Maria Ågren

The Gender and Work database [GaW]: an infrastructure for historical research into work, gender and everyday life

At the centre of this project stands the GaW database [Gender and work]. The database was developed at the history department, Uppsala University, in collaboration with CEDAR (the Demographic Database) at Umeå University. GaW offers information on women’s and men’s work activities in Sweden between 1550 and 1799. As a rule, such information is hard to come by and its collection requires painstaking analysis of (usually handwritten) historical sources. The major part of the sources used for GaW consists of court records. Storing this information in a database that can be accessed free of cost on the internet allows for inter-subjective scrutiny and reuse of data. In this way, the information can be used to answer many different questions, not least about people’s everyday lives. The database is constructed so as to support analysis according to the verb-oriented method. A major result of such an analysis is that before 1800, marital status (whether a person was married or not) seems to have been highly important for what types of work s/he carried out, maybe even more important than gender. This project develops and improves the usefulness and visibility of the database by (i) increasing its geographic and chronological coverage, (ii) linking information on people’s work to information on their marital status and age, (iii) facilitating the invitation of young non-Swedish scholars to the GaW research environment.
Final report
IN17-510:1, IN17-510:2, Final project report, January 2023
The Gender and Work database [GaW]: an infrastructure for historical research into work, gender and everyday life

Purpose and development
The purpose of this infrastructure project was (i), to expand the chronological and geographic coverage of the existing database GaW (Gender and Work); (ii), to enhance the quality and usefulness of existing data; (iii), to increase the international visibility and impact of the database and its methodology. All these objectives have been reached. In addition, a complementary RJ grant (IN17-510:2) has made it possible to technically rebuild the database. At the midterm evaluation in November 2019, the evaluation committee wrote:
"The verb-oriented method on which the project is based has had an international impact. The long-term collaboration with the computer linguists in Uppsala has great potential. [---] The panel for research infrastructure is very positive to the report.” [Trans. MÅ]
The GaW database supports research that relies on the so called verb-oriented method. The method is inspired by anthropological observation of (especially women’s) time-use in modern countries outside Western Europe and Northern America. Due to the fragmentary character of historical sources, it is however rarely possible to map what individual people did during a full day or a full week. Instead, the historical method identifies activity patterns typical of groups of people, e.g., men versus women, married versus unmarried people, etc. GaW is a relational database built by CEDAR (the Demographic Database), Umeå University, in close collaboration with historians at Uppsala University. The project has spawned several collaborative projects.

Objective 1: expanding the chronological and geographic coverage
At the time of application (April 2017), the database included some 23,000 observations of sustenance activities (verb phrases) from southern, middle and northern Sweden (plus a small number of observations from what is now Finland) in the period 1550 to 1799. In total, there were around 500,000 Swedish words, culled from court records, petitions, accounts and diaries. In the application, we wrote that ‘parts of the eighteenth century are too scantily represented. Also, southern and northern Sweden need better coverage as does Finland.’ We also wrote that ‘the nineteenth century in Scandinavia is often portrayed as a period when entirely new gender ideals and new practices appeared, [therefore] it is now crucially important for the GaW project to go beyond 1800.’
Today (January 2023), the size of the database is c. 40,000 verb phrases describing work activities. The chronological coverage has improved: 1550–99 9%; 1600–49 5%; 1650–99 21%; 1700–49 19%; 1750–99 18%; 1800–49 18%; 1850–99 10%.
Geographically, areas from southern and northern Sweden have been added (see Table 1). In total, there are c. 1 300,000 Swedish words, culled from court records, petitions, accounts, diaries and taxation material. A considerable share of this new data is the result of the project IN17-0510:1.

Objective 2: Linking of demographic information (marital status, age) to verb phrase data
We have complemented a large subset of data with demographic information from ecclesiastical sources (the data from Västerås, Snevringe and Tuhundra in Västmanland and from Skellefteå). As a consequence, for this data we know the marital status for 75 percent of the people – both women and men! – who performed the work activities (see Table 1). This makes the data considerably more useful.

Objective 3: facilitating invitation of non-Swedish scholars to the GaW research environment
In order to increase the international visibility and impact of the method and the database, we have used project funding to invite guest researchers, both junior and senior ones: Professor Carmen Sarasúa, Autonomous University of Barcelona; PhD Bob Pierik, University of Amsterdam; Professor and Director Karin Hofmeester, International Institute of Social History, Amsterdam. Some of these visits were delayed by the pandemic, and we had to ask for permission to extend the project period one year. In addition, PhD and Master students have been received by the project: e.g., PhD student Birgit Dober, University of Vienna; Master student Romain Rafai (Strasbourg University and Uppsala University).

