The relationship between the Validity of Natural Language Expressions and the Statistical Concepts of Accuracy and Precision

As you may be aware, I recently started reading a book that was quite the bestseller a few years ago (see e.g. my initial reaction to the quote: “The problem is that the pervasiveness of technology and mass marketing is screwing up a lot of people’s expectations for themselves: the inundation of the exceptional makes people feel worse about themselves, makes them feel that they need to be more extreme, more radical, and more self assured to get noticed or even matter“).

At first, I was very impressed that Mark Manson was apparently able to convey quite complex concepts in a very simple, colloquial language. I am now in the midst of going through the text with a more “fine-toothed comb”, to see whether I still find the ideas to be as impressive as the first time around.

Spoiler alert: Not quite.

Yet at the same time I am acutely experiencing a phenomenon that both enabled this to become such a runaway bestseller, but also is to some degree one of its primary pitfalls.

As I mentioned in the previous (“initial”) review, Mark is prone to making bold statements — and I might also add: in a very simple language, with little sophistication in his mode of expression. He begins by drawing a sketch of a drunkard literary genious — Charles Bukowski. We thereby are confronted with a wanton spirit, flailing about and occasionally hitting a nail on the head (I myself have not read much Bukowski, perhaps little bits and pieces, mainly short quotes noted in passing by some person apparently not secure enough to use their own words).

If we map out the validity of linguistic expressions, we may think of that as a scatter plot, with each expression located somewhere along axes measuring accuracy and precision (at least theoretically — I am not saying this might be simple or easy to measure). I think we can all agree that our immediate gut reaction would be to aim for high degrees of both accuracy and precision. See, for comparison, these images (source: “Accuracy and precision – Wikipedia“):

High accuracy, Low precision
High precision, Low accuracy

The point I particularly notice is what happens when our linguistic expressions are imprecise: In such cases, there may still be sporadic occasions of arguments that seem valid — for example, an expression that in a particular context seems to hit the nail on the head … but which, in a different context may appear completely baseless and untrue.

In this vein, making bold statements in a muddy or colloquial language seems to be less a work of art than making exact statements in a meticulous or scientifically explicit language. Although the former may seem easier to grasp, the latter appear more reliable.

In the meantime, I will continue to withhold judgement on Mark Manson’s words. I remain somewhat mesmerized, but unreliably so. 🙂

This entry was posted in remediary and tagged , , , , , , , , , , , , , . Bookmark the permalink.