University College of Cape Breton, Sydney, Nova Scotia
ABSTRACT: This paper employs Ian Hacking's notion of interactive kinds
to examine the recent construction of the kind, “depressed adolescent.”I examine first how adolescents themselves were constructed. I then trace how, in North America, we have moved in the past thirty-odd years from a situation of virtually no adolescent depression to the current situation where it is estimated that approximately one in four adolescents is depressed. I offer some reasons why we should be uncomfortable both with the exponential increases in this kind and with the way in which depressed adolescents are being treated at present. In conclusion, I tentatively suggest some ways of proceeding in the future.
he rate of depression amongst adolescents is shocking. Although exact figures vary considerably-between 1.3% and 28%, according to a recent
literature search (Navarette, 1999)-many suggest rates of depression in this age group of approximately 20-25% (Hendricks et. aL, 1999; Reynolds, 1990). Surely, if these estimates are anything close to the truth, they suggest an epidemic of quite staggering proportions. Hence, they raise a variety of questions: Are these figures accurate? On what are they based? How is depression defined and classified? Have these classifications changed over the past ten, twenty, or fifty years? Have rates of depression in adolescents changed during this period? What important consequences follow from these rates of depression? What is being done, and what can be done, to isolate and treat these "ill" young adults. Are these treatments "working"?
In this paper, I will suggest that "adolescent depression" is a construction of fairly recent origin. In arguing toward this conclusion, I will employ a conceptual device conceived and used by Ian Hacking in his work on multiple personality disorder and elsewhere (Hacking, 1995; 1999a; 1999b;1991a; 1991b; 1994; &1986). Very briefly, Hacking claims that for some kinds of entity, which he calls "human" or "interactive" kinds,! the classification which is employed constructs and alters both the behavior and the identity of those classified. In
(g2001, International Journal of Applied Philosophy 15:2, ISSN 0738-098X,
this, Hacking's thesis sounds very like what psychotherapists call "labeling theory" which occurs when an individual begins to change his or her behavior based upon the label or classification they have been given. That is, according to labeling theory, an individual conforms to their label. This is different from Hacking's thesis about interactive kinds, however, in at least two significant ways. First, Hacking's thesis talks instead of groups, and the way a classification can alter the behavior of groups (and, derivatively, of members of the group). Second, and most importantly, these groups, and their classifications, are not stable. They are, Hacking claims, subject to "looping effects." This involves a triadic relationship between the classified (e.g., depressed adolescents), behavior, and the classifiers (e.g., health care professionals, particularly, psychologists and psychiatrists). Hacking maintains that at times a looping effect occurs where those classified react to the classification and force changes in that classification. Over a period of time, then, an original classification gets radically altered through the work of both those classified and the ones doing the classification. lt is precisely this process which has occurred with respect to adolescent depression. There has been a dramatic increase over the past thirty years in the numbers of people categorized in this way, and this increase has more or less paralleled changes in the categorization itself; moreover, looping effects have begun with those so classified acting back vis-a.-vis the categorization.
Let me make something perfectly clear at the outset. In saying that adolescent depression is a construction, I do not deny that it is real. Lots of things can be both constructed and real, and adolescent depression is one of them. Some adolescents, just as some adults, suffer terribly from depression: it really can be a debilitating disease. Furthermore, at times therapy and/ or drugs can be of tremendous help in allowing such people to lead more productive, fulfilling, and "happy" lives. This being said, however, I will also argue that the way in which adolescent depression has been constructed is not inevitable. More importantly, we have some reason for concern regarding the creation of the "depressed adolescent" in its current form, and the way in which such young people are being treated. In particular, I shall suggest that we ought to be concerned whether we are creating, particularly through drug therapy, but also through some forms of "talk" therapy agents, who are more passive than they otherwise would be. In expressing this concern, I am again following the work of Hacking who claims, in his work on multiple personality disorder, that a (if not the) primary basis for judgement regarding a particular therapy is whether it leads to the patient's autonomy (Hacking, 1995,258-268).2 Indeed, Hacking goes so far as to claim that whether or not a particular classification is "true" in some objective sense is irrelevant; what is relevant is successful therapy, and this is defined normatively in terms of patient autonomy. While I do not wholeheartedly support Hacking on this issue, I do believe that at least some of the current treatments for adolescent depression are troublesome since, as I will suggest, there is some reason to suspect that the classification and its therapy for a number of "depressed adolescents" has failed according to Hacking's criteria; that is to say, it may be that they are becoming, through
therapy, more dependent and hence less autonomous. As a consequence, if we think that the enhancement of patient autonomy is at least an important therapeutic goal, we have reason to be wary of current treatments for depressed adolescents.
