User talk:Marie-Alice Le Corvec

Hi, I am the archivist for the Centre de recherche bretonne et celtique (CRBC) at Université de Bretagne occidentale (UBO, Brest). I work with Jean-Baptiste Pressac, who contacted Factgrid recently, on making available descriptions and data from archival fonds preserved by the CRBC. I'm currently working on a private family fonds (Miorcec de Kerdanet) who have more than 50 related individuals identified in their papers, with links to historical French and Breton institutions and agents of justice as well as writers and researchers from the 19th century. I'm currently looking for an open tool to generate a family tree people interested in the fonds might consult during their research.

Item:Q636357 - abstract art, one way or another...

I (Olaf Simons (talk)) am moving the following remarks of User:Martin Gollasch from the talk page of Item:Q636357 to the user page that should be addressed here:

Hello Marie-Alice, regarding the other items instance of Type of work, Abstract art, is more an art genre regarding Wikidata's abstract art. You could create a FactGrid item equivalent to art genre and declare that Item:Q636357 is an instance of the aforementioned.

Something is going wrong here: the German label is translated "abstract art", English and French label describe a special vine label. That needs definitely a clean up. You should not change the content of an item. Better create a new one. This is a total mess...--Martin Gollasch (talk) 10:13, 24 October 2023 (CEST)

Looking at the history I think I know what happened. The Iconclass perspective - "abstract art" went into a Wikipedia link on abstract art, and a German label of that topic.

Links to Wikipedia must go to an 100% match of the item. Further reading on abstract art (in Wikipedia) can be a property "online information", but not an external identifier.

I am, just by the way, not quite sure what kind of project we are entering here. I could imagine a research projet on wine labels with a clear research question and a defined body of items, but I would not consider to feed all the wine labels of the world into FactGrid. (Wikidata is presently flooded with scientific research articles - bibliographic information and this is a problem). We should focus on defined research questions and on projects that can be discussed in the field of historical research. I can well imagine a project on wine labels, but so far no one has introduced this project with its aims and potentials. --Olaf Simons (talk) 11:33, 24 October 2023 (CEST)

Dear Olaf,

Thank you for the remarks and the follow-up on the page I modified (changed from Abstract Art to [Richemont...]) and your questions on the research project.

Regarding the project : The Papeteries Armoricaines et Morlaisiennes were an important printing house located in Brest, which ran from the end of the 1920s till a few years ago, and which specialised in creating for mostly local shopkeepers original labels (mostly wine, but local food as well : peas, fish... even cleaning products). We have in our archival fonds files regarding customers and their requested labels, and are currently trying to work on a way to, at the same time, be able to study the geographic location of the customers asking for the PAM's printing/artistic services as well as study the subjects depicted in those labels and how the information is transmitted in them (what language, words, typography...). A book was published when the PAM closed a few years ago, highlighting potential research that could be made from the content of the fonds, and their labels : Image & Commerce de la fin du XIXe siècle à nos jours, dir. Maryse Cuzon, Philippe Jarnoux, Marie-Michèle Lucas & Florent Miane( url : https://www.univ-brest.fr/crbc/menu/%C3%89ditions+du+CRBC/Collection+Collectif/Image---Commerce-de-la-fin-du-XIXe-siecle-a-nos-jours). I can summarize the subjects concerned as follow : history of printing techniques, of visual popular culture and its influences (the colonial aspect of some of the Labels is particularly striking in that aspect), uses of regional languages in popular objects (here : breton) etc. I hope this helps, but if you need more info I would be happy to share them with you.

