Partnered With:

God V: What is "institution" anyway?

Vitalik Buterin

Read this article in 25 Minutes

Original title: What even is an institution?

原文作者： Vitalik Buterin

PANews

Special thanks to Dennis Pourteaux and Tina Zhen for their discussion that contributed to this article.

An alternative political compass, recently developed by Dennis Pourteaux, argues that the most important political divide of our time is not between liberty and authoritarianism or left and right, but rather how we think about "institutions". Is the system by which society operates today good or bad? Is the solution to improve them gradually, to replace them with something completely different, or to abolish institutions altogether?

V 神：到底什么是「制度」?

However, this raises a very important question: what exactly is an "institution"?

The word "institution" in political discourse conjures up images of the national government, the New York Times, the university, and perhaps the local public library. But the word is also used to describe other kinds of things. the term "the institution of marriage" is common in English and has generated more than two million search results on Google. If you ask Google point-blank, "The family is an institution," the answer is yes.

V 神：到底什么是「制度」?

ChatGPT's opinion:

V 神：到底什么是「制度」?

If we take seriously ChatGPT's definition that "a social institution is a pattern of behavior and norms that exists in society and is considered essential to its functioning," then the New York Times is not an institution - no one would argue that it is actually essential, and many consider it harmful! On the other hand, we can give examples of systems that Pourteaux's "anti-institutionalists" would probably approve of!

• Twitter

• Bitcoin or Ethereum blockchain

• English Language

• Substack

• Market

• Standards organization dealing with international shipping

This raises two related but somewhat separate questions:

1. Where is the dividing line between what people call "institutions" and what people don't?

2. What kind of world do people who consider themselves anti-institutionalists actually want to see? What should anti-institutionalists do in today's world?

Investigation experiment

Over the past week, I have conducted a series of investigations with Mastodon, where I have provided many examples of different objects, practices and social structures, and have asked the question: is this an institution at all? In some cases, I made different adjustments to the same concept to see the effect of changing certain variables. There are some interesting results.

Here are some examples:

V 神：到底什么是「制度」?

And:

V 神：到底什么是「制度」?

And:

V 神：到底什么是「制度」?

And:

V 神：到底什么是「制度」?

There are more interesting ones: New York Times vs Russia Today vs Bitcoin Magazine, solar system vs what if we started redesigning it, prediction markets, various social mores, etc.

Here, we are beginning to see some common factors. Marriage is more of an institution than romantic relationships, probably because it has the stamp of official approval, and more mainstream relationship styles are institutional than less mainstream styles (a pattern that repeats itself when comparing the New York Times, Russia Today, and Bitcoin Magazine). Systems where decisions are made by clearly visible humans are more like institutions than impersonal algorithmic structures, even if their outputs end up being entirely functions of human-provided inputs.

To further clarify, I decided to do a more systematic analysis.

What are the common factors?

Robin Hanson recently published an article in which he states:

On well-known topics, at least, most people want institutions to take the ideal form: the public recognises the elite, who monitor the experts and pick the details.

This seems to me to be an important and valuable insight, albeit in a different direction: yes, this is the institutional style that people are familiar with, and will not be surprised (they may be when they see the many "alternative systems" Hanson likes to propose), but it is also the institutional style that anti-institutionalists are most vehemently opposed to! Mark Zuckerberg's institutionalized oversight board certainly follows the "elite of crowd-approved oversight experts" template well, but it doesn't really make many people happy.

I decided to put this institutional theory, and a few others, to the test. I have identified seven attributes that seem to me likely to be important features of institutions, in order to determine which attributes are most closely associated with the idea that something is an institution:

• Does it have a "public recognition elite" model?

• Does it have an "elite monitoring expert" model?

• Is it mainstream?

• Is it logically centralized?

• Does it involve human interaction? (e.g. Intermittent fasting won't, because everyone can choose whether or not to do it individually, but the government can)

• Does it have a specific structure with a lot of intentional design behind it? (e.g. Company yes, friendship no)

• Does it have a role independent of the individual? (Democratically elected governments, for example, do this -- they even refer to their leader as "Mr. President," after all -- but a podcast named after its sole host doesn't do this at all.)

I went through the list and rated 35 possible institutions based on my polling of these categories. For example, Tesla gets:

•25% is "Mass Recognition Elite" (because it's run by Elon Musk, who does get a lot of recognition and support in practice as a celebrity, but that's not an intrinsic characteristic of Tesla, if Elon loses legitimacy, he won't be kicked out of Tesla, etc.)

•100% identify with "elite oversight experts" (all large companies follow this model)

•75% say "mainstream" (almost everyone knows it, many have it, but it's not quite a New York Times household name)

• 100% "logical centralization" (most things are 100% in this regard; As a counterexample, "dating site" gets 50%, because there are so many dating sites, and "intermittent fasting" gets 0%)

•100% "involved in human interaction" (Tesla makes products that sell to people, it employs employees, has investors, etc.)

•75% "intentional structure" (Tesla does have a deep structure with shareholders, directors, management, etc., but that structure is not part of its identity, e.g., equity proof consensus applies to Ethereum, voting and Congress applies to government)

•50% chose "personal-independent roles" (while roles in companies are often interchangeable, Tesla does benefit greatly from being part of Elon's universe)

The full data is here. I know that many people will have a lot of disagreement with the various personal rankings I make, and readers may convince me that some of my scores are wrong. My main hope is that I've included enough different systems in the list that individual differences or mistakes can be roughly balanced out.

