Reading Runes in the Exeter Book Riddles

Riddles and runes go together, at least in some of those found in the medieval codex known as the Exeter Book of Old English poetry (Exeter Cathedral Library MS 3501).

J. R. R. Tolkien puts their cryptic association to creative use when, in The Hobbit, the dwarves’ map reveals to Elrond in runic ‘moon-letters” a riddle describing how King Thorin Oakenshield’s company will discover the secret door and enter the Lonely Mountain of Erebor once they arrive to reclaim their stolen treasure-hoard from the dragon Smaug.

Moon-letters revealing the riddle of the secret door into Smaug’s lair from Peter Jackson’s film adaption, The Hobbit: An Unexpected Journey (2012).

As with Tolkien’s moon-letters, runes found in the Exeter Book Riddles serve to both obscure and illuminate their riddle and its solution. This is to say—if you are literate and can read the runic alphabet—you have an important clue to solving the puzzle. If not, the riddle’s solution is even further obscured from the solver.

While certainly not every Exeter Book riddle contains runes, and indeed most do not, there is a higher frequency of runes in riddles than elsewhere in the extant corpus of Old English poetry, suggesting that perhaps runes offered something useful to the playful, puzzling, at times comical, Old English riddle.

Moreover, there is a general instability in the consistency of runic characters, and this further adds another enigmatic layer of obscurity to a riddle, since no runic standard of writing—or carving—ever truly existed in any standardized form. Rather, form and style of runic inscriptions (as well as the orientation of runic characters) varied wildly in medieval England and Scandinavia, which makes reading runes especially difficult even to those with some runic literacy.

So how do runes enhance a riddle? If one can determine what letter a given runic character corresponds to in the Latin alphabet, how does this knowledge illuminate the riddle and its solution? By looking carefully at Exeter Book Riddle 19 and Riddle 24, we will now explore how runes operate within a broader riddling framework.

Riddle 24 on folio 106v, Exeter Cathedral Library MS 3501.

Ic eom wunderlicu wiht,         wræsne mine stefne,
hwilum beorce swa hund,         hwilum blæte swa gat,
hwilum græde swa gos,         hwilum gielle swa hafoc,
hwilum ic onhyrge         þone haswan earn,
guðfugles hleoþor,         hwilum glidan reorde
muþe gemæne,         hwilum mæwes song,
þær ic glado sitte.         G mec nemnað,
Swylce. A ond R         O fullesteð,
H ond I.         Nu ic haten eom
swa þa siex stafas         sweotule becnaþ.

“I am a wondrous thing—I change my voice:
sometimes I bark like a hound
sometimes I bleat like a goat,
sometimes I squawk like a goose,
sometimes I screech like a hawk,
sometimes I imitate the grey eagle,
the sound of birds of prey,
sometimes I utter with my mouth the kite’s voice,
sometimes the gull’s song,
where I gladly sit.

G names me,
also A and R.
O supports me,
H and I.

Now I am called as those six letters clearly show.”

The solution to Riddle 24 is higora, or ‘magpie’ in Old English, as the runes indicate when spelled out. In this riddle, runes function to obscure the solution from anyone unable to read these cryptic characters, but paradoxically they function also to illuminate the solution for the literate solver able to read the runes. However, as mysterious as the runes might appear to some, for those who understood them they aided in solving the puzzle.

Another riddle that uses runes is Riddle 19. In this enigma, there is a game of misdirection:

Riddle 19 on folio 105r, Exeter Cathedral Library MS 3501.

Ic seah on siþe    S R O
H hygewloncne,         heafodbeorhtne,
swiftne ofer sælwong         swiþe þrægan.
Hæfde him on hrycge         hildeþryþe
N O M         nægledne rad
A G E W.         Widlast ferede
rynestrong on rade         rofne C O
F O A H.         For wæs þy beorhtre,
swylcra siþfæt.         Saga hwæt ic hatte

“I saw, on a journey,
S R O H,
proud in spirit, head-bright,
running very swiftly over the fruitful plain.
It had battle-glory on his back,
N O M,
A nailed road,
A G E W.
Traveled the far-paths,
run-strong on the road,
brave C O F O A H.

The journey was the brighter, that very expedition.

