Role of Semantics in Legal Reality

2026-07-24
, , , , ,

In Programming Language (PL) theory, there are two primary ways to define what a program “means”:

The law maps to these exactly. The US and Indian legal systems formally aspire to Denotational Semantics – Objective Truth – but the Indian Evidence Act (IEA) and US Federal Rules of Evidence (FRE) are strictly Operational Semantics manuals: procedural execution guides, not truth manuals.

1. Denotational Semantics of Law: The Objective Truth

In denotational semantics, an expression \(e\) is assigned a mathematical value \(v\) via an evaluation function \(\mathcal{E}\). The meaning of \(e\) is just \(v\). It does not matter how long it takes to compute or what machine runs it. The truth is static.

The legal mapping. The denotational semantics of a criminal trial is the objective, historical reality of the event.

Let \(E\) be the syntactic event: “The defendant shot the victim.” The denotational function \(\mathcal{E}(E)\) maps to a strict Boolean: True or False. This mapping relies on Classical Logic (Law of Excluded Middle). \(\mathcal{E}(E)\) is objectively True or False regardless of whether any court can ever prove it.

In both the US and India, the goal of the legal system is to align the court’s verdict with the denotational semantics. The statute book provides the compositional rules:

\[ \mathcal{E}(\text{Murder}) = \mathcal{E}(\text{Act}) \land \mathcal{E}(\text{Intent}) \land \mathcal{E}(\text{Causation}) \]

Note the conjunction (AND), not addition. All three components must be True simultaneously. If the objective reality satisfies all three, the denotational value is Guilty.

2. Operational Semantics of Law: The Trial Process

In small-step operational semantics, meaning is defined by a transition relation \(\Gamma, e \to \Gamma', e'\). The program starts in an initial state and undergoes mechanical reductions until it reaches a final normal form.

The legal mapping. The courtroom is a virtual machine. The IEA and FRE are its Operational Semantics manuals. They do not define what is objectively true. They define what steps the machine must take to compute a verdict.

Under operational semantics, “truth” is not an objective mathematical value. It is the normal form the machine halts at after all permitted transitions are exhausted.

3. The Adequacy Theorem: Where Law Fails Math

In PL theory, an implementation is adequate if its operational semantics perfectly aligns with its denotational semantics: if \(\mathcal{E}(e) = \text{True}\), executing \(e\) must eventually halt and output True.

In law, the Adequacy Theorem is the core promise of justice: the court’s operational verdict will match the objective denotational truth. Legal realism acknowledges that operational semantics are inherently incomplete. The trial machine has strict resource limits:

The O.J. Simpson Case

Under the widely accepted reading of the trial evidence, \(\mathcal{E}(\text{Murder})\) was True. The operational machine, however, encountered a corrupted context: evidence of police misconduct introduced a plausible unsafeCoerce into the chain of custody. Because the FRE’s transition rules forbid building on an unsound context, the jury’s machine could not validly transition to Guilty. The operational result was Not Guilty – a failure of the Adequacy Theorem.

The subsequent civil trial operated under a different logical framework. The criminal trial operated under an intuitionistic evaluation constraint: the prosecution had to construct a direct, positive witness for Guilty. Disproving the defense’s alternative theories (\(\neg\neg\text{Guilty}\)) was legally insufficient. The civil trial, however, operated under classical logic with a lower threshold. By allowing Double Negation Elimination – simply proving the defense’s narrative was false (\(\neg\neg A\)) – the civil jury could legally derive liability (\(A\)). The shift from criminal to civil court was not just a change in probabilistic threshold; it was a shift from constructive to classical logic. It produced a liability verdict. The denotational value did not change; the operational machine did.

The Clinton Case

Clinton’s defense was a claim about the syntax of the predicate itself. His legal team argued before the grand jury that the term “is” in his deposition testimony – specifically “there is no sexual relationship” – should be parsed under a narrow syntactic reading that excluded past conduct. The strategy was to manipulate the operational rule for evaluating that specific input string so that the transition parse("is") -> present_tense_only would block the prosecution’s intended path to Guilty. The Senate impeachment trial was a parallel operational machine running under separate procedural rules – not an appeal of a lower court verdict, since no criminal conviction had been entered. Two machines, same denotational event, different transition systems, different halting states.

4. The Deep Lie: Type Soundness Failure in a Post-Truth Era

The previous examples assume the legal machine is functioning as designed, albeit struggling with epistemic limits or localized procedural attacks. But what happens when the machine’s operators reject the denotational goal entirely?

Historian of science Meera Nanda draws a distinction, in her analysis of India’s post-truth condition, between two distinct failure modes – a distinction this section adopts and re-derives in type-theoretic terms.1 The first, which she calls the “Big Lie,” is outright factual denial repeated regardless of evidence – the prototype being claims of a stolen 2020 US election, debunked by every audit and court that examined them, yet still believed by a large share of the public. A standard lie of this kind – perjury, for instance – is an operational violation: the witness injects a Void (Falsehood) into the evidentiary context (\(\Gamma\)). The system recognizes it as a violation because the type system (the rules of evidence) is still intact. Cross-examination (Cut Elimination) can isolate and expel the Void.

