they're different in a number of ways. I'll summarize the ones I think are most important for speech. Anatomically: 1) reduced gray matter volume in speech related areas (notably the left inferior frontal gyrus). 2) decreased organization of white matter tracks connected the rolandic and perisylvian speech / language regions bilaterally 3) increased volume of corpus callosum Physiologically: 1) Diffuse increased right hemisphere activity during speech production and language processing tasks 2) potentially reduced left hemisphere activity in inferior frontal regions and superior temporal regions. 3) increased cerebellar activity. *It should be noted that these functional measures are typically done in perceptually fluent speech -- so these differences exist even when people who stutter are not stuttering. You're correct to point out that we don't know whether these differences are related to the cause or consequence of stuttering. My dissertation will hopefully shed some light on this. I reject the notion that we have to jump to the idea of stuttering as a learned fear reaction. I am not afraid to talk in front of my class, I am not worried about stuttering in front of them, why then, if stuttering is the physical manifestation of the fear of stuttering, do I still stutter in front of my class? Why do I stutter at all? If we want to really see if people who stutter are more reactive in anxiety than their peers we can do that. The only way I know to objectively measure anxiety is to measure salivary cortisol levels. And they've done that--people who stutter don't have increased cortisol levels relative to their fluent peers. So, it's not about the anxiety, it's about something else--the motor system? the language system? the motor-language interface? I'm not sure what you're getting at with the physiologically different statement. Speech production in both children and adults who stutter is more variable than their normally fluent counterparts. This is interpreted to mean that the motor systems of people who stutter are less stable. I'm not sure I buy that reasoning for a lot of reasonings. But in an absolute sense, the kinematic data says they're different. There's evidence from muscle activity (EMG) in adults. Adults who stutter show evidence of tremor in the EMG but children who stutter don't.