How to print your own 95 Luther-theses

The 500th anniversary of the Reformation is fast approaching – time to get in the spirit and start printing your own 95 Luther theses!

You need:

I) A print shop:

Sign at the entrance to the Bodleian Bibliographical Press, Photo by Charlotte Hartmann


II) A template:


This slideshow requires JavaScript.

 Basel, Adam Petri. Digitised in the VD16 (link:

III) A typecase:

Photo taken in the Bodleian Bibliographical Press by Charlotte Hartmann


IV) A composing stick:

Photo taken in the Bodleian Bibliographical Press by Charlotte Hartmann



V) A skilled printer, who can help you with whatever challenges and obstacles this project will throw into your way:

Bodleian Bibliographical Press printer Richard Lawrence. Photo taken by Charlotte Hartmann


VI) A printing press:

The largest press in the Bodleian office. Photo taken in the Bodleian Bibliographical Press by Charlotte Hartmann
Once you have these essentials, it is time to get started!


As an inspiration, I’ll tell you my story of printing Martin Luther’s Ninety-Five theses. It all started towards the end of Michaelmas Term 2015 (meaning December)… Without any previous experience or any particular knowledge of printing, I more or less naively decided to have a try. Little did I know of the challenges and efforts that were awaiting me – but every single one was worth it. I got so much out of this project, not only gaining some pretty awesome new skills, but above all meeting some great new people and getting absorbed in the creation of such a meaningful piece of European history.

I did not typeset from the very beginning and only made this my own project by the time that about ten theses had already been set by various students of the History of the Book and Palaeography. My supervisor Prof Henrike Lähnemann had encouraged them to pick their favourite thesis for typesetting. Still, so far none of them felt particularly inclined to take responsibility for the remaining theses, so I stepped in. I had recently started helping out with an Occitan-print in the Bodleian Press and was eager to get my own project.

These first ten theses formed the print’s basis and I decided to maintain their layout so that I wouldn’t have to readjust them again. Since they were not in any particular order, but arbitrarily picked from all over the text, they had been numbered with their exact Roman numbers so as not to loose orientation. I also maintained a consecutive numeration throughout the text, although the template features an interval system (3x 25 theses and 1x 20).

When the Bodleian’s printer Richard Lawrence and I started planning the layout in more detail, we prioritised a clear arrangement of the text and therefore decided to ensure a distinct gap between numbers and text. We therefore measured the smallest acceptable space between thesis lxxxviij and the beginning of its text and then adopted this measure as standard for all other numbers (see below image). For this reason the overall layout of our print has a much more spacious appearance in comparison to other broadsheet prints of the theses.

Detail of set type featuring theses lxxxviij/88. Photo taken in the Bodleian Bibliographical Press by Charlotte Hartmann

Our line length is 22 points – this was as arbitrary or rather pragmatic a decision as agreeing upon a Roman typeface: The Bodleian office’s largest stock of leads (which are used in between the lines) is that of the width of 22 points and Roman type is a standard type today (there are six typecases of Roman type in the Bodleian). We knew from the beginning that this project would require both a large amount of leads and of type. Therefore, we simply went with the leads and type of which the greatest ressources were available.

As said before, the first theses were all set by different people – later I was mostly the only typesetter, although several people would occasionally contribute a thesis or two (only in the following term I got the invaluable permanent help of an Undergraduate, Walker Thompson). At first I only saw the disadvantages of this situation as I was making slow progress. Getting used to setting lines upside-down and mirror-inverted took some time getting used to, but with time I learned to know where to find each letter in his corresponding compartment without looking. I became much faster – at the beginning I simply set the lines and then added the space, later I was able to roughly gauge how many words would fit a line and how many lines a theses would need. So in the end I needed maybe 7-8 minutes for a thesis, in contrast to at least twice or thrice the time in the beginning.

Also, I came to realise that fewer typesetters have an even more significant advantage: fewer mistakes! Although help with typesetting was always very welcome, I noticed that this had a direct effect on the number and types of errors. The most common errors were a false dissolving of abbreviations and the omitting of ligatures (ſi, ſſ, œ, ӕ, fi, fl, &, the c-t ligature etc.). The tricky pointer with the ligatures is that they mostly do not have the exact same width as their single letters combined. This means that when this error is corrected, the entire line nearly always has to be readjusted with different spacing. Another point was the confusion of letters b and d. This mix-up happened either because they were incorrectly disseminated in the compartments, or because the typesetter(s) set them the wrong way round in the composing stick. The latter can happen very easily, because even though the lines and letters are set like one reads them from left to right, I already mentioned that they have to be set upside down and mirror-inverted, which means that the difference between d and b and even p and q can be quickly overlooked.

