Paleography and Mystery Words

If you’ve ever had the pleasure of skulking in archival holes and corners, you’ve no doubt encountered the joy and despair associated with paleography. When I was a kid I wanted to be a paleontologist; little did I know that my future would take me in the slightly-different direction of paleography. Perhaps not as exciting as chasing after a T. Rex (but almost as dusty), being a paleographer means excavating knowledge from handwritten documents slightly younger than fossils. An American history grad student’s first introduction to paleography can be fear-inducing: about to set off on his great European archival adventure, a new professor suddenly introduces him to the concept of difficult handwriting in a class he was auditing. The level of panic only increased when a fellow grad student already across the Pond confirmed the challenge, warning that he had spent several *weeks* in the archives deciphering the handwriting, literally l-e-t-t-e-r by l-e-t-t-e-r, for day after day.

Over the years I’ve researched in numerous archives and therefore gotten pretty good at deciphering the paleography written by French and English hands c. 1700 – enough to impress most HIS 200 undergraduates at least. But paleography varies greatly not only by person, but by place and time as well, so I’ll post what little I’ve learned over the years – undoubtedly serious guides to paleography will discuss these more authoritatively than I can. What follows is a combination of observations and tips:

  • Writer Identification: First off, handwriting is very useful not only because you can’t understand what’s being written without it, but also because it is usually easy to identify the author (or at least the writer) just based off of their handwriting. Much handwriting is distinctive: no doubt you, like me, can tell whether your mother or father sent you a letter/package just based off of the scrawl on the envelop address. This distinctiveness can also serve as a time-saving technique in the archives: at the British Library recently I was looking at incoming intelligence to an English Secretary of State – news accounts coming from a dozen different people in half as many countries. I was only interested in events occurring in the Flanders theater (time constraints), and in order to speed up the process I could identify the different authors’ handwriting at a glance, which made it easy to skip all the letters written by Mr. X from Naples and Mr. Y in Moscow. It’s a good idea to randomly skim through the other letters just to assure yourself that you’re not missing anything, but the ability to triage your letter-reading makes a big difference. Trying to decipher handwriting is mentally taxing over long spans of time, and it’s surprisingly difficult to skim for comprehension when you have to work to identify every third or fourth word. [I have vague memories of reading somewhere that scientists studying reading eye movements discovered that reading often occurs in chunks of words.] Identifying specific hands also helps if you’re looking at a volume where you know some authors never have anything interesting to say, whereas others are just goldmines of analysis and insight. Unfortunately for me, those few insightful individuals were also the ones with the tiny, cramped handwriting. (See more below.)

    Eugene of Savoy’s distinctive 3rd-grade handwriting

  • Type of Document: People involved with diplomatics (diplomaticists?) and careful historians will distinguish between documents that were in the author’s hand (holograph) or their secretary’s or second-hand copies made by someone else and enclosed in another letter, and between originals and copies and extracts. This (and any changes in handwriting over time) might tell you things like: how formal the document’s intended audience was, how hands-on or busy a particular author was, who else knew about the contents of the letter, the health of the writer/author at a particular period in time… With some of the worst cases of handwriting, I’ve often wondered how the recipient could even read it – I’d think you would want your communications to be legible, especially those sent to your boss.
  • Handwriting Psychology: If one wants to get psychological, they could say something like “in this portion of the siege journal the garrison commander’s handwriting gets really shaky and hurried, which just happens to coincide with a particularly heavy enemy bombardment.” But I’d avoid psycho-analyzing someone based off of their handwriting. The humorous incident with the BBC hiring a handwriting analyst to assess Tony Blair’s leadership abilities (Doodle Gate) comes to mind.

    Wm Cadogan’s signature

    Geldermalsen’s signature

  • Period-Specific Handwriting: It seems like the earlier in time you go, the more difficult the writing is. This isn’t necessarily linear and the breaks can be quite abrupt. For example, while in the Douai municipal archives I found it quite easy to read the French handwriting from the 1700s, but once I went back to the 1660s, it was impossible without a serious amount of effort. Later on in England, I was baffled by the 16th century handwriting that a colleague was reading – it was supposedly in English, but it looked Greek to me!
  • Country-Specific Handwriting: The period specificity is further complicated by the variations between different countries. In my c. 1700 experience, official French handwriting is a breeze to read – nice, evenly-spaced, slightly loopy but not too loopy, almost every letter distinct with few ascenders/descenders messing up the lines above and below. As is often the case, any marginal comments by the recipient will usually be much more challenging to decode, due to space constraints and the fact that marginalia are usually reserved for the writer alone.

