Monday, May 11, 2009

The Fallacy of Expert Witness


Andy Naselli reports
on a very helpful section from Greg Koukl's new book, Tactics. He writes:

One of the most helpful chapters is entitled “Rhodes Scholar.” It’s one of eight chapters in part 2, which is devoted to finding flaws in arguments. In order to discern if an appeal to an authority is legitimate or if it commits the fallacy of expert witness, Koulk crisply distinguishes between information and education:

The ["Rhodes Scholar"] tactic hinges on the difference between informing and educating. When an article tells you what a scholar believes, you have been informed. When an article tells you why he holds his view, you have been educated.

Here is why this distinction is so important. If you recall from chapter 4, an argument is like a house whose roof (what a person believes) is supported by walls (the reasons why he believes). You cannot know if the reasons are adequate to the conclusions—if the walls are strong enough to hold the roof—unless you know what those reasons are. If you know the reasons, you can assess them. Without them, you’re stonewalled.

Popular articles always inform, but almost never educate. As a result, you have no way of evaluating a scholar’s conclusion. You simply have to take his word for it. But scholars can be wrong, and often are. Their reasoning can be weak, their facts can be mistaken, and bias can distort their judgment. (pp. 167–68)

The key, Koukl aruges, is simple: “Always ask for reasons. Don’t settle for opinions” (p. 168).

So how should we evaluate whether we should believe a expert’s opinion? Koukl suggests two ways.

[1]The scholar may be in a special position to know the facts. However, if an authority is in possession of special information that guides his counsel, then he should be able to point to that evidence to convince us he’s on the mark.

Sometimes authorities give opinions that are outside of their area of expertise. . . .

In a court of law, the expert witness is always cross-examined. Credentials alone are not enough to certify his testimony; he must convince a jury that his reasons are adequate. . . .

[2] Sometimes a scholar is in a unique position to render a judgment. More than mere facts are in play here. Interpretation is needed.

In this circumstance, you face another pitfall. A scholar’s judgment may be distorted by underlying philosophical considerations that are not always on the table. . . .

Sometimes one’s destination is predetermined by where one starts. (pp. 168–69)

I have been listening to Greg's talk show, Stand to Reason (podcast here), in recent days and would highly recommend his show and this book! He will help you think through how to winsomely engage with those who don't see Christ as worthy to be trusted.

2 comments:

Unknown said...

I listen to Koukl's show also. I'm a huge STR fan.

I've heard him say that when someone quotes an expert, (example: "Dr. so-and-so concludes XYZ,") we might respond: "Oh, and how did he come to that conclusion?" - that simple question reveals what this post is after... what exactly are these facts that lead this expert to their conclusion...

Douglas said...

While reasons help people evaluate the claims, oftentimes, when it comes to expert witnesses, the general public will not have the background to properly asses the reasonableness of the explanations. I'm thinking in particular about medical and engineering witnesses. Asking for explanations and references is extremely helpful, but only if the person has the educational background to properly asses those sources. Sadly, in my opinion, most Americans don't have the critical thinking skills to go through that process when it comes to deep concepts.

As an engineer, I have the research skills to understand that the rubella vaccine is cultured in aborted fetal tissue (which sadly most doctors will deny vigorously until shown proof to the opposite), but I have little background to understand more detailed medical problems/issues. It doesn't take long before I have to start trusting people and relying on my bullshit meter and inferential analysis tactics like self consistency and majority opinion in the medical community to help me sort through the facts on medical issues. I simply don't have the medical background to understand the issues fully and make reliable judgments on my own. Just given the facts on complicated issues, apart from the opinion of expert witnesses, I'm convinced most people couldn't separate the fact from fiction. So, while people shouldn't settle for just facts, ultimately, IMHO, properly placing one's trust in an expert is the only way for most people to reliably evaluate technical questions.

MB