Lists


1. Lists with curly braces

This is a list of names, with some of the names bracketed and their rank or station given in the margin.

There are basically three ways to do this:

(1) Distribute the descriptions, i.e. repeat the term for each person to whom it applies. This is fine so far as it goes, especially with simple lists. The trouble with this approach, is that it often means turning plurals into singulars in order to make sense:

    <ITEM>Sir John Fennick. Knight.</ITEM>
    <ITEM>William Fennick. Esquire. Knight.</ITEM>

    etc.

    We can do this pl.->sg. change, but only if it can be done straightforwardly and only if it is the only practical way to deal with a lot of existing tagging.

(2) Use the descriptions as <HEAD>s of nested lists. This is usually the preferred method. Its only drawback is that unless you're willing to create little lists of one item each, it means treating labels applied to individuals differently from labels applied to groups:

  <LIST>
    <ITEM>
      <LIST><HEAD>Knights.</HEAD>
            <ITEM>Sir John Fennick.</ITEM>
            <ITEM>William Fennick. Esquire.</ITEM>
      </LIST>
    </ITEM>
    <ITEM>
      <LIST><HEAD>Burg.</HEAD>
            <ITEM>Robert Ellison. Esquire.</ITEM>
            <ITEM>John Scowen. Esquire.</ITEM>
            <ITEM>John Fiennis. Esquire.</ITEM
        </LIST>
      </ITEM>
   </LIST>
 

  (3) Finally, you can capture the appearance, rather than the logic, of the list by tagging the bracketed items as an embedded  TABLE. Apex is the vendor most likely to do this. There is nothing exactly *wrong* with this, except that is irritating in its agnosticism: it refuses to interpret so much that it misrepresents the original (which really is a list, not a table). Still, if it comes tagged this way, it is rarely worth the effort to change it:

<LIST>
 <ITEM>
   <TABLE>
   <ROW><CELL>
           <LIST>
              <ITEM>Sir John Fennick.</ITEM>
              <ITEM>William Fennick. Esquire.</ITEM>
            </LIST>
        </CELL>
        <CELL ROWS="2" ROLE="label">Knights.</CELL>
  </ROW></TABLE>
</ITEM>
 
 

Back to top

2. Genealogies as lists

 We wouldn't reject a book for resorting to <FIGURE>, but in the past when we've encountered genealogies in the books, we have adopted a suggestion made by Apex, namely to try to capture the content of these charts using nested LISTs. Siblings are items in a list headed by their parent; the next generation becomes a nested list, contained within the item headed by their parent, and so forth. So if Quentin had two sons (John and Bill), each son had two daughters (John had Joan and Jill; Bill had Betty and Beatrice), and Betty went ahead and had twins (Richard and Robert) the result would be:

<LIST>
<HEAD>Quentin</HEAD>
<ITEM>John
  <LIST>
  <ITEM>Joan</ITEM>
  <ITEM>Jill</ITEM>
  </LIST>
</ITEM>
<ITEM>Bill
  <LIST>
  <ITEM>Betty
         <LIST>
         <ITEM>Richard</ITEM>
         <ITEM>Robert</ITEM>
         </LIST>
   </ITEM>
   <ITEM>Beatrice</ITEM>
   </LIST>
 </ITEM>
 </LIST>

     Some genealogies include spouses (sometimes multiple spouses). It works to  tag those as <HEAD>s of lists of their children. In the sample above, if Betty in fact had four children, Richard and Robert by her first husband Randall, and Anne and Alice by her second husband Albert, that part of the list would be modified like this:

    <ITEM>Betty
         <LIST>
         <HEAD>Randall</HEAD>
         <ITEM>Richard</ITEM>
         <ITEM>Robert</ITEM>
         </LIST>
         <LIST>
         <HEAD>Albert</HEAD>
         <ITEM>Anne</ITEM>
         <ITEM>Alice</ITEM>
         </LIST>
  </ITEM>

Back to top


3. Format of tables of contents and indexes

(a) A table of contents appeared tagged as below, with each item tagged as a separate LIST--which is a definite misuse of list. If none of your lists contain more than one item, then they are not really lists.

<LIST>
<HEAD>LETTER I.</HEAD>
<LABEL><HI>TO Messer Perepollastre</HI>, an <HI>Italian</HI>,
the Author's Friend; exposing a Ca|lumny cast upon him, and
defending his Innocence.</LABEL>
<ITEM><HI>Page</HI> 1.</ITEM>
</LIST>

<LIST>
<HEAD>LET. II.</HEAD>
<LABEL>To Dr. <HI>Melgar</HI>, a Physician; in which are
handled the good and harm occasion'd by his Profession,
together with the Progress and several Interruptions of that
Art; as also the Author's thoughts of it from its first
Original.</LABEL>
<ITEM><HI>p.</HI> 11.</ITEM>
</LIST>

<LIST>
<HEAD>LET. III.</HEAD>
<LABEL>To a Lady, the Author's Neece, who fell sick
for the Death of a little Bitch.</LABEL>
<ITEM><HI>p.</HI> 32.</ITEM>
</LIST>


We simplified to this simple ITEM-ITEM-ITEM LIST:


<LIST>


  <ITEM><LABEL>LETTER I.</LABEL> <HI>TO Messer Perepollastre</HI>,
  an <HI>Italian</HI>, the Author's Friend; exposing a
  Ca|lumny cast upon him, and defending his Innocence.
  <HI>Page</HI> 1.</ITEM>



  <ITEM><LABEL>LET. II.</LABEL> To Dr. <HI>Melgar</HI>,
  a Physician; in which are handled the good and harm
  occasion'd by his Profession, together with the
  Progress and several Interruptions of that Art;
  as also the Author's thoughts of it from its first
  Original. <HI>p.</HI> 11.</ITEM>



  <ITEM><LABEL>LET. III.</LABEL> To a Lady, the Author's
  Neece, who fell sick for the Death of a little Bitch.
  <HI>p.</HI> 32.</ITEM>
  
. . .

</LIST>

Acceptable alternative would have been a LABEL-ITEM list
like this:

<LABEL>LET. III.</LABEL>
  <ITEM> To a Lady, the Author's
  Neece, who fell sick for the Death of a little Bitch.
  <HI>p.</HI> 32.</ITEM>
  
or simple <P>s:

  <P>LET. III.  To a Lady, the Author's
  Neece, who fell sick for the Death of a little Bitch.
  <HI>p.</HI> 32.</P>

(b) This table of contents started out as a two-column (LABEL-ITEM) list. We reduced it to a simple (ITEM only) list. Because:

(1) the chapter designations were being split across LABEL/ITEM, e.g. capitulo<LABEL><ITEM>ij.</ITEM>;
(2) some of the LABELs did not have corresponding ITEMs;
(3) the list makes perfect sense without the 2-column approach,
    and simpler is usually better.
    
<LIST>
<ITEM>The prologue declareth to whom this volume ap&abper;teyneth
and at whos requeste it was translated out of ffrenshe in to
englissh </ITEM>
<ITEM>After foloweth the &abpro;logue of the tra~slatour
declaryng the substau~ce of this present volume </ITEM>
<ITEM>After foloweth the book callid the Myrrour of the world
and speketh first of the power &amp; puissa~ce of god
<aBBR>Cap</ABBR>^o</ITEM>
<ITEM>Wherfor god made &amp; created the <aBBR>world</ABBR>
capitulo &para;.ij.</ITEM>
<ITEM>Wherfor god <aBBR>formed</ABBR> man to his semblau~ce
Ca &para;.iij.</ITEM>
<ITEM>Wherfor <aBBR>god</ABBR> made not man in suche wise as
he myght not synne capitulo &para;.iiij.</ITEM>
<ITEM>Wherfor and how the vij Artes liberal were founde and
of their ordre capitulo &para;.v.</ITEM>
<ITEM>Of thre maner of peple <aBBR>and</ABBR> how clergye
cam first in to ffraunce capitulo &para;.vi.</ITEM>
<ITEM>And first it speketh of gramaire capitulo &para;.vij.</ITEM>
<ITEM>After of logyke capitulo &para;viij.</ITEM>

(c) We have been looking at a lot of tables of contents and indexes lately, and have emerged with the following observations:

(1) The '2-column' (label-item) model of LIST is used more often than it really needs to be. We give preference to the simpler ITEM-ITEM-ITEM model. The commonest reasons for using the 2-column model are (a) the existence of heading(s) over the columns; and (b) a real pairing as one would find in a glossary (but do not normally find an index). Even (a) is not necessarily a reason to use the 2-column model if the heading is simple and can be simply distributed to all the items. "Page" for example can usually be distributed:

