标题：Fifty psychological and psychiatric terms to avoid: a list of inaccurate, misleading, misused, ambiguous, and logically confused words and phrases
source：journal.frontiersin.org | by Scott O. Lilienfeld, Katheryn C. Sauvigné, Steven Jay Lynn, Robin L. Cautin, Robert D. Latzman and Irwin D. Waldman
欲翻译的小伙伴，可评论『领稿』，并把已完成的译文『贴在评论』处或『发在自己的心理圈内』。请领稿的小伙伴，尽量在一个星期内完成翻译，谢谢！！~~
Ambiguous Terms
(35) Comorbidity. This term, which has become ubiquitous in publications on the relations between two or more mental disorders (appearing in approximately 444,000 citations in Google Scholar), refers to the overlap between two diagnoses, such as major depression and generalized anxiety disorder. A similar term, “dual diagnosis,” which has acquired considerable currency in the substance abuse literature in particular, refers to the simultaneous presence of a mental disorder, such as schizophrenia, and a substance abuse disorder, such as alcoholism (Dixon, 1999). Some authors have taken the comorbidity concept further, extending it to “trimorbidity” (Cornelius et al., 2001) or “quatromorbidity” (Newman et al., 1998).
Nevertheless, “comorbidity” can mean two quite different things. It can refer to either the (a) covariation (or correlation) between two diagnoses within a sample or the population or (b) co-occurrence between two diagnoses within an individual (Lilienfeld et al., 1994; Krueger and Markon, 2006). The first meaning refers to the extent to which Condition A and B are statistically associated across individuals; for example, there is substantial covariation between ASPD and BPD (Becker et al., 2014). The second meaning is a conditional probability referring to the proportion of individuals with Condition A who meet diagnostic criteria for Condition B. For example, in the case of the latter meaning, researchers might note that 45% of patients with ASPD also meet diagnostic criteria for BPD. The difference between these two meanings is hardly trivial, because they tend to be differentially influenced by base rates (prevalences). If the base rates of one or more conditions change, the covariation between them will not necessarily be affected but the level of co-occurrence almost always will be (Lilienfeld et al., 1994). Moreover, depending on the base rates of the diagnoses in a sample, two conditions may display little or no covariation but substantial co-occurrence. For example, although ASPD and major depression typically display only modest covariation (Goodwin and Hamilton, 2003), the rates of co-occurrence between ASPD and major depression in an analysis conditioned on major depression (that is, the rates of ASPD among people with major depression) would be extremely high in a prison sample, because most prison inmates meet criteria for ASPD (Flint-Stevens, 1993). Hence, the levels of comorbidity would probably be negligible in the first case but high in the second. If authors elect to use the term “comorbidity,” they should therefore be explicit about which meaning (covariation or co-occurrence) they intend.
Some authors (Lilienfeld et al., 1994) have further questioned the routine use of the term comorbidity in psychopathology research given that this term, much like “dual diagnosis,” presupposes that the conditions in question are etiologically and pathologically separable entities (but see Rutter, 1994; Spitzer, 1994, for demurrals). For example, although the high level of “comorbidity” between ASPD and BPD may reflect covariation or co-occurrence between two distinct conditions, it may instead reflect the fact that the current diagnostic system is attaching different names to slightly different manifestations of a shared diathesis, thereby falling prey to a jangle fallacy. To take an admittedly extreme example, how likely is it that a participant in a published study who simultaneously met diagnostic criteria for all 10 DSM personality disorders (see Lilienfeld et al., 2013) genuinely possessed 10 distinct disorders at the same time? Critics of the expansive application of the term comorbidity to descriptive psychopathology contend that these diagnostic conundrums are best explained by a flawed diagnostic system that is attaching different names to highly overlapping constructs.
(36) Interaction. As Olweus (1977) observed in the context of the person-situation debate, the term “interaction” has multiple meanings, some of them logically incompatible. For example, the familiar phrase “genes and environment interact for Disorder X” can mean any one of four things: (a) genes and environment are both involved in the causes of Disorder X; (b) the relation between genes and environments are bidirectional, because genes influence the environments to which people are exposed (by means of gene-environment correlations), and environments influence which genes are activated or inactivated (by means of epigenetic processes); (c) the influences of genes and environment are inseparable because of continuous transaction within individuals; or (d) the statistical effects of genes depend on people’s environments, and the statistical effects of environments depend on people’s genes. Only meaning (d) refers to a statistical interaction in the standard multiple regression or analysis of variance sense.
