“Thinking Through Law and Code” — Lawrence Lessig lecture notes and thoughts.
In 1999 Lawrence Lessig published his book “Code: And Other Laws of Cyberspace” in which he argued code is law, inviting us to reflect on the forces that regulate our behaviours, and how the law affects those forces. For him it is necessary to go over the relationship between machine code and law once more. The advent of blockchain encourages him to tread old ground, remind us of how right his predictions had been regarding the internet, and warn us not to make the same mistakes again. Indeed, a parallel exists between the early days of the internet and these early days of blockchain, both in their philosophies of decentralisation, and potential.
Lawrence Lessig explains the modalities that influence people’s behaviour.
1. Law: Laws are ex-ante rules that govern our behaviours. They can determine a priori which actions are out of the question, and which are permissible. For example, the law tells us we cannot kill another or steal. Rules like this align themselves with fundamental morals and we all follow this code without needing to try first.
2. Norms: Norms are community codes — unwritten rules so to speak, which we all, more or less, live by. They aren’t written in law but rather form the foundation of our social interactions in that they signal to us the nature of situations in which we find ourselves. Take speech norms for instance; on a daily basis people auto-adjust their language based on the situations they are in, choosing perhaps to use colloquial language in social situations and formal language at work. Gender is also a carrier of many norms such as male/female dress codes and codes of conduct.
3. Market: The market impacts our behaviour in that different prices or costs orient our decisions.
4. Architecture: Lastly and perhaps most importantly in the online world, architecture affects our behaviour. Our physical environment such as fences and roads determine where we can and can’t go. Doors indicate entrances, and bridges routes over water. Architecture in software is the limitations imposed by user interfaces. Our behaviours are steered by which menu options we find, and which buttons are clickable.
→ In the end, the law actually impacts all three other modalities. Let’s look at smoking as an example. (1) The law regulates who can and can’t buy cigarettes. (2) Governments persuade people smoking is antisocial and harmful with advertising campaigns. They manipulate public opinion to alter norms so smoking begins to be seen negatively. (3) The State makes nicotine products extremely expensive, driving people away from the habit. (4) Finally, the State can impose rules on the composition of tobacco products such as the amount of nicotine in each cigarette. The effect here reducing people’s appetite for nicotine.
In the digital world, architecture affects the environment that enables or disables privacy, for example. Lessig challenges the 1990s notion that cyberspace is not regulable because of it’s immaterial nature; a place where free speech and privacy are protected. This early (libertarian) conception of the internet, notably voiced by EFF co-founder John Perry Barlow, considered that the traditional principles of property and identity were not relevant online. Lawrence Lessig points to an underlying confusion; legal concepts are applied to us not because we are existing in tangible space, but as a result of the architecture we are in, and limited to. The freedom of the early internet made possible by it’s architecture, could therefore be curtailed by actors (business/governments) seeking to police and monitor it, by turning it into an architecture of control. Of course, this has very much happened (cc Snowden revelations). In the 2000s, this original idealistic view of the internet was compromised, resulting in less privacy and security, and more control.
Blockchain: The technology to bring about the next wave of paradigm shifts?
Many mistakes were made in the 2000s which affected the development of the internet. How can those mistakes be avoided with the development and implementation of blockchains? Blockchain code is unique because it has two-fold potential. On the one hand technical (decentralised, transparent, immutable, automated) and on the other political - with the potential to improve governance and bridge the gap between behaviour and governments’ ability to regulate behaviour.
Blockchain is an architecture designed not to have a regulator in the middle, but blockchain evading the law is a libertarian fallacy. The law will not go away and will use it’s courts and other tools to intervene, however crudely. One striking and close-to-home example of this is the recent Court of Justice of European Union’s (CJEU) ruling in the Dutch Pirate Bay case. Here, the CJEU boldly expanded the legal notion of ‘communication to the public’ (essential in determining copyright infringement) to apply to The Pirate Bay that simply hosts and indexes trackers uploaded by users. Liability was confirmed (in spite of safe harbour laws protecting online intermediaries) and a web blocking injunction was handed down. (ref)
In Lawrence Lessig’s view, there is a fundamental need for reciprocal understanding of machine code and law.
Lawyers must grasp the radical efficiency brought in by blockchain code. It could be a great benefit to the most disadvantaged people, eliminating the need for trust in order to participate in economic (and other) transactions globally. Software developers must in turn understand three important aspects of the law; 1) imperfections are not always weaknesses, 2) it’s not possible to be above the law, and 3) escaping the law is often plain wrong.