Nanda’s second category, the “Deep Lie,” is structurally different and harder to debunk, because it does not merely deny a fact within an otherwise-intact evidentiary system – it rewrites the rules for what counts as evidence in the first place. Her concrete example is India’s National Education Policy (NEP) 2020, which mandates incorporating “Indian Knowledge Systems” into curricula from primary school through the IITs. As a case study, she cites a 2021 calendar published by the newly inaugurated Centre of Excellence for Indian Knowledge Systems at IIT Kharagpur, which presented reinterpreted Indus Valley Civilisation artefacts – including the unmarked “unicorn” seals found at Harappa – as evidence for a Vedic-origin historical narrative, with no engagement with the archaeological or dating problems that reinterpretation raises.

In type-theoretic terms, this is a denotational attack, not an operational one – the equivalent of Type Erasure combined with axiomatic hijacking.

In a sound type system, a term \(e\) can only inhabit type \(A\) (Guilty) if it is constructed via valid rules. Nanda’s Deep Lie operates the same way: it does not smuggle a false fact past an intact evidentiary system, it asserts that a different category of testimony – ideological or theological narrative – is itself a valid type constructor for any proposition, on par with verified empirical evidence. It attempts to rewrite the operational semantics of the legal machine so that \(\Gamma \vdash \text{Ideology} : \text{Truth}\).

When an institution accepts this move, it suffers a systemic Type Soundness failure. The legal machine no longer evaluates evidence against objective reality (\(\mathcal{E}(e)\)); instead, it evaluates evidence against ideological coherence. On Nanda’s account, once this shift takes hold, empirical facts that contradict the preferred narrative are treated the way a Deep-Lie-captured system treats any inconvenient evidence: as claims that need not be forced to normal form at all, because the system has stopped recognizing normalization as the relevant test. The machinery of justice – courts, investigative agencies, and public discourse – undergoes a forced paradigm shift where unsafeCoerce is no longer a bug to be sanctioned, but a feature weaponized by the state to transition the machine to a politically predetermined halting state.

The Adequacy Theorem breaks down entirely here. It is no longer a gap between operational capabilities and denotational truth; it is a severing of the link. The machine still outputs a verdict, but the output is purely syntactic – a string matching “Not Guilty” or “Guilty” – devoid of any semantic mapping to historical reality. The law becomes a purely syntactic shell game.

5. Sequent Calculus: The Bridge Between the Two

The goal of a trial is to derive the sequent: \(\Gamma \vdash \text{Guilty}\).

The Cut rule is the operational engine of the prosecution’s case. The prosecution introduces lemmas – \(A\): “DNA matches”, \(B\): “Motive established” – to step toward the goal. Each lemma is a Cut: a derived fact inserted into \(\Gamma\) as a stepping stone.

Cross-examination applies Cut Elimination. In proof theory, Cut Elimination (normalization) shows that any proof using intermediate lemmas can be rewritten into a direct proof from the raw context alone. Cross-examination does the same: it strips the prosecution’s scaffolding to test whether the raw operational facts in \(\Gamma\) directly entail Guilty without the intermediate constructions. If they do not survive normalization, the lemmas were load-bearing fictions, not derivable facts. (In a post-truth environment, however, Cut Elimination fails because the ideological axioms refuse to be normalized by empirical challenges.)

The final verdict function. The final verdict function is a classical mapping imposed on an intuitionistic process. The operational machine evaluates the context \(\Gamma\) and attempts to construct a proof of Guilty. If it fails, it yields Nothing (an unproven state). However, the law (relying on the Law of Excluded Middle) forces a classical type cast: it takes the intuitionistic absence of a proof (Nothing) and casts it to a classical False, outputting “Not Guilty.” The law refuses to admit that it cannot compute the truth; it forces a halting state by legally equating “unproven” with “false.”

Summary

PL Concept Law Equivalent
\(\mathcal{E}(e) = v\) Objective historical reality
\(\Gamma, e \to \Gamma', e'\) Evidence rules (IEA / FRE)
Adequacy theorem Verdict = truth
unsafeCoerce / Void Perjury, evidence tampering
Principle of Explosion Systemic corruption / wrongful conviction
Unsound type system Wrongful conviction
Cut rule Prosecution’s evidence chain
Cut elimination / NbE Cross-examination, hearsay rejection
Classical vs. constructive Civil vs. criminal burden of proof
Type cast (Nothing -> False) “Not Guilty” verdict
Non-termination / Void Hung jury, mistrial
Two operational machines Criminal vs. civil trial
Type soundness failure Post-truth “Deep Lies” / institutional capture

The adequacy gap – the distance between what the operational machine can compute and what the denotational truth actually is – is not a failure of law. It is a consequence of epistemology. We have no direct access to \(\mathcal{E}\). We only have the machine, and the machine is all we have ever had. The rules of evidence are the type system that keeps the machine from hallucinating a verdict out of an unsound context. That is why perjury is not merely a procedural violation, and why post-truth “deep lies” are not merely political spin: they are attacks on the only instrument justice possesses for approximating truth.

References

  1. Meera Nanda, “Big Lies and Deep Lies in Post-Truth India,” The Wire, March 13, 2022.↩︎

Webmentions

Leave a comment

Comments are verified via IndieAuth. You will be redirected to authenticate before your comment is published.