The Inquisitor

A conversation with the Inquisitor character puts the user in the position of a heretic caught by the dreaded Spanish Inquisition. Of course, this particular Inquisition is not particularly deserving of dread, since it is based on the Spanish Inquisition in Monty python. As such, the Inquisitor characters tries harder to be amusing than religious, and there there are a goodly number of hidden gags (both from the original sketch and otherwise) which the user can see if certain keywords are entered. Cardinal Ximinez tries to prompt the Heretic with appropriate things to confess, but for the most part it's up to the user's imagination to fit things into the expected scene. Familiarity with the Monty Python sketch does help.

Currently the Inquisitor character has 34 keyword entries, with 89 distinct keywords recognized (counting all wildcards as a single keyword), and 124 possible responses. Unfortunately, some of the Inquisitor's paths of conversation are not as complete as they could be, due to the limitations of the authoring system. Indeed, a small number of keywords are only unpredictably accessible, since some of their conditions overlap. Due to the inability to set a priority among keywords, or to detect punctuation and therefore detect the end of a sentence, some related keywords will override each other unpredictably (such as the general "I confess." which should end the conversation, as opposed to the general keyword for a specific confession like "I confess my heresy.").

In general, I was quite disappointed with the limited capabilities of the authoring system, which is why the Inquisitor is perhaps less complete than it could be. Most of my good inspirations for conversation behavior would have required keyword priorities, better pattern matching (I managed to use some parts of standard Perl pattern matching, but others failed to work), or long-term memory (the latter would probably be the largets improvement). The Inquisitor as it stands is, I think, the best Inquisitor I can manage with the capabilities available.


Andrew Twyman, kurgan@mit.edu
Interactive and Non-Linear Narrative, Spring 1998