A brief introduction to deductive reasoning. At its most basic, deduction attempts to show that a conclusion necessarily follows from a set of premises or hypotheses. A deductive argument is valid if the conclusion must be true if the premises are true (and consequently its premises cannot be true if its conclusion is false). However, the premises don’t have to be true for the argument to be valid. A deductive argument is sound if it is valid and its premises are true.
Good deductive arguments are both valid and sound—that is, the conclusion must follow from the premises, and the premises must be true. Most punditry takes the form of deductive reasoning: all democrats are bad, this policy was written by a democrat, therefore this policy is bad. For the most part, such arguments are valid, which perhaps explains why they gain such traction in the partisan press; very few such arguments, however, are sound.
None of this is terribly consequential: learning basic logic structures is literally the first step in learning to think critically. But it is nevertheless worth keeping in mind as we ponder the war in Afghanistan—especially given the extreme volume of sloppy logic pushing eternal success. This week in the New York Times, for example, Carlotta Gall wrote an interesting piece about the pressures the Taliban are facing in Pakistan:
The Afghan Taliban are showing signs of increasing strain after a number of killings, arrests and internal disputes that have reached them even in their haven in Pakistan, Afghan security officials and Afghans with contacts in the Taliban say.
While the arrests have been conducted by Pakistan security forces, it is not clear who is behind the killings. Members of the Taliban attribute them to American spies, running Pakistani and Afghan agents, in an extension of the American campaigns that have used night raids to track down and kill scores of midlevel Taliban commanders in Afghanistan and drone strikes to kill militants with links to Al Qaeda in Pakistan’s tribal areas.
Others, including Pakistani and Afghan Parliament members from the region, say that the Pakistani intelligence agencies have long used threats, arrests and killings to control the Taliban and that they could be doing so again to maintain their influence over the insurgents.
What’s so fascinating about this is how vague it is. Taliban figures are definitely being harassed, but no one knows by who or for what purpose. Moreover, some of the commanders Gall quoted don’t seem attached to how this harassment is affecting Taliban operations in Afghanistan. For example, she paraphrases a commander from Kunar province as saying that he’s lost so many men he’s considering switching sides to the government for money and jobs (she then quotes a Pakistani official about what it means, but never directly quotes anyone from the Taliban). Yet just two days beforehand, the New York Times had run a story about how the Taliban were solidifying and expanding their control of Kunar province. (The eastern part of the province has seen a renewed push by U.S. forces, but that’s not what Gall was referring to. Weird.)
Anyway, so this is typical of what has become Carlotta Gall’s output at the Times: thinly sourced, with conclusions in the title and lede unsupported by the evidence she presents in her piece. It would be shaky grounds on which to base a deductive argument, as it might not form a premise that is true (which is necessary for a valid and sound deductive argument). Don’t let that stop Max Boot:
No one seems sure. But what is happening in Afghanistan is clear: U.S. forces are ratcheting up the pressure on the Taliban, as I saw for myself on my latest visit to Afghanistan a few weeks ago… This is the product of the counterinsurgency campaign that General David Petraeus is directing. Its progress is palpable.
To break down Boot’s argument into pieces:
- Reporter reports that Taliban figures in Pakistan are probably being harassed but doesn’t really know why or by whom;
- Max Boot acknowledges that no one really knows who, what, or why, but notes that the U.S. has been “ratcheting up the pressure on the Taliban;”
- Therefore, the harassment of Taliban figured in Pakistan is the result of General David Petraeus’ counterinsurgency campaign.
On a very basic level, this doesn’t pass a basic deductive logic reasoning test. The premises are disconnected from each other, and neither implies a conclusion, yet Max Boot used them both to justify the conclusion that General Petraeus is a timeless genius who is winning the war singlehandedly (and who, naturally, just needs a little more time to make it all work, though how much is anyone’s guess if he’s really quitting after only a year in charge). His argument is neither valid nor sound. It is hagiography disguised as argument—which is, again, sadly, typical of him.
But Max Boot isn’t the only one to engage in logically flawed thinking. In the Washington Post is a bizarre op-ed by Craig Charney and the normally smart James Dobbins about how Afghans are all optimistic about their future. In the interests of space, I’ll spare you the quoting from their very flawed survey and summarize the argument’s logic (thank you, Michael Cohen, for laying it out this simply):
- Things used to really suck in Afghanistan; but they suck a bit less today;
- Things suck in Afghanistan; but in other places people think their countries suck more;
- Americans think things are going badly in America; so they think things are going badly in Afghanistan;
- Therefore, things are going well in Afghanistan.
Even accepting all the premises are true—and there are many reasons to think they’re not—this is an invalid argument, with no reason to accept the conclusion must necessarily follow from the premises. That the premises are not necessarily true make the argument unsound. It is a terrible argument, in other words, as it relies on the structure of deductive reasoning without meeting the basic requirements of a deductive argument. Both men who wrote that piece should know better and should be embarrassed at the shallowness of their argumentation.
Sadly, this sort of thing is depressingly common when it comes to hyping the war in Afghanistan: sloppy logic, untrue premises, dishonest conclusioneering (neologism!), and thinly veiled hagiography of senior U.S. officials. It doesn’t mean that the war is automatically going poorly, though I see plenty of evidence to suggest it is. All it means is that the case for believing the war is going well is often on the shakiest of grounds, and not at all grounded in reasoning, logic, or sound argument.
Of course, while we philosophize the glories of poor deduction, we must consider Hume’s Problem of Induction. But that is probably a discussion for another time.