FactGrid:Career Statements

From FactGrid
Jump to navigation Jump to search

The present mess

Our present way to deal with career statements is still messy. The reason for this is historical. We started with an input from German address books - loading here everything from "baker" to "widow of a merchant" and "Doctor of medicine and senator" on a single Property Property:P165 - a property which eventually needed that extra broad label "career statement".

The second messy step came with the quickly reduce the risk of a language fork - the solution was here the Deepl-translation of the entire set of some 3,000 statements into English and French. It worked well with simple trades like "Baker" but created a mess with all the rarer names of historical trades.

A third insecurity came into the field with the neighbouring Property:P164 for "offices held" - here we began to link to very specific offices like Pastorate Altenbergen. Some of these specific terms also received the P2-statement "career statement".

On top of that we introduced a qualifier for specific positions Property:P166 since positions such as the Pastorate Altenbergen could be held by "Parish substitutes", "Pastors" or "Parish vicars" - which had to be qualified.

The various problems we created will not be solved that easily on the present number of roughly 4,600 items. We have:

  • unnecessary variations: Laquai, Lakei
  • compound statements of the sort "Carpenter's widow" (to be resolved as widow, Qualifier: status of deceased husband: Carpenter)
  • compound statements of the sort "Retired carpenter" (to be resolved as Pensioner, Qualifier: status of former occupation: Carpenter)
  • compound statements where people had different positions and occupations
  • academic titles which we could just state under their own property
  • honorary titles such as "Senator", or "Privy councillor" (de: Geheimrat) which should perhaps rather be seen as awards under that property
  • German labels where French and English words should appear

Eliminating unnecessary (?) variations: Laquai, Lakai

This is an easy case where we have just spelling variants such as in Laquai and Lakei but a complex problem where we have alternatives like Laquai and Valet de Chambre or Kammer-Diener. The early modern Kammer-Diener could be a high ranking title of honour.

The standard procedure is here the merging of items.

Dividing compound statements

We should resolve most of the compound statements.

  • "Carpenter's widow" should become Property:P165 Career statement "Widow" + Qualifier Property:P614 Status of the (deceased) husband "Carpenter"
  • "Retired Carpenter" should become Property:P165 Career statement Item:Q37181 "Retired" (or Item:Q37199 Pensioner) + Qualifier Property:P211 Previous Professional status "Carpenter". In addition we should set a P165 first level statement on "Carpenter" to note that former position.
  • "Inn keeper and master cooper" should become two statements, with the respective dates to state the simultaneity.

The complexity is here that we might actually preserve some of the compounds in the last case as they stand for a specific design of the production, e.g. in a factory that produces these two products.

The standard procedure is the division into regular statements and the deletion of the compounds.

No academic titles on P165 statements?

The first input brought academic titles on Property:P165 Career statements. We already have a Property:P170 for academic titles.

Academic titles on career statements make, nonetheless, sense as in the case of Dr. med., in German the usual statement for an active medical practitioner.

Should Honorary titles ("Privy councillor"/"Geheimrat") become "public awards"?

Adam Weishaupt receives such a title in 1786 - which brings the professor of law into a new position protected by Gotha's duke.

In a way these titles are awards which should appear under our Property:P171 public award, but in many lists e.g. of address books they will fill the statement of occupation. We have therefore left these statements on the Property:P165 for the time being.

From Deepl to authentic translations

The best solution is a translation with the look into contemporary dictionaries (see our collection historical dictionaries that went online at FactGrid:Authentic translation help).

It is the best solution as it allows the automatic matching of titles found in documents in different various languages.

The biggest problem is here that the different languages did not show the same differentiations. There are fields that received various specific terms in some languages (reflecting here the definitions defended by the old guilds), whilst other languages have just one word for the job that needed to be done.

The problem increases with the historical developments we are trying to grasp. Here we have terms that remain stable whilst the jobs were changing - and again these changes were not always noted in the various languages simultaneously and coherently. Some changed words and some did not.

Problem of the three properties: P165: Career Statements, P164: Positions, P166: specific positions

Wikidata has basically two properties: Occupation (P106) and Position held (P39]).

The FactGrid equivalents are Career statement Property:P165 and Office held Property:P164. Our "Career statement" property is more inclusive, allowing all the statements one might find on a personal record (such as retired, pensioner, widow, candidate of theology).

The additional Property:P166 came into use in order to state specific job descriptions on items such as Pastorate Altenbergen. It might make sense to fuse this Property into Property:P165.

Could we go a whole step further and reduce everything to Career statement Property:P165 statements? Not that easily if we still want to be able to run a simple count on all the shoemakers of a town. We might have to cleanse our career statements for that purpose of all statements that refer to specific positions in regiments etc.

Problem of historical changes and stable Q-Items

...this is an immense challenge. It needs specialists for this field.


It is not yet possible to run statistics on our career statements. At the moment we will get widely scattered fields of often extremely specific statements. Pies or bubble graphs will need succinct reductions to work with. A first pattern could be:

  1. Church
  2. Military
  3. Law, government and administration
  4. Crafts
  5. Commerce
  6. Agriculture
  7. Education and academia
  8. Artists and writers
  9. Servants
  10. Prostitutes and mistresses
  11. Landed property
  12. Unemployed / sick people / students/

We can run several patterns side by side - all they need is Properties and the respective statements on our career statements.

Wives are a difficult category - in a way they are a group of their own, in a way they belong to the trade of their husbands up to thee "professor's wife" who is eager to be noted in this condition.