    French letter and marginalia

    Things get very different if you go north of the border. The nice French secretarial hand disappears and you find the most disturbing Teutonic handwriting you can imagine, as if Satan himself was taking a dictation! To be fair, the ability to read difficult handwriting is related to the reader’s fluency with that language, but still. To give a sample (and there’s much worse, trust me):

    Moderately-bad Dutch handwriting (1710)

    A slightly-more-legible sample, very linear.

    Better Dutch handwriting

    But then, every once in a blue moon, you come across a Rosetta stone, a document that will help you unlock the key. Or you could just go study an early modern Dutch paleography guide, but my local library seems to have their copy checked out at the moment. The following is, for a historian trained in early modern France trying to give the Dutch their due, a fascinating find. I’ll let you look at it first, then divulge its secret.

    My Rosetta Stone

    For the uninitiated, what is so interesting, and bizarre, about this document is that the script is clearly in the Dutch style, yet the language it’s written in is French! The first line reads “Nous avons l’honneur de repondre a la letter.” I would even speculate that some individuals actually had totally different handwriting, depending on which language they were writing in. Now that’s weird.

  • Formulaic Language is Your Friend: Often times you can skip the beginnings and endings of documents because they don’t encode much useful information (depending on your purpose). The most basic examples include formulaic greetings and sign-offs. Another universal at the beginning of just about every early modern letter is a sentence or phrase in which the author informs the recipient as to which letters of theirs they have received and are responding to. Not surprising when you realize how haphazard early modern mail services were – we’re so spoiled with our ability to cc: or hit the Reply button. Similarly, the ends of letters will frequently segue rather abruptly into their signature, e.g. “resting safe in the knowledge that I am…..[signature].” More useful are the various phrases and particular word combinations that you will see used over and over again, especially long titles such as “Les Estats Generaux des Provinces Unies des Pays Bas.” You’ll see them often enough that you can identify them immediately (and abbreviate them if you’re transcribing) and skip on to the more important parts.
  • Compare and Contrast:  A common technique recommended by paleography guides is to literally go one letter at a time. Figure out all the letters you can puzzle out, and then look at the mystery letter as a shape, and then look for a similar shape in another word somewhere else on the page. Similarly, you can try to confirm a hunch by looking for that letter in another word. But be careful of traps: many letters can be easily confused with several others (especially vowels as well as common consonants like l, k, b, h, f), and many words can be misread if you mistake one vowel for another and then use that to decide whether the letter that follows is an l or a k. [See below for an example.]

There are various features that are particular challenges:

  • Certain letter combinations may be written differently than each of the letters in other contexts.
  • Certain letters might be written differently at the beginning of a word vs. elsewhere in a word.
  • Sometimes people just write their letters slightly differently. Sometimes their handwriting looks different in different documents, e.g. a letter vs. a letterbook where they make official copies of their correspondence. Go figure.
  • Abbreviations are rife, and sometimes period-specific. We’re all familiar with the ‘ye’ abbreviation, which stands for ‘the’, just as ‘ym’ stands for ‘them’, ‘yt’ is ‘that,’ ‘yr’ is ‘your,’ and so forth. I’ve already mentioned the early modern conceit of abbreviating months with Roman numerals according to their Latin root (e.g. 8re means OCT-ober, the 10th month). Then there’s the common use of superscripts: the aforementioned -m, but it could also mean thousands if combined with a number (12m usually means 12,000); a tilde-looking superscript character (~) above a consonant in the middle of a word usually means to double that letter (comon becomes common), and so on.
  • Other vestiges of the writing and conservation process might similarly interfere with your note-taking. Most frequent are damaged copies (ink bleed-through very common, tears, cross-outs and corrections, mis-cropped pages, water damage), poor copies (usually from microfilmed or photocopied versions), and the dreaded gutter margin that is too tight so that one or more words are illegible. For some reason archivists aren’t too keen on you pulling out your pocketknife or cracking the spine of the bound volume.
  • Spelling (orthography) was widely variable in the early modern period, which means that oftentimes we’re at a disadvantage when trying to guess words if we’re looking for modern standardized spelling. Sometimes you just need to know the interesting ways in which they spelled words in your period, a tame French example being doibt = doit. Other times it helps to sound out the words phonetically. I’ve noticed, however, that you really need to be careful to *account for every character.* On occasion I’ll guess at a word – I know five of the seven characters for sure and know just the word that would fit the context – only to discover two words later that it was actually a slightly different word, but one with the same letters, or one letter that is shaped very similar to the letter I thought it was. Sometimes it will take the second encounter, with a slightly different context or slightly more legible handwriting, or maybe it’s just a fresh look, to suddenly have that Eureka! moment of understanding. Dealing with mystery words is particularly tricky if the writer isn’t consistent about the spacing between words – often times they’ll separate the the article from the noun (I’m thinking of French contractions particularly), and sometimes even in English they’ll make two separate words into a single compound word, or the reverse (note the Gelder Malsen signature above). It may not sound confusing in the abstract, but if you string a line together where: 1) you can’t tell where three of the ten words end and another begins, and 2) you can’t read one out of every five or six words because each could be four different words, it becomes a real challenge!

