Value rather than Truth, as a conserved quantity in language analysis

Many phrases have a subjective 'good' or 'bad' attribute. It is subjective whether a person, hearing a sentence, feels it as 'good', 'bad', or 'neutral' but it is usually a value that is shared by the person's culture. So saying: "we defeated the Russian troops" might be 'good' for an American, it might be 'bad' for a Russian. Also you can usually imagine examples where a surface neutrality is changed by adding some context. So "the tomatoes are red" might be neutral except in a discussion of their ripeness.

I was telling some friends, the Cole family of Shallow Pond Rd, about this idea of value being in the universal soup where the phrases live and the Tim asked about the sender/receiver relationship within communication and how value worked there. I hadn't thought about it but you can imagine having problems communicating with people who are assigning different values to the phrases exchanged.

So I have been thinking more about value and its properties as a quantity preserved or lost when combining phrases. The Greeks and later British were darned clever to come up with a True/False quantity that was always present in a proposition and it was preserved and modified by rules for combining and operating on proposition. Tracking that quantity almost made you feel like you are doing something useful! Anyway, the rules became Bool's laws and those, in turn, became the basis of computer architecture. But I believe value has its own rules. For example
  
a and b

           G B N
     G   G N G
     B   N B B
     N   G B N

a but b
          G B N
     G   N B N
     B   G N N
     N   N N N
   
a although b
          G B N
   G    N G N
   B    B N N
   N    N N N

These don't work too well because there are a variety of un-semantic constructs. No one says "but" between two good things. It would be considered incoherent. Also it would be incoherent to say "but" before something neutral. There is another aspect of the incoherence where the 'a' and 'b' of the proposition are not on the same channel of value. "We won the battle of Stalingrad but the jasmine bloomed later than usual that year". You would need to combine two very different conversational contexts to make this sentence seem rational. So trying to have a rule based analysis just won't work for good/bad.

So another thing I was talking about with the Cole's was that the Greeks and British avoided subjectivity it choosing to analyze Truth. This pinned them against a wall of trying to explain communication, without people being involved. I would have been pretty easy to explicate 'Truth' by reference to how people's belief is modified.

But anyway, it becomes clearer that conversational context and the subjective response to a proposition must themselves be the subject to an algebraic treatment. If you want to be algebraic, maybe it is about operations on conversational contexts. Hmm...