# What Satoshi Did

Andrew M. Bailey and Craig Warmke[^1]

# Introduction

You may be familiar with Satoshi Nakamoto the software engineer. Renowned or reviled, he’s well known under that guise.[^2] You may also know Satoshi as a monetary designer – the creator of a new kind of digital asset – or even as an anti-authoritarian activist.[^3] In this article, we'll introduce you to Satoshi the social architect and founding father.[^4]

To situate Satoshi’s role as an architect and founder, we must inquire into the defining social features of Bitcoin which differentiate it from more familiar institutions and organizations. What *kind* of social institution is Bitcoin? And how did Satoshi shape Bitcoin’s institutional profile? Answering these questions requires careful attention to Bitcoin’s early history. 

We argue that although Bitcoin began life as an engineered network with just one architect, Satoshi Nakamoto, it became a leaderless institution akin to a natural language such as English or Tagalog. In short, Bitcoin’s protocol is a convention that solves a coordination problem: how to format, send, verify, and publish messages that encode digital transactions of value. In doing so, Bitcoin creates and sustains both an institution and an associated *network good* — a good that grows in usefulness along with its user base.[^5] The institution is the Bitcoin network; the good is its native asset, bitcoin.

Bitcoin’s founder enabled this transition through his own provably costly support of the Bitcoin network and asset as well as his subsequent departure from the project. By adopting and then departing from Bitcoin, Satoshi also resolved the important game theoretic challenge of inspiring adoption in a new network, where the rewards of participation themselves require adoption — a tricky bootstrapping puzzle. In this way, he enabled Bitcoin to become a “feral” institution, as we’ll call it: at one point engineered but now released into the wild and with a life of its own.

In Section One, we introduce and defend our focus on Bitcoin’s social layer. In Sections Two and Three, we use primary sources to tell the story of Bitcoin’s early history. We then connect that story to a more theoretical understanding of Bitcoin’s institutional profile in Section Four. That task continues in Section Five, where Bitcoin’s founding story unites with game theory—a pairing whose metaphysical results we explain in Section Six. There, we unveil our account of what Bitcoin has become: a feral institution. We conclude by relating Satoshi’s role to more familiar founding stories from early American history.

# 1\. The Social Layer

As the native asset of the Bitcoin system, bitcoin inhabits a distributed ledger maintained by a network of computers. Accordingly, it is tempting to think of bitcoin as a monetary and technical innovation and not much else. But doing so obscures another realm of innovation: its “social layer,” or “layer 0.”

The Bitcoin protocol, the machines that run it, and the asset they host all reside within the same level—layer 1.[^6] But layer 1 is not Bitcoin’s only layer. The machines on the Bitcoin network don’t magically coordinate by themselves. Rather, node operators set them up and tell them what to do. Here we find Bitcoin’s layer 0—its people. 

A bevy of academics have been quick to claim that Bitcoin’s layer 0 is overlooked and poorly understood.[^7] Many further argue that closer scrutiny of the social layer reveals that Bitcoin is something other than advertised – not automated, not decentralized, not trust-minimized, and, overall, not good.[^8] Though they differ on details, these scholars perform the same two-step dance, outlined as follows:

*Step one*. On the box, bitcoin is a money machine. It is a distributed network of computers that collectively and algorithmically regulate a new kind of money. Where traditional currencies have makers, mediaries, and managers – trusted parties who create money, govern its flow, and take custody of it along the way – bitcoin has none. Bitcoin replaces people with machines. No rulers, just rules.

*Step two*. Inside the box, bitcoin is just people. It isn’t a money machine after all. Although it uses code and automation, people decide which code to execute, when, and how. Rather than replace people, the machines veil their influence like a curtain. Trusted parties remain.

Or so they say.

This two-step dance is half right and half wrong. In truth, Bitcoin does have a layer 0 composed of people. But it nevertheless lacks trusted parties. No one has it  “on a leash.” We will not further critique this particular two-step dance. We do that elsewhere and argue that the dance fails to establish its conclusions.[^9] 

Yet the dance highlights a vital topic: Bitcoin’s people. Layer 0 shapes and constrains the layers built atop it and resists easy duplication.[^10] Although anyone may execute forks of Bitcoin’s fully open source codebase, doing so does not create an equally powerful monetary network. Bitcoin also requires people. They deserve scholarly attention.

Layer 0 may be approached in a variety of ways. We could study how people use bitcoin as a speculative instrument, money, hedge, or safe-haven asset.[^11] We could examine bitcoin’s status as legal tender in El Salvador or the Central African Republic.[^12] Or we might study Bitcoin’s developers, miners, or behind-the-scenes influencers.[^13] These are all fine projects.

Here, we will instead analyze Bitcoin’s social architecture, beginning with how Satoshi bootstrapped Bitcoin and sparked a trend towards leaderlessness. Whereas some might have founded a monetary network to build an empire or secure ongoing influence, Satoshi departed and thereby engineered something quite special. Attention to bitcoin’s social layer does not, therefore, undermine its claims to leaderlessness. It vindicates them.[^14]

Satoshi imbued Bitcoin with a novel institutional profile that doesn’t fall squarely within traditional taxonomies; it is not a church, a Westphalian state, a corporation, or a family, for example. We will show that Bitcoin’s history puts it closer to the category of natural language. Unlike other natural languages, however, someone engineered it.

# 2\. Bootstrapping: Satoshi as Architect and Marketer

We cannot provide a theory of what Bitcoin has become without careful attention to what it has been. In this section, accordingly, we will examine some key events in Bitcoin’s early history and the role its founder, Satoshi Nakamoto, played in its earliest days. After providing a history lesson in this section and the next, we will situate this lesson within a theoretical framework.[^15]

The Bitcoin network currently processes hundreds of thousands of transactions daily. It boasts tens of thousands of computers for that process and uses more energy than several small countries to secure the ledger of those transactions. But behind all the transactions, nodes, and energy users, we find people—lots of them. Worldwide, these people contribute to a network with far fewer and far less concentrated pockets of power than exist in the hierarchical structures of traditional social institutions.

After Bitcoin was launched in 2009, the network of participants grew not only in size but in distribution and diffusion. More nodes. More holders. More users.[^16] These kinds of growth differ even if, in Bitcoin, they occurred more or less simultaneously. Centralized institutions typically retain their centralized structures as they grow—consider the United States, Roman Catholicism, and Apple Inc., for example. But Bitcoin grew both larger and *less* hierarchical over time.

Satoshi himself is largely responsible for both kinds of growth and in very different ways. Both speak to his qualities as a social architect. The first kind of growth—*bootstrapping*—helped Bitcoin grow from nothing into a nascent monetary network. The second kind of growth—*dispersal*—distributed power over a wider range of people, which not only increased the network’s censorship-resistance and permissionlessness but set a precedent for doing so. Neither kind of growth was fundamentally technical. Both were social.

Let’s consider Bitcoin’s bootstrapping first, which itself has two major components: marketing and greed.[^17]

## Marketing

By mid-August 2008, Satoshi had narrowed down the name for his new money to “netcoin” and “bitcoin.”[^18] On consecutive days, he registered domains for each through  AnonymousSpeech, a Japanese service for anonymously buying domain names. We'll likely never know precisely when, how, or why, Satoshi chose “bitcoin.” But we can make an educated guess that he felt it better suited the brand he wanted to cultivate for the software. Satoshi used a range of effective marketing techniques to help bootstrap Bitcoin into a monetary network.

Satoshi first contacted the few people most likely to understand and appreciate his innovation. Around the time he registered the domains above, Satoshi emailed both Adam Back and Wei Dai, researchers responsible for key building blocks in Bitcoin's design. The pretext for Wei Dai's email–to double-check the citation for Dai's “b-money” page–likely wasn’t the main purpose. The email includes a link to a draft of the whitepaper, along with an in-text abstract.[^19] These are unnecessary additions for someone who needs a simple thumbs up or down on a citation. But they were necessary additions if Satoshi had wanted to alert Dai and convert him into one of Bitcoin’s first users and evangelists. On January 10th, 2009, soon after the network launched, Satoshi pinged Dai once more to inform him that Bitcoin was running and that it seemed to achieve nearly all the goals in Dai’s b-money paper.[^20] Satoshi ends this email by forwarding Hal Finney’s flattering assessment about bitcoin’s relation to b-money.[^21] Just two days prior, Satoshi had also notified Finney about Bitcoin’s software release.[^22] 

In between the early emails and the network launch, Satoshi posted a short, nine-page paper online that describes how to create digital cash from a globally consistent database, all without trusted parties.[^23] Entrepreneur Jack Dorsey has called the Bitcoin Whitepaper “poetry” and “one of the most seminal works of computer science in the last 20-30 years.”[^24] Satoshi distributed the paper to the very small group of people in the world who he believed could appreciate its clarity, concision, and depth—the mathematicians, technologists, and cypherpunks on the cryptography mailing list.[^25] The Bitcoin Whitepaper was neither a governance document, like a constitution, nor a formal specification of the Bitcoin protocol. It was, as venture capitalist Nic Carter has observed, an “introduction and teaser.”[^26] Satoshi distributed the whitepaper as if he were sowing seeds for a new community. But he planted the seeds in the most fertile soil—the subscribers to the cryptography mailing list.[^27]

It worked. As discussions about Bitcoin blossomed, Satoshi spun out a Sourceforge Bitcoin forum on December 10th, 2008\. Eventually, and thanks to Satoshi’s early efforts at community building, the discussions outgrew the online meeting place. Not one year later, on November 22, 2009, Satoshi created yet another, more expansive forum—Bitcoin Talk—for Bitcoin’s fast-growing base of users and developers.[^28] 

In addition to creating forums for the nascent bitcoin community, Satoshi also created the first Bitcoin logos—an early and important part of Bitcoin’s brand marketing. The first logo featured a gold coin with a ‘BC’ in the middle, a letter presumably for each syllable in Bitcoin’s name. When Bitcoin was around 13 months old, Satoshi created a second logo, asking for the community’s input in the very Bitcoin Talk forum that he himself had created. This time, the logo featured a lone ‘B’ inside a gold coin with parallel vertical lines similar to those in the dollar’s symbol.[^29] Satoshi’s request for input on the logo was not out of character. From the initial announcement, through the Bitcoin Talk forum days, until his departure from the community in December 2010, Satoshi argued, speculated, inquired, and schemed with those who expressed an interest in Bitcoin. He served as Bitcoin’s “Chief Community Relations Officer.”

Part of Bitcoin’s initial appeal, of course, was that it ran on free and open-source software. Satoshi made this clear when he initially announced the software’s release in early 2009 and to the very people who would care—not only the cryptography mailing list but also the online forum for the P2P Foundation, a non-profit advocacy group for peer-to-peer technology.[^30] 

Notably, Satoshi tailored his announcements to each group. The subscribers to the cryptography mailing list included many cypherpunks and technologists who had sought a working implementation of digital cash for decades—it had become a kind of “holy grail.” In his message to them, Satoshi mentions “electronic cash” first and describes Bitcoin’s hard-coded economics.[^31] But the P2P Foundation forum post mentions that bitcoin is “P2P” first, does not mention its economic properties, and instead describes how bitcoin decentralizes the trust we ordinarily place in monetary authorities.[^32] Satoshi had a keen eye for his audiences and how to cater a message for maximum effect. In each case, Satoshi reveals how Bitcoin furthers his audience’s unifying principles. In this way, he also built bridges to attract new Bitcoin users. 

Through education, content marketing, brand marketing, and by personally playing the role of a community relations officer and help desk, Satoshi marketed his creation to several target audiences in order to bootstrap a new monetary network. Without minimizing the contributions of other early Bitcoiners, Satoshi himself often seemed to serve as a one-man marketing department.

Research on early open-source software (OSS) projects suggests that early marketing efforts play a significant role in their success. Interdisciplinary scholars Andrew Whitmore, Namjoo Choi, and Anna Arzrumtsyan observe:

While the size of the project community and the internal characteristics of the project are indeed the dominant driver in the long run, in the initial stages of growth the diffusion of the OSS project is primarily driven by external forces attracting users to the project such as advertising or marketing efforts.[^33]

What kinds of marketing proves useful? Whitmore et al. continue:

These advertising or marketing efforts can take a wide range of forms. They can include localized efforts such as posts to message boards announcing the genesis of the new project or the creation of project web sites such as the official eMule project website. Alternatively, they can take the form of massive volunteer led marketing campaigns as in the case of FireFox. It is also feasible that online OSS development forums like SourceForge.net offer some marketing benefits for OSS projects.[^34]

Satoshi marketed Bitcoin along these lines. Message boards? Check: Satoshi created Bitcoin Talk. Project web sites? Check: Satoshi created bitcoin.org. Volunteer-led marketing campaigns? Check: Satoshi schemed with users about getting more publicity in places like Slashdot. Using open-source software development forums like SourceForge.net? Check: that’s where Satoshi first led the Bitcoin development community to discuss the project. Satoshi seemed to know the playbook to attract users to an open-source project like Bitcoin. He was also adept at explaining bitcoin to political libertarians without exclusively tying the project to their sometimes idiosyncratic views.[^35]

Yet Bitcoin’s early growth has a second root cause beyond Satoshi’s role as “Marketer-in-Chief.” His announcement to the cryptography mailing list included a tantalizing detail:  readers would have noticed straightaway that bitcoin had a limited supply—under 21 million coins—and that network incentives had been structured so that earlier network participants could more easily earn coins. This, paired with the difficulty adjustment to the mining protocol, would help bootstrap Bitcoin in another way: the network would market itself through its own incentive structure. 

