ErictheRed wrote on 02/23/08 at 02:44:32:
I disagree that 1.d4 d5 2.c4 c6 3.Nc3 dxc4 is at all inferior. But anyhow, the reason to prefer 3.Nc3 just boils down to this: what does White play after 1.d4 d5 2.c4 c6 3.Nc3 Nf6? If he just goes for the main line with 4.Nf3, there's no reason to prefer playing 3.Nc3. However, after 1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.e3!?, Black has no way of getting a main-line Slav. 4...Bf5?! is not good, 4...e6 it's a Semi-Slav, and 4...g6 is a Schlecter Slav (considered somewhat passive). So basically White avoids what most people consider to be "The Slav."
Your point about not being able to get a main line Slav is incorrect. After 4.e3 e6 5.Nf3 Nbd7, we have transposed to the Semi-Slav, which is in fact the main line Slav. However, it is classified in the Encyclopedia of Chess Openings and the Mega Database 2008 Opening Key under the rare move order 1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.Nf3 c6 5.e3 Nbd7.
Were you referring to the 4...dc4 variation or the 4...e6 5.Bg5 variation? The former one, while one of the main lines, is under a bit of a cloud these days, while 4...e6 5.Bg5 is actually not played as much as 5.e3, but it is the most fashionable line nowadays (mainly due to the efforts of Anand and many other strong GMs).
In conclusion: there are four "main lines" in the Slav as of now: 4.Nf3 e6 5.e3, 5.Bg5, 4.Nf3 dc4, and probably even 4...a6 (the Chebanenko) is something of a main line nowadays. So your point, while inexact, remains valid. What I think you meant to say was that by playing 4.e3, White cuts out the 4...dc4 variation, but cannot play the 5.Bg5 variation against 4...e6.
Returning to 3...dc4: How does Black obtain equality after 4.e4 b5 (4...e5 is not as strong) 5.a4 b4 and now 6.Nce2, with the idea of Nf3, Ng3 and Bc4, winning back the pawn with a superior pawn structure and stronger centre?
Oh for Pete's sake, the traditional distinction is between the mainline/classical Slav where Black aims to get his QB out (as in ...dc and ...Bf5) and the Semi-Slav (with early ...e6).
As to how Black might equalize after 6. Nce2, one could start by looking at ECO, which cites e.g. Nikolic-van Wely, Dutch ch 1999 and Gelfand-Huzman, Israel 2000. It doesn't share your assurance that 4...e5 is worse, however.