Studio Linda Liukas
Letter

No. 39 — Summer books 🔖 Read your way through 🏫 Library.json

PageRank for books

My name is Linda. I write a bi-weekly newsletter about computer science, childhood and culture - and there are 9 599 of you listening. If you enjoy this issue, please share it with anyone you think may find it useful.

Book discovery is a problem a bookstore owner and a computer scientist have very different approaches to. Caring about both, two thoughts:

On bookstores. Last week McNally Jackson posted a summer reading matrix with two axes: fast/slow and slim/epic. I rarely find future reading based on an algorithm, but from lists like these. And I like the categories better than the traditional genres of historical fiction, nonfiction, biographies, or the like. The matrix made me think of this quote from Teju Cole:

"I also find the stern distinction between fiction and nonfiction odd. It’s not at all a natural way of splitting up narrated experience, just as we don’t go around the museum looking for fictional or nonfictional paintings. Painters know that everything is a combination of what’s observed, what’s imagined, what’s overheard, and what’s been done before. Is Monet a nonfiction painter and Ingres a fiction painter? It’s the least illuminating thing we could ask about their works." - Teju Cole In Conversation With Aleksandar Hemon

I like how a reader’s taste or a publisher list form clusters of books that speak to one another. Especially useful is when someone finds a specific yet permissive theme: easy reads, with literary flourishes or books with giant, possibly magical libraries. And then there are publishers like 50 Watts, whose entire catalog I could buy.

Then, a thought on computer science.

Most book recommendation algorithms are plain boring: they know how to suggest things you already know you like. The recommendations lack nuance as they are often based on ratings of other readers - and if anything, a book runs away from a one-dimensional star ratings. (I’ve all but stopped rating books). However, books contain a lot of data and that’s what computers love.

What if there was PageRank for books, so instead of sales numbers or reader ratings the algorithm would look for acknowledgements and endorsements from agents, editors or other authors and surface unlikely connections? Or a k-means clustering algorithm for technology books none of my friends have read and that are over 50 years old? What if what Stephen Wolfram calls post-linguistic emergent concepts was let loose among a corpus of books and computers could help discover patterns on connections, turns of phrases or literary neighbours from books we humans can’t grasp due to our limited input speed?

I wish book discovery found a way to combine both human curation and the surprise of the unseen a computer can offer.

Meanwhile, of the 52 books I’ve read this year, I decided to veer more on the human curation side: inspired by the article on reading lists of pairs of books I offer eight recommendations, or four pairs.

Portraits: John Berger on Artists by John Berger | Cloud Cuckoo Land by Anthony Doerr
Portraits was maybe my favorite book of the year so far. It’s a collection of essays on art history that are both subjective and sharp. It shares the same quality of timelessness and power of a story as Cloud Cuckoo Land does.

A Bigger Message: Conversations with David Hockney by Martin Gayford and David Hockney | Wonderworks: The 25 Most Powerful Inventions in the History of Literature by Angus Fletcher
Both books felt like discussions (the other more formally than the other) and included a wink in the eye kind of curiosity.

To Paradise by Hanya Yanagihara | Ways of Being: Beyond Human Intelligence by James Bridle
Bridle’s book on nature and technology was love at first read. It’s sincere, searching and forgoes easy answers. Yanagihara was the kind of sweeping (literary/science) fiction that pairs beautifully with it.

Sea of Tranquility by Emily St. John Mandel | When We Cease to Understand the World by Benjamín Labatut
Both books have a quality of dancing, light, quick, all over the lines in terms of genres.

Books read in 2022

(Some of the linked books lead to Bookshop.org and I earn a small commission each time someone uses the link to purchase a book.)

January

February

March

April

May

June

July (so far)

Linked List

  • Read your way through a city is a New York Times (paywalled) series I’ve been loving. In it, writers give recommendations on books to advance read, bookstores to visit, literary pilgrimages and places to read in a city Leila Slimani on Paris and team Lars Kepler on Stockholm.

  • Tom Critchlow has outlined an interesting proposal called library.json. I, too, fell in the category of people who used to think book discovery ought to be a product someone "should build”. But maybe it’s a protocol. I’ve also been reading through some old RFCs and admiring the way giant and resistant things were built through protocols, clear communication and outlining ideas (related, Specifying Spring ‘83), not fancy UX.

  • Everyone has been gushing about Book Exploder and of course I’m eagerly awaiting for the podcast to launch in August, Susan Orlean and the process of breaking down some iconic books.

Classroom

I’m hoping to surface and share stories from all of you and I’d love to see your creations! Here are a few teachers using Ruby in creative, fun and inspiring ways.

Stories like these.. When a book inspires creation, like a little boy making a game.

Go Iowa! Can’t wait to see what else is in the teacher kit.

moi.e.class
A post shared by モイ・e・クラス (@moi.e.class)

These hardware drawings somehow always make me happy.

Thanks for reading Hello Ruby! Subscribe for free to receive new posts and support my work.