As template we chose a 1517 Basel print by Adam Petri – mostly because of its striking layout, the comparatively spacious arrangement with the indents and the initial at the beginning of the first thesis. Still, we made some fundamental changes from the template in that way that we resolved all abbreviations and special characters (except ligatures and the division between long s and short s). Abbreviations were certainly commonly used and therefore a customary part of both writing and printing practice (the Basel print features several hundred abbreviations and special characters), but we agreed that in order to guarantee a clearer reading experience and also to improve the print’s overall appearance, we would resolve them nonetheless.

As mentioned, the Basel print had a very recognisable layout with a distinct indent from the second text line onwards, which further supports the text’s clarity. The above mentioned resolution of all abbreviations was one of the main reasons why our 2016 version appears to be significantly longer than its template.

A major difference between our print and its template was the amount of time for the completion that was at hand. Whereas it can be assumed that the Basel print was most likely produced within a few days, ours was developed over months. This allowed us to apply the utmost care especially during the typesetting and correction process. The Oxford print went through several stages of correction by different people, both before and after printing, with several versions that were reviewed by hand on the paper itself. The following images give a few examples of different types of errors, which can occur both during the course of typesetting and the printing:

This slideshow requires JavaScript.

Examples of some of the correction prints with different types of errors. Photos taken by Charlotte Hartmann

We can see amongst other things that an entire line from thesis lxix appeared at the end of its succeeding thesis lxx. The most prominent mistake in the other image was the confusion of phrases of theses xcij and xciij. Instead of pax pax, & non est pax the typesetter set the phrase Crux crux, et non est crux, which should only appear in the following thesis. Also, this thesis has a wider indent than the remaining text. These are inadvertent mistakes, but nonetheless, they cost time and must be elaborately corrected. You can look for yourself, how many mistakes you can find, which we overlooked (I can assure you, there will be some!).

After typesetting and correction – finally printing!

We printed with two different kinds of paper – a pretty German type, which was very clean looking and also showed a watermark. And then the paper, which we came to call only The Wormpaper. The name speaks for itself… we found it in a drawer somewhere in the print-shop, Richard couldn’t remember its origin, so it had probably already been lying there for quite some time. We were directly taken by its distinct colour, rough texture and overall appearance. It looked worn and washed-out: in other words perfect for our purpose. The wormholes make it even more interesting. Still, the wormpaper had to be dampened before printing, otherwise the press couldn’t apply enough pressure for the imprint and the letters weren’t as clearly legible and sharp as we had wished.

Also, we actually printed the theses in two separate runs – first columns one and two, then 3 and 4. There was no frame in the office that was big enough two contain all four columns simultaneously. This of course presented some new challenges, since we had about 250 halfway printed copies and then needed to print the remaining half in the exactly appropriate space without having an overlap in the middle. What happened quite often was that the paper was not perfectly straight adjusted in the press and therefore the gap in between columns 2 and 3 got narrower towards one or the other end. Due to theses errors, about 50 of the 250 prints we got in the end are slightly less perfect than the others.

At last, we held the first print in our hands:

Charlotte Hartmann, Henrike Lähnemann and Walker Thompson – proudly present the first print of Luther’s 95 theses in Oxford. Photo by Richard Lawrence

Here are the final prints (wormpaper and German paper):

The “wormpaper”. If you look closely, you’ll see the typical mistake I described above: The gap between columns 2 and 3 gets narrower towards the bootom of the page. This is due to the two separate stages of the print – the paper was not correctly inserted in the press. Lawrence/Hartmann, Bodleian Bibliographical Press, Oxford 2016
The German paper. Compared to the slightly brownish wormpaper, it has a very clean appearance. Lawrence/Hartmann, Bodleian Bibliographical Press, Oxford 2016


2017 – Here we come!
Detail featuring the dates – although the Oxford print is from 2016, we kept the 500th anniversary of the Reformation in mind and set it as if it were 2017… (And also we actually did not expect to finish so soon..). Photo taken by Charlotte Hartmann featuring Lawrence/Hartmann, Bodleian Bibliographical Press, Oxford 2016.


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 )

Google photo

You are commenting using your Google 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