It turns out that carving things into stone is a pretty good way to preserve the canonical truth. But first you have to agree on the truth, which isn't so straightforward. (Photo by Luca Santos on Unsplash)
When many voices shout at once
None considers himself a fool;
Though the Sangha is being split
None thinks himself to be at fault.
Consensus is something we talk a lot about in the blockchain space. We tend to talk about it in the context of state and transactions: a blockchain is an engine that allows many untrusted systems to agree on things like the order of transactions and the state of a ledger. Of course, the idea of consensus is much older than blockchain, the Internet, and computers. If we really want to understand blockchain and why it’s such an important technology, a good place to start is by understanding consensus on both a technical and social level.
While blockchain can feel radically new and revolutionary, like all technologies it doesn’t enable anything that wasn’t previously possible or desirable.1 At least since the invention of language humans have found ways to reach consensus, sometimes over almost unbelievable time and distance. The way in which blockchain enables consensus and the underlying technologies it relies on are novel, but the idea of consensus is not.
Sometimes, however, making something orders of magnitude faster, easier, and cheaper can change the world. Blockchain technology has this effect on the process of reaching consensus. To understand why this matters—to see how far we’ve come, what a difference modern technology makes, and why consensus is so important to human society—we can look at previous technologies for reaching and maintaining consensus, and the role those technologies have played in society. There’s a lot we can learn about what worked and didn’t work previously!
Simply put, consensus formation is the process by which a group of people discuss, debate, and ultimately reach agreement. Consensus is about more than agreeing on facts: groups also need to come to consensus on beliefs and practices. Indeed, one definition of consensus is “group solidarity in sentiment and belief.” Humans are social creatures and consensus is fundamental to society. Couples, families, communities, firms, and countries all need to reach consensus in order to function. Sometimes the process is quick and painless. Other times it’s long, expensive, and difficult, and the results of a failure to achieve and maintain consensus are plainly visible: witness the recent United States election and the dysfunctional state of American government.
Consensus can be formal or informal, structured or unstructured, intentional or unintentional. Contrast, for instance, legal, judicial, and political processes to the way languages and social norms develop. All of these processes are examples of technologies of social scalability, tools that allow institutions to overcome inherent limits in human cognitive abilities and scale the number of people that can participate in or benefit from that institution.
For instance, when two people sign a legal contract, they don’t need to worry too much about the contract being enforced since they can trust the legal system to provide recourse if the other party breaches the contract. This allows both parties to focus on the substance of the transaction rather than the infrastructure. In the same fashion, I don’t need to worry whether or not you’ll be able to read this article since it’s published in a common, legible language (English) using a common stack of technologies (the Web). This allows me to focus instead on the content and the ideas.
There are many examples of technologies of social scalability and consensus formation. These range from concrete ideas and technologies like the postal service, telegraph, telephone, the Web, the automobile, and commercial aviation to abstract ideas as large as language, money, democracy, the nation state, and international law that enable consensus among billions of people. All are the outcome of long and successful processes of consensus formation, and all in turn enable easier consensus formation and even greater social scalability.
The best technology humans have come up with for successfully achieving, scaling, and maintaining consensus over enormous groups of people, and over enormous time and distance, is religion. The largest religions have existed for thousands of years and have billions of followers, larger than even the largest nation states and the most popular languages. If we wish to understand how consensus is achieved, how it’s maintained over time, how it evolves, and what happens when it fails, religion is a good place to start.
Other things being equal, consensus is less important when things change less often. We don’t tend to think of religion as being especially dynamic or contemporary so, naively, it’s not immediately obvious why consensus is critical to its survival and success. In fact, however, evolution is the only way to achieve longevity and all successful religions evolve. The oldest religions have perfected the balancing act of evolving just enough to stay relevant and legible, while not evolving so much that they lose their character.2 A robust consensus mechanism is key to this success.
I’m not a religious scholar, so I can’t explain precisely how religions establish, scale, and maintain consensus across billions of people and thousands of years. However, I recently came across a fascinating, modern example of this consensus process in Buddhism, involving debate about the validity of an ordination ceremony, that may help answer this question.