## Greed

Within days of Satoshi’s first announcement about Bitcoin’s software release, cryptographer and computer scientist Hal Finney wrote:[^36] 

As an amusing thought experiment, imagine that Bitcoin is successful and becomes the dominant payment system in use throughout the world. Then the total value of the currency should be equal to the total value of all the wealth in the world. Current estimates of total worldwide household wealth that I have found range from $100 trillion to $300 trillion. With 20 million coins, that gives each coin a value of about $10 million.

So the possibility of generating coins today with a few cents of compute time may be quite a good bet, with a payoff of something like 100 million to 1\! Even if the odds of Bitcoin succeeding to this degree are slim, are they really 100 million to one against? Something to think about...[^37]

Several days later, Satoshi echoed Finney:

It might make sense just to get some in case it catches on. If enough people think the same way, that becomes a self fulfilling prophecy. Once it gets bootstrapped, there are so many applications if you could effortlessly pay a few cents to a website as easily as dropping coins in a vending machine.[^38]

In other words, anyone could risk a little now for a potentially massive benefit later. These were direct appeals to the human desire for wealth. They worked.

In bitcoin’s earliest days, its native asset had no discernible value and no significant network effects. Bootstrapping past those barriers was no easy task, since adopting a new money is costly. As economist William Luther explains, “there is a systemic bias against monetary transition.”*[^39]* These costs are greater when a status quo money offers stronger network effects. In such a context, adopting a new money means that someone might choose to trade a more widely accepted money for a less widely accepted one. Think, for example, of a resident of the United States exchanging U.S. dollars for Turkish lira, which is legal tender in Turkey but not in the U.S. If he were to attempt to use the lira within the U.S., he would encounter friction and bias against the less widely accepted currency. Those transition costs are potentially even higher when switching would require leaving behind an established monetary asset for an asset that *might* become a widely accepted money sometime in the future. In such cases, some perceived benefit must attract new users by outweighing the disadvantages of unplugging capital from a stronger monetary network. Bitcoin attracted those users, in part, by offering them a chance at wealth.

Greed therefore played a role in the Bitcoin network’s early growth. An ever-increasing number of people bought bitcoin because they thought bitcoin may enrich them. bitcoin’s dramatic price movements drew frequent media coverage, which spread the word about its potential for wealth generation. These headline-grabbing price movements owe much to bitcoin’s core economic design – namely, the fixed issuance schedule with four-year halvings. In other words, Satoshi inserted a greed exploitation device into bitcoin’s economic design. 

The device first includes a fixed issuance schedule. The network issues bitcoin on a predetermined schedule, whether demand goes up, down, or sideways. Thus, bitcoin’s price lacks the moderating influence of an elastic supply. When demand goes up, the network cannot substantially increase the rate of bitcoin issuance. And when demand goes down, the network cannot substantially decrease the rate of issuance. Bitcoin’s relatively inelastic supply means that smaller changes in demand cause larger than usual changes in market price. 

The greed exploitation device also includes the halving schedule. This is a specific feature of bitcoin’s overall issuance schedule. According to the issuance schedule, the “mining reward,” or reward for solving the computational puzzle that creates new bitcoin, halves every 210,000 blocks, or once about every four years. Satoshi thus embedded a human-friendly geometric series into bitcoin’s issuance schedule. The original reward, in 2009, was 50 bitcoin (BTC) per block. After the first 210,000 blocks, 50% of the supply of bitcoin remained to be mined, and the mining reward was cut in half to 25 BTC. After the next 210,000 blocks, 25% of the supply remained. And so on. Because the amount of new BTC coming to market drops substantially and suddenly with every halving, each halving reduces the amount of marginal buying of new supply necessary to sustain or increase the price of bitcoin.

These two design choices—fixed issuance and the halving—served to induce price volatility.[^40] Price volatility, in turn, draws publicity. And, as we’ve seen, Satoshi cared about garnering publicity for his nascent network.

A Slashdot article in July 2010 was an early instance of media coverage of the network.[^41] Thanks to that publicity alone, the price of bitcoin increased tenfold in a matter of days, from $.008 to $.08. Within four months, the price had increased another sixfold. The Slashdot article was no lucky accident, however. Two weeks before its publication, Satoshi had schemed with bitcoin users on the Bitcoin Talk forum about how to label the latest software release so as to increase the chances of seeing a piece about bitcoin appear on Slashdot.[^42] 

The first Slashdot article concerned a software release and brought a significant number of new users to Bitcoin. Within months of Satoshi’s departure in December 2010, Bitcoin appeared in another Slashdot article—this time focused on bitcoin’s *price*:[^43] This second article explained how bitcoin’s price had, in two years, reached dollar parity from nothing. The bitcoin.org website buckled under the wave of curious newcomers, many of whom came from social media platform Twitter. Within four months, Wikileaks began to accept bitcoin donations in order to route around a banking blockade imposed by U.S. authorities. The price of bitcoin catapulted from $1 to $30. In the span of months, bitcoin buyers had caused one newsworthy spike in the bitcoin price (“dollar parity\!”), which in turn prompted another newsworthy spike to $30.

Bitcoin’s privacy features, censorship-resistance, and Wikileaks acceptance had initially drawn the attention of small groups of dedicated cryptographers and cypherpunks around the world. But bitcoin’s *price* had taken it mainstream. When bitcoin’s price crashed in June 2011, coverage of bitcoin’s price splashed across the websites of the *BBC,* the *Guardian*, *The* *Economist, Forbes, Fortune, NPR,* and *Vice*. Bitcoin’s price continues to command global attention today, particularly around the time of scheduled halvings. Through its first three halvings—a tiny sample size, to be sure—bitcoin has seen enormous volatility:

- From the first halving, in November 2012, to a year later, the bitcoin price rose from $30 to over $1,150.  
- From the second halving in July 2016, the price catapulted from just over $600 to almost $20,000 within 18 months.   
- From the third halving, in May 2020, the price rose from under $9,000 to almost $70,000 in about 18 months.

Since earlier halvings involve bigger drop-offs in the quantity of new bitcoin issued per block, we would expect earlier halvings to have more noticeable effects on price. Bitcoin’s price history confirms this hypothesis. 

Despite the fact that price stability is nearly universally invoked as a monetary virtue, we suspect that Satoshi deliberately architected the network incentives to generate price volatility in its early years—and thereby garner more publicity, which would in turn spur adoption. U.S. presidential elections, the Olympics, the World Cup—these headline events happen once every four years and draw enormous media attention. We have four-year cycles for sports and politics—why not for *money*?[^44] 

As each of the first halvings loomed on the horizon, bitcoin holders and speculators would likely engage in a predictable line of reasoning: 

*If bitcoin’s issuance will halve in a way that demand likely won’t, the price will likely increase in the weeks ahead. If news of the halving-induced price increase will likely draw even more buyers, the price will increase even more dramatically, thanks to fixed issuance. Therefore, we should buy bitcoin now.* 

This line of thought makes for a self-fulfilling prophecy, if enough people engage in it.[^45] It also kickstarts a bubble – a self-reinforcing, positive feedback loop that eventually tops out, runs out of steam, and pops. In *The Alchemy of Finance,* famed investor George Soros calls this self-reinforcing feedback loop “reflexivity:” “Rising prices often attract buyers and vice versa. How could self-reinforcing trends persist if supply and demand curves were independent of market prices?”[^46] We grant the point. At *any* time*,* bitcoin’s market price lies at the intersection of supply and demand. But *over* time, price itself affects demand. This feedback loop explains why bitcoin’s newsworthy volatility can lead to even more speculative trading of the asset. Bitcoin’s core economic design seems tailor-made to generate this sort of reflexivity, over and over.

Satoshi hoped that Bitcoin would become a freedom technology. In a 2008 email to the cryptography listserv, he wrote that “with cryptography . . . we can win a major battle in the arms race and gain a new territory of freedom for several years.”[^47] However, for Bitcoin to achieve this goal, it needed to become useful as money. To be useful as money, bitcoin would first need to acquire a network of early adopters. Some of these earliest adopters were in fact speculators. Thanks to bitcoin’s volatility-inducing design, more buying caused more volatility. Volatility drew headlines. Headlines drew more speculators and more people willing to use bitcoin as money.

In February 2024, developer Martti Malmia released a trove of Satoshi writings to the public for the first time. These emails show Satoshi not only as a careful engineer but as a marketer keen to harness human greed. In one exchange, Malmi and Satoshi discuss how they might use a fiat-to-bitcoin market to ensure an eventual non-zero price for bitcoin. Satoshi points out that bitcoin's verifiability could play an important role:

Since you know how many coins exist and how fast new ones are created... People would obviously be sceptical at first that the backing will hold up against an onslaught of people trying to get the free money, but as the competition raises the proof-of-work difficulty, it should become clear that bitcoins stay scarce. People will see that they can't just get all the bitcoins they want. It would establish a minimum value under bitcoins enabling them to be used for other purposes if, hopefully, other purposes are waiting for something to use.[^48]

In other words, Satoshi bet on the fact that people want scarce assets. Bitcoin's provable scarcity therefore plays nicely into human psychology. But Satoshi’s endgame wasn’t mere speculation: Satoshi hoped for bitcoin to find actual widespread adoption as money.

It’s difficult to imagine a more potent fuel for bootstrapping than greed. Greed enjoys wide distribution across the population and has powerful effects; few can resist the allure of wealth.[^49] In this way, Bitcoin’s economic design, tailored for headline-grabbing volatility early on, bore the seeds of Bitcoin’s tremendous growth. Yet greed could have hurt Bitcoin more than it helped if Satoshi had behaved differently. 

He could have sold his coins – those he had earned through mining, or special rewards minted out of thin air at no cost to himself, as has been the case with many other cryptocurrency tokens – whether to line his own pockets, or to pay for Bitcoin development or marketing. This might have accelerated Bitcoin’s early bootstrapping phase, but it would have come at a price: reducing Bitcoin’s credible claim to leaderlessness at layer 0\.

Bitcoin, to this day, has no marketing department, no insider rewards, and no protocol-supplied developer budget. Its users are its only marketers. As bitcoin’s user base grows in numbers and diffusion, so too does its informal marketing office, which functions as a feedback loop that trends towards diffusion. So, although bitcoin is not without advocates and network stewards, they are no more and no less influential than Bitcoin’s other users. Satoshi’s approach to bootstrapping – relying on the idealistic motivations of cypherpunks and the greed of all – ensured that Bitcoin would never be captured by a foundation, company, or government.

Satoshi didn’t stop there. He took one more important step towards leaderlessness at the social layer. He left.

# 3\. Departure: Satoshi as Founding Father

Many of us have fond memories of riding a bike without training wheels for the first time. In one moment, a parent steadies us as we pedal. Then, they let go and we ride by ourselves—at least for a moment. As young bicyclists, we had the potential to stay upright. But we couldn’t manifest the potential until the parental scaffolding withdrew. 

The story of Bitcoin is similar. Satoshi designed a protocol for peer-to-peer electronic cash. For Bitcoin to fulfill its promise, peers would need to populate the network. Through marketing and publicity, they did. But then something happened that few anticipated. Satoshi, founder of the Bitcoin project, effectively disappeared.[^50]

## Preparation

Before Satoshi resigned as Bitcoin’s unofficial leader, he minimized his own leadership footprint in several ways. He waited two months for the world to digest the Bitcoin Whitepaper before open sourcing the software that fulfilled its promises. He pursued no patents, claimed no equity, and, unlike the vast majority of other cryptocurrency projects, held no public or private sale of tokens prior to the network's launch. Instead, Satoshi included a newspaper headline in the first block of Bitcoin transactions to prove that he hadn’t given himself bitcoin before anyone else had the chance to earn it for themselves. He also refused to claim any glory—fame, awards, appointments, and respect—by keeping himself anonymous.

Satoshi mined bitcoin like any other miner. By protecting the fledgling network as an early miner, Satoshi earned what at the time of this publication amounts to tens of billions of U.S. dollars worth of bitcoin. But from the network’s transparent design, we have good evidence that Satoshi (i) mined bitcoin with an identifiable marker so that the world would know how much he had; (ii) mined much fewer bitcoin than he could have; and (iii) has never spent any of the bitcoin mined under the guise of that marker.[^51] 

There is no clear line between Satoshi’s preparation to exit and the exit itself. Before he left permanently, he would have to divest himself of all the positions of influence he had gained. This would take time. His divestment unfolded in orderly steps. It began soon after Bitcoin launched by way of delegating tasks and relinquishing power to others.

## Exit

For Satoshi to depart without undermining all he had built, others would have to take up the mantle of leadership in the Bitcoin community. But Satoshi had many mantles. He would need to hand over responsibility for the development and maintenance of Bitcoin’s protocol and the bitcoin.org website and discussion forum. He would also need to ensure that the community had the wherewithal to market and educate about Bitcoin effectively.[^52] 

By July 2010, Satoshi already had help with the bitcoin.org website from Martti Malmi, also known under the pseudonym “Sirius.”[^53] Then, in his last update to the bitcoin.org website on December 10th, 2010, Satoshi would include the names of a few other contacts besides his own.[^54] Around this time, Satoshi gave the domain to people “separate from the Bitcoin developers, to spread responsibility and prevent any one person or group from easily gaining control over the Bitcoin project.”[^55] Malmi also took responsibility for the Bitcoin Talk forum, with help from Michael Marquardt (also known as “theymos”). 

