Close Reading with TEI (Text Encoding Initiative)

This worksheet, and all related files, are released CC-BY.

By M. H. Beals; Adapted for HIST3814o by S Graham

  1. You will need the files in the Crafting Digital History GitHub module 3 folder. You need the folder tei-hist3907.
  2. Select the Clone or Download button and choose download zip to download that repository as a zip file.
  3. Unzip it somewhere handy on your machine. Inside will be the subfolder named tei-hist3907.
  4. Open the subfolder named tei-hist3907.

This exercise will explore a historical text and help you create a digital record of your analysis.



Vetting a Website

  1. Visit the Recovered Histories Website.
  2. Examine the site's layout and read its introduction.
  • What makes you believe this site is a trustworthy provider of historical texts?
  • What makes you believe this site is NOT a trustworthy provider of historical texts?

Finding a Source

  1. Visit the site's collections via the 'Browse' function.
  2. Locate the pamphlet Negro Slavery by Zachary Macaulay and open it.

This is an abolitionist pamphlet regarding the Atlantic slave trade, presenting and examining evidence of how it is run. When you approach a primary source like this, it is tempting to read through it from beginning to end, to get an overview of its contents, and then 'mine' or 'cherry-pick' good quotations to include in your assessments. However, we are going to focus on examining a very small part of the text in a very high level of detail.



Setting Up Your Workspace

You will use your own machine rather than DH Box for this work.

  1. Arrange your workspace so that you have the scanned text of the pamphlet easily visible on one side of your screen.
  2. Open the blanktemplate.txt file in Sublime Text, Atom, Textwrangler or Notepad++ (or any text editor that understands encoding) and have that on the other side of the screen.

The last lines will be </body></text></TEI></teiCorpus>. Everything you write today should be just above </body> tag.

Transcribing Your Page

  1. Go to the following tag:

    <biblScope>1</biblScope>
    

    and replace the number one (1) with the page number you are transcribing. Which page should you transcribe?

  2. Select a page in the document that you find interesting.

  3. Next, you will need to very carefully transcribe your page of text from the image into your document. Make sure you do not make any changes to grammar or structure in the text, even if you think the author has used poor grammar or misspelled a word. You do not need to worry about line breaks but should start every new paragraph (or heading) with a <p> and end every paragraph (or heading) with a </p>.

  4. Once you have completed your transcription, look away from your computer for 30-45 seconds. Staring into the distance every 10-20 minutes will keep your eyes from straining. Also, shake out your hands at the wrists, to prevent repetitive stress injuries to your fingers.



Encoding Your Transcription

You are now going to encode or mark-up your text.

  1. Re-read your page and highlight / colour the following:

    • Any persons mentioned (including any he/she if they refer to a specific person)
    • Any places mentioned
    • Any claims, assertions or arguments made

    Now that you have highlighted these, you are going to put proper code around them.

  2. For persons, surround your text with the following:

    <persName key="Last, First" from="YYYY" to="YYYY" role="Occupation" ref="http://www.website.com/webpage.html"> </persName>
    
    • Inside the speech marks for key, include the real full name of the person mentioned
    • In from and to, include their birth and death years, using ? for unknown years
    • In role, put the occupation, role or 'claim to fame' for this individual.
    • In ref, put the URL (link) to the Dictionary of National Biography, Wikipedia or other biography website where you found this information. If there is a & in your link, you will need to replace this with &amp;.



  3. For places, surround your text with the following:

    <placeName key="Sheffield, United Kingdom" ref="http://tools.wmflabs.org/geohack/geohack.php?pagename=Sheffield&params=53_23_01_N_1_28_01_W_type:city_region:GB"> </placeName>
    
    • In key, put the city and country with best information you can find for the modern names for this location
    • In ref, put a link to the relevant coordinates on Wikipedia GeoHack website. To obtain this, go to the Wikipedia page for this city and click on the latitude/longitude coordinates for the location. For large areas, such as entire countries or continents, just use the Wikipedia page URL.



  4. For claims or arguments, surround your text with the following:

    <interp key="reason" n="citation" cert="high" ref="http://www.website.com/webpage.html"> </interp>
    
    • In key, explain why you believe this claim is true or not
    • In n, put a full citation to the relevant source
    • In cert (short for certainty), put: high, medium, low or unknown
    • In ref, put the link to the website where you got the information to assess this claim.

