I do not think that intentionality or unintentionality is the defining
criterion in our 'theory-in-use' (t-i-u) definition of 'manipulation'.
Perhaps a more concrete t-i-u definition of 'manipulation' might be an
action which, if it's outcomes were fully understood by all parties, would
not be agreed to by one or another party. This definition can be thought
of in relation to systems dynamics. As the theme of the butterfly wings on
one side of the globe affecting weather on the other suggests, all action,
regardless of intention, can be assumed to have an effect.
To manipulate, in its more mechanical definition (thanks, Mike), does mean
to physically create some effect. To apply our psychological baggage to
the term, the question is, will the effect of the action be ecological for
each party in the system? If the answer is no, and that is known in
advance by the actor, but not understood by the parties who are going to
be affected, then that action might be called "manipulative" (in our t-i-u
sense).
Marilyn Darling
mdarling@warren.med.harvard.edu