A few days after he updated bitcoin.org for the last time, Satoshi made his last update to the Bitcoin protocol. In this release, Satoshi quietly removed his copyright claim and assigned it to the amorphous mass of “Bitcoin developers”—a group that anyone could join by voluntarily developing Bitcoin’s protocol.[^56] For this passing of the baton, Satoshi already needed to have cultivated a small community of Bitcoin developers. Sirius, for example, had already enjoyed commit access from August 2009\. Laszlo Hanyecz had similar access by August 2010, and Gavin Andresen by October 2010\. Within months, Satoshi would apparently anoint Gavin Andresen as a lead maintainer for the software.[^57] 

In two separate emails to Andresen in April 2011, Satoshi first passed along some final wisdom for distributing credit to developers and then gave Andresen “a copy of the cryptographic key to Bitcoin’s alert system, one that effectively gave Andresen sole control over security notifications.”[^58] And that was that. 

By this time, Bitcoin had matured in two ways. First, it had a bevy of developers, defenders, and evangelists. Second, Bitcoin had garnered headlines through its price volatility. The “dollar parity” Slashdot article had appeared two months earlier. And when Satoshi sent his last emails, bitcoin’s price had blown past $1.50. Bitcoin’s greed exploitation device had finally kicked into higher gear. This incentivized more people to join the network as miners, which made the network more secure than ever.

Bitcoin could ride on its own. 

## Potentiality and Actuality 

By design, Bitcoin’s protocol is leaderless. The protocol confers no special powers to Satoshi, or to anyone else, or to their computers. No master key controls the network. Soon after the network launch, when others began to run the software with Satoshi, Bitcoin was already formally leaderless – at the layer 1 level, that is. In this way, Bitcoin stands apart from institutions that have formal leadership posts like kings, presidents, popes, and ombudspersons. But formal leaderlessness is not enough to set Bitcoin apart from other social institutions. Layer 0—the aforementioned social layer—matters just as much as layer 1, the base protocol layer. Social leaders operating on layer 0 can still exert informal control and influence.

Bitcoin bootstrapped, in large part, through the greed of its users and unsupported by a marketing budget or developer fund or the aid of any corporation or foundation. Anyone could run a node to validate Bitcoin transactions. Anyone could transact on the network. Anyone could mine bitcoin. Anyone could modify the code and execute it as they wished. Bitcoin’s creator, furthermore, didn’t freely mint money for himself or other insiders. There is exactly one way to mint new bitcoin: to complete proofs of work – that is, to burn electricity and processor cycles in the discovery of new blocks and to claim the accompanying reward of newly minted bitcoin. No exceptions. So Satoshi had to pay for his bitcoin, just like anyone else. He did not make magic beans out of thin air and hawk them at the local market. He bought them from nature, and the price was energy. So although bitcoin had early adopters, it has no “insiders.”

These considerations together make a formidable formal case for Bitcoin’s leaderlessness. But the case would be hollow if Satoshi had stuck around. Imagine Tesla, Inc., were its CEO to formally resign but still speak constantly in public about the brand, roadmaps for future products, software updates, and so on. Bitcoin could have evolved under such oversight had Satoshi retained his social position.

Decentralization comes in degrees. By late 2010, Bitcoin already had serious potential for a high degree of decentralization—much more than any previous electronic payments system. But this potential was somewhat stymied because Satoshi still enjoyed high levels of influence over Bitcoin—not formally, over the protocol, but socially, over its network of users. This doesn’t mean that Satoshi actually used these powers to their fullest extent. But to inhibit bitcoin’s decentralization, Satoshi simply had to *have* these powers, however latent. Some of these implicit powers were arguably passed on to Gavin Andreson, the second lead developer in Bitcoin’s history; other powers dissipated into the night and now belong to no one in particular.

In contemporary metaphysics, philosophers say that a disposition or power is *masked* when its manifestation is blocked by some other condition — think of a fragile glass surrounded by packing material.[^59] Satoshi’s own presence masked the network’s nascent disposition towards a high degree of decentralization. The disposition really was there, given the protocol’s design and the size of the network running it. But the disposition couldn’t manifest until Satoshi left. In walking away, Satoshi activated Bitcoin’s layer 1 potential to be something much more interesting: he enabled a shift towards leaderlessness at Bitcoin’s social layer.

One striking effect of Satoshi’s pseudonymity – further enabled by his timely exit – is that Satoshi has himself become an increasingly abstract and amorphous figure. In time, few will remain who had any direct contact with him at all. Does Satoshi look like you? Or us? Is he a woman or a man? White or black? Is he a group of people or an individual? No one knows for sure. It is not for nothing that a statue of Satoshi erected in Hungary is vague and faceless, giving form to the slogan, “We are all Satoshi.”[^60]

# 4\. Bitcoin the Institution

With this review of Bitcoin’s early history behind us, we can work through some of its consequences and their import. We will first introduce some conceptual tools for understanding social groups and show that bitcoin is one solution to a coordination problem. This insight paves the way for subsequent sections, which will connect Satoshi’s history with game theory, a branch of economics that studies strategic interaction. Finally, with the disciplinary tools of metaphysics, a branch of philosophy that studies ultimate reality,[^61] we will argue that Bitcoin’s network of users is, as it turns out, a rather distinctive kind of thing.

## Conventions and the Institutions They Make

People cooperate, but rarely by accident. Our interactive behavior within a given domain is often coordinated or governed by rules that we are motivated to follow. These rules are known as *conventions*.[^62] Some conventions are arbitrary. In some such cases, many possible conventions in a target domain would benefit everyone, but unlocking these benefits requires that everyone follows one convention exclusively. A famous example of this phenomenon is picking a side of the road on which to drive. Arbitrary or not, conventions bring people together and structure their behavior. A group of people so structured is an *institution*.[^63]

Institutions are as varied as their conventional rules: they can be big or small, simple or complex, intentionally created or forged bottom-up through evolutionary processes. Familiar institutions include churches, Westphalian states, corporations, families, and natural languages.

In the landscape of institutions, where does Bitcoin lie?[^64] What kind of institution is Bitcoin? Satoshi’s approach to bootstrapping – and his departure – help us answer this question.

The history documented in the previous two sections suggests that Bitcoin became a novel kind of institution. Though Bitcoin was indeed engineered, it resembles a language more than a corporation, state, family, or church. Before we argue this point, however, let’s uncover some key points of similarity between Bitcoin’s network and natural languages such as English. The first point should be familiar to students of monetary theory and is not unique to Bitcoin. The second draws on Bitcoin’s distinctive founding story.

That Bitcoin resembles natural languages more than states, churches, or corporations is somewhat unsurprising. The latter kinds of institutions often have formal offices occupied by leaders who explicitly exercise power – popes, presidents, chancellors, and so on. Their layer 1s are formally leader-ful. Bitcoin’s layer 1, by contrast, is formally leaderless. It has no president within its protocol nor any master node in its network. Similarly, English has no prime minister, and Tagalog has no CEO. And yet English and Tagalog are institutions with people who follow their respective conventions.

## Bitcoin as Monetary Coordination 

Bitcoin’s protocol is a collection of rules about how to format, send, verify, and publish messages. Bitcoin nodes, at the direction of their users, implement the protocol. People run nodes to follow Bitcoin’s rules and make coordination possible.

Bitcoin’s protocol is a convention. It enables a solution to a coordination problem. In doing so, it creates and sustains both an institution and an associated *network good* — a good that grows in usefulness along with its user base.[^65] The institution is Bitcoin’s network. The good is its native asset, bitcoin. Languages exhibit a similar structure.[^66] Further reflection on language use will help us home in on the kind of institution Bitcoin’s network has become.

Suppose Casey wants Bennett’s chair and has nothing but corn to trade. Bennett doesn’t want corn—he wants pigs.[^67] To get Bennett’s chair, then, Casey needs to flip her corn into pigs first. But for her to do that, someone with pigs must want corn, or else Casey is out of luck. Trading without money requires what economists call a “coincidence of wants”—a network of people who desire each other’s available goods and services. Since what we want to acquire rarely aligns with what our neighbors want to offload, a society that relies on a coincidence of wants for trade won’t trade very much.

Money greases the skids of exchange. Without money, trade requires the kind of perfect match of desired goods and services that rarely occurs in practice. But if there’s something that most people would trade for—whether it be shells, gold, or dollars—then we have a coincidence of wants after all. Suppose everyone accepts gold. Casey and Bennet could then sell their goods for gold. Once Casey has her gold, she can then easily trade that gold for Bennett’s chair. Problem solved. Gold, in this example, serves as a *medium of exchange*—something that sellers accept in exchange for their goods and services because they can reliably exchange it later for goods and services.[^68]

Gold is not the only solution to this puzzle. The problem of coincident wants is a coordination game with multiple solutions. Today, dollars are a solution of choice for many around the world. And for a growing number of people that solution is bitcoin. As a result, bitcoin is *a* money: it is one possible monetary convention. Bitcoin’s protocol provides the detailed rules of that convention, and the group of people who follow it are an institution – distributed, unruly, and heterogenous, but united in this one respect.[^69]

Languages work similarly. They solve a coordination problem, not for the communication of value, but of information. Words that mean the same thing to different people grease the skids of information exchange. With them, we can say quite a bit to each other. Communicating with words requires mutual understanding – a coincidence of *meanings*. Human beings in communities have generally discovered that we are better off when we coordinate and follow the same linguistic conventions. These conventions constitute a protocol, and people who follow it constitute an institution. English does not uniquely solve the problem of how to coordinate meaning, of course. The problem of coincident meanings is a coordination game with many solutions. But the rules of English provide a convention, and its followers  generate a linguistic institution.

Monies, like languages, involve network goods. The internet – and many services that run atop it – are also network goods. So are telephones. If you’re the first person to hook up to the internet, you receive zero benefit until someone else decides to do the same. If you have a telephone and nobody else does, you likewise receive no benefit. But the more people who use phones, the more each user gains by using a phone, as well. Similarly, more language users make a language more useful. Bitcoin and other monies also grow more useful as more people use them.

Bitcoin is thus like natural languages in two important ways: First, it solves a coordination problem by establishing a convention (protocol), and its convention introduces and sustains a network good. Second, participants follow the convention in order to benefit from that good.[^70] They’re enticed to bring new participants to the convention in large part because it makes that good more useful to them. Every new English speaker benefits the rest, if only slightly – there is now one more person with whom they can chat. Similarly, bitcoin’s usefulness grows whenever another person accepts it in exchange for something of value – there is now one more person with whom they can transact. In each case, the result is an institution: a group of people whose interaction is structured by the constitutive conventions or rules they are motivated to follow.

Other familiar institutions induce participation in their defining conventions somewhat differently. Westphalian states, for example, deploy coercion – follow these rules, they say, or else.[^71] Of course, any resulting goods needn’t be network goods, in the aforementioned sense. They might be anti-network goods, providing utility to participants only when the network remains small. Religious cults often work this way. In any case, an institution’s defining convention may induce participation in a variety of ways.

Bitcoin and natural languages involve voluntary cooperation. In both cases, a network good attracts people to follow the defining conventions. Why might you learn Tagalog? To exchange information – to communicate – with others who know Tagalog. Why might you use bitcoin? To exchange value – to transact – with others who use bitcoin. More specifically, you might use bitcoin to route around authorities who might otherwise censor your transactions. For unlike other monies, bitcoin unusually overcomes corporate and state blockades.[^72]

To recap: Bitcoin’s network is an institution with a defining convention, and its users follow that convention to partake in its associated network good, its native asset. Bitcoin motivates future users with the benefits of coordinating with those who already play by its rules. So far, so good. But questions remain: How, exactly, did the first users come to adopt Bitcoin’s defining conventions when there was no network to speak of? And how did Satoshi’s work bear on all this? We’ll turn to these questions now using the disciplinary tools of game theory.

# 5\. Money Games

Some social theorists and historically-minded social philosophers use concepts and tools from thinkers such as Adam Smith, Karl Marx, Max Weber, or Émile Durkheim to understand institutions, their constitutive conventions, and their workings and consequences. There is wisdom in that approach. But in this Section, we propose to draw instead from elementary game theory in a way more reminiscent of the analytic tradition in philosophy.[^73] Some ideas from game theory will deepen our understanding of the challenges involved in launching a new money and how Satoshi resolved them.[^74]

## Money Games for Two 

Imagine two players, A and B, faced with two strategies in a single-round game: either to offer and accept *m* as a money in exchange for things they find valuable, or merely to barter (engage in direct exchange of valuables). If both A and B barter, they’ll both benefit a bit. If both A and B opt to use *m* as money, they’ll both benefit far more. But if they fail to cooperate at all—if one tries to barter with the other, who is trying to make a trade using *m* as money—they’ll make no trades at all and thus fail to gain any of the benefits of trade. 

