Association. (1968). Diagnostic and statisti-

cal manual of mental disorders (2nd ed.). Washington, DC: Author.

American Psychiatric Association. (1980). Diagnostic and statisti-

cal manual of mental disorders (3rd ed.). Washington, DC:

Author. American Psychiatric Association. (1987). Diagnostic and statisti-

cal manual of mental disorders (3rd ed, revised). Washington,

DC: Author. American Psychiatric Association. (1994). Diagnostic and statisti-

cal manual of mental disorders (4th ed.). Washington, DC:

Author.

American Psychiatric Association. (2000). Diagnostic and statisti- cal manual of mental disorders (4th ed., Text revision).

Washington, DC: Author.

American Psychiatric Association. (2013). Diagnostic and statisti-

cal manual of mental disorders (5th ed.). Washington, DC: Author.

Arkes, H. R., & Tetlock, P. E. (2004). Attributions of implicit prej-

udice, or “Would Jesse Jackson ‘fail’ the Implicit Association Test?” Psychological Inquiry, 15, 257–274.

Breslau, N., & Kessler, R. C. (2001). The stressor criterion in DSM-

IV posttraumatic stress disorder: An empirical investigation.

Biological Psychiatry, 50, 699–704. Carlson, E. B., Smith, S. R., & Dalenberg, C. J. (2013). Can sud-

den, severe emotional loss be a traumatic stressor? Journal of

Trauma & Dissociation, 14, 519–528.

Cascardi, M., Brown, C., Iannarone, M., & Cardona, N. (2014). The problem of overly broad definitions of bullying: Implications

for the schoolhouse, the statehouse and the ivory tower. Jour-

nal of School Violence, 13, 253–276.

Cicchetti, D., & Barnett, D. (1991). Toward the development of a scientific nosology of child maltreatment. In D. Cicchetti, &

W. M. Grove (Eds.), Thinking clearly about psychology:

Essays in honor of Paul E. Meehl, Vol. 2: Personality and psy- chopathology (pp. 346–377). Minneapolis: University of Min-

nesota Press.

15

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS

De Vos, J. (2010). From milgram to Zimbardo: The double birth of

postwar psychology/ psychologization. History of the Human

Sciences, 23, 156–175. Dixon, J., Levine, M., Reicher, S., & Durrheim, K. (2012). Beyond

prejudice: Are negative evaluations the problem and is getting

us to like each other more the solution? Behavioral and Brain Sciences, 35, 411–425.

Dovidio, J. F., & Gaertner, S. L. (2004). Aversive racism. Advances

in Experimental Social Psychology, 36, 1–52.

Dovidio, J., Kawakami, K., & Gaertner, S. L. (2002). Implicit and explicit prejudice and interracial interaction. Journal of Per-

sonality and Social Psychology, 82, 62–68.

Fox, S., & Stallworth, L. E. (2005). Racial/ethnic bullying: Explor-

ing links between bullying and racism in the US workplace. Journal of Vocational Behavior, 66, 438–456.

Frances, A. (2013). Saving normal: An insider’s revolt against out-of-

control psychiatric diagnosis, DSM-5, big Pharma, and the medi-

calization of ordinary life. New York, NY: WilliamMorrow. Furedi, F. (2006). Culture of fear revisited: Risk-taking and the

morality of low expectation (4th ed.). London, UK:

Continuum. Gable, S. L., & Haidt, J. (2005). What (and why) is positive psy-

chology? Review of General Psychology, 9, 103–110.

Glick, P., & Fiske, S. T. (1996). The ambivalent sexism inventory:

Differentiating hostile and benevolent sexism. Journal of Per- sonality and Social Psychology, 70, 491–512.

Goldsmith, R. E., & Freyd, J. J. (2005). Awareness for emotional

abuse. Journal of Emotional Abuse, 5, 95–123.

Goodman, A. (1991). Addiction concept involves theoretical and practical issues. Psychiatric Times, 8, 29–33.

Graham, J., Haidt, J., & Nosek, B. A. (2009). Liberals and con-

servatives rely on different sets of moral foundations. Journal of Personality and Social Psychology, 96, 1029–

1046.

Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., & Ditto, P.

H. (2011). Mapping the moral domain. Journal of Personality and Social Psychology, 101, 366–385

Grant, J. E., Potenza, M. N., Weinstein, A., & Gorelick, D. A.

(2010). Introduction to behavioral addictions. American Jour-

nal of Drug and Alcohol Abuse, 36, 233–241. Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of

mind perception. Science, 315, 619.

Gray, K., & Wegner, D. M. (2009). Moral typecasting: Divergent

perceptions of moral agents and moral patients. Journal of Personality and Social Psychology, 96, 505–520.

Hacking, I. (1986). Making up people. In T. Heller, M. Sosna, & D.

Wellberry (Eds.), Reconstructing individualism (pp. 161– 171). Stanford, CA: Stanford University Press.

Hacking, I. (1991). The making and molding of child abuse. Criti-

cal Inquiry, 17, 253–288.

Hacking, I. (1995). The looping effect of human kinds. In D. Sperber, D. Premack, & A. J. Premack (Eds.), Causal cogni-

tion: A multi-disciplinary debate (pp. 351–383). Oxford, UK:

Clarendon Press.

Haslam, N., Holland, E., & Kuppens, P. (2012). Categories versus dimensions in personality and psychopathology: A quantita-

tive review of taxometric research. Psychological Medicine,

42, 903–920. Haslam, N., & Kvaale, E. (2015). Biogenetic explanations of men-

tal disorder: The mixed blessings model. Current Directions

in Psychological Science, 24, 399–404.

