Episode 24

The Shugborough Inscription Part 3 – The 46 Names of the Goddess

Welcome to the Poison Room, a podcast where I write a lot about how much other people have written about 10 letters, and then read it to you.

I’m Silvie Kilgallon, and this episode we’re continuing to look at a book by Dave Ramsden, called Unveiling the Mystic Ciphers, which attempts to provide a solution to the meaning of the inscription on the Shugborough monument. When we left off, Ramsden had made a lot of unreasonable proposals about using our 1767 anonymous poem to interpret the monument, based on an unfounded assumption that the author of the poem actually knew Thomas Anson at all, and a very tenuous comparison of the poem to a passage it Apuleius’ The Golden Ass.

Despite that, he did make one reasonable proposal: that we should take the meaning of the D · M on the second line of the inscription to mean ‘Dis Manibus’, which we can translate as ‘to the Shades’, as in, ‘this monument is dedicated to the Shades’. Unfortunately, Ramsden then returns to unreasonable proposals due to an unfamiliarity with Latin grammar, especially when it comes to inscriptions, and asserts that the other eight letters of the inscription must constitute a single name, because he mistakenly thinks that the ‘grammar’ of ‘Dis Manibus’ requires it. This requires him to ignore the fact that each of the remaining 8 letters is separated by an interpunct, which, as far as the grammar of Latin inscriptions is concerned, is a major hint that each letter represents a separate word. But ignore it he does, and turns back to our 1767 poem and takes the reference to the ‘mystic ciphers that conceal her name’ as proof that these letters are the name of the woman on the relief of the monument, who he thinks is a shepherdess, and that her name is encrypted.

But that’s not all. The Shepherdess’s name, must, he asserts, be one that fits into a syncretic tradition in which goddesses from various different religious systems who all share certain attributes are taken as different representations of the same divine figure. He takes this figure to be part of the tradition of a mystery cult, in which the true name of the divinity being worshipped is never spoken outside of certain ritual contexts. This, he thinks, is the reason the name of the Shepherdess is encrypted. Though now that I think about it, I can’t say I can think of any cults off the top of my head that encrypt the name of their deity. They tend to just use an epithet instead. Though if anyone knows otherwise, lemme know.

Anyway. That’s the shaky foundation on which Ramsden is building his theory, and when we left of Ramsden was just about to get down to the business of talking about cryptography. And cryptography, as we all know, is intrinsically, cool. So let’s talk about it.

First, let’s establish some terminology: our original, unencrypted text is known as the ‘cleartext’. The encrypted text is called the ‘ciphertext’. The thing that allows you to move between the two states of the text is the ‘encryption key’. Simples.

Ramsden is working on the assumption that a Polyalphabetic substitution cipher was used, based on the types of ciphers available at the time and the type of ciphers best suited for encrypting a very short bit of text. So what is a polyalphabetic substitution cipher? Well, let’s start with how a basic substitution cipher works.

In a standard substitution cipher, you replace every letter of the alphabet with a different letter. The simplest form of this is known as the Caesar cipher, because it was used by… Julius Caesar. In his version, you take a letter from you cleartext, and replace it with a letter that’s, say three places further down the alphabet. So you’d replace an ‘a’ with a ‘d’, a ‘b’ with an ‘e’, a ‘c’ with an ‘f’, a ‘d’ with a ‘g’, an ‘e’ with a ‘h’, an ‘f’ with an ‘i’, a ‘g’ with a ‘j’ … [voice trails off and then comes back]… a ‘w’ with a ‘z’, an ‘x’ with an ‘a’, and a ‘y’ with a ‘b’ and a ‘z’ with a ‘c’. So whatever letter you want to write, you shift three places down the alphabet, and write that letter instead. Obviously you don’t have to shift by three places – you could shift by 5, 12, 18 – take your pick. Your limit is how many letters your alphabet has.

In the grand scheme of cryptography, this is a super simple one to solve – frequency analysis does the trick. If you know that, say, ‘the’ and ‘and’ are the most common three-letter words in English, then you can just scan your text for any three letter repeating sequences and work from there. But, you can make it far more complicated by using multiple different shifts. As in, for each letter in your plaintext, you shift a different amount of places down the alphabet - that is the ‘polyalphabetic’ bit of the polyalphabetic substitution cipher.

There are many ways you can do this, but Ramsden decides on two ciphers that he thinks Anson would most likely have used. The first is the Vigenère cipher, which is named after a guy who didn’t create it. It was actually created by Giovanni Batista Porta, who first published on it in 1563. Because nothing makes sense anymore, Vigenère did actually create a far stronger polyalphabetic cipher, known as the autokey, which he published on in 1585. But for some reason people forgot about his idea for 300 odd years and only thought about it again towards the end of the 19th century. So he gets credit for a cipher he didn’t create, and no credit for the one he did create (Dooley 2018: 39). So unfortunately, I don’t get to explain the autokey to you this time, because it’s not one of the ones that could have been used.

The other cipher that Ramsden thinks is a possibility, after the Vigenère cipher, is what he calls the Sestri cipher, and which literally everyone else on the planet calls the Beaufort cipher. Let’s start with the Vigenère cipher.

