Then I am confused how you could agree that at the microscopic level there is only cause/effect (determinism or randomness - but no choice) - which is what I think you agree with - ...
So far so good.
Yet, there is the caveat concerning "choice". What does "choice" necessarily imply?
...
yet at the same time introduce something that offers at the macro level precisely what is not possible at the micro level... without the macro-level being illusory.
First, the macro level is illusory (which is to say, at our level of observation, we do not (cannot) observe all that there is to see..).
Second, and more importantly, at the macro level (ours) deterministic causality is not exhaustive. Which is to say, although causality still obtains, the direction it takes is not strictly determined. As active agents, we have the power of election.
If you define "choice" as purely a macro observation or ability or whatever then you surely limit the understanding of what is really going on.
Yes.
And given that determinism is very much a micro-consideration, surely the discussion of free-will (given the thread title) and thus "choice" should go to those same levels... i.e. what is going on at the core... not just what is observed at the macro-level.
I agree with you on the continuity of applicability. However, see my first comment above regarding the term "choice". Also, again, as agents, we affect the macro level system (which is to say, although the 'rules of the game' haven't changed, the players have..).
You seem to want to accept the micro-level but ignore the implications for the macro: ...
Not at all; but there's more to the situation than the rules.
If at the micro level their are only things/actions that are either caused or random (such as radioactive decay) - how do you propose there to be something at the macro level that is both uncaused and non-random - i.e. the requirements for "choice"?
I don't agree with your definition of "choice" meaning uncaused and non-random.
In any case, at the macro level, we're talking about agents that have a special kind of comprehension of the system within which they operate. Though the system is causal, it doesn't necessarily follow that the behaviour of an agent is strictly such.