Say what I am called”

Riddle 19 contains one of the prosopopoetic riddling challenges, enigmatic formulae which conclude many in the Exeter Book collection and prompt the reader to solve the puzzle: saga hwæt ic hatte “say what I am called.”

In order to answer this enigmatic challenge, one must first understand the riddle of the runes. In this case, if one deciphers and reverses the runic characters, the letters spell out a number of Old English words that allows the solver to understand the riddle in its entirety. The runic words are decoded as follows:

S R O H = hors (horse)
N O M = mon (man)
A G E W = wega (way)
C O F O A H = haofoc (hawk)

Now the riddle becomes more comprehensible, though not totally, as the runic words create syntactic breaks in the poem:

“I saw a horse on a journey,
proud in spirit, head-bright,
running very swiftly over the fruitful plain.
It had battle-glory on his back,
a man.
A nailed road,
the way
traveled the far-paths,
run-strong on the road,
a brave hawk.

The journey was the brighter, that very expedition.

Say what I am called”

With these words semantically integrated into the riddle, some resemblance of sense is gained. The solver of Riddle 19 may now better comprehend the riddle’s meaning; however, its solution is by no means as clear for the literate rune-reader as higora is for Riddle 24. With the runes deciphered, Riddle 19 presents an image of a man riding a horse along a nailed road with a brave hawk. But this image seems to fall short of a proper solution to the riddle. Although answers have been put forth, none has proven satisfactory. Riddle 19 remains unsolved, a puzzle yet to be fully unriddled.

Richard Fahey
PhD Candidate
Department of English
University of Notre Dame

Further reading:

Bitterli, Dieter. Say What I am Called: The Old English Riddles of the Exeter Book and the Anglo-Latin Riddle Tradition. Toronto: Toronto University Press, 2009.

Murphy, Patrick J. Unriddling the Exeter Riddles. University Park, PA: Pennsylvania State University Press, 2011.

Spurkland, Terje. “Literacy and ‘Runacy’” in Medieval Scandinavia in Scandinavia and Europe 800-1350: Contact, Conflict and Coexistence, ed. Jonathan Adams and Katherine Holman. Turnhout, Belgium: Brepols, 2004.

What Lies Beneath: The Reliability of Watermarks as a Method for Telling Time

Rijksarchief te Gent K91 1r.

As a child of the 1990s, I remember well when US currency turned into a magic trick to show your other classmates in school: hold a bill up to the light and a president’s face would stare back at you. This “trick” was simply a watermark. In the late 1990s, American bills began to be printed with watermarks, those presidents’ faces, to deter counterfeiters. The watermark is not a modern innovation, nor was it created to prevent counterfeiting. It was a medieval creation, one used primarily to show the maker of the paper, but also the place in which the paper was made, and (less often) when the paper was made.

For this blog, I would like to discuss the risks of this third purpose of medieval watermarking- using watermarking as a means to date paper. I will rely upon two manuscript collections of the Rijksarchief te Gent, numbers K91 and K98, to demonstrate the uncertainty of watermark dating. Holding K98 is a “verzameling watermerken”- a collection of medieval watermarks beginning in 1386 and ending around 1500. The watermarks come from the Abbey of Saint Bavo’s in Gent, a prominent and powerful abbey within the medieval city.

The watermark collection is indeed just that- a bundle of loose leaves of paper, that for the most part, are not medieval at all. In examining the bundle of paper, one sees watermarks with an assigned date of the mark in the top left corner of the page, ranging from the late fourteenth to the late fifteenth century. However, most of the paper upon which these watermarks are imprinted or drawn are from the sixteenth and seventeenth centuries, with some drawings added later on carbon paper in the twentieth century.

Rijksarchief te Gent K98 1386.

As one can see below, the watermarks have been impressed on the paper, with the top sheet having an outline of the watermark done in pencil, and later sheets having only the watermark itself.

Rijksarchief te Gent 1389.
Rijksarchief te Gent 1389-1.
Rijksarchief te Gent 1390.
Rijksarchief te Gent 1390-1.