The first time I used a Vigenère cipher was when I was a young teen and a Girl Guide. The leaders of our pack asked each patrol to come up with a coded message. Once we’d done so, we passed our message to another patrol, and they had to try and solve it. I had no idea then that it was called a Vigenère cipher. Or that it would be… too hard to solve. What I knew was that it was a cool cipher I had learned by reading a scroll in Castle Harmondale whilst playing Might and Magic VII.

So I’m going to explain this cipher how I learned it, which was algebraically, rather than using other methods such as a big table full of alphabets, or a cipher disc.

Step one: write out your alphabet. Let’s assume we’re working with an English alphabet, since that’s what we’re using for the Shugborough monument.

Step two: write out the numbers 0-25 under your alphabet, so A is 0, B is 1, C is 2, and so on.

Step three: decide on a key. Your key is a word or phrase that you’re going to use to encrypt the text. The longer the key, the harder your cipher will be to break. Let’s go with a key of ‘Poison’.

Step four: write out the message you want to encrypt.

Step five: write out your key repeatedly under the message, until there’s a letter from your key under each letter of your text.

Step six: add together the value of the letter in your plain text, and the letter in your key. Find what letter that new value belongs to, and write it down. So, for example, let’s say that for reasons of funsies, we want to encrypt the phrase ‘big elephants can’t always understand small elephants.’ The first letter of our message is a B. The first letter in our Key is a P. B has a value of 1, P has a value of 15, added together is 16, and that’s the value of the letter Q. So Q is the first letter in the ciphertext. The second letter in our plaintext is an I. The second letter in our key is an O. I has a value of 8, O has a value of 14. Added together we get 22, which is the value of the letter W, so W is our second letter.

When you’ve got to the final letter of the key, you just start again from the beginning of the key – so in this instance, the seventh letter of our plaintext would be once again added together with the value of the ‘P’ from the ‘POISON’ key. That letter is also a P. P has a value of 15, which gives us a total of 30, which means we get to 25 and still have 5 left over. So we go back to the start of the alphabet. Now this is the part that might be a bit confusing. A has a value of 0, but if we’re counting around continuously, it also has the value of 26, ‘cause A is what comes after Z, which is 25. So A is 26, B is 27, C is 28, D is 29, and E is 30. So E is our letter. In practice all you need to remember is that skipping back to the start of the alphabet subtracts one from whatever number you’ve got left. So in this instance, we’d subtract 1 from 5, end up with 4, and discover that four is indeed the value of E.

Step seven: now you’ve got your ciphered message written out, rip up your working out, burn it, and just remember the key. Oh, and also communicate the key somehow to the person who you want to be able to decipher the message.

Job done. Simple, right?

So that’s the Vigenère cipher. Now on to the Sestri cipher. Don’t worry, now you know how the Vigenère cipher works, the Sestri cipher is super easy to explain. Just use the same numbered alphabet as you’ve already got, write out your message and then write the key out under it just as you would before. Except this time, instead of adding the value of the plaintext letter and the cipher key letter together, you subtract the value of the key from that of the plaintext. So the first letter of our key is P, still with a value of 15. The first letter of our plaintext is B, still with a value of one. 15 minus 1 is 14, which is the value of O, so O is the first letter in our cipher. That’s it. That’s the difference.

So there’s a couple of points I want to make here. Firstly: there is absolutely a good argument for this to be a plausible starting place, from a cryptologist’s point of view. It just makes sense that if you want to write something in code, you want to use the best code you can. But I don’t think, in this case, ‘this is the most secure type of cipher for the job’ is a good enough argument to look two polyalphabetic ciphers and only two polyalphabetic ciphers. Especially when one of them is the Sestri cipher. As I said before, literally everyone else calls this the Beaufort cipher, after Admiral Beaufort, who, uh, ‘reinvented it’ in 1857. Someone else, a guy called Giovanni Sestri, had published on it back in 1710, but his book, apparently, had gone mostly unnoticed (Kahn 1967: 202). Which I’d have to say still seems to be the case, because I went through about 6 books on the history of cryptography and only found one book that mentioned him, just to say ‘yeah, he published on it in 1710 and got ignored’. So it seems curious that Ramsden is choosing to include it here, since there’s not much likelihood of Anson having known of it. But perhaps he’ll have something to say about that by way of explanation. We’ll see.

Secondly: when mentioning points about Virgil, Ovid, and Apuleius, as well as for the interpretation of the Poussin painting, Ramsden has made sure to note that there had been copies of these books in the Anson library – though that itself is not proof that the book belonged to, or was read by, Thomas Anson. However, he’s made no mention of any books relating to cryptography that could be used to get a better idea of what kind of ciphers Anson might use – or whether he was interested in cryptography at all. The list that Ramsden uses to check what books Anson owned is a copy of a catalogue of the property from Shugborough Hall to be auctioned off during 1842 because… So, remember Viscount Anson? The guy who was called George Adams but who changed his surname to Anson when he inherited the estate after Thomas Anson’s death? Yeah. Him. His son, who was obviously also called Thomas Anson, had a wee bit of a gambling problem, and wracked up enough debt that they had to sell stuff to pay them off.

