Constructing Social Networks in Modern Ireland (C.1750-c.1940) Using ACQ

Associational Culture in Ireland (ACI) project at NUI
Maynooth explores the culture of Irish associational life from
1750 to 1940, not merely from the point of view of who, what,
where and when, but also to examine the ‘hidden culture’ of
social networking that operated behind many clubs and societies
throughout the period. Recently commissioned government
research on civic engagement and active citizenship in Ireland has
highlighted the paucity of data available for establishing ‘trends
in volunteering, civic participation, voting and social contact in
Ireland’ (Taskforce on Active Citizenship, Background Working
Paper, 2007, p. 2). The same research has also confi rmed the
importance in Ireland of informal social networking compared
to many other economically developed countries (Report of
the Taskforce on Active Citizenship, 2007). The objective of the
ACI project is to provide a resource to enable scholars of Irish
social and political life to reconstruct and highlight the role
that the wider informal community information fi eld played in
the public sphere in Ireland from the mid-eighteenth century.
The project will also provide long-term quantitative digital data
on associational culture in Ireland which is compatible with
sophisticated statistical analysis, thereby enabling researchers
to overcome one of the hindrances of modern-day purpose
based social surveys: the short timeframe of data currently
available.
Associational Culture and Social
Networking
All historians are aware of the importance of social networks
that underpin the foundations of social and political life in the
modern world (Clark, 2000; Putnam, 2000). However, given the
often-transient nature of much of these networks, they can be
quite diffi cult to reconstruct.
One way to examine social networks is to trace them through
the structures of a developing associational culture where the
cultivation of social exclusivity and overlapping membership
patterns provide insights into the wider organisation of civil
society at local, regional and national levels. To this end, the
ACI project mines a wide range of historical sources to piece
together as comprehensive a view as possible of the various
voluntary formal associations that existed in Ireland during the
period c.1750-c.1940.
The fi rst phase of the project concentrated on collecting
data on Irish associational culture and social networking
in the period 1750-1820; the current phase centres on the
period 1820-1880 and the third phase will focus on the
years between 1880 and 1940. The research results so far
testify to the vibrancy of associational activity in Ireland and
already patterns in social networking are becoming apparent
in different parts of the country. The results so far indicate
that particular forms of associational culture were popular in
different parts of the country from the mid-eighteenth century,
which in turn produced their own particular social networking
systems. In many respects these patterns were maintained into
the nineteenth century with similar continuities in patterns
of sociability even though these continuities were sometimes
expressed through different organisations, e.g. the ubiquitous
Volunteering movement of the later eighteenth century
gave way to local Yeomanry units in the beginning of the
nineteenth century. Associational culture among the urban
middling strata also appeared to increase somewhat moving
into the nineteenth century with the increasing appearance of
charitable and temperance societies in different parts of the
country.
Software Development
Given the vast range of data available in the sources and
the multiplicity of question types posed by the fi ndings, the
organisation and dissemination of the research was one of
the main priorities for the project. The desire to present a
quantitative as well as qualitative profi le of Irish associational
culture, which can be expanded upon by future research,
presented particular diffi culties in terms of the production
of the project’s fi ndings. A multi-skilled, partnership approach
was adopted, fusing historical and computer science expertise
to digitally represent the data recovered from the sources
and expose the resultant social networking patterns through
the construction of an appropriate online database for the
project. Encompassing over seventy individual data fi elds for
each association, the online ACI database is fully searchable by
organisation and by individual – the later feature in particular
allowing social patterns behind Irish associational culture to
be traced over the course of two centuries.
Bradley (2005) points out XML is well suited to document
oriented project materials (such as written text), whereas
relational database implementations are better suited to data
oriented project materials. Furthermore, he indicates that
projects may often use both, given that textual materials often
contain data materials that are more suited to relational models.
He argues that even if one begins with a text, the materials
may be data-orientated and better served by a relational
model where certain characteristics, for example, linkages
between data objects and powerful querying are more easily
expressed using SQL (Structured Query Language) than XML searching facilities, for example, XPATH or XQUERY. It was
necessary, for our project to use a relational model to encode
data orientated materials, even though data were derived
from newspaper articles which are typically more suited to
encoding in XML. Our relational objects are densely linked,
and it is often necessary to build SQL queries incorporating
multiple joins across many database tables.
The database contains over 100 individual database tables
related to associations, members, sources, relationships,
associate members, locations, etc. Authenticated data entry
and access is available using online forms. Test, rollback and
moderation facilities are available for different classes of user.
Research queries may be formulated using a specifi cally designed
English language-type formal grammar called Associational
Culture Query Language (ACQL), which is parsed by the
software system to extract and present information from the
database – parsing involves evaluating ACQL and converting it
into appropriately constructed SQL suitable for querying the
database. Users may construct ACQL manually or they may
use an online query builder (see Figure 1).
