search

tdwg / attribution

Joint TDWG/RDA group on metadata standards for attribution of physical and digital collection stewardship
13 stars 4 forks source link

concept:transcribed #34

Open qgroom opened 4 years ago

qgroom commented 4 years ago

transcribed

   
Definition Entered the text written on a physical object into a digital resource.
Existing concept  
Existing namespace  
Existing concept identifier  
Format string
Examples "transcribed"

Notes

RBGE-Herbarium commented 4 years ago

Are we defining what text is being transcribed, eg original label, annotations, and how much of the text has been transcribed? I'm assuming not, so we could potentially have multiple transcribers either transcribing exactly the same text (eg crowdsourcing) or adding text to an existing transcription.

dshorthouse commented 4 years ago

@RBGE-Herbarium In theory, we could have multiple entries (rows) for agent-action-occurrence combinations that would indicate that multiple people participated in the transcription activity, but it is presently out of scope to say what was transcribed and to what extent. However, it does remind me that our conceptual model from the Research Data Alliance has an optional reason term, but we do not have it anywhere here in this extension. Maybe this is the missing term that could contain, "Robert did this because...".

RBGE-Herbarium commented 3 years ago

Definition suggestion: Entered the text on a physical object into a digital resource.

dshorthouse commented 3 years ago

@RBGE-Herbarium Perfect! Much better.

wouteraddink commented 3 years ago

transcribing means writing it down in another form, which is not necessarily digitized. If a written label in Chinese characters was rewritten in Latin characters it is also transcribed. So if we want to exclude this, maybe better rename to: concept: digitallyTranscribed. With as definition: Entered the written, printed, or spoken text attached to a physical object into digital text.

dshorthouse commented 3 years ago

@wouteraddink Not sure we want to exclude this, but you've certainly identified more nuance. There is (assumedly) more intellectual effort in simultaneously reading text, hand-written or otherwise, off an object & typing it into a digital form AND translating it from its native language, errors in doing so notwithstanding. Likewise, not all digital transcription events are equal in effort. The label may contain typed text through manual production whereas another label may be entirely handwritten and challenging to both read and comprehend. Are you suggesting then that we need more than one term to accurately capture the action?

wouteraddink commented 3 years ago

I have not thought this through well, but I think in your example I would like to attribute the agent for three things: helping in digitisation, entering the text and translating the text. So we might need different terms for these. It might require more effort to type a long written text into digital text than a short printed text, but i see the action as the same. If duration of the action is recorded, that could in some cases give an indication of effort that was needed (not if for example if the agent was a machine).