The boundary between opinion and expertise

There’s a line I love from The New York Times 2020 Report, on staffing:

We should continue to employ a healthy mix of newshounds, wordsmiths and analysts.

I like it for the simple reason that it acknowledges that great journalists can take many forms — indeed many more than that simple (and text-biased) categorization. But if “analysts” is one category of great journalist, perhaps it then further splits into the explanatory journalists and the opinion journalists. Both are certainly analysts more often than newshounds or wordsmiths, I’d argue, (though certainly some opinion journalists do have wordsmith tendencies). Where, then, is the line between opinion and explanation?

If the answer seems obvious to you, consider these lines. Here’s Dan Kahneman:

if a large body of evidence published in reputable journals supports an initially implausible conclusion, then scientific norms require us to believe that conclusion.

Here’s Bertrand Russell:

when the experts are agreed, the opposite opinion cannot be held to be certain

Both of these sentences hint at what to me is a somewhat haunting idea: can it ever be rational to hold, confidently, an opinion that every other deeply informed person thinks is wrong? It seems like we would want the answer to be “Yes”, at least sometimes, or else we’d end up with too much group think. And yet most good analysts would acknowledge that the aggregated judgment of a group of experts is superior to any one person’s view. Why exempt yourself from that rule?

OK, I won’t attempt to answer that whole puzzle in this post. But with it out there as motivation, I want to clip together a couple decent treatments of the distinction between opinion and explanation.

First, here’s economist Greg Mankiw writing in Foreign Affairs:

When economists write, they can decide among three possible voices to convey their message. The choice is crucial, because it affects how readers receive their work.

The first voice might be called the textbook authority. Here, economists act as ambassadors for their profession. They faithfully present the wide range of views professional economists hold, acknowledging the pros and cons of each. These authors do their best to hide their personal biases and admit that there is still plenty that economists do not know. According to this perspective, reasonable people can disagree; it is the author’s job to explain the basis for that disagreement and help readers make an informed judgment.

The second voice is that of the nuanced advocate. In this case, economists advance a point of view while recognizing the diversity of thought among reasonable people. They use state-of-the-art theory and evidence to try to persuade the undecided and shake the faith of those who disagree. They take a stand without pretending to be omniscient. They acknowledge that their intellectual opponents have some serious arguments and respond to them calmly and without vitriol.

The third voice is that of the rah-rah partisan. Rah-rah partisans do not build their analysis on the foundation of professional consensus or serious studies from peer-reviewed journals. They deny that people who disagree with them may have some logical points and that there may be weaknesses in their own arguments. In their view, the world is simple, and the opposition is just wrong, wrong, wrong. Rah-rah partisans do not aim to persuade the undecided. They aim to rally the faithful.

Journalists and economists do different things, but you could imagine that Voice #1 is explanatory journalism, whereas Voice #2 is good, responsible opinion journalism. (Voice #3 perhaps is bad opinion journalism.)

The next example comes from Paul Bloom’s introduction to psychology on Coursera. Consider the differences between these two statements:

I think it’s fair to say that nobody believes this anymore.  Just about everybody agrees to at least some extent that these three claims are mistaken.

And:

More recently, the philosopher Dan Dennett [has argued,] “Perhaps the kind of mind you get when you add language to it is so different from the kind of mind that you’d have without language that calling them both minds is a mistake.”

So, is this true?

Well, there’s an extraordinary amount of debate about it. My comments about animals and their lack of a human-like language are controversial,  but this stuff gets really controversial, and people have different views. So I’ll give you a sort of a warning, that what’s going to follow is my perspective. But my perspective is this…

Statement #1 is Mankiw’s Voice #1. Statement #2 adopts Mankiw’s Voice #2.
None of this addresses the deeper question of when it’s rational for an analyst to have an opinion that differs from expert consensus. But Mankiw’s framework and Bloom’s examples do show how to present those different opinions responsibly, and how they differ from the more “textbook” or “explanatory” approach.