The Ever-Expanding Data-verse

We are back with the final installment of our data-collection adventure.  In our previous posts we have talked about a number of important sources we have been using to collect already existing radiocarbon measurements (what we would call ‘legacy data’). These have included lists produced and published by radiocarbon labs, large datasets compiled by researchers and publications from research projects. However, there are a few more very important sources that we have not talked about, until now….

The first is the Historic Environment Record (HER). A HER is a comprehensive database of the archaeological sites, historic buildings and historic landscapes of an area, and there are over 85 HERs across England, managed by local authorities. In Wales, the HER databases are run by the four regional Archaeological Trusts; Clywd-Pows (CPAT), Dyfedd (DAT), Gwent-Glamorgan (GGAT), and Gwynedd (GAT). In Scotland HERs are managed by local authorities and overseen by one over-arching body, Historic Environment Scotland. In Northern Ireland the Sites and Monuments Record (NISMR) is managed by the Department for Communities, and in the Republic of Ireland the Record of Monuments and Places (RMP) is maintained by the Department of Culture, Heritage and the Gaeltacht’s National Monuments Service.

The Heritage Gateway portal for searching English HER records (left), and the map search function on Canmore for searching Scottish HER records

These contain a wide variety of records, from prehistoric findspots with a singe artefact to large multi-period archaeological sites and historic landscapes.  Furthermore, HERs are an ever-expanding resource; as new sites within the historic environment are investigated, the results are deposited with the local HER. This means they have up-to-date and comprehensive details about the historic environment within their local area. More importantly (and continuing our open access theme from previous weeks), the information HERs hold are public records, and anyone can access them, either through online HER records (visit the sites for England, Scotland, Wales, Northern Ireland and Republic or Ireland) or by making an enquiry. HERs are also used in the development process, where the detailed data held by them is used to characterise the historic environment in particular areas ahead of excavation and development.  For Project Time, we contacted HERs across England, Wales and Scotland, and requested records for any site in their area which had recorded radiocarbon measurements for sites recorded as being ‘Neolithic’ or ‘Bronze Age’. Then, all we had to do was look at the radiocarbon measurements, find those that fell within the 3500–1500 cal BC range, and add them to our database (if we didn’t have them already from another source)! In Ireland there is no centralised database of radiocarbon dates and instead we used the data collated by Robert Chapple, as featured in our previous blog post.

The final data source we explored is the data produced by commercial archaeology units, who undertake excavation ahead of development as part of the planning process (as well as doing research excavations in some cases). We explored this data in two ways. The first has been using the enormous library of unpublished archaeological reports (also known as ‘Grey Literature’) produced by commercial units. In the UK, this is hosted by the Archaeological Data Service (ADS). Here you can search the library, containing over 60,000 reports from commercial excavations, and once again, it is accessible to everyone! Using this, we searched for just those sites that included ‘Neolithic’ and ‘Bronze Age’ archaeology, then reviewed the reports to find any radiocarbon measurements that may have been taken. This is, as you might imagine, quite a labour-intensive job, but thankfully the project undertaken by Bevan et al 2017 included an online crowd-sourced review of the grey literature on the ADS website up to February 2016.  As we have incorporated the Bevan et al 2017 data into our database (see earlier blog post), we only needed to review reports archived since February 2016, and add any new sites.

The ADS grey literature library

In the Republic of Ireland all site reports must be submitted to the National Monuments Service but these reports are not available online. However, one of the major instigators of commercial archaeological work is Transport Infrastructure Ireland (TII) and the reports associated with their work are all available online in the TII Digital Heritage Collections hosted by the Digital Repository of Ireland. The collection currently contains 2742 text objects, mostly archaeological reports. The only reliable way of searching through this resource was to open each report and to search for the word ‘radiocarbon’. There were no shortcuts in this case!

Finally, we have also been working with a number of commercial units directly who have been kind and generous enough to share their own databases of radiocarbon measurements. We have used this data to pinpoint sites that fall within 3500–1500 cal BC. We then used the reports of these sites, from the ADS library, from the units’ own open-access report repositories (such as Oxford Archaeology or Cotswold Archaeology ) or from reports sent to us directly by the excavators,  to collect all the required information.

