Thursday, November 20, 2014

What Should I Read? Two Ideas to Help Guide Us

What Should I Read? Two Ideas to Help Guide Us

I read a lot. More and more, I’ve been thinking about what I read and asking myself if I’m making the most use of that time? Is there way to improve return on investment?
I’m not looking to optimize it to the tenth degree but it would be nice to come up with a simple map to help me know where I can find the most value per unit of work.
I think I’ve found it by combining two very simple ideas.

Understand Deeply

The first is getting back to basics, a concept from Five Elements of Effective Thinking:
Don’t face complex issues head-on; first understand simple ideas deeply. … Be brutally honest about what you know and don’t know. Then see what’s missing, identify the gaps, and fill them in. Let go of bias, prejudice, and preconceived notions. There are degrees to understanding (it’s not just a yes-or-no proposition) and you can always heighten yours. Rock-solid understanding is the foundation for success.
The most useful knowledge is a broad-based multidisciplinary education of all of the world’s great basic ideas. From there, you could specialize. And a broad understanding is build not on ideas that are recent but rather on ideas that have lasted, and will last.
“The more basic knowledge you have … the less new knowledge you have to get.” — Charlie Munger
The depth at which we master the basics is an indicator of how well we understand everything (after that). Deep work on simple ideas increases understanding and helps us build true virtuosity. If we are to have a chance of understanding complex problems, we need to master the basics.
The slightest wind blows over a house without a foundation.

The Lindy Effect

The second idea is Nassim Taleb’s concept of the Lindy Effect. Taleb was building on something developed by Benoit Mandelbrot. In Antifragile, he writes:
For the perishable, every additional day in its life translates into a shorter additional life expectancy. For the nonperishable, every additional day may imply a longer life expectancy. So the longer a technology lives, the longer it can be expected to live.
The nonperishable is anything that does not have organic or avoidable expiration dates.
While produce and humans have a mathematical life expectancy that decreases with each day, some things increase life expectancy with each day.
The perishable is typically an object, the nonperishable has an informational nature to it. A single car is perishable, but the automobile as a technology has survived about a century (and we will speculate should survive another one). Humans die, but their genes—a code—do not necessarily. The physical book is perishable—say, a specific copy of the Old Testament—but its contents are not, as they can be expressed into another physical book.
When I see a toddler walking down the street holding the hands of their grandparents, I can reasonable assert that the toddler will survive the elder.
With something nonperishable that is not the case.
We have two possibilities: either both are expected to have the same additional life expectancy (the case in which the probability distribution is called exponential), or the old is expected to have a longer expectancy than the young, in proportion to their relative age. In that situation, if the old is eighty and the young is ten, the elder is elected to live eight times as long as the younger one.
Life Expectancy
This is where Taleb asserts “So the longer a technology lives, the longer it can be expected to live.”
Not all things age with years.
If a book has been in print for forty years, I can expect it to be in print for another forty years. But, and that is the main difference, if it survives another decade, then it will be expected to be in print another fifty years.
This is where Taylor Pearson helped me put something together that I was just too stupid to do myself.
He connects reading to the Lindy effect. Whereas before my heuristic was simply older is better, now I know older is exponentially better. More importantly I have an idea why.
Pearson writes:
If you were to look at a typical person’s reading list, the vast majority of books would be crammed into the recent, low-value portion of the curve while many fewer books would occupy the much larger high-value, older section of the curve.
So your ROI on reading and understanding a concept from 500 years ago is highly likely to be exponentially greater in the long run than one presented only 5 years ago.
What I’m trying to get at is that the more fundamental or closer to the source that you move, the better the ROI in the long run.
So lets combine these ideas.
The most useful knowledge is a broad-based multidisciplinary education of the basics. These ideas are ones that have lasted, and thus will last, for a long time. And by last, I mean mathematical expectation; I know what will happen in general but not each individual case.
In the words of Charlie Munger, “take a simple idea and take it seriously.”
(Update: I’ve added clarity to this post, see the comments.)

What do you think?

Is older better? Are older ideas simpler? Is it exponentially better to read older books? I want to hear what you think. Leave a note in the comments.


  1. Shane,
    #1 – I love this blog! I just discovered it a few months ago and I’m constantly bookmarking your posts.
    #2 Connecting reading with the Lindy Effect is an interesting perspective I haven’t seen before, but perhaps an over-generalization since a lot of great books have been released in recent years
    #3 Which leads me to this question: I know you recommend a lot of books on this site, but which ones should be the foundation of this multi-disciplinary approach? For example, what are your Top 10?
    This is a good post, but I feel like I’m missing an actionable next step.
    – Dennis
  2. Shane Parrish says
    I wasn’t as eloquent in this article was I would have liked. Thinking on it a bit more, I think it’s a good idea to elaborate.
    So ..
    If something is still ‘in print’ today and it’s been around for a long time, we can assume there is a reason. The most likely reason is that there is something useful to the book. We can further assume that whatever is useful in the book will continue to be useful in the future.
    If it’s useful in the past, useful now, and likely useful in the future, there is an argument to be made that we’re probably dealing with something simple – the basics. Anything fragile gets weeded out by time. … so you’re at least dealing with robust ideas … This is something we should be reading to maximize ROI for reading.
    This isn’t perfect, of course. But it seems like a decent heuristic.
    Most of what’s new and best-selling today will expire rapidly. If you’re reading things that ‘expire’, you get trapped into a Red Queen situation; you’re running faster and faster but staying in the same place. Or in this case you’re reading more and more but not getting much smarter.
    You read more and more of the new stuff (e.g., best-sellers) but your knowledge doesn’t improve because you’re learning things with expiry dates … (narratives, studies based on small samples, or something that’s niche and specialized). When reading anything recent, it’s hard to distinguish if what you’re reading is fragile or not. And the base rate for fragility would be huge – almost everything printed today will prove to be fragile. (thought: this is why book recommendations really work because someone has now increased the odds you’re going to read something useful.)
    (The niche and specialized will improve your knowledge, for sure, but only within one particular domain, it won’t increase your broad based worldly wisdom. So these are useful but possibly not in the sense of maximizing knowledge accumulation. And you’d want to think about half-life of knowledge here too.)
    Basic knowledge and ideas, however, don’t expire, which is why reading something like Seneca gets you out of the red queen. You learn more, you learn simple ideas, and those ideas don’t change over time so your knowledge actually increases.
    Nassim’s been right all along. I was just slow to put it all together.
  3. J.D. Bird says
    If the nature of facts is information presented as objective reality, facts don’t have half-lives or expiration dates. They exist before we’re aware of them and after we forget them.
    To reference Arbesman’s first example, “Smoking has gone from Doctor recommended to deadly”. Smoking was bad before and after the Doctor’s advice. The fact never changed, only the advice.
  4. Radhika says
    I’m not sure if I agree with you, Shane. What you’re saying makes sense. However, introductions in the present will be infinitely easier to read and comprehend than introductions in the past. There are books specifically targeted to merely “introduce you to this fact,” and most of the ones you find will be published recently. They can serve to help you build a mental model, and they will refrain from the past-corrected mistakes.
    For example, if you read a book that is an Introduction to Biology before DNA strands were discovered, imagine how wrong the information you’d be getting is. It’d be better to look for recent books that are targeted for your purposes.
    (And through this whole thing, I’m avoiding the half-life of facts, because that would simply make everything futile.)
    I’d like to see if you have any rebuttal to what is being said here, or if I’m totally off of what is being said in the post.
    • Shane says
      I’m speaking as a general guide.
      Of course, you’re right that there are things that will be missed and some subjects lend them selves to meaningful discoveries at different rates.
      Older > Newer when it comes to books in general (think averages, mathematical expectations).
      But maybe that works better with some subjects/disciplines than others as well? Hrm…
      I still think the heuristic holds (e.g., you’ll find more signal and less noise by reading the older stuff.)