Two points are worth noting here. First, psychologists routinely confuse meanings (a) and (d). For example, when researchers write that “All reasonable scholars today agree that genes and environment interact to determine complex cognitive outcomes” (Bates et al., 1998, p. 195), some readers may assume that they are referring to the standard statistical meaning of the term “interaction,” (McClelland and Judd, 1993), i.e., a multiplicative rather than additive relation between variables, such as that between genetic and environmental influences. Instead, in this case the authors appear to be saying only that both genes and environment play a role in cognitive outcomes, a scenario that does not require a multiplicative relation between genes and environment. Second, meanings (c) and (d) are logically incompatible, because if the effects of genes and environment are not separable, then clearly they cannot be distinguished in statistical designs. The bottom line: when authors use the term “interaction,” they should be explicit about which of the four meanings they intend.
(37) Medical model. Although many authors who invoke the term “medical model” presume that it refers to a single conceptualization (e.g., Mann and Himelein, 2008), it does not. Some authors insist that the term is so vague and unhelpful that we are better off without it (Meehl, 1995). Among other things, it has been wielded by various authors to mean (a) the assumption of a categorical rather than dimensional model of psychopathology; (b) an emphasis on underlying “disease” processes rather than on presenting signs and symptoms; (c) an emphasis on the biological etiology of psychopathology; (d) an emphasis on pathology rather than on health; (e) the assumption that mental disorders are better treated by medications and other somatic therapies than by psychotherapy; (f) the assumption that mental disorders are better treated by physicians than by psychologists; or (g) the belief that mentally ill individuals who engage in irresponsible behavior are not fully responsible for such behavior (see Blaney, 1975, 2015, for discussions). Similar semantic and conceptual ambiguities bedevil the term “disease model” when applied to addictions and most other psychological conditions (e.g., Graham, 2013).
(38) Reductionism. There may be no greater insult in psychological circles than to brand a colleague a “reductionist.” Indeed, merely accusing a fellow faculty member of “being reductionistic” is often an effective conversation-stopper at cocktail parties. The negative connotation attached to this term neglects the point, overlooked by many authors (e.g., Harris, 2015), that “reductionism” is not one approach. Robinson (1995) delineated multiple forms of reductionism, including (a) nominalistic reduction, i.e., reduction at the level of names (“A brain structure called the amygdala plays a key role in fear processing”); (b) nomological reduction, i.e., reduction at the level of scientific explanation (“The perception of edges is mediated in part by feature detection cells in the visual cortex”); and (c) ontological reduction, i.e., reduction by eliminating immaterial entities (“Neuroscientific data strongly suggest that there is no immaterial soul”).
More broadly, we can differentiate between two quite different brands of reductionism: constitutive and eliminative, the latter termed “greedy reductionism” by Dennett (1995). The constitutive reductionist believes merely that everything that is “mental” is ultimately material at some level, and that the “mind” is what the brain and rest of the central nervous system do. Constitutive reductionists (like nomological reductionists; Robinson, 1995), who appear to comprise an overwhelming majority of psychologists and neuroscientists, reject mind-body dualism, the claim that the mind is entirely separable from the brain. In contrast, eliminative reductionists go a large step further (Lilienfeld, 2007). They contend that the “mind” will eventually be explained away entirely by lower-level concepts derived from neuroscience, and that mentalist concepts, such as thoughts, motives, and emotions, will ultimately be rendered superfluous by neuroscientific explanations. For eliminative reductionists, the field of psychology will eventually be “gobbled up” by neuroscience. Although we do not attempt to adjudicate the dispute between constitutive and eliminative reductionists here, suffice it to say that “reductionism” does not carry a single meaning in psychology. As a result, psychologists who use “reductionist” as a handy term of opprobrium against their colleagues must be explicit about which form of reductionism they are invoking.
