If you are a nerd you have been waiting for George R. R. Martin to complete his A Song of Ice and Fire series. But if you are a next level nerd, what you’ve been waiting for is for Donald Knuth to finish The Art of Computer Programming.
If you’ve never heard of Knuth, The New York Times has a nice profile up, The Yoda of Silicon Valley-Donald Knuth, master of algorithms, reflects on 50 years of his opus-in-progress, “The Art of Computer Programming”. When you first encounter Knuth and his life you get a sense of what it means to live and breath the life of the mind (Paul Erdos seems in the same category).
But this got me to thinking: if human civilization collapses would The Art of Computer Programming make it through to the successor societies? Enough people have memorized large sections of the Bible and the Koran, and various other religious and mythic works, that we’d be able to reconstruct them (and they would be passed down orally in rough form). It is unlikely that all the books would be destroyed. Similarly, great works of literature such as Shakespeare are widely read and internalized by the public.
This is not the case for a lot of detailed technical knowledge. From what I know the paper we use today is relatively perishable. If our civilization collapsed, it isn’t assured that low volume publications wouldn’t simply disappear as the books degrade beyond recognition without being copied (and without our modern technology digital storage will disappear).
Though I do think religious and literary works have value, to be frank it seems that any sufficiently advanced civilization has to converge upon similar narratives to encapsulate the sort of normative framework around which a society can function. For example, cannibalizing other human beings “because you can” always seems to be understood to be in the “bad” category. Some level of generosity toward the downtrodden is usually classed in the “good” category. I don’t think this is arbitrary, I think it’s an interaction between social complexity beyond the tribal scale, and our cognitive architecture which has first-order “natural tools” to deal with clan-based dynamics, but not supra-clan systems.
In contrast, a lot of technical knowledge, what we bracket into “natural science”, is quite counter-intuitive, and has appeared in one single civilization, that of early modern Europe. I’m particularly thinking of the fruitful synthesis of mathematical formalism and empirical testing which has characterized natural philosophy since Galileo. The historical record is clear that proto-scientific thinking in various forms emerges in many societies, with disparate threads in the same culture even (e.g., empiricism and mathematical formalism were present, but not fused, in the Classical world). But the combination in early modern Europe that kick-started modernity as we know it is rare, and takes a fortuitous combination of circumstances to allow for its flowering.
I hope that the Long Now Foundation has figured out a way to inscribe various technical texts on long-lasting tablets (perhaps stone?) and store them somewhere!
Can’t we just bury a bunch of paper copies in Egypt?
This might sound strange, but if our civilization collapses, I actually think computers in general will survive, and (along with robotics) become even more important. Both because a collapse won’t affect everyone at the same time (barring something like an asteroid impact), and because that’s what happened with collapses in the past. The loss of the western Roman Empire and its excellent capabilities in organizing and managing great numbers of people helped promote the adoption of labor-saving innovation, particularly the increased use of water mills and new use of wind mills. Same thing if international trade disintegrated, and people suddenly had to maximize more local and regional resources more effectively on a smaller labor supply.
The ‘reboot’ pieces I’ve seen at Long Now tend to severely misunderstand the issue. Presentist, taking for granted technical achievements, difficulty of bootstrapping, virtue-signalling about climate change or racism or sexism (need to make sure white male books are not overrepresented), etc. It’s a little pathetic. If you don’t believe me check it out yourself: https://medium.com/the-long-now-foundation/manual-for-civilization/home
The most important books or knowledge to preserve for a post-apocalyptic reboot attempt are the likes of Foxfire books, Gingery Books, Marks’ Standard Handbook for Mechanical Engineers, Handbook of inorganic chemistry, etc.
If these San Fran delusional liberals gave their post-apocalyptic civ “Handmaid’s Tale” while mine had actually useful handson ‘how-to’ books, my post-apocalyptic civ would annihilate them upon contact. Though frankly they’d likely starve to death first, unable to do anything productive other than argue and virtue signal.
In the question of “What to save, for the apocalypse?”, I’d be perhaps come down on the preservation of texts that represent irreplaceable primary evidence.
We can probably re-reason coding or moral philosophy, even if they’re counter-intuitive. Eventually. May not even take that long – core insights of moral philosophy seem to turn up pretty quickly and independently after writing becomes widespread, and it seems most likely that Don Knuth’s is not a low probability mind, given the constraints – population in the world in countries at the frontier of computer science education, % of that population with access to that education, levels of access to actual syntactic devices – and how quickly he shows up in the information revolution.
Once you have the conditions for economic growth to that stage, it seems like they should be the relatively low hanging fruit, prima facie probable to re-emerge pretty quickly.
What it would seem is irreplaceable is rich ethnography and history (including some economic history); that which no one could ever reason and experiment their way towards understanding at all. We might be able to reason our way to the same insights as Marcus Aurelius’ Meditations, but we can’t re-reason our way to Phytheas of Massalia’s On The Ocean.
