When do interpretations make a difference? A partial answer to Fliess's Achensee question.

Interpretations seem to matter most (at least in the case of Mrs. C) when the analyst feels included in the patient's thoughts. During hours of this kind (Related hours, scoring high on the cooccurrence of first- and second-person pronouns such as me/you), the analyst's interpretations are more comprehensive and probing; he intervenes earlier and more often within the session; and his specific interpretations have a greater effect on the patient's associations, as measured by an increased frequency of specific discourse markers. Low-scoring hours, with relatively few pronominal references to the analyst, were characterized by more superficial interpretations; later and fewer interventions; a greater tendency for the analyst to misunderstand the patient's productions; and a decreased effect of specific interventions. Special care was taken to protect the findings from the dangers of suggestion (the Achensee question).