Read Old Books
grasek.com
grasek.com
The Lindy Effect applies to socially-integrated phenomena and informational phenomena --- not just books, but for example, cultural practices and genomes. It's less the age than the widespread adoption which signifies their value.
And what matters for these isn't age per se but quality or fitness. "Read old books" is for the most part a failure to comprehend the notion of "read good books".
That misdirection itself has a number of components:
- Old works may well be high quality and contain valuable information.
- Old works tend to be more deeply integrated into cultural knowledge and practices ... even where those works themselves don't have information or models of particular quality. The error-prone nature of early philosophical works and gross mischaracterisations of natural phenomena found in many religious texts come to mind. (Such works may well have value on the basis of other insights, or simply through the understanding conveyed in recognising their cultural influence.)
- The old works which have survived tend to be those which had been found to be most useful. This especially applies to books dating from before the introduction of moveable type to the Western world in the 15th century.
- There are vastly more books published today than in historical times. Total cumulative output of books through the 18th century was on the order of 1 billion copies. Assuming 1,000 copies per volume, that's perhaps roughly a million titles. There are over 1 million titles printed each year presently. One would presume that the average value and quality per title may have fallen somewhat.
That said, there are still excellent books being produced and published, though I'd suspect that the rate of production of seminal works, fiction or nonfiction, is roughly the same as it has ever been. Lists of "great books" or canon tend to number on the order of 100 titles, accumulated over 2000+ years of history, or at the rate of a new entry every 20 years. In practice, the rate has likely increased as human population, literacy, and knowledge have greatly increased, by roughly two orders of magnitude since the time of Ancient Rome, and looking at lists of great works, we find that though there are some ancient works (older than, say, 200 years), most date from the past century.
The other issue is that greatness takes time to recognise, and that authorial reputation greatly lags publication dates. From a list of most influential authors selected by decade from the 1950s through the 2010s, the youngest authors present tended to be in their 30s, whilst the oldest could be into their 70s (Cormac McCarthy, popular in the 2000s, born in 1933). See: <https://diaspora.glasswings.com/posts/3e1b15003c27013a5c5d00...>. Recognition of works can similarly lag considerably. The Great Gatsby is an example, originally published in 1925, but becoming massively popular two decades later during World War II after it was included amongst paperback books distributed to US forces fighting overseas.
I'd agree with the notion that trying to stay current on most publishing genres (fiction or nonfiction books, cinema, music) will involve going through a lot of low-quality content. But that's not to say either that all that's new is dross or that all that's old is gold. Rather, time tends to be useful for allowing quality filters to work.
One of my own semi-structured practices has been BOTI: best of the interval. In any given period (day, week, month, year, decade), I'll keep a loosely-structured list of what I think the best works to come to my attention are. It's similar to the notion of 43 folders, or of a round-robin database. The idea is that your list length is finite, but that items can enter and leave it. With time you'll accumulate a list of works you consider to be of particularly good quality. And over the length of a year or decade, that tends to be fairly effective. Tastes and capacity to evaluate both change, and what you once considered excellent might no longer strike you that way. But it seems a good way to accommodate both new works and classics.
And yes, for the record, I do read quite a few old sources, some well-known and widely admired, others my own discoveries in what I find to be an underappreciated literature. I'm not at all against the practice. But understanding why and how that practice works both makes it more effective and reveals its limits.