Red Plenty, by Francis Spufford

A fictional non-fiction book about linear programming and the Soviet Union that might have been.

A couple of friends have really strongly recommended this book, and 4/5ths of Spufford's Golden Hill was one of my favourite books ever, so I was excited to give it a go. I didn't quite like it, overall, and I'm not sure I can explain why.

I only realised (remembered?) at the end, when reading the footnotes, how non-fictional the book was -- for most of it I was imagining that it was basically fictionalisation of real events, but from the footnotes it's clear that a lot of the details are true to life. In principle I can see how narrativising real-life events could make them more memorable and sticky than regular non-fiction is, but in this case I think I just failed to realise I was reading facts.

To the extent I have a critique of the book as literature, it's that there are too many disconnected sections/segments, and none of the narrative arcs are followed with enough consistency to make the tale satisfying. Which makes more sense when you realise it's non-fictional fiction!, and had to follow the limited documents available about the real-life characters, and be hemmed in my reality's unfortunate tendency not to be neat and satisfying.

Spufford is still an astoundingly good writer. One of my favourite sections is a description of lung cancer from the point of view of the cell -- I'll see if I can find it extracted online somewhere, it was just so good, and just so Spufford. It's such a throwaway part of the book, and the fact that he does it so perfectly and seemingly-effortlessly just makes you wonder, yet again, how you could ever be a writer in a world where Francis Spufford exists.

In some way I had the same feeling about this book as I did about Susanna Clarke's Piranesi -- just, like, why is this extraordinary writer spending their extraordinary talent on this* (*thing that I personally don't connect with, dammit).

Here's a blogpost about the book:

Attention conservation notice:  Over 7800 words about optimal planning for a socialist economy and its intersection with computational complexity theory.

Here's Spufford's response, which includes the great line:

I wish the essay had existed before I wrote the book. It would have saved me months if not years of clumsy attempts to think through the underlying intellectual issue: whether, in any possible world, and not just under the hampering constraints of the Soviet environment, anything resembling the Kantorovich scheme for optimisation through prices could power a planned cornucopia

My friend the literary agent says that "books are how civilisation talks to itself", and I think this is true, but it's such an inefficient conversation! I'd never thought about this particular angle on it before -- that if you write a book other people will read and review it, and give you useful inputs that could have improved the book, but which you couldn't plausibly have got them to spend the time giving you (or known who to ask about it, etc) until the book came out. Maybe we would benefit from a kind of arXiv/pre-print system for books? Or maybe books are not long for this world, because they induce too much lag into the system, I don't know.

I tend to think the topic at the heart of Red Plenty -- whether there's a more centrally-planned but still price-based alternative to market mechanisms for allocating resources -- is newly relevant now, more so than most people realise. Amazon kinda can calculate shadow-prices and co-ordinate sellers and buyers in ways that were never possible before? Idk, this is one of those topics I know so little about I should probably try not to overstate my opinions on.