August 19, 2022

When incremental progress doesn't work

Welcome to my experiment in keeping up with writing and thinking and getting better at both.

Thinking

The first draft of a project (or a chapter, etc) is always hardest for me. People say things like:

• "just write 500 words a day, and in five months you'll have a book!"

• "write as you research, and you can put all the pieces together later!"

To some extent, this is great advice. As a nonlinear thinker (thanks, ADHD), switching from a beginning-to-end-draft-in-Word model to a write-chaos-scraps-in-Scrivener model was the key to me writing anything longer than 10 pages. But it relies on the assumption that writing progress can be incremental. That those 500 words every day will somehow slowly add up to a 50,000-word draft in a neat linear graph. Even if that draft is bad, it's still a complete draft, right?

That's where I think the model breaks down. Writing, for me, isn't purely incremental. Every long piece I've written has gone through a phase where that growing pile of words is nothing but a tangled mess, like a conspiracy theory map of string and pins. It has no argument, no through-line, no structure of any kind.

Incrementalist advice for building a house would say, "just cut and nail one piece of wood every day!" Six months later, you obviously won't have a house. You won't even have a shit draft/zero draft/minimum-viable-product house. You'll have a big pile of wood with some nails stuck in it. I'm exaggerating, but you get the idea.

I haven't figured out a way to skip over the chaos. I complain about it every time. People have offered to read drafts for me at this stage, which is very kind and useless. I don't think they fully understand that asking for feedback at that point would be like inviting someone to come tour your hazardous pile of plywood and nails, saying "there's a house in here, help me figure out what it looks like."

For shorter pieces, I can get away with outlining first, giving me a blueprint for those little scraps of writing, so they do come together into a coherent first draft. But for longer pieces, I can't outline it because I don't yet know I want to say. For longer pieces, I have to write my way into the thinking, especially when it's based on research I haven't finished. Maybe someday I'll figure out how to make incrementalism work for me to write books. But for now, I'll settle for using this newsletter to help shortcut some of the thinking aloud.

Reading

Privacy on the Line I read the 2010 edition of this, originally published in 1997. It's written by a researcher and a policy professor, both longtime privacy advocates, and explains basic privacy-related technology and policy concepts from the ground up, topics like digital signatures and wiretapping and the jurisprudence of privacy rights. They argue for stronger privacy protections and guaranteed access to related technology like encryption.

A lot changed in those ten years between editions, and even more has changed since 2010! It's a big challenge for me to figure out what to do with books like this, analytical books written between the mid-90s and the early aughts. They're a little bit too new and too analytical to be ordinary primary sources, but they're also very firmly rooted in a different technopolitical era than our own, which makes them complicated to engage with as secondary texts. This is the curse of working on recent history, I suppose, where the contours of the subject are much too jagged to be called historiography.

For example, the authors (in 2010) took great pains to seriously consider the national security risks often cited as justification for encryption regulations and government surveillance, even though they disagreed with them. They argued that the risks of terrorism--specifically Islamist terrorism, because remember, this was the early 2000s--were real but not worth the cost of eroding privacy protections. From my vantage point today, it's clear that they were still rooted in the post-9/11 us-vs-them model of security, where "external enemies" were the main threat, even as they took pains to try to distance themselves from 9/11 fearmongering. A book like this today would probably advocate for a different model of security entirely, one that considers security in an interconnected world more holistically. (Landau herself has written more recent articles in this direction).

Fascinating stuff. But what does this tell me about how people thought about cryptography in the 1990s? That's the hard part.

Tweet of the week: A thread on what it means to lose technology.

Doing

This week I started learning about lead paint 🫠 It is difficult to get a handle on the actual risk (gasp, the internet is bad with uncertainty?). On one hand, lead poisoning is a real risk. But on the other hand, it doesn't sound like you need to move out of your house immediately like you would for, say, a carbon monoxide leak. But also, lead is bad? I think the risk management part of my brain is still overloaded from two and a half years of poorly-managed pandemic.