Contrary-to-Duty Obligation

A CTD obligation is defined by its conditional trigger: the breach of a primary norm. The structure is a conditional sentence where the antecedent is a violation. For example: 'You ought to keep the gate closed. If you leave it open, you ought to post a warning sign.' The secondary obligation to post a sign only becomes active in the sub-ideal world where the gate is left open. This creates a temporal and logical dependency chain that standard monotonic logics struggle to represent without contradiction.

The classic illustration of CTD reasoning, formulated by Roderick Chisholm in 1963, exposes the inadequacy of Standard Deontic Logic (SDL). The paradox consists of four intuitively consistent sentences:

• It ought to be that Jones goes to assist his neighbors.
• It ought to be that if he goes, he tells them he is coming.
• If he does not go, then he ought not to tell them he is coming.
• Jones does not go. When formalized in SDL, this set derives a logical contradiction, proving that a more expressive logic—one capable of handling defeasible conditionals—is required for legal AI.

CTD obligations are inherently time-bound. The primary duty exists in an ideal temporal state, while the secondary duty exists in a sub-ideal, future state. For instance, a contract clause stating 'The Buyer shall pay by Day 30. If payment is late, the Buyer shall pay 2% interest per month' encodes a clear temporal sequence. A legal reasoning engine must model this discrete state transition—from the ideal world to the violation world—to correctly activate the interest obligation only after Day 30 has passed without payment.

Real-world legal instruments often contain recursive contrary-to-duty chains. A primary obligation may have a secondary remedy, which itself has a tertiary fallback if the remedy is also breached. For example:

• Primary: File tax return by April 15.
• Secondary (CTD1): If late, file by October 15 with a penalty.
• Tertiary (CTD2): If the extended deadline is also missed, pay an additional failure-to-file fee. Modeling these iterated preference structures requires a logic that can handle multiple layers of sub-ideal worlds without collapsing into inconsistency.

CTD obligations can be classified by their teleological function within a normative system:

• Compensatory CTDs: Aim to restore the counterparty to the position they would have been in had the primary duty been performed. Example: 'If goods are defective, the Seller shall repair or replace them.'
• Punitive CTDs: Impose a sanction that goes beyond mere restoration to deter non-performance. Example: 'If the non-compete is breached, the Employee shall pay liquidated damages of $50,000.' Distinguishing these types is critical for automated damages calculation and remedy selection.

CTDs are formally captured using a dyadic deontic operator O(B|A), read as 'It ought to be that B, given A.' The violation condition A represents the contrary-to-duty context. This contrasts with the monadic O(A) of Standard Deontic Logic. Modern approaches use:

• Preference-based semantics: Ranking possible worlds by their ideality, where CTD obligations hold in the best worlds among those where the violation occurs.
• Input/Output Logic: Treating norms as ordered pairs (antecedent, consequent) and defining operations for their detachment and iteration without logical explosion.