Text only | Skip links
Skip links||IT Services, University of Oxford

1. Before you start

You will have just had a brief introduction to the <teiHeader>, metadata, and in particular the <msDesc> element for storing bibliographic and descriptive information concerning manuscripts.

The poem In Flanders Fields by Lt. Col John Alexander McCrae was published, unattributed, in Punch magazine on December 8 1915. A manuscript copy of this poem, as submitted to Punch, is held by Libraries and Archives Canada.

We have an image of that manuscript copy where it appears some unknown scribe has further edited the poem (to its detriment) in red ink. Our task in this exercise is to create a <teiHeader> for a file where we will later (in exercise 6) transcribe the poem including its bizarre changes. For this exercise we are only creating the metadata including a description of the manuscript. The image is provided in your booklet on page ??.

2. Opening our starting file

  1. A minimal starting file has been created for us called ex3.xml in your Work folder. If you open this file one of the first things we hope you'll notice is that we are no longer validating against our normal Imaginary Punch Project schema. Instead we're validating against one called tei_ipp_ms because we need the manuscript description (and later transcription) elements available to us.
  2. You will also notice that we don't have much information to base our header and manuscript description on. The <sourceDesc> gives us a nice big prose paragraph of:
    <p>A single paper leaf presentation version of "In Flanders
    Fields," written in John McCrae’s hand, is found in the E.W.B.
    Morrison collection in Library and Archives Canada, Ottawa,
    Ontario, Canada. Major-General Sir Edward Morrison, was a
    newspaper journalist, who headed the Royal Canadian Artillery
    during the First World War. He was a friend of the poet,
    physician and fellow gunner McCrae. It is believed that McCrae
    presented Morrison with this copy of the poem. The original in
    container 16 has been withdrawn from circulation. Researchers
    are required to consult photocopies in container 10. A digital
    copy of this item is available. There is also a contact print,
    number C 26561. Ref Num: MG 30 E 81. </p>
  3. And the rest of the header is pretty meaningless. Your job is to make it into a decent header with a manuscript description.

3. Let's start with the header

  1. First change the title to something better. How about ‘In Flanders Fields’ or because this is the title of the electronic version you might prefer ‘A digital edition of "In Flanders Fields" by John McCrae’.
  2. Rather than giving the author's name within the title, you can supply a <author> element as a child of the <titleStmt> element.
  3. Moving on to <publicationStmt>, delete the paragraph that this contains and replace it with individual elements for:
    • <authority>: put your name here
    • <availability>: put a paragraph here stating that it is release under a Creative Commons Attribution license
    • <idno>: give the work an identification number while thinking about appropriate ways people give identification numbers for resources
    • <publisher>: If you don't want to claim it for yourself, claim that ‘TEI@Oxford’ is the publisher!
  4. Double check that your document is still valid. You should have no red underlining and a green square in the upper righthand corner. Remember to save your work occasionally.

4. Describing the manuscript

  1. Currently we have a paragraph as the contents of <sourceDesc>. To do this our document will have to be invalid for awhile.
  2. Wrap the <p> element inside <sourceDesc> with an <msDesc> element. Immediately, you will notice that the schema complains that you are missing a <msIdentifier> element.
  3. Add an <msIdentifier>element before the paragraph. Your document should become valid again, though it is pretty meaningless without child elements inside <msIdentifier>. Add some for country, settlement, repository, and idno based on the information in the prose paragraph. Mine looks something like:
    <sourceDesc>
     <msDesc>
      <msIdentifier>
       <country>Canada</country>
       <settlement>Ottawa</settlement>
       <repository>Library and Archives Canada</repository>
       <collection>E.W.B. Morrison</collection>
       <idno>MG 30 E 81 v. 16</idno>
      </msIdentifier>
      <p>A single paper leaf presentation version of "In Flanders
         Fields"...</p>
     </msDesc>
    </sourceDesc>
  4. When your document is valid and save your work occasionally.

5. Describing the manuscript: <msContents>

  1. Now we will temporarily make our document invalid again. Immediately after the close of the <msIdentifier> element insert an <msContents> element.
  2. Inside the <msContents> element add a child <msItem> element to describe the intellectual content of the items in our manuscript. Ok, we only have one item in this manuscript so we will only have one <msItem>, but you could have more of them if this was a longer manuscript.
  3. Inside the <msItem> and child elements for <author> and <title> filling them in appropriately.
  4. After the <title> add a <textLang> element with a mainLang attribute with the value of en; the content of the element should be English.
  5. After this element add a <note> element with the text A single paper leaf presentation version of "In Flanders Fields," written in John McCrae’s hand
  6. Mine looks something like:
    <msContents>
     <msItem>
      <author>John Alexander McCrae</author>
      <title>In Flanders Fields</title>
      <textLang mainLang="en">English</textLang>
      <note> A single paper leaf presentation version of "In
         Flanders Fields," written in John McCrae’s hand</note>
     </msItem>
    </msContents>
  7. Now your document will not be valid because you are halfway between creating a structured record (using <msContents>) compared to an unstructured record (just having a <p>). But it shouldn't be complaining about anything in the <msContents> part of the document before you move on. Save your work anyway!

