GIFted Researcher

It has come to my attention recently that I don’t really blog all that much about my actual research. There are reasons for this, mostly because I didn’t think anyone else was interested – but also because I am wary of making too much available to the public before I have completed my research (and published). I’m not against open access, but I am against people not giving proper credit for research as is wont to happen on the internets.

The following is an edited version of the sets I presented recently at Science Showoff and Bright Club Manchester. As a result, this is not an entirely serious or fully comprehensive overview of my research project. But nevertheless, I hope you enjoy. If you are interested in the more (serious) aspects of my research methodology, please do get in touch as I am happy to discuss it.

For those of you that don’t know, I am a doctoral researcher at the University of Sheffield (that’s a fancy way of saying I’m a PhD student, adopted in a moment of “Oh no, what have I done?” to convince myself I’m important and it isn’t all a massive waste of time).


665 years ago, an epidemic broke out in Europe caused by the bacteria Yersinia pestis. You may remember it from such other names in history as: the Black Death, the plague, the great mortality, etc. It comes in three forms: bubonic, pneumonic, and septaecemic. If you got the latter two, you were going to die. If you got the former one, you were probably going to die. Of course, many people did survive… but still, millions of people across Europe died during this epidemic.

That’s because in the medieval period people resorted to treating the infected with anything from prayer to placing live hens on buboes to draw out the illness.

The problem is (and yes, this is an actual problem, you can ask me why in the comments if you’re interested – although almost no one ever is), to date we’ve only identified about 1-2% of the victims in England (and that’s with conservative death estimates). So, my thesis question, in a nutshell*, is “Where the hell are they?”


While it’s not often then we go out to specifically try and dig up certain people, given the sheer number of people that died from this disease, we really should have comes across more of them by now.

It turns out, medieval people are like crazy good at hide and seek. You’ve heard of Richard III? Car park?


Yeah, he learned from the pros. It turns out most of the Black Death victims have been hiding in the place where you’d least expect them. In cemeteries.

I can sense the gasps of amazement, “OMG, the powers of deduction! Sherlock’s NOT DEAD! But he appears to have been recast. Nooo… Brandysnap Tiddlywink! I loved that guy…”


Please, give me some credit. Obviously, it’s not that obvious. Basically, what’s happened is that up until this point of time we’ve been expecting to find these people in massive plague pits (because, up until this point of time, we’ve found most of these people in massive plague pits), but consider this: although a really high proportion of the population is dying all across the country during this epidemic, in urban areas this equals a high number of dead people – because a high proportion of a lot of people, is a lot of dead people – hence plague pits. Whereas in rural areas – where 90% of the population is living at this time (because let’s face it, the countryside is glorious…


(…or more accurately because they’re all indentured peasants) this ends up being a low number, because a high proportion of less people is less dead people. This means there are less dead bodies to manage after an epidemic in villages and small towns, so you can bury them the proper way. Alone. For all of eternity. But because the plague doesn’t show up on bones, telling plague-y skeletons from non-plague-y ones hundreds of years later when we excavate a medieval cemetery is really difficult. It’s likely we’ve dug up loads of them already, we just didn’t know that’s who they were.

So I’m using maths to find them. Yeah, that’s right. An archaeologist who does maths. It’s like you’ve found the blog of a unicorn, right?


Wrong. We almost all use maths almost all of the time. WE DO SCIENCE, DEAL WITH IT! However, I must say, since I moved to Britain a few years ago and started to use maths instead of just math… my results have improved loads. Hands down, I’m a convert.

Now the reason that I can use maths to do this is that humans are really predictable. They all die. That’s right ladies and gentlemen, we’re all going to die. I’m fairly certain you all knew that, but if you didn’t, I apologise for ruining Christmas.

This predictability in death at a population level creates something called a ‘mortality profile’ that specific to different scenarios. All you really need to know is that an everyday (dead) medieval population looks like this:


This is called ‘attitrional’ and is caused by a high number of infant deaths, followed by a decrease, and then a general increase throughout adulthood until it peaks and then tails away, because nobody lives forever.

Conversely, episodes of mass fatality (like the plague) look like this:


This is called ‘catastrophic’ and is caused by the fact that the plague (as far as we currently know) doesn’t pick on little babies or old people as it’s a non-selective disease, and so kills regardless of your age or sex.

I am trying to find these catastrophic signatures in rural (or rather, non-urban) medieval cemeteries. But it’s complicated because cemeteries are usually a mishmash of over three hundred years worth of bodies – and sometimes we don’t even have the entire cemetery, if the building developers aren’t particularly worried about ancient curses – and for so many other not-boring reasons (but this blogpost is already mahoosive, so let’s move on).

In order to try to solve this problem, I create a model medieval village population with lots of numbers and then I create a cemetery based on this population, by killing them.


But I do it in a science-y way, with different mortality rates depending on the historical circumstances. I take all the information I have about the really real cemetery I am interested in, how long was it used for, who it was used for (church, hospital), whether this population was subjected to famine, disease, wraths of God, etc over its period of use – and then I apply different mortality rates, including different risks for males and females of different ages, and in five-year intervals I kill people off. I won’t lie. It’s kind of fun – but in like, a really shit version of the Sims, way.