So yeah, we have a catalogue of some of the books that were in the library in 1842, because they got sold off. Unfortunately the catalogue is not available online, and is kept at the William Salt Library in Stafford. Which: a) closed right now because of the ‘rona, and b) charges what I feel are very steep fees for getting copies of documents. So whatever. Can’t check those myself. So it is perfectly plausible that there were, or still are, books in that collection that relate to cryptography. However, our man Andrew Baker notes in his book that the library collection was catalogued in recent years, and revealed some books surviving from Thomas Anson’s time (Baker 2019: 222). If Ramsden made use of this new catalogue, he doesn’t say. Which, as a general rule, either means he didn’t know about it, or did check it, and didn’t find anything.

Anyway. Ramsden doesn’t provide any further justification for his cipher choices beyond their suitability from the perspective of a cryptographer, and I’m not going to put time into trying to do that for him.

But how does this idea that it’s a polyalphabetic substitution cipher help with figuring out the meaning of the inscription? We don’t know the plaintext. We don’t know the cipher key. Without any clues to help us narrow down the answer, we have 208 billion different combinations to work through. And lots of them – a lotta lot of them, will spit out intelligible answers, which means you still have to have a way of figuring out which of the potential solutions is the real solution.

One way we can try and sift through all the possibilities is by looking at the key that would be required to produce each potential plaintext. Traditionally with a Vignère cipher, the key you use is a short word or phrase that you can easily remember. So at least you know that you’re hypothetically looking for not just a plaintext that makes sense, but a key that’s intelligible, too. Of course, there are already problems here, because if, say, the key you chose to use was the first letter of each of the first eight words from a certain text – a possibility that Ramsden himself acknowledges when first describing how the cipher works, then… it’s not something that would jump out as intelligible. Point is, even if the idea that it’s a cipher was based on a reasonable presumption, the odds of getting to the right answer are still not great. And we’ve still got a hecka lotta combinations to work through.

But don’t worry, Ramsden’s got a plan:

208 billion is a daunting number, and a eight-letter ciphertext is too short to crack using letter frequency analysis. However, we reduce these odds dramatically by using the historical information available. First, we know from the anonymous poem that the “mystic ciphers” conceal a name. Second, we know that it is a female name. Third, we know that it is an eight-letter name.

Essentially what Ramsden did was drastically reduce the list of potential keys by drastically reducing the list of potential answers. He believes he has proof that the eight letters represent a single word, and that it is the name of a woman. And because he’s decided on… reaaaaaally sketchy evidence that it’s a name that belongs to a syncretic tradition, then his first step is:

to define the data by establishing a list of goddess figures common in the western cultures of Greece, Rome, Egypt, Mesopotamia and Israel, as well as those common to the Celtic cultures of the British Isles. Pruning that down to those that are 8-letters long, we are left with 52 names.

Okay, hey, that’s a very short list. Curious definition of “western” there, but sure, fine.

Then, to be more comprehensive in our analysis, we add all given female names common in England during the period of the monument’s creation.

Wait. What? Why? Why are we doing that? Where did this come from? What counts as common? Where’s this info about name frequency coming from? What exact time period are we talking about? Why? Why? Why? I have so many questions.

This was definitely one of those ‘put the book down and glare at it’ moments. But after I’d done that, I carried on, and he reveals that he’s getting these names from UK parish marriage records from the years 1530-1830.

Okay, that’s a start, but this bit of information is far from sufficient. For instance, given that neither the 1500s nor the early 1600s are anywhere close to ‘the period of the monument’s creation’, then I really want to know why he went back that far. This is the equivalent of someone using data from 1820 to investigate what names were common around 2020. So whatever. Whatever. Ramsden has magicked 434 addition names for his list out of nowhere. Fine. Fine! This is fine. This is totally fine. This is absolutely definitely totally how research should be done. FINE.

Given Ramsden doesn’t bother to explain his decision or the reasons behind it, I can only guess at them. So, here’s my guess: he’s adding these names because he ran through his ciphers with his original list of 52 names and didn’t find anything that he thought looked reasonable. Instead of rethinking what ciphers he was using, he decided to add more names. He could have, say, expanded his remit to include saints, or deities from more religions. But no. Instead, he goes for… common names in England over a three century period.

So now he’s got his list of names, he makes a table, and calculates for each name what the key would have to be in both the Vigenère and Sestri ciphers to result in O U O S V A V V. Remember, the idea is that he’s looking for both a name that would make sense, and a key that we’ll hopefully be able to identify. He provides the table of his working in an appendix at the back of the book.

At this point, honestly, I was tempted to go and check his list of goddess figures “common” in western cultures, because I think that’s a categorisation that needs a lot more explanation and discussion. ‘Common’ by what standard? Surely what matters more is whether they’ve been identified as part of a syncretic tradition, or fit an archetype of a ‘mother goddess’ somehow? And how did he cope with the transliteration of all these names from Egypt, Israel, and Greece into a Latin alphabet? Did he include all possible variants? I’d expect him to.