To give a concrete example of this last point about how permutations can pile up, I present the motivation for writing this post. I had long ago intended to write a post on paleography, but my doing it now was spurred on by a particular incident I had in the British Library last month. Here, then, is your homework: What the hell is this word?!?!

The entire sentence under question reads: “They would rather weary France than [Irsitube?tade?tate?buke?] him and if this war was soon ended, they would be too ready to split asunder within.” I’ve also included transcriptions of the other letters in the picture so you can get a sense of his handwriting style.

I probably spent fifteen minutes across two days, and interrupted two other archive denizens, trying to decipher it. Modern dictionary software makes it easy to use wildcards in order to figure out possible words, but since there are only 26 letters in the Latin (English) alphabet and even fewer combinations, while many have similar shapes (having the same descenders/ascenders), there are many permutations even if you narrow it down: ?ell? might be ‘hello’ or ‘bells’… Even worse, it might actually be ‘hilly,’ and you misread the ‘i’ as an ‘e’ (that happens more than you might think). And if they use slightly archaic phrasing or vocabulary or spelling, it’s even more difficult to guess the word based off of what sounds ‘natural’ to us today. With this particular word I just can’t get it to work, so hopefully someone here will see what it is.

The uncertainty with this word is particularly high because I can’t tell exactly how many letters there are (the beginning of the word especially), while the end of the word (or is it a separate word?) has a very common {consonant}-{vowel}-{consonant}-e ending, and there are several letters that look like those consonants – I can’t even tell if that first consonant is the same as the last consonant! Or maybe it’s actually two words, like ‘to be’! Rather than bore you with all the possibilities I’ve tried, perhaps a fresh look is in order.

This is a particularly vexing example because, for the first time in my experience, the one word I can’t read is actually the most important word in the sentence! It’s also the most important sentence in the entire letter for my purpose! And, to top it all off, it couldn’t be more relevant to the whole point of my book on Marlborough and battle. So your help would be greatly appreciated. Crowd-sourcing paleographical transcription.

This leads to the final lesson on paleography:

  • Sometimes You Just Can’t Win: In every period (heck, even today, as the stereotype of doctors’ handwriting suggests), there will always be a few people whose handwriting is atrociously bad, or a word or two that you just can’t figure out. These tend to be worse in the squeezed-in marginal comments someone might make, but sometimes entire documents just have to be abandoned because of their illegibility. Please don’t make the above word one of these!

Thoughts or suggestions?


National Archive’s brief guide:

English Handwriting 1500-1700: Excellent site at

Beinecke Early Modern Paleography blog: now apparently defunct, but useful nonetheless with a variety of sample pages of different handwriting styles. At

A simple Google Image search on “paleography” will also provide numerous examples of the plethora of handwriting.

There were also various letter-writing and paleography manuals published in the early modern period.


Tags: , ,

22 responses to “Paleography and Mystery Words”

  1. Gene Hughson says :

    Working from context as much as the picture, I’d suggest this: “They would rather weary France than [just take] him and if this war was soon ended, they would be too ready to split asunder within.”

