Ireland’s nineteenth-century administrative units appear simple on first glance. There were essentially two systems at play, and both involved townlands (the smallest, most basic territorial division) and counties (the largest). The administrative scheme prevalent for the first half of the century grouped townlands into civil parishes, civil parishes into baronies, and baronies into counties. While some overlap took place—parishes occasionally straddled more than one barony, and a few baronies crossed county boundaries—this system did possess a certain coherence. But with the advent of poor law administration starting in the 1830s, and with it the creation of poor law unions, a competing system of townlands/district electoral divisions/dispensary districts/poor law unions existed simultaneously with the earlier townland/parish/barony approach. As the presentation of social statistics was increasingly attuned to a goal—improving public health and measuring poverty—tied to poor law union administration, the censuses of the second half of the nineteenth century began to organize their published age and health tables based on unions and their dispensary sub-districts. From the 1871 count onward, for example, census officials presented age data using the dispensary district scheme. Unfortunately, they also continued to present other data—notably, the figures for languages spoken—using the parish/barony breakdown.

There are a number of lesser, but no less nagging, difficulties faced by historical geographers. The boundaries of parishes, baronies, dispensary districts, and poor law unions changed over time. Parishes in particular were susceptible to being merged. Baronies, unions, and districts were split, especially as administrative needs expanded. Finally, as any historian or genealogist who has researched this time period is aware, the language question impinges on the names of these units. Townlands in particular pose a problem as most of their names originated in the Irish language, yet were presented in a form amenable to English readers in official nineteenth-century documents. Although officials tried to impose some kind of consistency on these anglicized forms of Irish names, one quickly discovers in comparing the various townland indices that variations occurred.  Spellings of townlands in 1851 often differ from those in 1901, and even constantly reappearing terms (such as “knock-” for cnoc) will display some irregularity when it comes to the anglicized form. Anyone who has had to track down one of the hundreds of townlands in Ireland with prefixes such as “kill-” (Irish: cill) knows that this requires an unpleasant search through dozens of fine-print pages in the Townland Index listing names beginning with all sorts of variations such as a single “l” (kil- vs. kill-).

(Genealogists, often in the vanguard on these sorts of things, have had an electronic index available for some time . . .)

A firm handle on the administrative history of a given area helps combat these problems, but GIS can obviously aid immensely in researchers’ efforts to track these shifts. Because townlands formed the basic building block of both administrative systems in nineteenth-century Ireland, the goal for organizing linguistic data would be townland-level files that could be grouped into parishes and baronies as well as district electoral divisions and dispensary districts.

Such projects are immensely helpful to researchers not only because they enable data to be presented in geographic form, but also because they solve many of the problems associated with geographic information over time -- shifting administrative boundaries being the most obvious. The British data is available here.

However, when I turned to Ireland, no such data appears to be available at present -- at least not on such a researcher-friendly basis. The natural place to look for such data would be the Ordnance Survey Ireland, an old and venerable institution with a long (and much-discussed) history of mapping in Ireland. However, while the OS does produce historic maps, its shp files seem to be confined to modern data (queries made to OS have yet to be returned). Worse, I fear that OS charges for all of its data. This is a reasonable practice and one that provides access for some researchers, but not one that is practical for me at this time.

As it turns out, the researcher is not without recourse for Ireland when it comes to modern boundaries: the Central Statistics Office currently provides shp files for the 2006 census boundaries. Although not the historical information I am after, this is a great find.

Undaunted, my next direction has been to create the shp files myself using public domain maps and my trusty ArcMap program which I am slowly learning how to use. This summer has therefore been a crash course in rubber sheeting and boundary tracing to create my own files for the nineteenth century. As it turns out, ArcMap is indeed a great desktop tool. I was surprised to find that I could, with considerable time and patience, create my own map file consisting of Ireland's baronies, one of the major administrative units used up until mid-century (and in some cases, beyond) to organize population data. Here are the lovely results:

Admittedly, it doesn't look like much, but it represents about three weeks of patiently tracing boundaries and the rocky coastlines of the west (I can't decide if southwest Cork or western Galway has the more intricate coastline, but either way I'm glad to be done with it!). Best of all, it provides a great basis for presenting my findings on the census project as well as future data based on baronial units.

Next post I'll take a closer look at the actual territorial units at play in nineteenth century Ireland and the potential needs for further shp production!