I’m going to disagree with Johnson – for myself, at this point in time – but then I’m going to defend his point of view for his time.
Personally, I just don’t feel that way about books. When I’m looking at, as he puts it, “a book of science”, I’m pretty much only going to get something out of it the first time (with a few exceptions, which I’ll detail more later in this post). If it’s a textbook, and I learn differentiation by parts, then I’m not going to get anything more the second time. If it’s a technical manual, and I’m learning how to do something, once I know it, the book is only a reference for me. And as a reference (or if the book is already just a reference book), it sits on the shelf waiting for me to look up something specific to refresh my memory – but it’s a refresh, and not a re-read for information.
A novel, on the other hand (which is the type of “book of entertainment” I’m prone to), if sufficiently deep – Wolfe, Herbert, Twain, what have you – will entertain me all over again. If I have read it before, I’ll get more out of the depth the second time. Maybe it’s because I remember the entire plot, and can place things in context. Maybe I see the foreshadowing better. Maybe the first time, I only picked up the action, and the second reading, I’m learning more about the characters, and the third reading, I’m using those parts to hone in on the themes… for whatever reason, I enjoy rereading good books. So I disagree with him.
But.
The “books of science” that I’m reading aren’t the ones that Johnson was reading. He was writing really pre- industrial revolution (or at the head of it), and science was in its, if not infancy, at least it was a toddler. Texts were much more likely to be full of unanswered questions and issues to think about. A science book published today is far more likely to act as though a question has been settled – whether it has been or not – and to try to be a “complete” resource on the subject. So it’s not the same comparison. His “arithmetic textbook” sounds terribly boring to me. But was it really like the Calculus text gathering dust in my basement, or worse, the algebra textbook that I didn’t bother keeping? Or is it more like a book by James Gleick, or Douglas Hofstader, that talks about a few of the issues that are known, and goes off speculating? Today, those aren’t considered “texts” – indeed, they are (to warp Johnson’s phrases) “science books for entertainment”, intended for the lay audience, to popularize science, explain it, and build an interest. I suspect (though I don’t know for certain) that the type of texts available to Johnson might well be more like this.
And I’ll tell you this – I could reread Hofstader’s Godel, Escher, Bach: The Eternal Golden Braid a number of times, and get something new out of it each time. And I call it a “book of entertainment”. Would Johnson, perhaps, have referred to the same text as a “book of science”?
Look at http://www.uh.edu/engines/epi1636.htm”]this link for a description of science books – even one hundred years after Johnson – and compare that to the science books we know.
And the “books of entertainment” we read – novels, thrillers, science fiction, fantasy – what was the equivalent in Johnson’s time? He was writing in 1773. Novels (to the extent that we call them that) weren’t at the same stage in development then, and many genres we have didn’t exist. Would Johnson have referred to Plato as a “book of entertainment”? I don’t know that he would have.
So no, I don’t agree with him – I would prefer to reread entertainment – but I can see that with the literature of his time, I might well have felt differently.