The archival inventory reads as if this collection of watermarks has been consistently compiled from the late fourteenth century by St. Bavo’s monks as a means to keep up with the numerous watermarks employed by the abbey.  Instead, the collection is a later, early modern creation used a means of dating the watermarks used in earlier centuries. But how effective and accurate are these watermark dates? Thankfully for our purpose here, the manuscript K91 helps to answer this issue. K91 contains the yearly accounts of St. Bavo’s from the late 1300s to the late 1400s, allowing us to see how well these watermarks, dated by year, correspond to yearly entries in other documents from St. Bavo’s in the fourteenth and fifteenth centuries.

Rijksarchief te Gent 1480.

Above, I have included the watermark for the year from 1480. It is the letter “P,” with a cross attached to the top of the letter and a bifurcated descender. According to the watermarking dating system of St. Bavo’s, this letter should only appear after 1480. However, in consulting the K91 account book (1384-1417), I found something quite different. The “P” first arises in the year 1402 of the accounts, a whole 78 years before it appears in the watermark compilation.

Rijksarchief te Gent K91 32r.
Rijksarchief te Gent K91 35r.

The clearest example is the “P” on the opening page of the account for the year 1403, the image K91 35r. just above. Perhaps this “P” is an exception? A dating oversight by the monks? In fact, it is the rule. Throughout the years for which account books survive, (1384-1417 & 1461-1498), very rarely do the accounts line up with the prescribed watermarks for that year. One watermark nearly hits the collection date, the watermark of 1395, the bow. The bow does not occur in 1395, but it does show up three years later in 1398, a much more reasonable margin of error.

Rijksarchief te Gent K98 1395.
Rijksarchief te Gent K91 23r.

The explanation for the inconsistency of the bow watermark is quite simple and known to many medievalists already. The monks of St. Bavo’s bought paper from a particular paper maker and they used that paper for many years. This is evident in the appearance of multiple watermarks in the same year of accounts (see the year 1403 in K91). The problematic example of the letter “P” presents a larger issue because the watermark appears far earlier than it should. It is not the only example of this phenomenon either- the watermark of 1410, a crown with a cross, first appears in the account books in 1407.

Rijksarchief te Gent K98 1410.
Rijksarchief te Gent K91 60r.

Based on the examples above, the efforts of the monks of St. Bavo’s were misguided in their attempt to coherently construct a means of dating their earlier documents. We know the monks were not papermakers themselves, so they did not have access to the wire frames with watermark outlines. These monks used their own watermarked documents to create their dating system, but they certainly did not rely upon the account books. If they had done so, they could not have drawn the conclusions they did. Perhaps they used a no-longer extant corpus of charters, but even in that case, the medieval monks of St. Bavo’s did not use watermarked paper in such a systematic and chronologically coherent way. They used the paper when they needed it- sometimes this was paper from decades earlier, sometimes this was new paper.

Further Reading-

Brown, Michelle P. Understanding Illuminated Manuscripts: A Guide to Technical Terms. Los Angeles: Getty Publications, 1994.

Briquest Online-


What’s in a name?

Titular Confusion in the Economic Records of Late Medieval Burgundy

Queen Elizabeth II. The Earl of Grantham. The King in the North. Rank and titles from popular television shows like The Crown, Downton Abbey, and Game of Thrones capture the modern imagination with images of conniving courts, opulent balls, and fancy dinners. But what did rank and title signify to the medieval mind? This question has provided the backdrop for some of the most prominent works of medieval scholarship, and the question has continued to interest and influence current work. One medieval society, in particular, had a healthy respect for, and perhaps an unhealthy obsession with, titles and status: late medieval Burgundy, the Grand Duchy of the West.

Burgundy earned its name of “Grand Duchy,” by the power of its statecraft and economy. At any given point, the Duke had a massive entourage of counselors. For major affairs of state, such as making peace with France, the numbers could go into the hundreds of men (Russell, Congress of Arras). The Duke kept very careful economic records of his payments to his subordinates at court, in what is called the Recette Générale. In studying the Recette, one is immediately struck by the length and structure of the titles of each person receiving payment. Titles mattered to the people of Burgundy. The order in which their titles were listed mattered too:

Archives départementales du Nord série B 1945, f. 330v.

