way back on March 27, 2002, Brad Delong wrote:
I have a five-point scale along which to rank commentators: category I try hard to shoot straight all the time; category II use strong arguments where they are available, but get excited and use weak arguments when they are sure they are right but the only arguments for their side are weak; category III don't care whether arguments are strong or weak, leading or misleading, they'll just use whatever sounds best; category IV will baldly lie; and category V are sufficiently clueless that they don't know whether they're telling the truth or not. By this scale, Stephen Moore ranks no better than category III: even when the good, straight, accurate arguments are on his side, he's likely to prefer something twisted and misleading that makes a better soundbite.
Let me give some examples. For category I my favorite example is someone like Alan Blinder, who always tries to shoot straight. The most recent example of category V I have run across is Ben Stein: consider his recent claim that the 1929-1939 Great Depression was caused by the New Deal; now at some point in the past Ben Stein probably knew that the New Deal belonged to Franklin Roosevelt and that Franklin Roosevelt took office in 1933, but that knowledge seems to be so long gone that Stein doesn't even know that he has a bad temporal sequencing problem; he flunks the Turing test--it's not at all clear that there is a mind at work in there. For category II--well almost all of us are in category II. For category IV, my best recent example is Robert Scheer's claim that the Bush Administration was subsidizing the Taliban government last spring by giving them $43 million--a claim that it is impossible to put forward in good faith once one has looked at the record.
I would like to add to that, by characterizing the preferred style of various people's arguments, rebuttals and refutations: Category 1 peoples' instinct is to go for their opponent's most relevant and persuasive arguments, and to correct any obvious, easily correctable errors their opponent may have made, so as to get to the real substantive dispute at hand. category 5's refute their opponents weakest arguments, whether or not they are relevant to the important issues, so as to establish a quick "win" by *avoiding* the core substantive issues at hand.