    • jostwald says :

      Wow. That actually makes sense. “Just take” isn’t as sexy as I’d like/was expecting, but it does seem to fit the context. I’ll see if anybody else has any other ideas, but your reading has jumped to the top of my list so far.

  2. Rick Herrera says :

    An archivist at the Clements gave me a helpful suggestion in trying to decipher unclear handwriting: hold an imaginary pen above the word and trace it out in the air. The muscle movement will sometimes prompt your memory or strike a familiar chord in you memory.

    • jostwald says :

      Interesting idea. Thanks Rick!
      I’ve noticed among my undergrad students that practically none of them write in cursive (homework, exams…), and when asked, most say that they barely spent any time on it in elementary school. I wonder if that bodes ill for future historians?

      • sappling says :

        I think the digital age poses a lot of unique challenges to historians–namely how to preserve when technology becomes so quickly obsolete and what to preserve out of the deluge of information available.

    • sappling says :

      Nice trick! Will totally help in transcribing old census documents for genealogy.

  3. learnearnandreturn says :

    Because it’s so difficult to scan old handwriting, I’ve come to the conclusion that it’s worth transcribing the entire piece – fortunately I can touch type, so I just keep my eye on the MS and my fingers on the keys. It’s effectively a letter-by-letter approach, but although tedious and bad for the back, it’s fairly reliable.

    If I’ve been able to photograph the originals, I include scraps of illegible script in my file notes, in the hope that I will be able to read it later, once the context is clearer, or get help from others.

    Has anyone else come across a very annoying feature of private correspondence, as follows:
    Somebody writes a letter addressed to their closest correspondent, but with the expectation that it will subsequently be passed on to other readers. So anything private (i.e. the bits I’d really like to read) are underlined, and it’s the duty of the first recipient to obliterate those underlined pieces of text before putting it into general circulation.
    This is a problem I’ve encountered in long, expensive, descriptive letters sent home from the colonies to extended families. It was current practice in the early 19C – there’s a reference to ‘too many words underlined’ in a letter from Lydia to Kitty Bennet in Pride and Prejudice. I imagine there must be ways of reading below the ink scribbles – but not ones most libraries would countenance.

    • jostwald says :

      Interesting. I’ve come across various marginalia, underlining and strikeouts in correspondence. But I can rarely tell who did it. Sometimes it’s obvious it’s a correction by the writer. But other times it’s unclear – e.g. if you have an entire paragraph crossed out. Were those marks made by the writer? the recipient? a later reader? an archivist? a historian/biographer reading it decades later? And for what purpose exactly? (In a few cases you can tell it’s a later historian because their quotes match up with the underlined passages in the original documents.) But I can’t remember ever seeing a letter with portions excised or obliterated. Perhaps that’s because the content of my documents (written by/for generals-diplomats-ministers) are already considered state secrets, so they’re not likely to be shared too widely and they don’t often deal with embarrassing personal details. If some details are confidential, they’ll either communicate it by mouth or just use a cipher – sometimes the recipient has deciphered it in the letter, other times not. Other times, you’ll come across the occasional “I won’t divulge my thoughts in print but I’ll let you know my opinion when we meet later.” Something someone in the colonies isn’t likely to have the opportunity to do, I’d guess.

      For Marlborough’s papers, I know that on occasion he would write his real feelings to Godolphin and have a separate milquetoast letter to be shown to the cabinet or others. Often times, the recipient (e.g. the Lord Treasurer) would only read excerpts in a cabinet meeting, so he could pick and choose which sections became ‘public’ in that respect. On occasion authors will explicitly instruct the recipient to share the letter with other individuals (always identified), and there’s also the occasional reference to having seen someone else’s correspondence – the English Secretaries of State shared a lot of info with each other that they were receiving from their respective sources. Of course there are also the enclosures, i.e. second-hand copies of letters (or extracts).

      Personally my favorite is the occasional letter where the author tells the recipient to burn it since it’s for their eyes only. I don’t want to think about how many recipients actually did that though; it would be too depressing.

      FWIW, in 1710 Sarah Churchill threatened to publish all the Queen’s private correspondence with her.

