Undergraduate student contribution


Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies

roseList documents mentioning Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies
In my previous blog post, I discussed the need for geocoded bibliographic databases in print culture studies and the enormous value of the spatial queries that such databases would enable. What would such a database look like? How might programmers and encoders design a database that dynamically links data points about material books and stationers with spatial variables? In an effort to answer such questions, I shall here showcase how The Map of Early Modern London (MoEML) has chosen to geocode data pertaining to the early modern London book trade in a TEI-compliant XML database. My hope is that MoEML’s prototype will serve as a touchstone for print historians, regardless of their specializations, who are interested in incorporating spatial variables into their bibliographic data.
Print historians have used a variety of methods and languages to encode bibliographic data. The English Short Title Catalogue (ESTC) uses MARC formatting, a library cataloguing method developed by the Library of Congress, to encode their 460,000+ entries. The London Book Trade Database (LBTD) encodes its information as tab-delineated values in a Microsoft Access database. Other resources use other, often unspecified methods. Ultimately, the method used to encode data is superficial: what matters is that datasets are encoded (marked-up) and therefore interoperable with one another.
MoEML is a TEI-XML project and therefore uses TEI-compliant XML documents and databases to power its Geographic Information System (GIS) interface.1 For MoEML, as for other digital editorial projects, the Text Encoding Initiative (TEI) provides a published set of standards for using the Extensible Markup Language (XML).2 With these standards in mind, I recommended that MoEML adopt the following TEI-compliant XML document tree as a template for encoding and geocoding its database of early modern London books:3
<!-- Metadata -->
<docTitle><titlePart type="main">[Database title]</titlePart></docTitle>
<person xml:id="ABCD1" sex="[Insert sex value]">
<!-- Value of @xml:id should be a unique alphanumeric ID. -->
<reg>[Surname Forename]</reg>
<occupation code="[Appropriate MARC relator code]" scheme="MARC"></occupation>
<birth when="[ISO date of birth]"/>
<death when="[ISO date of death]"/>
<location notBefore-custom="[ISO date of first book associated with given stationer at given location]" notAfter-custom="[ISO date of last book associated with given stationer at given location]">
<address ref="mol:ABCD1">
<!-- Value of @xml:id should be the <title level="m">MoEML</title> @xml:id for given shop location. -->
                                    [Authoritative description of place]
<lem wit="[Source information of description]">
                                            [Primary source description of shop location]
                                        [Add additional witnesses as needed, each in a new 
<geo>[Geo-coordinates for shop location]</geo>
                                    [Name of book associated with given stationer at given location]
                                [Add additional sources as needed, each in a new 
                        [Add additional locations as needed, each in a new 
<p>[Biographical statement for given stationer]</p>
<list type="links">
<item>[Link to further information about given stationer.]</item>
                            [Add additional links to resources as needed, each in a new 
                [Add additional people as needed, each in a new 
At first glance, it may seem odd that we have chosen to use the <person> element as the root element (i.e., the sorting variable) in a so-called geocoded bibliographic database. Indeed, in my previous blog post, I lamented the fact that many bibliographic databases favour queries about stationers or material books over queries about print shops or bookshops. In practice, however, early modern print historians currently lack the raw spatial data required to support a truly geographical database (i.e., a database that primarily supports queries about locations of print activity).4 Before print historians can design a geographical data structure, they must first geo-reference existing data structures. MoEML’s geocoded bibliographic database therefore serves the transitional purpose of geo-referencing existing datasets about early modern stationers and their material books. Accordingly, we have purposefully chosen to imitate the document structure of existing datasets such as the British Book Trade Index (BBTI) and the London Book Trades Database (LBTD) by using the <person> as the root element in our database.
Each database entry contains biographical information about an early modern London stationer. First, we use the <persName>, <occupation>, <birth>, and <death> elements to tag the stationers’ names, occupations, dates of birth, and dates of death respectively. Second, and more importantly, we use the <floruit> elements to tag information about the stationers’ shop locations and the material books printed, published, and/or sold there. The data structure, which nests the <person>, <floruit>, <location>, and <listBibl> elements as //person/floruit/location[@notBefore and @notAfter]/listBibl, may be expressed as stationer X worked at location Y from date A to date B, where s/he was associated with a list of books. In this way, the document structure dynamically relates traditional bibliographic data points to new, complex geographic data points.
Significantly, this data structure allows us to encode locations of print activity in terms of both their toponyms5 and geo-coordinates. Within the <location> element, we use the <address> and <addrLine> elements to tag an authoritative description of the location. We also use the <geo> element to tag the location’s longitudinal and latitudinal coordinates. When paired, these two data types (i.e., toponym and geo-coordinates) bear witness to both the qualitative and quantitative aspects of geographic information.6.
Title page of STC 22290. Image courtesy of LUNA.
Title page of STC 22290. Image courtesy of LUNA.
Title page of STC 22340. Image courtesy of LUNA.
Title page of STC 22340. Image courtesy of LUNA.
Because of the chronological distance between modern print historians and the early modern London book trade, both a location’s toponyms and its geo-coordinates must be inferred from primary sources. For early modern print historians, book imprints serve as primary sources in which the location of a book’s printer and/or bookseller is often described in toponyms. For example, the imprint on the title page of a 1602 playbook of Shakespeare’s Henry V reads London : Printed by Thomas Creede, for Thomas Pauier, and are to be sold at his shop in Cornhill, at the signe of the Cat and Parrets neare the Exchange, 1602 (STC 22290; my emphasis). Even individual book imprints, however, are not reliable in and of themselves: multiple book imprints that describe the same location will often provide different or contradictory toponymic descriptions. For example, the imprint of a 1608 playbook of A Yorkshire Tragedy describes Thomas Pauier’s bookshop with less precision than the previous playbook: London : Printed by R[ichard] B[radock] for Thomas Pauier and are to bee sold at his shop on Cornhill, neere to the exchange, 1608 (STC 22340; my emphasis). Due to such inconsistencies, book imprints must be regarded as mere textual witnesses to the objective location of print shops and/or bookshops in time and space. Much like the so-called literary work, the objective location of an early modern print shop or bookshop is a form [that] is inaccessible (Williams and Abbott 6) to print historians and therefore can only be physically embodied (Williams and Abbott 5) in the imprints of material books and their digitization. We therefore use the <app> and <lem> elements to encode each primary-source description of a print shop or bookshop’s location as a single lemma. When combined, these lemmas aggregate data points that allow us to approach an authoritative description of the print shop or bookshop’s location. By using the <app> and <lem> elements to record our editorial practices, we are able to encode geographic data points as transparently and accurately as possible.
While not the only encoding language available to print historians, TEI-XML certainly provides print historians with an effective way to encode and geocode bibliographic databases. TEI tagging acknowledges the textuality of the data available to us. In this blog post, I have showcased how The Map of Early Modern London (MoEML) has used the TEI P5 Guidelines to develop a prototype for a geocoded database of early modern London books. MoEML’s prototype database serves as a touchstone for pint historians insofar as it demonstrates how researchers might incorporate and structure spatial data points within a traditional bibliographic database. By geocoding existing datasets in this way, print historians may lay the groundwork for new databases that primarily support queries about the spatiality of the book trade. Of course, such possibilities cannot be realized unless developers have access to raw spatial data pertaining to books. My next blog post will discuss how print historians and programmers can work together to harvest and cross-reference existing datasets to populate a geocoded bibliographic database.


