I came up with an additional modality of interaction: threats. Actually, I have had this idea for a long time; I simply forgot and then remembered it. A threat is an inverted deal: instead of requesting an action, the threatmaker demands it as a preventative for some a negative action. I can think of but two negative actions:

I will attack you.
I will share my information about you with everybody.

This latter could be considered ‘betrayal’; instead of defining betrayal as a single divulging of information, we define it as a broad divulging of information. I think I need another term. ‘Betray your aura-counts’? ‘Tell the world’? ‘Announce your aura-counts’?

Here’s a third possibility: I will badmouth you. This suggests its opposite as a deal-request: goodmouth me. I like that. But how can it be verified? How could an Actor ask others about it? Ah, yes, with the news! But would this be intrinsic to asking ‘What’s new?’ or would it be a special question, such as ‘What have people been saying about me?’. I think this last is best.

But should goodmouthing (or, for that matter, badmouthing) be specific to a 4Actor? In other words, is it better to specify that you want DirObject to goodmouth you to a specific person, or do you want to make the request general? Verifiability suggests the former, but utility suggests the latter. 

I think that I’ll go with the 4Actor-specific version, because that also permits a specific question: Has 4Actor spoken about me? 

Conclusion: I have four new Verbs to add:

Threaten attack
threaten betrayal
threaten badmouth
promise goodmouth
request goodmouth

Wait! There’s one other issue: should goodmouthing be confined to a single aura? That is, does the request apply to any aura or to a specific aura? This puts the recipient in a tricky situation: if the request compels him to lie, then he’ll be seen as a liar. No, that’s out of the question. The recipient must have freedom to choose an aura to goodmouth.

On a completely different topic, I have also been wondering about how to handle interstitial stories. I intend to use the same basic structure I used in the original game, using text as the medium of delivery. I am prepared to write new code to handle this, but should I integrate the system into the main engine? There are two ways to handle interstitial stories. The simplest way is to have the engine trigger them randomly and then simply jump out to completely different code. I fear, however, that this method might make it difficult to integrate the results of the interstitial stories into the main story. Still, the adjustments to pGood, pPower, and pTruth are probably enough. 

So I’ll have a lot of messy programming to tackle for interstitial stories.