First: there's a hell of a lot of wisdom about reading and how to approach it in Mortimer Adler's <i>How To Read a Book</i>. That's a perennial favourite of both HN and myself. In particular it addresses the question of <i>how to read different types of books</i> and <i>for different purposes</i>. In detail. In particular, Adler repeatedly stresses the notion of <i>reading as an inquiry</i>, in the sense that <i>you are asking questions of the book and the author</i>. If those questions are rewarded, all the better. If they're not ... you might be better of spending time elsewhere.<p><<a href="https://archive.org/details/howtoreadbook0000adle_y9v4" rel="nofollow">https://archive.org/details/howtoreadbook0000adle_y9v4</a>><p>I've referenced it nearly 30 times on HN: <<a href="https://hn.algolia.com/?dateRange=all&page=0&prefix=false&query=by%3Adredmorbius%20%22how%20to%20read%20a%20book%22&sort=byPopularity&type=comment" rel="nofollow">https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...</a>><p>Disclaimer: I'm only about halfway through it myself, though the book <i>is</i> on my e-reader ;-)<p>On filters, I disagree with Burkeman, it <i>is</i> filter failure, and Burkeman is himself failing to define what sorts of filters are needed.<p>In the context of the "haystack-sized piles of needles", it's useful to consider the question <i>how many needles do I need?</i>, and to simply select a sufficient (and not excessive) number. This <i>does</i> presume that the stacks are of equivalent quality, which can pose a slight challenge, though factors such as reputation and some occasional broad sampling usually address this.<p>In particular, though, <i>in an environment with a surfeit of signals and records</i>, what is needed are <i>cheap, fast, no-regret rejection tools</i>. I've noted this a few times in earlier comments, e.g.: <<a href="https://news.ycombinator.com/item?id=22208255" rel="nofollow">https://news.ycombinator.com/item?id=22208255</a>> and <<a href="https://news.ycombinator.com/item?id=31254795" rel="nofollow">https://news.ycombinator.com/item?id=31254795</a>>.<p>In the physical world, our senses and mental capabilities provide us with this: we're tuned to a limited set of stimuli (sight, sound, smell, taste, various touch, proprioception, and inner-state senses such as hunger, thirst, and need to void bowels or bladder). There's a heck of a lot of other signals the Universe provides, <i>we are utterly blind to them</i> for the most part, and our lack of capacity to sense these <i>largely</i> doesn't bother us. Even within our extant sense channels, we can only detect so much: visible light, but not radio or x-rays, the audible spectrum but not infra- or ultra-sonic frequencies, some smells with great sensitivity (petrichor, mercaptin) others not at all (water, carbon monoxide), etc.<p>In the world of data one of the most effective <i>and unbiased</i> rejection methods is random sampling. We usually think of this in terms of <i>data selected</i>, but it could as readily be expressed as <i>data rejected</i>. The reason statisticians sample randomly <i>is because measurement is expensive</i>, and quite often a highly accurate impression of a phenomenon can be gathered from only a few tens or hundreds of individual examples <i>if those themselves are selected without bias or are otherwise representative</i>.<p>For texts and readings, much depends on <i>why</i> you're reading something. If it's simply for entertainment, then the requisite question is "am I being entertained". If it's for <i>information</i>, then the question broadens, but I'd suggest:<p>- Is this actually informative? Is it telling me something I don't already know (rather than reinforcing biases or existing beliefs)? Is it accurate? Is the information <i>durable</i>?<p>- Is this <i>useful</i>? Can I use the information, and will it change a future action or decision of mine? Does it put the <i>past</i> in a more comprehensible light?<p>- If a skill, <i>is the skill taught useful and durable?</i> I and many others have often noted that the base set of Unix / Linux shell tools and editors provides a highly <i>useful</i> and <i>durable</i> skillset, one I first began acquiring 30 to 40 years ago, and still use daily. I've seen many <i>other</i> technical skillsets, including expensive training or books (some at an employers' cost, some at my own), come and go over that same time.<p>- How does this fit with an existing worldview or structure? I've begun specifically cultivating a few of my own ontologies and sense-making structures, and find that these provide useful lenses for assimilating and testing new information. Occasionally I'll find other authors have anticipated my own notions, and the force with which <i>that</i> realisation hits is profound. It's also a strong validation that I'm likely on a useful path, or at least one that others had previously found worth pursuing.<p>All of these boil down to "Is this worth my time?"<p>Note that most news, episodic broadcast media, and online content fares <i>abysmally</i> on the "durability" criterion. I don't ignore these <i>entirely</i> (and struggle immensely with online content), but am <i>highly</i> cognisant that these tend to be the equivalents of fast food. As an occasional treat or spice ... not necessarily bad. But if these are your primary fare ... reconsider your priorities, environment, and patterns.<p>A practice I'm finding useful is to at least periodically consider what has be "BOTI" --- best of the interval. On some basis (weekly, monthly, quarterly), I'll try to cast back over what I've read, heard, or viewed, and consider what's been the most insightful or useful. For the past decade, that title's gone to an <i>Aeon</i> piece by Michael Schulson, "If You Can't Choose Wisely, Choose Randomly", on the value of sortition: <<a href="https://aeon.co/essays/if-you-can-t-choose-wisely-choose-randomly" rel="nofollow">https://aeon.co/essays/if-you-can-t-choose-wisely-choose-ran...</a>>. And yes, that relates to my comments above on random sampling / exclusion rather neatly.<p>As for the first time I realised that there <i>was</i> in fact far more information than I could reasonably access, it was at a library. On entering uni, I remember my first visit to the main campus library, <i>literally</i> a tower of books, holding several million volumes (and part of an even larger collection across other libraries both on campus and at other campuses). What I did instead was to treat the collection <i>as a resource</i>, something to ask questions of, with a few (small) sections explored in detail, many others skipped entirely, and some sampled from selectively. A particular memory was of seeing the film <i>The Last Emperor</i>, a biography of Puyi, the last emperor of China, and largely a Japanese puppet. One scene of the film featured a <i>Time Magazine</i> photographer, and the thought occurred that there was probably a story from that time. I tracked this down and read the story, published in the 1930s, from a copy dating to that time. The experience of "oh, I can track that down" has stuck with me. And yes, losing on-campus access to academic stacks was an absolutely wrenching experience for me, moderated now through the Internet Archive, Project Gutenberg, Library Genesis, ZLibrary, and other resources.<p>And yes, if something's not worth reading, it's not worth reading, as noted in an exchange I had on HN itself over the past day or so.