The payoff matrix for these different strategies might look something like Figure 1\. Here and in what follows, rows and columns list player strategies, and boxes at their intersection show outcomes, with payoffs for A and B listed in each. Note that bigger numbers signify better outcomes. We will also refer to the so-called Nash equilibrium, which holds when no player can benefit by unilaterally altering their strategy.

|  | B uses *m* as money | B barters only |
| :---- | :---- | :---- |
| A uses *m* as money | A: 2 B: 2 | A: 0 B: 0 |
| A barters only | A: 0 B: 0 | A: 1 B: 1 |

*Figure 1\. Money or barter game.*

This game is easy to play and easy to win. To cooperate is to win; to deviate is to lose. There are two equilibria (either using money together *or* bartering together), unilateral departure from which will involve loss for the departing player. But the two Nash equilibria differ. One, the upper-left-hand box, is *Pareto optimal*: adoption of any alternative strategy would make at least one player worse off. For this reason, although using *m* as money is a convention in our sense, it is not an arbitrary one. Using money produces better outcomes for everyone. When both players know this, they will reason towards adopting *m* and will not easily deviate from that convention. 

Once discovered and occupied, the upper-left-hand box is a mighty fortress – *ein feste burg*. 

Satoshi hoped to launch a new money despite the existence of other well-established monies. Modifying our simple game will illustrate a difficulty that he thereby encountered. Suppose A and B have converged on using money in trade and on *m,* in particular. Barter is, effectively, no longer an option. But now some other money *m\** is newly available and every bit as useful to the players as *m*. The new payoff matrix is shown in Figure 2\.

|  | B uses *m* as money | B uses *m\** as money |
| :---- | :---- | :---- |
| A uses *m* as money | A: 2 B: 2 | A: 0 B: 0 |
| A uses *m\** as money | A: 0 B: 0 | A: 2 B: 2 |

*Figure 2\. Two money game.*

The payoffs are the same as those in another classic game – “left or right?” — where players must decide whether to drive on the left or right side of the road. Adopting some convention or other is optimal, no *particular* convention is optimal, and converging on either convention is a Nash equilibrium. Note also that conventions are *sticky*. Once you know that others drive on the left, for instance, you have a strong reason to follow suit. So in a multi-round game, attention to other players’ past moves can play a decisive role in selecting among strategies otherwise on a par.

Selecting monies is no different. In a multi-round game where A and B have already benefited from converging on *m*, switching to *m\** will prove difficult, even if *m\** is otherwise no worse than *m.* Arbitrary conventions do not easily displace one another, all else being equal. 

In fact, something even stronger holds. Consider a multi-round variation on our game, as in Figure 3, where *m\** is a *better* money than *m* in some respect — easier to move around or cheaper to verify as non-counterfeit, say. Given such advantages, both players would benefit from adopting *m\** over *m* as their money.

|  | B uses *m* as money | B uses *m\** as money |
| :---- | :---- | :---- |
| A uses *m* as money | A: 2 B: 2 | A: 0 B: 0 |
| A uses *m\** as money | A: 0 B: 0 | A: 3 B: 3 |

*Figure 3\. Better money game.*

Our “better money” game (Figure 3\) resembles the “money or barter” game (Figure 1). But despite the improvements *m\** promises over *m*, A and B need not select it. For although doing so would benefit all more than any alternative, attempted convergence carries the risk of failed convergence. In other words, when A and B each know the other player has previously selected *m*, A and B have a reason to stick with *m*, especially since each player wants to avoid being stranded on a monetary island, as in the lower-left and upper-right boxes. These multi-round considerations further strengthen *m*’s grip on the players.

The game theoretical factors discussed thus far can help explain why players converge on a *uniquely* *good* convention, one defined as optimal or as an equilibrium of some sort. But they do not alone explain the persistent power of conventions. Nor do they suffice to explain why players converge on one of two equally good conventions (as in the “two money” game), or on a less-good convention (as in the “better money” game). To do that, we must appeal to other elements of human psychology. These are not far off, luckily, and can be invoked with one useful word — *precedent*. Philosopher David Lewis explains:

We can explain the force of precedent just as we explained the force of salience. Indeed, precedent is merely the source of one important kind of salience: conspicuous uniqueness of an equilibrium because we reached it last time. We may tend to repeat the action that succeeded before if we have no strong reason to do otherwise. Whether any of us really has this tendency, we may somewhat expect each other to have it, or expect each other to expect each other to have it, and so on — that is, we may have first- and higher-order expectations that the others will do their parts of the old coordination equilibrium, unless they have reason to act otherwise. Each one’s expectation that the others will do their parts, strengthened perhaps by replication using his higher-order expectations, gives him some reason to do his own part. And if his original expectations of some order or other were strong enough, he will have a decisive reason to do his part. So he will do it.[^75]

Human societies establish precedents about what monies to use under specific circumstances. So the theory thus far illustrates what is already known: making a new money is hard. Convincing other people to use it, and so to act against precedent, is even harder.

We have already noted some familiar solutions to these problems — tools for manipulating precedent, if you will. To require that taxes be paid in *m* is, for example, to place a thumb on the scale and alter the payoffs in favor of *m*. Players who adopt *m* as money in such a scenario — by, for example, accepting enough *m* to pay taxes — enjoy the additional benefit of avoiding jail. The state may thus play a role in money selection, though it doesn’t follow from this that the role is necessary or decisive.[^76]

Another familiar solution involves selecting a money with significant non-monetary uses—money that can be eaten, enjoyed for its beauty, smoked, or used in electronics, say. We can think of a persistent non-monetary user of a money *m\** as a third player always willing to pay a positive price for *m\**, transforming B’s payoff in the upper-right-hand box from zero to something higher than zero. Such a player would protect others from suffering the worst case scenario in selecting *m\**. 

Satoshi knew that bitcoin, by nature, would lack significant non-monetary uses.[^77] Nor could he mandate adoption for monetary uses. Bitcoin would need to attract users in some other way.

## Money Games for Three 

Imagine, now, a multi-round game in which, in previous rounds, two of the players A and B have already converged on *m*. Players A and B save in *m*, denominate contracts in *m*, accept *m* in exchange for things they value, are disposed to offer *m* in exchange for things they value, and so on. Their money is *m*. A and B have played our simple games above, that is, and have converged on a relatively sticky cooperative strategy.

No surprises, so far; this is how conventions in equilibrium work.

Now imagine that a third player enters the arena and wishes to convince A and B to converge on another money. How might that happen? What would it take for A or B to make the switch? To carry out this feat, our new entrant C will have to arrange things so that A or B will find a benefit to switching to the new money that equals or exceeds the benefits of sticking with *m* as their money of choice.[^78] C will therefore want to ensure that the new money improves on the old in some respect. Presumably, C will want to benefit from any new arrangement somehow, too; otherwise, C would be better off simply joining A and B in using *m*. But the new money’s technical superiority alone will not move the needle. Somehow, C must overcome the stickiness of conventions in equilibrium as seen in our “two money” and “better money” games.

How in the world can C pull this off?

Our question is no mere game theory puzzle, nor is it specific to money. It finds expression in all domains involving network goods. Children who invent new languages don’t easily persuade others to learn them. Adults who invent new languages discover the same. Ludwik Lejzer Zamenhof, creator of Esperanto, learned this the hard way: when networks and their associated network goods are already entrenched, people don’t easily adopt alternatives, even when those alternatives boast technical superiority.[^79] 

The vocabulary of game theory helps illustrate this point: converging on a convention is even less likely after convergence on another. Figure 4 illustrates that equilibria of this kind can display significant stability and resilience, even in the face of superior alternatives. With three players, the payoff matrix is more complex. But the main points remain simple: C has unleashed a new money *m\**, which represents a significant improvement over *m*. As before, though, it does not follow that A and B have a sufficiently strong reason to switch to *m\**.

|  |  | B |  |  |  |
| ----- | :---- | ----- | :---- | :---- | :---- |
|  |  | Use *m* |  | Use *m\** |  |
|  |  | C |  | C |  |
|  |  | Use *m* | Use *m\** | Use *m* | Use *m\** |
| A | Use *m* | A: 3 B: 3 C: 3 | A: 2 B: 0 C: 2 | A: 2 B: 0 C: 2 | A: 0 B: 4 C: 4 |
|  | Use *m\** | A: 0 B: 2 C: 2 | A: 4 B: 0 C: 4 | A: 4 B: 4 C: 0 | A: 5 B: 5 C: 5 |

*Figure 4\. Three-player better money game.*

Let’s work through a few details before connecting any of this to Satoshi or bitcoin.

First, the three two-player games described earlier — “money or barter” (Figure 1), “two money” (Figure 2), and “better money” (Figure 3\) — are all games of pure cooperation.[^80] Payoffs for players are equal in every outcome. Our three-player better money game, by contrast, is a game of conflict since payoffs sometimes differ for different players. But the game isn’t one of *pure* conflict since payoffs also sometimes converge.

Second, *m* and *m\** are both network goods in that benefits to users grow with use. So, when only two of our three players adopt *m*, for example, adopters realize two units of benefit, but when all three converge on *m*, adopters enjoy three such units.

Third, *m\** is superior to *m* in the sense that it rewards multiple adopters more than *m* does. Exactly two players with *m* enjoy three units of benefit, whereas two players who adopt *m\** enjoy four units apiece. With *m,* three players net four units. Yet with *m\*,* they net five. As network goods, though, neither *m* nor *m\** benefit lone users. 

Fourth, A and B’s knowledge of past games gives them some reason to stick with *m*, despite the benefits of switching. Herein lies the rub. The stickiness we observed in our two-player “better money” game holds sway here, too. Introducing a third player makes no difference. Not yet, at least. Precedents have power. Past successes with *m* gain momentum, and the risk of going solo while reaping no benefits whatever repels players from *m\**.[^81]

But now imagine a new round where player C takes a serious risk. Player C has, at significant expense to themselves, begun to acquire *m\** with every indication that she’ll continue to do so. C has, furthermore, taken great pains to advertise the existence and advantages of *m\**. These activities introduce new information about various outcomes. They send a signal.[^82] When that signal is sufficiently strong, the game’s other players needn’t worry that going solo with *m\** will reap no benefits. C has effectively given A and B reasons to update their mental models. They can now either screen off certain possibilities as less likely and thus less salient,[^83] or they can screen in the other possibilities as more likely and more salient. Either way, a new game is afoot.

|  |  | B |  |  |  |
| ----- | :---- | ----- | :---- | :---- | :---- |
|  |  | Use *m* |  | Use *m\** |  |
|  |  | C |  | C |  |
|  |  | Use *m* | Use *m\** | Use *m* | Use *m\** |
| A | Use *m* |  | A: 2 B: 0 C: 2 |  | A: 0 B: 4 C: 4 |
|  | Use *m\** |  | A: 4 B: 0 C: 4 |  | A: 5 B: 5 C: 5 |

*Figure 5\. Three-player better money game, updated.*

Now things have gotten interesting. Sticking with *m* is no longer the most obvious strategy for A and B, even with their knowledge of *m*’s previous successes. Given their knowledge of C’s past and likely future commitment to *m*\*, A and B accordingly discount the risk of going solo with *m\*.* In the light of the discounted risk of choosing *m\**, and their knowledge that the risk is also discounted for other players, they can henceforth ignore certain potential undesirable outcomes—now grayed out—and reason in light of the choices that remain. Two results then fall into place. First, the lower-right-hand box in which everyone chooses *m\** now constitutes a unique Nash equilibrium; no player can do better by adopting *m* instead of *m\**. Second, this same box is also optimal in the sense that it provides the uniquely best outcome for every player.

Does C’s strategy in past rounds provide a *decisive* reason for others to think she’ll continue to use *m\** in the future? Perhaps not—a reason’s strength often comes in degrees. We might therefore depict varying degrees of confidence in C’s commitment to *m\** on a scale from full opacity (certainty) to translucency (some confidence) to full transparency (no confidence whatsoever). To the extent that C’s signals for *m*\* are credible and evince a standing preference, the shading will become more opaque and, for the other players, the way forward more certain.

Optimal equilibria along these lines are, as we’ve already seen, potent. They attract and retain adoption. With that thought in mind, we can see how C could achieve the target feat by making certain moves and signaling that she’ll continue to make them. C’s unilateral behavior in early rounds of the game can give new information — not about the payoffs within each box, which remain unchanged — but rather about which boxes contain less probable outcomes and therefore may be ignored. In other words, C’s early unilateral behavior establishes a *precedent* that informs future decisions for all players in the game.

Let us replace our variables with names and see how this all applies to bitcoin. Bitcoin is *m\**. The mighty U.S. dollar is *m*. Satoshi is C, and the rest of us are A and B. Some of the assumptions from our three-player game will hold in the case of bitcoin and the dollar. Both bitcoin and the dollar are pure network goods, for example — useful only if widely used. 

But is bitcoin superior to the dollar in the way and to the degree envisioned here? Do pairs who adopt bitcoin enjoy twice the benefits they’d enjoy from selecting the dollar, instead? That seems optimistic. The dollar has advantages of its own, including extraordinary price stability and a network of billions of users. So, when assigning payoffs to bitcoin adopters, modesty seems like the best policy. But for our purposes, a fairly weak assumption will do: bitcoin use provides *some* benefits to its users in comparison to the dollar; and, as a network good, its usefulness scales up with use.

## The Satoshi Precedent

Can we map what Satoshi did onto C’s above strategy? To some extent, yes. It will be helpful to break things into four steps: invention, advertising, use, and departure.

