The letters in our collection come from many sources, both institutional and private. Institutional collecting has been done, particularly by the project team, by and large in locations that are within easy access. This has resulted in some individuals being over-represented. This is especially true when the holding institution contains a very large collection based on a particular individual or family. The analysis below is therefore based on our dataset and not necessarily representative of Ireland in that period.
A map of the letter network can be viewed here.
There are currently 4374 letters in the database. There are 2612 unique names in these letters. Within this there are 1137 senders writing letters to 1768 recipients. Of the identifiable senders, 1264 are male and 321 are female (note these numbers differ from the number of letters written by males or females as that counts each letter written by a male or female, this is the unique number of males and females in the database). There are also a a significant number of letters written by individuals in which it is not possible to assign gender (for example, when a letter is signed via initials only or indeed, unsigned).
The people who sent the most letters are:
- Wilfrid Bliss Spender 178
- Thomas Percy Kirkpatrick 123
- Patrick Carey 68
- Marie Martin 61
- James Finn 53
- James Davidson 44
- Edward O’Farrell 43
- May Fay 41
- Andrew Philip Magill 39
- Willie Doyle 39
The average number of letters sent is 1.47 (SD = 8.33). The distribution of letters is shown here:
(Note the logarithmic scale to help display the high deviations.) This figure shows the majority of people in the database send or receive only one letter. However a small number of people send and receive over 100.
The people who received the most letters are:
- Robert Chalmers 190
- Hanna Sheehy-Skeffington 184
- Lillian Spender 151
- Matthew Nathan 132
- Thomas Percy Kirkpatrick 78
- Gregory Ashe 73
- Mary Martin 64
- Edward Carson 60
- May Fay 54
- Mabel Fitzgerald 53
Matthew Nathan and Robert Chalmers were the Under-Secretaries for Ireland from 1914-1916 and 1916 respectively.
There are a total of 2578 connections in the network displayed here. This is lower than the number of letters as often more than one letter is sent between the same pair.
A common idea in social network analysis is the notion of “six degrees of separation”. This means that on average, any two people in the world can be connected by a chain of six steps. This is more a rule of thumb than an exact number. For the letter network we can calculate this however, and we find there is an average of 5.56 steps between any two people in the network. Therefore this network can still be said to be small world.
A community in a network is a group of densely connected nodes. Here the “modularity” is used to find the number of communities. This compared how many modules there are to what you would expect if connections were made randomly. The figure below shows the growth of the modularity (it can never reach 1.0)
Initially, at two communities, the modularity is very low indicating it is not much different from picking two random clusters. By the time five communities are reached however, the modularity is beginning to reach it’s peak. We observe that it levels off just after ten communities. These communities are displayed in the network diagram by the vertices being grouped in different colours.