Table 1. Data analysis and registration
Area Years Transcribed from handwritten texts Analyzed (registered with metadata in database) Complemented with demographic information from ecclesiastical sources Published in database
Gotland, Martebo 1768-1798 Yes Yes Yes Yes
(1768-69)
Västergötland, Kålland 1697, 1700 Yes Yes No Yes
Västerbotten, Skellefteå 1771 Yes Yes Yes Spring 2023
Stockholm, mantalslängder 1740 Yes Yes No Yes
Västmanland, Snevringe, Tuhundra 18th and 19th centuries Yes Yes Yes Spring 2023
Västmanland, Västerås 18th and 19th centuries Yes Yes Yes Spring 2023
Södra Österbotten 18th century Yes Yes No Spring 2023

Results
Methodologically, an important result is that court records have proved to be as useful for the 19th century as they are for the period before1800. The types of work and the level of detail is comparable across time 1550 to 1880. Apparently, the low degree of legal specialization in Swedish courts means that a wide spectrum of work activities continues to be reflected in the records.
Empirically, the most important result is that it has been possible to measure the impact of marital status and gender respectively, and to show that with time the latter became more influential for what work people did.

Usefulness of the database
With its focus on handwritten and unpublished texts, the database has proved useful for The Swedish Academy Dictionary (SAOB). The database has also been used as training data by computer linguists and image analysts. University and upper secondary teachers have used the database for their own teaching purposes. The database has been used e.g., by postdoc historians at Swedish (Stockholm and Gothenburg) and Finnish universities to identify court cases of interest for their research problems.
Naturally, the GaW database is used within the Gender and Work research project, Uppsala university.

Unforeseen technical problems
Shortly after the original application had been approved, the system analysts at Umeå University (who constructed the database and service it) reported that the software Glassfish risked becoming obsolete, i.e., it was no longer being updated. It was deemed necessary to write entirely new code that could be used instead. When we reported this problem to RJ, Göran Blomqvist encouraged us to apply for extra money, which we did (22 March 2018) and duly received (IN17-510:2). As a consequence the tools (applications) for entering and retrieving data have been completely rebuilt. The process has been very slow, and there are still some parts that remain to be handled. The cooperation between system analysts and historians has however been excellent, and the result is very promising.

The integration of the infrastructure at the home university
The database is used in a number of ways within the organization. At the History department, C-level students are encouraged to use the database when they write their C-theses. Master students can do 5- to 10-week internships at the GaW database (see above). At the department of Linguistics, the course in Digital philology includes a lecture on GaW. Still, we feel that the university does not fully take advantage of the fact that it hosts a unique database, nor does it take financial responsibility for it. At the midterm evaluation in 2019, the evaluation panel wrote:
“One question concerns the long-term funding of GaW and, to that end, the project is having a discussion with Uppsala University about status as an incubator. The panel for research infrastructure sees it as very important that the long-term utilization of the database is secured and looks forward to that issue being highlighted in the final report.” [Trans. MÅ]
This recommendation was forwarded to the head of department and the vice-rector. The vice-rector did subsequently discuss the matter with the vice-chancellor, but nothing concrete came out of this. Since the GaW project is funded by VR until 2029, there is no immediate crisis, but the situation is unsatisfactory.

Accessibility and quality control
The database is available on line for everyone to use ( https://www.gaw.hist.uu.se ). Data is made available as soon as it has been quality checked. Quality control has several components: The registrator checks his/her own work; a moderator makes systematic checks after the registration project has been completed; every registration project is documented with the help of a specially designed form; external users are encouraged to report errors via the website. Quality control takes time and accounts for the time-lag between analysis and publication of data.

International cooperation
The database has inspired similar infrastructure and research projects in the Netherlands and the UK. There is ongoing cooperation with the IISH, Amsterdam, and with CAMPOP, Cambridge, about categorization of work activities.

Publications
See separate list

Links
https://www.gaw.hist.uu.se
Grant administrator
Uppsala University
Reference number
IN17-0510:1
Amount
SEK 4,264,000
Funding
RJ Infrastructure for research
Subject
History
Year
2017