MIT study explains why laws are written in an incomprehensible style
Laws and contracts are notoriously hard to read, and a new MIT study argues that one culprit is “center‑embedding” — long, nested clauses stuffed into the middle of sentences — which even non‑lawyers spontaneously use when asked to write laws. Commenters weigh this claim against other explanations: the need for precision under adversarial interpretation, centuries of precedent attached to specific phrasings, and the social function of legalese as an in‑group signal or “magic spell” that projects authority. Many point to ongoing plain‑language efforts and examples of clearer statutes as evidence that much legal writing could be simplified without losing enforceability, while others caution that true clarity is constrained by complex realities and shifting judicial interpretation.
Nature of legalese and precision
- Many see legalese as an attempt to use natural language like a programming language: precise, adversarially robust, and covering many edge cases.
- Others argue legal language is often still ambiguous despite being verbose and complex.
- Technical terms of art (“reasonable efforts”, “manslaughter”, etc.) serve as shorthand for long, litigated definitions, but make texts opaque to non‑lawyers.
Ambiguity, intent, and under-specification
- Several commenters distinguish harmful ambiguity (multiple plausible meanings) from deliberate under‑specification (clear principle, details left to judges).
- Some argue ambiguity can be socially useful, letting courts apply “intent of the law” in unforeseen cases; others see it as a power tool for the wealthy and for selective enforcement.
- Examples like “next Wednesday” and “this weekend” illustrate how ordinary time expressions are already hard to pin down, motivating more rigid legal phrasing.
On the MIT “center‑embedding / magic spell” study
- The paper’s focus on center‑embedded clauses as the main readability issue gets both interest and pushback.
- Several think the experiment (crowdworkers drafting laws) mostly shows people imitating the style of existing statutes, not necessarily trying to project authority.
- Some call the study underpowered and methodologically weak, and note it largely ignores centuries of legal‑linguistics scholarship and the role of judges and precedent.
Historical, institutional, and economic factors
- Case law and precedent strongly incentivize reusing exact, litigated phrases to reduce risk; changing wording can open new attack surfaces in court.
- In common‑law systems, statutes, regulations, and constitutional principles have different purposes and audiences, which shapes how “dense” each is.
- Commenters also mention professional gatekeeping, path dependence, political coalition‑building, and even “paid by the word” history as contributing pressures.
Comparisons to programming and formal languages
- Frequent analogies: law as code executed by human “compilers” (courts), written for adversarial interpreters.
- Some propose formal or domain‑specific languages for law (e.g., Catala), or highly constrained natural languages; others warn this would recreate a “Latin priesthood” of experts.
- There is interest in LLMs and knowledge‑graph tooling to navigate statutes and case law, but concern about hallucinations and persistent ambiguity.
Plain language and reform
- US “Plain Language” initiatives and clear drafting in places like Canada/New Zealand are cited as evidence that much legal text can be simplified without losing force.
- Lawyers in the thread report being taught to avoid unnecessary legalese, but entrenched templates and risk aversion slow change.