*Invention*. Satoshi created a new column or row in other players’ payoff matrices by furnishing a monetary alternative for potential adoption and endowing it with automated payoffs for adopters. Bitcoin’s formally leaderless design enables use without the permission of potentially overbearing authorities, for example – that’s a benefit. The introduction of peer-to-peer digital money therefore changed the parameters of the money choice game.

*Advertising*. Satoshi proclaimed the benefits of bitcoin adoption and educated other players about bitcoin’s potential payoffs. However, his advertising campaign did not primarily focus on bitcoin, the asset. You may recall that Satoshi never sold any of his own coins, as far as we know. His primary focus was, more exactly, a new game theoretical strategy. Of course, advertising a strategy involves saying why people might benefit from undertaking it – hence, our earlier discussion of greed. But Satoshi relied on the greed of potential new users to bootstrap the network, not on his own.

*Use*. Satoshi signaled, in a provably costly way, that he valued bitcoin. As discussed above, Satoshi did not simply mint new bitcoins at no cost to himself; instead, he bought them from nature. Satoshi traded dollars (or some other money) for electricity and computer chips and then used them to acquire bitcoin. Since the proofs of work required to produce new bitcoin are indeed proofs — there’s no forging these — Satoshi sent credible signals to the network and to the world that he was acquiring bitcoin at a cost. And because Satoshi mined bitcoin for so long, his behavior also evinced a longstanding preference to exchange other monies for bitcoin. This was no mere one-round trick: Satoshi played enough rounds for others to gain confidence in the game he was playing and to join in.

*Departure*. Once the network became self-sufficient, with participants who were themselves sending credible signals of their standing preference for bitcoin, Satoshi left. In other words, inventing bitcoin sent a technical signal: “here is what bitcoin could be.” Leaving bitcoin sent a social one: “it’s ready.” Bitcoin’s technical features endow it with the potential to be leaderless. Satoshi’s departure transformed that potency into actuality.

The use and departure points deserve further emphasis. For invention and marketing cannot by themselves eat into the strong network effects of incumbent monies. Coordination is also necessary. As noted above, most monies have gained traction, in part, by taxes or the non-monetary uses of the money itself. Bitcoin relied on neither. In the absence of these bootstrapping mechanisms, how could a new money gain a network of coordinating participants? Here is how William Luther puts the problem:

Less attention should be given to whether an item has intrinsic worth — though non-monetary uses might contribute to a successful launch. More attention should be given to the role of coordination. How are shared beliefs proposed, established, and perpetuated? What factors encourage or inhibit coordination?[^84]

We answer Luther’s queries with careful attention to bitcoin’s history, properly situated within the theoretical framework at hand. Satoshi inspired further coordination first by using bitcoin and then by leaving. This is how promise turned into precedent.[^85]

In summary, then, elementary game theory shows that creating a new money involves serious challenges, especially for actors without the full power of the state at their disposal to compel use of the new currency. Bitcoin’s early history shows how it met those challenges. Let’s take a closer look.

## Why Bitcoin Adoption is Real, but Incomplete

Bitcoin adoption has not and may never reach the lower-right-hand outcome illustrated in Figure 5\. Not everyone uses bitcoin as money, and with good reason. Some may have little need or desire for peer-to-peer digital money. Adopting bitcoin might even harm some. So the actual situation is considerably more complicated than our simple three-player game might suggest, involving many more players and more disparate payoffs. Perhaps a more realistic approach to modeling Satoshi’s strategy therefore does not presume that bitcoin provides an obvious — and obviously optimal — equilibrium for all players. Things just aren’t that easy.

Our theoretical framework remains helpful, however, because we can modify it to more closely capture the salient features of actual money selection games. Let’s construct another three-player game.

Imagine that Alice is a rather normal dollar user and in no particular need of bitcoin’s censorship-resistance or its other seemingly exotic properties. Indeed, Alice finds bitcoin inconvenient and distasteful, no matter how many others are using it. We’ll model this so that Alice benefits minimally across the outcomes in which she uses bitcoin. Boris, by contrast, is a Russian dissident frustrated by authorities who block access to his dollar-denominated bank accounts. For Boris, it’s bitcoin or nothing.[^86] So Boris shows an outcome of zero in every case where he tries to use dollars. Our third player — Satoshi — has, as before, already revealed his preference for bitcoin. So we shade as unlikely or less salient those outcomes where he selects dollars. Satoshi has therefore established a bitcoin-favoring precedent. We also hold fixed a background condition where at least one other player, not explicitly depicted in our game, also selects dollars. This guarantees that the dollar provides some benefit even when other players eschew it. The resulting payoff matrix is shown in Figure 6\.

|  |  | Boris |  |  |  |
| ----- | :---- | ----- | :---- | :---- | :---- |
|  |  | Use $ |  | Use bitcoin |  |
|  |  | Satoshi |  | Satoshi |  |
|  |  | Use $ | Use bitcoin | Use $ | Use bitcoin |
| Alice | Use $ |  | A: 2 B: 0 S: 0 |  | A: 2 B: 3 S: 3 |
|  | Use bitcoin |  | A: 1 B: 0 S: 2 |  | A: 1 B: 4 S: 4 |

*Figure 6\. Three-player bitcoin or dollars game.*

Note that exactly one Nash equilibrium appears in this payoff matrix, in the upper-right-hand box; no player may improve their lot by unilaterally departing from their strategy. But there is no optimal outcome. Every outcome has at least one player who could benefit from a departure.

Given this scenario, Alice will definitely select dollars. She’s better off under that strategy no matter what the other players do. Given our fixed background condition, Alice is never alone either. She will always have dollar-preferring friends — off-screen, as it were. 

Yet despite Alice’s preferences, and despite her network of dollar-preferring friends, both Boris and Satoshi have reason to select a bitcoin strategy. No matter what the other players do, Boris is better off selecting bitcoin. And given the setup, Satoshi’s fate is already sealed; he will select bitcoin, and the boxes in which he selects dollars are already blocked out. Should Satoshi worry about what Boris will do, and whether he’ll be left out in the cold? No. For Satoshi may reason, just as we have already done, that selecting bitcoin is a dominant strategy for Boris.

Here’s another way to process things: Boris and Satoshi know that Alice will not select bitcoin in any case. So their reasoning can simply ignore her. As a result, Boris and Satoshi may transition to playing a two-player game depicted in Figure 7\.

|  | Boris uses dollars  | Boris uses bitcoin  |
| :---- | :---- | :---- |
| Satoshi uses dollars |  |  |
| Satoshi uses bitcoin | B: 0 S: 0 | B: 3 S: 3 |

*Figure 7\. Two-player bitcoin or dollars game.*

Alice is not joining Boris and Satoshi in selecting bitcoin, remember. She refuses. The lower-right-hand box in Figure 7, accordingly, assigns a reduced payoff to both Boris and Satoshi than does the lower-right-hand box in Figure 6: three instead of four. This is as it should be, for without Alice onboard the bitcoin train, neither Boris nor Satoshi will benefit from using bitcoin with Alice. Despite this setback, however, the resulting two-player game in Figure 7 has an important feature, and one not enjoyed by the game in Figure 6: its lower right-hand-box is a Pareto optimal Nash equilibrium. Deviation from it will harm the player who deviates, and any deviation from it will harm someone.

Precedent is potent. It can, as before, bring to unique salience an option otherwise indistinguishable from the others; it can also allow players to screen off as irrelevant certain (or, in the case of a weak precedent, somewhat irrelevant) outcomes, and even point them towards a game with a clear winning strategy, like the two-player game in Figure 7\. Coordination is thus achieved with new benefits unlocked.

Moreover, the consequences of coordination spread — for Boris and Satoshi may now act as a coalition in the game of Figure 6, and in yet larger money selection games. Alice will not join that coalition. Nor will her legion of dollar-loving friends. For Boris and Satoshi’s purposes, that’s fine. A two-player coalition is enough to benefit its members. Perhaps more importantly, these potential benefits can attract a third member, and then a fourth.[^87] Anyone may now adopt its defining convention, join its associated institution, and benefit all who do the same. All this thanks to the early work of Satoshi, who made sacrifices to adopt a new money, and Boris, who had little choice but to adopt it.

The modulated money selection game now in view is much closer to what we’ve actually observed across bitcoin’s fifteen years of existence. Satoshi led, and the Borises of the world followed. The Alices have not. Since, presently, Alices greatly outnumber Borises, the dollar continues to reign supreme among monies. Yet its rule is not absolute, for the number of Borises is greater than zero, and by now, Boris has many bitcoin-using friends.[^88] As a result, new entrants to money selection games know that if they select a bitcoin strategy, they will have company. In this way, bitcoin adoption has launched from nothingness into a stable equilibrium.

The story we’ve told about the game theory of early bitcoin adoption matches what we’d want to say about adoption of a natural language. Imagine a two-player language selection game: Russian or English. Why might Boris start using Russian? Because someone else — Satoshi, let’s call him — has already sent a credible signal that he uses or is willing to learn Russian. The costliness of the signal—the resources required to learn—enhances its credibility and shows that the preference in view is stable and likely to last across multiple rounds of play. With this information in place, Boris can safely ignore the outcomes in which Satoshi drops Russian and instead learn the language himself. Again, Satoshi’s behavior establishes a precedent. As the Russian language’s network of users grows, so does the usefulness of the language itself. Of course, not everyone needs or wants to chat with Satoshi. Nevertheless, credible signals aid coordination among interested players of a game without any centralized coordinator.

Although we have assumed without argument that bitcoin benefits some of its users, we make a book-length case for this claim elsewhere.[^89] But one doesn’t need a book to make an informed guess that bitcoin offers significant benefits. That bitcoin has users at all suggests that they find it useful. And that its user base has grown since its inception suggests that, in true network good fashion, its utility has increased as well.

# 6\. Bitcoin is Feral

Bitcoin, like any other novel money, has faced serious barriers to adoption. But its inventor employed a distinctive strategy to increase its probability of adoption, culminating in his eventual departure. Together, the barriers and the strategy Satoshi developed to overcome them explain how the Bitcoin network has become a working coordination machine. Satoshi’s distinctive strategy also catapulted Bitcoin into becoming something rather different from what it once was. Though it began life as a carefully engineered institution, Bitcoin is now something more feral. 

Before explaining this finding in more detail, another table will prove useful. Let’s compare and contrast bitcoin with a particular natural language: English. Both English and bitcoin are associated with an interacting complex of elements:

- A coordination problem: players in a game who would benefit from acting together.  
- A conventional solution: a rule, backed by precedent, with players motivated to follow it.  
- New functionality or use associated with the defining convention: genuine benefits for those who follow the rule.  
- A resulting network good: benefits that scale up with use – the more players who adopt the rule, the better off they are.  
- An institution: a connected group of people whose behavior is structured by rules in the ways noted above, with some stability over time.  
- A creation story or origin: how the defining convention was first promulgated.

We can place these elements against each other for easy comparison and contrast in Figure 8\.

|  | English | Bitcoin |
| :---- | :---- | :---- |
| Problem | Coincidence of meanings | Coincidence of wants |
| Convention | *Use English*, which means obeying rules governing syntax, semantics, and behavior | *Use bitcoin*, which means obeying rules enshrined in the Bitcoin protocol for receiving and sending transactions |
| Enables | Information transfer,    exchange of ideas by way of words | Value transfer,  exchange of value by way of networked digital computers |
| Network good | The English language | Bitcoin the asset |
| Institution | The English network (people who use English) | The Bitcoin network (people who use Bitcoin) |
| Creation | Natural | Engineered |

*Figure 8\. Bitcoin and English complexes.*

The similarities are striking: Bitcoin resembles a natural language in important respects. This is evidence that Bitcoin the institution is similar in kind to natural languages like English. However, as the last row indicates, Bitcoin differs from natural languages in an important way: Bitcoin’s conventions have an architect. In the beginning, Satoshi made the whole thing up. Bitcoin’s protocol was engineered, and there is an important sense in which its network good (the asset) and institution (the network) are synthetic, too.

Consider the difference between formal and natural languages. Formal languages are designed for specific purposes and often include explicit syntactical rules and vocabulary. For example, logicians regularly deploy the language of first order logic, an engineered language with a strictly defined vocabulary of variables, predicate symbols, sentential connectives, quantifiers, and so on. A typical string in that language looks like this:

|  ∀*x*(∃*y*(F*y* ∧ G*y*) ↔ F*x*) |
| :---: |

There are other formal languages as well. Some, like the language of first order logic, are intended for use by humans. Others, like C++, are intended for execution by machines. They do not merely describe; they command and coordinate the intentions of coders with the behavior of machines. A few lines of C++ might look like this:

| \#include \<iostream\> int main() {   std::cout \<\< "hello world";   return 0; } |
| :---- |

By contrast, as far as we know, no single person or small group of people invented English. Instead, the language evolved over time and through the behavior of millions of users brought together by a need to communicate. English is orderly but unplanned. For that reason, English is a paradigmatic *natural* language. Formal languages, by contrast, were engineered, sometimes by one person or by a small group of people (as with Gottlob Frege and the language of first order logic, for example).[^90]

Like formal languages, Bitcoin’s own protocol is *formally defined*. The whole thing is engineered. But, as we’ll see, focusing on what Satoshi did illuminates a key point of difference between bitcoin and familiar examples of both formal and natural languages. Despite its explicit rules, Bitcoin’s protocol falls within a slightly different category than either formal or natural languages. The resulting complex is similarly distinctive. Bitcoin therefore enjoys a distinctive institutional profile.