1. A lack of clarity in law can be a feature! Ambiguous laws can help apply abstract legal principles to situations which may not have been originally envisioned. The All Writs Act, a United States Federal Statute dating from 1789 for instance, allows Federal Courts to “issue all writs necessary or appropriate in aid of their respective jurisdictions and agreeable to usages and principles of law”. Following a shooting in San Bernardino, California in 2016, the FBI got hold of a suspect’s locked iphone. In order to access the phone’s data and help the investigation track down suspects and piece together events, the FBI appealed to Apple to help them circumvent the phone’s encryption. Invoking the 1789 All Writs Act, the FBI served the tech giant a court order to force it to write new software that would allow law enforcement officials to hack suspects’ phones. (Apple, with support from other Silicon Valley behemoths, has strongly resisted the court order on a number of legitimate grounds.) (ref) This case illustrates the relationship between codified law and judge-made law. It can be more practical and beneficial for overarching rules to be vaguely worded with terms like “necessary” and “appropriate”, so that they endure and can be interpreted to apply to a variety of situations. Florian Glatz, blockchain lawyer and software developer defends this position commenting; ambiguous language can be helpful as much as detrimental as seen in the German constitutional case interpreting legislation from the 1950s to protect people from pervasive data gathering in the 1980s. Flexibility in language allows abstract rules to be adapted to a variety of situations which may occur in the real world, something binary machine code is not able to do. (ref) Lawrence Lessig references Meir Dan-Cohen who illustrates this idea on a positive note. Often, what people think the law says, is not necessarily echoed by court decisions and judgments. For example, people mostly believe ignorance of the law is not a valid defence, however in practice, judges often give reduced sentences on the basis that a wrong-doer didn’t realise they were breaking the law. Here, the legal ambiguity helps because it incentivises people to know the law, but also protects from punishment in the case of accidental wrong-doing. At the end of the day, lack of clarity can make the law more efficient by outsourcing precision to courts.
2. Evading the law is pointless. First of all, the law isn’t going away. Just because a system or transaction is automated, does not mean the law no longer applies. A vending machine provides a mechanised version of buying a fizzy drink from a shopkeeper, but consumer/merchant contract law still applies, as well as consumer protection policies. Even if we could think up a more complicated scenario, history tells us the State will use the law in however crude way to intervene in automated systems. (See here where the Australian Tax Office decided cryptocurrency should be considered an asset for capital gains tax purposes, and raided Craig Wright’s house).
Second of all, blockchains will need the courts to fill gaps in the implied agreements being carried out by these machine contracts. In the event of unexpected outcomes, disputes will follow that need resolving. Already, we have an example of an unexpected outcome from a smart contract: The DAO heist of 2016. The DAO was a smart contract on the Ethereum blockchain designed to pool funds in a decentralised way for subsequent investments. Unfortunately, the account got hacked and $50 million in Ether was stolen. (ref) Lukas Abegg, an IP lawyer and PhD student researching 3D printing and patent infringement discusses the heist based on three aspects of information: structural (form/shape), syntactic (letters, numbers) and semantic (content/meaning derived by humans). He explains that The DAO was created as pure code with no further legal framework or interference. The syntactic language of the code ruled the organisation. It transpired, someone used the code in a way that was not intended by the creators of the smart contract. Indeed, the INTENT was the semantic layer of the information in the machine code that was missing, which allowed the unexpected outcome. (ref)
Thirdly, in any contract, the State has an implied role of arbiter in the event of a dispute. Traditionally a contract partially removes trust between parties because it binds them in a legal framework that punishes breaches, but a court can also choose not to enforce a contract. Lawrence Lessig claims governments will intervene in the smart contracts they want and/or need to. Smart contracts might remove trust and the State as arbiter, but it is unlikely to accept being excommunicated across the board.
3. Escaping the law is often plain wrong. Sometimes, involvement from the state is wanted. Terms in private contracts are not always in the public interest. The public may want to resist pervasive private contracting which smart contracts would enable. The State has a place in enforcing fundamental public values. Towards the end of the nineteenth century, industrialisation took hold and mass production became an integral part of the economic order. Profits hinged on efficiency and manufacturers brought in uniform, pre-conceived agreements. (ref) The classical paradigm of individually negotiated contracts gave way to standard form contracts (‘take it or leave it’) which unfairly reflected the needs of the Offeror. In response to this evolution in contract theory, courts developed legislative tools to help signatories, such as the doctrine of fundamental breach and the doctrine of inequality of bargaining power. (ref)
Are there ways to mitigate the law / code conflicts? Which conflicts matter? Which are to be avoided? These questions are important because, if left to State forces to answer, their approach would come with “raw, stupid power”, to use Lawrence Lessig’s words. Lessig holds that the (USA) State is fraught with corruption and we (as citizens) must proactively preempt attempts to throttle or disproportionately control blockchain technology to protect business or government interests. If we do not ask ourselves which industries may be threatened by blockchains, we may fall prey to corporate interests determining through influencing policy, which uses are prudent, and which are not.
Watch the full lecture here.