|This story is an attempt to answer twitter user @ecritures' question of whether there is a list of Dutch female occupations.|
I have been working with Dutch occupational data for over a decade and have seen various studies on women's work from ca. 1600 onwards. I am not aware that these projects have presented lists of their occupations as a 'product'. One reason for that is that the whole idea of FAIR data (especially reuse) is only beginning to sink in. But even when trying to be FAIR, there is only so much time, and people choose to invest time in publications over time invested in data, as the latter is not valued in the academic career track.
There are though, across the world, various exceptions to the rule. Basically, these exceptions are bigger datasets, such as IPUMS or Historical Sample of the Netherlands, that in addition to providing respectively census and civil registry data, also provide, what I call, auxiliary files. These are data files of entities such as places and occupations that provide on the one hand the raw archival entry and on the other hand a cleaned and standardized representation. E.g. for places in the Netherlands they would contain the Amsterdam Code and for occupations they are presented with the HISCO code. For some time now, I have been trying to gather various datasets across the world that contain these lists of occupations and their coded standards, (I call this collection 'job hoard'), and provide them as Linked Open Data as part of the History of Work collection at the International Institute of Social History (IISG).
Before I show the query, I'd like to make a more philosophical point about 'lists'. Lists are great. But lists are also somewhat annoying. When cooperatively trying to create lists, people tend to have slightly different ideas of what the list ought to be about. The idea seems clear at first: "women's occupations". But then there is the afterthought of whether a teacher, an occupation that used to have mainly male incumbents, ought to be considered a "women's occupation" as well. This is an issue that comes into play with any topic list, e.g. should a Dutch place name list contain 'Batavia' (a place in a former colony) as entry? And then what usually happens is that what started out with good intentions ends up in different groups working on different lists and over time resources prove to be too scarce to maintain these lists.
Linked Data provides a new way of creating lists. As Linked Data resources are more easily connected over the web, we could work on queries over multiple sources, rather than actual lists themselves. A query is the means by which you generate a subset from a resource, say, all "women's occupations" from the list of all occupational resources. This has two advantages. For one, various lists can be retrieved from the same resource. Regardless of whether one's interest is in places, occupations or products, they would rely on the same underlying data sources and thus there would be a common interest (and focus for resources) to maintain and uphold those sources. Another advantage is that queries, much more than lists reflect the common denominator. In the example of place names, lists of Dutch place names with and without locations in former colonies is basically a single line in a query. Basically, it is the same query, with an additional filter. Again, much more commonality and thus easier to uphold and recognize each other differences.
In the query below the feature of having multiple filters to create a list becomes evident. What you see below is the output of a sparql query. For convenience, I added a box where you can fill out a first latter or part of a string, to look for a particular occupation. (For nerds: this is an API variable, so please use it). E.g. rather than returning strings that start by "s" (default), you could fill out "kaas" ('cheese') and you would be given all occupational titles, that have been donated to HistoryOfWork that start with 'kaas'. Please do try it out now by entering the text in the box on the left and pressing the 'run query' bottom to the right. Next, scroll down for the internals of the query.
Now please right click on 'try this query yourself' on the top right of the table above. A new screen will open, with code. This code is called a sparql query and extracts a subset of occupations from all occupations that have been made available via the HistoryOfWork collection. I have done my best to annotate extensively and guide you through the query.
Basically the query consists of three steps. First, we define some 'prefixes' which avoids us having to write URI's in full. So I can write for example hisco:MicroGroup rather than https://iisg.amsterdam/resource/hisco/MicroGroup. Next, we define what an occupation is, and what we want to know from it. E.g. here I want to know the 'name' and 'hisco code' of the occupation. Finally, I want to filter out any unwanted occupations. So, for this case, I just want occupations in Dutch to begin with. Next, I try and retrieve gendered occupational titles, e.g. occupational titles that end with 'ster'.
In the query itself, you will see that I mention an additional step: expand the query. Obviously, simply stating that occupations ending with 'ster' are women's occupation is practically flawed, but more importantly is a very crude assumption to a topic delicate as work and gender. Therefore, sharing and updating each others queries is of major importance. So please give it a try and provide your best query. We could store them on github for example using grlc. Please be in touch via @rlzijdeman and let me know your thoughts!