Post by Martin Doerr (19 August 2025)
Dear All,
I propose the refinement of the scope note of E31 Document. In the past. there had been competing and contradictory interpretations about a supposed implied truthfulness of the content of an E31 Document, also with implications for CRMdig, CRMsci and CRMpem.
Therefore here my attempt of a more elaborate clarification, which follows in principle the well-known and well received concept of "non-fiction" in library science.
To be discussed.
Martin
Post by Dominic Oldman (19 August 2025)
Hi Martin
Perhaps it could be more compact and self contained? - I have swapped some bits.
- This class comprises identifiable immaterial items that are intended to make propositions about reality, often called non-fiction. Deviations with respect to reality are typically unintended errors or poorly supported assumptions, but also different cases of bias discussed in the scientific discourse.
- These propositions may be expressed in text, graphics, images, audiograms, videograms or by other similar means.
- Typical examples are scientific records and studies, observational data, realistic portraits, depictions of landscapes or buildings, log books of ships and many others. Documentation databases are regarded as instances of E31 Document.
- For clarity, an historical novel which may incorporate or is based on facts is not considered a document because it is primarily fiction. Fiction consists of narrative mainly based on imaginary events, people, places, etc.
My Reading Notes:
I had to look up Sinuhe the Egyptian to see what you meant (it's a historical novel!) - but to keep it accessible - is a general statement about historical novels clearer? I wasn't sure about the IT document differentiation. I didn't know what the Freemason's article was, so it didn't add anything.
Hope this is helpful,
Dominic
Post by Martin Doerr (19 August 2025)
Dear all,
Here the missing references:
The "Taxil Hoax", a farce about the freemasons not intended to be taken seriously, but getting "viral" as being real despite of his confirmation that it was a hoax.
- Wikipedia https://share.google/2E2q3ooy4ckLh5N74
and Mika Waltari, "Sinuhe the Egyptian", 1945. This is a particularly nice example, because it creates a long story around a very small Egyptian text.
Post by Martin Doerr (20 August 2025)
Dear all,
Here my reworking, based on Dominic's comments. I prefer a more verbose form.
Best,
Martin
Post by Dominic Oldman (20 August 2025)
Hi Martin,
It's got bigger!
You give a comprehensive description of Documentation.
D
Post by George Bruseker (24 August 2025)
Dear Martin,
Thank you for your proposed refinement of E31 Document. I would like to share some concerns, as I remain unconvinced that this scope note alteration is a useful addition to the CRM.
My unease comes from the fact that E31 seems to be an epistemic rather than an ontological category. It feels a little like having “hero” or “villain” as subclasses of Person: whether something qualifies depends on perspective, not on universally observable qualities. This seems at odds with the CRM’s “neutral point of view” principle, which I regard as essential for an international standard aiming at a reality-based, shared representation.
The new scope note, as I read it, emphasizes the intention of the author as the criterion for classification. That is problematic: we rarely know the author’s mind, and even if we could, intention is not a stable ontological feature. A few examples illustrate the difficulties:
Samuel George Morton – Crania Americana (1839): a phrenological work that strongly influenced scientific racism. Was it “about reality”? By intention, yes; by our current understanding, no.- Herbert Risley – The People of India (1908): similarly framed as objective classification, but now widely regarded as pseudo-scientific and harmful.
- The Popol Vuh (16th century): intended as an account of reality, but referring to gods and cosmologies not aligned with the CRM’s “reality” domain.
- Johann Becher – Physica Subterranea (1667): a major work of phlogiston theory. Again, intended as truth claims about the world, but about entities that (in our episteme) do not exist.
In all these cases, deciding whether they are “documents” requires epistemic judgement, not neutral description. Moreover, for some, classifying them as “documents about reality” risks forcing modellers into positions they may find objectionable.
The cleaner modelling solution already exists: all of these works are unambiguously E73 Information Objects. Their treatment as “documents” in particular times or contexts can be captured through P2 has type or via event-based modelling (e.g., E13 Attribute Assignment). This allows us to represent the social fact of “taken as documentation” without building epistemic judgements into the ontology itself.
