Reflections on Recovered Memories: Comment on Patihis and Pendergrast (2019)

In the report of their survey of the general U.S. public regarding memory recovery in psychotherapy, Patihis and Pendergrast (2019; this issue, p. 3) contend that the “central . . . controversy is whether attempting to help clients to recover purportedly repressed memories of abuse leads to memory distortions that harm rather than heal clients” (p. 3). Several findings of particular concern emerged from their survey. First, 9% (8% weighted to be representative) of their sample of 2,326 adults reportedly consulted therapists who discussed the possibility of repressed abuse, and 5% (4% weighted) recovered memories of abuse in therapy. Second, there was a 20-fold increase in rates of recovered memories among respondents who discussed the possibility of recovered memories compared with individuals who did not engage in such discussions. Third, reports of recovered memories were spread across most types of therapies, implying their ubiquity.

Considerable skepticism over the years has been voiced regarding the notion that highly adverse and arousing events can be fully encoded yet somehow selectively repressed or dissociated—essentially banished from consciousness—only to be “recovered” in their entirety and force years later (Patihis, Ho, Tingen, Lilienfeld, & Loftus, 2014). Skepticism typically boils down to “this is not how memory works”: Highly salient emotional events are often remembered “all too well,” as individuals with posttraumatic stress disorder can painfully attest. Moreover, researchers who have studied complex false memories produced by suggestive procedures (e.g., misleading information, hypnosis, imagination) have expressed serious concerns about the iatrogenic origin of recovered memories in psychotherapy (see Lynn, Krackow, Loftus, Locke, & Lilienfeld, 2015).

In short, Patihis and Pendergrast (2019) are correct to be skeptical of efforts to recover dissociated memories, concerned about the accuracy of such memories, and wary of their aftereffects. To the extent that many recovered memories in therapy are iatrogenic—and there is good reason to believe they are—an astonishing and deeply troubling 42.6% of individuals surveyed who reported recovering memories of abuse indicated they had cut off contact with family members, and the great majority of reporters of recovered memories (92.6%) still believed their memories to be accurate. Patihis and Pendergrast’s research helps to understand why it is problematic to place unfettered confidence in recovered memories, contrary to the findings of Brand, Schielke, Brams, and DiComo (2017), who, in their guidelines for expert witnesses, claim “Research has established the reliability of recovered memories of trauma” (p. 303).

What Patihis and Pendergrast (2019) do not tackle in detail are questions about the central assumption that underpins recovered memories: that there is a close link of traumatic experiences with dissociation. Even among clinical samples, the correlations are highly heterogeneous and range from r = –.14 (n.s.) to r = .63 (p < .001; see Dalenberg et al., 2012); 41% of correlations reported in Dalenberg et al.’s (2012) review were medium or lower in magnitude (Cohen’s provisional standards, below r = .30), and only 5% equaled or exceeded r = .50, indicative of a large effect. Additionally, individuals diagnosed with a serious dissociative disorder do not invariably report a history of any trauma exposure (Lynn et al., 2014). Moreover, a clear causal link between trauma and dissociation is suspect because the majority of studies are correlational, and trauma is not corroborated and based on retrospective reports. In short, the empirical foundation on which efforts to recover memories is grounded (i.e., a clear trauma-dissociation link) is very weak—far too weak to be regarded as reliable.

The Patihis and Pendergrast (2019) findings imply that a need exists to understand how therapists who believe in dissociated/repressed memories instill this belief in clients and create expectations that current problems can be interpreted in light of past traumas. Although researchers have identified techniques that are particularly risky in terms of inducing false memories (e.g., repeated imagining, hypnotic age regression), the process by which memories come to be recovered and a personal narrative constructed regarding the genesis of psychological disorders has been little studied. In fact, researchers have paid scant attention to iatrogenic influences in psychotherapy, and systematic research on side effects of psychotherapy has only just begun (Hardy et al., 2017).

Nevertheless, the survey provides clues regarding the context in which recovered memories arise. Cognitive-behavioral therapy, which is decidedly not oriented toward memory recovery, was associated with the lowest percentage of recovered memories and/or clients coming to believe they had multiple personality disorder/dissociative identity disorders, whereas attachment therapy, associated with the highest percentages, is probably more likely to frame current problems in terms of repercussions of early abusive relationships and thus, implicitly or explicitly, frame the recovery of memories as a valuable pursuit.

Recovered memories could be classified and studied as a function of how they come to light and are interpreted. Some clients, for example, might simply forget an event they later recall or might reinterpret as “abuse” an event that previously was viewed more innocuously. For example, Patihis and Pendergrast (2018) reported that emotional abuse was the most prevalent of all types of abuse memories recovered, which raises the question of whether such reports were based on retrospective evaluation of events as abuse and whether the hypothesized “special process” of dissociation or repression was at play at all. Clearly, the variables that affect whether individuals construe experiences as abuse or arrive at more benign interpretations are worthy of study, as are the nature and aftereffects of recovered memories that arise (a) in versus out of therapy, (b) spontaneously versus in response to a “trigger” (e.g., television program on abuse), or (c) after a suggestive technique such as hypnotic age regression for memory recovery. Researchers could also evaluate the effectiveness of psychotherapies among clients who recover memories versus those who do not and whether memories recovered are correlated with psychopathology. The fact that memories are retrieved may not be inherently problematic; rather, clients’ and therapists’ interpretation of memories may be crucial in shaping an evolving narrative that fosters psychological resilience versus feeling forever powerless and hopeless. As many memories in psychotherapy are difficult or impossible to corroborate, recovered memories, regardless of their accuracy, could ideally be folded into a life story that encompasses feeling empowered and capable of forgiving (or at least better understanding) oneself and others.

Patihis and Pendergrast (2019) suggest that “clients entering therapy should be given information about the potential hazards of recovered memories of abuse as part of informed consent” (p. 17). We further suggest therapists provide information regarding the potential adverse effects of certain techniques (e.g., repeatedly imagining events, age regression) in sculpting memories. Some years ago, Lynn (2001) suggested that therapists provide informed consent regarding the metaphorical rather than the literal nature of inner advisors, personality parts, ego states, hidden observers, and the like in ego state and “parts”-based therapies. This suggestion was roundly rejected at the time by four commentators (see Hammond, 2001; Kluft, 2001; Spiegel, 2001; J. Watkins, 2001). We hope that Patihis and Pendergrast’s (2019) important article shifts the therapeutic community toward routine use of informed consent not only for recovered memories but also for techniques used to elicit such memories.