  1. Although MoEML does not use a traditional GIS software package, it does meet most broad definitions of a GIS. For instance, Tor Benhardsen (2008) defines a GIS as any computer-based capability for the manipulation of geographical data (Benhardsen 4). (TLG)
  2. For more information about TEI standards and practices, see the TEI P5 Guidelines. (TLG)
  3. See detailed explanation beneath the template. (TLG)
  4. Such data does exist in print sources such as Blayney’s The Bookshops in Paul’s Cross Churchyard (1990) and Pantzer’s appendices in the Short Title Catalogue, v.3 (1991), but it has yet to be transferred into a digitally readable format. (TLG)
  5. A place-name; a name given to a person or thing marking its place of origin (OED toponym n.2.a). (TLG)
  6. Martyn Jessop (2008) aptly claims that the qualitative place name and the quantitative map constitute the two, mutually dependent sides of geographic information (Jessop 41). (TLG)


Last modification: 2016-05-27 14:37:29 -0700 (Fri, 27 May 2016) (tlandels)
Export to RefWorks
RIS file (for RefMan, EndNote etc.)

MLA citation:

Landels-Gruenewald, Tye. “Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies.” The Map of Early Modern London. Ed. Janelle Jenstad. Victoria: University of Victoria. Web. 18 October 2017. <http://mapoflondon.uvic.ca/BLOG17.htm>.

Chicago citation:

Landels-Gruenewald, Tye. n.d. “Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies.” The Map of Early Modern London. Ed. Janelle Jenstad. Victoria: University of Victoria. Accessed October 18, 2017. http://mapoflondon.uvic.ca/BLOG17.htm.

APA citation:

Landels-Gruenewald T. (n.d.). Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies. In J. Jenstad (Ed.), The Map of Early Modern London. Retrieved October 18, 2017, from http://mapoflondon.uvic.ca/BLOG17.htm

TEI citation:

<bibl> <author><persName><surname>Landels-Gruenewald</surname>, <forename>Tye</forename></persName></author> (<date>n.d.</date>). <title level="a">Georeferencing the Early Modern London Book Trade: 2. Filling the Space in Bibliographies</title>. In <editor><persName><forename>J.</forename> <surname>Jenstad</surname></persName></editor> (Ed.), <title level="m">The Map of Early Modern London</title>. Retrieved <date when="2017-10-18">October 18, 2017</date>, from <ref target="http://mapoflondon.uvic.ca/BLOG17.htm">http://mapoflondon.uvic.ca/BLOG17.htm</ref> </bibl>