Motivational Interviewing (MI) is a clinical approach that began as a treatment for addictions and has spread to a broader set of health-relevant behaviors . Implicit in it has been a sensibility derived from Rogers's  person-centered approach to counseling which emphasizes that being responsive to and acknowledging patients' feelings is important for their growth and wellness. In line with this sensibility, as argued by Resnicow and McMaster , promoting patients' autonomy or volition is important within MI. As such, there is a similar interpersonal perspective to treating people within MI and SDT, and the intervention techniques used within these two approaches also have much in common. Indeed, investigators performing the clinical trials that have been discussed in this issue and were intended to test the SDT model of health-behavior change have indicated that they drew from MI for specific elements of their interventions (e.g., Fortier, Duda, Guerin, & Teixeira, ). In doing so, they were acknowledging similarities between the philosophies and aims of MI and SDT.
For example, both approaches can be thought of as being person-centered; both are non-judgmental and supportive; both provide information that is responsive to what the patients' appear to need; both buttress patients' attempts to come into deeper contact with their inner experiences and motivations; and both have emphasized patients' autonomy while at the same time working to promote patients taking responsibility for behaving in healthy ways. Further, as argued by Markland, Ryan, Tobin, and Rollnick , SDT can be viewed as a theory that explains the effects that occur when using MI treatments.
Still, there are significant differences in the nature of the two approaches. As pointed out by Vansteenkiste et al. , SDT is a macro-theory of human motivation. It began with basic laboratory research and was gradually applied to education , to work organizations , and to health care . Vansteenkiste et al. thus referred to SDT as a top-down approach to health-behavior change (hypotheses are derived from an overarching theory) . In contrast, the authors spoke of MI as being a bottom-up approach-that is, an approach to health-behavior change that is largely atheoretical and was based on intuition, trial-and-error, and clinical observations when treating addicted patients within the counseling venue.
We agree that SDT was applied in a top-down manner to the health-care domain because at the time that research was begun the theory was quite well developed. At the same time, it is important to recall that the theory's formulation was based on basic research, much of it from the laboratory. We did not start with a theory and then test it. Instead, the initial process of developing the theory involved asking interesting questions and, through research, gradually developing theoretical propositions. As these propositions accumulated, the theory was increasingly applied in the "top-down" manner referred to by Vansteenkiste et al . Nonetheless, throughout its history, the theory has continued to be expanded and refined as new data have shed light on new phenomena and processes.
This is very different from the history of MI, which was developed within the domain of health-behavior change and paid little attention to theory. However, Miller and Rose  recently published an article that is aimed toward development of a theory for MI. Its primary focus is on patients' change talk as the central mechanism for promoting health-behavior change. Change talk, quite simply, means having patients talk about their behavior change-planning when and how to do it, enumerating the advantages of doing it, guessing how it might affect the people to whom they are closest, and so on. Miller and Rose suggested that the amount of change talk is important, with more being better for yielding change.
We do not disagree that change talk can promote change and be empirically associated with positive outcomes. But fostering change talk of any type seems like a step back from specificity, and may foster controlled as well as autonomous processes. In Miller and Rose's article autonomy seemed implicitly to be somewhat important, but it seemed to have been moved into the background, while the cognitive activity of change talk per se became foreground (see ). To the degree that this is so-to the degree that the amount of change talk takes a more front-and-center place in the theory than autonomy-the similarity between SDT and MI is diminished. Furthermore, from an SDT perspective, if change talk were to be given prominence, it would be essential that the focus be not just on the amount of change talk but rather on the quality of the change talk.
More specifically, patients can engage in change talk that reflects autonomy, speaking about such topics as the options available to them, their personal values, and taking greater responsibility because they want to become healthier for themselves and for those who love them. Doing that is likely to be associated with greater internalization, maintained behavior change, and well-being. Alternatively, patients could engage in change talk in a controlled way, emphasizing what they should be doing to change or how their families, friends, or practitioners have been pressuring them to change. Such controlled change talk likely reflects less internalization and will result in less effective and less well-maintained change attempts.
Finally, the fact that there are different qualities of change talk among patients has important implications for practitioners' roles in the process. To the extent that practitioners are directive and controlling in trying to move patients toward a larger quantity of change talk, the results are likely to be less effective. SDT maintains that any such pressure toward a specified outcome will likely foster more interpersonal control, and lower autonomy and relatedness in the interaction. For MI to maintain considerable similarity with SDT it will be necessary, quite simply, for its advocacy of change talk to emphasize practitioners being autonomy supportive and to facilitate patients' engaging in talk about the possibility of autonomous and self-endorsed, rather than heteronomous or externally-driven, change.