Where should a document break?

What information does a document boundary separate?

Make your own investigation

Author a small source corpus and query. Compare chunk boundaries and record which qualifiers survive retrieval.

Split documents into overlapping word windows and rank lexical query overlap; inspect each top chunk with its source.

Case cards

Cut these cards apart. Use the original example first, then let a partner supply a changed case. Keep sealed/final cards with the teacher until the learner commits a rule.

Case 1

Document: Visitor guide

Text: Tours leave at nine. When rain is heavy, tours are cancelled.

My prediction and reason:

Case 2

Document: Ticket guide

Text: Entry costs eight dollars. Children enter free on Sundays.

My prediction and reason:

What this cannot establish

Lexical retrieval is not semantic understanding and can separate a qualifier from a claim.