Protocols as First-Class Concepts

Issue 17

Early on, during discussions that led to the design of the Summer of Protocols, Ethereum researcher Danny Ryan suggested that the goal should be to establish protocols as a "first-class concept for thinking about the world." At the time, the handful of us in the conversation all immediately resonated with the idea, but we couldn't quite put our finger on why it seemed like the right aspiration. Of course, there's a banal selfish motive here, in that all of us have stakes in the protocolized future. But then, bakers have stakes in bread-making, astronomers have a stake in astronomy, and tomato farmers have a stake in tomato farming. We don't see people routinely making the argument that any of those things should be a "first-class concept."

Why not just study protocols as a interesting but limited subject, restricted to its own "lane," without the foundational and totalizing aspirations suggested by "first-class concept"? What's wrong with just betting on protocols in some narrower way, hoping for money, power, or fame? Academics often talking about "staying in lanes" to gently remind themselves to stay humble and within the limits of their expertise, and mutually police intellectual over-reach. So why are we at the SoP not "staying in our lane," and instead engaging in what might be a kind of intellectual over-reach? Why are we gleefully riding roughshod all sorts of established disciplinary boundaries, wielding "protocol" as not just a lens, but a sledge-hammer?

The answer, I think, is that a strong new first-class concept really levels up how we think about the world, so when you sense one lurking in a muddy discourse, it's worth taking significant intellectual risks to uncover it. A strong first-class concept is something like an intellectual public good. A powerful new dimension added to all thinking about the the world. One that can even refactor (and to a degree, subvert) your understanding of existing, seemingly well-understood dimensions.

What's more, increased conceptual clarity tends to create a virtuous cycle leading to steadily increasing new agency. When you understand protocol better, the world starts getting protocolized faster, because you have newer and more powerful ways of acting on the world. Of course, like all new agency, protocolization is a double-edged sword, as full of dangers as opportunities (an aspect being studied by SoP researcher Nadia Asparouhova). Arguably, there is no such thing as a non-dangerous first-class concept. Part of the challenge of establishing a new first-class concept is to develop a deep understanding of both edges of the new double-edged agency it catalyzes.

So what exactly is a "first-class concept for thinking about the world" anyway?

It's hard to define but most people would agree that certain concepts self-evidently belong in the set: nation-states, corporations, the environment, computing technology, cities, the economy, poverty, war, religions, and so on. We might disagree about the relative importance of various elements and argue about whether this or that element truly belongs, but we don't normally argue about whether tomatoes should be central to geopolitical debates. Random topics can occasionally rise to strategic, global salience (high tomato prices are currently a sensitive political topic in India for example), but only a few topics and concepts are consistently deployed to frame conversations about the future of the world at large for extended periods. First-class concepts comprise the default scaffolding for thinking about the world.

In some ways, it is intuitively obvious that protocols belong in this set. They are ubiquitous, exist at all scales, and are an increasingly consequential and critical part of both established and emerging infrastructure. What is not so obvious is what they in fact are. In other words, the problem of establishing protocols as first-class concepts is not that they need to be championed and promoted from second-class or third-class, but that they need to be understood in less fuzzy ways. They are a nascent first-class concept. They don't need evangelism or championship so much as they need clarification and rigor.

The concept of nation may be fuzzy at the boundaries, but we have a well-defined set of examples, a well-developed language for talking about them, good traditions around studying them in theory, and working with them in practice. There are academics studying nationhood, and diplomats representing nations at the UN. Nobody is particularly confused about what they're all up to.

None of this is true for protocols.

One reason is that protocols form the interstitial dark matter around other first-class concepts. Protocol is a glue concept (the term literally means "first glue"). Protocols are involved in how nations and corporations interact, how the environment is managed, how pandemics are contained, and how computing technology is built and operated. What's more, protocol is also a fractal concept that applies at multiple levels of almost all domains. Conflict resolution protocols exist at the level of nations, corporations, and individuals. There are protocols for administering a single vaccine shot, protocols for studying the effects of vaccines on large populations, and protocols for distributing vaccines across the world. There are protocols for doing, reviewing, and funding academic research.

"Fractal glue" is not exactly an easy substance to study.

But we are making progress. Halfway through the Summer of Protocols, we are beginning to develop the beginnings of a map of the territory, usable definitions, and a growing shared lexicon of terms with which to talk about protocols. We're still at a pre-paradigmatic stage, but confidence is slowly growing for many of us that we're on to something here.

Of course, there's a dangerous assumption buried there -- that there is in fact a coherent paradigm lurking in the muddiness, waiting to be uncovered (even if it isn't a scientific one, the way chemistry buried beneath alchemy turned out to be). But that's part of what makes this an interesting adventure. Perhaps we'll learn that there's no there there and we should stop using the term protocol. Or perhaps we'll come up with an elegant conceptualization (or three) of the idea that levels up how we talk about everything.

Maybe we'll discover the everything is a protocol...

...and always has been.

If that happens, we'll have succeeded in our mission. It isn't going to happen within the scope of a single summer of study by a small group of ~40 people, obviously, but our goal is to begin the exploration.