And that is about it-and what is the result of all this, I hear you ask? Well, as we come to the end of our data collection, we have collected nearly 15,000 radiocarbon measurements, ready to be modelled! Come back to us in the new year, as we begin this next exciting phase of the project!


Back to the Source

Last time on the blog, we were talking about some of the really important sources of radiocarbon data that we have been using in our data collection for Project Time.  Between these sources, it feels like we have managed to collect a lot of (perhaps even the majority of) the available measurements for 3500–1500 cal BC. However, the radiocarbon data is not presented in the same way in all of these sources, and in some cases some details that we need are missing. Sometimes this might be details of the measurement itself (such as the δC13, which is important in assessing the accuracy and precision of the radiocarbon measurement, as we discussed a few blog posts ago, or it might be details of the sample used, or where on a particular site, or in a particular archaeological feature or context a sample originally came from.  It is really important we collect this information for every measurement, as we need details of the sample, and where the sample came from (and its stratigraphic relationship to any other samples on the site) in order to use them in the Bayesian statistical models we will be creating for our sites in the near future.

So, we have a big list of measurements from a range of different sources, but some are missing some of these fields of data – where do we get them from? Well, we go back to the source, and for this, that means the official lists of measurements published by radiocarbon laboratories.  These are publications in journals, such as Radiocarbon or Archaeometry, and they report all of the measurements produced by the laboratories in the time since the last list was published (as long as the submitter agrees) – these can include sites from Britain and Ireland, as well as sites from around the world.  A wide range of radiocarbon laboratories publish date lists, including the British Museum (with 26 published lists spanning 44 years), Harwell (11 published lists over 18 years) and the Oxford Radiocarbon Accelerator unit (with 33 published lists since 1984, spanning a whopping 785 pages!). A number of these, such as the Oxford AMS date lists, are free access too (continuing our open access theme from the last blog post). Within these date lists, we can find the radiocarbon measurement, lab code, sample details, δC13, and in most cases, a discussion of the measurements in relation to the site. 

A section from one of the Oxford AMS date lists, detailing measurements taken from Parc le Breos Cwm in Wales.

One slight problem is that not all of the radiocarbon laboratories produce published date lists and some labs produce so many radiocarbon measurements these days that it is not always practical for the lab to produce comprehensive date lists. However, for sites in England, many of the results can be found in the catalogues of radiocarbon dates published by Historic England (formally English Heritage). These provide full technical details of all the radiocarbon measurements funded by them, and currently span projects between 1970–2007 (which can be found here).  For sites in Scotland, Wales or Ireland, there is not a direct equivalent of the Historic England date lists, however, we can use the radiocarbon databases for these areas that we discussed in the last blog – these have references to the original publication of sites, which will have the radiocarbon measurement details in them (if we can’t find them on a published lab date list).

The published date lists have been vital sources during our data collection; it has been quite regular to find radiocarbon measurements mentioned in publications, project databases or other sources (which we will discuss in our next blog post), but with some of the details missing. These details are crucial because we need them when producing Bayesian models. The date lists are where we go to confirm the information we have, and to fill in any gaps.

You might think this wraps our data collection process up, but you would be wrong! Come back next time for the final installment of our data collection run-down.

The many strands of radiocarbon data

We finished our last blog post by promising to take you on a digital tour of the major sources of radiocarbon data we have been scouring during our summer of data collection. So here we are, to start that tour-but it is going to be split up across three posts, simply because there are so many important sources (and important people) that we need the space to get them all in!