      • Elie says
        Maybe another way to look at it: The newer biology book is actually the old biology book plus new additions of knowledge that has been pretty well tested. So it is actually the old book. I don’t think it’s really in conflict with what you’re suggesting.
        For the commenting system, have you considered using Disqus?
        • Radhika says
          I don’t understand your comment about the new biology book being the old biology book. Care to expand upon it?
      • mla says
        I think your point holds: a book on biology pre DNA is unlikely to still be in print. If it *is* still in print, then it must have some special attribute that keeps it around and probably makes it worth reading.
        • Radhika says
          This is an interesting comment and may hold true. I haven’t read any biology recently, but this seems logical.
      • Radhika says
        What is “general,” though? I think this is a good starting point, but needs to be expanded upon.
  5. Shane,
    Thanks for the follow-up comment. Still, my big question remains: Where to begin? Other than Seneca, which books should we focus on FIRST to lay that basic foundation of knowledge?
    I’m not sure what the cut-off point is between classics and best-sellers, either. For example, “Thinking, Fast and Slow” was published fairly recently, but it’s also highly rated.
    • Shane says
      Dennis, I’ll put something together in the (hopefully near) future on the blog. I want to give this a bit more thought first.
    • Radhika says
      Highly-rated doesn’t necessarily mean quality. (Think: The Hunger Games.)
      Wouldn’t the Mental Models page serve as a good starting point?
      Shane, maybe if you added books to the page as recommended reading?
  6.  says
    I think it would be helpful to look at a specific example. In my case, I’m a software engineer, so let’s consider the field of knowledge that is programming.
    Programming knowledge moves quickly. New programming languages and techniques are being invented all the time. I could learn one of these new frameworks and learn it well. Now I’m good at the framework, but when the next one comes along I might struggle to pick it up and transfer my skills, because I’ve learned at a superficial level. An example of this might be learning the latest Javascript framework (a new one seems to pop up once a week and I’m barely exaggerating)
    Or I can take a different tact: I can learn a really old language, such as Lisp. Lisp isn’t a popular language, but knowledge of it has hung around for over 50 years now. Many other languages have come and gone, but Lisp is still talked about and still taught in some places. The magic of Lisp is that it’s simple but powerful. It’s power comes from the fact that it supplies building blocks that let me then create the complicated programs I need. However, to use these building blocks, because they are simple, means that I have to understand at a deep level how a program fits together and the aspects that make a good program. This gives me a solid understanding of programming itself.
    Now I return to that fancy new Javascript framework with my knowledge of Lisp. I may not initially understand the syntax of the program, but after a while I will be able to recognize the building blocks it is built of because of the knowledge I gained from Lisp. Now I can learn the Javascript framework more quickly and have a build better programs with it because from Lisp I learned the principles that make a good program. It takes much longer to learn these same principles if I start with another Javascript framework because I’m not starting from the base level. Starting from another Javascript framework is like trying to build a foundation while a house already sits on top of the spot you want to build the foundation. It’s possible to do, but it’s much easier to build a foundation on a spot that has no house yet.
    The basic idea of this post is to give a shortcut to finding out what the foundational knowledge is that you want to read. Once you have the solid foundation, it’s easier to build a house on top. What tends to be foundational knowledge? That which has stuck around.
    Now, does this mean you should read old books exclusively? No. What it does mean is that you should read books that contain old ideas. Often even new books are just reiterating good ideas that have been around for a long time. So the new book will present an idea that has existed, but will update the language to fit contemporary times. The Bible is a good example here. It certainly wasn’t originally written in English. Pride and Prejudice is another. It maybe set in a society that no longer exists, but it’s characters and themes would still fit in easily to today’s society.
    Let’s change domains. Consider American Football. Is it likely that any one football game is going to be so great that people are talking about it 50 years later? Not likely. However, is it likely that people are still going to be talking about American Football in 50 years? Yes, very. So if you want to be able to converse about football, it makes sense to learn the rules and how the game works. Then you can nominally pay attention to the games but still be able to converse intelligently on the subject. You studied the foundation first, so the benefit from the rest comes easier. Conversely, if you don’t understand the rules then no matter how many games you watch it’s difficult to speak intelligently on what occurs.
