Question on Representing Magnetic Tapes in RiC
Valentín Mansilla
Dear all, I am pleased to participate in this very interesting forum for the first time. I would like to ask a brief question, as I would appreciate some feedback before proceeding.
I am currently cataloguing ethnographic magnetic tapes as part of my PhD project. I was wondering whether, from a RiC perspective, it would be appropriate to use the Record entity to represent the magnetic tape as a whole and the Record Part entity to represent each track, since the tracks often have mixed provenance, involving different agents, places, dates, topics, and so on.
Thank you in advance for your insights.
Richard Williamson
Very nice question, thank you for asking it! I think what I'd first
and foremost suggest is that RiC is very flexible and
non-prescriptive, allowing its users to make the judgement of whether
this is a Record / Record Part or, say, Record Set / Record situation
:-). There is quite a lot about this in RiC-AG at the following link,
as you may well have seen already :-).
https://ica-egad.github.io/RiC-AG/faq--record_or_record_set.html
Without knowing much about it, I would probably be inclined here to
consider each track a Record, and the tape as a whole either as a
Record Set or itself a Record (this being permitted too). For it seems
to me that each track can likely be viewed as 'independently'
documenting something, i.e. there is a certain
'completeness'/'self-sufficiency' to each track in terms of what it is
documenting, which is one aspect RiC-AG emphasises as relevant when
making this kind of judgement.
An aspect to consider as to whether the tape as a whole is judged as a
Record Set or Record could be whether it would make much difference to
its 'identity' if a track or two was lost; if it would not, it might
be best considered a Record Set, otherwise it might be best considered
a Record. It is a bit difficult for me to be more concrete without
knowing more, but if one views the tape as 'a collection of
ethnographic recordings illustrating the X culture', this would likely
be more 'Record Set'-like I think, whereas 'the ethnographic
recordings made by anthropologist A on his/her field trip in 1963'
could be viewed as a Record, i.e. as documenting that field trip or
that anthropologist's work. If a track is missing in the latter case,
this would likely be important to note, as that track might e.g.
change the perception of the anthropologist's work.
Not sure if that is helpful? Just ask again if not :-). Other thoughts
from anybody out there are very welcome!
Best wishes,
Richard
> You received this message because you are subscribed to the Google Groups "Records_in_Contexts_users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to Records_in_Context...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/Records_in_Contexts_users/CAO0WP%2BmqG9XAiQh6F0VUpDTucj3dur25ndSr_zBW6hBqu8hVNA%40mail.gmail.com.
pieters...@gmail.com
Good day Valentin
Short answer
Yes. Model the tape as a Record Resource and the tracks as Record Parts (or as separate Record Resources where tracks need independent curatorial lives). Capture track-level provenance as relations and Events in the RiC graph, and store the assertion metadata (who asserted it, when, evidence, confidence) on the relations themselves - RiC's intended pattern. For AI or algorithmic suggestions, capture the algorithm run as an Agent and write inference-provenance evidence that the relation references.
Why
- RiC-CM treats context as a graph where relations carry meaning. A tape is a single physical carrier (Record Resource) whose internal components - tracks - can legitimately have different provenances. Representing tracks as Record Parts preserves the physical whole/part relationship while letting you attach separate provenance chains to each part.
- RiC explicitly allows relations to have properties (start/end dates, asserting agent, evidence, confidence). That is the right place to record the differing provenance (who asserted that Track 2 was created by Agent X on Date Y in Place Z) rather than forcing all provenance onto the tape-level record.
- If some tracks are functionally independent (different creators, reused tracks, later editing), prefer modelling tracks as distinct Record Resources linked to the tape via hasPart/isPartOf. RiC is agnostic about granularity; both are valid depending on how the data will be used.
Practical modelling pattern (entities and relations)
Record Resource: (tape) - physical carrier; identifier, physical description, carrier type.
Record Part: (track) - component of the tape; track number, duration, format, technical note.
Agent: - person/corporate/collective (performer, recorder, donor).
Activity: - creation, editing, transfer, digitisation, accession, cataloguing.
Event: - custody change, migration (tape to digital), accession, appraisal, disposition.
Place: - recording place, custody locations.
Mandate: - donor deed, copyright, access restriction.
Relations: - typed edges such as RecordResource hasPart RecordPart; RecordPart wasCreatedBy Agent; RecordPart hasSubject Topic; RecordPart hasOrHadCustodian Agent; Activity generated RecordPart. Put provenance metadata (assertedBy, evidence, confidence) on the relation.
Where to record "who asserted what"
On relations and on Events: RiC recommends that assertions and their evidence live with the relation: the relation carries properties (assertedBy, assertedAt, evidenceRef, confidence).
For machine-aided assertions (audio analysis, AI diarisation), record an Evidence/Assertion relation pointing to the algorithm run (as an Agent), its parameters, a confidence score, and a reference to a supporting artefact (spectrogram, transcript excerpt). Heratio's AI inference provenance pattern fits here as the evidence object the relation references.
Also record Events for observable facts (custody transfers, digitisation). These become part of the Record Part's provenance chain and enable path queries like "who had custody of track 3 between 1982 and 1990?"
Groete / Regards
Johan Pieterse
082 337-1406
--
Silvia Gattafoni
Hi Valentín.
Following your observation that each track may involve different agents, dates, places, and circumstances of creation, it seems more appropriate to treat each track as an independent Record and the magnetic tape as the carrier that brings them together. While the recordings share the same physical support, their contexts of creation may be sufficiently distinct to justify describing them as separate records.
I understand that, from a RiC perspective, what matters most is not the physical carrier itself but the context in which the record was created. Therefore, when a track has sufficient contextual autonomy, it can reasonably be understood as a documentary unit in its own right. The tape remains important, but as the material entity that contains multiple records rather than as a single Record composed of parts.
That said, I am still learning my way around RiC myself, so I'll be very interested to hear what the experts think!
--