In general, one commits the straw man fallacy in a situation of criticism–when one challenges someone else's argument in anything other than its true and charitable form, one is in danger of committing the straw man fallacy.  Let me give an Al Gore example.

• Al Gore argues that curbing carbon emissions is critical to reducing the impact of climate change.  He points to numbers and charts and data and stuff like that.  I see what he's saying, he saying we should get rid of all of our cars!  

The claim after "I see what he's saying" obviously bears little resemblance to what Al Gore is saying.  It's funny, in fact, how often that claim–oh, I see what you're saying!–precedes a straw man.  It's like a straw man warning.  

Anyway, back to what I was saying.   The straw man fallacy admits of a couple of variations.  You might call the most common variation the "misrepresentation" form.  It consists in the distortion of an opponent's actual position.  Take the above example.  Al Gore argues for more sensible carbon policies, but he does not advocate the rapid elimination of the automobile.  Another form, recently discussed by Bob Talisse and Scott Aikin, involves selecting the worst of an opponent's argument for attack.  This one lacks the outright stupidity or dishonesty of the misrepresentation form, although it involves the false claim that the weak argument is the strongest one.  Talisse and Aikin call this "the weak man" argument. One other common form of the straw man, the one I see today in the work of a dear friend of the NonSequitur dot com, involves completely inventing an opponent and an opponent's argument, and then attacking that and claiming victory.  Call this the hollow man argument.

This is just what George Will does today.  He writes:

Reactionary liberalism, the ideology of many Democrats, holds that inconvenient rights, such as secret ballots in unionization elections, should be repealed; that existing failures, such as GM, should be preserved; and, with special perversity, that repealed mistakes, such as the "fairness doctrine," should be repeated. That Orwellian name was designed to disguise the doctrine's use as the government's instrument for preventing fair competition in the broadcasting of political commentary.

Because liberals have been even less successful in competing with conservatives on talk radio than Detroit has been in competing with its rivals, liberals are seeking intellectual protectionism in the form of regulations that suppress ideological rivals. If liberals advertise their illiberalism by reimposing the fairness doctrine, the Supreme Court might revisit its 1969 ruling that the fairness doctrine is constitutional. The court probably would dismay reactionary liberals by reversing that decision on the ground that the world has changed vastly, pertinently and for the better.

The only problem is that, as has been pointed out all over the place, no one advocates the fairness doctrine.  Will doesn't even name one person who supports the fairness doctrine in his article.  Yet he concludes:

If reactionary liberals, unsatisfied with dominating the mainstream media, academia and Hollywood, were competitive on talk radio, they would be uninterested in reviving the fairness doctrine. Having so sullied liberalism's name that they have taken to calling themselves progressives, liberals are now ruining the reputation of reactionaries, which really is unfair. 

This is really appalling, even for Will.  Normally he can muster at least a straw man.  But I wonder whether his inability to find someone to slime is a step forward or a step backward.