I. DEFINING DEPRESSION
Defining ailments of the mind is notoriously difficult, particularly when their etiology has, like depression, no clear physical basis. Therapists in North America rely upon the American Psychiatric Association's Diagnostic and Statistical Manual (DSM), which is now in its fourth edition (1994). There was a radical change in the structure of the DSM from its first two editions to the third and fourth. Prior to the third edition, published in 1980, therapy was still influenced heavily (perhaps excessively) by the Freudian psychoanalytic tradition and the diagnostic manual reflected this. DSM-III (1980), however, moved to a multi-axial system which attempted, theoretically at least, to be as free as possible from any particular psychological ideology.3 Its first two axes classify all psychological "diseases" with Axis I targeting all of these with the exception of personality disorders and mental retardation which are dealt with in Axis II. The third, fourth, and fifth axes look to general medical conditions, psychological and environmental problems, and" global assessment of function" respectively. DSM IV lists depression under Axis I as a kind of "mood disorder" and further subdivides depression into four different types: only three of the types, however, are thought to be psychopathological. They are: "major depressive disorder," "dysthymia" (formerly "depressive neurosis"), and "bipolar depression" (formerly manic depression) (APA, 1994).4 Bipolar depression is characterized by radical mood swings: elated mania and deep depression. It is least common in adolescents, and hence I shall say nothing more of it here.s Major depressive disorder requires a depressed mood or loss of interest or pleasure in almost all activities for at least two weeks. Also necessary here is the presence of at least four other symptoms from the following list: (i) marked weight gain or loss when not dieting, (ii) constant sleeping problems, (iii) agitated or slowed down behavior, (iv) fatigue, (v) inability to think clearly, (vi) feelings of worthlessness, and (vii) frequent thoughts of death or suicide (DSM-IV, 1994; Sarason, I. G. and Sarason, B. R., 1999, 280). A third type of depression is dysthymia which is less severe but longer lasting than major depressive disorder (APA, 1994; Navarrete, 1999, 138). It is characterized by mild but chronic depressive symptoms. Indeed, because dysthymia is so persistent, some have claimed it ought to be classified as a personality order (Sarason, I. G. and Sarason, B. R., 1999,278). Part of the reason it continues to be classified as a mood disorder has to do with the fact that antidepressants have had some success in its treatment (Markowitz, 1993).6 A diagnosis of dysthymia requires the presence of a depressed mood most of the day, more days than not, for at least two years (in adults) or one year (in children and adolescents). In addition, at least two of the following six symptoms must be present while in a state of depression: (1) poor appetite or overeating, (2) insomnia or
ROBERT SCOTT STEWART
sleeping too much, (3) low energy, (4) low self esteem, (5) poor concentration or difficulty making decisions, and (6) feelings of hopelessness. These symptoms must not be absent for more than a two year period (or one year in children and adolescents), and other diagnoses and/ or problems, such as bipolar and major depression as well as substance abuse, must be ruled out. Finally, the condition must be debilitating (Sarason, LG. & Sarason, B.R., 1999,279; DSM-IV, 1994).
Even a cursory examination of the descriptions of dysthymia and major depressive disorder can indicate why rates of depression can be high. Consider dysthymia for example, particularly in the case of adolescents. A fifteen year old makes the transition from a small junior high school populated mainly by a relatively homogeneous group, many of whom are friends, to a large high school. Due to her place of residence, she gets sent to a different school than many of her friends. Her grades drop and she seems "moody." Toward the end of the year, her parents send her to a school councillor who, during an interview, discovers that she often feels "blue." Upon further investigation, the councillor discovers that she suffers from insomnia, has gained fifteen pounds due to overeating and lack of exercise (which bores her), and has difficulty concentrating. This all-too-common description of a fairly "normal" teenager now has a label: dysthymia. She is also a statistic in the growing number of depressed adolescents.
Although the case just described is fictional, it does mirror surprisingly well actual first person accounts. In her memoir, Mockingbird Years, Emily Fox Gordon documents her life in and out of therapy (Gordon, 2000). Initially diag
nosed as depressed, she began therapy at the age of thirteen shortly after mov-'
ing from a more or less comfortable setting in the small college-town of Williams town, Massachusetts to New York City. Between the time of this move and the age of seventeen, she went through five therapists, and after what she describes as a "feeble" suicide attempt-she scratched her wrists with nail scissors-was sent for a three-year period, initially as an in-patient, to the Austin Riggs sanitarium. But by her own admission, she was not "ill": rather, she, like most of the residents of Austin Riggs, was bored and "smitten with the romance of madness;" (Gordon, 2000, 5) she had "swallowed whole the familiar ideology that connects madness to beauty of spirit" (Gordon, 2000, 63). Indeed, she claims: "1 wasn't interested in being happier but in growing more poignantly, becomingly, meaningfully unhappy" (Gordon, 2000, 63).
