Having run into many dead-end discussions on the topic of atheist versus agnostic I thought the following article summary offered a pretty good way to clear up the maddeningly ambiguous and contradictory definitions of "atheist" and "agnostic".

The article, written by Mano Singham for the latest New Humanist, is summarized on the Why Evolution is True website.  Without further ado, here it is . . .

The latest New Humanist has an short article by Mano Singham that tries to untangle the confusion around the terms "atheist" and "agnostic".  As we know, one can define these terms so they become essentially equivalent, even though some people choose the "agnostic" label simply because it seems less confrontational.

Singham first looked up the definitions of the terms in the Oxford English Dictionary, and found this:

atheist:  one who denies or disbelieves in the existence of a God.

agnostic:  one who holds that the existence of anything beyond and behind material phenomena is unknown and (so far as can be judged) unknowable, and especially that a First Cause and an unseen world are subjects of which we know nothing.

The problems here are obvious.  For "atheist," "denying" God is not necessarily identical to "disbelieving" in God.  The former is absolute certainty, the latter allows for some hedging, i.e., "I see no reason to believe in God."  For "agnostic," Singham says, "the definition fails to distinguish between not knowing something and there being nothing to know."  In other words, we can "know nothing" about God because either it's logically impossible to know (i.e., the [false] claim that "you can't prove a negative"), or because there could be evidence for God but none has appeared.  The two parts of the agnostic definition—those parts separated by "and especially"—are contradictory.

These problems are why many of us see those who call themselves "agnostics" because they adhere to the "no -evidence-has-appeared" idea, but want to distinguish themselves from the nasty atheists who say the same thing, as intellectual cowards.

Singham's solution: deep-six the term "agnostic," and redefine "atheist" to eliminate these ambiguities:

atheist:  One for whom god is an unnecessary explanatory concept.

He explains the advantages:

This definition leaves little room for agnostics because they will have to answer the question as to whether they think God is necessary as an explanatory concept for anything. If they say "no", they are in the same camp as atheists. If they say "yes", they are effectively religious and would be required to show where the necessity arises.

Although this sounds like a rhetorical strategy to force people to admit they're atheists, I actually like it.  It subsumes in a logical way both people like P.Z., who don't think there can be evidence for a god because the very concept is incoherent, and people like me, who think that in principle there could be evidence for a god, but none has appeared.  Likewise, it subsumes those who are certain that there is no god (#7 on the Dawkins scale) with those, like Richard himself, who are highly doubtful but not absolutely certain.  And it's not just conflation of wildly disparate views, for it separates people on a crucial axis: whether or not they think we need a god to explain and understand the world.

The only problem I see is that of pure deists, who may claim that although God isn't needed to explain anything, he's still up there anyway.