    This has been a long and winding explanation, but hopefully it grounds what Shane is saying above, whether you understand programming languages, religion, literature, or sports. Shane’s point is read those things that provide you foundational knowledge in whatever area it is you’re interested in (and foundational knowledge usually beings with simple ideas that are then built up into more complicated ones); a shortcut to determining what is foundational is to look at what has been around a long time, as it’s likely that idea is going to be around for a long time to come. But don’t apply how long something has been around to the physical item, such as the book. Instead, the time something has been around should be judged based on how long the idea contained within it has been around.
  7. Rod says
    I agree that classics are classics for good reason. Same applies to music, fashion, architecture, etc.
  8. Michael Han says
    It’s interesting how this resonates as a recontextualization of the old teaching. Here the teaching has been contextualized to a materialistic worldview, thanks to Taleb. The tangible has become the subject, and much easier for the moderns to grasp.
    The first point about the deep understanding is something that is much needed in our informational deluge, and it even sounds refreshing to hear it, however, the ancients would have found it more profoundly tautological than some of us.
    And the Lindy Effect seems one of those elements also implicit in major religions and philosophical systems. Among religions, for example, the factor of eternity is constant, and usually the ideas or words (logos in Hellenic sense) originating from the personal being have been considered having intrinsic value, simply because of the originator. Taleb’s assertion about the longevity of technology seems to loosely touch on the fact that it is the natural principles (aka informational) which allow that technology. Again, in a religious or even some philosophical contexts, those natural principles allowing the technology would have an ultimate originator. The contrast seems to be one of coherence. I’m not sure how far moderns could climb hanging onto inductive reasoning from a non-coherent, pluralistic view of the uni-verse.
    Your question of what is better seems to be coming from an ROI-driven, practical-minded context. Is it better in the eyes of those walking right now? Or is it better in the eyes of those who have closed their eyes? Whose context of understanding should we be assuming to understand what ideas? Yes, generally speaking, I have preference for older ideas and books, however, there are many old ideas which have survived but are inherently flawed and potentially dangerous, as they are errors of the soul rather than a diligent work of the mind. It’s encouraging to see an attempt to understand something with internally coherence and logical consistency.
  9. Gaurav says
    Hi Shane,
    I agree:
    1. Most New Books (Best sellers) contains more noise and perishable information.
    2. Old books which have survived are less likely to contain the noise.
    I agree with you basically but this heuristic can be modified some how to achieve better results. Ideas (Books) can survive for 2 reasons : Either they correspond to reality/knowledge (non perishable kind, example seneca) or they appeal to our primitive hardwired mind example: religious texts , concept of god. Unfortunately, lot of what has survived (old) belongs to second category also. We need some way to differentiate between the two category.
    Talebs recommended reading books (that you have mentioned in your blog) contains more books from the recent years than the books centuries/decade old and so would be mine.
    • Shane Parrish says
      Thank you for sharing these thoughts with me. I bet if you ask Nassim, he reads more old than new. Maybe he’ll chime in on the comments.
  10. Shane Parrish says
    Introductory textbooks might be an exception to this as they teach basic concepts that change slowly and have more staying power than new best-sellers. (i.e., you’re still getting basic knowledge in a new book.)
    There are other exceptions too. Can you think of any?
  11. Henry says
    This post took me back to two of your recent posts that I enjoyed. “The Half-Life of Facts,” and, “How to Predict Everything.” After reading the responses here I think most of the issues people have taken with this post are covered in the previous ones.
    I enjoy your thoughts enormously,
  12. Shane Goldman says
    Developing a reading list has become an obsession of mine lately. I have a growing list of books on my amazon and audible wish list. It is interesting to look at the satisfaction level I get after reading each book. Comparatively, I do find more comfort and rate the older books higher than the most recent; however, some of the best books I have read do include recent publications (Thinking, Fast and Slow andEmperor of All Maladies come to mind). The higher ratings for the oldies does correspond with the “wisdom” that comes from these books and the desire to delve further into the subject after reading.