It is, in part, for reasons such as this that researc.hers such as Barry Nercombe suggest that dysthymia is "a classificatory nightmare into which many problems have been dumped" (Nercombe, 1994, 61). More particularly, Nercombe suggests that clinical bias may be a factor here.
Clinicians affected by customary bias may unwittingly suggest symptoms to their interviewees or encourage them to endorse marginal complaints, whereas "halo" effects will incline interviewers to detect the remaining criteria for a favorite syndrome after the cardinal features have been identified. In this way, square pegs are jammed into round holes (Nercombe, 1994,64).
And, according to Kovacs (1986), meta-cognitive immaturity prevents chil
dren from exhibiting the introspective, self-monitoring thought required to
" . .
THE CONSTRUCTION OF THE DEPRESSED ADOLESCENT
report accurately on their own feelings, upon which so much rests in clinical interviews aimed at diagnosis. Perhaps this is a reason why children almost always report stronger feelings of depression than their parents and teachers (Reynolds and Johnston, 1994, 7).
Let us accept, then, at least for the time being, that rates of depression in adolescents may not be quite so shocking as the numbers suggest initially. Questions of course remain: have rates changed recently, and if so, what can account for the changes. I turn to this in the next section.
II. DEPRESSION IN ADOLESCENTS
According to statistics such as those found in StatsCan, depression was thought initially to be primarily an adult affliction, and diagnosed as such. But this has altered radically over the past thirty years. Currently, those aged 12-24 years are more likely to be treated for depression than any other group: "from 1994 to 1995, for instance, Canadians aged 15 were twice as likely as their 45-year-old counterparts to report having experienced a depressive episode in the previous 12 months" (McLaren, 2000, R1).The situation is not peculiar to Canada: similar statistics are available for the U.S.A. (Reynolds, 1999; Reynolds, 1985). Currently, then, there seems little doubt about it: teenage angst beats mid-life crisis hands down.
Part of the reason for this change stems from changes in DSM, particularly between the second and third editions. As stated above, DSM-I & II were highly influenced by Freudian psychoanalysis and Freudian psychoanalysis hardly dealt with childhood depression at all since its practitioners felt that depression required developments not yet experienced by the young (Kovacs and Beck, 1977; Mahler, 1961). In particular, Freudian psychoanalysts believed that children lacked the superego development necessary to manifest a true melancholia (Nercombe, 1994). Childhood depression became a serious and wide-spread area of research only in the 1970's (Reynolds and Johnstone, 1994). Indeed, reading work in the area from this time strikes one mainly for the sense that researchers themselves felt they were tackling something genuinely new. For example, Kovacs and Beck, two of this century's pioneers in adult depression, note repeatedly in their 1977 paper, "An Empirical-Clinical Approach toward a definition of childhood depression" how new and foreign childhood depression seemed at that time. Indeed, in his commentary of that paper, Anthony Nowels congratulates Kovacs and Beck for their "courage" in proceeding in this area which is full of "widespread disagreement and confusion" (Now ells, 1977, 27) Clearly, they were attempting to establish a paradigm of childhood depression where none existed previously. As Nowells notes: "They have not allowed the confusion and the morass of unknowns to keep them from getting to work. They appear to assume that we are all starting from a particular point where we can agree that: depression does exist, it can be measured, it exists in children, and it can exist in any child" (Nowells, 1977, 27). The main tenet of Kovacs and Beck was that childhood depression and adult depression were essentially similar, and that one can take the "'knowns' of the adult world and extrapolate backwards" (Now ells, 1977, 28).
Although this seems a perfectly reasonable hypothesis, there is a great deal of theoretical excavation which must be done here in order to make this extrapolation from adults to children and adolescents. As stated above, part of the reason here emanates from developmental issues and Freudian psycho
analytic theory. But part too lies in a deeper and less obvious area having to do
with the conceptualization of the child, and later of the "teenager" or adolescent. As Hacking notes, the very category "child" is subject to a fairly recent construction (Hacking, 1995,55-68; Aries, 1962). So too is the idea of a teenager and adolescent a recent construction. We need, then, to understand the concept of "human" or "interactive" kind and to see how interactive kinds such as children and adolescents get constructed. Finally, we have to understand what that construction tells us of the identity of these entities. What this discussion will show in part is that a claim commonly heard about adolescent depression-that it has always been with us but is only now being recognized-is false. The idea of adolescent depression has not always been with us: we have only recently developed the notion of an adolescent at all, and only even more recently have we created a classification which would allow for this age group to be designated as depressed.