Horwitz, A. V., & Wakefield, J. C. (2007). The loss of sadness: How psychiatry transformed normal sorrow into depressive

disorder. New York, NY: Oxford University Press.

Horwitz, A. V., & Wakefield, J. C. (2012). All we have to fear: Psy-

chiatry’s transformation of natural anxieties into mental dis- orders. New York, NY: Oxford University Press.

Laham, S. M. (2009). Expanding the moral circle: inclusion and

exclusion mindsets and the circle of moral regard. Journal of

Experimental Social Psychology, 45, 250–253. Lane, C. (2008). Shyness: How normal behavior became a sickness.

New Haven, CT: Yale University Press.

Lebowitz, M. S. (2014). Biological conceptualizations of mental disor- ders among affected individuals: A review of correlates and conse-

quences.Clinical Psychology: Science and Practice, 21, 67–83.

Long, M. E., & Elhai, J. D. (2009). Posttraumatic stress disorder’s

traumatic stressor criterion: history, controversy, clinical and legal implications. Psychological Injury and Law, 2, 167–178.

McConahay, J. B. (1986). Modern racism, ambivalence, and the

modern racism scale. In J. F. Dovidio & S. L. Gaertner (Eds.),

Prejudice, discrimination, and racism (pp. 91–125). Orlando, FL: Academic Press.

McNally, R. J. (2004). Conceptual problems with the DSM-IV cri-

teria for posttraumatic stress disorder. In G. M. Rosen (Ed.),

Posttraumatic stress disorder: Issues and controversies (pp. 1–14). New York, NY: Wiley.

Michel, J-B., Shen, Y. K., Aiden, A. P., Veres, A., Gray, M. K.,

Google Books Team, . . . Aiden, E. L. (2011). Quantitative analysis of culture using millions of digitized books. Science,

331, 176–182.

Mishna, F. (2012). Bullying: A guide to research, intervention, and

prevention. New York, NY: Oxford University Press. Moynihan, D. P. (1993). Defining deviancy down. American

Scholar, 6, 17–30.

Neary, A., & Joseph, S. (1994). Peer victimisation and its relation-

ship to self-concept and depression. Personality and Individ- ual Differences, 16, 183–186.

Olweus, D. (1989). The Olweus bully/victim questionnaire. Bergen,

Norway: Department of Psychology, University of Bergen. Olweus, D. (2013). School bullying: Development and some

important challenges. Annual Review of Clinical Psychology,

9, 751–780.

Petry, N. M. (2006). Should the scope of addictive behaviors be broadened to include pathological gambling? Addiction, 101,

152–160.

Pinker, S. (2011). The better angels of our nature: A history of vio-

lence and humanity. London, UK: Penguin. Potenza, M. N. (2006). Should addictive disorders include non-sub-

stance-related conditions? Addiction, 101, 142–151.

Salin, D. (2003). Ways of explaining workplace bullying: A review of

enabling motivating and precipitating structures and processes in the work environment. Human Relations, 56, 1213–1232.

Sears, D. O., Henry, P. J., & Kosterman, R. (2000). Egalitarian val-

ues and contemporary racial politics. In D. O. Sears, J. Sida- nius, & L. Bobo (Eds.), Racialized politics: The debate about

racism in America (pp. 75–117). Chicago, IL: University of

Chicago Press.

Shorter, E. (2013). How everyone became depressed: The rise and fall of the nervous breakdown. New York, NY: Oxford Uni-

versity Press.

Singer, P. (1981). The expanding circle. Oxford, UK: Clarendon Press.

Smith, P. K., del Barrio, C., & Tokunaga, R. (2012). Definitions of bullying and cyberbullying: How useful are the terms? In S.

Bauman, D. Cross, & J. Walker (Eds.), Principles of cyberbul-

lying research: Definitions, measures, and methods (pp. 29–40). Philadelphia, PA: Routledge.

Sniderman, P. M., & Tetlock, P. E. (1986). Symbolic racism: Prob-

lems of motive attribution in political analysis. Journal of

Social Issues, 42, 129–150. Sommers, C. H., & Satel, S. (2005). One nation under therapy:

How the helping culture is eroding self-reliance. New York,

NY: St Martin’s Press.

Sue, D. W., Capodilupo, C. M., Torino, G. C., Bucceri, J. M., Holder, A. M. B., Nadal, K. L., & Esquilin, M. (2007). Racial

16

HASLAM

microaggressions in everyday life: Implications for clinical

practice. American Psychologist, 62, 271–286.

Thompson, A. E., & Kaplan, C. A. (1996). Childhood emotional abuse. British Journal of Psychiatry, 168, 143–148.

Wakefield, J. C., Schmitz, M. F., First, M. B., & Horwitz, A. V.

(2007). Extending the bereavement exclusion for major depression to other losses: Evidence from the National Comor-

bidity Survey. Archives of General Psychiatry, 64, 433–440.

Weathers, F. W., & Keane, T. M. (2007). The criterion A problem

revisited: Controversies and challenges in defining and

measuring psychological trauma. Journal of Traumatic Stress,

20, 107–121.

Wright, J. (2006). The soft addiction solution: Break free of the seemingly harmless habits that keep you from the life you

want. New York: J. P. Tarcher/Penguin.

Zoellner, L. A., Bedard-Gilligan, M. A., Jun, J. J., Marks, L. H., & Garcia, N. M. (2013). The evolving construct of

posttraumatic stress disorder (PTSD): DSM-5 criteria

changes and legal implications. Psychological Injury and

Law, 6, 277–289.

17

CONCEPT CREEP: PSYCHOLOGY’S EXPANDING CONCEPTS