I must have spent ten hours trying to come up with a glyph to represent “defeated in dream combat”, and I’ve had great difficulty. Here are just a few of the ideas that ended up in the trash can:
I really like the mushroom cloud, but it’s difficult to recognize. I think that the best choice is the flame on the far right. Thus, I now have the following spaghetti diagram for conversation:
Compare this with the earlier one; the only difference is the replacement of “tell about promise” with “tell about betrayal” and “give bead to” with “tell about defeat”. These are all useful bits of information. However, they cry out for confirmation verbs. If Joe tells Mary that Jane badmouthed Mary to Tom, then Mary would want to be able to go to Tom and ask “Did Jane badmouth me to you?” That turns out to be easily implemented: all I need add is a single new verb, “ask question”, denoted with a big question mark inside a verb frame. That verb would take a simple main clause for its wordsockets. Here’s what the results might read like:
Mary asks Tom whether (Jane badmouthed Mary).
Mary asks Tom whether (Jane betrayed Tom).
Mary asks Tom whether (Jane lied to Tom).
Mary asks Tom whether (Jane told about defeat Mary).
Mary asks Tom whether (Jane told about aura count telling Mary).
This last sentence is too clumsy to retain: “Tell me if Jane told you that I told Jane about your shial aura count.” That’s too messy to include. Now, I could add wordsockets to handle a few of these situations, but I think it’s better to have one verb for “ask question" and then fill in just a main clause; that’s enough to verify the truth of the original statement.
Of course, we could get into a whole new can of worms by adding the verb “Who told you that?” But I think that’s a level of complexity to avoid just now. Later on, if the game lacks sufficient complexity, I can always put that in.
Why do any of this?
So now I must answer the question, how does this fit into an actor’s strategy? Why would an actor do any of these five bottom verbs? In every case, use of the verb increases both the DirObject’s pHonest and pGood values. They would also serve to diminish the DirObject’s pHonest and/or pGood values for 4Actor. If the statements are true, then the Subject always benefits from their revelation. And remember, you can’t tell somebody something that they already know, so it benefits you to blab sooner than the competition.
But is it ever advantageous to use these verbs falsely? They’ll still have the beneficial effects just mentioned, but they run the risk of being exposed as lies. The decision hangs on the probability that the listener will check up on your claim, which depends on how unexpected the statement is, and how much the listener trusts the speaker.
This brings up the matter of belief in the truth value of statements. A listener should react by first deciding how much belief to invest in the statement — this number is already in the engine and so can be properly recorded. If the listener’s belief is too low, the listener can respond with a new verb: “I don’t believe you”, or perhaps it should be thought of as “That’s not true.” Which in turn opens up a new can of worms.
How is the speaker to respond to “I don’t believe you”. Should some degree of certainty be attached to the statement, as in “Fred tells George that Albert lied to Jimmy — with low confidence.” This allows actors greater ability to refine their statements to avoid accusations of lying.
Is this too gray? Should I keep statements like this black and white, true or false? I have already established the use of uncertainty adverbs, such as are depicted in the previous essay. This definitely allows actors more elbow room to chat away about lots of things. And I can certainly handle the algorithms for computing results with intrinsic uncertainty.
I’m going outside to work up a sweat and mull this over.
