Steve Mitchell was one of the main spokseman of a group of analysts that
saved Psychoanalysis from a devestating epistemological fate.  Imagine
Eisman's ideologies (which is Logical Positivism with its
sharp dichotomy between legitimate and unlegitimate knowledge) were still
the main discourse of western philosophy and culture today. If they were,
if people thought that a valid theory is only an empirical one,
psychoanalysis  could not have survived. Fortunately, our culture has
changed. It seems now, as Docherty pointed out, that Enlightment itself is
not the great demystifying force which will reveal and unmask ideology.
Rather, it is precisely the locus of ideology, thoroughly contaminated
internally by the ideological assumption that the world can match our
reasoning about it. Briliantly, Mitchell (with Hoffman, Aron, Benjamin,
Stolorow, Orange, Atwood, Spezzano and Renik, to name just a few) was one
of the first to acknowledge this epistemological shift and to apply it to
psychoanalysis (of course, there is a price that psychoanalysis will have
to pay for being postmodern).
I think the best answer to Eisman's critique is the following quotation
from Mitchell's  book "Influence and Autonomy in Psychoanalysis" (1997):
"Because of the swing of pendulum away from scientism most of us were
brought up on, we are particulary vulnerable to a clinical state I have
observed in psychoanalysis that I have come to think of as the "Grunbaum
Syndrome". This may afflict psychologist-analysts more then others: I don't
know. I have come down with it several times myself.  It begins with some
exposure to the contemporary philosopher Adolph Grunbaum's attack on
psychoanalysis. Grunbaum wants to indict psychoanalysis for not meeting the
criteria he designates as necessary for an empirical, scientific
discipline. Since the analyst's interpretations operate at least partially
through suggestion, he argues, there is no way of testing their validity in
any independent fashion. What follows (after reading Grunbaum) is several
days of guilty anguish for not having involved oneself in analytic
research. There may be outbreaks of efforts to remember how analysis of
variance works, perhaps even pulling a 30-year-old statistics text off the
shelf and quickly putting it back. There may be sleep disturbance and
distractions from work. However, it invariably passes in a day or so, and
the patient is able to return to a fully productive life. The most striking
thing about Grunbaum's impact on psychoanalysis is the extraordinary plat
his critique has attracted despite its almost irrelavence to contemporary
clinicians. The reason virtually all clinicians suffering from the Grunbaum
syndrome put the statistics text back on the shelf within a day or two is
that clinicians tend to be satisfied with kinds of confirmation different
from the singular empirical one Grunbaum insists on" (p. 206-7).

Aner Govrin - Clinical Psychologist
Tel Aviv