    I look for recommendations at any source when developing my wish list and I have found two which I use extensively: for recent books and The Great Books for older classics.
  13.  says
    Building up from basic knowledge to a deeper understanding of the phenomenon being studied and its complexity is the way we progress in our journey to understand things. We could call it spiral learning model or whatever but the language of an old text may not always be the best to understand those basic ideas in today’s contexts – at least not for everyone. In fact, even though it may sound redundant but often i like to read multiple articulations of a same idea to understand it better – or improve my own articulation of it. So I won’t discard all/most new work as noise. Yes, we have to be selective so that’s a choice problem because we shouldn’t assume that we can understand everything deeply.
    Also, language is just a medium which we or any author uses to share the personal (often experiential/abstract) understanding one has of that particular phenomenon or concept. This means anyone else’s articulation would get limited by their language itself. So the actual experience of that phenomenon and/or exposure to the multiple articulations will help anyone develop a deeper understanding. Interesting point is that deeper your understanding of a subject is, smarter you become in terms of very quickly identifying the signal from all the noise around and good (doesn’t matter if they are new or old) ideas from crappy ones without much stronghold. So I think it’s just a function of how much you have read or where you are on your learning curve to understand the world. Trying to take a short-cut may not be a good strategy for everyone – though it may work for a lucky few.
  14.  says
    Another way to put it is, let’s not try to make a god out of Taleb :) He is definitely one of the top few thinkers of our times. I am an average guy and nowhere close on the IQ scale to either Taleb or the kids who were considered smart in their class or friend groups. Which is why I am such a plodder in my learning journey w.r.t. the smart people. Maybe that’s why I have such an opinion.
  15. Costa says
    My 2 cents,
    Since early this year, I’ve been trying two tools: “knowledge roadmap” and “compounded knowledge”.
    Extrapolating the same argument that old stuff is more signal than noise, I’ve been trying to draw a roadmap for some topics and attack those books that have succeed the test of time, let’s take the finance/economics topic:
    1776 – Smith – Wealth of Nations
    1867 – Marx – Capital
    1934 – Graham – Security Analysis
    1936 – Keynes – The General Theory of Employment, Interest and Money
    1938 – Williams – The Theory of Investment Value
    1942 – Schumpter – Capitalism, Socialism and Democracy
    1949 – Graham – The Intelligent Investor
    1952 – Markowitz – Efficient Frontier (Portfolio Management)
    1958 – Fisher – Common Stocks, Uncommon Profits
    1958 – Tobin – Capital Market Line
    1958 – Modigliani-Miller – Capital Structure
    1962 – Chandler – Strategy and Structure: Chapters in the History of the Industrial Enterprise
    1962 – Friedman – Capitalism and Freedom
    1964 – Sharpe – Beta / CAPM
    1965 – Fama – Efficient Market Hypothesis
    1973 – Black/Sholes – Option Pricing
    1977 – Chandler – The Visible Hand: The Managerial Revolution in American Business
    These books/articles usually have the basic ideas/building blocks needed to establish a solid mental model.
    And based on the concept that our time is a limited resource and we have a constraint regarding acquiring knowledge each day, I’ve been trying to sip a bit of knowledge from those guys who have already gave so much time thinking or such a vast experience. To this concept I refer to “compounded interest” of their time thinking about a topic or sharing their experience. These “compounded interest” is our “compounded knowledge”. Examples: memoirs like Churchill’s writings, Autobiographies, “History” books like Diamond’s “Guns, Germs and Steel” or “introduction books” like Russell’s “History of Western Philosophy“.
    I’ve came up with these ideas since I’ve been grasping for a while with the same question about improving ROI.
