Read in February 2026

Katherine Heiny. Single, Carefree, Mellow: Stories

I read the first half of this book while pulling an all-nighter in an airport lounge in Delhi. It was great for that purpose: The stories draw you in like a good gossip session. The connecting theme is infidelity and other romantically disastrous choices and situations. The characters in these stories and their values are mostly pretty mysterious to me—their choices seem to make them so much lonelier than they realize. But the stories are told well and I laughed out loud a few times.

Peter Smith, An Introduction To Formal Logic

I read this book with pleasure last year and enjoyed it even more the second time through. It’s an excellent philosophical introduction to logic in that it’s sensitive and alive to so many of the interesting philosophical issues raised by the technical material. Smith does a wonderful job of sticking in a disciplined way to what is appropriate in an introductory text, while also clearly indicating some of the philosophical choices and issues that await the reader beyond that threshold.

Paul Torday. Salmon Fishing in the Yemen

I read this book because it was said to be funny—won the Bollinger Everyman Wodehouse Prize for being funniest book of the year it was published (2007), in fact. It started out pretty funny, but the laughs became increasingly spaced out until halfway through I wasn’t sure if I cared enough to finish it. I finished the book anyway, laughed a few more times, and don’t entirely regret having read it.

The story is about an improbable attempt to bring salmon fishing to Yemen. It is told improbably by way of a series of selections from diary entries, letters, testimony from an official inquiry, memoirs, etc. from the central characters in the story. Just as the characters need to suspend disbelief in their project, so too does the reader need to suspend disbelief in the various literary techniques employed to tell the story. No one would ever, for example, talk so frankly about their romantic aspirations on the record to an official inquiry. But no matter. It’s all meant to be as silly as possible, and at that I guess it succeeds.

It seems they made a movie based on this book. From skimming the Wikipedia summary, I gather they modified the romantic aspect of the plot in the most predictably Hollywood way possible. I think I will skip it.

Shinzen Young. Natural Pain Relief: How to Soothe & Dissolve Physical Pain with Mindfulness

Moved here.

Jacqueline de Romilly and Monique Trédé. Petites leçons sur le grec ancien

Moved here.

Papers

In addition to books, I also read a number of papers in February. Most of my reading was around the subject of Tarski and truth, which I’ll write about separately later.

I also read George Boolos’s classic 1972 paper, “The Iterative Conception of Set”.

Sets are aggregates of objects considered extensionally. The “considered extensionally” just means that a given set is determined entirely by what it contains, no more or less. The set {2, 3} is the same as the set {3, 2} and the set {2, 2, 3} because all these sets contain exactly the same two elements; order or duplication doesn’t matter. We might specify the same set in different ways, e.g., {2, 4} and {x : x is an even number under 5}. No matter the different ways of specifying the elements here: these are the same set nonetheless. Notice the “x is an even number under 5” bit. It’s very convenient to build sets using predicates in this way. Now, wrap up these two points—extensionality and “comprehension” (the use of predicates to specify a set) in axioms—and you have naive set theory.

The problem, as mathematicians and logicians realized pretty early in the development of set theory, is that naive set theory hits paradoxes pretty quickly. Here’s a famous one: If we can use any predicate we like to specify a set, then why not the predicate ‘is not self-membered’ which picks out sets that are not members of themselves? Now, if {x : x is not self-membered} does not contain itself then the predicate tells us it must contain itself, but—wait!—if it does contain itself then the predicate tells us that it must not. Whoops!

There are various ways of blocking such paradoxes, but a very popular one is the iterative conception of set. On this approach, we ‘build’ sets in stages using axioms carefully chosen to block paradoxical sets. At each stage, we can use sets built up in earlier stages as material for subsequent stages. (And yes, there is so much work to do here spelling out the dependence relation and the quasi-temporal language of building.) The key move is a modified version of comprehension that allows predicates to range only over sets we’ve already constructed in earlier stages. This restriction blocks the paradox mentioned above, since the predicate used to build the set can no longer range over the set itself during its construction (it isn’t available at that stage to be ranged over).

Now, because I read everything in the wrong order, and because I don’t know this area very well (to say the least), I was originally confused by the point of Boolos’ paper. What was the celebrated Harvard logician doing telling his colleagues about something as commonplace as the iterative conception of the set as though it would be news to them? As he himself notes (footnote 3), the iterative conception goes back to Zermelo and Russell, working long before Boolos’s 1971 paper. (Though see Potter’s Set Theory and Its Philosophy (especially pages 36 and 37 for scepticism about how early it really appears. He thinks we don’t really get a clear description in print until a 1947 paper by Gödel.) I think my initial impression was confused because as Boolos points out in the same footnote, “philosophers, in the main, seem to be unaware of it”. There’s been a lot of philosophical water under the bridge since 1971, and Boolos’s paper played an important role in getting it flowing. So now the iterative conception of the set and its philosophical motivation is familiar to philosophers working in this area. But I guess that wasn’t the case when the paper was originally published.¹