6. Physical description

  1. Now we're going to add the structured information about the physical description of this single leaf of paper.
  2. After the closing <msContents> tag insert a <physDesc> element.
  3. Inside this put an <objectDesc> element with a form attribute whose value is leaf.
  4. Inside the <objectDesc> put a <supportDesc> with a material attribute whose value is paper. And inside this element a child <support> element with the human-readable content A single paper leaf.
  5. After the closing <objectDesc> tag, add a <handDesc> element.
  6. Inside this put two <handNote> elements one after the other. Give the first one an xml:id attribute value of jMcCrae and the second an xml:id attribute value of otherHand.
  7. In the 'jMcCrae' <handNote> add the content The hand of John Alexander McCrae. In the 'otherHand' one add content like 'The unknown hand of a later scribe whose suggested corrections would ruin the poem'
  8. Mine looks something like:
    <physDesc>
     <objectDesc form="leaf">
      <supportDesc material="paper">
       <support>A single paper leaf</support>
      </supportDesc>
     </objectDesc>
     <handDesc>
      <handNote xml:id="jMcCrae">The hand of John Alexander
         McCrae</handNote>
      <handNote xml:id="otherHand">The unknown hand of a later
         scribe whose suggested corrections would ruin the
         poem</handNote>
     </handDesc>
    </physDesc>
  9. Your document still won't be valid because of that paragraph below, but leave it there for now, you can see that we're extracting all the intellectual content from it and putting it in the correct places.

7. Making history

  1. The <history> element allows you to detail the history of the manuscript from its origin, through its provenance to its acquisition by the resource-holding institution. In this case we mostly have information about its origin.
  2. After the end of the <physDesc> element create a <history> element.
  3. Inside the <history> element create a <origin> element.
  4. Inside the <origin> element we want to put the text which refers to where this copy of this manuscript came from. In this case the text of the paragraph from 'Major-General Sir Edward' to 'this copy of the poem.' would be a good thing to put in here.
  5. Usually inside <origin> we might include <origDate> or <origPlace> to give the date or place of origin, however we've not clearly been given these. I've addded a sentence that says ‘It is dated Dec 8 1915 but this is the date of publication in Punch, probably not the original date of this copy.’ I've also marked up the date as a<date>and 'Punch' as a <title>.
  6. Mine looks something like:
    <history>
     <origin>Major-General Sir Edward Morrison, was a newspaper
       journalist, who headed the Royal Canadian Artillery during
       the First World War. He was a friend of the poet, physician
       and fellow gunner McCrae. It is believed that McCrae
       presented Morrison with this copy of the poem. It is dated
     <date when="1915-12-08">Dec 8 1915</date> but this is the
       date of publication in <title>Punch</title>, probably not
       the original date of this copy. </origin>
    </history>
  7. Your document still won't be valid because of that paragraph below, but leave it there for now. Remember to save your work.

8. And everything else goes in Additional!

  1. <additional> is used to store administrative information, details of surrogate copies, or bibliographic information for works concerning the whole of the manuscript.
  2. After the closing <history> tag create an <additional> element. Inside this create an <adminInfo> element and inside that an <availability> element.
  3. The availablity element stores information concerning the legal or physical restrictions on the manuscript. Inside it put a <p> element with the sentence from the original paragraph below: ‘The original in container 16 ... photocopies in container 10.’
  4. We could have also put this information where we store data about surrogates of the manuscript, but in this case it seems mostly about access control. However, we do have information about surrogates, in this case that there is a digital copy and a contact print available.
  5. After the close of the <adminInfo> element put in a <surrogates> element.
  6. Inside this put two <note> elements with the text from the paragraph that we've not used anywhere else in the manuscript description. I have one that says ‘A digital copy of this item is available’ and another that says ‘There is also a contact print, number C 26561.’
  7. Mine looks something like:
    <additional>
     <adminInfo>
      <availability>
       <p>The original in container 16 has been withdrawn from
           circulation.Researchers are required to consult
           photocopies in container 10. </p>
      </availability>
     </adminInfo>
     <surrogates>
      <note>A digital copy of this item is available. </note>
      <note>There is also a contact print, number C 26561.</note>
     </surrogates>
    </additional>
  8. We could also have put a <listBibl> with multiple <bibl> elements inside if these notes had been more like bibliographic citations for facsimiles of the manuscript.

9. Making it valid

  1. The document still isn't valid because of that original paragraph at the bottom. TEI manuscript description allows you to choose either a completely unstructured method, with multiple paragraphs, or a more structured method with the elements we have chosen. Why might you choose one over the other?
  2. After looking through that original paragraph to make sure you've put the information it contains somewhere in the manuscript description, delete that paragraph. This should make your document valid. If it doesn't, attempt to solve any of the problems, and save your work.

10. Other things to try

  • Although our <availability> statement said it was licensed as Creative Commons, consider also putting in a <ref> element with a target attribute which points to the appropriate Creative Commons license URL.
  • In recording the <handNote> elements we didn't indicate which had a scope of 'major'and which was 'minor'.
  • What else could you have put into this manuscript description? Check through elements like <physDesc> to see what else could be put into this description. Many of the elements don't apply to this very brief manuscript.
  • In exercise 6 we'll be transcribing the text of this manuscript. You can always start early if you have time left over.


TEI@Oxford. Date: 2010-07
Copyright University of Oxford