Let’s say an organization decides to do some action X. X could be anything, but for the sake of an example let’s say it’s a decision to fund something. Now everyone who observes this decision being made, could take one of the 3 perspectives:
- Did this decision-making process happen according to the laws that govern the organization (it was lawful)?
- Do I agree with this decision and what can I do about it?
- How do we implement action X?
If you take the 1st perspective you will be asking questions like:
- Was due process followed when deciding this?
- Was due process followed when deciding how to implement this decision?
- Was decision implemented correctly?
If you take the 2nd perspective you will be asking questions like
- If I don’t agree with the decision, how do I stop it and make an organization do something else? Or how do I correct for what was done?
- If I agree with the decision, how can I support it (vote for it?)?
If you take the 3rd perspective you will be asking questions like:
- What concrete steps do we need to take to implement this?
- What are the technical details / considerations involved?
- Who should implement this and how?
- Is it feasible?
The first position is a position of the judge. The second position is a position of a legislator. The 3rd is a position of an executive / administrator. These three positions correspond to three branches of government that a lot of governments consist of. The main feature of this tripartite model is typically advertised to be separation of powers to prevent collusion. I argue that this is false. The main feature and benefit of this model is that it represents reality of how decision-making has always been working1. Regardless if we are talking about an individual human being, a community, a democratic government or a dictatorship - in all of these systems decisions go through iterations of these 3 perspectives being applied to them.
If an organization does not consciously use something similar to this tripartite model it simply means they are doing it unconsciously or haven’t mapped parts of their processes to this model. For example, let’s take a community that does not have any formal process for making decisions at all:
- At some point some member of a community will propose to others to do something (like throwing a party) - legislation
- At some point someone will request someone to do the tedious part of that work (like buying all the food) - execution;
- At some point someone will judge how execution was done (like expressing disappointment about the food and suggesting that it should be done differently next time or simply saying that the food is good) - judging
This is also constantly happening at the level of an individual personality. Planning your time is basically your legislative branch, carrying out that plan is your executive branch. Judgmental voice inside your head when you don’t carry out your plan (don’t wake up in time or miss a meeting) is your judicial branch2.
I’m pretty sure you can find the same model at a biological level as well. Brain sends a command to you nervous system to lift an arm (legislative), your nervous system controls your muscles which lift your arm (executive). If suddenly and unexpectedly you were not able to lift your arm, you would freak out and start thinking that something is wrong with your body or brain. This “something is wrong with my body” is a judgment. Judicial branch here is probably conscious thinking. As is common for the judicial function (in organisations as well) it only kicks in when legislative and executive parts stop working in unison or when their work does not produce expected outcomes.
Software development involves a lot of decision-making so obviously we can find the same 3 functions as well. They come in various forms and at multiple points in the process but, if we imagine software development team as a governance system, it is not hard to put typical activities performed in one of these 3 categories. For example:
- Judicial: verification (does the product satisfy specifications, requirements?), validation (does the product meet the needs of a customer?);
- Legislative: requirements definition, user stories, design specifications;
- Executive: coding, testing, debugging, deployment, system administration;
So we are able to find tripartite model in systems that we don’t even see as governance systems. Therefore I think it’s fair to make a conclusion that tripartite model reflects the reality of how decisions are made.
Why is this important?
First of all, it can help us reason about any decision-making system and help improve it. Simply look for these 3 branches in any system and first of all, check if all of them are present, and second, if they perform their function well3. If you see a problem in a governance system and perform this exercise it can help understand the problem and help organise your work towards a solution.
Secondly, this model provides basics for how to reduce confusion and improve focus of the group, make meetings more productive and overall time in a group more enjoyable. To see this imagine (or remember) what happens when a group creates a plan and then part of the group starts trying to execute a plan while another part of a group keeps questioning the plan and / or tries to change it as it is being implemented? It’s a mess and the group will most likely not get this nor any other plan executed well. This creates frustration for participants of a group, which reduces productivity and make it more likely that participants will start leaving4. Now to solve this the group could agree on a time and place to create a plan, time and place to judge if “this is really our plan”, time and place to execute a plan and not tolerate these perspectives mixing. Once everyone agrees that “this is our plan” the group can execute it without distractions. People whose personal opinions disagree with the plan start focusing on the next plan, instead of undoing the execution of the current one. Or better yet, they can help execute the current plan, in order to help the group learn faults with it faster. The group will proceed with its imperfect plan which is better than not proceeding at all. That will allow them to learn their mistakes and create the next plan to be a little bit better.
Separation of powers in time
The typical idea behind tripartite governance system is that people in the three branches are different to prevent collusion. My arguments above present a case that this model might be useful even if people in the three branches overlap. I will go a step further and present a case that a system where the three branches consist of the same set of people has its own advantages and is maybe even better than typical tripartite systems.