Regarding the Abstract Art now modified page : we were interested in using the genres Abstract Art / Figurative Art to define labels for the goals of the project. But, after having realized the complexity of defining such items (necessity of creating other pages such as "art genre" etc), and re-thinking our use of the IconClass Classification System for visual art, we've decided to forego the need for such items in FactGrid, and will instead be using when necessary the IconClass propriety for Abstract Art ("O"). I knew I couldn't delete the page created, but I thought I could re-purpose it for the project (turning it into a wine label item) so as to avoid having an unfinished item floating around the database. However, if it creates "a total mess" as user Martin Gollasch says, I'm willing to find another way to register the wine label involved. What would you suggest ? (creating straight away a copy-paste item before deleting Item:Q636357 ...) Marie-Alice Le Corvec (talk) 14:37, 25 October 2023 (CEST)

Dear Marie-Alice

Item:Q636357 still has the same mess. In French and English this is a wine label, in German it is the term "Abstract Art" and the link to https://de.wikipedia.org/wiki/Abstrakte_Kunst and https://www.wikidata.org/wiki/Q128115 substantiate and settle the German interpretation of this item (Wikidata is linking back to the item with the logic you did not intend).

I will turn Item:Q636357 into a coherent item matching the French decisions, and I will create a new item for abstract art to deal with the apparent lack of the item on the subject.

All FactGrid projects should have project pages on the database (visible in the project space), so that all on board have an idea what this is all about. I also recommend that projects use the blog to publish results of their work.

Data are a nice subject matter to collect if they can be contextualised by others, if others can integrate these data in their research. Data are otherwise as terrible as dust accumulating under cupboards, sofas and beds. No one knows what to do with them, whether still usable, whether making sense in any way - and then people begin to make messy connections. See this item on "Baron" - it became our standard item for the aristocratic rank Baron/ Baroness (though we had such an item already. It was not a double record, just a messy thing. I dissolved it this morning and hope I was right, spotting the person that had to be identified here as Ivonig Picard or Yves Picart. We are working on a collective database, others are working with everything you create - so things you create can develop into directions you did not intend. I am trying to make sure that others know what an item is all about and what project can do with it.

What will others be able to do with your work? wine labels - how will these items by used by others? What will they demonstrate? I cannot tell. You are free to create these items if you feel, this is research others can use in their own research projects in the future, also if this is only a test case showing structures another project can use with entirely different materials.

So maybe it is time to define the project, to give it a name and to offer a couple of first searches on a project page I will create with that name; so that others can relate to this work. Another thing: The Wine labels - why are they in square brackets? Would it make sense to create them with labels like

Wine label "XYZ"

as we create them for lodges or letters?

So far all this is somewhat mysterious here on FactGrid, more difficult to understand than the input of Bretonic authors that everyone could immediately relate to as an enrichment. --Olaf Simons (talk) 15:48, 25 October 2023 (CEST)

PS. Abstract Art is now Item:Q640885. If you look at Item:Q636301 - it is not really user friendly. No one knows the Icon-classes, no one has any idea what is depicted here until you start clicking through the links. Search for a motif or style on the Query service - this is equally impossible. Maybe it is nice as it saves you of the problem to create items (that should then be connected to the respective icon classes) but otherwise it is isolated work, very difficult to relate to. --Olaf Simons (talk) 16:28, 25 October 2023 (CEST)

Dear Olaf, thank you for your help with Item:Q636357.

I wasn't aware there were project pages (I am discovering the database as I go) but we're going to create one ASAP with my colleague Jean-Baptiste (he already has some searches on hand for that).

As regards the brackets around the titles of our wine labels : in our internal work, we're used to following IFLA's recommandation when it comes to built-up titles like those of our wine labels; they suggest using brackets. If it makes more sense for the data here to have quote marks, we will change it. (as a side-note, IFLA reference : https://www.ifla.org/references/best-practice-for-national-bibliographic-agencies-in-a-digital-age/resource-description-and-standards/bibliographic-control/international-standard-bibliographic-description-isbd/)

Concerning now the "user-non-friendliness" of the IconClass IDs : it's an issue we're aware of; Jean-Baptiste is trying to find a way to solve it by showcasing straight-away the concept behind the ID. --Marie-Alice Le Corvec (talk) 11:23, 14 November 2023 (CET)

My preference is: create the items of image topics and we see how they will be used by other groups - and, of course, link them to Iconclass-IDs as that is the superb way to understand where we see similar motifs on other pictures. You are at the beginning here (which is an uncomfortable situation - it is nicer to have the objects to connect already existing), but we have seen that things develop once they show that they are yielding interesting research results. --Olaf Simons (talk) 12:58, 14 November 2023 (CET)