III. THE CONSTRUCTION OF HUMAN/INTERACTIVE KINDS
There is an age-old philosophical debate between nominalists and realists which can be traced back at least as far as the debate between Plato (and Socrates) and the Sophists. As a metaphysical realist, Plato maintained that we must be careful "to carve nature at her joints" and not merely "hack off parts like a clumsy butcher" (Plato, Phaedrus, 266b). In contrast, the sophist and nominalist Protagoras (in)farnously claimed that 'man is the measure of all things.' The debate, at one level at least, is concerned with classification, and in particular whether the classifications we make are constructed (nominalism) or discovered (realism). That is, the nominalist argues that the objects of our world which we happen to classify together "have nothing in common except our names." Alternatively, realists maintain that there exist "n<;ltural kinds" out there "which we have painstakingly come to recognize and classify correctly" (Hacking, 1986,227).
Hacking is a firm realist with respect to most of the objects which populate our universe, so called "natural" or "indifferent" kinds.7 Such objects are different than interactive kinds in that they care not how we classify them (indeed, whether we classify them) or what we feel and think about them. That is, for example, the bacteria responsible for tuberculosis is unaffected by the words we employ in classifying it (properly or improperly): "A century ago I would have said that consumption is caused by bad air and sent the patient to the Alps. Today, I may say that TB is caused by microbes and prescribe a two-year course of injections. But what is happening to the microbes and the patient is entirely independent of my correct or incorrect description, even though it is not independent of the medication prescribed. The microbes' possibilities are delimited by nature, not by words" (Hacking, 1986, 231).
THE CONSTRUCTION OF THE DEPRESSED ADOLESCENT
Human or interactive kinds, alternatively, are affected by words. The classifications we make, and the attitudes we have toward various interactive kindsfrom "homosexual" to "multiple personalities," to use two examples from the history of psychological classification-affects the behavior of those entities. This sterns from the fact that human behavior is intentional and, following Elizabeth Anscombe, such intentional actions must be "actions under a description" (Anscombe, Intention; cited in Hacking, 1986,230). "This is," Hacking claims, "no mere lingualism, for descriptions are embedded in our practices and lives. But if a description is not there, then intentional actions under that description cannot be there either; that, apparently, is a fact of logic" (Hacking, 1986, 230).
With respect to interactive kinds, then, Hacking believes that the idea of us is 'made up,' and the person-kinds we become changes over time as the descriptions under which our actions occur change. "We are not only what we
are but what we might have been, and the possibilities for what we might have been are transformed" (Hacking, 1986, 233). In one sense, Hacking's claims here amount to nothing more than the claim that how people are labeled affects who they are. But, as I said at the outset, there is more to Hacking's account than this. "Labeling theory" is typically about individuals, human kinds are about individuals within a fluid, changing kind. One possibility that this opens up has to do with the ways in which interactive kinds can react with and actually alter the labeling classification. We are to think of this label
ing process as a framework consisting of two vectors:
One is the vector of labeling from above, from a community of experts who create a "reality" that some people make their own. Different from this is the vector of the autonomous behavior of the person so labeled, which possesses from below, creating a reality every expert must face (Hacking, 1986,234).
Because there are two vectors at play here, we can have what Hacking calls
"looping effects." This is a "feedback cycle" which occurs when a classification affects those classified (labeling from above) and this in turn results in a change in the classification itself by the classified (labeling from below-see, e.g., Latus, 2000). For example, let us say, following people such as Arnold Davidson, that "homosexual" carne into existence as an interactive kind during the later part of the nineteenth century as the result of the labeling from above by "experts" who constructed "homosexuality" as a medical condition in need of treatment. However, the classification itself was subject to a looping effect as those classified as "homosexual" reacted with and to the classification changing the very category itself by successfully de-medicalizing the "condition" (See Davidson, 1990). Hence, Hacking claims it is "not that there was a kind a person who carne increasingly to be recognized by bureaucrats or by students of human nature but rather that a kind of person carne into being at the same time as the class itself was being invented. In some cases, that is, our classifications and our classes conspire to emerge hand in hand, each egging the other on" (Hacking, 1986, 228).
This is the process which Hacking describes in Rewriting the Soul with respect to Multiple Personality Disorder. And in general, there have been a
ROBERT SCOTT STEWART
wide variety of things which are said to be socially constructed beginning with Tile Social Construction of Reality (Berger and Luckmann, 1966) and extending from Postmodernism, quarks, deafness, and teenage pregnancy, to name only a few (see Hacking, 1999b). In this paper, I examine the social construction of adolescent depression beginning first, following the work of Thomas Hine (1999), with the construction of the adolescent and teenager.