One of the nonhuman members of Project Time consults Gathering Time, one of a suite of important sources of existing radiocarbon data (because dynamic photographs of data collection are difficult)

It might sound like the aim of Project Time, to collate all the existing radiocarbon measurements from Britain and Ireland within 3500-1500 cal BC, is quite a daunting task (it certainly did to us…). However, we are not the first project that has sought to collate radiocarbon data; this means there are a range of existing data sources for us to consult, which focus on particular areas, time periods, or types of archaeology. Perhaps even more useful, however, is the fact that these are open access sources, which means the data is freely available-so if you have a deep yearning to undertake a review of radiocarbon data, you can uses these sources too!  So, here, in no particular order, are some of the most important datasets and datelists that we have been using:

  • ‘The Archaeological site index to radiocarbon dates from Great Britain and Ireland’ is a huge dataset, containing over 15,000 radiocarbon measurements. Originally compiled by Cherry Lavell for the Council for British Archaeology, and added to by people including the Scientific Dating Team at Historic England (then English Heritage), it is hosted on the Archaeology Data Service website. Here, you can query the database for records based on a range of variables, including keywords, dates, or site types. Have a look for yourself here.
  • The Scottish Radiocarbon Database includes legacy data. The best way to access this information online at the moment is by searching for individual sites within the National Record of the Historic Environment (CANMORE), and then clicking on the link on the “C14 Radiocarbon Dating” link. For example, the results from the timber enclosure at Meldon Bridge  can be found here. We have added more recent dates to the dataset by collating lists of radiocarbon dates from Discovery and Excavation in Scotland.
  • For Wales, there is the Wales and Borders Radiocarbon Database, which was created, and continues to be added to by Steve Burrow and Sian Williams, at the National Museum Wales (which can be found here) . 
  • In Ireland, Robert Chapple has spent years collating his own set of radiocarbon dates, and has made this freely available to everyone here. This collection includes more than 10,000 radiocarbon measurements and we have been concentrating on the results that calibrate between 3500 and 1500 BC.
  • There are also a number of previous projects that have collected RC data and have made their datasets available to researchers.  These can cover multiple time periods (such as the study by Bevan et al 2017 examining Holocene population fluctuations in Britain and Ireland), or they can focus on specific periods and practices, such as ‘Gathering Time’, ‘Cultivating Societies’  and the ‘Beaker People Project’.

Proof that the radiocarbon datelist hosted on the ADS really is for everyone

It’s really important to acknowledge that all of these sources are the result of a huge amount of work by those involved in these projects (and in our final publications, they will all be fully referenced). But perhaps more important is the fact that the data is made freely available – after all, archaeology should be for everyone, but to do that, we have to make our data accessible!  So that’s it for now, but stay tuned for the next thrilling installment of our tour of radiocarbon sources, and as always, if you are part of an archaeological project, please get in touch on twitter (@projecttimearch) to let us know if you have data that you think we might have missed.

The Long Hot Summer of Data Collection

Things have been a little quiet here on the blog over the summer, but that is because we have been hard at work undertaking one of the largest things on our project ‘to do’ list: data collection. Although we are going to be submitting new samples for radiocarbon dating from a number of sites we are re-examining as part of the project (see our sites page), the majority of the radiocarbon measurements we will be using are what are known as ‘legacy data’. These are all of the radiocarbon measurements that have been generated by archaeological excavation and research in the eight decades since radiocarbon dating was developed and widely used.  We are aiming to collate as many radiocarbon measurements that calibrate within the period 3500 to 1500 BC from sites in Ireland and Britain as possible. In particular, we are collecting a number of details about radiocarbon measurements-let’s take this measurement from Wyke Down 1 as an example:

Lab CodeMeasurement (BP)ErrorδC13Sample
BM-23964140±80-22.9‰Oak Charcoal, Heartwood

Here we can see there are a number of pieces of information we are interested in for each radiocarbon measurement:

  • The laboratory code for the radiocarbon determination (this allows us to refer to the specific preparation and measurement methods used by the laboratories)-for this example, ‘BM’ tells us this sample was submitted to the British Museum.
  • The uncalibrated radiocarbon measurement (in this case, 4140 BP, where BP stands for “Before Present” and where present is 1950 AD)
  • The error range of the measurement (in this case ± 80). We use the uncalibrated measurement and the error range to produce a ‘calibrated’ date – you can read more about calibration here on Oxford Radiocarbon Accelerator Unit’s website .
  • Quality control measurements , most notably the δ13C measurement (and we also collect the percentage collagen yield and C:N ratio for bone samples). These are important for assessing the accuracy and precision of radiocarbon measurement – you can read more about quality control measurements (also referred to as corrections for isotopic fractionation) on Beta Analytic’s website
  • Details about the sample submitted, including the material (such as charcoal, animal bone, human bone, carbonised residue), and more specific details (species, specific parts of trees, skeletal elements). In this case, we have the material (charcoal), the species (oak), and where from the tree the sample is from (heartwood).  In this case, the fact that this sample is from oak heartwood is really important, because the measurement obtained from this sample identifies the year that the specific tree ring grew (as opposed to the moment the tree was cut down).  Given that Oak can live for hundreds of years, this measurement may, therefore, overestimate the age of this deposit-hence why it is so important we collect these details!

One final, but vital piece of information we need to collect is the archaeological context – where did the sample come from, how did it get there (ie, what type of deposit), and how do radiocarbon measurements from a site relate to each other. This information allows us to think about how multiple measurements from a site relate to each other, and can inform on the sequence of events at a site.  The combination of these measurements and our ‘prior beliefs’ – the archaeological relationships between the contexts from which the measurements come from – is a core component of Bayesian statistical modelling, which we will be using to model our dates once our data collection has finished (more on Bayesian modelling in future posts!). 

Where the action has been happening! Nick and Penny’s desks (complete with a feline supervisor)

So, “where have we been getting all of this data from?” I hear you ask. Well, most of this work is desk-based and in the next few blog posts, we will be taking you on a tour of some of the most important sources of radiocarbon data we have been looking at, many of which are freely available to all of us! So join us next time for a radiocarbon data adventure, and in the meantime, if you are part of an archaeological project, please get in touch on twitter ( @projecttimearch ) to let us know if you have data that you think we might have missed.

Crab apples, 3500 BC to 2000 BC

Hello, I’m Penny and I work as a researcher with Project TIME. I am an archaeobotanist, so I look at plant macro-remains (usually charred fruits and seeds) that are preserved in archaeological deposits. These are good raw materials for radiocarbon dates from excavated sites.  

I’m really excited about starting Project TIME’s database, it’s a great opportunity for me to also research lots of archaeobotanical results from the period between 3500 and 2000 BC. I’m particularly interested in looking at contexts where apple remains are found. Crab apples are native trees in Britain and Ireland and their fruit was probably utilised by people from the earliest times. However, several archaeobotanists have noticed that their remains seem more frequent in the period covered by Project TIME, particularly from approximately 3400 BC onwards (for example, see Whitehouse et al. 2014). 

Continue reading “Crab apples, 3500 BC to 2000 BC”

Project TIME is go! Or research in the cold climate of a global pandemic…

So, we’re off! After seemingly months where recruitment of the team has stalled due to the pandemic, and Seren changing jobs and arriving at the wonderful Manchester Metropolitan University (hurrah!), we are finally ready to get going… But, managing a research project that needs a lot of access to museum archives during the time of a global pandemic is not as easy as you might think. Lots of the archives that we hope to visit are currently closed and not accepting research visits, and some of the museums have closed their doors entirely, possibly forever. Throw into the mix an international research project at a time when international travel is not really possible, and we have had to significantly change how we are working on the project. This said, we can still get access to some archives, under very limited conditions. Because of this we are focusing on sites that we can get access to collections and details of the material culture. First up on our list are the sites located around Wyke Down, a really significant series of places for our time period, excavate by a number of researchers including the fabulous Martin Green, whom we hope to interview soon in our podcast series. Given all the difficulties of research work at the moment, we can’t tell you the thrill of receiving in the post an unpublished student thesis with details of the faunal remains from one of the sites. Or the excitement of the team when we start to think about actual sample selection for 14C measurement from the — did we mention this? — *really stonking* series of sites excavated by Martin and his colleagues. This is what makes archives work in archaeology so exciting, being able to revisit colleagues’ work with new techniques! We can’t wait to share some more of the research process and the results as we go.