#d-84

An unrelated post on a serious topic reminded me of this bit from "the Robots of Death":

D-84: I heard a scream Doctor: That was me D-84: I heard a scream Doctor: That was me D-84: I heard a scream Doctor: That was me

So, the question is kind of, who's being obtuse in this scene?  Hint: it's not D-84.  D-84 is a robot, and just doing the best he can.

People (and presumably robots) say things for a reason.  We intend things by our utterances.  Clearly D-84 must intend something by "I heard a scream" otherwise he wouldn't be saying it.  The Gricean Maxim of Quantity says that utterances should be as informative a possible.  When D-84 says he hears a scream, he's adding the information "D-84 heard a scream" and "a scream occurred." to the common ground.  Once it's there, it's not informative to add it again.  So, the second time he says "I heard a scream" he must either be being a stupid robot who doesn't understand how language works (as the Doctor assumes) or he must be trying to convey additional meaning.  

Life Tip: ASSUME PEOPLE ARE TRYING TO MAKE SENSE

(this is also called the Cooperative Principle.)

So let's assume D-84 is a rational language user.  This means he correctly interpreted the Doctor's response of "that was me" meaning "the scream that you heard was a scream I made."  The Doctor is acknowledging the existence of a scream, and explaining its source.  Which means that the existence of the scream is in the common ground - it has been asserted by both conversation partners.  Which in turn means that in order for a further assertion about the existence of a scream to be meaningful, it must refer to a different scream.  Consider if we phrased it this way instead:

D-84: input: exist!scream Doctor: acknowledge: exist!scream D-84: input: exist!scream

Again, either D-84 is broken or there have been two screams.  But the Doctor is also repeating himself.  In order for "that was me" to be meaningful, it must not already be in the common ground.  Which means that the Doctor believes that D-84 did not understand "that was me."  Which means that the Doctor believes there was only one scream.  Which means that the Doctor did not understand "I heard a scream."  Which means D-84 needs to say that again too.  Ideally, the conversation should have gone more like:

D-84: I heard a scream Doctor: that was me D-84: I heard your scream, but I also heard a scream other than you.

Also, does the Doctor think that D-84 wouldn't recognize that he was the one screaming?  Presumaby the Doctor would know if he himself screamed, and wouldn't need to be told about it.  But the issue is D-84 is a robot, and thus has limited (although very good) use of language.  K-9 does exactly the same thing in with his "we are in a car" observation in "School Reunion."  The problem here is not robots failing to understand pragmatics, it's robots not having a good way to cancel presuppositions, and organics not realizing that robots do understand pragmatics in spite of that.

This can be legitimately played for laughs with robots because, well, they're robots.  When a human being does this, sit up and take notice.  Conversational partners intend their contributions to be meaningful.  If you encounter something like this, where on the face of it it is blindingly infelicitous, assume they're trying to make sense and find out what else they could possibly mean.