In a word, bitcoin is *feral*.

Here’s what we mean. Bitcoin certainly began as a tame creature. It was, as we’ve said, created by Satoshi the software engineer. For a time, everything about Bitcoin was entirely in Satoshi’s hands. But some important choices he made in those early years let Bitcoin break loose. He chose a bootstrapping strategy that ultimately turned marketing over to a distributed team of users rather than any budget baked into the protocol or any associated foundation or corporation. Then Satoshi left. The one way to guarantee that Bitcoin was no longer in his hands was for those hands to disappear. They did.

Some non-human animals would never have existed without human influence in the form of selective breeding – dogs and horses, for example. Many of the resulting beasts live with us still as treasured companions and co-laborers. But sometimes they escape. Sometimes, we let them go. Either way, the resulting species is not entirely wild; its members bear the marks of co-evolution with us. But it isn’t entirely tame either. Even hundreds of years after escaping from the Spaniards, for example, descendent mustangs who freely roamed the Americas bear the mark of their tame ancestors.

Bitcoin is like this. It’s a mustang who got loose and never returned. It couldn’t have existed in the first place without selective breeding – engineering, really. It required the careful oversight of an inventor like Satoshi. But it did not remain a domesticated creature. The engineer is gone, and the inventor is no more. Bitcoin has thus been set free.

Bitcoin’s network is a feral institution. English, by contrast, is wild and always has been, so far as we know. It never had an inventor. Bitcoin did. But then its inventor left and let Bitcoin out of the barn. By altering Bitcoin’s layer 0 in this way, Satoshi unleashed the potential for leaderlessness that was already present in its layer 1\.

We observe one more similarity between natural languages and the feral institution of Bitcoin: it concerns their definitive conventions and how they evolve over time. English has rules, and those rules can change. Some such rules are semantic; they concern meaning. How do you change the meaning of an English word? English has no CEO to direct changes; no foundation unilaterally updates a canonical lexicon. Indeed, there is no canonical lexicon at all. Instead, millions of English users freely deploy their own overlapping idiolects. If you wish to change the meaning of an English word or add a new word to the lexicon, enough people must start using words in the relevant ways.

Bitcoin has rules. And its rules, too, can be changed. How do you bring about one of these changes? There is no CEO and no foundation. Instead, thousands of users each freely run whatever software they like. If you wish to institute a fork, hard or soft, you must persuade enough people to run it. If you wish to move the herd, you must persuade the herd.[^91]

We’ve argued elsewhere that bitcoin’s native asset exhibits a distinctive monetary profile.[^92] It has no centralized makers, has a non-zero marginal cost of production, has no non-monetary use, can be directly possessed, and is natively digital. Here — see Figure 9 — is how it stacks up against a few other familiar monetary offerings.

|  | Physical dollars | Digital dollars | Gold | Bitcoin |
| :---: | :---: | :---: | :---: | :---: |
| Centralized makers | ✓ | ✓ | ✗ | ✗ |
| Costly production | ✗ | ✗ | ✓ | ✓ |
| Non-monetary use | ✗ | ✗ | ✓ | ✗ |
| Direct possession | ✓ | ✗ | ✓ | ✓ |
| Digital native | ✗ | ✓ | ✗ | ✓ |

*Figure 9\. Bitcoin and other monies.*

Bitcoin’s native asset, bitcoin, blends features of traditional commodity monies and traditional fiat monies – it is neither a mere element of nature nor a pure creation of arbitrary human will. It is, instead, a synthetic commodity money.[^93]

The hypothesis of this section – that bitcoin is a feral institution – coheres well with this result. We should expect that a network creating and sustaining a synthetic commodity money should fall somewhere between nature and artifice. 

# Conclusion

Bitcoin began life as an engineered product. Its inventor’s distinctive approach made it something more, however. In spending resources to produce bitcoin, Satoshi resolved a key game theoretic problem that besets anyone would-be monetary founder. And by departing from public view, Satoshi enabled bitcoin to activate its potential as a leaderless network. The history and game theory described thus far have consequences for the metaphysics of bitcoin – for what it really is. Bitcoin is a new kind of thing: a feral institution. Once engineered, it now roams freely.

George Washington, Thomas Jefferson, Satoshi Nakamoto. Some might find this grouping absurd. But our arguments thus far reveal the parallels. Washington and Jefferson are American founding fathers. Their overall legacy is a mixed bag, but this much we know: they were instrumental in building the world’s longest-standing modern democracy. Apart from their efforts in the Revolutionary War, they helped craft the Constitution. In so doing, they limited state power and hence their individual power within. Decades earlier, Jefferson had also drafted the Declaration of Independence, itself a sustained argument against the institution of hereditary monarchy. It was his conviction that just powers of states were derived by consent of the governed, not by inheritance.

Washington and Jefferson were social architects. They knew that formal protections against an overreaching state would not suffice; layer 0 mattered, too. In 1807, therefore, President Thomas Jefferson informed the New York legislature that he would refuse a third term. He cited his “illustrious predecessor” George Washington as setting the precedent he would follow. Relinquishing power was, for both Washington and Jefferson, the culmination of a career aimed at building a nation without lifelong rulers.[^94] Their actions, in turn, informed social conventions around higher political office in the United States for centuries to come.

Bitcoin’s software and the American Constitution thus serve similar conventional roles: both encode checks and balances and outline a process for reaching consensus. Whereas the Constitution aims for just political consensus, Bitcoin’s software aims for financial consensus about who has what quantities of Bitcoin’s native asset. The checks and balances within the Bitcoin protocol work through clever incentive design: the software lures people to compete for bitcoin rewards in a way that secures Bitcoin’s global ledger and hence the bitcoin rewards whose possession it documents.

The American founding fathers adhered to the Constitution that they themselves designed. Likewise, Satoshi mined bitcoin like any other miner, as his own software required. There were no exceptions in Bitcoin’s defining conventions that assigned anyone any formal leadership status or special powers to mint coins or spend them. 

Strikingly, both Washington and Satoshi withdrew from their leadership roles under similar circumstances. Washington left office in part because he feared exacerbating the partisanship of the burgeoning political parties. He writes in 1799, for example, that he is “thoroughly convinced” that he would “not draw a single vote from the Anti-federal side.”[^95] Satoshi was similarly embroiled in controversies during his last days publicly contributing to Bitcoin.[^96] So he took his leave.[^97]

In summary, Satoshi discovered how money could operate without central authorities, encoded this discovery into software, freely gave it to the world from behind a mask, bootstrapped a community devoted to its success, relinquished control, and—left. Anyone may now use the feral institution he left behind.

That is what Satoshi did. He set Bitcoin free.

