Much of the confusion about the proper regulation of smart contracts stems from the fact that both code and law are expressed in language. Natural (human) and formal (computer) languages are profoundly different, however. Natural language in the form of a true legal contract expresses human meaning and expectation. Code simply acts, and when code acts contrary to the understanding of the parties to a contract, courts must have a theoretical and legal basis in order to intervene--which this Article provides.

Present scholarship on the governance of smart contracts centers on logistical problems relating to the effects of automation on operation and execution, most notably problems of inflexibility and lack of enforcement discretion. However, automatic execution is nothing new in contract law. Rather, it is the legal interface between contract law and code that must catch and hold our attention. We focus on the point where the ‘natural language’ of contract law crosses over into the ‘formal language’ of computer code. Natural language contract terms are made accessible to a human and receive some sort of confirmation to establish the contractual magic, a set of bespoke legal rules between two parties encapsulated in some document or through behavior that makes the intention of the parties unmistakable. The formal language program portion of a smart contract executes, sometimes in accordance with these expectations, sometimes not. This Article asserts that human expectations determine the legal obligations of a contract, and that code merely executes it. It then explores the legal bases and ramifications of this human-centered law of smart contracting.