Frequently Asked Questions

What is the scope and purpose of the MMM portal?

The MMM portal contains most of the data found in three sources : Bibale, Medieval Manuscripts in Oxford Libraries, and the Schoenberg Database of Manuscripts. The emphasis is on the history and provenance of manuscripts, rather than textual or artistic traditions. The data have not been corrected or amended, and they have been harmonized rather than merged. This is why you will sometimes see differing and even contradictory details about the same manuscript. But references to the same person, organization, place, work, or manuscript from different data sources have been combined, as far as possible.

Why are some elements from the source datasets not included in the MMM Portal?

The MMM project focused on the history and provenance of manuscripts, and elements from the source datasets were selected on the basis of their relevance to this research area. Some elements, especially those unique to one of the datasets, were not mapped into the MMM data model. Two examples are (1) Bibale’s distinction between manuscripts which are simply recorded (« attesté ») and those which are actually preserved today (« conservé ») ; and (2) the bibliographies for manuscripts in Bibale and the Oxford catalogue. Links to the original records in the source datasets have been provided, so you can always see the full information about any given manuscript in the data source.

Where can I find more detailed information about the MMM Portal?

As well as the « About » and « Help » sections of the Portal, you can also watch this Schoenberg Institute video on YouTube or read the Project’s blog, which has links to publications and conference papers.

What is the difference between « Collection » and « Owner » ?

The “Collection” information comes only from the Bibale database, whereas “Owner” information may come from any of the three data sources. This is why they produce different numbers of manuscripts for the same person. For example, Claude Fauchet as an “Owner” owns nine manuscripts but his «Collection » contains only one manuscript. Vossius owns seventy manuscripts, though his collection comprises only five. Use « Owner » to get the fullest picture of a specific person’s manuscript holdings.

Why are some aspects of manuscript ownership and history categorized as « unknown » ?

This reflects the contents of the source datasets, where specific data elements may not be specified or the information (e.g., « Production Place ») is not known. It is possible to select  “Unknown” in the relevant filters.

Is it possible to select multiple production places when filtering manuscripts?

The place-names are arranged in a hierarchy, and it is possible to select multiple values from the same level of the hierarchy. When a parent place is selected, all the underlying places are automatically selected. If you select « Europe », then you do not have to select separately each country and town within Europe. In some cases it would be good to be able to bypass this automatic selection and select only one place at a time; we are working on an enhancement to enable this.

Is it possible to combine values from within one filter, e.g., to find those manuscripts owned by both Phillipps and Beatty?

The default combination is « OR » within one filter. So selecting both Phillipps and Beatty in the « Owner » filter for Manuscripts will produce a list of manuscripts which were owned by either Phillipps or Beatty. We are working on an enhancement which will give you the choice of applying « AND » within one filter (to find manuscripts which were owned by both Phillipps and Beatty).

« AND » is the default combination between two or more different filters. Applying « OR » between two different filters  would contradict the general purpose of faceted searching.

Why are all the different stages in the migration of a manuscript not shown on the map?

It is possible to visualize the place of production and the last-known location of a manuscript on the maps, together with the arc joining these two places. But the information in the source datasets is not full or consistent enough to show the intervening stages in the history of a manuscript’s ownership over the centuries, in chronological order. You can see this information displayed, however, in the events and owners columns for each manuscript.

How do I find manuscripts of works in the same textual tradition?

The MMM project is primarily focused on the history and provenance of manuscripts. Information about works in the same textual tradition, but which are nevertheless different works (e.g., the various chansons de geste featuring Guillaume d’Orange), has not been aggregated or linked. You will need to search for them by their individual titles.

How do I find and select multiple relevant owners (e.g., popes)?

Searching the list of owners in the « Owner » filter will show the number of results which match your search string. You can then step through the relevant results by clicking on the arrows, and selecting all the appropriate owners.

Why are there sometimes different numbers in the filtered results and the results table?

When you filter on “Europe” as the “Production Place” for manuscripts, the number [83,881] is displayed in square brackets. The results table, on the other hand, shows 72,400 manuscripts. This is because some manuscripts have multiple places of production, usually because the exact place is uncertain or unknown. The number in square brackets shows the count of production places, while the results table shows the number of manuscripts. In the case of « Europe », this means that there are an average of 1.159 production places recorded per manuscript.     