This example struck me for several reasons. First, because the religion in question, Buddhism, is one that appears especially conservative to me. When I think of Buddhism, I think of timeless images of ancient statues and of monks and nuns meditating and reciting mantras by candlelight. Second, the doctrinal point in question is as old as the Buddha. I had naively assumed that in the 2500 years since then Buddhism would have worked out the finer points of doctrine, at least around practices as important as ordination. Third, while the debate seems insignificant to the uninitiated, several of the most prominent Buddhist scholars and practitioners in the world participated, demonstrating the importance of this question to the Buddhist community today. Finally, this debate is especially interesting because it relates to women’s rights and gender equality, both important, modern topics that have relevance far beyond Buddhism.
In brief, the debate involves the ordination of four nuns (bhikkhunī) during a single ordination ceremony at an Australian monastery, Bodhinyana, in 2009. The debate centers not on the doctrinal fitness or eligibility of the nuns—indeed, all had substantial training prior to the ordination—but rather on procedural details of the ordination process. Following the controversial ordination, a number of scholars from the Thai Forest Tradition (of which Bodhinyana was a branch) and other Buddhist schools weighed in on the ceremony’s validity. While the debate was overwhelmingly polite and respectful, this should not lead us to believe that it wasn’t heated. It appears to have caused a large divide in the Buddhist community, and in the end, the Bodhinyana Monastery and its senior monks, including those who conducted the ceremony, were expelled from their lineage. As one commentator put it:
[T]he Buddhist world watches with dismay the clamor arisen among senior bhikkhus (monks) in the wake of the recent ordination of bhikkhunis (nuns) into their lineage. - Ven. Sudhamma Bhikkhuni
The details of the debate are abstruse and not terribly relevant for our purposes, but they’re laid out in great detail in this letter from Ṭhānissaro Bhikkhu, one of the foremost scholars and prolific translators of early Buddhist texts.3 As a renowned scholar of the monastic code and head of another Thai Forest Tradition branch monastery, he was asked to review the validity of the ceremony in question.
What struck me in particular about Ṭhānissaro’s letter is its familiar language, including words such as “protocol,” “validate,” and “trust.” He refers to the ordination itself as a “transaction,” a word he uses no less than 56 times in the letter. Here’s one example, which should look familiar to any blockchain developer:
“The passage… lists five ways in which a transaction statement is rendered invalid, thus invalidating the transaction as a whole”
According to Ṭhānissaro the ordination is a transaction, and that transaction is only deemed valid if it follows certain rules—i.e., conforms to a protocol. He goes on to say:
[W]hatever a “transaction that is not a transaction” claimed to accomplish would automatically not count as accomplished… this would mean that the candidates accepted through such a transaction would not count as genuine bhikkhus [monks] or bhikkhunīs [nuns]
This is eerily similar to the way we talk about transaction validity in blockchain. Whatever change an invalid transaction “claimed to accomplish,” e.g., transferring a balance from one account to another, would automatically fail to be applied to the ledger. In the case of Buddhism, the ledger in question is the set of ordained monks and nuns (along with the lineage of who ordained whom).
Ṭhānissaro explains that there are certain, rare exceptions to the protocol:
[A]n Acceptance transaction that does not mention the preceptor would break with the authorized pattern… [and] would grant an exemption from following the authorized form
In blockchain, we have a word for an exceptional transaction that doesn’t follow the ordinary rules of the protocol but is still deemed valid: an irregular state change.
Ṭhānissaro’s language is echoed by several other commentators who opined in response to his letter. For instance:
[A]n ordination is said to “fail”… because of faults in the proclamations - Bhikkhu Bodhi
We might say the same thing of a blockchain transaction: that it fails because of faults in its “proclamations,” e.g., trying to transfer an amount that’s greater than the sender’s balance at the time the transaction is processed.
Here’s another, subtler example:
[An] ordination ceremony culminates in a motion and three proclamations identifying the candidate and his preceptor, and asserting that the candidate seeks ordination and is qualified; if all remain silent, the proclamations become the decision of the group, and the ordination is a completed legal transaction. Permission came for lumping together proclamations for multiple candidates, when two men planning to ordain under one preceptor quarreled over whose proclamations would be spoken first… Being first presumably would determine not only who felt more favored but which student would become the more senior of the two, a matter greatly impacting the co-students’ daily lives. - Ven. Sudhamma Bhikkhuni
This quote sheds light on the details of the ordination process and on how the group of monks reaches consensus. The emphasis on “if all remain silent” sounds eerily similar to the idea of rough consensus, pioneered by the IETF and adopted by blockchain communities including Ethereum as a governance mechanism.4 It also reminds me of certain Byzantine agreement protocols where participants in a consensus process gossip proposals and acceptance or rejection of those proposals to one another through several rounds until the group reaches consensus—resulting in silence on the network.
Additionally in Buddhism as in blockchain the order of transactions matters! If one monk is ordained before another, even as part of the same ceremony, he becomes more senior, a matter of great importance in the community of monks.
In closing, Ṭhānissaro explains why it’s so important to follow the rules of the protocol, even when they seem obscure and “heartlessly legalistic”:
We have to remember, though, how the Buddha instituted the Saṅgha [community of monks and nuns]. He created no overarching organization to administer or police the survival of his Dhamma and Vinaya [teachings and monastic code]. Instead, he established rules, protocols, and other patterns of behavior, entrusting each local Community with the task of governing itself in line with those forms. The act of adhering to the authorized forms for Community transactions is one of the few ways we have of showing to ourselves and others that we are deserving of the Buddha’s trust.
This statement particularly resonates for me. In a few short sentences, the author has captured the reason a protocol exists and the central role it has played in Buddhism. While some religions are highly centralized and hierarchical, Buddhism is famously and perhaps uniquely decentralized.5 There are thousands of Buddhist schools, sects, subsects, and movements around the world. Without a central agency to oversee or “validate” each individual school and its “transactions,” the schools must instead validate one another, and the protocol is the only way they have of doing this. They may disagree on the finer points of doctrine, but as long as they adhere to certain central tenets, in particular those specified by the Buddha, they can rest assured that they and their “transactions” will be recognized as valid by other schools. As another of the commentators points out, consistency matters:
[A]n important criterion is consistency. Given that according to the discourses the Buddha himself presented consistency as a criterion of truth, it would be reasonable to expect that the Buddha was coherent in his views - Bhikkhu Anālayo
Reaching consensus and maintaining consistency today is facilitated by technology such as airplanes, telephones, and the Internet. How on earth did Buddhism achieve and maintain consistency for thousands of years before these technologies existed?
It achieved this feat using an old but remarkably robust technology: putting everyone in a room together from time to time (or, in the case of Buddhism, a cave). Starting three months after the death of the Buddha, the international Theravāda community has met every so often to come to consensus on a number of important points of doctrine, the most important of which is the contents of the canon, i.e., the protocol. Six times over the past 2500 years, a group of hundreds or thousands of senior monks have met and recited the canon in its entirety. This process can take as long as nine months as the canon is 20,000 pages long and all of the monks must chant every word. It ends when the monks agree unanimously on the contents:
These collective recitations which were performed by the monks at all these Dhamma Councils are known as the ‘Dhamma Saṅgītis’, the Dhamma Recitations. They are so designated because of the precedent set at the First Dhamma Council, when all the Teachings were recited first by an Elder of the Saṅgha and then chanted once again in chorus by all of the monks attending the assembly. The recitation was judged to have been authentic, when and only when, it had been approved unanimously by the members of the Council (Chaṭṭha Saṅgāyana, The Six Dhamma Councils)
The most recent such council occurred in 1954 in Myanmar. The final product of that council is available on the Web, but previous councils used older technologies to record and disseminate the resultant canon. After the council of 1871, the entire canon was carved into 730 enormous marble slabs that are still on display in a Burmese pagoda. What more permanent way is there to record canonical truth than to literally carve it into stone and put it on public display? And the fact that the canon was divided into 730 contiguous stone blocks is the icing on the cake to fans of blockchain.
While the Buddhist process for achieving and maintaining consensus has been quite successful, it hasn’t been entirely without issues. Unlike the Bible or the Quran, the Buddhist canon has traditionally been considered more open. Various texts were added to it over time, and in modern times, texts that seem less relevant to the world today have also been removed. Unfortunately, the details of who added or modified what, when, have been lost to the sands of time. As a result, certain elements of the canon are regarded as older or purer, and as doing a better job of conveying the original words of the Buddha, than others. This is referred to as “stratification,” which also sounds quite familiar: layers are piled on top of older layers, and the older layers become more authoritative, trustworthy, and harder to change over time.6
Reviewing the sources of gender discrimination in the Buddhist canon and the Thai Forest Tradition, another commentator describes this process of stratification in detail:
The Pali texts are not the only textual renditions of the Buddha’s early oral teachings; and the Pali texts, as well as other early texts such as those of the Middle Indo-Aryan languages, have their strata. This means that they exist in layers that developed over time, just as is true with the modern English language… Scholars of the Pali language and other ancient languages can… read such differences of time and place, called stratification.
From this study of the stratification of the Buddhist texts, we understand that both the Pali Suttas and the Vinaya [monastic code] have taken shape and come to be in their present form over a vast period of time. The last Buddhist Council, in which the Pali Canon was set in its current form, took place in Burma remarkably recently, just a little over 50 years ago…
The Vinaya is commonly spoken of within the Thai forest tradition as one of the oldest and purest teachings of the Buddha. But the Vinaya shows such textual stratification. It developed a little differently in different times and places; some parts of the Vinaya where [sic] “closed” earlier, while other parts continued to develop. - Going Beyond Gender Ambiguity in Theravada Forest Tradition, Tathaaloka Bhikkhuni
Tathaaloka Bhikkhuni goes on to argue that, while Buddhism has a long history of patriarchy and male chauvinism, this was not the case in Buddhism’s earliest days. She points out that sexism and chauvinism in Buddhism are instead a product of the patriarchal cultures through which Buddhism has been transmitted, and that the process of stratification has obscured the Buddha’s original attitude towards women and his female disciples.
The route the canon has taken from the original teachings of the Buddha to us today is anything but a straight line. It has meandered and been filtered through many countries, cultures, languages, traditions, and other religions, coming into contact with, affecting, and in turn being affected by each of these. Along the way, commentators, scribes, and translators have also made alterations, some intentional, some unintentional. Different branches of the canon have evolved independently, sometimes in isolation for hundreds of years. This independent evolution has given rise to an enormous breadth of schools and traditions, many of which survive and thrive today. Branches and evolution are unsurprising given the enormous time and distance, cultural as well as geographic, that the canon has traveled. What’s perhaps more surprising is the fact that, in spite of this meandering journey, “Buddhism” still exists as one distinct, intact, legible religion. This is testament to the difficult and often dangerous work that Buddhist monks did to translate and carry the Buddha’s teachings all across Asia, and to the success of the consensus process.
It’s also interesting to note cases where independent, isolated branches of Buddhism were later reconciled, partially or fully. For instance, the Theravada school evolved largely independently in Sri Lanka and Myanmar, two especially active centers of scholarship and commentary. While the canon was eventually reconciled at the Dhamma Councils and a single canon was agreed, these different evolutionary paths and the process of stratification means that many differences in interpretation remain among different schools and traditions. This makes it difficult and contentious to determine the proper protocol for transactions such as the Bodhinyana ordination.
It doesn’t take much work to translate ideas such as stratification, forks, and reconciliation to the language and context of blockchain. I’ve heard blockchain compared to religion many times. Satoshi Nakamoto is often called a Christ figure. Religious canons like the Bible are sometimes referred to as the original blockchains as they don’t change often, they’re rather expensive to get something into, they get harder to change over time, and they’re about as permanent as any form of human knowledge can be. I never took those ideas very seriously or gave them much thought, but I’m beginning to. I’m beginning to wonder what else I’ve been missing.
However, these are all very abstract ideas. What do they mean for blockchain communities today? What can we learn from the history of Buddhism and the Theravada canon, and from the Bodhinyana incident and other examples of conflict and reconciliation in religious communities?