                       Page
Chapter 1. Of metaphor. 3
Chapter 2. Of simile.  15
Chapter 3. Tropes.     23

<LIST>
<ITEM>Chapter 1. Of metaphor. <HI>Page 3</HI></ITEM>
<ITEM>Chapter 2. Of simile.  <HI>Page 15</HI></ITEM>
<ITEM>Chapter 3. Tropes.     <HI>Page 23</HI></ITEM>
</LIST>

whereas in the following example,  the column headers force one into the 2-column model, since no distribution is feasible:

Term       Signification
-------------------------------------
giblet     a rabbet cut in masonry.
giff-gaff  mutual help, give and take
gigot      a leg (of mutton)

<LIST>
<LABEL ROLE="label">Term</LABEL>
  <ITEM ROLE="label">Signification/LABEL>
<LABEL>giblet</LABEL>
  <ITEM>a rabbet cut in masonry.</ITEM>
<LABEL>giff-gaff</LABEL>
  <ITEM>mutual help, give and take</ITEM>
<LABEL>gigot</LABEL>
  <ITEM>a leg (of mutton)</ITEM>
  
  
(2) Whichever model is used, when it meets a typical index, it frequently gets in trouble with nesting. A typical if fairly complex index entry looks like this:
Roman
  women,
    their dress 106
    their slaves at the toilette 107
    the ornaments they wore in their hair and at their ears ibid
    their high head-dresses 109
    dyed their hair yellow, and powdered it with gold dust ibid.
    their cosmetics, paint, and coating for the face 109
    their false teeth made of box 110
    were long unacquainted with the use of linen and silk ibid
    their extravagance in ornamenting their shoes 147
  knights,
    the speech of Caesar to them on their having neglected to marry 246
    fined by him for this neglect 247
    some of them married children to fulfil the letter, and avoid
       the spirit of the law, which obliged them to marry 251
      
We have seen some of these wrongly captured entirely within one LABEL, with ITEM reserved for the last page number, thus:

<LABEL>Roman
  women,
    their dress 106
    their slaves at the toilette 107
    the ornaments they wore in their hair and at their ears ibid
    their high head-dresses 109
    dyed their hair yellow, and powdered it with gold dust ibid.
    their cosmetics, paint, and coating for the face 109
    their false teeth made of box 110
    were long unacquainted with the use of linen and silk ibid
    their extravagance in ornamenting their shoes 147
  knights,
    the speech of Caesar to them on their having neglected to marry 246
    fined by him for this neglect 247
    some of them married children to fulfil the letter, and avoid
       the spirit of the law, which obliged them to marry</LABEL>
<ITEM>251</ITEM>

which is obviously wrong. Minimalist, but OK, would be to treat the whole block as one ITEM:

<ITEM>Roman
  women,
    their dress 106
    their slaves at the toilette 107
    the ornaments they wore in their hair and at their ears ibid
    their high head-dresses 109
    dyed their hair yellow, and powdered it with gold dust ibid.
    their cosmetics, paint, and coating for the face 109
    their false teeth made of box 110
    were long unacquainted with the use of linen and silk ibid
    their extravagance in ornamenting their shoes 147
  knights,
    the speech of Caesar to them on their having neglected to marry 246
    fined by him for this neglect 247
    some of them married children to fulfil the letter, and avoid
       the spirit of the law, which obliged them to marry 251</ITEM>
      
A little better would be to break out the pieces of the block as separate items, but not attempt to capture the hierarchy, thus:
<ITEM>Roman women, their dress 106</ITEM>
<ITEM>their slaves at the toilette 107</ITEM>
<ITEM>the ornaments they wore in their hair and at their
      ears ibid</ITEM>
<ITEM>their high head-dresses 109</ITEM>
<ITEM>dyed their hair yellow, and powdered it with gold
      dust ibid.</ITEM>
<ITEM>their cosmetics, paint, and coating for the face 109</ITEM>
<ITEM>their false teeth made of box 110</ITEM>
<ITEM>were long unacquainted with the use of linen and silk ibid</ITEM>
<ITEM>their extravagance in ornamenting their shoes 147</ITEM>
<ITEM>knights, the speech of Caesar to them on their having
      neglected to marry 246</ITEM>
<ITEM>fined by him for this neglect 247</ITEM>
<ITEM>some of them married children to fulfil the letter, and avoid
      the spirit of the law, which obliged them to marry 251</ITEM>
      