And part of me thought… ‘checking this out is not a good use of your time’, and then another part of me said ‘Must. Be. Thorough.’ Soooo I figured, ‘okay, the Greek Calliope is a “common” figure, right? And her name is usually spelled with a C, but the closer transliteration is with a K. And it’ll be easy to check - since I can just check whether he’s listed it with C as the first letter and with K as the first letter, and if he’s managed it with that one, I’ll give him the benefit of the doubt and assume he handled that issue, and if he didn’t, then… well, I will not give him the benefit of the doubt about the other names. So off I hop to the appendix. Aaaaaaand… LOL. Not on there. At all.

Obviously, my next thought was ‘well… what Greek names did he include? And then I thought ‘Am I really going to flick through an entire table of 486 names just to look for a handful that are Greek in order to see what he deigned common enough to include? On a kindle screen with tiny text? In a list where he didn’t bother to distinguish in any way which are the 52 “common” goddess names?’ And then another part of me said ‘Research. Must. Flow.’

So after going through all the names on that list, I’m even more puzzled as to what he thinks a “common” name is from these ancient religions. Y’all. I’ve been doin’ classics for quite a while now, and I did not see a name on that list that I recognised as a “common” goddess in Greece. There were five obviously Greek names that I found whilst scanning the list, and not one of them is the name of a goddess, common or otherwise.

And then, dear audience, I had a thought. A… slightly worrying thought, about what the absence of Calliope, amongst others, might mean. I flipped back through the table and realised the Roman goddess for victory, Victoria, was also missing. This was… weird.

So I counted. 432. The list in the book is 432 names. There are 54 names missing from the list. That is just over 10% of the list that’s just not there.

I put the book down, and I went to bed.

So, the next day I picked the book up again, and flicking through it, realised that although his claim that there was a full table of all the names in the appendix was false, there are two much smaller tables within the chapter itself. So obviously, I checked those names against the appendix. The first table was four names that were all goddesses – including Calliope! It also included Despoina – which is a very standard epithet used for Persephone; Rhiannon – who is… not a goddess, but is kinda supernatural, and Shekinah, which he’s pulling from Jewish Kabbalistic literature.

The other table is his shortlist of names, which we’ll discuss in a bit, but it contains 16 names. Of those names, four – the names of goddesses, of course – do not appear on the appendix table. But I guess at least we’re now missing only 46 names of goddesses? And that means it’s only 9.4% of names that are missing!

So here’s where we are at this point, y’all:

Ramsden is using the rules of Roman inscription writing to translate the D M of the inscription, and then throwing those rules out the window when it comes to the remaining eight letters. His justification for doing so is not based on any evidence in the monument itself, but on a line in an anonymous poem by someone he cannot prove ever even met Thomas Anson. Based on a very flimsy reading of just a select few lines from description of the monument in this poem, he has decided that these 8 letters represent a name of a goddess or feminine figure who is part of a syncretic tradition.

He has decided that of all the encryption methods available at time the, it is most likely that a Vignère or Sestri cipher was used, and that they length of the key used to encrypt the text was eight letters. Using an unknown criterial to define “common”, he has compiled a list of 52 8-letter names of “common” goddesses from Greek, Roman, Egyptian, Mesopotamian, Israeli, and Celtic cultures.

Without providing any explanation, and using largely unspecified criteria, he has also thrown in some common feminine names procured from marriage records in England over a three-hundred year timespan. And he has provided in an appendix a table of names and the keys that would be required to encrypt them to result in the cipher text O U O S V A V V, which, it turns out, is missing just under 10% of the names that should be on it.

Anyway, this was kind of the point at which I stopped caring about trying to take Ramsden’s argument seriously. This research is sloppy and, let’s be real, in all probability the reason he added 434 “common” names from between 1530-1830 is because his initial list didn’t turn up anything interesting, so he expanded his search to include women’s names that don’t even fit his own stated criteria. When your list of common English names that don’t have anything to do with syncretic traditions is over five times longer than your list of actual goddesses’ names, you probably done gone wrong.

And if I wanted to be really cynical, I’d suggest that he chose the time frames for names that he did so as to make sure to include the name that he wanted it to be.

But let’s get back to his attempt to make something of his supposed ciphertext.

Upon close inspection of the 104 keys that we have produced through our reverse engineering of the 52 goddess figures’ names, the initial result reveals that there is not a single name which produces an easily recognisable one-word key.

Well colour me surprised.

We then continue to produce keys for the comprehensive list of names, generating a grand total of 972 keys to examine. Again, we find that none of the results consist of a one-word key.


The next step in our analysis of the keys is to look for patterns that may indicate the use of multiple words or an acrostic taken from a poem or song.

Interesting. I wonder how you go about checking all the acrostics. Like, do you start with Ovid, Virgil and Apuleius – the texts Ramsden has cited to back up his interpretation of the monument, and… look at the parts where Ovid talks about the ilex, and any mention of Arcadia or shepherds in Virgil, and any mentions of syncretic stuff in Apuleius? And then… expand that kind of thing to the whole library?