One possible lesson for the blockchain community is that, in blockchain as in biology, diversity is beneficial. As the Dalai Lama put it:
I find this idea—which is so emphasized and explicitly pointed out in Buddhist literature—very attractive: there is a diversity of mental dispositions and receptivity interest, and spiritual inclinations existing among humanity. In Buddhist literature, all Buddhist schools of thought follow the same teacher, Buddha Shakyamuni. However, there is such diversity in the teachings that are attributed to the Buddha—some of them may even seem to contradict each other—that we are prevented from falling into dogmatism. All these various teachings are aimed toward sentient beings’ diverse mental dispositions, needs and spiritual inclinations. And when I understand the truth of this, I am able to truly appreciate the richness and value of other traditions, because it enables me to extend the same principle of diversity to other traditions as well. - from The Dalai Lama, The Good Heart: A Buddhist Perspective on the Teachings of Jesus (Boston: Wisdom Publications, 1996), pp 71-73
Diversity in practice and teaching allows Buddhism to make itself approachable to the “diverse mental dispositions, needs and spiritual inclinations” of people everywhere. It has likely made Buddhism antifragile, capable of adapting itself to different cultural settings and evolving in response to threats. Embracing diversity may be one way that religions like Buddhism have enjoyed such longevity. In blockchain terms, this could mean that having many chains and many protocols is desirable in order to serve diverse attitudes and needs, and to adapt to local regulatory environments.
I think the best lesson from the Bodhinyana incident is that, important as it is, scripture and protocol are not intended to be interpreted literally. In the case of the Bodhinyana ordination, the rule in question, which prohibits multiple nuns from being ordained in a single ceremony, was originally intended to protect nuns and ensure that they received sufficient training and instruction as their order was new at the time and there was a shortage of lodging. As several commentators argued, the situation in 2009 was very different, as the nuns had received ample training and had the support of a large, active community, and therefore the original offense “hardly applies in present times” (Ajahn Brahm).
One of the reasons Buddhism and other major religions have managed to stay relevant is precisely the way in which they’ve balanced the need to evolve and adapt themselves to modern, local preferences against the need to maintain ancient traditions. While they may appear monolithic to the untrained eye, the most successful religions are very much living, evolving projects that don’t stand still for a single moment. In this way they are very much like companies, countries, and code repositories.
In order to accomplish balanced evolution, there must be room for interpretation, compassion, and judgement in light of present circumstances, all things that don’t encode well into a protocol. Several of the commentators spoke to this point:
Vinaya [monastic code] is not a system of absolute black and whites, with immediate invalidation of anything that transgresses a minor rule. On the contrary, Vinaya, as laid down in the texts, is a highly flexible instrument, which clearly tries to be as reasonable and contextual as possible. - Ajahn Sujato
[T]he pragmatic principle of adjusting to circumstances [is] a recurrent feature in the formation of rules as documented throughout the Vinaya. In the end, tradition–which I personally highly value–only stands a chance to survive if it is able to adjust to changing circumstances without loss of what is essential - Bhikkhu Anālayo
The Buddha emphasized flexibility and pragmatism in his teaching, an idea known as upāya and often translated as “expedient means.” Upāya dictates that a teaching must be geared towards the audience and must be guided by wisdom and compassion. As scholar Peter Nelson writes:
[T]he Buddha asserted that his spiritual teaching was only a useful ‘pointer’ to the truth—not the truth itself. This meant that the dhamma—because merely provisional—was neither fixed nor dogmatic. Rather, being an expedient ‘means’ (upāya) to an end, this teaching was—like Socrates’ celebrated dialectical method—dynamic and contextual. This pragmatic philosophical stance gave the Buddha enormous flexibility—it meant he was free to adjust or change his teaching to suit the level of his audience.
