As those who read my blog regularly will know, my research focuses on the effects of ageing and neurological disorders on medieval handwriting. In a recent publication, Jane Alty and I examine the shaky handwriting of the thirteenth-century scribe known as ‘The Tremulous Hand of Worcester’. In the article, we trace the contours of the Tremulous Hand’s writing visually, to conclude that he most likely had a condition called essential tremor. To strengthen our conclusions, we compare the script with examples of modern-day people with the same condition.
In the article, we acknowledge that we cannot know the quill’s speed when the Tremulous Hand was writing. Knowing the speed of writing can help to classify distortions in writing. A tremor is ‘a rhythmic oscillation of a body part’, and it can be classified using several characteristics including the oscillation’s amplitude and frequency. A tremor’s amplitude can be ascertained from static handwriting samples: it is the extent of the movement to the left and right away from the letter’s baseline. In contrast, the frequency is the number of these movements away from the baseline in a certain amount of time (usually seconds). As Jane Alty and Peter Kempster have pointed out, it can be just as useful for diagnostic purposes to use plain language to describe the amplitude of a tremor (‘fine’ or ‘coarse’) and frequency (‘fast’ or ‘slow’) (p. 624). However, to know the tremor frequency in hertz, we would need to know the speed of writing. This can be ascertained for modern-day people with tremors, using digitising tablets that capture the person’s hand movements through time and space. However, it is impossible to recover precise temporal information from static samples of medieval handwriting alone.
So, how would we go about recovering some information—knowing at least something about the speed at which scribes wrote? There has been much head-scratching on this subject amongst medievalist scholars. Interest gained momentum and some precision in the nineties with the works of Michael Gullick, J. P. Gumbert, and E. A. Overgaauw (all published in 1995!), and has recently been revived by Daniel Wakelin. Book historians have been lured by the idea of creating not just an image, but a film, of the scribe at work. Overgaauw offered us a protagonist for this movie: the sixteenth-century German scribe Hans Ried, who wrote a huge compilation of romances known as the Ambraser Heldenbuch . Overgaauw explained that the book comprises 239 leaves and took Reid up to 12 years to finish. Some believe this scribe was a ‘raffinierter Faulpelz’, or ‘clever sluggard’, alleging that he took as long as possible in order to continue to receive payment. If this timescale is correct, Ried would have copied an average of less than half a column per day – or just over a column a day if he was not always actively working on the book. But was he really a sluggard?
As Overgaauw’s article progresses, he calculates writing speed using contracts in which scribes promised their patrons a number of folios per day (p. 215). Insight can also be gained by scribes’ colophons* in books, where they add a date at the beginning and/or end of each text in a compilation (p. 212-214). However, Overgaauw and others working on scribal speed remind us that just because we know that a scribe was working on a piece for three years, we cannot assume that he was always working on it. We also don’t know that he was only working on that piece. Daniel Wakelin has pointed out that changes in ink colour throughout texts written by some scribes witness short bursts of writing, rather than extended periods at their desk. They undertook, ‘short stints, frequently and irregularly interrupted’ . For comparison, if I said that it took me two days to write this blog post, you would not assume that I was constantly writing the post and not spending considerable time reading, writing other things, or making tea. If you did, you would have certainly over-estimated my commitment to my social media presence. J. P. Gumbert showed that patrons made some effort to keep scribes from procrastinating, though. For instance, one chaplain of Bruges promised to write and gloss a Life of St. Donatian. He was instructed to do two folios per week and show it to his overseers at the end of each week. His rate would work out at 0.28 folios per day – but he never finished it. (pp. 67-68)
Even if we can say with relative certainty that a scribe wrote eight folios in four days, it is difficult to be any more precise from the evidence preserved in contracts and colophons. We cannot know for sure if the scribe was working on that manuscript for every one of those four days. If we were to try to calculate the number of lines per hour, or even the number of letters or strokes per minute, we’d have to account for the scribe reloading his quill with ink, and—as Michael Gullick has pointed out—for the fact that writing speed would vary according to the time of year (i.e. in bitter winters, cold fingers could slow or stop production). Scribes also had to pause to change or re-cut their quills (Gullick, p. 52). This is not to mention time taken to rest the eyes, have a break, go to the toilet, take a meal or a drink, or just the slowness caused by fatigue. So, though we might gain an impression of the number of pages a scribe wrote per week, or even per day, it is hard to be any more precise. As Gullick puts it: ‘there are too many imponderables—the more minute the calculations, the greater the likelihood of error’ (p. 49).
Gullick, a calligrapher-scholar, began to investigate the question of scribal speed from a more practical—and, I think, promising—angle. He asked the calligrapher Donald Jackson for some advice, and Jackson worked with a quill to try to estimate the speed of a twelfth-century scribe. However, Jackson stopped short of describing the rate of his work more precisely than, ‘twenty-five lines an hour’ (p. 50). Certainly it’s also true that calligraphers, no matter how expert, are not equivalent to medieval scribes in terms of their training and the way in which writing fits into their daily lives. So, do we give up? This is a question that I am currently investigating. I suspect that, in fact, that it is at the ‘how many strokes per second’ level that we might make the most accurate calculations. The ‘actual speed of forming letters’ should be easier to determine than ‘the speed of producing a book’ . Because, at that level, we do not have to account for scribes visiting the toilet.
*A colophon is a brief note added by the scribe, usually at the beginning or end of the text that they were copying.
 Alty, Jane E., and Peter A. Kempster. ‘A practical guide to the differential diagnosis of tremor.’ Postgraduate medical journal (2011).
 Overgaauw, Eef A. ‘Fast or slow, professional or monastic. The writing speed of some late-medieval scribes.’ Scriptorium 49.2 (1995): pp. 211-227.
 Wakelin, Daniel. ‘Writing the words.’ The Production of Books in England 1350–1500 (2011).
 J Gumbert, Jan Peter. The Speed of Scribes. Scribi e colofoni (Spoleti: 1995) pp. 67-68.
 Gullick, Michael. ‘How fast did scribes write? Evidence from Romanesque manuscripts.’ Making the medieval book: Techniques of production (1995): pp. 39-58 (p. 43)
 J. P. Gumbert emphasises the importance of making the distinction between these two elements of ‘speed of writing’ (p. 58)