The picture above is a standard entry in the Recette Générale. It reads (translated from middle French) “to the Reverend father in God, the bishop of Bethlehem, counselor and confessor to monseigneur the duke.” The order of the confessor’s titles here is consistent with the economic records in Lille: first came the appellation, here “reverend father in God,” but it could also be something like “sir,” or “my lord.” Following next generally was the name of the individual, which is omitted in this instance. Luckily we do know the name of this confessor was Friar Laurens Pignon from payments in earlier records. After the name came the confirmation of any title held by the payee and the place it was held. Here it is “the bishop of Bethlehem.” Lastly, the record includes the position (if any) the person in the entry held at court: “counselor and confessor of monseigneur the duke.”

With such careful attention placed on the titles, it is especially interesting when the accounts strayed from their standard formula. One mild variation is the exclusion of the episcopal title in payment to the confessor Pignon from two years earlier:

Archives départementales du Nord série B 1942, f. 56r.

As one can see, the expected “reverend father in God, the bishop” is missing. The receiver general has replaced the longer title with an abbreviated “monseigneur of Bethlehem.” Such a title does not indicate his episcopal role, although the normal positions at court are included, “counselor and confessor.” The omission of the first half of the bishop’s title is not especially surprising here. The confessor had appeared multiple times earlier in the accounts with his full title included (f. 27v-29r). In all likelihood, the explanation here is an instance of rushed record keeping. There was only one monsieur of Bethlehem, after all.

A more interesting “error” in the accounts concerning the confessor occurs in only one year of the Burgundian accounts: 1432, the same year that Pignon transferred from his old diocese of Bethlehem to the more prestigious diocese of Auxerre.

Archives départementales du Nord série B 1945, f. 60v.

Here, the title proceeds according to the expected formula: “to the Reverend father in god, the bishop of Auxerre, counselor and confessor of monseigneur the duke.” According to the careful naming conventions of the Recette Générale, the title of Auxerre should be found throughout the rest of the year. In fact, the opposite is true. The title of Auxerre appears one additional time in the accounts of 1432. However, the Duke of Burgundy pays Pignon ten more times in the year, always with the title of “Reverend father in god, the bishop of BETHLEHEM, counselor and confessor of monseigneur the duke.”

Archives départementales du Nord série B 1945, f. 103r.

The reversion by the Recette Générale to Pignon’s old title is undoubtedly an aberration- in the following years until his death in 1449, Pignon is identified as the Bishop of Auxerre. More importantly, the title of Bishop of Bethlehem no longer correctly referred to him- A fellow Dominican by the name of Dominic filled the post in 1433 (Eubel, Hierarchia Catholica vol. 2, 118).

There are a number of explanations for the oversight. The easiest one to dismiss would be the intentional disrespect of Friar Pignon by the Recette Générale by referring to him in a less prestigious way. No surviving evidence suggests that Pignon ever had issues with the Duke’s accountants. Indeed, he even helped them perform their duties in rare instances. Another explanation is that the members of the Recette Générale did not know of Pignon’s promotion- there were many members of the accounting body, sometimes with multiple receiver generals in a year. Perhaps the lower levels of the administration simply did not know of Pignon’s advancement to a higher bishopric. A third possibility is that the Recette Générale knew Pignon too well- again, Pignon shows up in the economic records every year he was in the Duke’s service, from 1412-1449. In 1432, he had been at court for 20 years and had been the Bishop of Bethlehem for almost a decade. To my mind, this is the most sensible and likely explanation.

This example of Laurens Pignon is meant to show something simple about the use of titles in the medieval period. Even a culture highly cognizant of standing, title, and rank made mistakes in this regard. The members of the Recette Générale knew Laurens Pignon, and had known him for many years. To them, he was the Bishop of Bethlehem, the counselor and confessor to the Duke. The promotion to Auxerre undoubtedly felt real to the confessor, but to the rest of court, it perhaps took more time to register.

Further Reading-

Blockmans, Wim, Antheun Janse (eds.). Showing Status: Representation of Social Positions in the Late Middle Ages. Turnhout: Brepols, 1999.

Duby, Georges. The Three Orders: Feudal Society Reimagined. Chicago: University of Chicago Press, 1982.

Eubel, Konrad. Hierarchia Catholica medii aevi vol. 2. Regensburg: Monasterii Sumptibus et typis librariae Regensbergianae, 1913.

Russell, Joycelyne Gledhill. The Congress of Arras, 1435: A Study in Medieval Diplomacy. Oxford: Oxford University Press, 1955.