    • Bill Roberts says :

      Excellent post and thread! The issues aren’t just European. John Ericsson’s hand, for example, is easy to identify but very difficult to read. In addition, I have often cursed the practice of “cross-writing” to save paper.

      • jostwald says :

        Thanks. I personally haven’t come across much cross writing at all, except to fit one last line on a letter. More often, the writers usually squeeze the rest of the text (perpendicular) into the side margin. So I’m not sure under what conditions it was used: When paper/postage cost a lot? When you didn’t care if the recipient had to work hard to read your letter?…
        For those who don’t know, a Google Image search on the term will show lots of examples. Many of the examples seem to come from the 19C for whatever reason – you wouldn’t think paper/postage would’ve been more expensive in the 19C than in the 17C, but who knows.

      • sappling says :

        I HATE cross-writing. I haven’t come across many examples but it is so hard to decipher. I tried doing it in my own handwriting, and that was better, so I imagine that when receiving such a letter from a contemporary, it would not be such a pain.

  4. Gavin Robinson says :

    I reckon it’s ‘Irritate’ but I’d need to see more of the page for comparison to be sure. Probably not ‘Just take’ as the writer seems to use the long s except at the end of words. The 5th and 7th letters are almost certainly the same character and compared with ‘taking’ are much more like t than k.

    • jostwald says :

      Unfortunately I don’t have a better photograph of the letter. In the archives I checked and the first letter does look almost exactly like an ‘i’ at the beginning of a few other words; but I suppose it could be a j or even a p.
      The problem with ‘irritate’ is that from the context it seems the word should be a stark contrast between wearying and ???ing the enemy. ‘Irritate’ doesn’t seem nearly enough of a contrast. Here’s a bit more context, which I think supports my belief that the mystery word(s) should have something to do with the contrast between annihilation and attrition (to put it in modern terms):
      “But ye States are for shunning a battle, some ascribe it to weakness of spirit, otherse to prudence because ye consequences are uncertain. The temper of ye Dutch is to be slow, & ye present policy is judg’d to be rather to protract ye war than to make a speedy end to it. They would rather weary France than ??????? him and if this war was soon ended, they would be too ready to split asunder within.”

      • Gavin Robinson says :

        I’ve just checked the OED and there’s an obscure alternative meaning of ‘irritate’ that I’d never heard of before. In civil law it means ‘To make void, render of no effect, nullify’ and is synonymous with the term ‘defeat’ in English law.

      • jostwald says :

        Interesting. You’ve prompted a future post.

  5. David Underdown says :

    ON the numebring of months – the official New Year’s Day in England was 25 March until the adoption of the Gregorian Calendar on 1 January 1752. Under that scheme October really is the 8th month.

    • jostwald says :

      That’s an interesting idea (I’ve posted a few posts in the past on the Old Style/New Style divide). Do you have any evidence for it? I’m a bit skeptical: 1) how would they deal with the last week of March – was it month I? are they counting months starting on the 25th/26th?; 2) I don’t recall seeing contemporaries use Roman numerals for the early months (April-August), i.e. those months without Roman (numerical) etymology. I could be wrong though.

      • David Underdown says :

        Goods points – I should have been less definite. The clash in numbering apparent to us would have been less so at the time mgiht be abetter way of putting it.

    • Paul Lockhart says :

      I can attest that documents generated by/in early modern states that hadn’t yet adopted the Gregorian calendar but *did* recognize 1 January as New Year’s Day (the Oldenburg monarchy, for example, before 1700, and much of the Protestant Germanies (ducal and electoral Saxony, for example) frequently referred to October as “VIII ber” — and there October was clearly regarded as the tenth month.

      • jostwald says :

        Thanks for the confirmation. I think awhile back I posted a c. 1700 table (in English) listing all the different German states and whether they used OS or NS. What a mess.

  6. Erik Lund says :

    Watch those dots in Eugene’s letters. Sometimes they’re a code, although they can’t read it at the Kriegsarchiv. (Or couldn’t in 1994.)

    Also, in the German-writing world, you’ll often encounter writers who will render a German word in illegible Buchstabenschlussel land a French or Italian word in italics. Thank you, ancient dudes, from this researcher.

  7. Paul Jones says :


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

%d bloggers like this: