What is Social Knowledge?

This article is about a subtle phenomenon that is in the blindspot of both Information Idealism and Coordination Idealism.

Logic

General Logic

When formally studying knowledge, the most common tool is logic.

In general, logic is the field that concerns itself with representing knowledge in terms of logical statements to then formally reason about them.
For instance, we can write down our knowledge as "P = It rained this morning" and "P -> Q = If it rained this morning, then the birds won't sing today".
With logic, we can deduce from this that "Q = The birds won't sing today", without needing to understand the meaning of the statements. Just from the application of the rules of logic: P and P -> Q together imply Q.

Epistemic Modal Logic

The technical sub-field of the logic that concerns itself with social knowledge is Epistemic Modal Logic.

Epistemic Modal Logic is an extension of regular logic. While regular logic is focused on knowledge of the true facts of the world, epistemic modal logic is focused on knowledge of what people know about the world.

For instance, in epistemic modal logic, you could say:
1.a. Alice knows that if it rains, the birds won't sing for the day.
1.b. Bob knows that it rained this morning.
2.a. But Alice doesn't know that it rained this morning.
2.b. And Bob doesn't know that birds stop singing when it rains.
3.a. Thus it is known to neither Alice and Bob that the birds won't sing today.
3.b. And if they pooled their knowledge, they would both deduce that the birds won't sing today.

Social Knowledge

What is Social Knowledge?

Some skills and pieces of knowledge are not universal: driving a car, editing a video, some historical facts, etc.

Given that they are not universal, we can not assume that either everyone knows them, or no one knows them. For instance, when we say "We know how to go to the moon", we do not mean that everyone single human being knows how to go to the moon. We actually mean that something like .01% of the population, if they were to work together, could go to the moon.

Thus, there is a form of knowledge that is about who knows what. This is social knowledge. It is not about knowing what's up, it's about knowing who knows what's up.

Why do we care about Social Knowledge?

Why do we care about knowledge in the first place? To do things in the real-world. If we know how to drive, not only can we drive, but we can also build much more ambitious plans that involve driving.

When we solely consider plans that involve ourselves, things are simple: we know what we know and can do, and we plan accordingly.

But Social Knowledge becomes relevant when our plans involve others.

When we plan for a group, what matters is not what we know, but what people in the group know. If we know that at least one person in the group can drive, deal with taxes, etc., we can start building plans that rely on those skills.

In short, when we need to rely on the knowledge of others, social knowledge is what matters.
Furthermore, most of our knowledge is social: we directly know very few things. Few of us know how to play music, plumbing, operate computers, etc. But almost all of us know people who do.

This is what I mean by Social Knowledge being in a blindspot. It is our main form of knowledge, and the closest I have seen it formalised is in a niche subfield of logic.

Social Knowledge Management

The theory

So far, so good. It seems like social knowledge is a straightforward: it is when we know what someone else knows.
But in real-life, things go much deeper. Consider a company, a state or a military. We can not have each employee/citizen/soldier go to Mr CEO/President/General and tell him what they know. There is just too much knowledge spread across everyone for this to be tractable.

Thus, we start getting into nested social knowledge. Now, the CEO can not know what the workers individually know. But he can appoint people whose job is to know what the workers know.
And this is why the CEO must get middle managers, the president must get ministers, the general must get lieutenants, etc.
And to appoint them well, the person at the top must have social knowledge of what the middle layers know. The CEO must know that the middle managers know what the workers know.

In practice, the chains of transmissions go even deeper. The general must know what the lieutenants know of what the captains know of what the sergeants know of what the privates know.
The president must know what the ministers know of what the secretaries know of what the low-level civil servants know.
The CEO must know what the C-Suite executives know of what the VPs know of what the middle managers know of what the minions know.

The job of the president is almost entirely social. It is not to directly evaluate what civil servants can do, and more to ensure that his ministers have good knowledge pipelines where information flows from the lowest level of civil servants to the highest level of government.

This is how organisations can scale and work in practice. Else, there is just much more information that needs to be processed than what a single person can handle.

The practice

What I just described is a bit of a theoretical ideal.
I am describing big organisations that grow rationally, optimising for coherent planning and a neat flow of information.

In practice, social knowledge becomes complicated super quickly, even within small groups.
We just know too much to be able to track it by listing it all on a sheet of paper.

Even in a small team of 5 people, as people work more together, they will get a more refined understanding of what each of them can know: they will have seen how long it takes each of them to perform various tasks, how much they struggled with certain topics, etc.
Within that team, there might be a pair who almost always works together. They will have an even deeper understanding of what they can do, even compared to the rest of the team. We expected that as a result, regardless of their expertise, just because they know each other well, they will be able to coordinate better.
Thus, when planning, we will take this into account, and will likely put those two people together whenever possible.

The example above already features deep nested social knowledge. We are saying that we know that people in the team of 5 know that the pair know a lot about each other.
There is no need for presidents or generals to get into the complexity of nested social knowledge. It arises naturally whenever we need to work together.

Social Knowledge matters at all scales, from the smallest groups to the entirety of humanity.

Conclusion

I often think of social knowledge as a pre-condition for reliable group planning.
If there is no person or entity that knows what others can do, it becomes simply impossible to make good plans that leverage the skills of all members of the group.
Conversely, to the extent that we can build plans that leverage the skills of many members, it is only because there is some social management in place, in one way or another.

I have more to say about the topic, but I needed a sharp explanation of the basics first, rather than just pointing at vague intuitions.