The software engineering process began with all partners
involved in a collaborative process concentrating on the
construction of a Specifi cation of Requirements (SR)
document which essentially formed the contract between
the ACI historians and software engineers – this key
document was used to derive all phases of the software
development process, for example, System Analysis, System
Design Specifi cations, Software Specifi cations, and Testing and
Maintenance. These phases were implemented using a rapid
prototyping model, where successive passes through a designdevelopment-
testing cycle provided new prototypes which
could be evaluated against the SR document. As is typical
of such development projects, there was some requirement
drift, but the rapid prototyping approach ensured insignifi cant
divergence between prototypes and expectations outlined in
the SR document.
In order to produce the SR document, the ACI historians
composed a series of research questions, which were used
to extract the information categories necessary for the
construction of social networks. During the collaborative
process, these research questions and sources were used to
identify and model the relationships between the associations
and individuals in the sources. A key requirement was that
ACI historians wanted to be able to answer specifi c research
questions related to social networking, for example, “Are there
any illegal women’s associations/clubs active in Ireland after
1832?” It was necessary to revise these research questions
during the Analysis phase, however, as it became apparent that
the historians did not expect a yes/no answer to this question,
but rather a list of the associations and clubs, if they existed.
Most research questions went through a revision process
where the cardinality or result context was explicit, i.e. “List
the illegal women’s associations active in Ireland after 1832”.
This research question could be reformulated in ACQL as
follows:
LIST THE BODIES (NAME) (
WHERE THE GENDER IS “Female” AND
WHERE THE STATUS IS “Illegal” AND
WHERE THE FOUNDING DATE
IS GREATER THAN 1st OF January 1832
)
The parsing software then converts this
ACQL query into the following SQL:
SELECT BodyName.bodyName
FROM Body, BodyGender,
BodyName, BodyStatus
WHERE (Body.idBody = BodyName.
Body_idBody) AND
(Body.idBody = BodyGender.
Body_idBody AND
BodyGender.gender =’Female’ AND
(Body.idBody = BodyStatus.
Body_idBody AND
BodyStatus.typeOfStatus =’Illegal’ AND
(Body.foundingDate > ‘1832-01-1’)
)
);
This query is then executed by the Relational Database
Management System (RDBMS) and the results are returned
to the environment for presentation to the user.
We examined over one hundred questions of this type
ensuring that our relational model provided appropriate
answers. We developed ACQL to map research questions
into SQL queries, thereby removing the requirement for the
user to have knowledge of the database, table joins, or even
SQL. Researchers need not have an intimate knowledge of the
relationship between the database tables and their associated
fi elds to perform searches. Another advantage of this approach
is that the underlying database structure could change and the
users would not have to change the format of their queries
implemented in ACQL.
Ongoing and Future Developments
The success of this project depends on the research
community using and contributing to the information
contained in the online database. We are heartened, however,
by the reports of Warwick et al (2007) who have shown
that when information aggregation sites have user-friendly
interfaces, contain quality peer-reviewed information, and fi t
research needs, the research community scholars are more
likely adopt the digital resource. We believe that ACQL is a
crucial component in making this system usable by professional
historians interested in social networking. In particular, as the
volume of data within the database expands in digital format,
the potential for developing further social analysis tools such
as sociograms will be initiated. All in all, the ACI database,
by providing quantitative and qualitative data on specific associations, regions, and groups of individuals, comparable
with international data sources, will greatly aid historical and
social science researchers to establish Irish trends in civic
participation, social inclusion, marginalisation and grassroots
organising in modern Ireland.
References
Bradley, J. (2005) Documents and Data: Modelling Materials
for Humanities Research in XML and Relational Databases.
Literary and Linguistic Computing, Vol. 20, No. 1.
Clark, P. (2000) British clubs and societies, 1500-1800: the origins
of an associational world. Oxford: Oxford University Press.
Putnam, R. (2000) Bowling alone: the collapse and revival of
American community. New York: Simon & Schuster.
Taskforce on Active Citizenship (2007) Statistical Evidence
on Active Citizenship in Ireland, Background Working Paper.
Retrieved March 25, 2008 from the World Wide Web: http://
www.activecitizen.ie/UPLOADEDFILES/Mar07/Statistical%20
Report%20(Mar%2007).pdf.
Taskforce on Active Citizenship (2007) Report of the Taskforce
on Active Citizenship. Dublin: Secretariat of the Taskforce on
Active Citizenship.
Warwick, C., Terras, M., Huntington, P. and Pappa, N. (2007). If
You Build It Will They Come? The LAIRAH Study: Quantifying
the Use of Online Resources in the Arts and Humanities
through Statistical Analysis of User Log Data. Literary and
Linguistic Computing, Vol. 22, No. 1. pp. 1-18.