# Bibliography

Alden, Lyn. “Proof-of-Stake and Stablecoins: A Blockchain Centralization Dilemma.” Last modified November 2021\. [https://www.lynalden.com/proof-of-stake/](https://www.lynalden.com/proof-of-stake/).  
Alden, Lyn. *Broken Money*. Timestamp Press, 2023\.  
Bailey, Andrew M. “Digital Value.” *Philosophy & Digitality* (forthcoming).  
Bailey, Andrew M., Bradley Rettler, and Craig Warmke. *Resistance Money: A Philosophical Case for Bitcoin*. New York: Routledge, 2024\.  
Bailey, Andrew M., Bradley Rettler, and Craig Warmke. “Bitcoin’s halving is a major spectacle — that’s the whole point”. *Blockworks*, 2024\.  
Bailey, Andrew M., Bradley Rettler, and Craig Warmke. “Philosophy, Politics, and Economics of Cryptocurrency I: Money Without State.” *Philosophy Compass* 16, no. 11 (2021).  
Bailey, Andrew M., Bradley Rettler, and Craig Warmke. “Philosophy, Politics, and Economics of Cryptocurrency II: The Moral Landscape of Monetary Design.” *Philosophy Compass* 16, no. 11 (2021).  
Bailey, Andrew M., and Craig Warmke. “Bitcoin is King.” In *Cryptocurrency: Concepts, Technology, and Issues*, edited by J. Liebowitz, 175-197. London: Taylor & Francis, 2023\.  
Baldwin, Jon. “In Digital We Trust: Bitcoin Discourse, Digital Currencies, and Decentralized Network Fetishism.” *Palgrave Communications* 4 (2018).  
Baur, Dirk G., KiHoon Hong, and Adrian D. Lee. “Bitcoin: Medium of Exchange or Speculative Asset?” *Journal of International Financial Markets, Institutions and Money* 54 (2018): 177-189.  
Beltramini, Enrico. “Against Technocratic Authoritarianism. A Short Intellectual History of the Cypherpunk Movement.” *Internet Histories* 5 (2021): 101-118.   
Butler, Simon. “The Philosophy of Bitcoin and the Question of Money.” *Theory, Culture & Society* 39 (2022): 81-102.  
BBC. “Bitcoin Becomes Official Currency in Central African Republic.” *BBC News*. Last modified April 27, 2022\. [https://www.bbc.com/news/world-africa-61248809](https://www.bbc.com/news/world-africa-61248809).  
Bier, Jonathan. *The Blocksize War: The Battle Over Who Controls Bitcoin’s Protocol Rules*. Self-published, 2021\.  
Blau, Benjamin M., Todd G. Griffith, and Ryan J. Whitby. “Inflation and Bitcoin: A Descriptive Time-Series Analysis.” *Economics Letters* 203 (2021): 109848\.  
Bouri, Elie, Peter Molnár, Georges Azzi, David Roubaud, and Lars Ivar Hagfors. “On the Hedge and Safe Haven Properties of Bitcoin: Is it Really More Than a Diversifier?” *Finance Research Letters* 20 (2017): 192-198.  
Brunton, Finn. *Digital Cash: The Unknown History of the Anarchists, Utopians, and Technologists Who Created Cryptocurrency*. Princeton, NJ: Princeton University Press, 2019\.  
Carter, Nic. “Bitcoin’s Existential Crisis.” Medium, May 9, 2018\.   
[https://medium.com/s/story/what-is-it-like-to-be-a-bitcoin-56109f3e6753](https://medium.com/s/story/what-is-it-like-to-be-a-bitcoin-56109f3e6753).  
Cross, Troy, and Andrew M. Bailey. “Carbon-Neutral Bitcoin Adoption for Nation States.” In *Financial Cryptography and Data Security*, edited by Matsuo et al., 55-65. Berlin: Springer, 2023\.  
W. Dai. “b-money.” [http://www.weidai.com/bmoney.txt](http://www.weidai.com/bmoney.txt),  1998\.  
De Filippi, Primavera, and Benjamin Loveluck. “The Invisible Politics of Bitcoin: Governance Crisis of a Decentralised Infrastructure.” *Internet Policy Review* 5, no. 3 (2016): 1-28.  
Dick, David. “How Many People Does it Take to Make a Dollar?” In *The Philosophy of Money and Finance*, edited by Sandberg and Warenski. Oxford: Oxford University Press, 2024\.  
Dodd, Nigel. “The Social Life of Bitcoin.” *Theory, Culture & Society* 35, no. 3 (2018): 35-56.  
Dowd, Kevin, and David Greenaway. “Currency Competition, Network Externalities, and Switching Costs: Towards an Alternative View of Optimum Currency Areas.” *The Economic Journal* 103, no. 420 (1993): 1180-89.  
Epstein, Brian. “What Are Social Groups? Their Metaphysics and How to Classify Them.” *Synthese* 196 (2019): 4899-4932.   
Ewald, William. “The Emergence of First-Order Logic.” *Stanford Encyclopedia of Philosophy*. Last modified June 3, 2018\.   
[https://plato.stanford.edu/entries/logic-firstorder-emergence/](https://plato.stanford.edu/entries/logic-firstorder-emergence/).  
Guala, Francesco. *Understanding Institutions*. Princeton, NJ: Princeton University Press, 2016\.  
Hayes, Adam. “The Socio-Technological Lives of Bitcoin.” *Theory, Culture & Society* 36, no. 4 (2019): 49-72.  
Hayward, Andrew. “Twitter CEO Jack Dorsey: Bitcoin Whitepaper is ‘Poetry.’” *Decrypt*, April 17, 2020\. [https://decrypt.co/26705/twitter-ceo-tech-crypto-currency](https://decrypt.co/26705/twitter-ceo-tech-crypto-currency).  
Hazlett, Peter K., and William J. Luther. “Is Bitcoin Money? And What That Means.” *The Quarterly Review of Economics and Finance* 77 (2019): 144–149.  
Hendrickson, Joshua R., and William J. Luther. “Cash, Crime, and Cryptocurrencies.” *The Quarterly Review of Economics and Finance* 85 (2022): 200-207.  
Held, Dan. “Planting Bitcoin \- Soil.” Last modified January 6, 2019\.   
[https://www.danheld.com/blog/2019/1/6/planting-bitcoinsoil-34](https://www.danheld.com/blog/2019/1/6/planting-bitcoinsoil-34).  
Johnston, Mark. “How to Speak of the Colors.” *Philosophical Studies* 68 (1992): 221-263.  
Kapilkov, Michael. “Previously Unpublished Emails of Satoshi Nakamoto Present a New Puzzle.” *Coindesk*, November 26, 2020\. [https://www.coindesk.com/markets/2020/11/26/previously-unpublished-emails-of-satoshi-nakamoto-present-a-new-puzzle/](https://www.coindesk.com/markets/2020/11/26/previously-unpublished-emails-of-satoshi-nakamoto-present-a-new-puzzle/).  
Kavanagh, Donncha, and Paul John Dylan-Ennis. “Cryptocurrencies and the Emergence of Blockocracy.” *The Information Society* 36, no. 5 (2020): 290-300.  
Lewis, David. *Convention: A Philosophical Study*. Oxford: Wiley-Blackwell, 1969\.  
Lipman, Martin. “On Bitcoin: A Study in Applied Metaphysics.” *Philosophical Quarterly* 73 (2023): 783-802.  
Lopp, Jameson. “Who Controls Bitcoin Core?” Last modified October 1, 2018\. [https://blog.lopp.net/who-controls-bitcoin-core-/](https://blog.lopp.net/who-controls-bitcoin-core-/).  
Luther, William J. “Cryptocurrencies, Network Effects, and Switching Costs.” *Contemporary Economic Policy* 34, no. 3 (2016): 553–571.  
Luther, William J. “Is Bitcoin Intrinsically Worthless?” *Journal of Private Enterprise* 33, no. 1 (2018): 31-45.   
Luther, William J. “Getting Off the Ground: The Case of Bitcoin.” *Journal of Institutional Economics* 15, no. 2 (2019): 189–205.  
Narayanan, Arvind, and Jeremy Clark. “Bitcoin’s Academic Pedigree.” *Communications of the ACM* 60, no. 12 (2017): 36–45.   
Nakamoto, Satoshi. “Bitcoin: A Peer-to-Peer Electronic Cash System.” Last modified October 31, 2008\. [https://bitcoin.org/bitcoin.pdf](https://bitcoin.org/bitcoin.pdf).  
Palmer, Daniel. “Craig Wright Called ‘Fraud’ in Message Signed With Bitcoin Addresses He Claims to Own.” *CoinDesk*, May 25, 2020\.   
[https://www.coindesk.com/policy/2020/05/25/craig-wright-called-fraud-in-message-signed-with-bitcoin-addresses-he-claims-to-own/](https://www.coindesk.com/policy/2020/05/25/craig-wright-called-fraud-in-message-signed-with-bitcoin-addresses-he-claims-to-own/).  
Parkin, Jack. “The Senatorial Governance of Bitcoin: Making (De)centralized Money.” *Economy and Society* 48, no. 3 (2019): 463-487.  
Popper, Nathaniel. “A Bitcoin Believer’s Crisis of Faith.” *New York Times* January 14, 2016\. [https://www.nytimes.com/2016/01/17/business/dealbook/the-bitcoin-believer-who-gave-up.html](https://www.nytimes.com/2016/01/17/business/dealbook/the-bitcoin-believer-who-gave-up.html).  
Raskin, Max. “A Global First: Bitcoin as National Currency.” *The Wall Street Journal*, June 15, 2021\.  
Ritchie, Katherine. “The Metaphysics of Groups.” *Philosophy Compass* 10, no. 5 (2015): 310-321.  
Rizzo, Pete. “The Last Days of Satoshi.” *Bitcoin Magazine*. Last modified April 26, 2021\. [https://bitcoinmagazine.com/technical/what-happened-when-bitcoin-creator-satoshi-nakamoto-disappeared](https://bitcoinmagazine.com/technical/what-happened-when-bitcoin-creator-satoshi-nakamoto-disappeared).  
Rizzo, Pete. “10 Years Ago Today Bitcoin Creator Satoshi Nakamoto Sent His Final Message.” *Forbes*. Last modified April 26, 2021\. [https://www.forbes.com/sites/peterizzo/2021/04/26/10-years-ago-today-bitcoin-creator-satoshi-nakamoto-sent-his-final-message/](https://www.forbes.com/sites/peterizzo/2021/04/26/10-years-ago-today-bitcoin-creator-satoshi-nakamoto-sent-his-final-message/).  
Salter, Alexander, and William J. Luther. “Synthesizing State and Spontaneous Order Theories of Money.” *Advances in Austrian Economics* 18 (2014): 105-123.   
Sarkar, Arijit. “Bitcoin Was Almost Named Netcoin by Satoshi Nakamoto, Hints Domain Data.” *CoinTelegraph*. Last modified January 9, 2022\. [https://cointelegraph.com/news/bitcoin-was-almost-named-netcoin-by-satoshi-nakamoto-hints-domain-data](https://cointelegraph.com/news/bitcoin-was-almost-named-netcoin-by-satoshi-nakamoto-hints-domain-data).    
Schelling, Thomas C. *The Strategy of Conflict*. Cambridge, MA: Harvard University Press, 1960\.   
Schneider, Nathan. “Decentralization: An Incomplete Ambition.” *Journal of Cultural Economy* 12, no. 4 (2019): 265-285.   
Sedgwick, Kai. “Bitcoin History Part 19: Wikileaks and the Hornet’s Nest.” *Bitcoin News*. Last modified February 6, 2019\. [https://news.bitcoin.com/bitcoin-history-part-19-wikileaks-and-the-hornets-nest/](https://news.bitcoin.com/bitcoin-history-part-19-wikileaks-and-the-hornets-nest/).    
Selgin, George. “Synthetic Commodity Money.” *Journal of Financial Stability* 17 (2015): 92–99.  
Soros, George. *The Alchemy of Finance*. New York: Simon and Schuster, 1988\.   
Srinivasan, Balaji. *The Network State*. Self-published, 2022\. [https://thenetworkstate.com](https://thenetworkstate.com).  
Swartz, Lana. “What Was Bitcoin, What Will It Be? The Techno-Economic Imaginaries of a New Money Technology.” *Cultural Studies* 32, no. 4 (2018): 623-650.    
Taylor, Luke. “The World’s First Bitcoin Republic.” *New Scientist* 253 (2022): 14\.   
Walch, A. “In Code(rs) We Trust: Software Developers as Fiduciaries in Public Blockchains.” In *Regulating Blockchain: Techno-Social and Legal Challenges*, edited by P. Hacker, I. Lianos, G. Dimitropoulos, and S. Eich, 58–82. Oxford: Oxford University Press, 2019\.   
Wan, Samuel. “Was the CIA Behind Bitcoin Creator Satoshi Nakamoto’s ‘Disappearance’?” *CryptoSlate*. Last modified May 25, 2022\. [https://cryptoslate.com/was-the-cia-behind-bitcoin-creator-satoshi-nakamotos-disappearance/](https://cryptoslate.com/was-the-cia-behind-bitcoin-creator-satoshi-nakamotos-disappearance/).    
Warmke, Craig. “What Is Bitcoin?” *Inquiry* 67 (2021): 25-67.  
Warmke, Craig. “Electronic Coins.” *Cryptoeconomic Systems* 2, 1 (2022).  
Warmke, Craig. “Hash Handcuffs.” (manuscript).   
Warren, Micah. *Bitcoin: A Game-Theoretic Analysis*. Berlin: De Gruyter, 2023\.  
Whitmore, Andrew, Namjoo Choi, and Anna Arzrumtsyan. “Open Source Software: The Role of Marketing in the Diffusion of Innovation.” *Information Technology and Control* 38, no. 2 (2009): 117-125.  
Zamani, Efpraxia D. “The Bitcoin Protocol as a System of Power.” *Ethics and Information Technology* 24, no. 1 (2022): 1-13.  
Zeelenberg, M., and S. M. Breugelmans. “The Good, Bad and Ugly of Dispositional Greed.” *Current Opinion in Psychology* 46 (2022): 101323\.   
Zook, Matthew, and Joe Blankenship. “New Spaces of Disruption? The Failures of Bitcoin and the Rhetorical Power of Algorithmic Governance.” *Geoforum* 96 (2018): 248-255.

[^1]:  For helpful conversation, critique, or comments, we thank Natalie Smolenski, Tuur Demeester, Dan Hughes, Bradley Rettler, Troy Cross, Trey Walsh, Simon Butler, Austin Krauss, Kaushik Swaminathan, Maximilian Brichta, Milan Stanojevic, Ben Blumson, Micah Warren, and various anonymous bitcoiners (they know who they are). We also thank the Bitcoin Policy Institute for research support.

[^2]:  Narayanan and Clark, “Bitcoin’s Academic Pedigree,” Beltramini, “Against Technocratic Authoritarianism”.

[^3]:  Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 3\.

[^4]:  We’ll drop “Nakamoto” from here on out and refer to Satoshi as an individual (although the moniker may well have been operated instead by a collective) and use the pronouns that Satoshi apparently preferred.

[^5]:  A network good is one where additional consumption (participation) drives up the utility of consumption (participation).

[^6]:  For a description of how Bitocoin’s Layer 1 works and its ultimate nature, see Warmke, “What is Bitcoin?” and Warmke, “Electronic Coins”.

[^7]:  Some representative samples include Baldwin, “In Digital We Trust,”, De Filippi and Loveluck, “The Invisible Politics of Bitcoin,” Dodd, “The Social Life of Bitcoin,” 45-52, Hayes, “The Socio-Technological Lives of Bitcoin,” 51, Parkin, “The Senatorial Governance of Bitcoin,” 463, Swartz, “What Was Bitcoin, What Will It Be?”, Walch, “In Code(rs) We Trust,” 64, Zamani, “The Bitcoin Protocol as a System of Power,” 13, and Zook and Blankenship, ‘New Spaces of Disruption?,” 248-249. For penetrating critique of ‘decentralization’ talk that often obscures bitcoin’s layer 0, see Schneider, “Decentralization: An Incomplete Ambition”. For a much friendlier treatment of bitcoin’s politics from a social scientist, see Butler, “The Philosophy of Bitcoin,” 92-97.

[^8]:  Ibid.

[^9]:  Bailey, Rettler, and Warmke, *Resistance Money*, especially Chapter 2: Section 2.6 and Chapter 11, Sections 11.3.2 and 11.3.5.

[^10]:  For penetrating discussion of another factor that blocks easy network cloning, see Alden, “Proof-of-stake,” on stablecoins and their role in blockchain forks. Alden also gives a useful analogy to Wikipedia: cloning the entirety of its archive wouldn’t result in a site with equal traffic or usefulness – it’d lack the editors, users, and brand name of Wikipedia itself. So also with bitcoin and its network.

[^11]:  See, respectively, Baur, Hong, and Lee, “Bitcoin: Medium of Exchange or Speculative Asset?,” Hazlett and Luther, “Is Bitcoin Money?,”, Blau, Griffith, and Whitby, “Inflation and Bitcoin,” and Bouri, Molnár, Azzi, Roubaud, and Hagfors, “On the Hedge and Safe Properties of Bitcoin”.

[^12]:  Raskin, “A Global First: Bitcoin as National Currency,” Taylor, “The World’s First Bitcoin Republic,” BBC, “Bitcoin Becomes Official Currency”. See also Cross and Bailey, “Carbon-Neutral Bitcoin Adoption for Nation States”.

[^13]:  Lopp, “Who Controls Bitcoin Core?”.

[^14]:  Our main focus here is on bitcoin. But see Bailey, Rettler, and Warmke, “Money Without State,” “The Moral Landscape of Monetary Design,” and Bailey and Warmke, “Bitcoin is King,” for comparative treatment of bitcoin in the context of other cryptocurrency offerings.

[^15]:  Our history will begin with bitcoin. For a broader perspective on the quest for digital cash and bitcoin’s place within it, see Brunton, *Digital Cash*.

[^16]:  Bailey and Warmke, “Bitcoin is King,” 186-188.

[^17]:  For comparative treatment of bitcoin’s bootstrapping stage from the perspective of monetary economics, see Luther, “Getting off the Ground”.

[^18]:  Sarkar, “Bitcoin Was Almost Named Netcoin”.

[^19]:  “Wei Dai/Satoshi Nakamoto 2009 Bitcoin Emails,” Gwern, accessed March 5, 2023, [https://www.gwern.net/docs/bitcoin/2008-nakamoto](https://www.gwern.net/docs/bitcoin/2008-nakamoto).

[^20]:  Dai, “b-money”.

[^21]:  “Wei Dai/Satoshi Nakamoto 2009 Bitcoin Emails,” Gwern, accessed March 5, 2023, [https://www.gwern.net/docs/bitcoin/2008-nakamoto](https://www.gwern.net/docs/bitcoin/2008-nakamoto).

[^22]:  Kapilkov, “Previously Unpublished Emails”.

[^23]:  Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System.” See Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 1, Section 1.3, for a taxonomy of these trusted parties and now bitcoin eliminates each.

[^24]:  Hayward, “Twitter CEO Jack Dorsey: Bitcoin whitepaper is ‘poetry’”.

[^25]:  “The Cryptography and Cryptography Policy Mailing List,” accessed March 5, 2023, [https://www.metzdowd.com/mailman/listinfo/cryptography/](https://www.metzdowd.com/mailman/listinfo/cryptography/).

[^26]:  Carter, “Bitcoin’s Existential Crisis”.

[^27]:  Held, “Planting Bitcoin”.