The Buddha also emphasized that we should not blindly adhere to protocol, tradition, scripture, or math, but that we should instead always trust our own experience:
Don’t go by reports, by legends, by traditions, by scripture, by logical conjecture, by inference, by analogies, by agreement through pondering views, by probability, or by the thought, “This contemplative is our teacher.” When you know for yourselves that, “These qualities are unskillful; these qualities are blameworthy; these qualities are criticized by the observant; these qualities, when adopted & carried out, lead to harm & to suffering”—then you should abandon them… When you know for yourselves that, “These qualities are skillful; these qualities are blameless; these qualities are praised by the observant; these qualities, when adopted & carried out, lead to welfare & to happiness”—then you should enter & remain in them. - Kālāma Sutta
In the blockchain community we tend to adhere to strict protocols and be “heartlessly legalistic,” to borrow Ṭhānissaro’s phrasing. Indeed, in communities including Bitcoin and Ethereum there is widespread desire that the protocol be “neutral” and free from all human interpretation and social governance. No human interpretation of a protocol means no compassion, judgement, or forgiveness. It means no adjusting or interpreting the protocol to suit the audience. It means interpreting the protocol itself as the truth, rather than as a “pointer” to the truth.
In Autonocrats & Anthropocrats I argued that we have a powerful tool at our disposal in blockchain, and while we should use it to augment human judgement, we should not attempt to outsource too much agency to an algorithm or a protocol. If we want blockchain and the things we build on it to be around for a long time, we’d do well to learn the lesson of flexible protocols like the Buddhist canon that leave room for plenty of interpretation, and have stood the test of time and touched the lives of billions of people. Technology alone is never a solution. The Dalai Lama put it best:
[S]cience, technology, and material development cannot solve all our problems. We need to combine our material development with the inner development of such human values as compassion, tolerance, forgiveness, contentment and self-discipline.
I’m sure there are many other examples of lessons we can learn from Buddhism and other religions. What can we learn from the relationship among various religious schools, practices, and traditions? What can we learn from their history of conflict and cooperation? How do we avoid the mistakes religious communities have made? What successful examples can we turn to?
I leave it as an exercise to the reader to come up with other examples of lessons we can draw from scholarship, practice, and history in the religious world. I look forward to continuing the conversation in the comments below, on the forum, and in future articles here.
This is probably a slight simplification. It’s more of a process of feedback loops, combination, and gradual evolution. Technology starts out by making it easier to do something we already wanted to do, but along the way we often discover other things that the technology also enables that we didn’t originally intend (often in combination with another technology). To put it another way, a technology often has second- and third-order effects that were unforeseen. The Internet was famously funded by United States Department of Defense and was first used to share research and access to mainframe systems. While its creators could probably have predicted email and instant messaging they could hardly have foreseen smartphones, social media, and sharing economy. We have civilian nuclear power because of the Manhattan Project and modern jet travel because of the German Luftwaffe, both part of World War II. Early photography later enabled film and, of course, also helped lead to social media. The first clocks were used for administrative purposes and to make more precise astronomical measurements, but they also made navigation much easier. Makes you wonder what blockchain might someday be used for! ↩
In addition to the following examples from Buddhism, we can also see this process playing out today in living color in debates in the Roman Catholic Church, which is thousands of years old, on topics such as homosexuality, sexual morality, abortion, and women’s rights. ↩
If you’re curious to learn more about this incident and the subsequent debate, in addition to this letter and the response letters that I quote heavily from below, see the archived dhammalight website for lots more information. ↩
Note that this mechanism is not without challenge and controversy. Automatically interpreting the silence of stakeholders as tacit agreement may work well for small groups and low stakes decisions but it tends not to work very well in the case of highly contentious decisions, especially when stakeholders are loath to raise concerns in public. One highly contentious issue that the Ethereum community has struggled to conclude through a process of rough consensus is progpow. ↩
The Roman Catholic Church, by contrast, is highly centralized. It’s nearly as old as Buddhism and it, too, has been wildly successful. It has been called “the world’s oldest and largest continuously functioning international institution” (Wikipedia). So it appears that either approach can work. ↩
For more on the evolution of the Theravadin canon, see The Selfless Mind, Personality, Consciousness and Nirvana in Early Buddhism by Peter Harvey, Canonicity by Jonathan Silk, and Buddhism Through Its Scriptures. ↩