The Argument Sketch (Photo credit: Wikipedia)[/caption]
There are many ways to argue deceptively or mistakenly, and often these ways involve logical fallacies - defects in arguments that come in two general sorts, formal and informal.
The former are defects in structure, errors in the pattern of the argument that render it invalid, usually independent of the argument's specific content.
The latter are often violations of procedure in argument rather than structure, attempts to rhetorically thwart the goal of a critical discussion, often to 'win' the argument rather than reach a common understanding, rather, to obscure truth than reach it.
Fallacy theory is a complex subject, and not all logicians agree on the definitions I've just given, but that's cool. Anyway, fallacies can be used to distract and mislead, or they can be used in reverse, in labeling argumentative tactics, to reduce their effectiveness by calling attention to them when abused in an informal discussion or a structured debate.
In some debates, such labeling will be done fairly frequently, in others, more subtle counterstrategies will be used rather than explicitly pointing out the fallacies by name. In this case, knowing thine enemy and naming it is useful, subjecting the flawed argument to scrutiny, and lessening its sting.
But that is not enough.
To argue that naming a fallacy shows the claim of an argument false is to commit the fallacist's fallacy, as it it entirely possible for an argument may be flawed, the conclusion not logically following from the argument's premises, but that conclusion may by some chance be correct despite the argument made for it.
It is not enough to label an argument. It's also important to show not only that something is a fallacy, but why it's one as well. It's often necessary to show the argument flawed in some manner, but also why that fact is relevant to the claim being made, why the argument doesn't follow (even with true premises), and that the suspect argument must therefore be thrown out as inadequate support for the claim.
That's especially the case with logical inconsistencies and outright contradictions. If two or more arguments work at cross-purposes, logically at least one of them must be rejected, but the natural human inclination is to throw them all out, as reflecting poorly on the arguer's credibility and maybe resulting in loss of the argument.
Even fallacious arguments sometimes assert things that are true, but they cannot reliably show those things to be the case. Logic alone proves nothing, even if the argument is perfectly valid or uses strong inference, for that, we need data as well.
We need the best evidential data available to support our premises, which must not only be true, but acceptable, clear, relevant, as non-circular as possible, and as free from unwarranted assumptions and presumptions as can be had.