Loftus hadn't thought of memory doctoring as a good idea in more than a decade. And now the idea struck her quite differently. It was no longer a mystery. It was her craft. She could do it.
Her first idea, cooked up over lunch with a friend, was to plant a memory of sitting on your grandmother's lap and being told that you were her favorite grandchild. Wouldn't that feel wonderful? But then she realized what would happen at the end of the experiment. As a research psychologist, she was ethically required to tell her subjects that the memory wasn't true. She couldn't bear to do that. So she dropped the idea.
Then another idea came along. Loftus had two junior colleagues studying imagination and memory. They were demonstrating that the act of imagining an experience increased people's confidence that the experience had really happened. Loftus tacked on a second experiment to see whether similar imagination exercises could increase healthy behaviors. The subjects were asked to imagine flossing and eating vegetables with dinner. It seemed to work: 24 percent of subjects later reported more flossing, and 40 percent reported more vegetable consumption.
In February 1997, Loftus unveiled her new line of thinking at the annual meeting of the American Association for the Advancement of Science. She explained how imagination could distort memory, and she conceded that this was often harmful. But rather than renounce the whole idea, she proposed to "harness the power of imagination and put it to some good use" by inducing healthier behavior. Instead of tampering with memories, she was proposing to take one component of the memory-tampering recipe—imagination—and use it for a completely different purpose.
Soon she was being flown to Washington, D.C., for a workshop organized by the National Institute on Drug Abuse. Its purpose was to figure out how behavioral science could inform behavior therapy. She presented her findings on imagination and memory, and the group discussed how this research could be clinically applied. Based on the presentation, Loftus and a colleague, Giuliana Mazzoni, wrote up a proposal and published it in the Autumn 1998 issue of Behavior Therapy.
Loftus had two models to work from. One was the accidental brainwashing recipe of the recovered-memory therapists. The other was her copy of that recipe, refined in the laboratory. All she had to do was tweak it.
The tweaked recipe, outlined in Behavior Therapy, was a procedure called Expert Personalized Suggestion. The patient's behavior would be changed through the power of suggestion, in the form of guided imagination exercises. She would envision herself flossing, for example, and would be promised better flossing habits as a result. This promise would be made credible by a tailored analysis of her personality. And she would accept the analysis and the suggestion because they came from an expert.
Loftus speaking at Beyond Belief, a conference of the Science Network, November 2006
The authors described this as "a novel procedure that capitalizes on what past research has intimated about the power of an authority figure, and the power of personalized suggestion, to influence people's thinking about their past." The past research, much of it conducted by Loftus, had focused on the danger of these powers. But the new procedure exploited them. It incorporated some of the tricks she had learned from the recovered-memory therapists, starting with pseudo-customized diagnosis. The diagnosis was actually scripted, and the expertise, as presented to the patient, was fraudulent. In fact, Loftus and Mazzoni called the whole thing a "therapy simulation." But instead of hurting people, it would help them. It would get them to floss, take calcium supplements, and avoid cigarettes.
There was one catch: The procedure wouldn't work if patients knew it was fake. They had to believe that the therapy, expertise, and personalization were real. "We probably would not wish to include the deceptive aspects of the methodology because it is not ethically acceptable to deceive a client," the authors conceded. But maybe this problem could be fudged: "Suppose after a short interview with [the patient], the clinician tells her that he proposes trying an imagination exercise that he is quite confident, given her interview, will lead her to increase her consumption of calcium. In a real sense this is a true statement," since imagination exercises, on average, did change behavior. If the simulation worked, it wasn't fake—was it?
With each innovation and rationalization, Loftus was building a technical and moral case for therapeutic trickery. But one ingredient was still missing from her recipe: memory.