Discussion

Thirty-one years of deciding what is ‘reasonable’, and no degrees of truth in sight

March 5, 2026 · 5 replies · 150 views

Law Analytic Islamic Islamic jurisprudence H. L. A. Hart Lon Fuller #degrees-of-truth #legal-thresholds #reasonableness #vague-standards

I spent thirty-one years on the bench in Punjab, the last ten as a sessions judge. A large part of the work was applying words that have no sharp edges: "reasonable doubt", "sudden provocation", "sufficient cause", "unsound mind". Now I read that fuzzy logic was invented for exactly this kind of vagueness, and I want to test that claim against what a judge actually does.

Here is my difficulty. The standard is vague, but the verdict is not. The accused is convicted or acquitted. I never once found that a doubt was "0.7 reasonable". What I did was draw a line in the case before me and give reasons for drawing it there, reasons that would have to be defended on appeal.

That looks less like degrees of truth and more like a forced sharpening, followed by an obligation to justify the sharpening.

The fiqh tradition I read in retirement knows this problem well. Many measures are left to custom ('urf), and the jurists argued at length about when custom may fix a boundary that the texts leave open. Hart, too, saw that legal language has an "open texture": a rule about vehicles in the park is clear for cars and unclear for a child's bicycle.

So, to the logicians: is there any place in law where degrees of truth genuinely fit? Or is the judge's job precisely to turn a vague standard into a crisp decision?

Like ReplyFollow5 replies

Members reply here. Reading is always free.

5 replies

Hamza Qureshi 7 months ago

Fellow

Your Lordship (forgive the habit), Fuller's reply to Hart is worth adding. Hart's vehicle case was meant to show that penumbral cases need judgement because the words run out. Fuller asked about a truck mounted on a pedestal as a war memorial: obviously a "vehicle", obviously not what the rule is about. His point was that even the clear cases are clear because of the rule's purpose, not its words. That makes the judge's sharpening less like picking a number and more like reading an aim.

Helpful · 3
Oskar Nyberg 7 months ago

Fellow

From vagueness theory: what you describe sounds like supervaluation followed by choice. The statute leaves a range of admissible sharpenings; the judge must pick one and give reasons; appeal courts police the range. That's not degrees of truth.

Where degrees do fit, I think, is sentencing and damages, where the output is itself on a scale. A sentence can be proportionate to a degree. A verdict can't be guilty to a degree.
Deepa Nair 7 months ago

Fellow

Indian constitutional law is full of this. Article 19 permits "reasonable restrictions" on freedoms, and the courts have built structured tests to decide reasonableness: legitimate aim, suitable means, necessity, balance. Each stage is still a yes or no, but the final balancing stage feels like a weighing of degrees. So maybe the degrees live inside the reasoning while the output stays binary.

Helpful · 1
Thandeka Mokoena 7 months ago

Fellow

The South African constitution writes that structure into its text. Its limitation clause asks whether a limit on a right is reasonable and justifiable in an open and democratic society, and lists factors to weigh. Deepa's point holds there too: we weigh in degrees and decide in categories. I'd add that the legitimacy of the decision depends on showing the weighing, which is something no fuzzy controller has to do.

Helpful · 2
Khalid Mahmood Awan 7 months ago

Fellow

Thank you all. Deepa and Thandeka have given me a better formulation than my own: degrees in the reasoning, categories in the verdict, and a duty to show the weighing. Oskar's point about sentencing is right too, and it reminds me that the fiqh tradition was most comfortable with discretion (ta'zīr) precisely in punishment, not in findings of fact. I had not seen the connection before.

Helpful · 3