When first introduced to bitcoin I experienced an epiphany of sorts in realizing I didn’t understand what money was. One of my first assignments was to educate myself on what makes money money. This gave me a new understanding and appreciation for bitcoin and blockchain technology as a whole.
I believe the Bitcoin, Ethereum, and open-source blockchain communities are collectively suffering from a similar ignorance revolving around the idea of consensus. I call this affliction ‘consensus maximalism’ and hope this post can help shed light on this issue and share a different perspective on this concept.
Only known photo of the world’s earliest blockchain.
Defining the Term
Merriam-Webster provides 2 variations of the primary definition for the term consensus. They are as follows:
Most notable about these definitions is neither includes any reference to absolute agreement. In lieu of absolute, total, or complete the definitions contain the terms general and most. This is a common oversight. I’ve observed this misunderstanding frequently in the Bitcoin and Ethereum space … I call it ‘consensus maximalism’.
Consensus Maximalism vs. Consensus in the Real World
I would define ‘consensus maximalism’ as the belief that to reach consensus there must be absolute agreement by all parties as to the best path forward. The fallacy is this level of rigidity would make any real-world consensus relatively impossible. Real progress requires a real-world definition of consensus.
Dr. John Toussaint, a former doctor & CEO, leadership guru and author famous for his contributions to organizational transformation work in the healthcare space was the first to bring to my attention that consensus is most often never about reaching full agreement. It is best described by the following excerpt:
…synthesizing the wisdom of all the participants into the best decision possible at the time. When you consent to a decision, you are giving your permission to the group to go ahead with the decision. You may disagree with the decision, but based on listening to everyone else’s input, all the individuals agree to let the decision go forward, because the decision is the best one the entire group can achieve at the current time.
After all, the term is rooted by the word ‘consent’. Very few systems, organizations, or communities have ever reached unanimous agreement on any decision. They always require a minority to ‘consent’ to moving forward in spite of their views for the benefit of the majority. Participants often have the option to ‘exit’ should their convictions be strong enough.
The Irony of Consensus Algorithms
The reality is that even in distributed systems consensus is never 100% at any time. This is displayed by orphaned blocks, alternatives cryptocurrencies, and of course hard forks. It is the great irony of this space that the communities supporting these systems struggle so greatly to achieve consensus in their effort to innovate upon consensus-mechanisms (blockchains).
In my view, as with many things in life, there is likely a smaller minority of ‘consensus maximalists’ who act as the squeaky wheels. This is not to say that this is all bad. In fact I believe it has some great benefits including:
- Increasing the difficulty to change highly distributed protocols with larger consensus pools (Bitcoin Core’s relative conservatism as an example)
- Encouraging outliers & dissenters with great conviction to ‘exit’ or support minority forks for communities & protocols. This drives creativity & innovation. (Every alt-coin ever)
It stagnates progress to the limits of acceptance encouraging participants to ‘exit’. Nothing is more painful then being in the deadlocked state of consensus failure. Be it the decision as to where to go to dinner…
You: “I want to eat out … where would you like to go?”
Partner: “I don’t know, nothing sounds good.”
You: “Ugh… I don’t even want to bother.”
…or the debate on how to scale the world’s most promising blockchain.
The failure to reach consensus often blinds the minority participants from the shared purpose/goal of the entire population. Gun controls a good example of this. Nobody wants mass shootings but many people have strongly held opinions as to how to impact the problem… The ugly reality is as a result many of these camps end up fighting one another rather than focussing on actual experiments/solutions.
This unfortunate reality has reared its ugly head several times in recent years … examples including the bitcoin blocksize debate and the DAO / Ethereum hard fork.
The Good News
One benefit that blockchains have is the almost instant ability for those that do not consent with the path forward to fork. Ultimately anyone supporting or interacting with the longest chain should have nothing to complain about… by doing so they are ‘consenting‘ to the decisions of the majority of hashpower.