For many years a thought has routinely come to mind:
An unliterary man may be defined as one who reads books once only.
I had forgotten who said it or where I read it, but looking it up now it comes from CS Lewis.1
I thought of this often when encountering comments made by readers of important books, where the reader either completely misunderstood the text or seemed to lack the requisite background and contextual knowledge to accurately interpet it.
I recently encountered an even more severe version of this principle. Gaston Bachelard, in The Poetics of Space, describes the first reading of a book as a "too passive" affair.
But every good book should be re-read as soon as it is finished. After the sketchiness of the first reading comes the creative work of reading. We must then know the problem that confronted the author. The second, then the third reading...give us, little by little, the solution of this problem. Imperceptabily, we give ourselves the illusion that both the problem and the solution are ours. The psychological nuance: "I should have written that," establishes us as phenomenologists of reading.2
Criteria for what counts as a "good book" may vary from person to person. But I can't imagine any genuinely good book with respect to which it wouldn't be worth establishing oneself as a "phenomenologist of reading".
How many books can you think of that you'd like to read again?
I can't think of one I've read multiple times, where I didn't gain a great deal from the effort.
Lewis, C.S., On Stories: And Other Essays on Literature, Houghton Mifflin Harcourt (October 28, 2002), p. 16. An additional remark from the same section fills things out a bit: " We do not enjoy a story fully at the first reading. Not till the curiosity, the sheer narrative lust, has been given its sop and laid asleep, are we at leisure to savour the real beauties. Till then, it is like wasting great wine on a ravenous natural thirst which merely wants cold wetness." pp. 16-17. ↩
Bachelard, Gaston, The Poetics of Space, Boston, Beacon Press (1964), p. 21. ↩