The Concern
The English reading sometimes breaks one Chinese sentence into several English lines. At other times, one English line gathers a relation that is distributed across more than one Chinese clause.
This is not merely a display problem. If the page pretends every English line maps cleanly to one Chinese sentence, it creates a false witness. The reader may infer a one-to-one relation that the text itself does not support.
The Formal Problem
The object to be represented is not simply a sequence of lines. It is a set of relations among source clauses, translation lines, concepts, predicates, and editorial decisions.
A responsible alignment must distinguish source segments, English reading lines, semantic units, syntactic units, editorial grouping decisions, and claims of equivalence, dependency, or partial correspondence.
The Proposed FOL Solution
First Order Logic can make the alignment accountable by turning each pairing into a claim rather than a layout assumption.
This does not require claiming perfect identity. It allows weaker, more honest relations such as PartialAligns, Expands, Condenses, Splits, Groups, and UnresolvedAlignment.
Editorial Practice
The immediate page should remain readable. The formal layer should make the invisible decisions inspectable.
- Preserve the Chinese source order.
- Preserve the English reading order.
- Give each segment a stable address.
- Declare whether an alignment is direct, partial, expanded, condensed, grouped, split, or unresolved.
- Allow later correction without breaking the published witness.
In AMOS terms, this treats alignment as a record of accountable transformation, not as a decorative convenience.
Next Implementation Shape
A later edition can add hidden or visible identifiers to each row, such as zh-de-38-001, en-mlt-de-38-001, and align-de-38-001. Those identifiers would let the page support a future alignment table, proof layer, JSON record, or challenge register without rewriting the reading itself.