Better still is nested tagging that reflects the actual nested structure, at least approximately:

<ITEM>Roman
   <LIST>
   <ITEM>women,
       <LIST>
       <ITEM>their dress 106</ITEM>
       <ITEM>their slaves at the toilette 107</ITEM>
       <ITEM>the ornaments they wore in their hair and at their
              ears ibid</ITEM>
       <ITEM>their high head-dresses 109</ITEM>
       <ITEM>dyed their hair yellow, and powdered it with gold
             dust ibid.</ITEM>
       <ITEM>their cosmetics, paint, and coating for
             the face 109</ITEM>
       <ITEM>their false teeth made of box 110</ITEM>
       <ITEM>were long unacquainted with the use of linen and
             silk ibid</ITEM>
       <ITEM>their extravagance in ornamenting their shoes 147</ITEM>
       </LIST>
   </ITEM>
   <ITEM>knights,
       <LIST>
       <ITEM>the speech of Caesar to them on their having
             neglected to marry 246</ITEM>
       <ITEM>fined by him for this neglect 247</ITEM>
       <ITEM>some of them married children to fulfil the letter,
             and avoid the spirit of the law, which obliged
             them to marry 251</ITEM>
       </LIST>
   </ITEM>
   </LIST>
</ITEM>

We have also seen some wrong attempts at this that started the nested list a little too late, e.g.

   <ITEM>women,their dress 106
       <LIST>
       <ITEM>their slaves at the toilette 107</ITEM>
       <ITEM>their cosmetics 110</ITEM> ...  ]


(In each case, the lowest-level LIST could also be done in the 2-column way, but there is no good reason to do so.)
Back to top


4. Changes to the model of LIST

We have put into effect a number of minor changes to the EEBO/Evans DTD.

They consist of three changes to the model of LIST.  The TEI LIST element essentially offers two different models of list: the 'one-column' list consisting of ITEMs and the 'two-column' list consisting of LABEL-ITEM pairs. (The 'columns' do not have to be rendered as actual columns, of course.)

Two changes have to do with the second (LABEL+ITEM) type of  list. In brief:

(1) A ROLE attribute has been added to the LABEL and ITEM elements, with one possible value being ROLE="label". This will bring these elements into line with the CELL element (in TABLEs), and it will allow two-column lists that have headings over one column or both to be tagged properly. (LABEL ROLE="label" and ITEM ROLE="label" are equivalents for the  TEI elements HeadLabel and HeadItem, respectively.)

(2) LIST is now allowed to appear within LABEL. This is fairly radical and perhaps apt to be abused, but it will allow some of the more complex lists to be
captured, e.g. those that pair 'bracketed' lists of things with single things.

This simple two-column list illustrates both changes:

========================================
Distances from Manila to various cities
---------------------------------------
City                  Km
----                  --------
Sydney, Australia     6258
New Delhi, India      4765
Chicago, USA \        
Atlanta, USA /        13099
========================================

could be captured as:

<LIST>
<HEAD>Distances from Manila to various cities</HEAD>
<LABEL ROLE="label">City</LABEL>
  <ITEM ROLE="label">Km</ITEM>
<LABEL>Sydney, Australia</LABEL>
  <ITEM>6258</ITEM>
<LABEL>New Delhi, India</LABEL>
  <ITEM>4765</ITEM>
<LABEL>
<LIST>
<ITEM>Chicago, USA</ITEM>
<ITEM>Atlanta, USA</ITEM>
</LIST>
</LABEL>
  <ITEM>13099</ITEM>
 </LIST>

Finally, (3)  TRAILER has been added to LIST for those rare lists that have (as it were) their HEADs at the bottom. This brings LIST into line with TABLE, which allows for both HEAD and TRAILER.

Back to top


5. Syllogisms as lists

We have tagged some syllogisms using LB tags.

<P>All Cretans are liars
<LB>I am a Cretan
<LB>Therfore I am a liar
</P>

The following is preferable and  can appear embedded in running prose (unlike the LB solution):

<LIST TYPE="syllogism">
<ITEM>All Cretans are liars</ITEM>
<ITEM>I am a Cretan</ITEM>
<ITEM>Therfore I am a liar</ITEM>
</LIST>

Back to top