Today's Politico.com commemorates the sitting of the Constitutional Convention in 1787:
Delegates from the large states found themselves at odds with the small ones over how to apportion the legislative branch. The compromise called for a bicameral system, with proportional representation in the lower house and equal representation (two senators per state) in the upper house. All revenue measures would originate in the lower house.
Proportional representation, proportional voting or just plain "PR" debuted some time in the middle to late 1800s. This later PR - and the one we learn about in Intro to Comparative Politics - is about parties receiving the same percentages of seats as they win in votes.
When talking to non-poli sci majors about PR, I often get: "Don't we already have that?" They mean proportional representation a la Politico. Or more precisely, apportionment of seats in proportion to state population. Every member of the U.S. House theoretically has the same number of constituents. In the old days, before we all used single-member districts, there was the question of how many Reps each state should elect. The answer was: in proportion (to population). One of my first tasks at FairVote was to find in the Federalist every occurrence of the term "proportional representation." There were five, six or seven - I don't remember - but this 'old sense' of the word was the intended one.
With winner-take-all elections, proportionality of population doesn't translate to proportionality of seats to votes for either party. Whence one key normative question: what's a fair electoral system?
Several Founding Fathers came up with formulas for determining what proportion of Reps each state would get: the Jefferson, Hamilton and Webster methods, for instance. More interesting: these formulas are extremely similar to later formulas used in the other kind of proportional representation (to apportion seats by party, you might say on the Continent).