> Bob and Sally are things that people already understand and relate to.
People understand they're people's names. But that's not a relevant detail for the equation, so why add that to the problem? It complicates, rather than simplifies, my understanding of it.
> !(Milk || Sugar) == (!Milk && !Sugar), duh.
I don't get what the 'duh' is. You've added nothing to the equation! They're just different names!
for a similar example, see [1]. summary: a study presented people with a logical problem. they failed miserably if the question was about abstract math-y stuff, but got it right most of the time if the variables were like `X is underage` and `X can buy beer` (explained in the "Policing social rules" section)
---
[0] as a bonus, we can quickly sanity-check if our reasoning is correct – if there's `no (milk or sugar)` in my coffee, there's obviously `(no milk) and (no sugar)` in it!
This brings up an interesting idea, though - are there any programming languages that "learn" what the user wants by example? What would that even look like?
Bob and Sally are things that people already understand and relate to. a and b are not, so there is an extra layer of indirection there when attempting to reason about them.
!(Milk || Sugar) == (!Milk && !Sugar), duh.