How do I connect Events and Owners for a manuscript?

In the « Manuscripts » perspective, you can click on a specific « Transfer of Custody » event to see all the available information about that event. This will display the persons and organizations involved in the event, in the « Custody surrendered by » and « Custody received by » fields.

Can I reduce the number of columns which display in the results table?

We have assumed that all the data relating to a manuscript, in particular, should be displayed in the results table. We are working on an enhancement which will allow you to select columns to be displayed in the results table.

Where does the information for the last-known location come from?
Last-known locations are mostly taken from the Schoenberg Database, but they also come from the manuscript « labels » or shelf-marks recorded in the Bibale database.

Why do places outside a filtered geographical area still display in the map visualization?
This occurs in the « Manuscripts » perspective when a manuscript has multiple production places. All those places are displayed in the map visualization, even when one or more are located outside the geographical area which has been used to filter the results.

Why do some authors have multiple entries, while other authors have only one entry?
Many authors appear in different forms in the data sources for the Portal. We have tried to harmonize as many of these as possible, but there are still some authors who have not been harmonized. They appear as two or more different names in the Portal. If you select the different forms in the Author filter, this will produce a combined set of results.

Why are there repetitions and inconsistencies in some elements of the manuscript descriptions?
This is usually because we are combining data about the same feature from two or three different sources. For production dates and places, in particular, it is quite often the case that (for example) the Schoenberg Database has multiple differing values for the same manuscript from different sales catalogues. We have chosen to display all these variations, rather than attempting to merge them or identify a single « most-accurate » value.  

Why are some authors linked to many works, even though they are the author of only one of these works?

This arises from the structure of some entries in the Schoenberg Database, where multiple works and multiple authors in a single manuscript are not individually linked. In these cases, the author is labelled « possible author ».

Is it possible to track the manuscript tradition of the different versions of a text?

The MMM portal is designed for tracking the history and provenance of manuscripts, rather than the textual tradition of a work or related works. Many works have been matched and harmonized across the data sources, so that it is possible to find all the manuscripts carrying a specific work in a specific language and to see when and where they were produced. But this matching does not usually extend to versions of the same work in different languages, so the textual tradition can best be seen by searching for the work in different languages using the various versions of its title (e.g., City of God ; De civitate Dei).

Why are pie charts and percentages not displayed for some filter results?

We have chosen to display pie charts with percentages where the data are suitable for this kind of analysis. If you think pie charts would be useful in other areas, please let us know.

Can the date slider be made more precise ?

We are considering replacing the date slider with « from » and « to » boxes for date ranges.

Why do some manuscripts have multiple map locations for their production place?

When a manuscript has two or more production places recorded in the data sources, it is located at each of the production places on the map, even when one is more specific than the other(s).

Can I export the geographic coordinates of the production places?

The CSV export  via the Yasgui service does not contain geographic coordinates for places in the MMM data. It is possible to find coordinates by searching the SPARQL endpoint, since they are present in the MMM data as well as in the place vocabularies used by MMM. Otherwise, the best way of exporting a map is via a screen capture from the relevant MMM visualization.

How can I get help with constructing SPARQL queries for the MMM data?

The project’s GitHub site has a guide to constructing SPARQL queries, together with the MMM Data Model and schema.

In the SPARQL CSV export file, why is information about a single manuscript presented across multiple rows?

This comes from the design of the underlying SPARQL queries. Because there is so much information about each manuscript, often with multiple values for the same element (e.g., production date), it would be difficult and possibly misleading to combine all the information on to a single row. Once the CSV file has been exported and downloaded, it is possible to combine multiple rows using software like OpenRefine or Google Sheets.

Can I see some worked examples of queries against the MMM Portal, and of queries against the SPARQL endpoint?

Worked examples of queries against the MMM Portal (and the source datasets) can be seen in the project’s list of sample research questions. There are some sample SPARQL queries in the MMM SPARQL tutorial, available on GitHub.