On Knowledge Codification

What is knowledge codification? When I first heard this term, I thought of simply documenting knowledge. The weight of its meaning can be measured in memory, be it in bits or glyphs. The formal definition is slightly different; knowledge codification is defined by who will consume the knowledge rather than the one producing it. If a diary is written, it's not codified until it is made "easily accessible and manipulable" by users.

If a piece of documented knowledge was incomprehensible and jargon-filled, it wouldn't be codified. I also think that what if the original piece of work wasn't comprehensible but made comprehensible by some third party? Then, the knowledge is codified? What tools help with this, and how can we classify them? Next, there are varying degrees of comprehensibility; it it at one point, and at which point, does the producer of knowledge consider their work codified? How large of a group should the work be accessible to? At what point does comprehensibility depend on the consumer's ability to decode the knowledge produced? These questions are also the basis of collective sense-making. Moreover, if intelligibility is a barrier to collaboration, how do we dismantle it?

I think I was first really acquainted with the importance and idea of knowledge codification when I read a paper by Juhász, Sakabe, and Weinstein titled "Codification, Technology Absorption, and the Globalization of the Industrial Revolution." The paper paints the landscape of Japan and the 1870 Meiji Restoration — a time of great upheaval of the economic status quo. Unlike other countries of its stature, Japan is (still) very ethnically, culturally, and linguistically homogeneous. Canada, the US, and the UK are some of the world's largest economies; they're also primarily English-speaking, with migrants bringing their native languages from around the world. As the (region of the) language of global order was converged upon, a mono-linguistic society, more or less, meant that more knowledge was available to more people who could make sense of it. It also meant that nations with very different languages and weren’t linguistically included were economically excluded — stalling their progress.

Common language lets the pluralistic society survive. It lets companies hire, neighbours become friends, and more people understand each other, breaking them free from Babylonian destiny. Once upon a time, the only way to have a "common language" was for each speaker to learn a single one. Still, today, with the digital translator and soon technology that will allow translation to happen in real-time, it won't be necessary to take on the burden of language learning, the long, intensive undertaking that it is. That said, even if the Japanese could not learn from Anglo-technical works, they quickly built their capacity to learn independently. As of 1870, a disproportionate amount of codified knowledge was in French and English, but 40 years later, as of 1910, Japanese works had caught up. Considering this, we can attribute this to the rapid economic growth of the era that broke free from “Malthusian equilibrium.”

Patents serve as relics of the work done by its developers, even if, at the time, there isn't a product to show for their work, for whatever reason that may be. Thus, patents are an essential resource for manufacturing capacity. Indeed, if Japan was going to make a comeback after its defeat by China in the First Opium War and if it was going to be prepared for conflict in the decades to come, it “needed a strategy to absorb Western science.” There was also the technical report, which was objectively a patent, but functionally, an instruction manual with no promise or intent of future commercialization — just a guide on how things worked and how to build them.

If a country is an incredibly effective manufacturer, it is best to match its capacity to existing technologies and start building them immediately.

With that, let’s now look at TSMC. While the degree of truth is unclear, by the nature of human memory, we can assume that Moriss Chang's former employer, the semiconductor division of Sylvania Electric Products, takes stock in TSMC's knowledge capital. This is a direct transfer of knowledge, where the knowledge used is only retrieved by Chang, as it would be illegal to codify and share with others, but this is essentially the function of the patent.

Relatedly, part of Japan’s restoration came from recognizing that codifying knowledge (alone) wasn’t enough — there had to be an educated workforce who could use it. Funding was poured into education, and the investment paid its dues. They hired international instructors, most from Britain, who could serve as translators and teachers for Japanese students. While Britain is also the hallmark of the Industrial Revolution, it is an interesting question of the spread of what they could teach and the advantage they brought in being from a country with (the most) novel ideas to share. Furthermore, a weak pedagogical foundation meant almost no instructors in advanced STEM fields. The Japanese also funded the inverse: funding their students to study abroad.

It was the building of codification capacity that influenced education in interesting ways. When they entered school, young agricultural workers grew erudite, could develop numerical literacy, and could “compute bond yields,” producing more crops. In a decade, young Japanese men and farmers were reading technical books, and the benefits of that were visible and widespread. With government pressure, children were returning to school: a trend seen across many nations post the peak of the Industrial Revolution, but one that, in Japan especially, paid out its initial investment many times over because, compared to before, there was now knowledge to teach.

A suit was filed earlier this year with TSMC as the plaintiff and Apple and Samsung as the defendants, claiming they infringed on TSMC's IP. I'm not sure what the churn is at TSMC, but investigating how employees flow through organizations is always interesting from an IP angle. There's a lot of ambiguity with a former employee bringing in their best practices and serving as a spy; how are they distinguished? What does this mean legally? Relevant statutes from the age of the Cold War to parry fascism include the Smith Act (Alien Registration Act of 1940) from the Truman presidency, the Hatch Act of 1939 that prohibits federal employees from engaging in certain political activities, and the McCarran Internal Security Act of 1950 (also known as the Subversive Activities Control Act), which gave the DOJ authority to arrest employees if their employers suspected them of espionage.

To conclude, science, research, discovery, and development rely on the codification of knowledge. When you think about it, the codification of knowledge is at the heart of it all. The goal of the researcher is to build on existing knowledge. Therefore, the body of knowledge they can work with, the extent to which they can research, and the scope of what they do research depends on the codification of preexisting knowledge.