I'm reading Braudel and came across this quote. "But the laws or so-called laws of economics probably last only as long as the desires and realities of the period they reflect or interpret more or less faithfully." (182)
This should be required reading for any economist, and more generally any social scientist. It seems this should actually be obvious, conditions and institutions change so much that many of the assumptions, including behavioral assumptions, should only apply to a limited range of situations. If something really unexpected happens, it should be asked if there has been a shift in fundamental assumptions.
What Braudel was getting at though, was that Say's law did not hold in the pre-industrial economy. People were too poor and the income that workers were given was insufficient for more than basic needs so they could not stimulate demand for manufactures since they could not afford any. Agriculture remained too uncertain and productivity could not be increased enough to drive costs down, other factors constrained agricultural growth on the supply side despite a demand for it that could not be filled. Only when prices came down far enough in the industrial revolution could the poor participate in the manufactured goods market and the economy really take off (of course something must have happened in agriculture to but he doesn't address that in this section).
This was just a particularly interesting detail I came across that wouldn't fit into the book review on Sunday.