User talk:Velatrix/Radiocarbon calibration

Original research and other concerns

@Nicolas Perrault III: Thanks for the useful templates and for bringing attention to the problem of uncalibrated dates in articles. However, I have some reservations about this page. There doesn't appear to have been an attempt to gain consensus for the advice given and much of it contradicts established consensus.

My main concern is that we should not encourage editors to calibrate dates themselves because it's original research. Radiocarbon calibration may have become a routine procedure in archaeology, but it's still a scientific procedure. Selecting a calibration curve requires expertise. Identifying and compensating for reservoir effects requires expertise. Most significantly, reporting calibrated dates is not straightforward. Even many published archaeologists get it wrong. The advice given here—to simply state a range—falls far short of the accepted best practice, and arbitrarily "rounding" a calibrated date to a single value is completely inappropriate. Doing our own calibrations breaks verifiability because the cited source will not actually contain the date reported. Adding steps to reproduce the calibration in a HTML comment doesn't help our readers; it only highlights that the article is no longer a pure summary of published information. Asking editors to deduce whether or not a date is calibrated when not explicitly stated in the source adds another layer of original research.

I also object to advising the use of "cal BP" and "uncal BP" in article text. It's not our usual practice and it's an unnecessarily confusing detail for most readers. The vast majority of articles shouldn't even need to state individual radiocarbon dates: they should be based on summary chronologies in secondary and tertiary sources, generally using BCE/CE, or BP for early prehistory.

I'd suggest much more simple advice: don't use uncalibrated dates in articles, and if you come across them, try to find more recent sources that use calibrated dates. – Joe (talk) 07:06, 11 December 2018 (UTC)

Notified WikiProject Archaeology of this discussion and pinging @Doug Weller:. – Joe (talk) 07:18, 11 December 2018 (UTC)

For years I've been seeing editors changing BP dates to calendar years without any apparent understanding of BP. I usually presume that they want to make the dating clearer to the reader, but that results in our articles showing a sort of certainty not usually held by the original researchers. I feel strongly that we should adhere to the sources and not mess with them in any way. Any alteration of the source is by its nature original research. I agree about not using "cal BP" or "uncal BP". Uncalibrated dates aren't terribly helpful and I'm not convinced they help our readers. Doug Weller talk 09:24, 11 December 2018 (UTC)

I generally agree, but less so about not using "cal BP" and "uncal BP" in article text at all - does that extend to notes? Certainly Joe's point is broadly right but in practice lots of specific dates are always going to appear. While on the subject, I also don't much like "years ago", which ought to make things easier for the non-expert, but imo somehow doesn't. The whole subject is rather a nightmare, though mitigated by not actually mattering very much for the understanding of most of our readers. Apart from these points on the general thrust, the page as it is needs a lot of editing for clarity etc. Just checking User:Mike Christie has spotted the notice on this. Johnbod (talk) 14:24, 11 December 2018 (UTC)

I've always tried to give dates in the form used in the sources, without adding anything that is not in the sources. If the source says "calibrated" or "uncal" or something similar, then I include the qualifier in the article. I include any range given in the source. I agree with Doug that dates should be in the form used in the sources, and that we cannot interpret or modify them. Linking to WP articles about radiocarbon dating and terms used in connection with it will let readers find more information about the process, but we should not be interpreting dates beyond what the sources say. - Donald Albury 14:47, 11 December 2018 (UTC)

I was hoping Nicolas Perrault III would respond, but we seem to have a consensus at least that dates should not be changed from the source. That renders all the advice here moot. Should we move the page to Nicolas' userspace? Or to the Wikipedia namespace and tag it with {{failed}}? I think the two templates are still useful but I will edit them to remove the reference to this page and the inappropriate advice to calibrate dates. – Joe (talk) 12:49, 17 December 2018 (UTC)