On Reading Old Books Badly

There are two ways to misread an old text and they fail in opposite directions. The first is to read it as though it were written yesterday, applying present categories to a document that does not contain them. The second is to read it as though it were entirely alien, a curiosity with nothing to say, so exotic that no claim in it needs taking seriously. Both are errors. The second is more common among careful readers, because it feels like rigour.

The first error is easy to spot and everyone can name examples. A modern political vocabulary imported into a medieval dispute. A psychological framework applied to a text written before that framework existed. The result reads well and describes nothing that was actually happening.

The second error is subtler because it wears the costume of scholarly caution. It treats the historical distance as total: they thought differently, their categories were not ours, we cannot really know what they meant. Applied consistently this makes every text unreadable, so it is never applied consistently. It is applied selectively, to claims the reader would rather not engage with, and the historical distance becomes a way of declining the argument without addressing it.

The test is whether the caution is symmetrical. If you can extract a usable insight from a text when it agrees with you, and reach for incommensurability when it does not, the incommensurability is doing motivated work.

What makes old texts hard is real, though, and worth stating properly. The problem is not that people thought differently. It is that words moved. A term that appears in a fourteenth-century text and in a modern one is often the same word carrying a different load, and the continuity of the spelling hides the discontinuity of the sense. Reading such a text without checking the load is how you end up confidently discussing a position nobody held.

The corresponding discipline is unglamorous and there is no shortcut. Establish what the key terms meant to the people using them, which usually means reading around the text rather than only in it. Find out what argument it was answering, because almost every text is a move in a dispute and reads as nonsense without the other side. Notice what it assumes without stating, since the unstated assumptions are where the distance actually lives.

Do that work and the strangeness usually resolves into something recognisable. Not identical to us, and not alien either. A person with different premises reasoning carefully towards conclusions that follow from those premises. This is the most useful thing an old book can show you, and it is the thing both misreadings prevent: that intelligent people, arguing well, arrive somewhere very different if the starting assumptions differ. Which raises an obvious question about your own.

There is a further hazard specific to the material this archive spends time in. Esoteric and religious texts were frequently written to be opaque, sometimes for protection, sometimes to restrict access, sometimes because the author was working out an idea rather than presenting one. Opacity in these texts is not always depth. Some of it is deliberate obstruction and some of it is confusion, and the reader who treats every obscurity as a hidden teaching will find hidden teachings everywhere, including in the parts that are simply muddled.

The corrective is to hold both readings live: this passage may be encoding something, and it may be badly written. Only the surrounding evidence decides which, and sometimes the surrounding evidence does not decide.

The aim is not to be generous to old books, nor severe with them. It is to read them at the right resolution: close enough to hear the argument, far enough to see what it was arguing against.