Actually (e:ajay) you are completely wrong. Pew did a study that showed during both 2004 and 2005 that 30% considered themselves to be Republican and 33% considered themselves to be Democrat. You can read the study here ->

Even with CBS's "objective survey" being weighted it came out to 28% Republican 37% Democrat. That = the textbook definition of skewed polling. To suggest that somehow this was a representative sample is amusing to me.
Anyhow, onward and upward.
I occasionally put myself through watching Chris Matthews' trainwreck of a show. (Hint - there is a reason why his ratings are among the worst in the talking head circuit.) This blithering fool Carter devotee (strike 1) actually suggested that it was a pity that the potential for civil war didn't start sooner.
"The problem is it took a little time for this (the potential for civil war) to take shape."
THE PROBLEM, Chrissy? Take a bow sir. You've now exposed yourself to the suggestion that you are willing to allow Iraq go down in flames as long as Democrats can gain politically from it in an election year. Any of you people who think that Democrats are going to somehow be smelling roses in November are kidding yourselves when the most prominent among you are begging, pleading and praying for our action in the Middle East to fail. That suuuure is going to be buying you votes, baby! P.S. I'm already deep in your squishy grey matter between your ears, lefties. Iraq is not a failure and won't be unless we do what you want, which is to give up. Of course, there is no meaningful public support for a pullout so you just might have to find a "sympathetic" judge to rule in your favor somehow suggesting that war is unconstitutional!
I actually pity the Democrats lately, because the DNC and the liberal bankroll has no idea how to handle itself right now. Say what you want, but Republicans are immeasurably better than liberals at political strategy. Admittedly this is the time where Democrats could clearly make a case to the American people that just might make a difference in November. Instead, during the darkest hours for the RNC, its clear that a) the likely Democratic nominee in '08 won't win a general election, b) nobody on DNC side is suggesting a new or innovative idea that voters would agree on with respect to how to handle national security that would be different from the Republicans, c) there is no apparent strategy or platform for the '06 elections other than what Carville and Begala are amusingly suggesting - "Bash Bush!!!"
Lastly, my dear friends, many of whom may be incensed by my right-leaning, well articulated sensibilities at this point in this journal post. You know that this is all in good fun and I love my leftie friends, but you have to admit that its hard to be a credible liberal when you get your news from a fake news show, or in the case of NPR's Nina Totenburg when she was on "Inside Washington" recently, you say that you don't root for Americans to win medals at the Olympics and that the Salt Lake Games were somehow spoiled by American "nationalism." Well gee, apparently rooting for your country during the Olympics is now a crime!
(e:Joshua), your understanding of statistics is deeply flawed.
Try this experiment: toss a coin 10 times. Count how many times you get heads and tails. See how often you get exactly 5 of each.
"Sampling" is a very complicated subject. The poll you originally griped about had a sampling error of 3.3%; this means that the results were "probably" within plus/minus 3.3% of the "actual".
There are so many variables in sampling, which is why there are business built around it (Gallup and Zogby, for instance). I can go on and on about how difficult it is to get a right sample, but it'll just (a) bore the others, and (b) never convince you otherwise anyways... ;-) :-D