As we all know, printed books are made up of parts with which we are very familiar and know almost instinctively how to use.

We also almost instinctively repress any knowledge that the physical and paratextual parts of a printed book contribute to its meaning.³

And, of course, just as there are parts of a printed book, digital editions are comprised of parts:

No single edition has to have all these parts, of course. What is almost more important than the parts seen by users — readers of the digital edition — are the parts of it at work behind the scenes. Categorizing data into code elements manipulated by programming, search categories, and/or database fields channels interpretive possibilities.⁴

After our experience in the aughts with digital editions that no longer work because code has changed, we want to make sure that the digital edition is encoded for longevity. It needs to be "archival quality," by which I mean only that libraries could take the digital edition into their collections as a digital asset.

Beginning with the California Digital Libraries and the Brown Women Writers Project (now at Northeastern University), scholars have agreed upon, used, and together developed code for digitizing texts in order to preserve documents crucial to understanding our cultural heritage.

The TEI code used to create digital editions has been developed by the Text Enoding Initiative Consortium. The group has created tag, element, and attribute names written in XML code (eXtensible Markup Language).

Why did digital humanists decide to create tag names and encode documents in XML?⁵ XML code is readable by humans as well as machines — computers executing programs — which means that, even if no one is alive who knows how the code works, they will be able to understand what it contains and, better yet, be able to easily re-purpose it for use in newer technologies.

XML code consists of words, letters, and numerals that are most often not arbitrary — that make sense semantically to human readers. Therefore, in order to distinguish the content of the edition (poetry, drama, etc.) from code, codes are enclosed in angle brackets.

There are three types of code inside angle brackets:
element or tag
attribute
value
A value is text that adds information to the element. In some cases, the attribute value is a Path.

The image above is slightly misleading insofar as, in XML, tags must be opened and then closed using a forward slash:

<div type="essay">...some portion of the digital edition typed here...</div>

Library quality, scholarly digital editions are built in XML (eXtensible Markup Language). A minimal XML document contains a root element with another element or tag nested inside.

Here you see a div tag, short for "division." In XML, every tag is opened and then closed. Close tags are just like open tags—in both cases, angle brackets enclose the element name. But close tags are distinct insofar as they also contain a forward slash just before the element name. The content, the text of your edition, goes inside—not inside the angle brackets, but rather between the open and close tags.

To be clear: the element name that is part of the XML code goes inside angle brackets: <title>. The chunk of text marked by the element goes between the open and close tags:
<title>The Handmaid's Tale</title>

XML code must also be properly nested, as explained in detail in the textbook dictionary.

To the left, the code says that a division (div) contains a poem with a title at the head of it made up of line groups (lg) that are stanzas.

Here is what a bit of XML code — coded according to the TEI Guidelines — looks like in the oXygen software often used by digital editors.

XML code describes features of the text that it represents; it doesn't care at all about presenting a pleasing, readable view of it to readers.⁶ As anyone knows who has ever used WordPerfect, our apps and software obsolesce as soon as the commpanies that create them go out of business. Because XML is not owned by any company, and because it is sheerly descriptive, it requires the use of programming to transform it into something presentable, readable, and usable for the software of the moment: right now, browsers (Chrome, Safari, etc.).

The code used to create web pages, while in the same family, differs from XML. But using a computer program to transform XML into another language for the sake of presenting it online seems like a waste of time.

Why not just code your digital edition in the code that browsers can read and present to viewers, skipping altogether the laborious process of coding in XML? Why do libraries prefer storing TEI/XML, and why do granting agencies often require it?

Because in 20, 30, or 100 years, instead of reading web pages on the Internet, we may be inserting chips into our brains. XML code can be transformed by a computer program into the code on those chips, whatever it may be.However, the code used to create web pages, while in the same family, differs from XML.

You may already know that your browser, whether Chrome, Safari, FireFox, or Edge, is a piece of software. It reads code and then transforms that code into the pages you see when searching the web.

You can open any files that live on your own computer using your browser, but it won't make much sense of them unless they are written in Hypertext Markup Language or HTML.⁷ Immediately below, you can see the HTML code for an edition of Robert Bloomfield's letters and poems: code on a server, identified by the URL or web address, is displayed in a browser for user-readers.

Yes, Bloomfield did spell imagination with two m's!! The screen shows a web page because your browser is reading the HTML from a server that is accessible to the Internet.

Although the two markup languages are in the same family, HTML code differs from XML code: the only HTML tag that indicating that a group of words constitute a title appears in the head of the HTML document. This "title" appears on a browser tab, not in the document itself. The content that will appear ON the web page is in its "body."

But instead of "title" and "author" tags inside the HTML "body," we have only a heading one (h1) code telling the browser to make the words inside the tag appear as the biggest words on the page, and a heading two (h2) code telling browser to make the words big but not as big as h1. Furthermore, the heading tags (h1, h2, h3, h4, etc.) may contain ANYTHING. A progam searching the web will have no idea that either of these tags contains a title or an author.

The browser software makes the HTML code above look like this image (to the right) whether you open the web page — that is, a page that ends with .html⁸ — on your computer, as I have done here, or whether you, an anonymous user, go to the HTML / web page on the Internet.

Stick with me; we're coming down the home stretch. Only two more components of creating digital editions remain to be discussed.

The plain web page for The Handmaid's Tale, above right, looks pretty awful. There is a kind of code for adding color and style: Cascading StyleSheets or CSS. To the right, you can see the effects of the CSS code I wrote for this web page.

If you go to the CSS Zen Garden website, you will find a host of web pages that look VERY different:

As hard as it is to believe, each one of these pages is made from EXACTLY the same HTML code and contents. Only the CSS code differs. The site is designed to let you download all the css files to see the code created by the web designers, and I have stolen bits from here and there. But using other people's code, referred to in the code itself, is a compliment.

The LAST component needed for creating an archival-quality digital edition is the programming that will transform the XML document created using TEI tags into the HTML document which will "call" the CSS files you have created to style it.⁹

The programming language typically used is XSLT (eXtensible Stylesheet Transformation Language), a language which makes use of XPath. XSLT can transform an XML page into HTML, text files, metadata records, spreadsheets, ePubs, and PDFs.

This textbook is introductory: you need only learn TEI encoding in order to use the DigEd system to publish your digital edition. In other words, you do NOT have to learn to code in HTML and css, nor to create XSL Transforms, for three reasons:

1. The software that we will use for coding in TEI, namely oXygen, contains built in XSLTs, and this textbook teaches you how to use them.
2. This open access textbook is written for people who do not want to become web designers or programmers, offering CSS and XSLT files (already made) to use and/or tweak for your digital edition (the "DigEd system").
3. The textbook comes with support from its author: when the code and programs in the DigEd system don't work as you and I expected, I will troubleshoot them with you.

That said, this textbook teaches XSLT programming as well as coding in TEI/XML, HTML, and CSS. A beginner might simply wish to use the codes and programs provided here without trying to understand how they work.¹⁰ An intermediate learner may wish to go through the tutorials so that they can learn to "read" the code and programming, and perhaps tweak it for their own needs, but not actually write any of it from scratch. But you can, based upon these tutorials, write in TEI / XML, HTML, CSS, XPath, and XSLT. And expertise can be achieved through practice as well as making use of the many resources available on the Internet for learning to code and program, discussed in the Introductory Essay.

---

To summarize, the image below shows the full digital editing workflow: a TEI / XML master document is created which represents the physical documents that you are editing. The TEI master is then transformed into HTML, the web page displayed to viewers visiting your website, rendered aesthetically pleasing with the help of CSS.