We find the types of fragments and short words that any random sampling of text may generate. Seeking to refine our search for a more focused analysis, we take all keys with a four-letter word or longer and examine them more closely.

Wait, what happened to the potential acrostics? How did we rule those out? What texts did we check? This. Is. Ridiculous. Ramsden has just… quietly dropped any mention of potential acrostic formulations at this point because it’s inconvenient to him. It would be a heck of an effort to check all the books that Anson owned that would be potential candidates for the cipher key. So he just… doesn’t say anything about them, and hopes you won’t notice. This is not. Good. Scholarship.

So, what result does Ramsden get after ignoring any potential acrostics and just looking for words? Nothing. Nothing that he recognises as a name, or a word in general. So what do we do now? Well, we start looking for words of four or more letters that appear within the list of potential keys. This brings his list down to 16 names. The words contained in the various potential keys are: Ohio, Oops, Maumau, Toes, zest, foot, thou, thou, odor, owen, warp, rest, pews, aloe, warp, and warp.

I had a suspicion that Ramsden only looked at English words, and a very, very quick check seemed to confirm that, since one of the other keys contained ‘vino’ which is the singular dative and ablative form of the Latin word ‘vinum, meaning ‘wine’.

Now, if I read out the sixteen names we’re left with, I’m preeetty sure most of you will figure out immediately what name Ramsden is going to settle on. Here we go: Angerona, Ataegina, Debranua, Neverita, Bettrice, Bridgget, Magdalen, Magdalle, Arabaell, Chatheri, Cisseley, Farbrick, Mildreda, Ruderfor, Sisselie again, and Sysselly again again. Three different spelling.

Can y’guess which one it is yet?

You may have noticed that this list is not in alphabetical order. Ramsden has conveniently arranged it into two groups: the first eight have a “spiritual” connection. The last eight do not. So we rule those out, and wonder why we bothered to include random popular women’s names in the first place if we’re then just going to dismiss those without a spiritual connection.

So we’re left with Angerona, Ataegina, Debranua, Neverita, Bettrice, Bridgget, Magdalen, and Magdalle. So we’re gonna talk about the answer it clearly is now, right? Hahah, no. I am a master of Homeric retardation of the plot.

So before we get to what Ramsden decides the answer is, we’re gonna look at these eight finalists. Because who doesn’t want to learn about obscure goddesses?

Angerona and Neverita are Roman goddesses. Neverita is super obscure. Like… ‘mentioned-once-in-one-single-also-really-obscure-text-and-no-other-evidence-of-her-exists’ obscure. The text she’s mentioned in is from the 5th century CE, written by a guy called Martianus Capella. It’s called ‘The Marriage of Mercury and Philology’, and I love the description of it given by Nancy Thomas de Grummond, as ‘a compendious pedantic allegory… regarded as eccentric, tedious, and superficial in its discourse on the seven liberal arts. (Thompson de Grummond 2006: 4-5)’. I want it.

Anyway. Not exactly the kind of text you turn to for established religious figures that actually exist outside of one single allegorical text. So that’s Neverita. What about Angerona? Given the whole vibe Ramsden’s got going with his theory, Angerona would have been a really good fit for him.

Pliny the Elder makes passing reference to her in his Natural Histories, and here’s what he says:

Rome itself, whose other name is held to be a sin to utter except at the ceremonies of the mysteries, and when Valerius Soranus divulged the secret religiously kept for the well-being of the state, he soon paid the penalty. It seems pertinent to add at this point an instance of old religion established especially to inculcate this silence: the goddess Angerona, to whom sacrifice is offered on December 21, is represented in her statue with a sealed bandage over her mouth. (NH 3.65 = 3.29. (26)).

Another source, Macrobius’ Saturnalia mentions that instead of a bandage over her mouth she might be depicted with a finger pressed to her lips.

There are other explanations for Angerona’s silence in other sources – Macrobius offers several different explanations in his Saturnalia which is a text that takes the form of an imaginary dialogue, and is, by his own account, a collection of things that he thought were worth knowing, and includes a lot of quotes from other writers. In the case of Angerona, Macrobius offers three sources which say different things about the goddess. The first is from Flaccus, who says that Angerona is so called because she banishes angores - i.e. anxiety. To this explanation Macrobius adds a detail from a guy called Masurius, who stated that an image of Angerona with a sealed mouth was placed on the altar of Volupia – another obscure Roman goddess – to indicate that those who concealed their pain and anguish would eventually come to great voluptas. Which means something like pleasure or delight, in either a physical or a spiritual sense (Saturnalia I, X: 7-9).

So… a goddess connected to a mystery cult, but possibly also connected to secret grief? Well that’s clearly not relevant to Thomas Anson at all. Nope! But, oh, hey – look! Angerona begins with an A, and there’s an A in the Shugborough inscription! And Volupia begins with a V, and there’s three of those, and two of them are right next to the A! Coincidence???

Yes. It’s a coincidence. And I’m sure Ramsden would agree with me on that one. After all, the four letter word found in the key that would make her name the cleartext is ‘Ohio’.