Here is the relevant table:

V 神：到底什么是「制度」?

But the correlation turns out to be misleading. "Human interaction" is almost certainly a necessary attribute of an institution. The correlation of 0.57 suggests this to some extent, but it understates the strength of the relationship:

V 神：到底什么是「制度」?

Literally, every single thing that I've labeled as obviously involving interaction has more people thinking of it as an institution than anything that I've labeled as not involving interaction. The dot in the middle is my hypothetical example of an island where people with odd birthdays are not allowed to eat meat before 12:00. I don't want to give it 100% because not eating meat is a private activity, but the question still strongly suggests some social or other pressure to follow this rule, so it's not really 0% either. This is a place where Spearman's coefficient is better than Pearson's, but I'd rather just show the chart than spout outlandish numbers. Here are the other six:

V 神：到底什么是「制度」?

The most surprising finding for me was that "role independent of the individual" was by far the weakest correlation. Twitter run by a democracy is the most institutional, but Twitter run through a paid governance plan is just as institutional as Twitter run directly by Elon. Being independent of the individual adds a guarantee of stability, but being independent of the individual in the wrong way feels too foreign, arbitrary, or uninstitutional. Dating sites are more independent of individuals than professional matchmaking agencies, but matchmaking agencies are seen as more like institutions. Experimenting with highly role-driven and mechanically credible neutral media (such as this neat design, which I actually think is really cool) just feels alien - maybe bad, but maybe good, and if you find today's system frustrating, you're open to possible alternatives.

"Elites recognized by the public" and "elite supervision experts" are highly correlated. The second one is taller than the first, although Hanson and I may have different meanings for "recognize." The lower right corner of the "intentional structure" chart is empty, but the upper left corner is intact, indicating that intentional structure is necessary but not sufficient to make something institutional.

That said, my main conclusion is probably that the word "institution" is a mess. Unlike the term "institution," which refers to a set of concepts (e.g., "high modernism"), the term appears to have a number of different definitions:

1. This structure fits the familiar model of "the public recognises the elite and supervises the experts"

2. Any large structure deliberately designed to mediate human interaction (including financial markets, social media platforms, dating sites, etc.)

3. Widespread and universally regulated social customs

I suspect the anti-institutionalists focus their skepticism on (1), especially since the example of (1) has been seized upon by the wrong tribe. It doesn't seem to matter to anti-institutionalists whether the structure is individualistic or role-driven: personalities (" Klaus Schwab ") and bureaucracies (" awakened scholars ") can equally come from the wrong tribe. Anti-institutionalists generally do not oppose (3), and in many cases actually want (3) to replace (1) as much as possible.

Support for (2) may be closely related to Pourteaux's distinction between "techno optimist" and "techno minimalist". Tech minimalists don't see Twitter, Substack, bitcoin, Ethereum, etc. as part of the solution, although there are "bitcoin minimalists" who see the bitcoin blockchain as a narrow exception, otherwise they would like to see a world where more outcomes are determined by things like family. The "techno Optimist anti-institutionalists" are specifically involved in a political project to either try to replace (1) with the right (2) or to reform (1) by introducing more of the right (2) elements.

What is the direction of travel for anti-institutionalists or institutional reformers?

It would be wrong to attribute a plethora of deliberate tactics to anti-institutionalists: anti-institutionalism is a movement that is more united in opposing than supporting any particular alternative. But what is possible is to recognise the pattern and to ask the question of which paths forward make sense for anti-institutionalists.

From a linguistic point of view, even the use of the word "institution" at this point seems more likely to confuse than enlighten. (i) a desire to replace structures containing elite roles with structures that do not contain them, (ii) a preference for small and informal structures over large and formal ones, (iii) a desire simply to replace the current elite with a new elite, and (iv) a socially liberal position that individuals should be driven by their own ideas, Rather than being driven by incentives created by other people. The word "institution" masks this divide, and may focus too much on what is being dismantled rather than on what is to be built upon.

V 神：到底什么是「制度」?

Different anti-institutionalists have different goals. Sure, the people who make the powerful, sharp criticisms of the New York Times on Twitter agree with you about how society shouldn't work, but are you sure they're going to be your Allies in deciding how society should work?

The challenge of avoiding structure altogether is obvious: there is a prisoner's dilemma and we need incentives. The challenges of small-scale and informal structures are often obvious: the gains from economies of scale and standardization -- although sometimes the other benefits of informal methods are worth losing. The challenge of simply swapping elites is obvious: it cannot socially scale to cross-tribal consensus. This is more credibly neutral if the goal is not to worship a new set of elites forever, but to keep the elite in high turnover forever (see Balaji's founder-heir dichotomy), but then it begins to approach the realm of avoiding universal elite worship.

It's interesting to create formal structures without an elite, not least because it hasn't been fully explored: There is a strong case to be made that institutions with elite roles may be an unfortunate historical necessity in times of more restricted communication, but modern information technology (including the Internet and newer and scarier things such as zero-knowledge cryptography, blockchain and the DAO) can rapidly expand our available options. Still, as Hanson points out, this path has its own challenges.