How can we separate our decision-making system into 3 branches if all branches consist of the same set of people? First of, we can separate these three branches in terms of time. For example:
- Schedule a meeting or meetings where we will try to reach consensus on what to do next. This will be our legislative time;
- Schedule a meeting or meetings where we will try to reach consensus on judgments necessary to proceed. This is our judicial time;
- Most of the work towards group goals is done by individuals outside of meetings. This is our executive time;
Executive time can of course involve more meetings, but they should be about implementation details from the perspective of legislative time.
This would be the first step. The next step would be to add a bit more structure to legislative and judicial meetings so that it is clear what their outputs are. You could model each of those meetings as games with clearly defined goals and rules. This would make it easier for people to switch perspectives between judicial and legislative kind of thinking. For example, someone might naturally be leaning towards judging legitimacy of the way the system works currently. But if you tell them: “today we play a game where the goal is to develop the best plan for the next steps” and then structure the game where they are rewarded for contributing to this goal of the game, then they are likely to be able to switch their perspective and be helpful, given that you also have a game where they are able to express their natural tendency towards judging in a meaningful way for the group.
Advantages of allowing everyone to participate in all three branches
No person is only a judge, or only a legislator or only an executive. Everyone has opinions about activities of each of these branches, which also means that they probably have something to contribute to each of them. Inability to express all these opinions in a productive way will gradually grow friction between individual and the group. Expressing them in time and place where they are heard and potentially even integrated makes time in the group more enjoyable for the individual. Making all decisions together (or at least being allowed to), having your concerns heard can grow a sense of unity which a sign of a good team.
Another related advantage of not dividing people into branches is quality of communication. Good communication between people who pass legislation and people who execute it, is critical. Otherwise, you get problems like legislation which is unfeasible in practice or executives misinterpreting the legislation. But if there’s no enforced split between people who pass laws and people who execute, it’s much more likely that they will understand each other. It gets rid of “us against them” attitude if people are able to go into a meeting of the other side and have their influence expressed there.
This model in practice
Purpose of legislative is to decide what we want to do using human-level language that every participant would understand. Purpose of executive is typically to translate human-language decisions into lower-level, typically more technical actions. In DAOs this typically looks as making on-chain transactions.
- EdenFractal (as it has been working so far):
- Judicial - Respect Game and resulting Respect distribution;
- Legislative - Eden+Fractal
- Executive - Msig;
- AW Fractal
- Judicial - Respect Game and resulting Respect distribution;
- Legislative - Agenda game
- Executive Msig;
- Other potential configuration:
- Judicial - Respect Game and resulting Respect distribution;
- Executive - OREC;
- Legislative - Respect trees or some other alternative;
Why is Respect Game a good judicial branch:
- In this game you do have to judge each others work. Respect game makes judgment less stressful by requiring to rank people instead of “good” or “bad” type of judgment;
- Respect game can judge legislative contributions as well as executive ones. Assuming legislative as well as executive powers granted to an individual are proportional to their Respect (which I think should be the case), then bad legislators or executives can have their power reduced to minimum. This achieves the main purpose of judicial branch - to regulate the other two branches;
- Giving contributors executive and legislative powers proportionate to Respect means that executive and legislative branches can function in lightweight mode most of the time with only interested, proactive people participating, but in case of any contraversial issues it is likely to trigger more Respect holders issuing their vote in these branches. So judicial branch is Respect game and resulting Respect distribution which has power to take over both legislative and executive in case it’s needed. This is a useful feature for judicial branch for obvious reasons.
TODO
- Before the internet and blockchains you would have had to gather everything together in order to play these games. With the internet you don’t have to be in the same place anymore in order communicate, you only have to be at the same point in time. Maybe that’s one of the reasons why the separation happen.
- Universality of tripartite governance model
- Separation of powers in time;
- Or to be more accurate and a bit less bold: this model can help make sense of all decision-making systems even if they are not formalized in any way as well as build new ones;↩︎
- One could make an argument that legislative, judicial and executive correspond to ego, superego and the id in the Freudian structural model of the psyche https://en.wikipedia.org/wiki/Id,_ego_and_super-ego↩︎
- Note that these branches are likely to be named differently in different systems and might not even be separated. The point is that you look at the structure of a decision-making system and see which pieces of it perform which of the three functions (legislation, execution or judging). Same piece might be performing multiple functions as well.↩︎
- Setting goals and seeing yourself achieve them is one of the most basic ways to regulate your emotions. Not being able to do that (not producing expected outcomes) is stressful and disapointing.↩︎