The final explanation Macrobius gives for Angerona is from Julius Modestus, who says that sacrifices are offered to her because, upon the fulfilment on some unspecified vow, she cured the Romans of angina.

Bear in mind that this is what the ancient sources said about Angerona, and her origins. And ancient Greeks and Romans were decidedly bad at figuring out where they’d got some of their ideas from. There’s still not a definitive scholarly consensus on what the precise deal with Angerona is, but, again, in terms of what sources and knowledge would be available to Anson and his fellows, Pliny is the most likely source, followed by Macrobius.

Oh, and for the record, the wikipage on Angerona is a mess and I did not try and fix it.

Let’s look at Ataegina next. Ataegina, or Ataecina, with a C rather than a G, as appears to be the more common spelling, is a goddess native to Lusitania, on the Iberian Peninsula. Today that area is Spain and Portugal. She is apparently one of the most well documented deities of those indigenous to the area, though that’s still only around 40 inscriptions related to her. Apparently some of them are curses though, which is always cool (Simón 2005). Most of the dedications and altars to Ataecina have been found in the western areas of Spain and Portugal, and a notable amount associate her with Proserpina, or give Prosperpina as an epithet (Fernández Corral 2016; Abascal Palazón 1995). So from the few hours I could afford to spend poking around trying to find out more about her, there doesn’t appear to be much more that we know about her, unfortunately. Now let’s talk about Debranua. Oh dear Debranua. Debranua is a celtic goddess of, apparently, speed and fat, and she was born on the 19th of February… in 2008. On that day she appeared on a list of supposed ‘Celtic Deities’ on a website called lowchensaustralia.com. The site has a bunch of pages all listing various gods names, and it looks like it was designed in 1995. But… when you click the link to the ‘site index’ you get to a much shinier home page. And when you get there, the true nature of this website is revealed.

It is a website about dogs. It has sections dedicated to dog grooming, dog health, dog breeding, etc. and one dedicated to… dog names. If you go to the ‘names’ section of the website, you can see they have lots of “articles” (by which I mean, pages that are purely lists) for dog names. Prior to this website being published there is no mention anywhere on the internet of her existence. She doesn’t appear in books like the Oxford Dictionary of Celtic Mythology. I have no idea where this dog website got this list from, but… she doesn’t exist. Except on the multiple lists of names of Celtic goddesses on the internet that either copied this list, or copied from another page that had copied from this list. And of course, that meant she ended up on Wikipedia, too. For a while, at least. Someone eventually cleared up the page she was listed on and removed her. But not before her name made it into a couple of books, including the Druid’s Primer, published in 2012, and The Witches Guide to Wands, published in 2015. Soooo… that’s embarrassing.

So… where did Ramsden get her name from? From Wikipedia, before they fixed the page? Or one of the multiplicity of sites that seems to have copied and pasted the list of names from this dog website? WHO KNOWS. But in terms of quality of research, this is… ridiculous.

Anyway. We’ve covered Angerona, Ataegina, Debranua and Neverita. These four are not listed in the appendix table. All the rest of them are. So the rest of them are probably not from his goddess list.

What about Bettrice and Bridgget?

Well, when it comes to Bridgget (spelled B-R-I-D-G-G-E-T, just in case you were wondering how that could possibly be an eight-letter name) ,there is both a goddess in Irish myth called Brigid, and she was associated with fertility. And there’s also a St. Brigid. These two are definitely why Ramsden is now placing Bridgget on his shortlist, but I strongly suspect that this name is one of the ones he got from the marriage records. Partly because all the figures missing from the table in the appendix were goddesses, but Bridgget appears on that list, but also because a huge tell-tale sign for those candidates drawn from the ‘popular names’ list is that each single name appears there about 20 odd times, just because of the amazing array of different spellings of the name.

I’m guessing that the same is true for Bettrice – it’s one of quite a few variant spellings for Beatrice, or Beatrix, and there have been a handful of Saints with that name. So: probably from the popular name list, but turns out there’s a saint with that name so let’s keep her on the shortlist.

Okay. Time to get to the last two names. Magdalle and Magdalen. Magdalle is a variant of Magdalen. Which leaves us with Magdalen. I’m assuming that for many of you, you recognise the name, and have some inkling of why this should perhaps set alarm bells going for you. If not, lemme give you the low-down.

Mary Magdalene is a biblical figure from the New Testament. She appears in all four of the canonical gospels, but her name is only used explicitly 16 times – that’s not 16 different appearances, her name might be used several times within the same story, and frequently the stories in which she appears are shared across all four gospels (Ehrman 2006: 185).

In Matthew and Mark, Mary Magdalene was one of the women watching the crucifixion from a distance (Matt. 27:56; Mark 15:40). Luke doesn’t mention her specifically by name, but notes that ‘the women who had followed him from Galilee’ were watching at a distance, and that group includes Mary Magdalene (Luke 23:49). In John, Mary Magdalene is standing next to Mary, Jesus’ mother, at the crucifixion, and they’re close enough for Jesus to be able to talk to them (John 19:25).