But, that’s not all. I also make sure that while I’m busy killing them off, my model population is busy… getting busy. When I first ran my model simulations, I had to up the birth rate significantly, because I killed my first village off in less than fifty years. It turns out they were really taking this saying to heart.


And as time passes, people from these new generations I create will die too. Some will die shortly after birth, but some will live well into old age. It’s the circle of life.


What I end up with is a model cemetery profile of what I should expect to see if all of these historical variables (famine, plague, etc) are true. I then compare this to what I actually have from a really real cemetery with skeletons that have been aged and sexed – and I run tests to see if they are similar. And, yeeeah. That’s actually my PhD. I said I did maths… I didn’t say it was any more complicated than what you learned in Year 3.


Seriously, it’s just graphs with lines instead of bars. I even colour-code all the plague deaths in green. Because who doesn’t love a bit of colouring in?

It is important though (and I may be downplaying the level of maths I use a little bit). This type of modelling will hopefully allow us to investigate specific catastrophic incidences that until now have remained unstudied. In the case of this particular research project, this will give us a better understanding of the Black Death across the entire country, instead of just places where it’s really obvious, like say, a giant mass grave in the centre of London.

Why should you care? Because understanding diseases from our past – especially diseases that caused massive epidemics – is really important to understanding diseases that currently threaten the human domination of the globe (up to and including Yersnia pestis).

Now, when I presented these sets and began writing up this blogpost, I had ended it with a joke about Plague Inc, and Madagascar – but given the current situation of the on-going outbreak, I don’t really think it’s appropriate so [redacted].

In conclusion, if you’re going to play hide and seek with the dead, if you want to win, use maths.

*I previously had a nutshell on either side of that GIF, making it even funnier – but WordPress doesn’t like it anymore. 😦

6 responses to “GIFted Researcher

  1. Alison, this is splendid. Very interesting and enjoyable read. I’m struck by your insight about the normalicy of rural plague death. My own image was, like you said, massive plague pits. I’d never thought about its impact in the countryside. It seems very melancholy, yet also very touchingly human, to imagine lots of death that, in the big picture, was part of a huge, continent-wide event, but was simply another part of daily struggle for many or most.
    I’m also thinking about current vectors of death (HIV, gun violence, etc.) that are widespread, but will be hard to see in the future because the victims are so dispersed.
    Very interesting piece, those were some lucky attendees at the Science Showoff!

    • Thank you John. I am glad you enjoyed this post. It certainly does make you think – before I began this PhD I’m not sure I really understoond the scale of catastrophic incidences in the past. The point you picked up on current vectors is a great one. Although I am currently researching a medieval epidemic, it is my hope that these types of modelling methods will be applied by archaeologists (and other researchers) to many different scenarios – especially as I came to this project by way of researching the application of archaeology to forensic investigations. It is certainly a promising area of research where other catastrophic incidences are concerned (e.g. natural disasters, warfare/conflict, disease morbidity, etc) and I expect beyond that as well.

  2. This was a fascinating read! Sharon Dewitte of USC gave a presentation on work with similar (albeit urban) populations at the AAPAs last year, which was the first I’d heard of the bioarch approach to the plague. Seems like a period that’s ripe for investigation…

    Out of curiosity, how do you control for differential representation of cemetery populations? You mentioned in text that not all cemeteries are fully excavated, which seems like it would lead to biased samples in the ‘real world’ data you’re comparing to your model(s). Do you resample or something to that effect?

    • Thanks for the comment, I’m glad you enjoyed the post. I’m familiar with DeWitte’s work and it is indeed very interesting. I’m hopeful more people will see the potential of palaeodemography!

      You are spot on about the models. One of the biggest problems is that they create ‘ideal’ cemetery populations, but thankfully since data is so easily manipulated it allows you to make changes without too much difficulty. In the simulations I run, in addition to presenting total numbers in mortality profiles, I also use random number generation to achieve samples that would reflect a partially excavated cemetery (these can also have variables applied if there are knows about the areas of the cemetery that form the sample).

  3. ‘I am trying to find these catastrophic signatures in rural (or rather, non-urban) medieval cemeteries’ is most of your evidence ‘archaeological’ or do you have good documentary evidence (e.g. Parish registers etc). In the geoscience world researchers are looking at major volcanic eruptions e.g. Laki killed over 6 million souls; on the east coast of England, death rate was 2/3 times ‘normal’ because of it. There was also the medieval Little Ice Age – this probably overlaps your period – was the mortality signature different to the plague?

  4. Thanks for your question. The models that I am building up for comparison are based on both historical documents and recent results regarding population response to crisis (e.g. fertility following epidemics) – these are then compared against the archaeological data. So in short, both. I am not specifially creating matching cemetery profiles based on parish records (although I would love to do this in the future – the reason I am not right now simply being time). This has been done in the past though, so existing ones will be use for comparison (although it is inherently problematic, as most of these are from later outbreaks – and therefore a population that has already been impacted by the earlier epidemics). It is certainly a possibility for investigation in more detail into later instances though (e.g. 17th century) – as has been done in France recently.

    The different mortality signatures is something that I’m still not sure on, however early results suggest that they are different (as the risk categories in these incidents are not the same). The fact that while people were dying of the plague other people were still dying of other causes (and some of these which can also be classed as ‘catastrophic’) is certainly a complication!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s