In short, I think E31 complicates rather than clarifies. The CRM’s strength has always been to stick to the neutral “facts of the matter.” For that reason, I would argue against expanding or refining E31 and instead rely on the mechanisms we already have.
Best regards,
George
Post by Dominic Oldman (24 August 2025)
A comment....
A document can be nonfiction or fiction. The act of assigning E31 Document to a document requires that a decision is made. Let's say I have an archaeologist's notebook which has text and sketches which documents the aspects of an ancient site. The scope note says "implicit or explicitly expressed intention". So either there is something empirically explicit that is evidence of the purpose/intent of the document, or there is something implicit that is also evidence of the intention. It could be that it is part of an official process or job, but implicitly it could be the context in which it took place. I don't think the scope note says that we have to subjectively guess what the creator was thinking in their mind (George is this what you are saying? - "we rarely know the author’s mind"). There is no separation of mind from matter.
The intention should be empirically apparent. The work is assessed in terms of its overall context which will inevitably indicate the intention of its creation. The archaeologist documents a site while being paid or funded to document the site as part of a formal archaeological process/project. In other words, intention is empirically 'baked into' the concept of this immaterial concept. Perhaps some more words are required?
This issue seems to overlap with other past conversations.
?
D
Post by Martin Doerr (25 August 2025)
Dear all,
Thank you for your comments. I suggest to discuss this orally in the next meeting. The positions are clear enough and well respected.
Please let me remark that all categories in the CRM or any ontology have undecidable border cases, as I clearly showed in my recent paper about Identifiable Individuals. The question is therefore not, if they exist, but if they are significant in practice.
Since librarianship and science has a well-established tradition in this respect, I believe the SIG has to decide if the ambiguous cases exceed the utility of the concept.
Kind regards,
Martin
Post by Athina Kritsotaki (25 August 2025)
Dear all,
just a comment: the epistemic nature isn't also confirmed by the argument that the E31 are also kind of propositions, statements about real, non fictional things (instances of E89 too)? I understand George's argument about the dependence on the intentionality to define such things, but should we consider that part of the identity criteria of the man made things is the functional use of these objects (resulting from a human activity) ?
just a thought,
BRs,
Athina
Post by George Bruseker (25 August 2025)
Dear all,
It seems indeed a great topic for a live discussion and useful that it has been raised. In the meantime, I'm sure there are many other points of view that should be heard. Glad to hear Athina and Dominic's. Look forward to hear other thoughts and to discuss them fully together.
I certainly appreciate that it is key to capture the notion of document / documenting / documentation in the modelling, it's just a question of which is the best way.
Best,
George
Post by Oyvind Eide (26 August 2025)
Dear George, and all,
OK, I will say something, but I promise it won't help you. Due to the change of dates for the October meeting, it is now colliding with teaching, so I cannot be there -- so I am here instead.
The understanding of intentionality and the difference between fiction and non-fiction are surely not in line with important areas of research today. I have colleagues within ten meters from me who would have things to say about the concept of document too. Thus I would be happy to sit down with my pipe and explain why this is crap.
However, someone also said that modelling is simplification. And then the question is not what some significant researchers write about this, or even a state of the art understanding (hard to find that one in literary studies, but in principle), but rather if the simplifications are useful for the purpose.
Which boils down to two things:
1. Is a more fine grained and up to date model possible to develop, given the significant variation of theoretical positions, even within one discipline, and surely across the number of disciplines involved here?
2. Is a more complex model necessary to handle the use cases?
Given my understanding of the discussion and the use cases as they are outlined, both questions would have to be answered with "probably not". If this is true, then two questions remain:
1. Is the model as it is suggested harmful?
2. Will future use cases generate the need for non-monothonic changes to the model?
These two questions I cannot answer.
There are other interesting topics here too, for instance including epis and onto. Those I leave for people who have more to say about those parts.
Regards,
Øyvind