Address eventDate do's and don'ts

[tucotuco]

Issue borrowed from tdwg/dwc#100.

[debpaul]

@tucotuco a great opportunity to use the data examples we have in our respective piles - to show what research can and can't be done if dates are in question. Also, to show how one might fix. And, to show what programming (scripting) can and can't fix in this regard. Might be a chance to connect with researchers (collectors) on this point as well so that future data is unambiguous in this respect.

[ekrimmel]

Comments from DwC Hour #13 suggest that it could be helpful to clarify that for specimens, eventDate should be the collecting event. For extant organisms users may assume that the collecting event coincides with the time in which the organism was alive, but for paleo/zooarch objects this is not true.

[dagendresen]

If dwc:eventDate should be strictly conforming to the "collecting event date" (domain = Occurrence?), then event-core datasets (when using the Darwin Core archive model) might need a separate (new?) eventDate term with domain = Event ...?

[ekrimmel]

Ah! I didn't mean to suggest that DwC recommend restricting the use of eventDate. What I meant was that in our documentation here we could clarify that for paleo/zooarch occurrences eventDate is typically not the same as the date when an organism was alive or in use.

[tucotuco]

Yet, I have counter examples in the zooarcheological specimens published to VertNet, where they do use the chronometric date range as the eventDate so that the meaning of when the organism occurred "in nature" is preserved. There will need to be a clear resolution on this issue going forward. The mix of concepts is not a good thing, and the use of only a collected or observed date is limiting to the questions people want to investigate.

…

On Mon, Jul 2, 2018 at 11:56 AM, Erica Krimmel ***@***.***> wrote: Ah! I didn't mean to suggest that DwC recommend restricting the use of eventDate. What I meant was that in our documentation here we could clarify that for paleo/zooarch occurrences eventDate is typically *not* the same as the date when an organism was alive or in use. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#107 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAcP6z1c2T96mUIjdrSTqST5uwkvJFASks5uCjSugaJpZM4P2F1f> .

[debpaul]

seems clear that separate fields are needed where taxon is not extant to differentiate date collected/observed from time the specimen itself would have existed. Both are important to capture. @tucotuco, since as you say "there will need to be a clear resolution on this issue going forward, " what's the next step to getting it addressed (getting a field/fields added)?

[tucotuco]

The next step is for someone (it can't be me, by convention) to follow the procedures in section 3 Change Process in the TDWG Vocabulary Maintenance Specification [1] for submitting requests for new terms in the Darwin Core, paying special attention to section 3.1 Justifications for Change.

[1] Vocabulary Maintenance Specification: https://github.com/tdwg/vocab/blob/master/vms/maintenance-specification.md

[hollyel]

So, are we suggesting that we need new terms so that we could, for example, maintain eventDate as the collecting/observation event and then a new term as the date the specimen was alive? Would the proposed chronometric extension then be an expansion on that new term for more thorough information with a paleo/zooarch context? (or the opposite, although I would observe it seems most records tend towards eventDate being the collecting/observation event, at least for paleo)

Where does that leave @dagendresen's concern though?

[tucotuco]

I think that is why it is a big community discussion. The meaning of eventDate is pretty clear and entrenched, it just is not as useful as it could be.

…

On Mon, Jul 2, 2018 at 6:22 PM, hollyel ***@***.***> wrote: So, are we suggesting that we need new terms so that we could, for example, maintain eventDate as the collecting/observation event and then a new term as the date the specimen was alive? Would the proposed chronometric extension then be an expansion on that new field for more thorough information with a paleo/zooarch context? (or the opposite, although I would observe it seems most records lean towards eventDate being the collecting/observation event, at least for paleo) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#107 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAcP6580NDc6L2tiyuVCphHsjfxuE-J0ks5uCo8ggaJpZM4P2F1f> .

[dagendresen]

How about simply use the rule that if the dwc:eventDate is used as an attribute for an dwc:Occurrence then the eventDate has domain dwc:Occurrence (i.e. date is describing the observation or sampling date, and not the time the organism was alive). And when dwc:eventDate is used as an attribute for an event then dwc:eventDate simply has domain dwc:Event, and the type of event decides what type of date is described...? Minting new date terms for all types of events seems a never-ending story to me. But perhaps the time the organism was alive is specially important enough?

[tucotuco]

Still further consideration about best practice for eventDate can be found at tdwg/bdq#86 (comment).