On the side of what technical knowledge to preserve though, also this touches on the questions of finding exceptions to “Steam Engine Time”; those cases where we can identify that we have the technical framework for the idea or innovation for long stretches of time prior to which it’s actually invented.
That is, much of our technical knowledge is counter-intuitive, however we understand well the processes by which people came to reason along those lines, and we know that independently people come across the same ideas after the right precursors are in place. Probably, you have mass printing due to the Gutenberg press, you pretty soon after get a phase change from proto-scientific thinking to science. You have access to the fossil record, pretty soon after theory of evolution by natural selection. Explicable.
But then you have those ideas, which don’t crop up until hundreds or more years after they are technologically possible and after their immediate precursors, for no good reason that we can understand. Gene Wolfe gave the example of Epaminondas’ knotched shield – https://medium.com/the-polymath-project/gene-wolfe-a-science-fiction-legend-on-the-future-altering-technologies-we-forgot-to-invent-a3103572a352. Intuitive and straightforward to us conceptually, but apparently no one actually thought of it for ages.
Those seem like technical ideas you’d want to preferentially find and preserve. If you could reliably believe you could identify them!
Well, as a nerd you have read *A Canticle for Leibowitz*, right?
.. but then there’s also Larry Niven’s (iirc) depressing idea that you only get one shot at technological civilization per planet, because that first shot burned up all the easily accessible energy dense fossil fuel.
I don’t think that one would really need anything like a societal collapse in order for knowledge to become ‘lost’. Apart from stuff like how to sail a square rigger efficiently.
Say the ‘mathematical formalism’ cited above. If one were to require a complete explanation of ‘mathematical formalism’, as in, per an analogy, not just how to do it, as in drive the car, but why it’s done the way it’s done, how the car works, that isn’t something that might get lost, it’s mostly lost right now. Very few physics professors could explain why physics works the way it presently does down to the nuts and bolts, as in what does one need to assume to believe mathematical formalism would actually work on the physical world, even if they could explain how it works. Physics professors are a pretty smart bunch too.
I also think that once one were to get started, one would see lots and lots of ‘lost’ knowledge, and I’d say that quite a bit is the result of what used to be called a ‘pious fraud’. One of the reasons I think that ‘pious frauds’ have gone out of style is because of the ‘pious fraud’ that one needs to be explicitly religious to create a pious fraud.
Please see https://en.wikipedia.org/wiki/Arctic_World_Archive
Not far away from the famous Svalbard seed fault, I think.
@AlanL: Yes, Canticle for Leibowitz is a great book. However, I think that electric generators would be reinvented much sooner, because there would be many specimens available as technofossils. Note that some types of electric motors work also as generators if you start rotating them from the shaft.
“Enough people have memorized large sections of the Bible and the Koran, and various other religious and mythic works, that we’d be able to reconstruct them (and they would be passed down orally in rough form).”
Lots of people have the contents of TAoCP memorized as well. I mean, not like they could sit down and recreate the exact text, but there are enough people who need the relevant knowledge on a regular enough basis that they could probably recreate it if necessary. Plus, the tech-sector has made recreating algorithms from memory a semi-ritual in job interviews.
The real issue, I think, is how long it would take for that knowledge to be useful again. People might be able to recreate TAoCP from memory to pass to future generations, but they might not bother, since it isn’t really useful until society has both the ability and need to create automated computing machines. If that time seems to be several generations off, I could see the knowledge dying out simply because people don’t see the point of putting the effort in to preserve it. Preserving stuff like basic medical, engineering and agricultural knowlege would probably seem like a better use of time.
@simplicio You could preserve some of the algorithms in verse (cf. Pāṇini and his grammar), and then employ them in (quasi-)religious rituals. See also the history sections in Volume 4A (Combinatorial Algorithms, Part 1) of TAoCP, documenting existence of many algorithms long before the advent of mechanical computation. An earlier fascicle containing all such sections together (“History of combinatorial generation”) can be found at Knuth’s home page at: https://www-cs-faculty.stanford.edu/~knuth/fasc4b.ps.gz
or via this page: http://www.antiquark.com/2004/12/knuth-history-of-combinatorial.html
.. but then there’s also Larry Niven’s (iirc) depressing idea that you only get one shot at technological civilization per planet, because that first shot burned up all the easily accessible energy dense fossil fuel.
If the world warms enough there will be one area where it will be relatively easy to find untapped surface mineral deposits and fossil fuels for a second industrial revolution – Antarctica. Considering it’s roughly the size of Europe, in an “Eocene hothouse” style environment that would be more than enough to bootstrap civilization again.
I’d be shocked if this hasn’t been used as a post-apocalyptic setting actually.