use of Boundary-Sigma_Boundary and Boundary-Tau_Boundary

[gianmarc...@gmail.com]

Dear All,

I was wondering if anyone here knows of any published example of the use of Boundary:Sigma_Boundary, and Boundary:Tau_Boundary (figures 4 and 6 in this page: https://c14.arch.ox.ac.uk/oxcalhelp/hlp_analysis_oper.html).

In particular, I'd like to read about practical applications of those, and in which situation(s) they can be employed.

Thank you

Best

GmA

[Richard Staff]

Hi Gianmarco,

The converse (Tau_Boundary:Boundary) is often used in archaeological situations where you have a destruction layer, indicating that the majority of 14C ages are likely to closely predate the destruction, but with a decreasing probability of dating older material.

In an environmental context the Tau_Boundary & Boundary combination (both ways around) was applied by, e.g., Vandergoes et al. 2013 (https://doi.org/10.1016/j.quascirev.2012.11.006), to account for 14C analyses of plant macrofossils dated immediately below and immediately above their tephra layer of interest, accounting for the likely close chronological relationship between the 14C-dated materials and the date of the volcanic eruption, but acknowledging that there would be a decreasing probability of the 14C ages falling significantly earlier and later than the eruption. The same was also done by Egan et al. 2015 (https://doi.org/10.1177/0959683615576230). 

We also used a similar approach (Boundary:Tau_Boundary) (Liu et al. 2016, https://doi.org/10.1016/j.quaint.2016.02.059) to account for sampling bias when trying to date the "first appearance" of millet in different regions - the assumption being that our sample selection was more likely to result in 14C ages relatively close to the "true" first appearance of the species, with a decreasing probability of the 14C ages falling significantly later. (I'm sure that this is an assumption that could be debated!)

I don't remember seeing Boundary:Sigma_Boundary applied, so would be equally interested to hear from others if this is something that they have/they have seen applied.

Best wishes,

Richard :-)

---

From: ox...@googlegroups.com <ox...@googlegroups.com> on behalf of gianmarc...@gmail.com <gianmarc...@gmail.com>
Sent: 29 May 2019 06:41
To: OxCal
Subject: use of Boundary-Sigma_Boundary and Boundary-Tau_Boundary

 

--
You received this message because you are subscribed to the Google Groups "OxCal" group.
To unsubscribe from this group and stop receiving emails from it, send an email to oxcal+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/92ed7faa-4656-473d-a543-b46150b2ea19%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[gianmarc...@gmail.com]

Dear Richard,

thanks for the detailed reply and for the references, which I will read ASAP.

In effect, my question meant to wrap my head around how model 14C dates from a floor level in a prehistoric context.

There is a lot of ethnographic and ethnoarchaeological literature about relaxation in maintenance level (in settlements) in anticipation of a site abandonment. Therefore, assuming that in a dwelling the floor will not be swept as carefully as before (or, indeed, will not be swept at all) in anticipation of abandonment, I was wondering if it makes any sense using the pair Boundary:Sigma_Boundary in order to "reflect" a situation in which older events are more "represented" than more recent ones. It is somewhat like when dealing with coin hoards, where older coins are more abundant than coins close to the date of deposition of the hoard.

I do not know if the above makes sense, and I am open to suggestions.

Should that make any sense, I am wondering what would be the implication of using Boundary:Sigma_Boundary as opposed to Boundary:Tau_Boundary.

Thank you again.

Best

Gm

To unsubscribe from this group and stop receiving emails from it, send an email to ox...@googlegroups.com.

[Richard Staff]

Cheers Gianmarco,

There is also the Zero_Boundary<->Boundary combination which gives you a linear rise/fall, rather than the exponential rise fall (using Tau_Boundary<->Boundary) or the Gaussian rise/fall (Sigma_Boundary<->Boundary).

I guess the important thing is to be able to justify (based upon your archaeological assumptions) why any of those distributions is the most likely (as you have started to describe below), and build that in to your model prior.

You can (and probably should!) run a series of models, using those different model configurations and assess what difference it makes to your event(s)/parameter(s) of interest (presumably here, the ultimate site abandonment). It may well be that the output for that event is relatively insensitive to the choice, which you can argue as giving robustness to the quoted age range.

I can imagine (based on my own experience!) that it could be hard to be certain as to which of those three rising/decreasing Phase distributions (the result of the Boundary combinations) is ultimately 'correct'.

I guess too that you could also use, e.g., the model agreement indices to see which configuration better matches your data. Though this might not be a foolproof assessment, especially if you only have a relatively limited number of 14C dates.

Good luck!

Richard

---

From: ox...@googlegroups.com <ox...@googlegroups.com> on behalf of gianmarc...@gmail.com <gianmarc...@gmail.com>

Sent: 29 May 2019 11:17
To: OxCal
Subject: Re: use of Boundary-Sigma_Boundary and Boundary-Tau_Boundary

 

To unsubscribe from this group and stop receiving emails from it, send an email to oxcal+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/oxcal/97a4cf91-25ad-47ab-88fc-641ffd3c0008%40googlegroups.com.