3 ms·
Depends; you can could consider the mean to that of the whole world, or the whole of History, and seek exceptionalism relative to that. In case, we would seek
by ElFitz 7y ago
Depends; you can could consider the mean to that of the whole world, or the whole of History, and seek exceptionalism relative to that.
In case, we would seek to have a macro-scale program delivering exceptional results to all of our current participants.
- brighton36 7y agoIf we consider the mean to 'that of the whole world' then, our task is done. A child from today needs to merely watch television for an hour a day, in order to be wiser than the average human being across time. School would be unnecessary. Your last sentence isn't reasonable. We would test the performance of participants in your proposed program, and a bell curve would immediately arise.
- ElFitz 7y ago> A child from today needs to merely watch television for an hour a day, in order to be wiser than the average human being across time. Depends what you mean by wiser. It really seems we absolutely don't have the same definition. > Your last sentence isn't reasonable. We would test the performance of participants in your proposed program, and a bell curve would immediately arise. So what? Nowhere did I say it would deliver equally exceptional results to all the participants. Just that it's goal would be to deliver exceptional results to all current participants, compared to the means of the rest of the world and all of human history. How does your inevitable bell curve contradict or invalidate any of it? How does having internal variations invalidate any of it?