In Matthew and Mark she’s there waiting at the entrance when Joseph of Arimathea brings Jesus’ body to the tomb (Matt. 27:61; Mark 15:47). Again, Luke doesn’t mention her individually, but mentions the women from Galilee waiting there, and then going away to prepare the spices and perfumes that were required for the ritual preparation of a body (Luke 23:55-56). All three specify that she’s one of the women who goes back to the tomb on the dawn of the third day and witnesses the angel roll back the stone blocking the entrance to the tomb and announce to them that Jesus has risen (Matt. 28:1-7; Mark 16:1-7, Luke 24:1-10). Mark also adds the detail that Mary Magdalene was the first person Jesus appeared to after his resurrection, and also decides that this is the appropriate moment to mention that Jesus had, at some point in the past, driven seven demons out from her (Mark 16:9). That same detail is given much earlier in Luke, which gives it in a more chronologically logic place, and also explains that Mary and several other women had joined Jesus in travelling around along with the 12 disciples after he had driven the demons out (8:1-2).

In John, only Mary Magdalene is mentioned as going back to the tomb on the third day, where she discovers the stone rolled back. In this account, there’s no angel there to greet her, so she assumes that the pharisees have had Jesus’ body removed, and runs back to tell Peter and another disciple as much (John 20:1-2). Peter and the other disciple inspect the tomb, and then leave again. But Mary stays behind, weeping. She looks into the tomb and sees two angels inside, and then Jesus himself, who she doesn’t immediately recognise, probably because her facial recognition algorithm sensibly removes dead people from the list of possible subjects. But then he says her name, and she recognises him. He tells her to go tell the disciples that he’s back, and off she goes (John 20: 11-18).

These are the only mentions of Mary Magdalene in the gospels. If you’ve seen Jesus Christ: Superstar or The Last Temptation of Christ, or read The Da Vinci Code you’ll be familiar with the idea that Mary was the lover, or even wife, of Christ, and, in the case of the Da Vinci Code, that they had children together. But, as anyone whose already interested in the subject of Mary Magdalene will know, embellishment on her role outside of that given in the canonical gospels in long and varied. You may have heard that she was a reformed sex worker, that she anointed Jesus’ feet with oil and then dried them with her hair, that her brother was Lazarus, who was raised from the dead by Jesus.

None of these things come from the canonical gospels. In fact, in John, it’s specifically a different Mary – Mary of Bethany – who anoints Jesus’ feet with oil (12:1–8). And it’s also Mary of Bethany who is the sister of Lazarus. So some of the confusion is just from people getting their Marys mixed up. And the sex worker thing seems to have come from men deciding that a word that simply meant ‘sinner’, when applied to a woman, must mean ‘sex worker’ (Ehrman 2006: 188). And over the years, others have thrown in other conflations. For instance, you may have heard some variation on ‘let he who is without sin cast the first stone’. This comes from a story related in John’s gospel where some pharisees approach Jesus with a woman who they have caught in the act of adultery. According to the laws of Moses, the punishment for such an act was death by stoning. But the pharisees want to know Jesus’ take on it, primarily in order to get him to say something they could use against him as proof of him preaching things that were contrary to the accepted religious teachings of the time. So they ask him what he thinks should happen to her, and Jesus says… nothing. He doesn’t reply. He just starts writing on the ground with his finger. The pharisees persist with the question, until Jesus eventually replies ‘let he who is without sin cast the first stone’. Gradually, all the pharisees drift off in silence, until only Jesus and the unnamed woman remain. Jesus tells her that he will not condemn her either, and to go and sin no more (John 8: 1-11). This woman is also not Mary Magdalene.

So, essentially, Mary Magdalene is a very interesting figure, with a very common name, about whom we are told very little. Over the centuries, this has lead to lots of different stories being associated with her, when they aren’t about her at all. And she’s certainly not a goddess, but she is a saint.

And all of this was confusing enough, but then, along came a book called Holy Blood, Holy Grail. Which really pushed the idea that she was the wife of Jesus and that they had children int the public consciousness. And that’s where Dan Brown got it from, and it ties into a whole Priory of Zion blahblahblah conspiracy-esque thing which is all nonsense and will absolutely be a subject for another time. But this is why it should set alarm bells ringing for you. Because there’s so much nonsense out there and especially when someone is bringing it up post the publication of Holy Blood, Holy Grail you have to be a little bit suspicious of where you think this is going.


Those are our eight finalists. Just to remind you, here’s the names of all eight of them, with the four-plus letter word that appeared in the cipher key that would be required to transform any of these names into what we actually see on the Shugborough inscription.

Angerona (Ohio), Ataegina (Oops), Debranua (Maumau), Neverita (toes), Bettrice (zest), Bridgget (foot), Magdalen (thou), and Magdalle (thou).

So, how is Ramsden going to solve this? Well, firstly, he states that:

The words in most of them are hardly compelling, and feel more like we are grasping at straws.

Strongly agree.

However, upon close inspection, one of the keys which contains the word THOU is very noteworthy. The letters BTHOUOQH contain the triad OQH. In examining the tradition of divine or semi-divine feminine figures, we know that the title Queen of Heaven is almost universally applied, from Isis to biblical references to Astarte and on to the Blessed Virgin Mary. Now we have an interesting pattern with a thematic tie to our subject.

