Historic context/origin of a Collection Object

How do you record the historic context of a Collection Object ?

A great number of our collection objects – particularly in entomology – were acquired as purchases or donations of historic collections. The history of these collections is often crucial for the cultural and scientific value of the specimen and there is often a lot of accessory documentation available (e.g. publications, biographies, and notes).

However, it not so evident how to record this optimally in Specify.

The obvious first solution for us was to use the Accessions for this. However, we ran into problems and wish to indicate the historic context independently from the accession. For many old objects we do not have accession data, but on the labels it is noted to which historic collection it belonged previously (e.g. “ex collection Lamarck” etc). On the other hand, several important historic collections were acquired in several portions, sometimes spanning decades, differing thus in dates and often also accession mode (gift, purchase, legacy).

A solution would perhaps be to have a new related table named “HistoricContext” that would basically resemble the Accession table (and also be shared among the whole institution!), but instead of a numeric identifier it should have a mandatory text field as identifier to make it easier reusable (e.g. named title, content example like “Darwin Crustacea collection”)

The Projects table could be repurposed for this, however, for us it lacks the necessary attachments and references links and it is not shared among the whole institution.

If you have similar problems, how did you resolve this?

Interesting to hear about the challenges.

I have no solution to this but I have a somewhat similar issue as I’m looking for a solution that could handle both natural collections and cultural heritage collections. I’m not sure what the limitations (of Specify) might be. I assume one criteria is to be able to have different data models.

(I fully understand that this might be suboptimal but we have limited resources and it would be great just to be able to have one system for our collections.)

Yes, this is typically information that is stored associated with the legal acquisition of material for your collections - an accession. There is no need to repurpose another table for this. You could just use the accession table and change the formatter for the accession number to be plain text and insert a textual identifier for the accession.

You could just use the accession table and change the formatter for the accession
number to be plain text and insert a textual identifier for the accession.

That’s what we did, but it forces the user to compose an identifier (e.g. like “Lamarck collection purchase 1879”). However, the users hate this as for different types of accessions you have to compose different identifiers.

Additionally, this does not resolve the problem of a collection that was acquired in several instalments or for accessions for which only the mode or the date is known. For the latter we have to compose an accession number. For many users this is too complicated.
This should be done by the application.

An compromise would be to add a new text field to the Accession table named “context title” that is also searchable with he magnifying glass and which can be added to the combo-box type searches. There are some large spare text fields in the Accession table, but these cannot be searched.

At NHMD we are doing what Andy suggests, and in cases where we don’t know exact dates we use estimated or arbitrary ones, but at the same time explain this in a text field. If an accession comes in several installments, we would register the first of these and then again explain the later installments in a text field.
We do, however, encounter one problem, which is not easily solved, except by writing a comment in a collection object remarks field. Sometimes an object that is part of an accession of one collection has a history of formerly having belonged to another collection, which we would also very much like to register. For example we can have materíal from the collections of Otto Frederik Müller and Otto Fabricius registered as two accessions (as you would expect). But much of Müller’s collection was inherited by Fabricius, so it gets registered under the accession of Fabricius. It is probably a problem that will require so much engineering to solve that it is best just to make remarks, and accept that a search will not be quite easy to perform. I just mention it in case somebody has thought of something clever for cases like this.

Hi Tom,
good to hear that you have the same problems. Specify was more intended to deal with new collections, so the problems of historical collections and legacy data pose often difficulties. As I said, in this case it’s the users who have difficulties in being obliged to create an identifier according to a schema and not having something automatic like a accession number.

Hi Peter,
I totally agree with you that it is not an especially good idea to have users create their own identifiers. Not without a rigorous standard, at least, and such a thing always seems difficult for people to follow.
Your suggestion:

It appears that our Specify support people here in NHMD have actually just recently done this, and I have just been a little too slow to realize how well they did it before now. The field is called ‘Accesssion name’, is searchable when using the magnifying glass, and it doesn’t replace the accession number field, which is a rigorous (semi)automatic one.

I have also no solution, but I have been thinking about this problem off and on for a long time now. When reading this now, I was first interested in @AndyBentley 's solution with accessions, but. like @Schuchert and @TSchioette , I do not like having to use the Accession Number for something that it is not for.

Also, apart from the other problem @Schuchert identified:

we have a bigger problem in that historical specimens have often passed through more than one other collection before they finally arrived here at MEL and any Collection Object can only belong to one Accession. So, you cannot use Accessions to record the full provenance of a specimen.

When we were implementing Specify 12 years ago, we were thinking of using Exsiccata and Exsiccata Item for this, as those have the right relationship with Collection Object, but we never made serious work of that.