  16. Meatball says
    I think it depends on what kind of knowledge you are talking about. This principle of old being better than new, let’s assume it is correct for philosophy and history and human social patterns, probably does not hold true for hard science such as theoretical physics, neurobiology, or genetics. On the other hand, I think it’s pretty obvious that it’s useless to read ONLY recent history. Rather, reading a lot of history old and new would quickly reveal the strikingly familiar patterns that continue today, thanks to the relative stability and predictability of human nature and social psychology. But frankly I don’t think any of this is a groundbreaking new idea — which is a good thing if we were to believe that anything worth saying has already been said by someone a long time ago. As a good example I highly recommend Chuang Tzu, whose insight has been exceeded by very few in the subsequent 2500 years.
    A more fundamental question that bothers me is what imperishable knowledge means to a perishable life like ours. Why do we want it and what does it do for us? Why do we spend our finite life on acquiring knowledge and understanding of the infinite ways of the world/universe rather than on other pursuits? What is the point? (I’m not saying there is no point. I just wonder what the point is.)
  17. Vipul Vivek says
    1. So do we read any old book just because it is old? 2. All old books were new once. So do we read less new books now and later should we give less importance to these new books that would become old then? 3. All old books that are out of print don’t necessarily mean that they don’t have value. That way Enlightenment and Renaissance would never have happened. What if we’ve simply forgotten about or ignored some of the old books? 4. Basic ideas in old books may evolve over time and come to be presented in a better way in a new book. For instance, should we start learning calculus from Newton’s Principia Mathematica or pick up a new high school text?
  18. Praveen says
    Hello Shane!
    This is one newsletter that I literally wait with all enthusiasm to arrive in my mailbox. Great job really and thank you so much!
    I have a question here, and maybe slightly off topic, so please bear with me.
    Recently I gave up reading all works of fiction after giving it a bit of thought. With all the great works of non-fiction out there, the things that REALLY matter, I thought the ROI of reading fiction was probably the least!
    But having said that, there are works of fiction which have stood the test of time that you have explained above (I’m sure there are quite a few books of fiction out there that have survived 40 years or so as well).
    What do you think about reading fiction? Do you read them? Or do you have a proportion (1 fiction for every 10 non-fiction :) ). WOuld really love to hear your thoughts on this.
    Many Thanks,
    • Radhika says
      I’m not Shane, but a few months ago, I began to believe fiction was worthless, too. I now recently bought 3 works of fiction out of my 6-book-buying total. (Not a deliberate count.)
      Fiction is good for understanding people, the times, psychology, and perhaps history. Not all fiction accomplishes this, of course, but books of significance, aka: “classics” (although my definition is different than the mainstream definition) do uphold a viewpoint that non-fiction may overlook.
      For example, if you’re interested in the society of America is the 60’s, To Kill a Mockingbird represents the racism and the newfound obsession of mental health. The Virgin Suicides (only 20 years old, but still a classic) represents the American suburbia in the ~60’s. On The Road represents the Beat generation in the 60’s. These books encompass parts of society that non-fiction will never be able to fluidly represent.
      It is harder to sift through what is important in fiction (although that’s debatable, considering the PopSci/PopPsych dominating shelves), with a little work and an interest in people, societies, and eras, you can learn quite a bit. I therefore believe that you should pick up as much fiction as interests you, and as much non that interests you.
      Hope this changes your mind. Feel free to disagree.
      • Praveen says
        No, I fully agree. I cannot say for example I did not enjoy reading some of George Orwell’s works. In fact, I absolutely loved it. I think I’m just going to be much more selective here on. Many thanks for sharing your thoughts.
    • Shane Parrish says
      I was never a big fiction reader. However, that’s something I’ve changed my mind on.
      I think reading some fiction is a good idea.
      • Praveen says
        Now this is exactly why I think I arrived late on the scene (only discovered this blog a few weeks ago) and have a lot of catching up to do… I do agree on most of the points you have made in that article. Thanks.
  19. Excellent post. Another way to describe what you’re getting at is to understand “principles.” Timeless principles, that is. This is one of the reasons that Covey’s the 7 Habits has had such a profound, long-term, wide effect. It is based on 7 timeless principles that Covey uncovered in his research–he focused the old.
  20. Alejandro Dinsmore says
    You’ve been reading some Murakami books recently and it seems to me as an author he would support your ideas about reading older books that have survived the test of time.