Searching an old medical text for familiar words can leave relevant passages out of view. A source may describe a similar complaint in different language, while a familiar character may play a different role in its sentence. I organized eye-symptom terminology and its source passages so that finding a term and understanding the passage could be examined together.
The work began with making the material easier to navigate. In developing the manuscript, I focused on whether the terminology and categories could be applied across texts, and whether a reader could trace each judgment back to its source. The current manuscript concerns source-based terminology research; it does not test the effectiveness of screen-reader access.
A term match and a passage category answer different questions
Collecting passages that contain known search terms is a quick way to build a corpus. It can also exclude relevant material before the reading begins. I kept lexical matching separate from passage classification: one asks whether a specified expression occurs, while the other considers what the local passage describes.
That separation avoids treating a failed search as evidence that relevant content is absent. It also leaves room for a passage containing a familiar word to fall outside an expected category. Records that did not fit, or could not be classified securely, retained a reason. The categories are annotations for exploring historical texts, not claims that an old expression corresponds to a modern diagnosis.
Keeping a route back to the source
Preparing text for reading can remove differences in character forms or spacing. To keep those editorial choices inspectable, I separated the raw passage from the normalized display form and linked it to a location in the source. A category can then be reconsidered without losing the passage that prompted it.
AI assistance was used in structuring and classifying the material. Checks of source positions and strings help establish where a passage came from; they do not prove that its meaning has been interpreted correctly. I concentrated on preserving the evidence another researcher would need to read the source and examine the classification.
After the major revision
I submitted the strengthened manuscript to Digital Chinese Medicine on 5 September 2026. The status confirmed that day was With Editor. Submission of a revision is separate from acceptance.
Detailed findings and terminology data remain outside this public account while the manuscript is under consideration. Once a version of record and DOI are confirmed, I plan to describe the methods and examples that can be shared under the source material’s rights conditions.