7AAVIN01 Chap.6 Encoding a Humanities Text: Columbus's Diary
Encoding a Humanities Text: Columbus's Diary
Encoding a Humanities Text: Columbus's Diary works through the main Week 2 workshop, in which groups mark up an English translation of Columbus's journal in stages. It covers document analysis, from metadata and high-level structure to chunk-level units, building a template in logical order, and in-line markup of places, people and other terms a reader might search.
It then distils the lessons of the class feedback: containers for repeating entries, no numbered element names, dates as content rather than names, neutral attributes, two ways to encode footnotes, and what to leave for software. The workshop is also an exercise in reflection, which the report rewards.
Groups divide the excerpt, agree element names first, compare their files and then aggregate them, and every change to the scheme is noted. The feedback shows how small naming choices decide whether software can sort, count or extract anything later: one former student's encoded directions could be extracted automatically with a short script.
In a real project you would usually adopt TEI or DITA rather than invent elements, and you would study the document's history, status and audiences before tagging. Worked examples mark up an invented diary sentence with places and people and propose a full document hierarchy, and a clinic page sets out the errors the class feedback meets most often.
What this chapter covers
- 01
Document analysis before tagging
- 02
Metadata and high-level structure
- 03
Templates built in logical order
- 04
In-line markup of places and people
- 05
Name or person
- 06
Containers for repeating entries
- 07
Footnotes and line breaks
- 08
Automating markup
Worked example · free
Structure a diary for XML
- 1Use one root, such as journal, containing a header for title and editor and a body.
- 1Inside the body, place the introduction, the entries and the notes in that order.
- 1Wrap each day in one repeated entry element rather than numbered names such as entry1.
- 1Put each date inside its own date element, or in an attribute of entry, never in an element name.
Key terms
- Document analysis
- The stage of identifying a text's metadata and structure before any markup is written.
- In-line markup
- Tags placed inside running text that mark words or phrases without breaking the flow of a sentence.
- Container element
- An element that groups a repeating set of related elements, such as a date and its entry text.
- Normalised reference
- An attribute value that gives a standard form for a name, such as a place spelt in another language.
- Element name
- The label inside a tag that says what kind of thing the element is, such as entry or place.
Encoding a Humanities Text: Columbus's Diary FAQ
Why should I not number element names in XML?
Names such as entry1 and entry2 are neither helpful nor sustainable, because software cannot treat them as one kind of thing. Use one element name and, if the original numbering matters, record it in an attribute.
Should I mark up a name or a person?
It depends on the question. Tagging a name captures a string shared by everyone with that name; tagging a person identifies an individual, even an unnamed one. Some encoders mark both, and the choice should be stated.
What are the two ways to encode footnotes?
You can mark pointers in the text and gather the notes in a separate section, or place each note's content where it belongs. The second lets software generate pointers and numbering automatically, at the cost of a busier editing view.
How should places be marked up so they can be searched?
Tag each place, add a type attribute such as city or region, and give a normalised reference so that variant forms, including names in other languages, point to one standard key. That lets software group every mention of the same place.
Should every line of a diary be tagged?
Usually not. Line tags suit poetry, where line endings carry meaning. In a prose diary the breaks are accidental and change with page width, so paragraphs are enough, and navigation links can be generated later rather than encoded by hand.
Can markup of a historical text be automated?
Partly. The feedback shows regular expressions one student used to tag notes, dates, directions and distances, offered with a warning that they were untested, and mentions projects that automated encoding. Automation still needs checking, because patterns misfire on irregular text.
Assessment move
Choose a short public-domain letter or diary page and complete both stages yourself: draw the hierarchy, then add in-line tags for places and people. Keep lab notes in two parts, practicalities and reflections, and use the reflections as raw material for the critical reflection section of your report.
Compare your file with a classmate's and list every difference in element names; each difference is a consistency problem a project would have to solve.