[^28]:  “Welcome to the new Bitcoin forum,” Satoshi Nakamoto, accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/posts/bitcointalk/5/](https://satoshi.nakamotoinstitute.org/posts/bitcointalk/5/).

[^29]:  “New icon/logo,” Satoshi Nakamoto, accessed March 5, 2023, [https://bitcointalk.org/index.php?topic=64.msg504\#msg504](https://bitcointalk.org/index.php?topic=64.msg504#msg504).

[^30]:  “Bitcoin open source implementation of P2P currency,” Satoshi Nakamoto, accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/posts/p2pfoundation/1](https://satoshi.nakamotoinstitute.org/posts/p2pfoundation/1/) .

[^31]:  “Bitcoin v0.1 released,” Satoshi Nakamoto, accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/emails/cryptography/16/\#selection-33.41-33.62](https://satoshi.nakamotoinstitute.org/emails/cryptography/16/#selection-33.41-33.62).

[^32]:  “Bitcoin open source implementation of P2P currency,” Satoshi Nakamoto, accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/posts/p2pfoundation/1](https://satoshi.nakamotoinstitute.org/posts/p2pfoundation/1/) .

[^33]:  Whitmore, Choi, and Arzrumtsyan, “Open Source Software,” 99\.

[^34]:  Ibid.

[^35]:  “It's very attractive to the libertarian viewpoint if we can explain it properly. I'm better with code than with words though.” See: “Bitcoin P2P e-cash paper,” Satoshi Nakamoto, last accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/emails/cryptography/12/](https://satoshi.nakamotoinstitute.org/emails/cryptography/12/).

[^36]:  This section draws heavily from Chapter 3, Section 3.4 of Bailey, Rettler, and Warmke, *Resistance Money*.

[^37]:  “Re: Bitcoin v0.1 released,” Hal Finney, accessed March 5, 2023, [https://www.mail-archive.com/cryptography@metzdowd.com/msg10152.html](https://www.mail-archive.com/cryptography@metzdowd.com/msg10152.html).

[^38]:  “Bitcoin v0.1 released,” Satoshi Nakamoto, accessed March 5, 2023, [https://satoshi.nakamotoinstitute.org/emails/cryptography/17/\#selection-103.0-111.57](https://satoshi.nakamotoinstitute.org/emails/cryptography/17/#selection-103.0-111.57).

[^39]:  Luther, “Cryptocurrencies, Network Effects, and Switching Costs,” 565\.

[^40]:  Whether that volatility is with us to stay is an open question; for discussion, see Bailey et al (2024): Chapter 5, Section 5.5.4.

[^41]:  “Bitcoin Releases Version 0.3,” SlashDot, Accessed March 5, 2023, [https://news.slashdot.org/story/10/07/11/1747245/bitcoin-releases-version-03](https://news.slashdot.org/story/10/07/11/1747245/bitcoin-releases-version-03). 

[^42]:  “Beta?,” Satoshi Nakamoto, accessed March 5, 2023, [https://bitcointalk.org/index.php?topic=217.0](https://bitcointalk.org/index.php?topic=217.0).

[^43]:  “Online-Only Currency BitCoin Reaches Dollar Parity,” SlashDot, accessed March 5, 2023, [https://news.slashdot.org/story/11/02/10/189246/online-only-currency-bitcoin-reaches-dollar-parity](https://news.slashdot.org/story/11/02/10/189246/online-only-currency-bitcoin-reaches-dollar-parity). 

[^44]:  Bailey, Rettler, and Warmke, “Bitcoin’s halving is a major spectacle”.

[^45]:  The idea here is not at war with the sensible thought that ‘halvings are priced in’, note. Bitcoin’s issuance schedule is indeed known in advance, and marginal buyers will discount their bids accordingly. Halving events provide no new information about supply, in other words. But new bids and excitement around them can fuel and provide new information about demand.

[^46]:  Soros, *The Alchemy of Finance,* 30\.

[^47]:  “Bitcoin P2P e-cash paper,” Satoshi Nakamoto, accessed March 5, 2023,  [https://satoshi.nakamotoinstitute.org/emails/cryptography/4/\#selection-43.8-43.106](https://satoshi.nakamotoinstitute.org/emails/cryptography/4/#selection-43.8-43.106).

[^48]:  “Satoshi \- Sirius emails 2009-2011”, Satoshi Nakamoto, accessed March 5, 2023, [https://mmalmi.github.io/satoshi/](https://mmalmi.github.io/satoshi/).

[^49]:  Zeelenberg and Breugelmans, “The Good, Bad, and Ugly of Dispositional Greed”.

[^50]:  Rizzo, “10 Years Ago Today Bitcoin Creator Satoshi Nakamoto Sent His Final Message”.

[^51]:  “A New Mystery in Patoshi Timestamps,” Sergio Demian Lerner, accessed March 5, 2023, [https://bitslog.com/2020/06/22/a-new-mystery-in-patoshi-timestamps/](https://bitslog.com/2020/06/22/a-new-mystery-in-patoshi-timestamps/).

[^52]:  We owe many of the dates and references here to Rizzo, “The Last Days of Satoshi”. 

[^53]:  “Sirius,” Bitcoin Wiki, accessed March 5, 2023, [https://en.bitcoin.it/wiki/Sirius](https://en.bitcoin.it/wiki/Sirius).

[^54]:  “Bitcoin: Contact/IRC”, Satoshi Nakamoto, accessed March 5, 2023, [https://web.archive.org/web/20101215111454/http://www.bitcoin.org:80/contact](https://web.archive.org/web/20101215111454/http://www.bitcoin.org:80/contact). 

[^55]:  “About bitcoin.org,” Bitcoin developers, accessed March 5, 2023, [https://bitcoin.org/en/about-us\#own](https://bitcoin.org/en/about-us#own). 

[^56]:  “Release v0.3.19,” Satoshi Nakamoto, accessed March 5, 2023, [https://github.com/bitcoin/bitcoin/releases/tag/v0.3.19](https://github.com/bitcoin/bitcoin/releases/tag/v0.3.19).

[^57]:  Popper, “A Bitcoin Believer’s Crisis of Faith”.

[^58]:  Rizzo, “The Last Days of Satoshi”.

[^59]:  Johnston, “How to Speak of the Colors,” 233

[^60]: ​​ This is the final line in a notorious message signed by one of bitcoin’s earliest pseudonymous miners. See Palmer, “Craig Wright Called ‘Fraud’ in Message Signed with Bitcoin Address He Claims to Own”.

[^61]:  For some options and useful framing about bitcoin’s metaphysics, see Lipman, “On Bitcoin”.

[^62]:  Guala, *Understanding Institutions* and Lewis, *Convention*. Guala treats conventions themselves as institutions. We find it more natural to speak of groups of people structured by conventions as institutions; but nothing will hang on this terminological point.

[^63]:  We take no stand here on the ontology of groups – whether they are individuals, mereological sums of their participating people, sets, something else entirely – or whether group talk can be understood as plural quantification in disguise. For a useful survey of some options, see Ritchie, “The Metaphysics of Groups”. See also Epstein, “What Are Social Groups?”.

[^64]:  For a very different approach to a cluster of related questions, see Kavanagh and Dylan-Ennis, “Cryptocurrencies and the Emergence of Blockocracy”.

[^65]:  Put slightly differently: a network good is a convention where additional consumption (participation) drives up the utility of consumption (participation).

[^66]:  On the comparison of monetary networks and languages, see Dowd and Greenway, “Currency Competition,” 1188\.

[^67]:  Parts of this section closely follow Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 3, Sections 1.2 and 3.4.

[^68]:  It is generally agreed that money solves a coordination problem along these lines. How this lines up against the actual history of money is a matter of significant controversy, though. See Szabo (2002) and Graeber (2011) for two important but wildly differing approaches.

[^69]:  Though it doesn’t matter to our argument here, it’s worth noting that Bitcoin solves the problem of coincident wants under rather extreme constraints that other monies do not obey. Above all, it is designed to allow exchange of value without trusted parties of various kinds. For a detailed explanation of how Bitcoin makes good on this promise, see Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 1\. For discussion of how bitcoin — a mere digital artifact, some might say — could enjoy value in the first instance, see Bailey, “Digital Value”.

[^70]:  In both cases, there are two levels of convention. We can express the first quite simply: “speak English,” or “use bitcoin as money.” But there is, one level deeper, considerable complication. To speak English is to be disposed to follow all sorts of phonetic and syntactical and behavioral rules, some far from obvious. Similarly, to use Bitcoin is to use software that follows all sorts of cryptographic and network rules, some of which are also far from obvious.

[^71]:  On the idea of a *network state*, which needn’t be coercive or geographically concentrated, see Srinivasan, *The Network State*.

[^72]:  For more on how this works, and a detailed assessment of whether it is ultimately for the best, see Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 7\. On threats to Bitcoin’s censorship resistance, see Warmke, “Hash Handcuffs”.

[^73]:  Lewis’ *Convention*, is perhaps the most influential text in this crossroads tradition. It is Lewis’ most-cited work and has commanded wide attention from linguists, economists, and philosophers; it is a significant influence on this section.

[^74]:  Our treatment will be elementary and focus on the game theory of Bitcoin adoption. A thorough and definitive treatment of Bitcoin’s game theory, with a focus on mining, may be found in Warren, *Bitcoin: A Game-Theoretic Analysis*. 

[^75]:  Lewis, *Convention*, 36-37.

[^76]:  Salter and Luther, “Synthesizing State and Spontaneous Order Theories of Money”.

[^77]:  Selgin, “Synthetic Commodity Money,” 93\. For further discussion, see Bailey, “Digital Value”.

[^78]:  See Dowd and Greenaway (1993) and Luther, “Cryptocurrencies, Network Effects, and Switching Costs”.

[^79]:  For further discussion of the Esperanto case and its lessons, see Bailey, Rettler, and Warmke, *Resistance Money*, 53ff.

[^80]:  Lewis, *Convention,* 14\.

[^81]:  These flashing asterisks are now known as “Schelling points”; see Schelling, *The Strategy of Conflict*.

[^82]:  On which, see Lewis, *Convention*, Section IV. 

[^83]:  Much of what we’ll say below could be formally regimented in terms of the *expected value* of various strategies, defined as the sum of the payoffs of each outcome under that strategy, each payoff weighted by its probability. Our main points do not require that machinery, however. So, for simplicity, we’ll instead speak more informally about outcomes being screened off as less salient or unlikely and model this as the relevant outcomes (and later, an entire player) being excised from the payoff matrix.

[^84]:  Luther, “Is Bitcoin Intrinsically Worthless?,” 43\.

[^85]:  For more on how early coordination around bitcoin was achieved with special attention to early bitcoin forum posts, see Luther later (2019). The story we tell here is consonant with Luther’s, though ours places greater emphasis on Satoshi’s role in achieving early coordination around bitcoin.

[^86]:  Being outright blocked from use, as is the case with Boris, is not the only way for a money to fail someone, though it may be especially salient for someone considering bitcoin. For thorough and lucid discussion of other monetary failure modes, see Alden, *Broken Money*.

[^87]:  Luther, “Getting Off the Ground,” 200: “To get off the ground, early bitcoin advocates recognized, you do not need to convince everyone to join the network. You just need a small, reliable user base to anchor the currency. Then, others can join on the expectation that the core group is committed to accepting bitcoin. In doing so, the additional users expand the network and, in turn, encourage still others to join. It is that old Mengerian story, with one very important exception: at the outset, it is coordination—not some non-monetary use value—that gets the ball rolling.” See also Dick, “How Many People Does it Take to Make a Dollar?,” 55\.

[^88]:  Hendrickson and Luther, “Cash, Crime, and Cryptocurrencies”.

[^89]:  Bailey, Rettler, and Warmke, *Resistance Money*.

[^90]:  Though see Ewald, “The Emergence of First-Order Logic”.

[^91]:  For a vivid illustration of how the rules for Bitcoin’s feral network evolve – proposed changes here are no less contentious than in the case of English users who would phase out certain of its naughty words and involve significant campaigns to influence the direction of its herd of users – see Bier, *The Blocksize War*.

[^92]:  See Bailey, Rettler, and Warmke, *Resistance Money*, Chapter 3\.

[^93]:  Selgin, “Synthetic Commodity Money”. 

[^94]:  “Letter From Thomas Jefferson to New York Legislature, 10 December 1807”, Thomas Jefferson, accessed March 5, 2023, [https://founders.archives.gov/documents/Jefferson/99-01-02-6955](https://founders.archives.gov/documents/Jefferson/99-01-02-6955).

[^95]:  “George Washington Letter to Jonathan Trumbull,” George Washington, accessed March 5, 2023, [https://www.gilderlehrman.org/sites/default/files/inline-pdfs/T-05787.pdf](https://www.gilderlehrman.org/sites/default/files/inline-pdfs/T-05787.pdf).

[^96]:  Rizzo, “The Last Days of Satoshi”.

[^97]:  Two other possible factors in the timing of Satoshi’s departure include Wikileaks’ decision to accept bitcoin payments (Satoshi warned of the ‘hornet’s nest’ that decision would arouse), and bitcoin developer Gavin Andreson’s decision to meet with the CIA to discuss bitcoin. See, respectively, Sedgwick, “Bitcoin History Part 19,” and Wan, “Was the CIA Behind Bitcoin Creator Satoshi Nakamoto’s ‘Disappearance’?”.