Dr. Loveman takes us to task for not discussing the association between child abuse and chronic symptoms in adult life. However, this evidence would not have altered our argument. We do not doubt that, in many patients, chronic unexplained illness has its origins in the embodied effects of traumatic past experience. We argued that the embodiment of the emotions is normal and that the problem lies in the chronicity of the symptoms. In the etiology of any chronic illness, it is important to distinguish between the initial and the perpetuating causes. The factors triggering the onset are not necessarily the same as those maintaining the illness. As Dr. Wallace reminds us, nociceptive response is one such factor. Failure of the emotion to enter consciousness is another. Success in therapy requires us to attend to the causes of chronicity, directing our efforts to those in each patient that are most amenable to change. The approach described by Dr. Patil is in accordance with this thinking, although cognitive and psychobehavioral approaches do not exhaust the possibilities.

The conceptualization of somatoform disorders does not consider the multifactorial nature of chronic illness, and our dualistic ontology forces us to classify the disorders as either mental or physical, psychiatric or medical. We do not agree with Drs. O'Malley and Kroenke that the association between symptoms and functional impairment constitutes empirical validation for the somatoform disorders. Instead, this emphasizes that chronic symptoms, whether explained or unexplained, have a profound impact on patient's lives. Neither do we share their fear that deconstructing the concept of somatization will imperil a biopsychosocial approach. Our intention is to enrich physicians' grasp of patients' suffering by understanding the experience of illness in the patient's terms, without creating stigmatizing barriers.

Smith and colleagues [1] demonstrated improved physical functioning in patients with unexplained symptoms who volunteered to undergo psychiatric evaluation. We have no doubt that selected patients with chronic illnesses, including those that are "unexplained," can benefit from psychiatric intervention. Establishing a strong patient–physician relationship, frequent nonsymptom-dependent visits, the use of touching, and "harm reduction" by limiting unnecessary testing are very much in accordance with our own thinking, and evidence supports the effectiveness of these approaches in all types of illness, regardless of diagnosis [2]. However, it is difficult to believe that a single consultation letter from a psychiatrist would produce such profound behavioral change in practicing physicians. We await replication of Smith and colleagues' results in a study that does not include significant enrollment bias.

An effective intervention does not prove that the diagnosis is meaningful. Drs. Loveman, Wallace, and Patil point out the importance of considering the multidimensional nature of suffering in patients with chronic unexplained symptoms by simultaneously invoking genetic, neurologic, developmental, social, psychological, and physiologic processes. The term somatizer does not accomplish that goal; instead, it restricts our view and hampers our ability to be of help. The fact that physicians find the concept of somatization helpful is surely not a sufficient justification for it if patients find it offensive. Perhaps we physicians should try to imagine how we would feel if we heard ourselves described as somatizers or somatizing patients. We might, like William James' hypothetical crab, upon hearing itself described as a crustacean, respond with indignation: "I am no such thing; I am MYSELF, MYSELF alone" [3].