Was there ‘scientific method’ in the early Middle Ages?
This was a question posed on Twitter yesterday. It is a good question. I am also going to give a paper that addresses this at the University of Kent next week so it is hard for me to compress quite what I want to say into a tweet. But let’s see if I can sketch some parameters for thinking about the idea.
First of all: what is ‘scientific method’? Each field and discipline has its own norms and rigour and processes. But mostly it works something like this:
(1) have a problem
(2) think of a hypothesis that addresses the problem
(3) design an experiment that will help you to assess the hypothesis (hopefully in such a way that it might generate bonus data or ideas)
(4) carry out experiment and write down data
(5) work out how data and hypothesis relate to each other
(6) try again as appropriate
And, let’s not forget, there is an onus on making the data and ideas available in such a way that both can be interrogated later on.
Now, we know well that most people think that there was obviously no science before the Enlightenment because everyone had Religion and Religion demands Faith not Reason. (There will occasionally be sarcastic capitalisation ahead – sorry). The Greeks and Romans might have had it, but they were a long time ago. There has been a modest amount of revision, as there often is. Notable contributors here include David Lindberg and Edward Grant, who in their different ways argued that there was enough attention to inductive method and mathematics, especially after the Twelfth Century Renaissance when Christians discovered Greek knowledge via the Muslims and Discovered Reason. But it’s not the same as calling it ‘science’, as (a) people at the time thought of what they were doing as part of a larger programme of ‘philosophy’ aimed at understanding God and (b) lots of it wasn’t very good and didn’t always involve very much by way of method in the modern sense. Or even reality.
Is it possible to characterise what people were doing, then, rather than ‘what they were not doing relative to modern scientific method’? I think so (obviously – it’s my job at present) and I think it is instructive.
Let’s start with my stock example for these hastily written essays: Bede! Many Bede scholars like to think that Bede had a ‘science’ (or proto-science) because he wrote three books on nature, mathematics, and calendars (On the Nature of Things, On Times, and On the Reckoning of Time – all available in translation with commentary by Faith Wallis and Calvin Kendall). Much of what Bede wrote has not always satisfied hardcore historians of science. This is because most of it is derivative – mostly what he does in those books is create tapestries from other authorities such as Isidore of Seville. Also, he doesn’t obviously do a lot of hypothesising or writing up of experiments.
But we can spin Bede’s work in other ways. He faced two principal problems with the natural world: how to distinguish the natural order of things from the miraculous, and how to use mathematics and astronomy to establish a reliable calendar to calculate Easter dates for forever. Bede had a wide-ranging literature on these subjects with many contradictory ‘authorities’ he could use, some of them very old indeed. So he spent a lot of time considering the logic of their statements, crunching numbers to see how things worked, and even a little bit of time outside checking that things worked. He did not write up his experiments and clearly fudged some of the data. But he did write up his conclusions in books that sought to justify his ideas and to refine the existing paradigm wherever possible. We might therefore offer this positive ‘early medieval method’:
(1) have a problem
(2) find out what people said about the problem before and assess their authority and accuracy
(3) do a bit of observation and number crunching as appropriate just to make sure
(4) write a treatise about how to address the problems. (Include arguments).
It’s a bit sloppy by modern standards and not very data-heavy. But as a process in which the logic of model is tested and refined, it worked for them. And it addressed the problems they needed addressing.
Looking at one author can be misleading: this was a Europe-wide way of doing things. Similar efforts at processing authorities and data can be found from Ireland and across the Frankish kingdoms in the eighth century. Given the limited set of problems they were interested in, it is also striking that there were efforts to improve data (i.e. there are lots of records of eclipses in annals etc) and refine ways of dealing with it. Then, importantly, students were taught how the mathematical systems worked and they were trained how to investigate nature. There was little sense that they just had to accept the answers of previous generations – especially where the answers seemed wrong.
And so people disagreed about interpretations and argued about how best to interpret unusual phenomena. In the letters of Alcuin, it is clear that Charlemagne would write to him with astronomy problems, and that Alcuin did not approve of the models of explanation being used in his absence from court, so he even sent Charlemagne back a series of questions to ask so-called experts to catch them out. A generation later, in 837, Halley’s Comet was visible in the night sky and there are sources that show people monitoring its process and struggling to interpret why it seemed to disappear and then reappear that April. (Scott Ashley’s recent solution, by comparing similar records in the Chinese Xin Tangshu 新唐書, is that the comet just happened to be followed by an observable nova). What people didn’t do was go ‘oh, it’s probably just God’ – they read, debated, and reasoned. And then, in Louis the Pious’s case, went off and prayed just in case.
Now, this is a very Western-centric response to the question, because I am writing quickly in my lunchbreak and not doing extra research. But let’s not forget that other societies were engaged in these pursuits too. If you want a good non-European counterpart to Bede, go and read up on the Buddhist monk and scholar Yi Xing 一行 – a scholar who was involved in a massive enterprise to improve the accuracy of calendrical methods relative to longitude using gnomons across the Tang Empire. (You will find some useful starting points here by Jeffrey Kotyk). Or look at the wealth of astronomical records for his time in the Jiu Tangshu 舊唐書 and Xin Tangshu. You have lots of the same issues of applying Science to illuminate and negotiate Faith, but on a much bigger scale and with some greater precision.
All of this is to say that, if you go looking for examples of the scientific method in the Middle Ages, you probably won’t find them in a way that will satisfy purists today. Or you at least have to be prepared to defend that it just isn’t quite the same. But that it isn’t quite the same is reason to take it seriously rather than to dismiss it. Why? Because it shows us that science isn’t just one thing. It changes in response to the problems we set, our politics, our beliefs, and technologies. We like to believe that science and its data is neutral because it is independent of us, but as soon as we try to do anything with it, it becomes part of the subjective human world.
 Eg Lindberg, The Beginnings of Western Science (1992) or Grant, The Foundations of Modern Science (1996).
 In this context do read Faith Wallis’s essay ‘Reframing Bede’s Science’ in DeGregorio (ed.), Innovation and Tradition (2006) because, actually, you just need to. But what I say below doesn’t really derive from that.
 Selections from these letters can be found translated by Allott in Alcuin of York (1974).
 Scott Ashley, ‘What did Louis the Pious see in the night sky? A new interpretation of the Astronomer’s account of Halley’s Comet, 837’, Early Medieval Europe 21.1 (2013), 27-49.