066X.59.1.20 experience subthreshold symptom levels. Resilient individuals, by contrast, may experience transient perturbations in normal functioning (e.g., several weeks of sporadic preoccupation or restless sleep) but generally exhibit a stable trajectory of healthy functioning across time, as well as the capacity for generative experiences and positive emotions (Bonanno, Papa, & O’Neill, 2001). The prototypical resilience and recovery trajectories, as well as chronic and delayed disruptions in functioning, are illustrated in Figure 1. In the loss and trauma literatures, researchers have tended to assume a unidimensional response with little variability in possible outcome trajectory among adults exposed to potentially traumatic events. Bereavement theorists have tended to assume that coping with the death of a close friend or relative is necessarily an active process that can and in most cases should be facilitated by clinical intervention. Trauma theorists have focused their attentions primarily on interventions for PTSD. Nonetheless, trauma theorists and practitioners have at times assumed that virtually all individuals exposed to violent or life-threatening events could benefit from active coping and professional intervention. In this section, I discuss how the failure of the loss and trauma literatures to adequately distinguish resilience from recovery relates to current controversies about when and for whom clinical intervention might be most appropriate. This failure also helps explain why in some cases clinical interventions with exposed individuals are sometimes ineffective or even harmful. The Grief Work Assumption Traditionally, mental health professionals in the industrialized West have understood grief and bereavement from a single dominant perspective characterized by the need for grief work (Stroebe & Stroebe, 1991). The conception of grieving as work originated in Freud’s (1917/1957) metaphoric use of the term to describe the idea that virtually every bereaved individual needs to review “each single one of the memories and hopes which bound the libido . . . to the non-existent object” (p. 154). Theorists following Freud emphasized even more strongly the critical importance to all bereaved individuals of working through the negative thoughts, memories, and emotions about a loss (see Bonanno & Field, 2001). As researchers began to devote more attention to the bereavement process, however, it became apparent that, despite the near unanimity with which mental health professionals endorsed the grief work perspective, there was a surprising lack of empirical support for such a view (Wortman & Silver, 1989). What’s more, recent studies that have directly examined the legitimacy of the grief work approach have not only failed to support this approach but actually suggest that it may be harmful for many bereaved individuals to engage in such practices (see Bonanno & Kaltman, 1999). A more plausible alternative would be that grief work processes are appropriate for only a subset of bereaved individuals (Stroebe & Stroebe, 1991), most likely those actively struggling with the most severe levels of grief and distress (Bonanno et al., 2001). The idea that grief work may characterize only the more highly distressed bereaved individuals (i.e., those exhibiting either the recovery or chronic symptom trajectories) is further supported by data indicating that the practice of engaging a wide array of bereaved individuals in grief counseling has proved remarkably ineffective. Grief-focused interventions typically target both acute or prolonged grief reactions as well as the absence of a grief reaction (e.g., Rando, 1992). Two recent meta-analyses Figure 1 Prototypical Patterns of Disruption in Normal Functioning Across Time Following Interpersonal Loss or Potentially Traumatic Events George A. Bonanno January 2004 ● American Psychologist 21 independently reached the conclusion that grief-specific therapies tend to be relatively inefficacious (Kato & Mann, 1999; Neimeyer, 2000). A third meta-analytic study reported that grief therapies can be effective but generally to a lesser degree than usually observed for other forms of psychotherapy (Allumbaugh & Hoyt, 1999). In one of these analyses, an alarming 38% of the individuals receiving grief treatments actually got worse relative to no-treatment controls, whereas the most clear benefits were evidenced primarily with bereaved individuals experiencing chronic grief (Neimeyer, 2000). In summarizing these findings, Neimeyer (2000) concluded that “such interventions are typically ineffective, and perhaps even deleterious, at least for persons experiencing a normal bereavement” (p. 541). Trauma Interventions and Critical Incident Debriefing Although for centuries practitioners have linked violent or life-threatening events with psychological and physiological dysfunction, historically there also has been confusion and controversy over the nature of traumatic events and over whether to consider psychological reactions as malingering, weakness, or genuine dysfunction (Lamprecht & Sack, 2002). The inclusion of the PTSD category in the Diagnostic and Statistical Manual of Mental Disorders (3rd ed. [DSM–III]; American Psychiatric Association, 1980) resulted in a surge of research and theory about clinically significant trauma reactions. There is now considerable