    When you are happy with your work, hit save your work, give it a useful name, make sure it has .xml as the extension, and save it and the .xsl file to your repository.



Alex Gill has made The Short and Sweet TEI Handout which you might want to explore as well. When you embark on encoding documents for your own research, Northeastern University has some questions to think about to help you decide what kinds of tagging you'll need; these templates from HisTEI might be useful (open the whole project with OxygenXML for full functionality, but you can copy those templates in any editor).

Viewing Your Encoded Text

To see your encoded text, make sure your .xml and .xsl file are in the same folder. Open either Internet Explorer or Firefox. The following will not work in Google Chrome because it has different security settings.

Making sure both your (page number).xml file and your 000style.xsl file are in the same folder (or both on your desktop), drag the icon for (page number).xml into your browser window.

If you now see a colour-coded version of your text, Congratulations! If you hover over the coloured sections, you should see a pop-up with the additional information you entered.

If your text comes up only in black, with no paragraph divisions or headings, or doesn't come up at all, something has gone wrong. Re-open your .xml file and check that you have:

  • Placed <p> at the start of every paragraph, including the start of the page.

  • Placed </p> at the end of every paragraph, including the end of the page.

  • Made sure all your <persName>, <placeName> and <interp> tags are properly enclosed in <>s.

  • Made sure you have both an open <> and close <\> tag for each tag you use.

  • Made sure you attribute values are fully enclosed in "".

  • Made sure you have a space between the " of one attribute and the start of the next attribute.

  • Made sure you do NOT have a space after the = of an attribute.

If your text still does not appear formatted, you may need to remove the text one paragraph at a time (pasting it somewhere handy), refreshing your browser window, until it appears. This will help you identify which paragraph (or sentence) has the error within it.



If you still don't see your text

If you do not see the colour-coded version of your text, this might not necessarily mean that you've done something wrong. Some browsers will not perform the transformation, for security reasons.

In which case, we can do the following:

  1. If you are on a Windows machine using Notepad++, go to 'Plugins' >> Plugin Tools. (If you are on Windows but aren't using Notepad++, Sublime and Atom probably have a similar functionality, but you will have to search to figure it out.)
  2. Select 'XML Tools' from the list, and install it.
  3. You'll probably have to restart the program to complete the plugin installation.
  4. Open up the 1.xml file in Notepad ++.
  5. Under 'Plugins'>>'xml tools" select 'XSL Transformation settings'.
  6. In the popup, click on the elipses: ... to open up the file finder, and select the 000style.xsl stylesheet.
  7. Click 'transform'. A new tab will open in Notepad++ with a fully-formed html file displaying your data according to the stylesheet.
  8. Save this new file and open it in a browser!

You can also check 'validate' from the XML Tools menu in Notepad++, which will identify errors in your XML. If you're still having errors, a likely culprit might be the way your geographic URLs are encoded. Compare what you've got with what's in the 1.xml reference document.

Advanced: If you install a WAMP or MAMP server, and put your xml and xsl files in the WWW folder, you should be able to see the transformation no problem at localhost\myxml.xml (for example). (You can also use Python's built in webserver if you have Python on your machine — all Mac users for instance do.)

You can access the CND.xml, transformed into a CSV on my GitHub. If you right click and choose 'View page source', you'll see the original XML again! Save-as the page as whatever-you-want.csv and you can do some data mining on it.

More on transformations

I made a file I've called SG_transformer.xsl. Open that file in your text editor. What tags would it be looking for in the XML file? What might it do to your markup? What line would you change in your XML file to get it to point to this stylesheet? Write all this down in your open notebook. It is a good habit to get into to keep track of your thoughts when looking at ancillary files like this.

If the nature of your project will involve a lot of transcription, you would be well advised to use an XML editor like OxygenXML, which has a free 1 month trial. The editor makes it easy to maintain consistency in your markup, and also, to quickly create stylesheets for whatever purpose you need. There are also a number of utility programs freely available that will convert XML to CSV or other formats. One such may be found online on Google code. But the best way to transform these XML files is with XSL.