Strongly disagree.

But the solution becomes clearer when we realize that the clue has been before us all along, in the anonymous description of the Shepherdess’s Tomb left to us by one of Thomas Anson’s circle. At the very moment that the author refers to the “mystic ciphers” that conceal the Shepherdess’s name, the author formally alludes to Apuleius’s Metamorphosis. And not just any part of the Metamorphosis, but the exact moment when the protagonist pleadingly addresses Isis by name. And in addressing her, he begins, “O Queen of Heaven, be thou…”, continuing on with a long list of her syncretic names from various classical cultures. The poetic muse has, in fact, hidden a hint at the key without actually writing it down. BTHOUOQH appears to stand for some form of the phrase “Blessed be Thou, O Queen of Heaven.” Would Thomas Anson have understood? Certainly. Would strangers be familiar with the cipher? Not unless they had generated every possibly key and painstakingly examined them, and then only if they understood the bigger ideas at play in the monument itself.

So there you have it. Obviously the Shepherd’s monument is a secret shrine to Mary Magdalen. Congratulations, y’all. We have reached the top of Mount Not-How-That-Works. Because, of course, as we saw last episode, Ramsden has miserably failed to prove that the author of the poem had any idea what Anson’s intended meaning of the monument was, and, in fact, in trying to prove that the author of that poem understood syncretic traditions, Ramsden himself managed to offer a compelling counter-argument against the idea that the passage concerning the Shepherd’s monument was about syncretism.

Now, if you’re somehow not utterly convinced by this explanation of the key, then fear not! Ramsden’s got some explanations that’ll totally sway you. Probably. Possibly. Not.

But that’ll have to wait until next time, which, I promise, will be the final episode on this particular book.

Oh, and before I forget. Ramsden doesn’t indicate in his shortlist which cipher – Vigenére or Sestri – produced the key with a four-letter word in for each name. And obviously, with the goddesses, I can’t check. But I can check with Magdalene, and it was the Sestri cipher. That really well known one.

Thanks for listening. If you like the show, please subscribe on whatever podcatcher you’re using. Rate and review the show, especially on iTunes.

If you have questions, comments, corrections, feedback, want to suggest a topic, want to tell me what cool or actually useful thing you learned from playing video games, etc. You can find the podcast on twitter @poisonroompod,
or send an email to: poisonroompodcast@gmail.com

Alternatively, wrap a bandage around your mouth, press a finger to your lips, and strike down any who dare utter the sacred name of Rome.

Transcripts of all episodes are available at poisonroom.com, where you can also see the references and bibliography. As always if the sources are publicly available, they’re linked to.

You have been listening to The Poison Room, a podcast that doesn’t think Debranau is even that good of a name for a dog, to be honest.

The voice of Pliny the Elder was: Murphy Terrett.
The voice reading extracts from Ramsden’s book was: Inari Porkka.
The voice in your ears has been: a compendious pedantic allegory… regarded as eccentric, tedious, and superficial.


Abascal Palazón, J. (1995) ‘Las inscripciones latinas de Santa Lucía del Trampal (Alcuéscar, Cáceres) y el culto de Atecina en Hispania.’ Archivo Español de Arqueología 68: 31-105.

Fernández Corral, M. (2016) ’Nueva Árula a Ataecina en Oña (Burgos)’, Zephyrus Vol. 78: 203-209.

Simón, F. M. (2005) ‘Iberian Religion.’ Encyclopedia of Religion, edited by Lindsay Jones, L. (Ed.), 2nd Edn, vol. 6: 4249-4254. Macmillan Reference USA,

Baker, A. (2019) Thomas Anson of Shugborough and the Greek Revival, (unpublished, available online).

Dooley, J. F. (2018) History of Cryptography and Cryptanalysis, Springer.

Ehrman (2006) Peter, Paul, and Mary Magdalene: The Followers of Jesus in History and Legend, Oxford University Press.

Kahn (1967) The Codebreakers: The Story of Secret Writing, The Macmillan Company.

MacKillop, J. (1990) Oxford Dictionary of Celtic Mythology, Oxford University Press.

Thompson de Grummond, N. (2006) in ‘The History of the Study of the Etruscan Religion’ in The Religion of the Etruscans, Thompson de Grummond, N. & Simon, E. (Eds), University of Texas Press.

Rackham, H. (1906) Pliny: Natural History Vol. 2 Libri III-VII, Harvard University Press.

Ramsden, D. (2014) Unveiling the Mystic Ciphers, CreateSpace Independent Publishing Platform.

De Bernardo Stempel, P. (2013) ‘Celtic and Other Indigenous Divine Names Found in the Italian Peninsula’ In Hainzmann M. & Mathieu N. (Authors) & De Bernardo Stempel P. & Hofeneder A. (Eds.), Théonymie celtique, cultes, interpretatio - Keltische Theonymie, Kulte, Interpretatio. pp. 73-96. Austrian Academy of Sciences Press.