So here is an updated version of my original article. In the nine years since I first wrote this article, I too have learned a lot about the symbols.

The impossible maths

The first idea is that the pair of symbols usually found on symbol stones represents a person’s name: SYMBOL1 [mac] SYMBOL2, where SYMBOL1 is the personal name and SYMBOL2 is the patronymic.

No one, ever, has presented any evidence which lends support to this claim.

The maths of the situation cannot be denied. A symbol cannot represent a Pictish name, and symbol stones do not display a personal name.

Names in most cultures are many and varied. Ireland, for example, during this period has many hundreds of common personal names, and the same is true of other groups of the period. For example, a wide choice of personal names is found in both the Bodmin Manumissions (Cornish) and also in the Redon Cartularies (in Brittany). We should expect no less of the Picts. Pictish names recorded in the literature include Brude, Nect, Bile, Cano, Derelei, Drust, Artbranan, Talorc, Uoret, Uorguist, Uepogeni, just to name a few. Further lists of examples can be found on the web and in print.1

How many symbols are there? That depends on how symbols are defined, something which isn’t always clear, but the number is somewhere around 60. This number is obviously too small to represent a complete body of hundreds of common personal names.

Personal names aren’t used in equal frequency for all members of the group at any period. Some names are always going to be more popular than others. So, if symbols represented names, then we should expect some symbols to be more frequent than others, and this indeed is the case. The problem is though, the frequencies are way too heavily skewed. On CI stones, the 3 most frequent symbols, the crescent/V, the Pictish beast, and the double-disc/Z, make up 40% of all symbols. It is inconceivable that 40% of all Picts, all over Pictland, used only three personal names throughout all the centuries of symbol usage, and in fact, the evidence that we do have, even though limited, tells us this definitely wasn’t the case.

The evidence is quite stark. Each Pictish symbol does not represent a personal name.

Reviewing the proposals

To understand the current state of play, I will need to review what has led to the idea of names.

There have been three key works which have touched on the question of names. They call upon two ideas, the first being that the symbols as names are a novelty introduced under Roman influence to copy Roman name inscriptions; the second being that they are Christian gravestones which begin to have inscriptions in the late 7^th century.

W.A. Cummins2 (1999) tried to associate specific symbols to certain names on historical grounds. The data in his study was insufficient to be convincing, but this was an essential exercise to put the idea to the test.

Ross Samson (1992)3 wrote a large and detailed article in which he explored the social context and patterns of distribution of the symbols. He proposed that the idea of a memorial is a possibility worth considering. He then took a detailed look at the frequencies of names in other language groups of the time, and he demonstrated as invalid, on the basis of their frequency alone, the idea that one symbol can equate to one personal name. He did however go on to propose that the usual pairing of two symbols might reflect a two-theme name. For example, Brandub could be represented as two symbols for two ‘themes’, one bran ‘raven’, the second dub ‘black’.

The main focus of Professor Katherine Forsyth’s paper (1997)4 was to explore the implications of presuming the symbols are a form of writing. Many aspects of the symbols were presented, several of which fit with the idea of them being language, but Forsyth concluded that the symbols cannot be assigned to any form of written language known to man. Despite this, she still supported Samson’s assertion, published a few years earlier, that the symbol stones should be memorials with names on them. So why did these two authors still try to reach for this solution, despite not being able to find any supporting linguistic or mathematical evidence?

The answer lies in their own timeline, and the elephant in the room – the dating of the symbol stones. Both authors were writing in the 1990s, when it was promulgated by the art-historians that the Picts had been taught to carve CII symbol stones by the Northumbrians, thus pushing the CII stones out very late in dating, as Forsyth says in her paper, up to the 10^th century, and consequently dragging the dating of the CI stones upwards with them into the second half of the first millennium. Because of this assumption of very late dating, both authors then thought of the symbol stones as falling easily within the wider social context of the memorial stones that were being raised by neighbouring groups which appear to begin in the later 7th century. So, it seemed reasonable at that time to presume that the Picts were copying the Christian memorial tradition of their neighbours, and that the symbols were therefore created to represent the names on these memorials.

Since these papers were written, the archaeological dating of symbol stones, particularly by the Gondek & Noble team, is gradually becoming clearer. It is now impossible to reconcile the early dating of the creation of the first symbols (late centuries BC to early centuries AD) into the earliest possible beginning of the Christian memorial stone tradition (late 7^th to 8^th century AD). In other words, the symbols cannot have been created to represent names for the Christian memorial tradition, centuries before that tradition began.

The Roman inscription influence

Gondek and Noble in 2017 made this statement – with references:5

The Pictish symbols probably communicate names or identities (e.g. Samson 1992; Forsyth 1997; Lee et al. 2010) and the monuments are part of a wider northern European phenomenon of carved stone monumentality which became popular from the 5^th century onwards (HANDLEY 2003, 177; EDWARDS 2007,30; FRASER 2008,2).

It’s worth having a closer look at these references. Samson proves that symbols aren’t names by their frequency, but then he tries to redeem the situation by suggesting a di-thematic solution, a solution which was rejected at the time of publication and which has never undergone serious discussion since that time. If his paper is going to now be quoted as valid, that discussion must first be had.

Forsyth only asserts that the symbols should be names, but provides no supporting evidence.

Lee et al.’s paper6 claimed to have used maths to prove that the symbols are language (not names). The authors’ mathematical and linguistic methods and conclusions were immediately and roundly rejected. Under these circumstances, it cannot continue to be referred to with the unstated implication that it supports any argument.

Handley7 notes that there are only a few localised groups of Christian epitaphs in west and north Britain. He argues that because elsewhere in western Europe the pattern of stone memorials waned in the early 3^rd and 4^th centuries then increased in the late 4^th and 5^th centuries, we should presume the same happened in Britain. However, this work is today considered to be too early for small gravestones at known monasteries which seem mostly to be from the late 7^th or 8th century. As Edwards8 points out, stone memorials have the same problem with dating as symbol stones, with only two dated with any certainty to c625 and c824, but again this is now out of date, and we know from contextual dating of monasteries that the memorial stones have a date range 650-800 AD.

In other words, none of these references quoted by Gondek & Noble in any way begin to provide evidence for the symbols as names. And the consequent assertion that these assumed personal names are then on a memorial stone is a non sequitur.

Dating the symbols

Recently (2026), Prof Gordon Noble has written another paper9 on the subject, in which he continues to assert that the symbols are names, although I notice he has now softened this idea to ‘names or identities’.

Current consensus is that this was a system that expressed names or identities of some kind and that this was an elite form of expression found in both settlement and burial contexts.

But there is no ‘consensus’, as there never has been any evidence raised to support this idea of names.

And the only person who has written about the issue is myself, arguing for the last ten years that the symbols are easily proven to not be names.

Under Noble’s scheme, it is no longer being claimed that the Pictish symbols were created for names on Christian memorials in the 7^th century. Now it is being claimed that the symbols were influenced by Roman inscriptions, created in the wake of the Roman invasion of Britain. So, the next step is to recap what is known about the dating of the symbol system.

Over the last few years, archaeology such as done by the REAP team has gone a good way to point to much earlier dates for CI stones. Rhynie, with its group of symbol stones, an elite palisade enclosure and a major settlement on the nearby Tap O’Noth, demonstrates occupation spanning the 3rd to 6th centuries AD. The sea stack of Dunnicaer with its symbols has produced finds from the 2nd to late 4th centuries. Silver chains, a couple with mature CII forms of symbols on their clasp, are to be dated to the late 300s or mid 400s when large amounts of silver were incoming. The Sliding Cave at Wemyss in Fife with Pictish symbols has been dated to 240-400 AD.

These archaeology dates for symbols are the earliest so far attested, they are wonderful work by the Aberdeen team. Prof Noble prefers to regard the simple symbols on cave walls or the sea stack as the earliest symbols undergoing creation, rather than the alternative, that they are informal symbols in ritual spaces just like simple crosses or scrawled modern names often found in such places. Although the symbols are simple, they are already fully formed, and they can occur in pairs, following the established rules of ordering and pairing. This aspect of the cave symbols is critical to our understanding, as Prof Forsyth pointed out and I agree, it means the symbols had undergone a significant period of development prior to their carving. Which means that the creation of the symbol system likely reaches back into the late centuries BC, long before the Romans arrived.

Symbols start to be carved on stones across the land, as far as we can tell, in the 2^nd or 3^rd century, and like in Gaul and southern Britain, the practice of carving on stone was doubtless influenced by the Romans. The Romans did at times raise memorial stones to certain people, including an inscription.

But, symbols are not the same as a symbol stone, although the two are often conflated. Symbol stones have their own particular function, which the symbols are enabling. Take for example a modern street sign. It has a function of guiding people to places. The alphabet used on the street sign was not created for the street sign, it has its own history of development. Symbols are to a symbol stone as the alphabet is to a modern street sign.

The issue here is that the symbols appear on the Pictish stones fully formed, a mature structured closed set of symbols, acting in pairs according to strict rules, which implies a considerable development time for the symbols themselves, before being carved on symbol stones. The symbol system is much older again than the symbol stone practice influenced by the Romans.

There is no reason then to presume the Picts were using the symbols to represent names, as the symbols predate the Roman influence of carving names on stones.

There is a proviso however. We might reach for the argument that the Picts were the first people, before the Romans, before the Christian churches, to come up with the idea of creating a system of symbols representing names. It’s a one in a million chance, but in history weird things do happen. So, I’ll keep exploring the topic to make sure we have a solid conclusion.

The symbols and inscriptions on name stones at Christian monasteries

In the late 600s, the Christian church begins a new practice, in which burials were gathered into the space around the church, and along with this, small memorial stones begin to be carved, some carrying a personal name.

Southern Britain has 31 name stones, at five large monasteries. Lindisfarne has most with 23, mainly dating to 750 - 800 AD with a possibility of a few decades slightly earlier or later. Hartepool and Billingham have 5 name stones, dating to 650 – 750 AD.

In contrast, Pictland’s gravestones only have one Latin inscription at Dull. Pictland does have at least 80 small gravestones, with around 60 at St Andrews alone, and the rest spread at notable church sites. All these stones are around one metre high, and none carry an inscription. If symbols represent personal names, then we might expect the symbols to come into their own at this point. In fact, the reverse is the case, this is the moment when the symbols disappear.

Pictland does take up the practice of inscriptions, in ogham and Latin, but all inscriptions (except Dull) are on medium to large stones, not small personal burial stones. Although Pictish ogham is infamously difficult to interpret, a few do contain a form of mac as part of a name, but the bulk of inscriptions on large stones appear to refer to saint Ethernan. So, while the Picts along with everyone else are beginning a practice of inscriptions on their stones, they aren’t using them to put a name on a personal memorial stone, but using the inscriptions for a grander purpose.

Symbols and inscriptions occur together on large monumental stones at Largo, Scoonie, St Vigeans, Logierait, Brodie, Dyce, and Fordoun. However, there is no correlation between the inscription and the symbols, they are not parallel texts, they are not presenting the same information in different form.

If symbols were names, we’d expect these gravestones to continue the practice of putting symbols as personal names on stones. The fact that this doesn’t occur points to the symbols not being personal names. It further points to the symbols not being associated with a particular human at all, which also rules out personal ranks or status identifiers.

Symbols as ranks

Is it possible that some symbols represent names, and others represent ranks, or signifiers of office?

As a test we could try out the Pictish equivalent of king/queen/prince, so symbol pairs would be in the form of say, King Nectan, Queen Vanora, Prince Talorc.We could at the same time add regional affiliations, such as King of-Pictland, Prince of-Fife.The first thing to note is that these rank+name pairs have an internal linguistic structure, rank (first symbol) + name(second symbol). So, the rank symbols should occur either always as first or always as second symbol, depending on Pictish syntax. This is not the pattern we observe with symbol pairs. The table below shows the numbers of pairs where the symbols in question occur in the top or bottom position. Only those symbol pairs have been counted which can clearly be confirmed to have no mirror+/-comb to avoid unnecessary complications in the data. A trend is visible: the crescent/V tends to occur as the top symbol, while the beast and the double-disc/Z tend to occur as the lower symbol. I have included the other two rodded symbols, to test the case of the rod denoting a rank, but there is not enough information to provide a clear picture for those symbols.

What proves definitively that neither these high frequency symbols, nor the group of rodded symbols, are ranks, is that they tend to occur together as a pair. Crescent/V + double-disc/Z would then read for our test case, kingly-queen, or something similar. Crescent/V + Pictish beast would read kingly-prince, or something similar. Pictish beast + double-disc/Z would read princely-queen, or similar. Which makes no sense of the data, and demonstrates that the symbols do not signify a rank.

That said, there is something odd that happens near the end of the CII period as the internal structure and layout of the CI symbols decay, where certain CII symbols seem to attract certain motifs. As yet I can’t totally rule out the possibility that some of these symbols may be being re-purposed to denote rank in the dying days of the CII Christian period.

Di-thematic names as a symbol pair

Ross Samson presented a proposal which deserves more close analysis. He suggested that as many Celtic names are di-thematic, then each symbol may represent a ‘theme’, making the symbol pair the full and entire di-thematic name (with no patronymic element). For example, the name Brandub would be represented by a symbol for bran ‘raven’ and a second for dub ‘black’. The idea is very seductive, as it might explain the imperative for symbols to be depicted in a pair.

The main reasons why Samson’s idea of di-thematic names was rejected at the time of publication was that many Celtic names aren’t di-thematic. For example, Talorc (tal ‘forehead’ + orc ‘pig’) is di-thematic, but Talorcan has a diminutive suffix –an giving it three elements. Fergus/Forcus (fer ‘man’ + gus ‘virility’) has two elements, but Drust and Brude only have one element, although Drustan has -an which is a diminutive suffix or a reflex of ancient –genos. The structure of real names is demonstrating that the proposal of di-thematic names has inherent difficulties.

Symbols in pairs – or not

The first way of looking at the proposal of di-thematic names, is the question of whether symbols really do always come in pairs. If personal names aren’t necessarily di-thematic, then do symbols sometimes come in other combinations?

Currently, we have found just over 200 stones with a symbol pair.

There are a couple of stones that on first appearance only have one symbol, Dun Echt and Kintradwell 3, both with one symbol plus a mirror+comb. However, the Dun Echt stone is very eroded on the top half, and the Kintradwell stone is red sandstone found in a marsh and badly eroded with ‘much edge damage’. The presumption must be that the top symbol in both cases has just eroded away.

Three Class II cross slabs have a symbol alone without its pair, Dunfallandy (26295), Fordoun (36458) and Kirriemuir 2 (32300). In the same vein, a couple of stones also have a mirror+comb without any symbol pair. At the same time, other rules, such as the order of the symbol pair are also breaking down. These changes are all happening right at the end of the symbol tradition in the late 700s, after maybe 7 to 10 centuries of the symbols being depicted according to strict rules. So, it would be wrong to treat these degrading instances as ‘normal’ data.

There is only one Class I symbol stone with three symbols, that being Knockando 1 with a spoked wheel over 2 crescent/V symbols. One of those is very faint, as if it isn’t part of the original pair. On CII stones, only St Madoes holds three symbols, but, this stone too comes from the same very late CII period when the rules are breaking down.

All other symbol stones have either a symbol pair, or a multiple of symbol pairs. All Pictish names are not di-thematic, yet all Pictish symbols come in a pair, except for a few degrading instances at the end of the tradition. Symbols and names do not match.

Combining symbols in pairs

A second way of looking into the di-thematic proposal is to look at the frequency of symbols compared to the frequency of name ‘themes’.

Let’s take the name Brandub. Breaking this name into two ‘themes’ gives not two personal names, but two common words: bran ‘raven’ and dub ‘black’. And this is the case for most di-thematic Celtic personal names. So, could symbols simply represent words? A natural language usually contains up to 300,000 common words, with about 30,000 in everyday use, and, as Forsyth once pointed out, the body of about 60 symbols is way too tiny to cover everyday words in a language.

Recall that the three most frequent symbols account for roughly 50% of all symbols. Looking at the case (see table below) where all themes/symbols would be free to combine at random, excluding the case where both symbols in a pair are the same symbol, then 61% of CI stones and 67% of CII stones would contain one or two of the three most frequent themes. This is too highly skewed to represent a body of names.

While looking at how these top three symbols actually combine, something else of interest shows up: they aren’t pairing randomly at all. They actually like to combine together in a pair, to be more precise, the Pictish beast and the double-disc/Z symbols prefer to combine with the crescent/V.

Inspecting all the symbol pairs on stones, CII symbol pairs include at least one of only 7 themes/symbols, a very small number! Those are the crescent/V, Pictish beast, double-disc/Z, double-disc, serpent/Z, cauldron, and salmon. If we treat the rodded and non-rodded symbols as versions of the same symbol, this collapses to just 6. (This leaves out the Ulbster and Golspie stones as its difficult to be certain if each symbol has a specific pair in the series.) To put it another way, every CII symbol pair contains at least one, sometimes two, of only 7 symbols, a very narrow set. If these 7 symbols are one part of a di-thematic name, then all Pictish names would have to hold one of only 7 themes/words, an implausible situation.

The same spread in the CI symbol pairs is wider than in the CII pairs, as expected from the flatter frequency distribution of all symbols, but still gives all symbol pairs containing at least one of only 15 themes/symbols, and this coverage is again heavily skewed towards the same high frequency symbols.

This usage of a tiny set of symbols at the core of all symbol pairs demonstrates that it is implausible that the symbols could represent personal names, neither as a full name for each symbol, nor as two symbols of a di-thematic name.

The structure of a symbol pair

Next, we can look at how the symbols function, and what the rules that govern them are, to see if the idea of personal names might fit.

On late CII stones, we find the mirror+comb appearing on stones by themselves, possibly referring to a female burial. But, does this mirror+comb always refer to a female? Let’s assume so for the moment, and see how that might work with symbols if they were personal names.

A mirror+/-comb often modifies the main symbol pair on a symbol stone. Around a fifth of all symbol stones carry a mirror+/-comb, so we should have a reasonable pool of male/female names if that’s what the symbols represent.

If the mirror+comb was marking a female name, then we might expect the accompanied symbols to come from a ‘female’ set of symbols, but, there is no obvious differentiation into male and female symbols at all.

Next, the same symbol pair can optionally carry a mirror+comb (see an example above of the crescent/V and Pictish beast). Let’s say just for example that this pair represents Bran-dub, a male name, then Bran-dub(mirror+comb) is somehow a female name just because a mirror+comb is added. Which is not the way names work.

Celtic names do sometimes have the same male or female form. But not always, and the few Pictish female names we know of are not the same as male names.

In another blog, I discussed how the mirror+comb moves the natural precedence of the top symbol to the bottom symbol. In this case BRAN-dub is male, and bran-DUB(mirror+comb) is female. Again, this is not following the patterns of a Celtic language.

A single mirror without a comb means that the natural order of the symbol pair has been reversed. In this case, if Bran-dub is male, then Dub-bran(mirror) is female. This makes nonsense of the entire symbol set.

Over and again, it feels like every time I discover some new aspect of the symbols, they are again demonstrating that they cannot be personal names.

Another interpretation of a symbol as a name

At this point, I have demonstrated that the symbol stones do not hold names, neither as a full name, nor as a di-thematic pair. I am now, somewhat perversely, going to suggest that a symbol may be a name under certain circumstances.

The druidic institution relied on many years of oral learning, a highly complex memory feat, and to do this, they would have employed a sort of memory system to reflect the structure of their worldview.

We see traces of this memory system within the organisation of the Welsh triads, and also within such things as Irish ogham kennings, and associations held within stories and art. For example, Esus is associated with an axe, a willow, white, death, S, a bull and cranes – all items connected within an integrated environmental picture. Bran, the raven, the guardian, is recognised by his association with the alder, shields (made from alder), protection, purple and F. Along the same lines, Nectan might be associated with water/well/wisdom/salmon/hazel/C, being the ancient deity who ‘owns’ the well at the centre of all rivers in which swims the salmon and above which are the hazel trees that drop their nuts into the water to give wisdom to the salmon and to anyone who ate the salmon in turn. In other words, one symbol could represent a complex picture with layers of interrelated meanings at many levels - trees, weapons, birds, animals, just to mention a few, and also on one level, names of deities.

Take for the sake of argument, the stone at Aberlemno, considered to represent the Battle of Dunnichen. If the Picts, say, wanted to place Dunnichen (Dun of-Nechtan) on the stone as symbols, then they would in this case need the symbol which at some level refers to Nechtan/salmon/hazel/C, plus a second for ‘dun’. A particular symbol could then be called on to represent the name Nechtan, but only if or when it is required.

And, notice that once more we are dealing with typical Celtic di-thematic names, only this time not personal names, but far more consistently, placenames. Which is where my research has also drawn me, that each symbol stone is referring to items of significance in the vicinity, which is also how placenames work.

And, if or when required, a symbol pair might display a di-thematic personal name like Brandub ‘Black Raven’, because both these elements or themes would be included in the coverage of their respective symbol. This means that while the maths and social context tell us that the great majority of symbol stones do not depict a personal name, it is just possible that a few symbol stones may well do so if needed.

A friend once commented that the symbols are as easy as counting the fingers on your hands, and, as complicated as the entire universe, because that is exactly what they represent, simplicity and structure within the ultimate complexity. A symbol may be a name if required in certain circumstances. But hard cold maths tells us that, as a general rule, the pair of symbols on a symbol stone does not, cannot, just represent a personal name.

This story has so many intiguing twists and turns, and clear-cut answers are difficult to come by.

I have to thank the members of the Pictish Symbols: Art and Context group for their input and interest in this research, I couldn’t have done this without their help. Special thanks to Fiona Campbell-Howes and Linda Sutherland who valiantly put me right on several scores. All errors and opinions mine alone of course.

The crosses of Eigg

This story for me started with a very odd distribution map – of the cross with a square centre and a ring connecting the four arms. This type of cross is peculiarly ‘Pictish’, in that it is rarely found outside Pictland.

But the map doesn’t just say ‘Pictish’, it shows a distinct focus in the southern region of Angus, Gowrie, and Fife. Which makes the two exceptional locations really stand out. First there is a small group at Loth in the north, which suggests we have here another church or monastery that we’ve lost sight of – like we did with the major monastery of Portmahomack. But that is another story for another day.

The second exception is on the western isle of Eigg.

It’s not that there aren’t other types of crosses in the western islands, there are a few, but they’re not of this sort. So who went all the way across Scotland to carve this Pictish cross, almost as a sign that Eigg belonged to the churches of south-east Pictland?

There are two crosses on Eigg with the square centre and ring. One is a simple cross inside a circle, likely an early gravemarker, you can see it on page 40 of Fisher’s ‘Early Medieval Sculpture in the West Highlands and Islands’.

The second has a cross filled with interlace and a panel of complex key pattern at its base – also typical of Pictish sculpture. The other side has icons drawn in the typical Pictish style – a horseman and his hounds with a tiny bird, following a group of animals, including a lion. The other animals might be a boar, a horse, and a bear, but are too eroded to positively identify.

These crosses come from a relatively short-lived monastery established at Kildonnan on Eigg in the early 700s. Oan, church head of Eigg, is recorded as dying in AD 725, and Cummene in 752. For once, we have a nice date range of the early 700s for a Pictish stone.

Eigg lay on what was likely the southern boundary of the territory of Cenel nGartnait, a group with Pictish elite names. The region also has pagan symbol stones, square barrows, and traces of the Pictish language in placenames Applecross and Ob Apoldoire with a form of Pictish aber or apur. All these factors make it clear that this area centered on Skye was self-identifying as ‘Pictish’.

But, we’re still left with the question of why the interest in this small rocky island of Eigg? And why this particular form of cross that appears to connect Eigg to south-east Pictland?

The martyrdom of Donnan and his companions

The one significant event that is recorded for Eigg is the death of Donnan and his companions, as first noted in the Annals of Ulster 617.1, about a century before the later monastery and crosses.

Combustio martirum Ega,—combustio Donnain Ega h-i .xu. kl. Mai cum .cl. martiribus; & occisio Torche, & loscadh Condiri.

“The burning of the martyrs of Eigg - the burning of Donnán of Eigg on the 15th of the kalends of May 17 April with 150 martyrs; and the ravaging of Tory Island, and the burning of Connor.”

This is a rare case of an early Christian saint being killed, although the annals do not tell us who was the aggressor. This entry says he was burnt, another later one says he was beheaded. One oral tradition relates that he was killed by a Pictish queen who was mad at him for stealing her sheep. Another that he and his companions were slaughtered by sea-raiders after saying mass at Easter. Whatever the details, this slaying was rare in the insular Celtic church, making Donnan and his companions important proto-martyrs.

Was this event considered of such importance that nearly a century later the Picts were drawn to establish a monastery on Eigg? Does this event lie behind why a Pictish carver came to Eigg to put up such a special cross?

There is only one thing to do, we have to look at Donnan to see what his connection might be to the Picts, and to Eigg.

Donnan’s origin

Donnan himself has no genealogy and no claims about his origin. This isn’t unique, but it certainly is rare, especially when dealing with such a well-known saint. It immediately raises a red flag.

His dedications are mainly in western Scotland. He has no dedications in Ireland, which above all seems to point to Donnan, as the first red proto-martyr, having no significance for the Irish, which is unexpected had he been an Irishman or a student at one of their monasteries.

It has been suggested by Rekdal, and taken up by other scholars, that these western dedications are found in areas of later Norse settlement, and perhaps are a form of ‘expiration cult’, even though Donnan’s death was far too early for the Norse to be the culprit. However, if these really are Norse dedications made several centuries after Donnan’s death, then we face the quandary of whether any earlier Celtic dedications to Donnan exist at all. There seems no way to be sure that there ever were.

And, if we take these thoughts to their conclusion, no genealogy, no origin, perhaps no early dedications, could this hint that the martyrdom of Donnan himself may not have been ‘real’? For the moment, I’m going to assume he was real, and these are just glitches.

Donnan’s name and the Cenel nGabrain

Although we don’t know where Donnan came from, we might have a clue in his name. The name Donnan is a hypocoristic form of Donn, a fairly common name in earlier pagan times, but quite rare in later Christian times. It is found however in some names of the Cenel nGabrain ruling class, and because certain names concentrate in particular elite families, this may give us a clue to Donnan’s origins.

Two kings of Dál Riata have the name Donn, Domnall Donn d.696 and his brother Máel Dúin mac Conaill d. 688. Their father had been co-ruler with Dúnchad mac Conaing d.654. In one genealogical tract Domnall Donn is named as Eochaid Buide’s son, Eochaid (608 – 629) being the king at the time of Donnan’s death in 617, although this is perhaps a mistaken parentage for Domnall Donn d.696. The use of the name Donn is suggestive that Donnan was part of this family, and that maybe the two brothers Domnall Donn and Máel Dúin were named for him. Or, at a stretch, that he was a third brother.

If this was the case, it might be proposed that Donnan, as a member of Cenel nGabrain, was intruding into the Cenel nGartnait territory of Eigg, and was killed for it.

Donnan being a member of Cenel nGabrain could explain why his dedications are only found in western Scotland, not Ireland. It brings to mind the traditional story of the relationship between Donnan and Columba, with Donnan asking Columba to be his soul friend, but Columba refusing and warning Donnan not to go to Eigg, a sacred island of the Picts. It also may answer why his martyrdom was of particular interest to the Ionan community, for he may have been viewed as one of their own Dal Riatan kin.

The number of Donnan’s companions

The Annals of Ulster (AU) gives the number of Donnan’s companions as 150 – a typical number of disciples claimed for important saints and druids. But this number is likely a later addition to the entry, which of itself is retrospective to a period before the AU began recording events by at least half a century or more.

Does this bring the earliest entry about Donnan in AU into disrepute? Perhaps, but the number part of the entry may be a later addition, so the basic record of his death may be still valid.

The Martyrology of Tallaght numbers his companions as 52, but even this reduced number is suspect as being unrealistically large.

For comparison, Columba is said to have founded his church on Iona with 12 companions. In AD 806, at a time when the annals were contemporary, 68 monks of Iona were slaughter by Vikings, although we’re not told how many, if anyone, survived this attack. So the famous Ionan monastery was relatively small, perhaps with only a few dozen resident monks, and we shouldn’t expect Donnan’s group to be any larger.

Eigg has an area of 11.8 sq mi. and its population today is around 100, however its inhabitants are not reliant on self-sufficient farm settlements. Prior to the clearances, Eigg’s population peaked at around 500. Kildonnan, above the eastern port, is the location of the 700s monastery and the presumed scene of Donnan’s death, and it has a limited fertile area around it. It’s doubtful Eigg would ever have been able to reliably sustain a founding group of 52, little alone 150.

The archaeology carried out in 2012 by Professor John Hunter found good evidence for the monastery of the early 700s. This monastery was substantial in its day, with a number of its clerics noted in the Irish annals. The reconstruction shows 7 houses which, if it was of similar to Donnan’s earlier group, would require at least 7-8 people per house, even assuming all houses were dormitories. The claim that Donnan had 52 companions really seems far too many to be realistically valid.

So why do we have this list of 52 companions? A list which goes so far as naming those companions.

The names of his companions

The names of Donnan’s 52 companions are thought to be Irish in the main, although actually they were recorded in Latin, but can be readily transferred into Irish, which isn’t quite the same thing. As Alex Woolf said, the ‘annalistic and hagiographical texts make it quite clear that Pictish names might easily be Gaelicized’.

If Donnan was of Cenel nGabrain origin, then we might expect most of his companions to also be from this same kindred with Gaelic names, which may go some way to explain why the names appear ‘Irish’.

Prof Clancy calls this roll-call of Donnan’s companions ‘extraordinary’:

In itself this is extraordinary. Although these are “bare names”, the list deserves to be better known, particularly on Eigg. From no other monastery in Scotland do we have such a substantial roll-call of its inhabitants at one moment in time.

It certainly is ‘extraordinary’. It’s extraordinary in its very existence, extraordinary in its exaggerated number, extraordinary in its recorded names. Perhaps if it sounds too good to be true, chances are it isn’t.

The validity of this list of names also depends on the dating of the entries in the Martyrology of Tallaght (MT). Pádraig Ó Riain has argued that the Martyrology of Tallaght was an augmentation of an older martyrology that originated on Lindisfarne c. 700 and went to Tallaght via Iona. In his estimation, it was kept on Iona between during the period 715 – 740, a century after the martrydom.

This dating provides us with an alternative solution. The list, rather than being Donnan’s martyred companions, could be the Liber Vitae of the later Eigg monastery. There is another comparable list of 223 names attached to Fintan of Tech Munnu’s entry in the Martyrology of Tallaght which could be the Tech Munnu Liber Vitae, i.e. the list of people who had donated to the community, or went there on pilgrimage, and who wanted to be remembered by the community at the site. The proposed dating of 715 – 740 when the MT was on Iona coincides with the dating of the later Eigg monastery in the early 700s. The list of names may be so clear and detailed because it was contemporary, but represents the current Liber Vitae rather than his martyred companions.

Broadening the view

So far, I’ve been presuming that Donnan was real, and was indeed killed at his church on Eigg. And this is the only possibility considered by scholars today. To that end, I’ve been testing the validity of the earliest records in the AU and MT. These entries are retrospective by at least a century, and the details of his companions patently irrelevant, the number unrealistic. His dedications tell us only that he was culted in the Norse regions of western Scotland. So far then, belief rests on the suspect validity of a retrospective entry, and a possible connection via his name to the Cenel nGabrain of Dal Riata. It’s really not a lot to hang our hats on.

Is it possible that the story of Donnan is a Christian myth, a story serving another purpose? Let’s now keep going with a look at the broader context for Donnan.

The date of the event – a day of magic ritual

One issue to consider is the very precise date 17 APR 617, or as the annal entry has it, the 15th of the calends of May. I did have a look at the skies at that date, and it turns out to be the date of a minor lunar standstill, which only happens every 18.6 years. Which could be why the year 617 was retrospectively chosen as special.

But, the date has another really key significance. On the Coligny calendar, there are three days that are marked with TIOCOBRIXTIO, the most special ‘magical’ sacred days of the year. These days weren’t the normal popular gatherings of Beltane or Lugnasad, but seem to have been moments where the druids and elite gathered for special rituals, as described for the Feis of Tara. One magic day is the full moon of the winter solstice, the second the full moon of the spring equinox, and the third day marks the rising of the sun in the beautiful star cluster, the Pleiades. Unfortunately because these were ‘druidic’ days their ritual significance was lost sight of very early on, but their very importance meant they were all sained by major Christian festivals, Christmas, Easter, and it seems, this day of Donnan.

At the date of 17 APR 617, the stars have progressed a little in the sky so that the sun no long rises with the Pleiades but with its entry into the constellation Taurus.

The western church also created a special feastday of ‘all European saints’ on the 20th of April. Hennig suggested that this European saints day was based on the date of Donnan’s martyrdom, but it seems more likely that the reverse is true, that it was a means of saining the third most sacred day in the Celtic calendar.

So Donnan’s feastday seems to be saining one of the most sacred days of the calendar. Does this mean he wasn’t real? On one hand, we could think his story is mythological and he was given this feastday because it was special. Or, we could think that he is real, and like St Patrick, he did something to interfere with a sacred pagan ritual, only this time it got Donnan killed. It makes the story even richer in detail, but still doesn’t quite give us a definite answer as to whether he is real or mythical.

The date of the event – the link to Donnotaurus

There is another astronomical aspect to consider. At this time in April, the sun is moving into the constellation we today call Taurus, which has the Pleiades inside it. It is likely that the Celts knew the constellation as Donnotaurus, ‘Lord Bull’. We find a Gaulish personal name Gaius Valerius Donnotaurus, and the equivalent in Ireland is the famous Donn Cúailnge, often translated as the Brown Bull of Ulster. This is the famous bull that Maeve wished to own, and so set in motion the cattle raid, the Táin Bó Cúailnge. The same name and date, for both martyr and constellation, may not be coincidence.

Donn, guide of the dead

Donnan and Donn are fairly common names in earlier pagan stories, but they’re rare in later Christian times, for good reason. Donn is a deity whose house the spirits of the dead go through to reach the afterlife, Tech Duinn, the House of Donn. In Ireland this is localised at Bull Rock, a small rocky island with an arch off the west coast. Donn is then a form of Celtic psychopomp, a guardian and guide of souls.

Many of the old deities were fairly seamlessly transferred into hero-ancestors under the new Christianity, but certain deities had a function that was antagonistic and not easily reconstituted. The deity Donn found himself in this situation. The pagan dead travelled to the west with the guidance of Donn and their spirits were eventually reborn, but death for the Christian meant burial until the Day of Judgement when the soul would go to heaven. Under this new scheme, Donn no longer had a place. Worse, he stood in opposition to a fundamental Christian belief.

Such a major deity could not easily be eradicated from cultural memory, so he eventually became a form of the devil. The Metrical Dindshenchas relates:

… the souls of sinners visit Tech Duinn before they go to hell, and give their blessing, ere they go, to the soul of Donn. But as for the righteous soul of a penitent, it beholds the place from afar, and is not borne astray.

Just as Bull Rock off Ireland’s west is a localisation of Tech Duinn, Eigg may have been considered the localisation for Cenel nGartnait. Not only is Eigg a sacred isle in the west, it also has several caves around the coastline, and today at least one natural arch, making it a good candidate for a localised Tech Duinn.

Eigg was definitely a key place of burial for the Pictish elite. It has a wonderfully large group of at least 16 Pictish square barrows laid out in a regular line along the western bay of Laig, at the opposite side of Eigg to the church of Kildonnan. This is the only square cairn cemetery in western Scotland, with a couple of isolated examples on Uist. Pictish square cairns elsewhere have been dated prior to around AD 600, making the Eigg cairns near contemporary with Donnan’s period.

The importance of Eigg as the site for Pictish elite burials cannot be overstated. It places Eigg in its time at least on a par if not more important than Iona itself.

Recent archaeology at the site of Kildonnan itself also showed it to have a long tradition as a burial site (John Hunter, The Small Isles, p137-138).

The results were astonishing in demonstrating that the site was of long-standing burial tradition starting in Neolithic times, thus by the time of Donnan’s arrival at the end of the 6th century the area had already been a ‘special place’ for some four millennia. This is almost certainly the reason why it was chosen for monastic activity.

Perhaps the most extraordinary ritual space is found at Struidh. Here a ritual platform is surrounded by a massively thick wall, and the inside leads you to an underground cavern. The cliff here is horizontal banded rock, but the platform below the cliff faces an extraordinary uprising of vertical larva tubes. Unfortunately, there is no dating available although there is a suggestion that the platform and cavern is Neolithic, but once more we have good reason to view Eigg as an entry to the otherworld, perhaps of long-standing.

Here then, we have a possible explanation for the sacred aspect of Eigg, as the Pictish localisation of Tech Duinn, the island in the west from which the spirits of the dead depart for the afterlife to be reborn. An exceedingly sacred place in the pagan worldview.

This connects Donnan as saint to his namesake Donn as deity of the dead. But, does it tell us that Donnan was a real Christian martyr who came to grief on a sacred island, perhaps attracted there because of his namesake?

Or does it tell us that Donnan is perhaps a Christian retelling of an ancient story of Donn and his dead? It is but a small step from the deity, Donn, gathering the dead souls around him, to a new version of the ancient story, Donnan, the saint surrounded by his dead.

Sunniva and the bones of the dead

There is a story parallel in some ways to Donnan’s. Sunniva was an Irish princess who lived on a sacred island off the coast of Norway, but she and her companions were killed by aggressive pagans from the mainland opposite. Rekdal wrote that the two stories are so similar they may be a result of cultural exchange.

Sunniva is supposed to be an Irish saint, although neither she nor her companions have any genealogy. Sunniva is a Germanic/Norse name meaning ‘sun gift’, interesting in itself, as we have Sunna as a form of Norse sun goddess. And we have stories from Roman times onwards about groups of women living on sacred islands. So Sunniva also has deep roots in Norse lore.

But what is fascinating about Sunniva, is that her story was triggered by people finding a mass of bones on her island. From there her story spread to become an important Nordic proto-martyr cult.

What if Donnan’s story too was triggered by people discovering a pile of cremated bones, possibly in one of the caves around the shoreline of Eigg? We already know of the elite Pictish burials on Eigg, but Eigg could well have been a place where other people brought their cremated dead, to the House of Donn, for rebirth. Finding bones on Eigg is to be expected.

There is of course another case of bones being found in a cave on Eigg. This is the story that all 400 people of Eigg were hiding in a cave from the Macleods, but they were discovered and the entrance to the cave set alight so that they were all smothered by smoke.

Massacre Cave has a tiny 1 metre high entrance, originally behind a waterfall, a 7 metre crawl space leading to a very large, high-roofed cavern about 70 metres wide and 6 metres high. It is almost a given that this exceptional cave has long been a ritual space.

Some time back, the local people cleared out the cave and reburied the bones, but recently, some more bones were uncovered and found to be of a teenager living between 1430 and 1620, which appears to support the tale of the massacre. Sometimes coincidences are just that, other times it raises stressful questions – and suspicions.

As Prof Hunter said, the small island of Eigg has a “dramatic historical background and legacy of wild superstition”.

Donn and Dal Riata

So far I’ve been focusing on Donn in his aspect of guardian of the dead. And we’ve seen how Donn’s role led to him being ostracised from the stories under Christianity, even to the point he becomes a form of the devil, leading souls to hell.

Why then would we find at least two historic kings of Cenel nGabrain using the name Donn as part of their own, Domnall Donn and Mael Duinn ? The answer is that a deity never had just one aspect, a deity was a complex being, with many functions and roles.

Donn had a specific connection to Dal Riata. Apart from the historical kings, we also find him as Donn Mor of Monad, son of Ruadh of the coastland of Alba who was slain in the battle of Cronnmhoin by the great Fianna warrior and sometime rival of Fionn mac Cumhaill, Goll mac Morna. (Watson CPNS p392 and Duanaire Finn)

Monad, according to Watson was a term used to refer to the kingdom of Dal Riata.

Donn is again part of the Fianna in the Silva Gadelica i.145 where he is said to have been the chief of the Fianna of Eire and Alba for 27 years, ‘till Dubh son of Dolar slew him in the battle of Cuire over in Alba’.

These stories of Donn as king and leader of the Fianna in Monad, trace the importance of Donn in myth and history to the Dal Riatan kingdom. It explains why their kings might continue the name, and why Donnan himself may have been a member of this group. It also hints that Donnan may have felt a particular attraction to Eigg, perhaps he felt it his role to Christianise the island of his namesake, despite Columba warning him off.

Donn appears in other stories of course, but there is one story I will finish on, as it seems to follow a familiar theme. This is the story of ‘the enchanted stag’ from Duanaire Finn (p108). Donn, of divine race, carried off a hundred maidens from the house of Aed. Aed’s wife loved Donn and in her jealous rage she turned the maidens into a herd of deer, and Donn himself into a mighty stag. Donn then defied Fionn to hunt him (again note the link here to the Fianna). Finn refused, but gave his hunting dogs to Oisin, who finally killed the great stag with his herd of deer.

Once more we have Donn, killed by the Fianna, the god of the dead surrounded by his companions.

Making up my mind …

Does all this mean Donnan is another telling of the same story of the great deity Donn? No, not necessarily, people are always complicated and do amazing things, and have really strange things happen to them. On the other hand, quite possibly. But, at the very least, there is a deep crossover here between Donnan and Donn, playing out in a way which is intriguing but hard to make out clearly. Perhaps in the end it doesn’t matter, perhaps we need to accept that all the stories, Donnan’s included, are part of the same Donn/Donnan rich tapestry, something to be celebrated.

That said, I still haven’t answered why there is a particular form of cross on Eigg which seems to connect the later monastery of Eigg to the church of south-east Pictland. Certainly there is a Pictish connection to the Cenel nGartnait, but why south-east Pictland? well I’d hate to have no mystery to think about…

From the very start, this finding was extraordinary, at least some of us thought so. Because the place where it was found was extraordinary – beside a narrow passage between lochs and marshes, the only pathway through the marshes for some distance, at the centre of Pictish activity with forts, nemeton, islands, and symbol stones, and, at the foot of the Dun of Nechtan with its sacred wells, at the head of the river Lunan with its salmon.

I went to visit this place afterwards. There was nothing to see but a small quarry in the middle of a scrubby area drained of its original lochs and marshes. Somewhat underwhelming.

But with such an important place, the story had to keep going, and recently it has, with biometric research into those buried there (love the science!).

I’m going to suggest that far from just another group of burials, the elite burials themselves, and the importance of the location, point to it being a key ritual druidic burial site.

And yes, there you go, I did say the dreaded ‘D’ word. For those of you who have resisted the urge to run for the hills, who aren’t afraid of the past and human endeavour, welcome to this blog.

Restenneth ‘the point of fire’

The cist burial site is near Restenneth ‘the point of fire’.

In Ireland, the word tinne ‘fire’ is found in a number of placenames, but in Scotland it is exceedingly rare. This alone points to a unique place, a place of ritual fires, a place of key significance.

Restenneth isn’t up on a hill as we think of for say Beltane fires, it is down on the flat, near a narrow passage through the lakes and marshes in this area.

The passage through the marshes

This ancient narrow passage near Lochhead quarry lies in the middle of a very very soggy area between marshes and shallow lakes. Today, most of these marshes have been largely or completely drained, but in the past this whole area was very wet and difficult to move around in. This narrow gap allowed movement between the areas of Forfar to the west and Aberlemno to the east.

In AD 685, the Northumbrian king Ecgbert and his army were ambushed and killed at Dun Nechtain, considered to be the same named Dunnichen, the hill beside this passage. It is likely that the Restenneth passage played a key role in this event. If Ecgbert’s army had just passed through this passage making for the Pictish centre of Turin and Aberlemno, he would have found himself in the extensive marshes between two hills, Dunnichen and Turin Hill, and the Picts could have easily blocked his retreat back through this narrow passage, trapping him to his doom.

The battle of Dun Nechtain is often said, then and now, to have been a pivotal point in both Pictish and Northumbrian history. Some 40 years later, a stone church may have been raised at Restenneth with a dedication to St Peter. The king of that time is reported as inviting Northumbrian clerics and stonemasons in to build churches in the new fashion of stone, and at the same time he promised to dedicate Pictland to Peter. The implication is that this church at Restenneth was part of that scheme, given to Northumbria precisely because it was near the spot where their king had been killed.

The Sidh of Nechtan

In Irish mythology, the Sidh of Nechtan is the dwelling place of Nechtan, the ancient god of waters. At his sidh, there is a sacred well in which the salmon swim, eating the hazel nuts which give it wisdom.

But this well is also the centre of Ireland, and the whole world. From the well of Nectain, all the rivers of the world run out, crossing the globe. It is the mythic omphalos of the world, and as such it could be repeated and localised in other similar places where rivers flow outwards across the land.

In Pictland, it was possibly localised at Dun (Sidh) Nechtain with its wells, sitting above the source of the Lunan river, famous for its salmon. (I’ve written about this here.) For years, people had searched in vain for a physical fort on the hill of Dunnichen, the large hill and ridge below the point of Restenneth, thinking ‘dun’ here meant a fort, when it rather refers to the residence of Nechtan. But Dunnichen is a critical symbolic point in the land, the source of water, power, and above all wisdom.

Restenneth, the place of fire, Dunnichen with its sacred wells, the Lunan river with its salmon, all point to this place as the ritual centre of southern Pictland. And it’s no surprise then to find Pictish settlements, forts, church, symbol stones concentrated around this point. It is the political, military, and philosophical centre of their world.

So can you imagine our excitement when it was announced that a cist cemetery had been found right at this critical point!

A shore of dorn stones

Years back I remember reading about the whole shore line of Loch Forfar near this passage being found to be covered by fist-sized water-rolled pebbles (although I seem to have lost this reference through countless computers). Not a type of beach, but something that gave the impression of having been carefully engineered by humans.

These type of rocks called ‘dorn’ pebbles do tend to show up in odd places. For example, the frontage of Newgrange has chunks of white quartz and black ‘dorn’ pebbles.

It’s possible that sometimes these were used in slings as a weapon. But they also show up in ritual contexts, and I don’t think anyone quite knows why.

Once, in Ireland, I went to see Dind Rig, a place of great importance for early Leinster. Here famously Labraid Loingseach, the ancestor of the Leinstermen, built an iron feasting house, invited his rivals to the feast, then roasted them in it.

Dind Rig isn’t a on a raised hill, it is on the pointy end of a ridge running beside the river, and a deep narrow ditch has been cut across the neck of the ridge to form a sort of island. Climbing down and up this really quite steep ditch was a challenge, and a bit pointless as when I got to the top of the fort it was overgrown with brambles. Getting back down off this island meant only one thing – I had to slide down on my back it was so steep. And of course the inevitable happened, momentum took over, and down I went. So there I was lying flat on my back in the bottom of the ditch of famous Dind Rig, covered in mud and brambles, cracking up laughing. Then, I thought, woah what on earth am I lying on?! The whole ditch was covered with dorn pebbles. Not one here one there, but totally covering the floor. I walked the entire way round that ditch surrounding Dind Rig, and the whole thing was covered with dorn pebbles. Why? I’ve no idea, but it was extraordinary. There are so many puzzles with no easy answers, what is clear is the ritual intent involved.

The cist cemetery

Roll on some years, to recent research on the biometrics of the Lochhead burials. Here’s a video with Dr Lucy Koster telling us some of the findings.

First, the cemetery itself has at least 20 burials in stone cists. Large groups of cist burials aren’t all that common in Pictland. Where the rest of the population was put is unknown. So, this large group of cists is notable.

The cists here are very well-made, lined carefully with thin slabs.

The dates for the main activity range from early 400s to late 600s, meaning it was not a burial of a particular event such as battle or plague, nor a cemetery that was local and short-lived for only a few generations. This was a significant place of burial over a lengthy period of time – in fact, the end point is around the time when Christian burials start to be gathered into a churchyard.

The cists are fairly regularly spaced with no intercutting, meaning that each new burial was placed knowing the locations of earlier burials, presumably because there was some form of marker above ground. It means that this cemetery was a space that was maintained for 300 years, a length of time most modern cemeteries would struggle with. And people repeatedly returned here to place a special burial, on average one burial every 15 years.

The cists are aligned NE-SW with heads at the western end, a common Iron Age direction in Scotland. Again though, the consistency of direction is at a rare level of organisation maintained over several centuries in this important place of burial. It suggests strongly that someone or some group oversaw the management of this burial site, and there are only two groups that could plausibly be, the druidic institution, or the Christian church.

All of these factors point to a place of elite burial for a known and acknowledged purpose, between 400 to 700 AD, and managed by a knowledgeable group wielding considerable and protracted social power.

The people

12 out of 17 burials where sex could be determined are women. This pattern of a preponderance of women seems to be a common phenomenon for Pictish burials. I’ve written about the overwhelming and growing evidence for the visible roles of Pictish women in these early centuries, especially obvious in this southern region of Angus, Gowrie and Fife. So this comes as no surprise, but it still requires explanation.

We know that women played a major role (along with the men) in both the druidic institution and the early centuries of the church prior to the 7^th century when things start to change across Europe and women start fading from view.

The isotope analysis showed that while some people were local to southern Pictland, others appeared to have traveled from the west coast. This amount of movement across known political borders does suggest we are looking at either druidic figures or churchmen, both of whom were not tied to their family group’s obligations and responsibilities under the law, and so could move around as their studies and skills demanded.

But all of the people also were shown through their isotopes to have lived their later lives in the local area. Which suggests these people have lived as a social group for some time before dying.

And most importantly, none had a familial genetic relationship up to a 3^rd degree of relatedness, nor did they share a matrilineal lineage. If this cemetery was the chosen burial site for the local lords and their families and their relations, then this would have shown up in the DNA, but it didn’t. The only group that fits this genetic diversity – travelling across borders and not related at all, living together as adults, buried at a philosophically important place in the land since AD 400, near no settlement or church, is a druidic or clerical group.

In the written history of early church groups, we do see plenty of cases where a monastery is run ‘in the family’ for many of its members, especially as spiritual heads, Iona being a case in point. In which case we might expect to find some genetic relationships if the Lochhead burials came from a particular church group.

We do have to ask whether it’s possible that this ‘in the family’ tradition may have extended back into the druidic class, but we don’t have enough data to say either way. What we do have is a tradition where the druid was said to be chosen because of their expertise, knowledge and wisdom, which if correct and actually applied, may tell against their choice being influenced by family connections. The early church may have fractured this druidic tradition of expertise with claims of ‘holiness’ and a close relationship to God – which could be subjective, and vulnerable to influence from political power and family connections.

The druidic bureaucracy famously met in sacred places called ‘nemetons’, and held independent status in ritual activity and as mentors to kings, the first to speak in an assembly even before the king. In contrast, the early church had not just ‘monasteries’ as their sacred spaces, but wanted to build churches across the land to convert the general population. This need for land to build a church, and the need for political agreement and safety to carry out conversion, meant that the early church needed to build a relationship with the local region’s political elite, and again this promoted selection of influential church leaders who had familial connections with other high status political families. So, it seems that we might expect an early church group to show signs of some genetic relatedness, especially for its leaders. Again the evidence seems to tell against the Lochhead burials being part of a church group.

So far, we see the evidence tending towards ‘druidic’ because of the relatively early date, the Iron Age orientation, the sacred site of Nechtan, the lack of a church, and the lack of familial relationships.

Evidence of druidic presence

There are two massive druidic nemetons in this southern region of Angus and Gowrie. The first, not far from the burial site, is Finhaven, ‘Wood Sanctuary’ (see Watson CPNS p250), on the mountain ridge above Aberlemno with all its pagan and Christian symbol stones (for its location see my map above). The second is to the west in Strathmore at Meigle and Nevay, a nemeton that continued on through the centuries as a key Christian site. Here at nearby Finhaven we have a major druidic centre, beside a major political centre, the likely source of these druidic burials.

Let’s sum up the case for a druidic burial site. We have a large cemetery with elite and carefully constructed burials. The location lies at a critical point in the land, a point of fire, a passage between east and west, the centre of the land and its rivers, in the presence of the deity Nechtan. The cemetery was organised and maintained by a powerful group, for three early centuries between AD 400 and 700. The burials are aligned SW-NE as was common through the Iron Age, not E-W as we might hope for Christian burials. The people here were both local and from other political regions, but had lived locally for their adult lives, indicating they were part of a religious group whose members could cross boundaries and cut familial obligations. They displayed no genetic relatedness as we might expect of a later church group. There is no trace of an early church nearby, but there is a large nemeton, close to the centre of Pictish power and activity. The early date, the sacred place, the alignments, the lack of genetic relatedness, the nearby nemeton, all point to a sacred place of druidic burial.

In her video Koster relates that it is thought this cemetery was a means of ‘bringing people together’. It is unclear what social mechanism was imagined to be present that pulled diverse people across borders, breaking familial obligations, to be buried here, who these powerful people doing the pulling were, why someone might be chosen to be pulled, why someone in particular might be chosen for burial, or for what critical purpose this whole exercise could be. It is a forced description, not an explanation, disconnected from the culture of the day. The simple fact is that we cannot hope to understand the archaeology without an honest and broad acknowledgement of the entire context of the Picts.

What is the purpose of burying a druid at a critical place in the land?

What was the philosophy that gave motivation to burying people here? We know some things, we don’t fully understand many others. We know for example, that burying a person, even in a Christian cemetery, is said to result in the person’s spirit being held there, unable to pass on to the next life, at least for a while till the next person is buried.

In Irish lore, we find a similar context, where a famous warrior is buried at a border point, to guard the land in death as he did in life. To say nothing of all the legendary heroes that will rise again when the land is in need.

This idea, that burying someone traps their spirit, may be behind the general lack of burials during this period.

So my feeling is that this cemetery likely has a similar purpose, for the druid, in death as in life, to protect and sanctify this crucial omphalos of the land, the source of life-giving water, the source of all wisdom, to keep it safe, to keep it pure. To remain there, protecting, guarding, until the next druid is buried and takes his or her turn.

But that’s my sense from all my reading. There could be some other philosophical rationality driving this tradition. We need to stand up, take courage, think, look with honesty at everything, warts, druids, and all.

If the notched form of the symbol refers to a symbolic broch, what are we to make of the unnotched form?

Logically, it probably means that it refers to the same or similar philosophical item, but unfortunately we still don’t really have a solid handle on what role a broch plays. So, here I’m going to look in more depth at the context of the unnotched form, and start from scratch to see where it leads us.

The southern brochs

The unnotched form of the symbol only occurs to the south, with a dividing line around the level of Easter Ross. This puts it outside the area of the real brochs in the north.

There are a group of structures called the southern brochs, mainly found around Angus and Perth, extending into Lothian. They are generally considered to be of poorer quality workmanship, but seem to fit into the general idea of a large thick-walled round tower. There are two disc-over-rectangle symbols in this same region, a very late one from the 700s at Meigle on a small grave marker, and a very early one, probably prior to AD 200 below East Lomond Law at Westfield which may be contemporary with the actual brochs. Apart from this, the disc-over-rectangle symbol again doesn’t occur in the region of the southern brochs, another pointer to a possible relationship between broch and symbol.

The accompanying symbol

Perhaps the most striking thing about the unnotched disc-over-rectangle symbol is that each instance occurs with a different accompanying symbol.

This ‘once only’ pattern extends all over the mainland of Pictland, even to the couple of northernmost unnotched forms. In contrast, the Orkney notched forms always occur with a crescent/V.

And, while not exact, the accompanying symbols largely follow the same north-south pattern as the symbol set’s ordered list.

Another strange aspect of this symbol is that both variants can occur as part of a string of 4 symbols, on a stone that appears to mark out a focus of the accompanying symbol in other local stones.

When we add this 4-symbol marking of a special region for the accompanying symbol, together with the unusual pattern of a ‘once only’ accompanying symbol, and the general pattern following the ordered list, then the implication seems unavoidable that these accompanying symbols are somehow marking the local region or the local people.

Disappointingly, there seems to be no known regional placename that might align with these accompanying symbols, apart from Gowrie and its Gabor, the Pictish beast. But, that doesn’t mean there won’t be other related names, we need to keep an eye out for them as we go.

Details of the locations

Let’s now look at the findplaces on the Pictish mainland, in the hope we can find further information about the symbol. Here I will repeat the instances of the disc-over-rectangle symbol (unnotched).

The disc-over-rectangle findspots are notably very active places in the landscape. Dyce has multiple stones, a ford, an ancient chapel, two known settlements within a mile or so, and a massive silver chain found nearby. Inverurie has its concentration of stones, with the junction of two rivers, fords, stone circles, two large ritual mounds, at the base of Bennachie. Tillytarmont, again with multiple stones, is a long thin water meadow between two rivers falling together, a major fording place, with large cairns and mounds, an obvious place of ritual, not settlement. Kintradwell is a ritual space on a narrow strip of coastline to the north of Golspie with multiple symbol stones. Sandside House is reputedly a place of numerous cist burials, overlooking a sheltered harbour on the north coast. Drumbuie is at a river flowing into Loch Ness, a major ford and the entry to the road heading north, beside Urquhart castle, a druidic nemeton, multiple cist burials, overlooked by a large barrow cemetery. The list of important places, as evidenced by known archaeology and marked out with multiple symbol stones, is striking. In contrast, there are only a few single instances that don’t appear to be at a major ritual place, although further archaeology may still discover otherwise, Advie and Corrachree.

Multiple symbol stones The majority of instances occur at a place with a number of other symbol stones. And this fact alone points to these places as key points in the landscape, as the majority of Class I symbol stones usually occur by themselves. Rhynie is a good example of this type of situation with multiple symbol stones, with a ritual palisade enclosure, a later chapel, a ford, cist burials, barrow burials, ritual square enclosures, an archaeology we now have a good handle on, in a region replete with earlier circles, cairns, stones from the Neolithic onwards (but none of our symbol though). The settlement area is not at the ritual space of Rhynie but nearby up on the Tap o’Noth.

Boundaries Although we don’t really know how the Picts divided the land, it is reasonable to see most if not all of these locations as a boundary point. However, some instances do not sit at a boundary, including the Spey river instances, and Corrachree, Westfield, Meigle, and Dun Echt. It may be that this symbol isn’t marking a boundary as such, but a place of importance to a local region – which would often naturally be a boundary location.

Burials Some of these locations have multiple burials, Sandside House, Drumbuie, Tillytarmont, while Dyce and Meigle have burials at churches. The Meigle instance is actually on a late personal memorial stone found in multiples at churches. Burials are not all that commonly found in Pictland, so finding burials at a number of these sites underlines the significance of the site. Many of the sites have not undergone archaeological investigation, so the list could be larger. However, it seems that burials are not a necessary part of these sites, instead it is the importance of the site itself, including some boundary locations, which attracts the burials to it.

Fords Most of these stones are at a ford, or a harbour, and this is undoubtedly the most consistent aspect of the symbol’s findspots. A couple of stones aren’t, namely Westfield at the base of East Lomond Law in Fife, and Corrachree which is in the middle of a small very active region on the inner River Dee.

Settlements None of these findspots are at a known settlement, although some have a settlement nearby, or we can assume from the route position and activity that there must have been a settlement nearby. It is quite clear that the findspots have a ritual significance, not a mundane one, especially when we see places like Tillytarmont and Drumbuie which are soggy river meadows.

So far, the picture we have is of a special ritual space, at or near a ford or harbour, with multiple symbol stones, and often with cist and barrow burials. What this tells us is that these places are key to the ritual, and probably political, organisation of the land of the Picts. And that it is the disc-over-rectangle symbol that is the key marker of these ritual places.

This disc-over-rectangle symbol is marking a special ritual space, but that still doesn’t explain exactly what the symbol is, or means.

Things with a similar form

Another way of tackling this problem is to look for things that the symbol may represent. We’ve already seen that the notched form of symbol looks like a broch, but are there other things that it could be?

To begin with, the Westfield stone shows us that the symbol is drawn as half of the double-disc symbol, but again, as we don’t know for sure what the double-disc symbol means, that doesn’t really help.

When we have a symbol that has as its focus a circle, we might suspect that we are dealing with some sort of ‘sun’ symbol. Bronze Age carvings from Sweden often have a similar depiction of the sun as a wheel tied to the boat which is drawing her across the sky.

The sun when seen at sunrise or sunset over water also has this shape – the sun’s disc over a column of light.

But, can we make a connection with our symbol to this picture of the sun? It’s hard to think of an immediate way to do this. However, it is the strongest picture I’ve come up with, so I can’t rule it out.

Metalwork

One item that does have a similar form to a broch, although of course much smaller, is a smelting furnace.

Although a broch itself would not have served this function, it is notable that most of the big brochs have a settlement of small cellular houses around their base, with lots of evidence for metal-working. For places in the far north of Scotland and Orkney and Shetland, making high-value metal-work items was an excellent way of creating a product which was light and easy to transport, and of sufficiently high value to make the journey to the client still profitable. We might ask then whether the broch was a symbolic ‘furnace’ itself, not in practice but as symbolic of a smith deity.

Along the same lines, there was a ‘bag-shaped’ building discovered at Portmahomack which looks like the disc-over-unnotched-rectangle, a building that was used as a craft-working area, here again with metal-work (upside down in the photo below).

The other location that falls into this metalwork category is the symbol stone of Knocknagael, close to the massive craft-working and metal-working centre of Culduthel, just outside Inverness. Similarly, a small settlement near the Dyce stone also showed evidence of metal-working. I don’t know of any other cases of metal-working near other findspots, although many won’t have had archaeological work done so we can’t say for sure. Overall, I suspect that smithing might be another activity which is attracted in to the sacred space, rather than its prime function.

Making sense of the complexity

At this point, we have a number of aspects of the findspots, all of which indicate the disc-over-rectangle is sitting at a special ritual place - fords, shores, cist burials, barrow burials, multiple symbol stones, metal-working, and a unique accompanying symbol. But, it’s still not clear which aspect is key to the significance of the symbol itself.

Language

In purely linguistic terms, if we have one symbol that always accompanies another symbol that is particular to one region or tribe, then we may be looking at a term along the lines of Irish úa which means ‘grandchild, descendent’. This term was used very early on for Irish sept names, but fell out of usage until the 10-11^th century when it began being used for today’s surnames, eg. Ui Néill, Ui Fidgente, Ó Néill, O Ruairc.

Alternatively, we could be looking at some other term denoting a place belonging to a sept or region, for example, the Ford of the Goose (sept), the Burial place of the Eagle (sept), the Smithy of the Boar (sept).

The Ford

There is another issue that is likely muddying the waters. Which is that all symbols are probably multi-valent, with layers and levels of different but associated meanings. Understanding the linking association is the tricky part, as that will be driven by the Celtic worldview, which is largely opaque to us today. So the use of a symbol on a particular stone may be referring to a different association than the next instance of the same symbol.

Let’s take, for instance, the ford association, which seems to be the most consistent factor across the findspots, and explore what it could mean if this symbol represents a ford.

The ford holds a significant role within the Celtic worldview. The ford (or shore) is a liminal space where battles are fought by single warriors for the control – physical and philosophical – of the land itself.

Many battles between warriors are fought at a ford, it is the ritual space where battles must be fought. In the Táin Bó Flidhais, at one point Fergus goes to Ailill Find to ask for harbour, but after a heated exchange, Aillil is unwilling. ‘Out of the fort with you then’, says Aillil. Fergus responds by telling Aillil to send a warrior to do battle with him in the ford. Aillil decides to go himself to the ford where he is killed by Fergus’ warrior Dubthach. The interesting aspect of this incident is how the warriors deliberately place their battle at the ford, rather than at the fort or outside it.

The deity of battles and warriors is the Morrigan, and in this capacity she is seen at the ford in some stories. The heads of dead warriors are said to be her ‘mast’, the fallen whom she gathers up. And we might recall here Sueno’s stone, with its mound of heads in the well of a broch, bringing us back around to our symbol. I would doubt that the purpose of a broch is solely to contain the heads of dead warriors, but it may be one of the functions of its well.

Another interesting story that shows the importance of the ford to warriors and its association with the Morrigan is in the Táin Bó Cúailnge. Lóch’s brother had been killed in battle with Cú Chulainn. ‘Come to the upper ford,’ said Loch; ‘it would not be in the polluted ford that we shall meet, where Long fell.’ But when the two warriors meet, the Morrigan tries to interfere with the outcome of the fierce battle at the ford. First she comes at Cú Chulainn as an eel and wraps herself around his ankles, flattening him on his back, while Lóch hacks at Cú Chulainn to turn the ford red - Cú Chulainn breaks free by breaking the eel’s ribs. Then she comes as a she-wolf, driving the cattle away – he puts out her eye. Then she comes as a hornless heifer leading the herd of stampeding cattle that churn up the riverbed so that he cries out that he can no longer see the ford – he breaks her legs.

It is also at a ford that Cú Chulainn has his final meeting with the Morrigan. Here she displays another of her aspects, as fate goddess, foreseer. The Morrigan tries to turn him away from traveling on to his final battle and death but she is unsuccessful. Cú Chulainn has seen an otter, a doburcu ‘water dog’, an animal that is his kin, and knows that his death is near.

Now this is where trying to fully grasp the meaning of a symbol becomes difficult. If we think of this symbol sitting at a ford because it signifies a ford, then we may be at the wrong level of meaning. By which I mean that perhaps the top level of meaning refers to the Morrigan, and secondly, it can then refer to the ford as the place where her warriors do battle under her auspices.

Here’s my outline of some of the key aspects of the Morrigan. They seem to be connected through the idea of fate, fate in battle, fate of the people of a region, fate in control of the land, and the fate of the warriors through whom she watches fate play out.

The use of the symbol at a ford by the Picts could then refer both to the Morrigan, the Great Queen, and to places such as the ford or shore as her liminal space where the fate of warriors and the land is decided.

That the ford was sacred to a goddess we know from dedications on the continent to the goddess Ritona, also known as Pritona. Her name is based on the Gaulish word ritu- ‘ford’, also found in Welsh (and therefore likely in Pictish) rhyd ‘ford’.

As goddess of the dead warriors, the ford would also be an appropriate place for certain burials of the Morrigan’s dead. The presence of the symbol at a ford does not necessarily imply that a Pictish battle has already been fought here, although it does imply that the ford is the place where such battles for the land are appropriate, and therefore a type of boundary space, both ritual and real. For the Celtic worldview does not make a distinction between ritual and real, it is all a seamless part of their universe.

The symbol may have a meaning as straightforward as ‘ford’, or ‘descendent’. But chances are it has multiple levels of meaning that will take years of study to re-discover.

So for the moment, that’s my take on the disc-over-rectangle symbol. With the notched form in Orkney it represents a broch, or whatever the broch itself represents, but which may include the fate of the dead warriors with their heads in the well. And in the south, the symbol sits on fords or shores that are marking ritual places in the land between regions, accompanied by that region’s particular symbol. These places are usually fords, likely under the aegis of the Morrigan, a place where battles, fate, and the future are decided.

But, although that is my take for today, it must be an ongoing discussion.

First, let’s recap what we have so far.

1. The symbol comes in two variants, one with a notched base, the other with an unnotched base.

2. The notched symbol occurs in the north from Sutherland up to Orkney.

3. The unnotched symbol occurs in the south from Loch Ness down to Fife.

4. This north-south divide is a unique distribution pattern for any symbol.

5. This symbol’s use continues from early simple CI stones, through to CII stones.

6. The notched/unnotched variant may be a RH/LH version of the same symbol

7. The unnotched form occurs with a wide variety of other symbols, its accompanying symbol only occurs once. This ‘once only’ pairing pattern is not seen with any other symbol.

8. Both notched and unnotched occur as special stones with 4 symbols, marking a region with the accompanying symbol, showing the variants to be forms of the same symbol

9. The accompanying symbol of the unnotched variant has a strikingly similar layout across the land to the symbol ordered list.

10. The notched form in Orkney has a different pattern, always accompanied by a crescent/V.

11. It is not a ‘mirror-case’, as no such thing exists, mirrors being kept in a cloth bag.

The north-south division

We see from the distribution maps, the disc-over-notched-rectangle symbol occurs in the same northern area as the brochs. Conversely, the disc-over-unnotched-rectangle symbol only occurs in the complementary southern areas where brochs are not found. This sharp alignment of symbol and broch means it is worth investigating whether the two may be related, the real broch and the symbol as a broch.

While other symbols sometimes are limited to a certain broad range, none of them have such a clear delineation as this disc-over-rectangle symbol.

Summary of the context of the notched symbol

Drawing the symbol

Like the cauldron symbol which is drawn as a form of abstraction viewed from the top rather than side on as we would see in a photo, so the disc-over-notched-rectangle symbol can be seen as a form of abstraction of the shape of a broch.

The roof By depicting the symbol in such a way, the opening at the top is presented as the most important feature of the broch. And this depiction also shows us that the broch did not have a roof. It has lately become the fashion to portray a broch with a conical wood and thatch roof, although this remains highly controversial. The broch is basically shaped like a big chimney, and if the wind catches in the doorway (flue) at its base, the sparks from the hearth would reach to the stars. The last thing we want on top of a chimney-like structure is a combustible cap.

The internal scarcement The same problem is encountered when considering if the small ledge roughly half way up the internal wall was there to support an internal upper floor made of timbers large enough to span the interior. Brochs usually have an internal staircase to this ledge, so it must represent some function that people could reach. But a wooden upper floor would be extremely dangerous and prone to being set alight every time the door of the broch caught the wind.

The Greens (photo STAMS) version of the symbol appears to show the internal ledge and separation of the broch wall into lower and upper sections.

Light and Sky Having a capped roof on the top of the broch, and a secondary floor half way up the tower, would create a serious problem of light. Without windows, the whole place would be dismally dark, and it’s hard to imagine daily activities being performed in pitch black. But if we imagine a broch without a roof, without a full floor at the second level, then we may have something quite magical – a huge tower with an opening to view the northern sky. And as we see that the disc-over-notched-rectangle symbol tells us that it is the open top of the broch which is its key aspect, this opens up all sorts of possibilities, rituals, science, philosophy, centred on the highest part of the northern skies.

The broch – its form and function

If the symbol does represent a broch, then we next need to look at the actual broch itself, and what is known about it.

A broch is a large tower of stone, with extremely thick walls and a relatively small inner space, inside which there is often a natural well, and a central hearth. The broch is entered through a ground level small entrance, and there are no windows. The lower half of the wall usually contains a gallery and cells, reaching to a narrow scarcement, which appears to be some sort of structural support for a ledge, walkway, or floor half way up the inside of the tower.

There are somewhere between 100 and 500 brochs in Scotland, depending upon the definition and state of archaeology. Many brochs are located near the sea, but a number are inland. Dating of the brochs has recently pushed their beginning back a few centuries BC, but it remains that most were built in the 1st centuries BC and AD.

Determining the purpose and function of the broch has long been a thorny issue. Some ideas proposed have been high status dwellings, defensive structures, offensive structures, but none of these have found easy acceptance. The situation is not helped by the general lack of finds within the earliest levels of brochs (a notable exception being The Cairns in South Ronaldsay) indicating most were not initially used as daily habitation spaces. In later centuries, when the brochs had started to deteriorate, usually after a hiatus during which windblown sand accumulates for a century or two, we have the ‘Pictish’ layer, where the internal space of the broch is partitioned off with small living spaces at ground level.

A group of smaller buildings frequently surrounds the outside of the broch, often with evidence of fine metal-working, a trade which was perhaps one of the few cash trades which could withstand the expense and hazards of long-distance trading from the far north.

Brochs are so plentiful that it feels like there was a sudden ‘broch-mania’ in the north and north-west of Scotland for a short period of time. And then it stopped, as if there was simply enough brochs for their purpose, or the motivating factor behind their construction had passed.

So why would the Picts have a symbol which portrays the broch?

The Kintradwell symbol stone was found 100 yards west of Cinn Trolla broch, and may have stood at the entrance to the Pictish settlement around the base of the broch. The informal stone of Gurness and the bone of Burrian were also found in a broch. All other the disc-over-notched-rectangle symbol stones sit at a likely place for a real broch – above a good anchorage bay, or a position which could oversee a portage between sailing points. In other words, the disc-over-notched-rectangle ‘broch’ symbol may be standing in for the real broch at a point in time when building a broch is no longer feasible.

Dating the brochs and the symbols

The relationship between the real broch and the symbol that looks like a broch comes down to a question of dating. So far, archaeology has found the earliest known symbols carved in the Fife caves and the Dunnicaer seastack dating from the third or fourth century AD, but even here the symbols are fully formed and some occur in pairs following the syntax rules. In other words, the symbols themselves are far older again than the symbol stones, and the symbols may indeed originate in the same period as the brochs 100BC-100AD, or even some centuries earlier. If the disc-over-notched-rectangle symbol is from the same or later period than 100BC-100AD, then the symbol might be a depiction of the real broch.

We can assume that all disc-over-notched-rectangle symbol stones in Orkney come from the same period as the secondary level of habitation of the brochs which archaeology shows has distinctive Pictish artefacts. In other words, the symbol stones are later than the brochs themselves in Orkney (regardless of whether the symbols originate with or before the real brochs). So here we see that the Picts are carving ‘broch’ symbols amongst already existing real brochs. Why would they do that? It may be that the Picts are no longer in a position to build a real broch, lacking resources or expertise or will, and a symbolic broch can perform the same ritual function as a real broch.

In other words, why bother to build a real broch when carving a ‘broch’ symbol does the same job!

But if, as well may be, the Pictish symbols as a group are older than the real broch, the reverse is true, that the broch is likely a material realisation of the symbol. That is, the symbol is what is dictating the shape of the broch. And if this is the case, it follows that the real broch’s meaning and function is derived from the meaning of the disc-over-rectangle symbol, not the reverse.

But what was that ritual significance? We now seem to have two unknown but related factors, the broch and the ‘broch’ symbol.

Heads and brochs

There may be another clue to the purpose of a broch. On Sueno’s stone at Forres in Moray, with its representation of serried rows of warriors and battles, two scenes show the aftermath of battle with bodies and severed heads. One section shows severed heads under a sort of thick arch, and a line of headless bodies. The arch may depict a cave, as there are several caves along the coastline here which are renowned for their ritual usage over centuries, including symbols and the organised deposition of heads and various bones into different parts of the caves.

A second scene shows a line of decapitated bodies, with the severed heads under what appears to a broch. This makes sense, as we know from archaeology and stories, that the heads of warriors were often placed in a well, and most brochs have a deep internal well. These broch wells seem to be more for a ritual purpose than practical, as the way down is steep and it would have been an unsatisfactory source of water for a community. Were these particular heads of the Forres warriors placed in a well of a broch, and, is this part of the ritual function of a broch’s well?

The language of the broch

There are two words that eDIL gives for a ‘tower’. The first is tor ‘tower, fortified building’, in other words, a broch. The other, tuir ‘pillar, column’, is according to eDIL sometimes confused with tor.

One of the Gaelic names for Orkney is inistura ‘the island of towers/brochs’. It is no wonder that we find a whole cluster of disc-over-notched-rectangle (‘broch’) symbols on Orkney.

The broch in mythology

As we might expect with these extraordinary buildings, they are mentioned in Irish stories and mythology. The great god Lugh was conceived in a tower owned by Balor, the Fomorian king, on Tory Island, off the north shore of Ireland. Its name Toraigh means ‘place of towers’, although it’s also said to mean ‘place of steep rocky heights’ which seems to be a denial of its mythological importance.

Tory Island was also where the Nemedians defeated the Fomorians at Conand’s Tower, in one of the earliest battles of Ireland.

But perhaps the most famous ‘tower’ place is Mag Tuired, the Land of Towers, which was the site of two early battles between the Tuatha De Danann first with the Fir Bolg, then with the Fomorians and Fir Bolg. These two battles are part of the mythology of Ireland that relate how the land became inhabited, so of course they are localised in Ireland. But, the Tuatha De Danann are said to have come from ‘the north’, the Fir Bolg after defeat went to ‘the north’, and the Fomorians were located on the north-west coast of Scotland. Which has always made me wonder if the original Mag Tuired of myth was Orkney, the Land of Towers.

Which makes me wonder if the three symbol stones in Orkney, all with the crescent/V and disc-over-notched-rectangle, Greens, South Ronaldsay, Birsay, plus the two bones found in the Broch of Burrian, can be read as Land (crescent/V) of Towers (disc-over-notched-rectangle).

The picture so far …

So far, we have a symbol, the disc-over-notched-rectangle, that has the same distribution pattern as real brochs, and looks like a broch, although it’s hard to tell given the similar early dating of both broch and symbol as to which came first.

It’s possible that on Orkney, the notched symbol was used to symbolically stand in for the real broch, at a time when building a broch was no longer feasible.

The problem though is that even if the broch and symbol effectively have the same philosophical intent, we still know little about the meaning and function of either.

But, we also see that this notched symbol is related to the unnotched form, the two having complementary distributions. Both can function as a group of 4 symbols. On the Pictish mainland both notched and unnotched forms appear to mark a region where the accompanying symbol has a special presence, although the notched form in Orkney only occurs with the crescent/V. So, how do we meld the two contexts of these notched and unnotched symbols into one picture that makes sense?

The next thing to do is to look in more depth at the southern unnotched form of the symbol, to see what its context might tell us.

It comes in two variations, one with a notched base, one with an unnotched base. Does this make them variations on the same symbol? In this blog, I’m going to look at the distribution patterns of these two symbols, to look at the ways in which these two variations can be quite different, and yet similar enough to tell us they are the same symbol.

The mirror case nickname. Now some of you will know this symbol as the ‘mirror case’. When people didn’t know what so many of the symbols meant, they gave them monikers – just something to call them so that we could communicate with each other when talking about them. This does not in any way mean this symbol is a ‘mirror case’, it’s just a nickname. In point of fact, there is no such thing as a mirror case shaped like this, as early texts tell us mirrors were held in a cloth bag. So, here I’m going to call it ’disc-over-rectangle’ to remove any false assumptions.

The complementary distribution pattern

Beyond anything else, the distribution pattern is the most striking aspect of this symbol. The notched form of symbol only occurs in the north, while the unnotched form only occurs in the south.

The notched form in the north has the same geographical coverage as the brochs. And given that the symbol looks like a broch, the connection between broch and symbol seems an obvious idea to investigate, but, for the moment, I don’t want to leap into possible meanings of this symbol, before looking at what other evidence we can gleam. After all, we don’t actually know the role a real broch played, so two unknowns may not get us far anyway.

Most instances of this symbol are Class I, with only two late Class II examples at Dyce and Meigle. Some of the Class I instances are also late, for example, Inveraven and Sandside House both with a double-sided comb. But this continued usage into the Christian period does indicate that this symbol remains of significance throughout the symbol-carving period, no matter which worldview is prevalent.

Understanding this north-south divide is key to this symbol, and I’ll be following the similarities and differences across the divide next.

The disc-over-notched-rectangle - the northern variant

There are 7 instances of this symbol, five of them on CI stones, one on bone, and one graffiti. Two are on the mainland, the rest in Orkney.

All examples from Orkney, except for the graffiti of Gurness, pair the disc-over-rectangle symbol with the crescent/V. However, all the symbol stones in Orkney regardless of their symbol types, have a crescent/V as one of the symbol pair, so this isn’t unique to this symbol alone. But this pattern does make the notched form of symbol notably different from the disc-over-unnotched-rectangle which pairs over a wide range of other symbols.

All examples of this notched variant are paired with a RH symbol, which is a strong indication we have here only the right-hand version of the symbol.

Drawing the disc-over-notched-rectangle symbol

The symbol is usually fairly plain, a circle over a notched rectangle. The circle has one inner circle, both inner and outer circles may be a single or double line. Only the Birsay stone has further internal decoration within the circle.

This lack of internal decoration in the disc is in stark contrast to the symbols in the south without a notch.

Most of the Orkney symbol stones appear to be finely carved, and despite the relative plainness of the symbol itself, its paired symbol is delicate and complex, indicating that this group is relatively late (by which I mean somewhere after the 4^th century AD). The Sandside House stone in Caithness also fits into this group, a beautifully worked and carved stone, with a late double-sided comb.

In contrast, the Kintradwell and South Ronaldsay examples are fairly simple and plain, suggesting these may be earlier than the Orkney set (or just simple and plain!), showing a continued usage of this symbol over a lengthy period.

The disc-over-unnotched-rectangle – the southern variant

This unnotched southern variant is markedly different to its northern variant. And a number of its aspects are odd. To begin with, it seems to often pair with the rarer LH form of symbols, and does so without any need to call in the moderating mirror+comb.

When it pairs with a RH form of symbol, it often calls on the mirror+comb as if to correct something. Only 3 instances pair with a RH symbol without any moderators, but we can have rare but valid cases of a LH symbol over a RH, so it’s possible here too the disc-over-unnotched-rectangle is LH.

What does this mean? If anyone thought trying to figure out the rules that govern symbols is easy, this shows just how difficult and confusing it really is. It may be that this form of the unnotched symbol is the left-hand LH form of the symbol, and the notched variant is the RH form. If this is the case, then why would a symbol only use the RH form in the north and the LH form in the south? That doesn’t really make much sense, and it is different to the patterns displayed by other RH/LH symbols which generally show a random mix.

Alternatively, it may be that this unnotched form prefers to be the dominant symbol in a pair, while the notched form is the nondominant form – especially as some other symbols create a notch in the symbol to express their subordinate state. At this point, I can’t be sure which scenario, RH/LH or dominant/subordinate, is correct.

Drawing the unnotched symbol

Some of the examples of the disc-over-rectangle have fairly plain discs like the northern notched variant. But, notably quite a few examples - perhaps from a later date – add decoration in the disc. Does this internal decoration signify a difference in meaning between the two variants, or is it just the art developing, or a local artistic trend? Another question I can’t answer as yet.

Four Symbols instead of a pair

Symbols come in pairs as a general rule, but a distinctive group of stones has instead the normal pair extended to 4 symbols. In all but one possible case, the notched or unnotched form of the disc-over-rectangle symbol occurs as part of this group of 4.

This peculiar set up of 4 symbols tells us that both the notched and unnotched variant can perform the same function, in addition to looking similar, and having a complementary distribution. They are definitely either variants on the one symbol, or two closely related symbols.

All of these stones with 4 symbols that include the disc-over-rectangle seem to be at the end of a series of stones with one of the other symbols in the group. In other words, a group of 4 symbols which includes a disc-over-rectangle is marking a run of symbol stones with one of the other symbols.

There is only one other CI stone that has 4 symbols on it, which is KINTORE1 with a symbol pair on reverse faces, fish+cauldron | crescent/V+beast.

Other CI stones have odd mixes, INCHYRA is an odd-shaped pillar stone with two pairs on adjacent faces, double-disc+fish | fish+snake, with an eroded possible third pair, along with ogham. KINTORE2 is quite eroded but seems to have possible beast + double-disc/Z and on the other upside down a possible beast with mirror, while ROSEISLE has a goose+fish and on the reverse an earlier pair upside down, a crescent + crescent/V, so that we can safely say that both KINTORE2 and ROSEISLE, and perhaps INCHRYA, are all examples of a stone having a new pair of symbols added at a later date, rather than an original group of four symbols.

A couple of other stones have a peculiar arrangement, but again these do not fall into this group of 4 symbols. For example, DINGWALL has an odd series of several crescent/V, double-disc/Z and a triple-circle, while KNOCKANDO has two crescent/V plus a spoked wheel.

There are quite a few CII stones with a Christian cross which have more than just the one pair of symbols, however, most of them occur either with pairs in separate areas on the stone, or with a longer string of symbols.

Given then, that we only have KINTORE1 as another possible stone with 4 symbols carved as a group, we see that this grouping of five stones each with 4 symbols, each including a disc-over-rectangle symbol, must be significant.

As a linguist, when I see this, I can’t help but wonder if this symbol is acting as ‘and’. We know that the general rule is to present symbols in a pair A B – but what if we needed a third symbol A B C? Then we might express this as A B and C, with the disc-over-rectangle acting as the ‘and’. But, that’s an observation, I can’t prove it, and there are other possibilities anyway.

Inverurie and the Snake/Z

The Inverurie stone with 4 symbols ends a sequence of snake/Z symbol stones which start at the Picardy Stone near the source of a burn now called The Kellock, which is a tributary of the River Urie. Today the Urie is said to arise in another small tributary just to the south. This burn then flows past the Newton House stone, then the Brandsbutt stone, both with a snake/Z symbol, to finish at Inverurie, where the Urie falls into the Don. This river’s modern name ‘Urie’ is derived from the word for ‘fresh’. The symbol of the snake/Z then starts with the rise of a significant Aberdeenshire stream, and follows it to its ‘inver’ where it falls into the River Don at Inverurie, where we find the Inverurie stone with its 4 symbols, snake/Z+disc-over-rectangle, flanked by the common symbols of crescent/V and double-disc/Z.

There are of course other symbol stones with the snake/Z symbol. But it is the location of the 4-symbol stone with a run of the same symbol that is of significance.

It also pays to remember that we don’t know how many symbol stones have gone missing over the centuries, so it’s always possible that these sets of symbol stones originally had more members. In this case, it may be that the run of snake/Z continues up to Tillytarmont with its arch+snake/Z stone, and where we find another stone with a disc-over-rectangle + duck probably marking the start of a new ‘region’.

Here, for a broader context, is the distribution map of the Class I snake/Z symbol. The grouping in Aberdeenshire is quite clear. But, like all symbols, while there is sometimes a concentration of a certain symbol in a region, there are always some other instances elsewhere to muddy the water.

Dyce and the cauldron symbol

At Inverurie, the end of the snake/Z group, the Urie falls into the Don. The cauldron symbol appears to be associated with the River Don. This cauldron set along the Don ends at Dyce, today on the outskirts of Aberdeen, another stone with 4 symbols, the cauldron + disc-over-rectangle, the others again being the common pair of crescent/V+double-disc/Z.

This Dyce stone with its 4 symbols is a Class II stone with a Christian cross, and it would appear to be the latest stone carved at Dyce, with two earlier CI symbol stones. The Monymusk stone is also a Class II stone. This mix of old and new symbol stones tells us that these sets of a symbol were continually added to down through centuries.

Dandaleith on the Spey

This stone with its 4 symbols sits on the River Spey as it turns northwards towards the coast, and, where the River Fiddich flowing north falls into the Spey. 3 other symbol stones close by each hold a single pair, each one a disc-over-rectangle with one of the other 3 symbols on the Dandaleith stone. It’s as though the Dandaleith stone is ‘summarising’ the other three symbol stones. All of these stones sit on fords, just as many of the snake/Z and cauldron stones do.

Birsay and South Ronaldsay

A similar picture presents on Orkney. The Birsay stone has an eagle as its accompanying symbol, and this symbol is repeated at Knowe of Burrian and Oxtro.

On the South Ronaldsay stone the accompanying symbol is the crescent/V, and on the other face rectangle+crescent/V, but it’s hard to know which symbol (if any) is the accompanying symbol.

The pattern of the accompanying symbol

It’s at this point that things get even more extraordinary.

Ignoring (for the moment) the Orkney notched examples all of which pair with a crescent/V, then all the other examples each pair with a different symbol! Once, and once only. That is odd, because 50% of symbols are of only three types, the double-disc/Z, the crescent/V, and the Pictish beast, which means that we should find about a half of all disc-over-rectangle examples paired with one of these three symbols – which they aren’t, by a long shot.

Also, a lot of symbols seem to ‘like’ to pair with certain other symbols. But this doesn’t happen here. In fact, the disc-over-rectangle is the ONLY symbol that shows this strange ‘once only’ pairing pattern.

Some of the accompanying symbols are quite rare. The Tillytarmont stone is the only example of the duck symbol that we have.

If we add this peculiar pairing pattern to the observation above, that certain stones with 4 symbols seem to be marking a local set of the accompanying symbol, then we have to ask whether we’re looking at some sort of territorial - political or philosophical - marking in the land.

To which I would add, that in my video about the Pictish beast, I noted the peculiar pattern of southern Pictland below the Mounth being divided to the west by the Pictish beast in Gowrie (with the Meigle disc-over-rectangle+PictishBeast) and the east by the double-disc/Z in Angus and Fife (with the Westfield Fife stone with disc-over-rectangle+double-disc). So here again, we find the disc-over-rectangle paired with the symbol that appears to be of special significance to the local area.

While most symbols cannot be easily relegated to a particular region that might suggest a tribal identity, there is still something odd that shines through at times, like for instance, that the southern Fife caves have a preponderance of the double-disc, and the northern Moray caves show the reverse, a preponderance of the crescent. Something geographical is happening, the question is what exactly?

In the next map I’ve marked the accompanying symbol to show the distribution of those symbols and the area around them that potentially might be under their sway, just as the cauldron is on the River Don for instance, largely by thinking about the other examples of that symbol.

Now, if you’re been following my previous blogs on the syntax of the symbols, something about how these accompanying symbols are laid out across the land might seem to have an odd familiarity to it. If we ignore the animal symbols (boar, salmon, duck, eagle) which I can’t as yet place in the ordered list, these accompanying symbols are roughly following the order of the symbols in the ordered list – not exactly, not precisely, but largely.

(All the symbols appear to be arranged in an ‘ordered list’, which then dictates what their natural order on a symbol stone should be. My blog about this is here.)

It’s worth remembering too that on the ordered list as I’ve come up with it, a number of symbols only have a relative place, so the list itself may at times be a little out of sequence.

This cannot be a coincidence.

This is telling us something geographical is going on here, now we have to find the appropriate explanation. Are these the names of territories – as for example, Gowrie is named for the gabor, the Pictish beast? I would have hoped that if this were the case, then we might have spotted other names that suggested an association with a symbol before this.

Are these the symbols that belong to a territory regardless of what the name of the territory was? That seems likely. But we’d have to find a lot more evidence before we could be confident of that solution.

Or, is this some sort of philosophical marking of the land in its sacred spaces? That too seems a likely solution, but, how to progress the idea further?

Or, is there third factor driving both the order and the symbol location to bring them roughly together? That too is a possible scenario, but again how to test the idea?

Slowly, slowly, the Pictish symbols are giving up their secrets. Bit by bit the puzzle is coming together, and eventually we will see through all these questions and complexities to the whole picture. We just have to keep at it.

In the next two blogs, I will look more closely at the northern notched form, and then the southern unnotched form, to try to further tease out their significance.

Recently Fiona Campbell-Howes wrote her own blog about this dilemma, and it’s great when someone does this every so often, as having the aspects of the puzzle pulled together in one place does trigger the reader to focus on the topic, and hopefully to have another long hard think about it all. It certainly had that effect on me, so thanks Fiona, and here’s a link to her blog if you haven’t already read it. My own blog will pick up the discussion.

And yes, please hang on to the end, while I try to come up with a long overdue solution.

Bangor/Banchor as a placename

The heart of the dilemma is that there have been two quite different interpretations of this placename, and researchers cannot come to a consensus on which is right. This debate has been on-going now for a whole century. That’s an awful lot of head-scratching. But this lack of ability to clearly understand this placename also betrays something else fundamental about the dilemma – the simple fact that neither of the proposed interpretations sits quite right.

The first proposed interpretation of beannchor is that it means ‘horn-cast’, that is, a cow’s horn which has been cast off, and the great placename scholar W.J. Watson pointed out that most instances in Scotland occur on a significant loop or horn of a river, the benn ‘point/horn’. But, as Fiona said, rivers have many such loops, why just these ones? And why, in the case of Fife, no river at all? The second interpretation came from the Welsh linguists, who pointed out that there is a word bangor in Welsh that refers to a twisted rod that is used to strengthen the top of a wattle fence, and from this it was proposed that the term had been extended to mean the wattle fence itself, and further because four bangor’s were associated with a monastery, that this wattle fence referred to the wattle fence around a monastery, and then by extension to the monastery itself.

As you can see, I don’t feel particularly confident in either explanation.

Where are these banchor places?

For me, I always like to turn to the most basic data first. In this case, we actually have a lot of banchor instances to examine – 9 in Wales, 18 in Scotland, and 24 recognised in Ireland.

So, the placename Bangor/Banchor/Beannchor is relatively common, which makes it all the more frustrating that we seem to have lost the sense of what exactly it means.

Where are these bangor places found? Most in Britain are beside a loop or large bend in the river. A few are not. The Fife bangor stands out as an exception as it is on no river. In Wales, the famous monastery site of Bangor has a river but no loop, and six examples are at small farms. In contrast, only one of Ireland’s banchor’s is in a river loop, but, almost all of them abut on a river, loch or ford. Despite the few exceptions, there seems to be a persistent association with water.

Also, the majority of them are found outside the main settlement areas. Only a few instances, such as Bangor in Gwynedd, Bangor in Ulster, and Banchory-Ternan, Banagher in Offaly, have become modern towns. Fiona in her blog pointed to banchor having a similar pattern of distribution to the term carden. another placename which seems to be related to wooded areas. This trend - to find a banchor near water and beyond the main settlement areas - is striking.

An interesting observation on the Irish banchor’s is that they seem to be on higher ground. A thanks is due here to the work on the Irish banaghers by James Scully and Kieran Keenaghan, whose website is inspirational and has helped shaped my thinking.

Survey maps showing places called Banagher/Beannchor or a version thereof show that while their altitudes may vary, in all instances the place overlooks the surrounding landscape.

Another point to note is that the majority of banchor’s are not associated with any man-made structure or activity, except for the four ancient monastery sites. Most banchor’s appear today simply as a place in the landscape.

How old is this term?

We can say this term is quite early, because it is connected to three very early Christian monasteries and we have records of when they were founded - Bangor in Ireland in AD 555, Bangor in Wales in 525, and Bangor-is-y-Coed in Wales in 560.

But if the monasteries were simply established at places already called bangor, then the term could stretch even further back in time, pre-dating the arrival of Christianity.

Welsh or Irish in origin?

The dating evidence of a term at least as old as AD 500 may point to it having had a very long history in both the Irish and Brittonic languages. As it stands, with the wide and prevalent usage over Britain and Ireland, there is no immediate reason to presume that the term was originally confined to one linguistic region and at some later date introduced to another. This is especially so as most banchor’s are in remote places, and we find no ready association of the placename with any recognisable human element that may be responsible for introducing it (the four monasteries excepted).

A Lack of descriptors

One of the most striking aspects of the banchor Scottish examples is that they rarely have any additional descriptor attached – most places are simply called banchor.

There are a few exceptions, but these places occur near each other, and are likely being given these extra descriptors to make it clear which of them is being referred to. Banchory-Ternan and Banchory-Devenic occur on the lower part of the River Dee in Aberdeenshire. Edinbanchory and Corbanchory occur within a couple of miles of each other. East and West Banchory sit on two adjacent loops on the River Isla. In Wales, Bangor-is-y-Coed and Bangor Teifi are presumably given descriptors to distinguish them from the famous settlement of Bangor itself.

The Irish examples do have a descriptor in 13 of the 24 examples, the Welsh 8 out of 9, but they are refinements on the location, for example, Coolbanagher (Cúil Bheannchair) ‘the nook of Banchor’, Drumbanagher ‘the ridge of Banchor’, Rossbennchuir ‘the point of Banchor’, Cefn Bangor ‘the ridge of Bangor’, Bangor Teifi ‘Bangor on the river Teifi’.

Most Irish examples are a smallish area or townland, not a distinct ‘spot’ except for those which have become towns. This raises the unsettling suspicion that banchor may not even be a ‘placename’ in these cases, but rather something in the land which is notable akin to a road or farm or a ridge. For example, at Drumbanagher, there may never have been a ‘Banagher’ placename per se, only a ‘banagher’ item that the ‘drum’ is above or on.

What we don’t generally find is a form of saint’s name, with the possible exception of the doublet which distinguishes Banchory-Ternan and Banchory-Devenic. The church was never backward in putting their name on things – Kil X, Both X, Suide X, Tobar X and so on. This simple lack in itself tells against most of the banchor placenames being primarily named for an unrecorded church with a wattle fence.

Furthermore, the banchor placenames without a possessor tell us that these places didn’t belong to any one person, or any group or institution. (There is one exception, Maevemanougher Ireland, ‘Maeve’s Banchor’, Maeve being the great mythological queen.) A banchor has the appearance of an intrinsic element in the landscape.

Monasteries and churches at a Banchor

There is no doubt that a church does occur near a few Banchor places. Fiona pointed out one in northern Pictland on the upper Findhorn, and one near Newtonmore, to which I would add Banchory-Ternan which has evidence of an earlyish (8-9 century?) chapel with a simple cross-marked stone and stories of a handbell. These churches however are all at road junctions, two important enough to become modern towns, exactly where we’d expect a church to be located, making a causal correlatation between the church and the banchor unlikely.

Fiona also mentions two churches one or two miles away from a Banchor. The problem here is that Scotland has a chapel within walking distance of most settlements, so unless many or most banchor’s include a church at that site, it’s very risky to claim any causal connection.

Overall, despite there being four significant monasteries at a Bangor – in Ireland at Bangor, Ulster, in Wales at Bangor and Bangor-is-y-Coed, in Morbihan Brittany at Bangor, this is very few compared to the complete body of 51 banchor placenames. I have noticed a trend, that because we have monasteries at four Bangor places, that when discussing other instances of Bangor, it is noted that there is ‘thought’ to have been a church present, and then this suspicion is repeated by the next researcher as a potential presence of a church without realising there is no evidence for a church except for a bangor placename. We need to be constantly alert to this type of circular thinking.

It does leave the question open of why these four early monasteries were founded at a Bangor, two in Wales, one in Ireland, one in Brittany. Is it just coincidence? Or is there a third factor driving the connection? Only one of these monasteries lay on a significant loop in a river from which it will have got its name, Bangor-on-Dee, also known as Bangor-is-y-Coed ‘Bangor below the forest’. Bangor in Ireland has no river, Bangor in Gwynedd has a river but no loop or bend, and Bangor in Brittany is on a loop of a small stream.

Around AD 613, 1200 monks from Bangor-on-Dee were slaughtered by the pagan Northumbrian king, Æthelfrith. At the same time, the monastery itself seems to have been destroyed as a few survivors are next recorded at Bardsey Island. We cannot underestimate the impact of this horrific mass murder, the worst such event in the history of Christianity in Britain. Even the English scribe, Bede, writing a century later, emphasises the catastrophic nature of this event for the early church. Perhaps it was this horrendous martyrdom that the other early monasteries of the day were commemorating by using the same title, Bangor. This may explain why, while Bangor-is-y-coed is on a typical river loop, the other Bangor monasteries are not.

The consistent lack of churches at most banchor’s, apart from the four monasteries, means that we can safely rule out any relationship to a wattle fence around a monastery or church.

The Welsh common noun bangor

One proposal for the meaning of bangor is that it relates to a wattle fence around a monastery. Although only 4 out of the 51 bangor’s are at a monastery, the idea still needs further examination to make sure we can safely dismiss it.

Welsh has a common noun bangor, which Bedwyr Lewis Jones explained as ‘an upper plaited rod to strengthen a fence’. The GPC describes it as ‘the upper row of rods, thicker than the rest, in a wattle fence, that strengthens, or locks the inner parts together’. Fences to this day are strengthened at the top like this.

The plaited rod is not the same as the wattle fence in common speech. So to expect that a specific and technical noun like this is then extended to the whole fence and then to a monastery, then spread as a popular placename all over Britain and Ireland even without a monastery or church, is so unlikely it must be rejected.

It’s also relevant to note that this technical device to strengthen a wattle fence is not used in building a house with wattle-and-daub. It is a specific device for a fence.

The lack of churches at most Bangor’s has led to a suggestion that the wattle fence may just be a wattle fence around a house. When artists draw an imagined picture of a settlement for the archaeologists, they almost always surround the house or fort or church with a wattled fence or hedge. It is a necessary part of daily life, to keep animals in their proper place. The fence stops the pigs eating the vegies, the goats and deer eating the precious apples, the kids from traipsing dung through everything, and the crucial freshwater spring from turning into a quagmire. It doesn’t seem likely that just a few houses then would be named for having a wattle fence around them, when probably they all did.

When we look at the full context of all the bangor’s, there’s no logic to support a technical term for part of a wattle fence being used early and widely as a placename all across Ireland and Britain.

The linguistics

So this is the part where angels fear to tread. I need to address the linguistics of the dilemma. But it’s not going to be an etymological deep dive, it’s going to look at the proposed linguistics in light of the data above.

The benn/ban element

In both Welsh and Irish, the word is considered to be composed of two parts, ban/benn and cor. The benn (Irish) or ban (Welsh) element is straightforward, a common term in both languages meaning point, or pointed, peak, summit, having a sticky-out bit. Irish benn can be used for a cow’s crescent horns, as in Finnbennach, the famous white horned bull of queen Medb. This application of benn to describe a cow’s horns lies behind the Dindshenchas explanation, that banchor means a place where a cow cast off (cor) its horns (benn).

The cor element

It’s the cor element that poses difficulties, and from that the meaning of the composite term.

Irish cor has many translations into English, eDIL has organised them into 20 groups, and takes some 7 screenfulls of notes to cover the translations. Some of these meanings are put, set up, cast, throw, twist, coil, detour, leap, throw, and so on. Languages often have these basic words that have such wide and general meanings. The problem is that different languages chose different basic verbs, so translating between languages becomes problematic. For instance, the general verb ‘come’ in English requires a choice between 256 different Indonesian verbs! If I had to come up with the basic sense of Irish cor, I would say ‘put to the side’.

So here we have the literal translation of Irish beannchor, ‘point put-to-the-side’. And, here’s the problem – what on earth does the composite term mean? The Dindshenchas explanation as the ‘horn-cast’ may well be demonstrating that even medieval linguists had lost a secure grasp on its meaning. Or perhaps the story was more appealing.

Perhaps we should put it into the same category as the Welsh proposal of ‘High Choir’. Or the Irish idea that it was a hill on which a crime was committed.

Turning to the Welsh case, a lot of ink has gone down trying to make sense of this element. The Dictionary of the Welsh language reports that bangor as part of a fence includes côr (3), with the meaning ‘plaiting, binding; boundary, limit’. But, the entry for côr consists of only two lines, saying that côr is never found on its own but might be the element in a few words such as our bangor, and amgor (a surrounding border), cored (weir), ysgor (score). My overwhelming impression is that côr has either been lost in Welsh as a productive term, or that it never existed.

Looking for equivalent Welsh terms for actions that Irish cor covers, a multitude of different verbs are employed. The difference between the development of the two languages here seems quite stark. But, that still doesn’t quite rule out that bangor was used within a landscape sense related to a river loop in Britain prior to around AD 500. The shift in usage in the Brittonic of Wales and in other language regions may have developed after this period.

I’ve already thrown out the ‘wattle fence’ option. But what about the ‘horn-cast’ option? It too has never quite sat right. A banchor undoubtedly often sits beside nice loops or bends on a river in Scotland, but does bangor really just refer to a loop of a river, or is something else going on here? And if it is referring to a river loop, then what’s so special about just those loops when rivers frequently wind around all over the place? In Ireland, the banchor’s rarely sit at a river loop, which led the Irish scholar Joyce to suggest the benn here refers to a pointy hill or rock, the ‘vertical’ benn, but that idea has not found any supporting evidence since. How could Ireland and Scotland have two such different applications for patently the same term? The story really isn’t making sense yet.

Looking for a third option

Sometimes, when a dilemma is intriguing us, when we’re faced with a puzzle difficult to solve, we tend to get fixated on what we are told are the options to choose from. Sometimes we need to sit back for a moment and look around to see if something else is staring us in the face, daring us to notice.

I have to give a shout out to Andy Sweet, whose comment on Fiona’s blog made the same outside-the-box observations as I am about to build on:

… there may still be some river-specific significance to the Banchor names, just different to what Watson concluded. Also if the placenames were referencing the river bends one might expect the places to be slightly closer to the river than many of them actually are.

So, I’m going to try here for a third option, with a closer examination of Bangor examples. I will divide these examples into a) river loops b) river bends c) lochs and d) fords.

Banchor on a river loop

Many paths in Scotland follow a river route. The river provides guidance, leading people on through farms, fields and villages to their destination. Pictish symbol stones are concentrated along river valleys, indicative of where the focus of settlement and activity was. But what happens if the path comes to a big loop in the river? The path takes the quickest way, straight across the neck of the looped area. It was faster than walking the long way round the point, and it often provided solid drier ground given that river loops tend to be soggy and tick-ridden river meadows. So perhaps in the river loop situation this is what a banchor is - a path that turns aside from the river, that takes a shortcut, a detour, across the neck of the loop.

This may be why only certain river loops have a banchor placename. Because it is the path detouring the loop that is the referent.

This meaning of cor as a ‘detour’ is number 9 in eDIL, ‘twist, detour, circuit in road’. Several examples are presented where a deviation in a pathway is demonstrated, including: Cur sligheadh ‘a detouring path’, do bretha cor n-imruill fothuaid ┐ fodess do na sluagaig ‘he led the hosts astray’.

This also explains why banchor placenames are always on the inside of a river loop, and not close to the river but up near the neck.

If we accept that the banchor is a detour, it may explain some banchor’s that didn’t quite sit right as just referring to a river loop itself. The term banchor is quite ancient and has had centuries to move into the language as a lexical item, with a set meaning of ‘detour’, rather than the sum of its parts, in which case the ‘detour’ may sometimes end up detouring some other obstacle.

I will put below a few examples here to illustrate the river loop context. The main difficulty is that we only have relatively recent maps of routes, so for the moment, it’s safest to pick fairly isolated examples where the path is unlikely to have significantly moved due to the mechanics of the land itself.

BANCHOR ON THE UPPER FINDHORN

The first ‘river loop’ case is Banchor, Strathdearn, on the river Findhorn (NH 76269 23981). I’ve chosen this because, as the next map shows, it is off in the hills, so hopefully undisturbed linguistically, and yet it is at the junction of two routes through the hills, a feature that is found at a few other banchor places.

The next map is a closer view. Here we see the nice loop in the river (the benn), and a path (the cor) which diverts from along the river bank to cross the neck of the loop. I’ve placed a pink dotted line near the path. It is so much easier and quicker for the path to cross here than to follow the bank of the river loop the long way round. Note also that this benn has some internal markings of swampy ground but no sign of fields, also typical of many such looped spots.

The placename Banchor always occurs on the inside of the loop or promontory, it never occurs on the other side of the river at the point of the benn, which we might expect to happen sometimes if the point was what the benn was referring to.

The Dalarossie church is on the opposite bank and to the east of Banchor. Dedicated to St Fergus it was first mentioned in 1224. If the banchor name is ancient, from as early as AD 500, that allows up to 700 years between the two. Also, the river here is fast-flowing, not easy to cross, so connecting the two is logistically problematic. The church however might be expected at a junction of routes like this, and as Fiona pointed out, there is a chapel near the Banchor at Newtonmore, also at an important junction, and at Banchory-Ternan with another chapel and route junction. The causal link here would be the church at an important junction, rather than a link between a chapel and the Banchor.

BANCHOR and LITTLE BANCHOR on the MIDDLE FINDHORN

Banchor, a fair way inland on the River Findhorn is a fascinating example. Here we see another form of the benn, this time not so much a loop but a turn in the river which creates a promontory as a point. The path here cuts diagonally across this promontory.

The placename is marked above a steep bank, which would have blocked the path from following the river edge, causing a detour. This blockage of the river bank path by a natural obstacle is also a feature found at a number of other banchor’s. It’s a nice clear example of a banchor in action.

Again there is no chapel on this Banchor. The nearest chapel is on the other side of the river, more than 2 kilometers away as the crow flies, but far more than that to reach it over a ford, more like 4 kilometers. The Findhorn is a notoriously dangerous, spating river, risky to cross at the best of times. The Princess Stone sits 4 miles downstream, possibly overlooking a ford, with a story of a drowned princess, very apt for this dangerous section of the river. The chapel near Dunearn is also on a different route on the other side of the river, again indicative of there being no association with the Banchor.

But it gets even more interesting, because immediately upstream of Banchor is a second small loop, and another path crossing its neck, this time marked as Little Banchor. At first I had taken this to be a modern farmlet named after the bigger Banchor next to it, but as Little Banchor also marks another path across a river loop, I suspect this may be another valid Banchor in its own right. If this is so, it might suggest that Gaelic speakers knew the meaning of banchor perhaps until quite recently and we just weren’t quick enough to catch it.

EAST and WEST BANCHORY on the River Isla

Just to the west of Coupar Angus are two wonderful river loops, East Banchory and West Banchory. The modern road is an ‘old military road’ and is present as far back as maps go, as are the two Banchory farms. This route is quite crucial as it starts at a ford (with chapel) over the river to Coupar Angus, then passes west across the two river loops, past the end of the Cleavan Dyke and the Roman camp, then tracks along the north bank of the Tay all the way to Dunkeld. The logistics of this road points to it following an ancient route.

BANGOR-ON-DEE (BANGOR-IS-Y-COED)

This part of the River Dee in Wales has a couple of lovely loops, the monastery in one, a modern racecourse in the adjacent one. In this flatter worked over area it is hard to know about ancient pathways, but we should assume that one went to the monastery, and so it would have passed along the neck of both loops, much as the modern road does today.

Early records show the presence of a bridge over the River Dee at this point, and it seems likely this dangerous river crossing was the purpose of the monastery being placed here. The river majorly floods in winter and spring. Removing many tons of flotsam that has caught on the bridge during winter is a constant modern concern.

A number of other banchor’s are at fords too. But the one thing I have noticed while looking around on the internet for descriptions and photos of banchor’s, is the constant mention of a problematic river that floods and spates. The Findhorn with its three Banchor’s is a prime example of this type of river. Add to this that most banchor’s are on large significant rivers, not mere burns or streams, and it suggests that we may be dealing with not just any river, but one that presents a significant flooding problem at those locations.

A closer look shows the church of St Dunawd on the supposed site of the ancient monastery, and it sits on the route (High Street/Graig Lane) crossing the neck of the river loop as expected of a bangor path. The church site is, as expected, close to the crossing point where we now see the modern bridge.

THE LOST BANCHOR near Inverness

Watson mentioned a lost Banchor which he thought may have been ‘below Dochgarroch’ (by which I think he means downstream). Given the pattern we’re looking at, I’d suggest this was the path across the river loop downstream from Dochgarroch.

The A82 runs further inland, but traces of an older path can be seen on the map – look for the D of Dunaincroy. The mix of public road, farm track and path is characteristic of old routes that have neither been lost nor become main modern roads. The element -croy (cruaidh) suggests the possibility of a word for a paved (literally hard) road: cf. Croy Hill near Falkirk where the Romans’ Military Way runs along Croy Hill beside Antonine’s Wall. It could be argued either way, that if this meaning for -croy is credible, it makes another routeway term less likely or more likely to have survived. (Thanks to Bill Patterson for this information.)

Banchor on a RIVER BEND

Sometimes the river is more of a bend around a point, rather than a rounded loop. The second example above of Banchor on the middle Findhorn shows the path cutting across this loop which is arguably more like a bend.

BANGOR TEIFI on the RIVER TEIFI in WALES

Bangor Teifi does not have a history of a monastery. The river loop is a significant turn in the river, and a path crosses this promontory through Bangor Teifi. Again we find the bangor not down at the point of the loop, but up on the detour across it.

This detour is forced by the presence of the fairly steep hill Bryn Teifi and the narrow gorge the river flows through. The Teifi is yet another river notorious for flooding.

TULLYBANNOCHER on the River Earn

The bend here is the area enclosed by the green road, which is acting like a modern banchor. The path had to detour across the bend, first being forced upwards by the small steep loop (marked on this map with ‘weir’). It then takes advantage of that detour by skirting the hill of Drumlochlan before returning to the riverside.

One of the issues here is that when another landscape element is added to the banchor placename, it’s hard to tell where exactly the original banchor was, except to say ‘nearby’.

Even today, modern councils are still busy applying flood protection measures at Tullybannocher on the River Earn.

Banchor at LOCHS

There is another group of banchor names, definitely not on a river loop or bend, but on or near a loch. These are found in Scotland, Wales, and Ireland. All of these lochs are small, narrow lochs, between high mountains. They actually present a consistent picture. All of them, being small and taking water off mountains, are subject to rising water levels, and even flooding and overflow.

Watson suggested that the ‘horn’ in this case could be the pointed ends of the lochs, but that has never sat well.

What is exceptional about these small mountainous lochs is that they have a narrow footpath clinging to the edge of the loch, and I’m going to suggest that it is again this path that is the ‘banchor’. The path is presumably just high enough to avoid the rising water levels.

In Wales, at the southern edge of the massive Cadair Idris lies a small narrow loch, Llyn Myngul or Tal-y-llyn Lake. Bwlch bangor ‘the pass of bangor’ is here but not marked on a map that I could find. Part of the loch’s edge is flanked by steep slopes, making the path difficult. (The green dots on the map mean that cyclists must share the path with walkers.)

There are three similar small narrow lochs flanked by mountains in Scotland, Loch Venachar, Loch Beannacharan, and Loch Beannachairan. There is also a lost ‘Lacus de Benchoir’ near Peterculter, between Banchory-Ternan and Banchory-Devenich, another stretch of river prone to massive flooding as recent events have shown. All have a narrow pathway along their flank.

The next map is LOCH BEANNCHARAN in STRATHFARRAR

Next is LOCH BEANNACHARAIN in STRATHCONAN

In Ireland, there are several banchor’s near a loch, but because most are marked as large general areas it’s hard to tell exactly what is the original referent. There are several Drum-banagher, similarly to Cefn-bangor in Wales, suggesting the banchor is along a ridge or high place overlooking a watery obstacle.

Cnoc Beannchar (Hill of Banchor) sits above Banagher Lough, Donegal, the loch again being small, in high moutains, today with a tourist walking path around it. “It is stated that there has been crimes committed on this mountain which it has derived its name from”.

PENALLTFANGOR in WALES

Penalltfangor in Wales, ‘the head of the cliff of the bangor’ isn’t on a loch, but may have relevance. Here the path arrives at the lip of a high cliff dropping down into a narrow, steep-sided gorge of the River Cywyn, before picking up the path on the other side. To get over this gorge, the path takes a wild dive with zig-zags to the bottom of the gorge (marked in green dashes on the map). This presumably is the ‘bangor’ in question.

Banchor at FORDS

BANAGHER OFFALY on the RIVER SHANNON

Banagher, county Offaly in Ireland, is one of only a few possible crossing points over the mightly river Shannon. It is also known as Áth Cróich. Here a large esker (a ridge of gravel thrown up by an ancient glacier) provided access through the marshes on the riverside in all seasons, even when flooded in winter. (Great photos here).

On the banagher.ie website is a fascinating document. In the early 1600s, Matthew De Renzy, a learned Irish scholar who lived nearby, wrote of this ford over the Shannon, and in several places refers to it as ‘the benghar’. By using the article ‘the’, he is implying that his banchor is something specific in the land, not just a placename. In one of his items, he seems to explain that the banchor is the passage on the esker giving access over the river:

‘At the Benghar there ought a towne or a good fort to be made, to keep that passage of the Shannon, for that in no other place can come any horsemen near the river to take passage out of Connaught but only here, by reason of the impediments of the bogs and woods; from this passage it is but 30 or 34 miles to Galway all hard and fair ground.

This isn’t the only banchor at a ford, in fact I highly suspect a number of other banchor’s were also some kind of raised causeway at a ford, especially in Ireland where banchor’s so often abut a river. However, if so, this did not refer to a hurdle clíath ford, in fact there seems a lot of confusion about what that would technically be. But I do seem to be coming close to a tóchar which was a raised path sometimes paved with stone or wood.

BALLYBANAGHER

Ballybanagher (Beal Ath Beanchor ‘the mouth of the banchor ford’) is today known as Corrofin ‘the White Ford’. Cora means ‘stone-fence, palisade, wall, row or serried rank’, and by extension the ‘weir’ enclosed by the stone wall, and the ford over which the stone wall provides a type of causeway or crossing of the river. It raises the possibility that a banchor was a type of stone wall acting as a type of ford.

BANGOR, GWYNEDD WALES

Which brings me back to Bangor in Wales, the site of the early monastery. Bangor has a river, but no loop or bend to suggest a bangor. Again I wonder if we always get blinded by the presence of a Christian site, when the logistical reason for the position of Bangor is that from here people have crossed the Menai Straits into Anglesey. Was the original bangor, like the cases of Bangor-on-Dee, Ballybanagher, and Banagher Offaly, referring to a ‘ford’ or crossing over the Menai Straits?

Before the modern bridges were built across the Menai Straits, crossing here was highly dangerous. At low tide, people could walk or wade across, a heart-stopping undertaking. The constantly shifting sandbars made even a ferry boat crossing perilous. Professor Delyth Prys once took me on an outing to see an ancient church on the coast opposite the Lavan Sands and told me that scholars suspected there was once a type of causeway across the straits here marked by the church. Now, I’m beginning to suspect this may not have been merely a sand crossing (similar to the crossing at Lindisfarne), but that it may even have been a form of stone causeway used at low tide.

Why might we have forgotten about these bangors crossing such dangerous waters? We seem to so easily forget the past once a modern bridge has removed the danger before us.

Oh you’ll take the high road …

In the end, my proposal is that a banchor is a high path that usually skirts a watery obstacle, especially one that is subject to flooding. On river loops and river bends, a banchor cuts across the neck of the bend, avoiding the longer, wetter, route following the edge of the bend.

On small narrow mountainous lochs, especially ones that are prone to sudden rising water levels and flooding, a banchor can refer to a narrow path that skirts along the side of the loch.

A banchor may also be a raised path to ford a river or strait.

There are still other banchor examples whose context isn’t totally clear. Yet the range of accompanying terms such as ridge (drum/cefn), hill (cnoc), point (ross), rocks (carrick), marsh (waun), cliff (allt), pass (beal/bwlch), ford (ath), and that most banchor’s are on elevated land, suggests that here too we are looking at a path that diverts over higher ground to avoid an obstacle, usually a water obstacle.

An element of confusion has been inserted by four early monasteries using the term Bangor, provoking the notion that this church usage has been the driving force behind the common placename. But it seems the reverse is true, that at least the Welsh monasteries may have been drawn to these sites to offer divine protection for dangerous waters over which a banchor allowed passage.

A special thanks to Bill and Fiona for their reading and commenting on this research over many drafts and sticky issues. All errors on me of course.

But, the Pictish silver chains all seem to be restricted to the northern section of Aberdeenshire, Buchan, and Moray, between the Mounth and the Moray Firth. No silver chains have (yet) been found in Pictland below the Mounth, or above the Moray Firth line. Which raises the question of why just this region? Were the ‘silver chain Picts’ not in control of other regions? Did they not have enough silver to go further? Were the northern Picts the only group that had access to large amounts of silver? Or was there a philosophical reason why the silver chains were only relevant in this region?

A second key point is that while the massive silver chains can be dated from the AD 300s to mid-400s, we also see this concept of the silver chains coming from much earlier, as expressed with the earlier bronze/findruine chains, in the case of Skye around 790 – 539 BC. And we see their elite significance recorded in Irish stories and texts.

In other words, the silver chains are contemporary with the use of Pictish symbols.

In which case, we might ask the next question – if the chains represented such a critical element in their worldview, might there not also be a Pictish symbol representing a silver chain?

In searching for an answer to this question, one symbol stone in particular stands out, Sandside House. Given that Pictish symbol art is quite abstract, as we see for example with the cauldron symbol which is depicted as a large circle with two small circles as the handles, could it be that the triple-oval symbol is an abstract depiction of a chain?

The triple-oval symbol

The earliest form of the triple-oval symbol likely appears on the very simple stone of Wester Balblair. This stone has an unsmoothed surface, both symbols have no internal decoration, and the ends of the V rod don’t yet have floriations – all indicators of a very early date, in fact, this must be one of the earliest symbol stones of all. The shape of the triple-ovals are more pointed at the end, more like lozenges or triangles than rounded ovals.

Other examples of this symbol are similar to Balblair, being nearly triangular. Only the late intricate stones of Keiss and Sandside House have fully oval forms.

A key point though is that all instances, except Wester Balblair and the two simple instances in the Sculptors Cave, have some sort of round connector running through their centre – which could represent the connecting links in a chain. Here we see that while each stone’s decoration is individual, it is still expressing an underlying concept.

Comparing the landscape context of the triple-oval to the silver chains

How can we go about testing the idea that this symbol might be acting as a symbolic silver chain? One way would be to see if the instances of the triple-oval sit at points that might be considered an entrance, or guardian point, to various regions, just as the real silver chains do.

First, here is a distribution map of the triple-oval symbols.

Next, we can add the findspots of the real silver chains (silver stars) to this distribution map.

The two distributions of symbol and chain are largely complementary, which is an excellent start, as it implies that when a silver chain is needed, but a real silver chain is not available, then a symbol chain can take its place.

The proviso here though is that the data is quite sparse for both the silver chains and the triple-oval symbol, so there may be a number of other explanations for this complementary distribution. Still, it’s a good start.

The triple-oval symbol as guardian of the land

The next step is to look at the triple-oval findspots in detail, to see how they might compare to the pattern observed with the silver chains – near the coastal entrance to a region, usually not far up a river estuary.

The most northern three triple-oval symbols are in the north of Caithness – a region which I have already spoken about as potentially having Lugh as their ancestral deity.

The Sandside House stone sits beside a small harbour, an excellent location for boats entering the land.

And it is the most north-western attested symbol stone, close to the modern boundary of Caithness and Sutherland. The placename Reay has several possible derivations, one being the Norse Ra, a boundary marker. This is precisely the same situation we found most of the Pictish silver chains at – an entrance to the land, and a boundary between regions. The Sandside triple-oval symbol, and its harbour and boundary location, fit well with it being symbolic, and functioning, as a silver chain.

The next example is Keiss. This location too can be seen as an entrance to northern Caithness, this time in the east. Keiss lies on the broad Sinclair’s Bay, from which boats may have traveled to and from Orkney. Perhaps what makes this location special, is that the symbol of its pair is a cauldron, which usually sits beside a shallow loch or marsh, in this case the Loch of Wester, which may have provided a sheltered place to pull boats into. This fits well with a silver chain type of location.

The third triple-oval symbol sits at the farm of Skinnet, on the Thurso River. It is the only case of a Christian cross with this symbol. Oddly, if we were looking for a guardian location, the most obvious place would be at Thurso itself, yet Skinnet is five miles inland on the Thurso River. It may be, as related to me by a local historian, that this cross has been moved inland under later pressure from the Norse.

These three locations, Keiss, Skinnet (Thurso?), and Sandside, represent three key entrances to Caithness from the north. Points where it could be considered necessary to call on a deity for protection with a triple-oval symbol as a symbolic silver chain.

The fourth example of a triple-oval symbol is at Beauly, just inland on the river. Beauly has another early Pictish stone, settlements and cemeteries, as an important entry point for movement northwards. This isn’t far from the findspot of the Torvean silver chain, again just inland on the river Ness, putting both symbol and chain in a parallel landscape context. The Inverness silver chain location is the furthest north-west in the silver chain distribution. In other words, where the real silver chains end, the triple-oval symbol begins.

There are two triple-oval symbols on the walls of the Sculptors Cave on Moray coast near Burghead, a ritual cave with activity dating between the 2nd and 4th centuries AD. This date also overlaps with the proposed dating of the silver chains, late 4th to 5th centuries AD. In contrast, there are no examples of this symbol in the southern caves of Fife. It may be that these Covesea symbols are guarding the northern coast of this area of Pictland, or alternatively, it might be that here they are performing the strengthening function seen in the older structures (which I will discuss later). Or, maybe both, at the same time.

The southern triple-oval symbols of Glamis and Aberlemno

The Glamis triple-oval is a fragment whose exact findspot is unknown. It might be argued that Glamis sits near the start of the Dean Water, which flows westward to Meigle through what appears to be a large nemeton. Perhaps this might be regarded as the entrance to the western part of southern Pictland. But it’s unclear.

The Aberlemno example is extraordinary, although again its original location is not known. The triple-oval here is overlain by the Z-rod of another symbol, and the stone itself is badly laminated, and gives the impression of being a re-used ancient stone, perhaps from a chambered cairn. Aberlemno is undoubtedly a key place in the Pictish landscape, and again is beside a large ancient nemeton. Perhaps we might view this central Pictish place of power and religion in the same light as the centres of Votadini territory – Edinburgh and Traprain Law – but again it’s unclear to me.

The key thing is that these two examples of the triple-oval at Glamis and Aberlemno are found in the southern Pictish region, where no massive silver chains have been found, suggesting the triple-oval might be standing in as a symbolic silver chain in this region. Both stones would appear to be very early Class I (pagan) stones, perhaps dating to the early centuries AD.

Summarising the case for the triple-oval as a symbolic silver chain

There seems to be a case building for the triple-oval symbol functioning in the same way as a silver chain. But the examples are too few to make a firm conclusion. The northern examples of the symbol all seem to be at spots that we might expect of a silver chain, but, it’s hard to draw a causal link here, as the spots are naturally important entrances to the land, where we might expect a symbol stone regardless of the symbols on it.

For me, I think this is still a distinctly possible identification of the triple-oval symbol, but not proven. One to keep in mind as research goes forward. Hopefully, more evidence either way will present itself in the future.

The Norrie’s Law plaque

While my colleague Bill was reviewing this blog, he made a startling observation. He pointed out that the plaque found in the Norrie’s Law silver hoard in Fife is a similar shape, although slightly more teardrop. Is this another instance of a Pictish ‘lozenge’? It wouldn’t be the only time an outline of a symbol was found with internal symbols – the bronze plaque of Monifieth was shaped as a crescent symbol inside which are the same two symbols as the Norrie’s Law plaque, with another crescent on the reverse.

The original reports of the hoard imply that many of these same plaques were found originally, although only one survives today.

The dog-head symbol is rare and is only paired with the double-disc/Z symbol. The same pair is found on this plaque, on the Monifieth plaque, on a symbol stone at Rhynie, and in the Dovecot cave on the coast of Fife. The odd lack of arms on this dog-head symbol might point to it being primarily an otter, although in Celtic terms an otter is a form of dog – a water dog. The double-disc/Z symbol on the other hand is common as the second most frequent symbol. Given the context of this series of blogs, I can’t help but think this important pair of symbols reads as CUNO-MAGLOS ‘hound lord’, which we have seen is a major epithet of Lugh in his aspect as guardian hound of the land, and a name used by several kings including Maelgwn of Wales, and Bridei mac Maelcon of Pictland.

Other ideas about the triple-oval symbol

For those of you who have read other articles of mine, you might be saying, but haven’t you written before that the triple-oval symbol could be functioning in the same way as the earlier ovals/triangles/lozenges on Neolithic stones, as tokens invoking a strengthening of the stone itself? Yes, I have.

So, I’m going to present that theory here too. Then we can compare and contrast the evidence.

At the end of this discussion, I’ll have an appendix with details of each symbol stone.

Lozenges or diamonds are a common motif in Neolithic passage mounds, in Ireland and Britain. They are also prevalent in the buildings of the Ness of Brodgar. Commonly, they are found on the narrow edge of very large, load-bearing stones, even when those stones are themselves buried deep in the structure, not intended to be seen. The implication is that these are not a ‘nice’ decoration, but a form of protective device, to give strength to the vulnerable stones that need it most.

Lozenges are made of two triangles, and it is the triangle which is the only geometric form that cannot give way and deform under pressure. Its strength and rigidity is still relied upon in modern architecture (for example the Sydney harbour bridge depends on a series of triangles to support its massive weight). So it is easy to see what the basis of this symbolic carving is. However, it is not known how this fitted into the broader philosophy of Neolithic times. Presumably a deity was invoked along with the symbol to give it context and significance.

Lozenges at the Ness of Brodgar

Triangles and lozenges are notably found on many stones at the Ness of Brodgar in Orkney, almost to the exclusion of any other form of carving. Some are found on the side edge of large flagstones, as we see in other Neolithic buildings. Originally thought to be part of burial cists, they are now understood as elements within a building.

And, there are many other stones with lozenges or triangles. The buildings at the Ness were huge and exceptional, but had an inherent problem of being built on unstable ground, and the assumption is that the lozenges are still continuing the practice of strengthening the stones with this motif.

Here is an example found below the buttress of Structure 10.

Structure 10 had been raised over the site of two other collapsed buildings. To try to stabilise this new building its corners were fortified with buttresses, and under one of these buttresses the above stone with lozenges was found, presumably an invokation to help strengthen the buttress and its building.

The website records the fate of Structure 10.

In the centuries before the construction of Structure Ten, subsidence had been a problem at the Ness of Brodgar complex…

Undoubtedly to negate (or delay?) the impact of subsidence, huge stone slabs were used as foundations for sections of the building’s huge walls.

But it made little difference.

Around 2800BC – perhaps less than a century after its construction – Structure Ten collapsed. Its south-western corner, in particular, was badly affected.

A continuation from the Neolithic?

Many ancient lozenge carvings would still have been visible to the Picts in passage cairns. The question then is whether the Pictish symbol is still being used in a similar way, to invoke protection and strength to some stone or building?

Here is a summary of possible relationships between the ancient lozenge and its protective function.

Just inside the mouth of the Covesea cave, it splits into two passages, and it is on the roof at the junction of the archway between these two passages that a triple-oval is carved. Other symbols are usually carved on reachable cave walls, not on the roof. The symbol is carved at a critical point in the cave structure, the point most vulnerable to collapse, a situation which recalls its protective function in the ancient Neolithic cairns.

At Keiss, Sandside House, and Wester Balblair, any buildings of a settlement would have faced the difficulty of their foundations being on unstable and mobile sand. If the Skinnet stone has been moved from somewhere at Thurso with its sand dunes, then it too may have faced this difficulty. The triple-oval symbol in these cases could be argued to be performing a function similar to the Neolithic lozenges, as a protective and strengthening invocation.

Glamis also has sand blown in from the edge of Loch Forfar and its surrounding marshes, but the soil here is a mix of loam and sand, so it’s not as clear whether any building would have been thought necessary to have a strengthening symbol. However, there is the odd tale where Glamis castle was built on unusually low ground because previous attempts to build it higher up had been thwarted by its stones falling down. And it’s fascinating to wonder if this story has a basis in some earlier Pictish experience which was later overcome by the triple-oval symbol’s presence.

In the end, there does seem to be good reason to consider that the triple-oval symbol is functioning in much the same way as it has since Neolithic times – as a protective and strengthening symbol for structures.

But if this is so, it’s just the beginning of the conversation. For there is no doubt this symbol would also have been placed in a larger ritual and religious context to provide it a rationalisation for ongoing usage.

Comparing the two ideas about the triple-oval symbol – protection or strengthening

Now comes the quandary. Which of the two ideas above best suits this triple-oval symbol. Is it a protective motif called upon to help strengthen a stone or building from breakage and collapse? Or it is a symbolic form of a massive silver chain, used to protect the entrance to a region? Is one or other of the two ideas right, or both wrong, or both right?

At first glance, the two proposed functions don’t seem to fit together neatly. However, they may have one thread in common – their connection to the deity Lugh.

I’ve discussed in a previous post how the silver chains may be associated with Lugh in his aspect as a hound, Lugh’s silver chain being a magical elite collar. The massive silver chains appear to have been ritually placed at the entrances to various regions in Pictland and Lothian, giving rise to the assumption that Lugh is being called upon to act as a guard dog of the land.

On the other hand, the use of carved lozenges as a protective device is often found in Neolithic passage mounds, where it is placed on the narrow edge of any weight-bearing large flat flagstone. So, it’s reasonable to propose that this same motif is found in Pictish places which have the same vulnerability. And it is with this flagstone that we find another hint of Lugh’s role.

When Lugh arrives at the court of the Tuatha De Danann, their strongman and champion, Ogma, challenges Lugh in a unique contest of strength - Ogma hurls a massive flagstone, which needed eighty oxen to lift, out over the wall of Tara. Lugh responds by turfing it back in.

This contest of strength between Lugh and Ogma involving a massive flagstone is peculiar, not seen elsewhere in stories, which emphasises its significance. It may not be much to hang our hats on to provide evidence that Lugh is the deity invoked by the lozenges to strengthen the flagstones, but for the moment, it’s the best I can find.

The ancient Celtic worldview must have been complex and mature, its deities having major roles that we can only glimpse dimly today. So, it’s not impossible that Lugh may have held both ‘strength’ functions, as protector of stones, and protector of the land.

Of course, that just complicates any attempt to understand the symbols, as they too may have been multi-purpose with complex back stories. Perhaps in the end, choosing between options is unnecessary, as both may well have been appropriate.

I can even imagine a scenario where the more triangular Pictish symbols are still functioning to strengthen stone, while the two more oval symbols at Keiss and Sandside House, both much later stones in the far north, are functioning as a symbolic silver chain.

As always, it’s an ongoing work. No doubt in the future, we’ll find more evidence to help us understand better.

That’s the end of this article. Below I’ve put more details about each location of a triple-oval symbol.

Subscribe now

Share

APPENDIX : DETAILS OF EACH TRIPLE-OVAL SYMBOL

Numbers here are the identification number in TROVE

COVESEA CAVE, MORAY 1358778

In the Covesea cave on the Moray coast near Burghead, a ritual cave full of human bones, a pair of symbols has been carved, one of which is the triple-oval symbol. The symbol pair sits, not on the wall like other symbols, but on the roof near the entrance to the cave, on the edge above where the passage splits off into two. In other words, this triple oval is precisely in the same sort of location, possibly for the same reason, as Neolithic lozenges. The implication is that it is performing the same function - as a protective symbol giving strength to the roof at its most vulnerable point.

There is another unpaired instance of a simple triple-oval symbol high on the rear wall in the west passage. There is no known instance of the symbol in the southern Fife caves by contrast.

KEISS (BIRKLE HILL) 9131

The Keiss symbol stone was found “on the summit of a sand-blown rise”, with a shell midden as its base. This large sand dune together with a second one to its south is known as the Birkle Hill. The symbol stone had been re-used as paving in a structure built into the top of the sand dune, although it’s unclear whether this structure is Pictish or Norse. Several cist burials were also found, which are presumably Pictish, given that there was no “indication that fishing was found nor was there any indication that fishing had contributed to the diet of the inhabitants”.

The Linglas broch 9130 stands on the second part of the large sand dune at Birkle Hill. Although there is no dating for this particular broch, two nearby brochs of Keiss 9318 and Kirk Tofts 9333 both have artefacts that show an origin in the 1st centuries BC and AD, with a subsequent modification in the 2^nd century and again in the 3^rd century. Here we have a potential contextual date for the Pictish stone, either 2^nd or 3^rd century AD.

The key thing of interest here may be that we are dealing with buildings on sand, an inherently unstable ground, and because of this, the Linglas broch, like the other Keiss brochs, is thought to have a solid base.

Keiss also has the fresh-water Loch Wester behind it, making it one of the rare sites for pulling in boats from the open ocean. This also makes it a good candidate for the type of place we find the Pictish silver chains – at the entrance to a region.

SANDSIDE HOUSE 7247

This beautiful stone is finely incised with a double-sided comb and mirror, indicating a relatively late date, possibly the late 600s. It is the only instance of a comb symbol above the Dunrobin stones.

While the Keiss stone was beside a broch, this symbol stone has a disc-over-notched rectangle as the accompanying symbol, which may symbolise a broch.

The stone was located above Port Mor at Reay, the beautiful bay here being an ideal safe haven for boat traffic along the northern coast, and a place from which to set off to Orkney. Nearby is a source of fresh water, Lady’s Well. The stone is at the site of an ancient settlement on the sand links by the sea shore.

This whole area of Reay and its hinterland is covered with old sand dunes, creating a strange moonscape. The sand dunes here are notoriously mobile, and like at Keiss, would have been a constant problem to build on.

Although Keiss and Sandside House symbol stones are both associated with buildings built on sand, this is not always the case for other similar settlements and cairns. MacLaren[1] gives an overview of different types of situations while writing about Lambsdale Leans, inland from Reay:

Lambsdale Leans is a natural mound, of elongated shape and composed largely of sand, into which are set the burial and structural remains. Many of the known long cist burials in the north of Scotland also occur in natural mounds. These, as at Reay, and at Keiss (Laing 1870, 38–41) and in the Birkle Hills, near Murkle (Tress Barry 1895), are often sand dunes. In other cases, for example Loch Watenan and Dairy Park, Dunrobin, an artificially constructed cairn covered the cist (Gourlay 1984; Ashmore 1981, 350; Close Brooks 1984, 99–102). Not all mounds utilised in this way were of natural origin; a significant number of long cists have been recovered from broch mounds, such as Crosskirk and Green Tullochs, both in Reay parish (Fairhurst 1984, 101–3; Anderson 1873, 185).

Sandside Bay is the harbour nearest the north-western border between Caithness and Sutherland. This is the furthest westerly point of any symbol stone. Again, it presents the same type of picture as the Pictish silver chains, an entrance to the land, and at a regional boundary.

SKINNET 318992

The Skinnet CII cross slab was found built into the wall of a small chapel of St Thomas.

Skinnet chapel is not like Keiss and Sandside House, built on sand. It is on a broad gentle slope of the river valley, today a farm. There is a well near the road which flows past the chapel, and over on the opposite side of the river are several large chambered cairns. Trees have today been planted around the chapel ruin, but they are only bent shrubs, as the wind howls down the valley. An opinion expressed by a local historian is that this important cross slab may originally have come from Thurso, where we might expect such a stone but where none have been found. It may have been brought down the valley from Thurso, under pressure from the Vikings as they took over the port of Thurso. Thurso and its bay has large sandhills along its shoreline and in its estuary. So it may be that the Skinnet stone also relates to the sand dunes of Thurso.

On the other hand, if the Skinnet stone is in its original location, and that the triple-oval here is acting as a symbolic silver chain guarding an entrance to Pictish territory, it indicates that by the time of carving, Thurso was already in the hands of the Norse.

Unlike Keiss, Wester Balblair and Sandside House, there is no evident settlement near the farm of Skinnet. It only has the ancient chapel. Today, the road from Wick passes through Halkirk just to the south of Skinnet, but local opinion suggests that the road originally continued to a ford on the Thurso River just below the chapel.

WESTER BALBLAIR 12674

Two Pictish stones have been found close to each other at Wester Balblair, the other one being a standing man with club, of the so-called ‘formidable man’ type, suggesting a major Pictish centre.

Both stones appear to be very early, in fact the accompanying crescent/V symbol has no floriations on its rod ends, suggesting this may be one of the earliest symbol stones known. The only other instance of a V rod without floriations is an example found in the same Covesea cave as the simple triple-oval symbol.

There is an “extensive linear prehistoric site” along the low sand ridge of Balblair wood, “it once comprised 13 hut circles, 2 chambered cairns, burnt mounds and a field system of clearance cairns and trailing banks occupying an area of 750m x 200m (maybe more) along the SW edge of the wood”. To emphasise the sandy nature of this river terrace,there is today a massive sand and gravel quarry directly to the south of the stones’ location, which has eaten away a large part of this settlement.

Not far to the north-east is the extensive Pictish barrow cemetery of Tarradale. Although as yet there is no direct evidence that these two sites are linked as the living and the dead areas, this seems likely.

On the east end of this settlement terrace is a small fort, overlooking the river ford. Beauly’s estuary is again notorious for its sand dunes – and seals which birth on them at mid-summer.

Again, we have the triple-oval symbol at a place on which the Picts are building, a place of mobile and unstable sand.

But again, this Beauly location is a similar context to the Pictish silver chain locations – near the entrance of a major river to a Pictish region.

GLAMIS

A fragment with a triple-oval symbol was found in the Manse gardens at Glamis. The stone is made of local sandstone, so presumably came from somewhere in the general area. (Information from David McGovern June 2020).

The geographical location of Glamis appears, at first glance, to be quite different to the sandy settlements we have seen at Keiss, Beauly, Sandside House, and perhaps Skinnet/Thurso. Yet “Strathmore is a fertile area and the soil is a rich loam, with much sand.” The sand in this area has been thrown up from the eastern shore of the large lake and marsh of Forfar. There is even a sand quarry on the estate. Does this put it into the same situation as the other sandy locations? There is a big difference of sandy loam to sand dunes – unless the difference isn’t the driving factor for the Picts. Is it enough to presume the symbol is relevant to the stability of the ground for building on or staying strong?

If this is the case, it also indicates that there was an important early Pictish building at the famous well of St Fergus, today behind the modern church. The lack of erosion of the symbols on a piece of soft sandstone suggests the stone was originally sheltered indoors. The fragment also stands in stark contrast to the two massive cross-slabs also at Glamis, again suggesting a different context for the triple-oval fragment. This wouldn’t be a new situation, as there are other symbol stones which likewise appear to be acting as part of a structure rather than a stone with its end in the earth.

There is a story about Glamis castle, just to the north of Glamis manse, that suggests an odd history of an unstable location to build, and it’s tempting to associate this with a symbol for stability:

The Castle is set in the low plain of the Dean Water, unusual for such an old stronghold as Glamis. Tradition holds that attempts to build on higher land were foiled by the fairies who threw down the stones as soon as they were put up.

In all cases of the triple-oval symbol, we seem to have a constant theme, a theme of instability or weakness, of stones that need to be stay strong, of buildings on sand that need to be strengthened. A theme which goes back to the Neolithic, to the time when engineers were building the great stone monuments of their day. Builders who understood how to construct passage mounds and cairns which are so well engineered they stand today 4000 years on, Builders who also understood the structural weak points of load-bearing stones, and sought to strengthen them with the blessing of the lozenge/triangle motif.

The new Aberlemno6 stone

Recently an extraordinary symbol stone was found near the church of Aberlemno. Along with a number of other large flat stones it had been re-used to pave a later area, possibly of a large building.

The stone is strange in a number of ways. It has a crescent/V symbol, over a double-disc/Z symbol, with a large mirror and comb. It also has a small and simple triple-oval symbol under the upper left arm of the Z rod. Uniquely, it also has another large V rod, seemingly without a crescent. The stone’s surface was not smoothed in preparation as most symbol stones are, in fact it had previously laminated badly, yet the symbols were placed on the stone to respect the lines of breakage. Again, this is unique.

It is possible that this stone originally came from an ancient passage mound with an importance of its own, otherwise it’s not clear why it would have been chosen for a symbol stone. If the stone was originally from a Neolithic monument, then it’s even possible that the triple-oval here was a Neolithic carving rather than a Pictish one. In support of the idea that it was an ancient motif already present long before the other symbols were added, is that its right-most curve is re-used for the middle curve of the double-disc symbol, and this goes against the general rule of symbols, that symbols should have negative space between them.

The implication is that the triple-oval is on this stone for the same reason as the Neolithic lozenges – as a protective device to strengthen the stone and prevent further breakage. Whether it was added by the Picts or their Neolithic predecessors remains a question.

Aberlemno is undoubtedly one of the most important political and religious centres of the southern region of Pictland. So, the triple-oval symbol can’t be guarding the entrance to the region itself. However, we might recall that two major power centres, Traprain Law and Edinburgh, both had massive silver chains guarding them, so perhaps Aberlemno falls into this category.

Subscribe now

Share

[1] MacLaren, Alistair, A Later Prehistoric house and Early Medieval buildings in Northern Scotland, p24

The high level of silver

Although it doesn’t seem all the silver chains have been recently tested for their level of purity, it is generally agreed that most if not all silver chains have a very high level of around 95% silver, the same level found in good quality Roman silver items.

We also know that silver coming into Pictland was subsequently diluted over the years to make it stretch further, once the source of Roman silver had dried up in the 400s. Even before the Romans left Britain, they had been having trouble meeting their silver needs, resulting in a dilution of the silver content of their coinage.

Interestingly, Roman silver was largely mined in Spain. And it was then used to buy goods like silk coming in from China. The empire was leaking silver fast.

We can assume then that the silver of the chains was sourced during the later Roman years, or as raided goods in the years immediately after the Romans left. And, we can assume that the chains were made almost immediately after the silver was sourced, as there is little to no dilution of the silver. So, we’re looking at a date from AD 300 to 450 thereabouts.

This general dating is currently accepted as a first line.

Raid, trade, or paid?

Hoards of silver, and the massive silver chains, have questions over how this Roman silver came to be in Pictland, and this question has relevance for dating.

Hoards tend to be composed of hack silver – silver hacked into pieces to deliberately conform to a system of weights. These can be thought of as payments for services rendered – either as mercenaries, or help in defending borders. A scenario for this could be protection money from native British elites, rather than central Roman authorities, as the empire crumbled and then collapsed. The PAID option.

Given that we find a bronze chain and carvings of dogs wearing a chain collar at Caerleon on the Severn Estuary, it’s a fair possibility that either the Picts or the Votadini were mercenaries. If the silver is connected to this military service, they weren’t paid in silver coins as was the usual case for the Roman army, as the denarii by this later stage weren’t of the required purity of the chains. Nor do we find any other evidence for a lot of denarii coming into Pictland after AD 200. To accept that the chains came from military payments, we’d have to propose that they were paid in hack silver, which seems a little odd.

The TRADE option is one that we have little to no evidence for, but which is a distinct possibility. Both areas of the chains, Lothian and Aberdeenshire/Buchan/Moray, are good cereal production regions, and the Roman army was always in desperate need of grain to feed their troops. Grain for hack silver makes good sense.

The RAID option is also possible. Both the later years of Roman Britain, and the years immediately following their withdrawal, saw a number of recorded ‘barbarian’ incursions. Could raiding parties collect enough silver to make the chains? It seems unlikely that a smash and grab raiding party could focus enough on finding large amounts of silver. But there is another possible scenario, that the Picts and Votadini took many hostages on their raids, and then received payment for their release. This scenario is supported by finding hack silver with Christian motifs in the Traprain Law hoard. According to church law, Christian silver vessels were only allowed to be destroyed for this for one reason – buying the release of Christian hostages.

These options of raid, trade, or paid, all seem to point to the later Roman years. The connections to Caerleon are in the early part of the 300s, before the Romans deserted this major camp. Trade in services or grain would have been at their height in the 300s too. Raiding was sporadic, but may point to a slightly later period of the late 300s into the mid 400s.

The form of the silver chains

One of the interesting aspects of the silver chains is that they were all made to a fairly strict pattern.

They are of nearly identical length, within only a couple of centimetres difference. The size of their links varies a little, which means that a different number of links is required to make the chain of the required length.

A chain is composed of a series of double links, with a pair of links slightly larger at one end, and a single larger link at the other end. A flat clasp with an opening is at the double-link end.

The same style of chain but in bronze is found at the High Pasture Cave on Skye, and at the permanent Roman fortress of Caerleon. These chains are much smaller, and much earlier, early 300s for Caerleon and 790 – 539 BC for Skye, meaning that the philosophy behind these chains has a lengthy history.

The long history of these chains is also supported by the Irish stories which tell us that they were made either of silver or findruine – bronze with a tinned coating that appears to be silver. Findruine is found in archaeological contexts prior to the influx of Roman silver, or as an alternative to the real thing when unavailable. The history here feels like the Picts were suddenly able to make a real ‘silver chain’ according to their legends, and when the opportunity arose, they took it.

This standardised form of massive silver chain indicates that someone, one day, decided to materialise a story, and created these chains all with a near identical pattern, shape and form, but now in ‘real’ silver, and appropriately super large for a deity. In other words, it’s beginning to seem likely they were all made at the same time. Or, at least within a generation by the same school.

The purpose of the silver chains

As my previous blogs have discussed, the silver chains all seem to have been deposited in similar contexts, with the primary purpose of guarding the land. In Lothian, they were deposited at the power centres of Traprain Law and Edinburgh, then at the entrances and junctions of routes heading north into Lothian. In Pictland, with only one possible exception, they sit near the estuary of major river systems, and on borders between regions, again guarding entrances to the land.

But, here’s the most interesting thing about this distribution - it’s coordinated. The findspots aren’t random. It’s like someone has sat down with a map and asked, where do we need to deposit a silver chain? Then gone about making and depositing them in one ‘silver chain event’.

Once again, the logic seems to be that the chains were made as a group, to be deposited as a coordinated group. Which means that their dating is going to be very tight indeed. All we need to do is discover what event brought in so much silver, and what event or decision led to the need to protect the land with such extravagant gifts to the gods.

Previous dating concerns

The dating of the silver chains to around AD 300 to 450 has only recently been accepted. But frankly it’s always been obvious. Everywhere else in the world happily gave their silver hoards this timeline, only in Scotland were they previously claimed to be from many centuries later. So why all this angst?

The answer lies in the fact that two of the chains carry Pictish symbols. And the dating of Pictish symbols has been the hairiest story of them all. It is only in recent years that the date of the earliest symbols has been significantly pushed back by archaeologists.

Prior to the archaeological dating, James Fraser, in From Caledonia to Pictland p. 114, writing in 2009, said this:

In fact, the prevailing art-historical argument maintains that the symbol stones significantly post-date Christianisation, even if the symbols themselves developed earlier. As such, they speak of the persistence of cultural traits of a certain kind in southern Pictland, from the Roman Iron Age through the Early Historic period. There is no reason to link them with pagan religiosity.

The religious agenda as a driving force in dating the symbols is clear from that statement. Many people, just 20 years ago, believed that all symbols came from the 9^th to 10^th centuries, putting them safely into ‘real’ Roman Christianity. Gradually, slowly, painfully, archaeology has pushed the dates back, 8-9^th C, to 6-7^th C, to 3-4^th C. And the story isn’t over yet. The earliest symbols identified so far are informal ones on cave walls, but they are already fully formed with governing rules, indicating that the symbols themselves come from much earlier again.

The problem hasn’t really been that we didn’t have a good handle on the date of the symbols, but rather that other people took as fact the dating as current in their day, and then used it to interpret their own archaeology and history. And this was what happened to the dating of the silver chains. Their dates were pushed out into later centuries – because the symbols they carried must be from a later date.

There have been a few brave scholars who questioned this decision. Cummins in his book ‘The Picts and their Symbols’ p.82-97 discusses the silver chains and places them in their logical timeline and context. Laing also questioned the dates. Only with the recent exhibition of silver was the dating silently, without comment or acknowledgement, re-aligned.

But as we will see, this still remains an issue to some degree. Miller in his thesis, just this year of 2025, while accepting that the chains were probably early, suggested that the clasps with symbols could be much later additions. In doing this, he separates the dates of the clasps with symbols from the dates of the chains.

We need to look more closely at the symbols on the chains, to try to discover how their dating may be playing out.

The symbols on the clasps

Two clasps carry Pictish symbols. Most writers present this as an exception, implying that most clasps didn’t carry symbols. But that’s not the case.

Of the 15 silver chains found, only 9 remain to us today. Of these 9 chains, only 5 have clasps surviving - 2 with symbols, 3 without symbols.

In other words, we don’t know how many of the other 10 chains originally carried symbols. All of the chains inside Pictland have lost their clasp, except the Parkhill chain with its symbols, so it’s entirely possible that all Pictish chains originally carried symbols.

Three of the eight Lothian chains have clasps without symbols, and this ties in with the lack of symbol stones in the area.

The one outstanding question though is why the Whitecleugh chain found so far south also carries symbols. It seems likely, given their link number, and their type of clasp, that the Whitecleugh and Parkhill chains were made by the same hand. I have no ready answer for this, except to say the ownership of all the Lothian chains remains a question.

The symbols on the Whitecleugh chain

The Whitecleugh chain has two recognisable symbols, the double-disc/Z and the notched-rectangle (without a rod).

We can tell that each symbol has its upper end at the break in the clasp. The double-disc/Z symbol always has the end with the five floriations on its rod pointing to the top of the symbol.

And we can also tell that these two symbols are acting as a pair, the usual case for symbols - because the notched-rectangle has no Z rod.

Symbols are drawn from an ordered list, and occur in their pair in the same relative order as on the list. In this case, the notched-rectangle is above the double-disc/Z.

But, in a symbol pair, one of the symbols holds what I have called ‘dominance’ over the other. Usually this is the upper symbol, in this case the notched-rectangle. But if, for some reason particular to a stone or artefact, dominance is needed to go to the second symbol, one way to do this is for the upper symbol to lose its rod, which is why the notched-rectangle on the silver chain has lost its Z rod – it is handing dominance to the lower symbol, the double-disc/Z.

There is only one other instance of a notched-rectangle without its Z rod, on a symbol stone fragment at Westfield below East Lomond Law in Fife.

Nor is there any other instance of these two symbols in a pair, despite the double-disc/Z being a high-frequency symbol at around 10%. This means that, although both symbols are frequent and well-attested, the symbol pair on the Whitecleugh chain is quite an unusual choice.

The notched-rectangle symbol mostly occurs in the north, in the Spey valley and Moray coastal region. The Whitecleugh and Westfield symbols, both uniquely without their rod, are the only CI instances that occur below the Mounth.

On CI stones, the double-disc/Z has a focus in the Aberdeenshire region, from Aberdeen through to Rhynie. It does however have a wide scatter elsewhere, even if not so concentrated. This means that the Whitecleugh chain holds two symbols that both have a focus in the same northern region as the Pictish silver chains – another pointer to the southern chains being part of the same ‘silver chain event’ as the northern chains.

The symbols of the Parkhill silver chain

The Parkhill silver chain has two symbols on its clasp – the ogee and ‘two triangles’. The ‘two triangles’ is not known from any other example, but we do have a number of other symbols which only occur once, and we don’t know how many symbols are not attested at all, so there is no reason to think this is not a valid symbol.

The ogee symbol is rare, only found on three CI stones, two at Kintradwell near Brora, and the other at Drimmies, upstream of Brandsbutt, ABD.

It’s possible that most of the northern symbol stones, like the two from Kintradwell, are part of a later Pictish expansion into the region, although it’s hard to be precise here. The Drimmies stone is a later CI stone, as shown by its double-sided comb. This possibly means that all these ogee examples on symbol stones may well post-date the silver chain.

So, the symbols on the two chains don’t really help us with dating the chains as such. But they are interesting in their own right, the Parkhill chain has two rare symbols, and the Whitecleugh chain has a unique combination of symbols.

The case of red enamel

The symbols on both chains are filled with red enamel. Scotland’s Early Silver says that this may date the chains to the 300s. Miller, on the other hand, says that red enamel was early, and during the Roman years other colours such as blue and yellow were added to red, but with the withdrawal of Rome, red enamel once again becomes the norm. From this, he falls on the side of thinking the clasps with symbols are much later in any dating sequence, and that the symbol clasps were made later than the chains, perhaps ‘as much as 300 years later’.

The issue is we only have these chains, the Norries Law plaque, and a couple of hand-pins, with red enamel, so the data is too rarified to draw a general pattern through time, especially as the use of red enamel is a constant over the centuries. In addition, the Picts in several ways show that they are not falling in line with Roman fashion. I don’t think the red enamel can help us with more a precise dating for the chains.

Keeping the chains for centuries before burial

If, as Miller suggests, the date of the chains may be separated from a considerably later date of the clasps with symbols, there is an implication that the chains were used as some kind of regalia, and held above ground for many centuries before being buried. However, I would expect after several centuries of usage wear patterns would be fairly obvious.

I also come back to the fact that the chains were not wearable regalia fit for a human, but symbolic chains with immense mythic power. And, that they were all deposited in a coordinated pattern over northern Pictland and Lothian, a pattern we wouldn’t expect to find if they were randomly buried after several centuries of usage as regalia. The function of the chains as shown by their deposition sites is to guard the entrance to a region, and the logic here is that they were all made, and deposited, with a clear purpose, to perform a critically important ritual role, in coordination, at the same time, ‘the silver chain event’.

The La Tene style internal decoration

The dating of the silver chains to AD 300-450 thereabouts puts the chains into the pre-Christian period, or at least into the period before Christianity took hold in a political sense. So, we might expect the symbols on the chains to be similar to pre-Christian Class I symbols, but they’re not.

I have to confess to having triggered this bit of chaos – when I pointed out that the double-disc/Z on the Whitecleugh chain, and on the Norries Law plaque, do not have the type of internal decoration that we expect of a Class I symbol. We seem to have a silver chain from the Class I period, but with symbols that may be from the Class II period. It’s a quandary, which I guess I should try to fix as I started it.

We have loads of examples of the Class I double-disc/Z symbol. All of their discs are divided with at least one internal circle, usually in turn with a central dot.

When we get to Christian Class II examples, this internal circle is lost completely, and the discs are filled with a few different patterns. And one of those patterns is what George Bain in Celtic Art: The Methods of Construction p59-62, calls the ‘three-coil spiral’.

The double-disc/Z of the Whitecleugh chain has a two-coil spiral, but the Class II double-disc/Z of Strathmartine 5, Ulbster, Meigle 2 and Kirriemuir 2, all have three-coil spirals. The three-coil spiral also occurs within other CII symbols and other decoration, for example, the ‘sunburst’ of Hilton of Cadboll.

These spirals are not found on any Class I stones, but they are one of the most basic circular forms of La Tene art. And while La Tene art flourished in the years BC, elements of it still persisted for centuries, and can be found in manuscripts such as the Book of Kells and the Book of Durrow, although it must be pointed out that this is rare.

The Gaulcross pin also has the three dots decoration found on the Parkhill chain, again tying them together.

I was hoping that the silver pins with La Tene style artwork might help with a more precise dating as they are found over Britain and Ireland, but every paper I opened said the same thing - it’s just too hard to pin down a precise dating context. Proposed dates for this sort of pin range from the 3^rd to 7^th centuries. Which isn’t terribly helpful.

Dating Class I and Class II symbols

Where does that leave us with dating the silver chains?

Filling the discs of the double-disc/Z is definitely a form found only on Class II stones. So, unless putting a symbol on metal was considered differently, we should place the Whitecleugh symbol, and the Norries Law plaque, into a Class II context.

But at the same time, the style of La Tene art on this group of silver items suggests we are in an earlier phase of artwork than that of most Class II stones. The dating of the silver chains and their symbols would seem to be right at the point where Class I artwork is giving way to what will eventually become Class II artwork, but is still ancient La Tene style.

And this is where the difficulty lies. How do we date the Class II period?

We can safely date the final flourishing of stones such as Hilton of Cadboll to the 700s, as the Portmahomack monastery was destroyed around AD 800. But when did the period start?

It is likely that the Class II stones began to be raised in the period when the political class formalised the practice, which happened in the second half of the 500s in Ireland, and in the first half of the 600s in Northumbria. That’s my best guess.

But, from the very start of the Class II stones, they displayed quite significant changes in their artwork. A new style of cross peculiar to Pictland, interlace, key pattern, and symbols filled with different decorations, including the double-disc/Z with triple-coil spirals that we’ve seen above. There is no gradual change here, it has already happened before the CII stones begin. We seem to be missing the period of this shift in artwork from CI to CII. A shift which presumably happened on unattested metalwork, wood, or documents, well prior to the AD 500s/600s of the CII symbol stones. Which puts the silver chains into this interlude.

Similarly, there is a problem with dating the Class I symbol stones, which is especially obvious in southern Pictland. Here the symbol stones have a single-sided comb, the ancient Celtic type. But all the Class II stones carry a double-sided comb, the new Roman fashion. What this tells us is that there is a significant time lapse in the south, between the last Class I stone being carved and the first Class II stone being raised with its new styles of art and symbols. And the silver chains with their peculiar symbols fit within this strange interlude.

So now we can ask the question from the other direction – can we date this time interlude in the south? The answer I suspect is obvious, that this is the same interlude we find in the archaeology, from the early 200s onwards for two or three centuries, when population numbers decline severely. And remembering that the south isn’t just missing symbol stones from this interlude, but also it has no known silver chains.

Pulling the dating together …

To sum up, this is what we have so far. The silver chains themselves can be dated from around 300 to 450 thereabouts. The symbols on the clasps, along with those on the Norries Law plaque, seem to come from somewhere after 200 to before 500, but probably closer to the earlier date, while new art styles were being developed on silver items.

In other words, there is no reason to be concerned that the symbols of the chains were not the usual form of Class I symbol, as art styles had already begun to shift, possibly in the 4th century. This period seems to be a time when so much of the Pictish world was changing, under stress from the Roman presence and its decline.

An unexpected aspect that has come out of this research is that all the massive silver chains seem to have been made, and ritually deposited, at one time. They were all made to quite a standard size, and in the same form. Elements, such as the clasps, number of links, and length seem to tie chains to the same maker. The chains were deposited in such a coordinated manner over part of Pictland and Lothian, that it appears their manufacture and immediate deposition are part of the one ‘silver chain event’.

The silver chains belong to a long-standing myth centuries old, but the sudden opportunity to gather enough silver to make an enormous chain suitable for a deity in order to call upon his protection for the land, that was a major change. The symbols themselves were ancient, likely created in the centuries BC, but now they could be put on stone, and on the great silver chains. The rules of the symbols and their sacred outward form were held steady. But now their internal decoration could express other ideas - first simple geometry, then filled with ancient La Tene artwork in all its glorious swirls on silver, then intricate interlace and key pattern on the Christian stones. Some things would always remain the same, other things were allowed to evolve and fly.

And sorry, this doesn’t really pin down the dates of the chains and their clasps with symbols to any greater precision than what we started with, from the 300s to mid 400s. But, at least we can be happy we didn’t find anything to gainsay that!

*My great thanks to Bill Patterson for his scholarship, constant advice and input into this research. Errors or oddities all mine of course.

Subscribe now

Previously, I looked at the Gododdin silver chains to discover that they sat on routes, usually junctions, or at the centres of power at Traprain Law and Edinburgh. Then, I thought about the philosophy behind the silver chains, discovering a linkage to the deity Lugh in his aspect as a guardian dog with a silver chain as collar. One question to watch here will be how well this picture holds up within Pictland itself.

The Pictish chains, so far found at least, are all in the north-eastern region from Aberdeen to Inverness. None so far have been found in southern Pictland, or further to the north above Inverness. The implications of this odd distribution doesn’t as yet have any ready explanation, except to say that this region also has the largest concentration of early Class I pagan symbol stones.

All but the possible Gaval chain are in similar positions - at or inland from the river’s estuary, close to a major river, Nigg for the Dee, Parkhill for the Don, Torvean for the Ness, and Banff for the Deveron. At first glance, this pattern appears to be quite different from the chains of the south, where they are mostly marking an inland route, each seemingly guarding an entrance to the region of Lothian. But, the same function can be seen here in Pictland, where they are guarding the entrances – via the river mouths – to inland regions. The difference lies in the geography, not the position.

The Parkhill silver chain (20319)

The find location is on the north side of the River Don, in an area with several lochs and low-lying mosses. At this point, we can remember that the Gododdin chains are also found in a boggy place near a road or junction, and here we have much the same pattern. Near here the river Don was likely forded by a N-S route, heading up to Oldmeldrum and Fyvie, or possibly out to a northwards coastal path. At the same time, moving inland along the River Don, the route takes us through Kintore and Inverurie, to the foot of Bennachie, with the densest concentration of symbol stones in Pictland. From there the fields open up to eventually reach Rhynie and Tap o’Noth with its extraordinary settlement of hundreds of roundhouses.

This chain’s location is also at the furthest eastern point on the River Don that we find symbol stones.

The immediate area is quite busy. There are three long-term locations nearby.

A long-term settlement of 4,000 years was recently discovered which lies on the south side of the river. The settlement included iron working and would have been active during the Iron Age period when the silver chain was deposited, in the late 300s to early 400s.

At Goval on the north side of the river, not far from the chain’s find spot, another long-term settlement has been found, with dates from 200 BC to AD 400, again covering the period of the chain’s deposition. It was discovered during the preparation for the Aberdeen bypass, a lengthy piece of archaeology which only found this one late Iron Age settlement. It’s reasonable to assume that this settlement would have been connected with the first, by a route and ford over the River Don, with the silver chain deposited in the nearby moss.

A third long-term site is at Dyce, where a collection of symbol stones has been found, and where a later church dedicated to St Fergus was built. The symbol stones themselves are interesting as one is quite early, on a rough boulder with simple outlined symbols and a single-sided comb. A second symbol stone is dated later, on a prepared slab, with delicate and intricate symbols. The third is later again, a rare Christian symbol stone with ogham. The stones could span a period of AD 100 - 800, altogether showing how this location has long been held sacred.

Opposite the symbol stones was a large gathering area marked by Cothal comhdhail and Logie. The placename ‘Logie’ in Pictish terms often lies in this type of location – a large low flat area just inland from the river’s estuary, a place of festivals and trading. To sum up, this Dyce-Parkhill region is a busy boundary location, at a junction of a N-S route with the E-W route heading along the River Don into Aberdeenshire, and it’s no surprise at all to find a silver chain guarding it.

The Parkhill silver chain has two symbols on its flat end link – the ogee and ‘two triangles’. The ‘two triangles’ is not known from any other Pictish symbol example, but we already have a number of other symbols which only occur once, while we don’t know how many symbols are not attested at all, so it’s likely this is a valid symbol.

This chain has the same number of links as the Whitecleugh chain, which is the other chain with symbols on a surviving clasp. This suggests that both chains, Parkhill and Whitecleugh, were made within the same Pictish context, although it is still to be resolved why the Whitecleugh chain with symbols was located so far beyond the borders of Pictland. A number of other silver chains have their end link missing, so it’s possible other chains also carried symbols. In fact, this Parkhill chain is the only one with a surviving clasp in Pictland, so it’s quite likely all Pictish chains carried symbols.

The Nigg silver chain (20251)

This chain is today only a fragment, 9 links out of about 30. Yet, it turns out that, of all of them, this chain has the best recorded circumstances of its discovery, thanks to a letter from a surgeon named Troup, which you can read as the last entry on this website.

The chain was found on the northern headland of Nigg Bay, Aberdeen, which sits above a small sandy bay, and below the estuary of the River Dee. As described by Troup, ‘the place where it was found was very solitary on a grassy declivity rising 20 feet above the level of the rocks and sea mark on the North Branch of the land adjoining to the extremity of the Bay of Nigg’. It was found about 2 feet below the surface, between the ‘pan and the mould’, that is between the top soil and the hard layer of compacted soil under it.

To begin with, I confess this find spot did have me puzzled and questioning its validity. But with help from a number of colleagues, a fascinating picture can be drawn which points to this find spot being quite significant. So my thanks in particular to Colin Taylor, Fiona Campbell-Howes, Bill Patterson, Andy Sweet, Linda Sutherland, Sam Fleming.

The surgeon Troup was the brother of the gentleman farmer of South Kirkhill, which is Churchill on Roy’s map, the land above St Fitticks church (and isn’t to be confused with the more modern farm of South Kirkhill, marked on the OS first edition maps).(Thanks Fiona for this sleuthing!)

I have placed a red star near the place that I think is being described, above the ‘rocks’ near the extremity of the bay. ‘Sea mark’ means a buoy used to mark a passage or hazard, which suggests a position close to the tip of the headland. (The harbour here was only built in recent years.)

Intriguingly, there are two ‘stones’ marked on this early map, which raises the question of whether this chain, likes others, was buried near an ancient standing stone.

Trying to figure out the story of this chain has been an adventure. So much is written, and repeated, it is a study of how wrong a story can become with the telling. The chain was not buried 18 feet down, but was found around two feet below the surface, similar to that reported for some other chains such as Torvean and Traprain. It does not have double links alternating with single links, as the photo of it shows. Some reports say it is a diluted alloy of bronze and silver, but others say it is 95% silver, which is the purest form of silver used for any item, and the level found in elite Roman silverware. Which is correct, I still can’t say. It was not ‘in a bag’, at least not one that was found - it was ‘found in a lump, as if dropped into the earth from one’s hand’, which may imply it was originally in a bag, or may not. It is interesting that the chain was not laid out in a line.

Finding the Nigg chain

Troup’s letter says that the chain when found was ‘about 2 feet long, with a thimble at one end (such as sailors use for their ropes)’, by which is meant a flat clasp as other chains have. In other words, the chain was complete when found.

The fascinating part of this story is the description of people’s reaction to finding it.

The chain when taken out of the ground had a most infernal smell and the farmer imagining it to be the plague that had been there buried, used it very roughly, and tho it was about 2 feet long, with a thimble at one end (such as sailors use for their ropes) yet the thimble and other links were all broken in a thousand pieces by the farmer, his servants, his children and others, in order to try what metal it was.

As Troup goes on to describe, the silver of the chain had been badly tarnished black to the point of being unrecognisable as silver, and the silver itself was very brittle from long pressure underground, so the chain was in a parlous state. Tarnished silver puts off hydrogen sulfide, making it smell like rotten eggs, the ‘most infernal smell’.

The date and purpose of the chain

Troup finishes his letter by offering some thoughts on what had been found.

Some think it a prisoner’s chain, others to the Church, such as were used for vessels to hold incense, others to the Spanish Armada, or to the Surinam Dutch fleet wrecked on this coast and pillaged from their ships some person there hid it.

The purpose and date of the chain are intertwined. The chain is from the 300s to early 400s, not in much later centuries as was claimed for so long. From the earliest discussions of these chains, they have been victim to the belief that Pictish symbols were from much later in ‘Christian’ centuries, and because two chains carry symbols they too must have been much older. So much blood on the carpet. Yet, finally, with the newest study leading into the recent silver exhibition in Edinburgh, the date has been suddenly and silently corrected, to the same period of other pure silver items and hoards, either the late Roman period or soon after when large quantities of pure silver could be sourced.

The importance of Nigg Bay

Nigg Bay is a small sandy cove, separated from the estuary of the River Dee by this same headland where the chain was found. This bay and its associated inlet would have been something of a rarity as a refuge at the time, as there are very few such landing sites for small boats along this stretch of coast. It is no coincidence that in recent years Nigg Bay has been further developed into a small harbour.

This small but critical landing place may be why the ancient church of St Fitticks is down near the beach, providing a place of prayer or thanksgiving for sailors. The church is otherwise not clearly connected to any early settlement, notably out on its own in a fairly desolate spot. There is no known saint by this name, to confuse the matter further.

But, with the estuary so close, why wouldn’t boats have simply sailed up the river mouth? While this is possible in favourable conditions, there are real perils. A dangerous rock platform surrounds the point, today guarded by the lighthouse. River estuaries on this coast also have shifting sandbars, large waves, awkward currents, and can be hard to navigate. The river Dee ‘has often a surge rising quickly at its mouth, by the wind blowing in shore, dangerous to boats’, according to the statistical account.

From Nigg to the River Dee

Nigg Bay then was an important small landing beach providing access to the River Dee region. From the bay, there were two possible passages into the River Dee’s estuary.

The first possibility is the line that follows the current railway below Torry. In previous times, the sea came further inland, past Balnagask, and it seems to have been the case that a small branch of the Dee formed a channel below Torry, emptying into Nigg Bay, effectively making this whole headland of the silver chain an island. Evidence for this situation can be found in three distinct hirsts of pebbles, one of them being 150 yards inland, pebbles that would have come from the River Dee itself.

The earlier presence of this channel is supported by an event in 1804. Workmen were digging along the burn that still follows the original channel when they found a timber boat of ‘considerable size’ buried nine or ten feet below the surface, and about a hundred yards from the high-water mark, perhaps a boat that became stranded in the channel.

The second possibility is a foot passage, today St Fitticks Road, which starts near the old church of St Fittick and follows the base of the headland, going north-west across into the river estuary behind the mouth, where presumably travelers on foot could find a ferry. Roy’s map c.1750 also shows the coastal path coming north from Stonehaven going past Balnagask and Churchill above St Fitticks church to terminate near this point.

This region is also the far north point of the Mearns, a region which is divided in two by the Mounth.

To sum up, the original find spot of the Nigg chain is at a location guarding the headland of a small, sandy, bay, a rare safe landing place for sailors along the coastline, a place from which they could enter the inner estuary of the River Dee, and from there, the Aberdeen region and its hinterlands.

At the same time, it is also guarding the terminal of a long coastal path coming north from Stonehaven, before it enters the area of Aberdeenshire. This point also marks the northern boundary of the Mearns.

The lost Banff silver chain

There is a story of a lost ‘silver chain of great mass and intricate workmanship found in a moss near Banff’.

This description is contained in the only source about the chain, a letter from John Stuart to Daniel Wilson, dated ‘Aberdeen 10 Feb 1853’.1

There has been some confusion about whether this refers to the fragment of a woven chain of silver which is part of the Gaulcross hoard found some miles to the west near Fordyce. This small piece of Roman hacksilver has nothing to do with the kind of massive double-linked silver chains we are referring to as ‘silver chains’ in this study.

The importance of this Banff find spot is that once again we have a massive silver chain, as an isolated deposit, in a bog, beside the River Deveron, somewhere near the mouth of the river at Banff.

Banff is also beside the NW boundary of Buchan. Of the five silver chains in Pictland, three are on the coastal boundary points of regions, suggesting these points already marked political boundaries in Pictish times.

The lost Gaval silver chain

There is a record of a ‘silver chain’ being found at the recumbent stone circle at Gaval, near Fetterangus, just to the north of Old Deer. The card entry reads tersely ‘Silver chain found near this stone’. The mention is fleeting, we have no description or picture to prove the identity, and no idea what happened to it. However, it’s difficult to imagine any other artefact that this could be, other than a massive silver chain. Can we use what we now know about how the silver chains function to see whether this may have been an appropriate location for a massive silver chain?

This would be the only instance of a Pictish silver chain which is not found near the mouth of a river. But we have seen other ritual deposits near ancient artefacts, including for example, the Borland chain which was found beside another stone circle. The Gaulcross silver hoard was also found in association with ancient standing stones.

Another pointer to this being a massive silver chain is that it was found as an isolated deposit, as all silver chains are, not part of any larger hoard.

The place where the chain was found is Gaval, presumable from gabhal/gobhal ‘fork’, implying there was a fork or junction in the road here. There is a fairly large area with associated Gaval names here, all on the west side of the North Ugie Water, suggesting this location was once a significant estate. Oddly, we’ve also seen this name as Goval, the Pictish settlement beside the Parkhill chain.

To the north of the stone circle is Kindrought, Cinn drochaid ‘Bridge Head’, indicating the presence of an ancient route over the river here from Gaval. If a route had then continued north to the prominent Mormond Hill, it isn’t today identifiable on a map, but remains a possibility nonetheless.

The recumbent stone circle where the silver chain was found is on a fascinating alignment – due south of the eastern summit of the great hill of Mormond. Whether this alignment, and possibly a northwards route over Kindrought to Mormond Hill, had relevance to the deposition site of the silver chain is a question to consider.

A small hill is on the edge of Gaval and Fetterangus, today with an observatory on top, with one of those dials which points to all the places in view. It doesn’t appear to be a large hill on maps, but it is a ‘moderate’ climb, and from the top the land spreads out, north to the sea on the Moray coast and east to the coast of Buchan. It would have made an excellent beacon hill, connecting the east coast with the north coast.

Below Gaval is the famous monastery of Old Deer, a significant early Christian site. The name Deer, Doire, means ‘Oak place’, similar to Derry in Northern Ireland and Brigid’s Kildare. There can be little doubt that a large druidic oak nemeton, marked by two of the rare Class I symbol stones in Buchan, was the precursor of the later monastery.

One early pagan CI symbol stone was found at the later monastery of Old Deer, placed upside down with a simple cross carved on its rear. But another small CI stone was found at the church of Fetterangus, meaning that it comes from somewhere in the general area of Gaval. And, it is one of the strangest of all. It has a normal cauldron symbol at the top of the stone, and below it a possible eroded rectangle symbol. But in addition it has something unique – two interlocked rings. The rings have some sort of internal decoration which can no longer be make out clearly. These rings are not a valid symbol, nothing like this is seen elsewhere, and the rings don’t lie in a relationship of ratios to the cauldron symbol, as normally drawn. I can’t help but wonder if these two interlocked rings are somehow connected to the presence of the actual silver chain. Or perhaps a memory that the function of a ritual chain is appropriate for this place.

It’s hard to put a date on this stone. It may be from anytime from AD 100 to AD 800, although it seems fairly early with an unprepared surface and a simplistic symbol. Its date perhaps covers the same period of the silver chains, AD 300 to mid 400s. There is no obvious causal linkage between symbol stone and chain, just to say that they are both evidence of activity, perhaps ritual activity, at the location.

To sum up, the Gaval chain’s find spot is part of a very active location with a long-term critical religious focus. Although not at a river estuary as are the other Pictish silver chains, the very importance of this site, with its oak forest, its nemeton and monastery, its recumbent circle aligned to Mormond Hill, and its symbol stones, all seem to point to an appropriate deposition site for a silver chain.

The Torvean, Inverness, silver chain (13495)

The Torvean silver chain was found 2 ft deep in the side of a large flat Pictish cairn. It weighs 2.88 kg, making it the heaviest Pictish chain, fitting for a chain that guards the entrance to a power centre, akin to Traprain Law. The large fort to the north is Craig Phadrig, likely a royal seat in the north.

The location of the chain was low down in a gap in the large esker that runs along the west side of the River Ness. The Caledonian Canal was dug through this gap to divert the river in recent times, and the chain was found in that process.

The silver chain is then close to the ancient route along the Great Glen, here following the River Ness. And likely also a junction with a route to a major power centre.

To the NE, the esker forms the prominent detached hill, Tomnahurich, Hill of the Yew Trees, where a carved stone ball was found. Tomnahurich has wide views over Inverness and the surrounding areas, and is said to be the resting place of Thomas the Rhymer, who waits with his men and horses to come to Scotland’s aid once again. To the other side along the esker ridge is a small fort, of unknown date, 100 feet long and 50 feet wide (13549), sitting on top of the highest point - another good beacon position.

A large building ‘circular in shape with at least four concentric outer ditches’, measuring about 30m, was found just to the east of the chain’s find spot.

Over the River Ness from the chain’s find spot, a ‘cow’ symbol stone sits at Lochardil, suggesting that it marked a road from Lochardil to a river crossing near the chain, and this road continues past the building noted above and on to Craig Phadrig (13486) about 2 miles above.

Not far to the south of the Lochardil stone was the Knocknagael stone with a boar and disc-over-rectangle symbol, and a massive craft-working settlement, although seemingly abandoned around AD 200. Other boar symbols are further south, with the final one at Dunadd, all marking the Great Glen, and the original territory of the Caledonians. It looks like there may be a territorial boundary between the cow symbol of Lochardil and the boar of Knocknagael. If so, this makes this chain, like Banff, Parkhill, and Nigg, another chain on a Pictish territorial boundary.

In summary, the Torvean silver chain is right in the heart of activity in the northern Pictish region.

Craig Phadrig and Columba

Craig Phadrig is a large vitrified fort originally built between 400-200 BC. Dates from a secondary palisade ditch gave 416–556 AD, which suggests this may have been the location where St Columba came to see the Pictish king Bridei mac Maelchon, who ruled 555 to 584 AD, although this identification has frequently been questioned.

The Life of Columba, written by Adomnan about a century after the events, tells us about St Columba visiting the Pictish King Bridei mac Maelchon in his fort near the River Ness. Columba had to ‘climb a steep path to King Bridei’s fortress’ (Adomnan, II 35 p184). Craig Phadrig is about 2 kilometres from the river and while not on an immensely steep hill, still provides a decent climb. We also know that the River Ness was a fair distance from the fort, as Bridei at one point sent riders from the fort with a message to Columba who was waiting at the river’s edge (II 33 p181), something you wouldn’t do if it was just a matter of walking or running to the bottom of the hill in a matter of minutes, as would be the case for the other proposed candidate, an assumed fort under the Castle of Inverness. A ride of 2 kilometres though would make sense as faster to saddle up and ride than walk.

Columba was not staying at the king’s fort, yet food and shelter must have been provided him and his companions, beside the river, in a place not too far from the fort. All this suggests we have here a bruiden, a hostel established by the king, which provided hospitality and safety to travelers, traders and visiting dignitaries. It is intriguing to think that the large round building near the chain’s location, on the river bank, may have been this hospice.

Later, when about to leave and return to Iona, Columba went to ‘the long loch [Loch Ness] at the head of the River Ness’ to start sailing for home (LOC, II 34 p183) and engaged in a contest of wind and waves with Bridei’s druids, something so familiar from Irish stories. To be a great saint in these days meant to be an even greater druid! Columba wins of course, but it also means that where he had been staying was downriver, somewhere along the River Ness, between the end of Loch Ness and the estuary of the river. We seem to be in the generally correct location.

The Loch Ness Monster

Then we come to the most famous story of all (II 27 p176), in which a water monster has attacked and killed a local man, but despite the ‘frozen terror’ of all the onlookers, Columba sent one of his men across the river to fetch his boat. The monster ‘was lying low on the riverbed, its appetite not so much sated as wetted for prey. It could sense that the water above was stirred by the swimmer, and suddenly swam up to the surface, rushing open-mouthed with a great roar towards the man as he was swimming midstream’. Columba raised his loud voice commanding the monster to go back at once, the monster fled in terror and the man arrived back on Columba’s side of the river with the boat, boat and man intact. It’s interesting that Columba’s boat is on the opposite side of the river at the start of the tale, is this a story-telling device? Or is it part of the ritual politeness of being a guest at the royal hospice? Either way, it indicates that there was some sort of river crossing at this point, and it’s wonderful to think that this place may have been the very location of the first written account of the ‘Loch Ness monster’.

The High Pasture Cave chain

We can’t leave the subject of Pictish silver chains without considering the case of the same type of chain which was found at the ritual site of High Pasture Cave at NG 5943 1971, on Skye.

This chain was bronze rather than silver, and much smaller, although given everything else we’ve seen, I’d suspect it was originally findruine, bronze with a coating of tin, making it shine like silver. It comes from an earlier period, the Iron Age, around 790 – 539 BC, a period when silver was not generally available.

At this point, we can remember that another bronze chain was found at Caerleon on the Severn Estuary, along with a carving of a dog wearing a chain collar.

The site is a complex ritual building linked by a stairwell down to a large cave, and the chain was found as a ritual deposit in the wall of the stairwell put there during its construction. The cave itself has ‘extensive views’ over the sea and hills of Skye.

From a small sandy cove below the cave, a path leads across to Broadford, then on to the passage over the kyle to the mainland. Perhaps this chain too is guarding an entrance from the sea, leading to the mainland.

There seems no reason to assume the function of this double-linked chain is any different to the later silver versions. Its early presence also tells us that significant mythology lay behind the form and function of a double-linked chain.

CUNOMAGLOS and MAGLOCUNOS

Bridei mac Maelchon, King of Picts

The Torvean silver chain was deposited beside a route heading up to Craig Phadrig, likely the king’s fortress where Columba met the Pictish king, Bridei mac Maelchon. The chain is dated to the late 300s to 400s, while the specific date of Columba’s visit is not specified but would have taken place in the later 500s, after Columba moves to Britain c.563 and before the death of Bridei c.586, so these two events are separated by one or two centuries. Despite this time gap, (and thanks to Fiona for pointing this out to me), there is an odd connection here, the name Maelchu, from earlier Maglocunos, ‘Hound Lord’.

Those readers who’ve read my previous post on ‘What is a massive silver chain’ will remember that in the region of the Severn Estuary there are several places associated with dogs – a bronze double-link chain and picture of a dog with a chain collar at Caerleon, a mastiff with a chain collar at Pagans Hill, votive dogs at the shrine of Lydney, and a dedication to Apollo Cunomaglos at Nettleton Shrub. The native name here, Cunomaglos ‘Hound Lord’, is the same name as Maglocunos but earlier, the elements just reversed, a linguistic shift which seems to happen fairly early on.

And, I have already spoken about Kentigern of Traprain Law, a name possibly with the same meaning, although here using a different term for ‘lord’.

Now we have a third instance in Bridei mac Maelchon, again a notable character. Is this just a coincidence? Or is something deeper going on? Let’s take a look at some other instances.

Cynfael (W.) and Conmáel (Ir.)

The early form of Cunomaglos turns up in Irish as Conmáel and in Welsh as Cynfael. Not all names switched their elements, some becoming fossilized in their older form. (Rhys has a relevant section p.218-233)

I can’t find anything about a person called Cynfael. But the Irish Conmáel was in legend the first king to have been born in Ireland. He was the son of Éber Finn, one of the first Milesian (human) incomers. He reigned for 30 years (a Celtic ‘age’, putting this into a mythic context), and was killed by Tigernmas, another legendary figure. What is fascinating here, is that we have an ancient form of name and a story which appropriately places it in the earliest phase of Irish human history.

MAGLICUNAS at Nevern in Wales

In the region of the Demetae, around the coast to the north from the Severn Estuary with the dedication to Apollo Cunomaglos, another stone with a double inscription was found at Nevern (Nanhyfer), in Irish MAGLICUNAS MAQI CLUTA[RIGAS] and Latin MAGLOCVN(I) FILI CLVTORI. (Picture here.)

Prof David Stifter points out that its language makes it one of the earliest ogham stones in Wales, although he doesn’t explicit give a date. (Ogam of the month SEP 22).

The archaic retention of all endings, the lack of vowel affection (-CUNAS and CLUTA-, not younger *-CON and *CLOT-), and the lack of syncope of the middle vowels mean that it probably belongs to the oldest layer of Irish inscriptions in Britain. … Therefore, while the inscription is in an Irish script, uses Irish endings and the Irish word MAQI, the names are gaelicised British Celtic.

This places this Maglocunos stone sometime shortly after AD 360, when a series of raids resulted in the Déisi from Ireland settling in Dyfed and bringing the practice of ogam with them, an initial period when the Irish and British languages are mixing. We can also note that although it’s not known exactly when the Romans abandoned their permanent fort at Caerleon, this is the likely period of that change too. So the Nevern stone, with a date in the later 300s is now heading towards the period of the Gododdin and Pictish silver chains.

Clotri is on record as an early king of Dyfed, so the memorialised Maglicunas may have been the son of a king, perhaps an unrecorded king himself.

A side note here. The important church and monastery at Nevern has a tradition of being founded by St Brynach in AD 540, a fascinating and prolific Irish missionary saint.

Maelgwn Gwynedd

A generation before the Pictish king Bridei mac Maelchon d.586, another king was named Maelchu, Maelgwn in Welsh, the infamous king of Gwynedd, north-west Wales, who died c. 547.

There has been a suggestion that the Pictish king may have been the son of this Welsh king, an idea for which we have no evidence but can’t rule out completely. But, it seems more likely that we have the same name being used by two hugely important kings of their day. In other words, this name is a high status name, with a mythological significance that supports the legitimacy of the king.

The interesting thing is that we may have a possible connection to the Votaddini at Caerleon, through their chain and mastiff with a chain collar. A connection which may continue with the son of Clotri, a king of the Dyfed region. But the connection is also present with Maelgwn, as he is the great-grandson of Cunedda ‘Good-Dog’ (reigned c.450–460), who lead his people out of Manau Gododdin to re-establish themselves as kings of Gwynedd, and the Pictish Bridei s. Maelchu, also in a region guarded by the silver chains. It’s a tangled web, hinting at stories untold.

All of which begs the question, is it possible that the Pictish king Bridei mac Maelchon is somehow also connected to the Votaddini? Had the Votaddini also established a Pictish rulership in the north? Is this why we have a silver chain at Torvean? So many teasing questions, that may be covering up a history which is far more complex than we can imagine.

Here’s a summary of the dates for forms of the name Cunomaglos. At the very least we can say that this name was popular in these early centuries AD. All, apart from Apollo, are the names of kings or sons of kings. Which is appropriate if this is a calling of the deity Lugh, himself often associated with sovereignty.

• Conmáel, legendary first human king born in Ireland

• Apollo Cunomaglos, shrine inscription 200s?

• Maglocunos son of Clotri (king of Dyfed), inscription second half of the 300s

• Maelgwn Gwynedd, king of Gwynedd d. 547

• (Kentigern, grandson of King Lot of Traprain Law, floruit 500s?)

• Bridei s. Maelchu, reigned c. 554 to 586

The function of Pictish silver chains

To sum up, the Pictish chains are functioning in the same way as the Gododdin chains, guarding the entrance to a region. In the Gododdin context, these points are usually on routes and junctions, or at power centres. In the Pictish context, they mainly sit at or near the estuary of a major river system, guarding the region from the seaward side.

I have not been able to discover any explicit placenames that implicate Lugh near the Pictish chains, but we have to assume that the philosophy of the silver chains which supported their guarding function was understood by the Picts just as well as the Gododdin.

The possible association of the massive silver chains with the name Maglocunos / Cunomaglos, with Lugh, kings, and the Gododdin in particular, hints at a story that is complex, with many twists and turns lost to us. But a story that we need to bear in mind as we go forward. Who knows what else lies ahead?

Postscript: I do apologise for the length of this blog, it started off, as usual, as just a small quick one, but kept leading me on to discover one exciting thing after another. If you’ve made it this far, well done, and thanks.

Subscribe now

Here, I’m going to ask the hardest question of all, what on earth is a silver chain meant to be? To be acting in a guardian role, to be worthy of having so much silver invested in it, the silver chain must have had a powerful mythic significance.

What is a silver chain?

The function of a silver chain has long been a puzzle. This video talks about the chains being ‘regalia’, but admits their opening is too small to fit around a man’s neck.

The idea of putting such a massive heavy chain around your neck has always been a bit mind-blowing – imagine hanging 6 tubs of butter around your neck and being comfortable with it! But lately it’s also been pointed out that the internal width of a chain is too small to fit a man’s neck. In which case, it’s been suggested the chain was worn by a woman or a child. Personally, I can’t imagine any woman thinking such a massive clunky heavy object was something to aspire to wear. And a child would end up howling the place down.

So if these chains were not some sort of elite necklace, what could they be?

We could just avoid the question by saying the chains were simply made for deposition. Which is possibly true anyway. But even so, they were hugely massive and expensive ritual objects, seemingly with a key function of guarding the land, so there must have also been a powerful story behind them to enable them to function in this way.

We don’t have any contemporary local documents or even mythology to give us a clue. But the Irish were particularly good at writing things down, and it’s reasonable to expect that something so amazing as a silver chain would not have escaped their attention and may have been included in their writings.

And that is exactly what we find. Silver chains are mentioned in two contexts, birds and dogs.

Birds with silver chains

The poignant description of a pair of birds, usually swans, linked by a silver or gold chain is found a number of times in Irish myths. They are usually women who have taken a swan form. This picture in found in ancient European art, where the swan is attached to the boat of the sun by a chain. (A summary is in Anne Ross, Pagan Celtic Britain pp 305-311).

Here is one example. As people are traveling to Bru na Boine where the deity Lugh resided, they are led on by a great flock of wondrous birds. Cu Chulainn is about to be born, fathered by Lugh himself.

The flight of the birds, and their song, captivated the Ulstermen with their beauty. There were nine score birds in all, with a silver chain between each pair of birds, and each score flew its own way. And two birds flew out in front, a silver yoke between them.

I think we can discount this context as irrelevant, as no bird would have been able to take to the air attached to one of the Gododdin massive silver chains.

Dogs in silver chains

A ‘silver chain’ is mentioned several times in the context of a dog’s chain. All the dogs with silver chains are part of a retinue of a king or hero. And it is here that I think we have a potentially relevant context. Dog’s, after all, are guardians.

Fraech and his seven hounds with silver chains

Secht mílchoin i slabradaib argait ‘seven hounds in chains of silver’

Fraech and his retinue are described as they come to the court of Maeve and Aillil. His renown and stories filled Eriu and Alba, for he was ‘the hero who is the most beautiful of the men of Eriu and of Alba’. In another part of the story, Fraech goes to Alba to retrieve his stolen cows – another interesting connection to Alba.

And there were fifty gold-hilted swords with them, and a soft-grey mare under the seat of each man, and bits of gold to them; a plate of silver with a little bell of gold around the neck of each horse. Fifty caparisons of purple with threads of silver out of them, with buckles of gold and silver and with head-animals (i.e. spiral ornaments). Fifty whips of findruine, with a golden hook on the end of each of them. And seven chase-hounds in chains of silver, and an apple of gold between each of them. Greaves of bronze about them, by no means was there any colour which was not on the hounds.

The magnificent sight causes one of the first recorded deadly crowd crushes!

The manipulation and play that the young hero who is in it makes--I have not before seen its likeness. He shoots his pole a shot’s discharge from him; before it reaches to earth the seven chase-hounds with their seven silver chains catch it.

At this the hosts come from the dun of Cruachan to view them. The people in the dun smother one another, so that sixteen men die while viewing them.

Cano meic Gartnáin

Da milchoin ar slabrad airgit ina laim deis ‘two hounds in silver chains in his right hand’

The story of Cano tells of his exile and return to Scotland, which the annals place in AD 670. It’s thought Cano was a real person perhaps based in Skye, so it’s interesting to again see a connection with Scotland, perhaps even Pictish Skye at this period. A place where an earlier bronze double-linked chain was found in a ritual deposition.

Cano’s retinue is described in some detail, similarly to Fraech’s. This time it is his servants who each have in their hand a pair of hounds with a silver chain .

Each servant carried a set of fidchell across his back with gold and silver pieces. A bronze lute in his left hand, and a pair of hounds with a silver chain in his right hand.

King Conaire’s gift to Da Derga

trí noí con ina slabradaib airgdigib ‘three times nine dogs in silver chains’

King Conaire describes his gifts to Da Derga, a high status hospitaller who ran the king’s feasting house. Again, we see hounds in silver chains as a particularly elite object to be found only in the highest company.

I gave him a hundred kine of the drove. I gave him a hundred fatted swine. I gave him a hundred mantles made of(?) close cloth. I gave him a hundred blue-coloured weapons of battle. I gave him ten red, gilded brooches. I gave him ten vats ... good and brown. I gave him ten thralls. I gave him ten querns. I gave him thrice nine hounds all-white in their silvern chains. I gave him a hundred race-horses in the herds of deer.

Cu Roi’s gift to Fer Chertne

X soga soleicdi (.i. milchoin soleicthi) asa-slabraduib finruine no airgit hi segraide oss n-eng ‘ten hounds eagerly loosed from their chains of findruine or silver to hunt the deer’

In this example, we again have a long list of worthy items, given to Fer Chertne, the poet, by his king Cu Roi, which includes this verse:

Cu Roi granted me
Ten holdings of Daire’s sons
Ten slavewomen
Ten golden bridles
Ten noble horses
Ten bordered garments
Ten cauldrons
Ten straight swords tusk-hilted
Ten fair pairs of victory swords
Ten prows
Ten hardy swarms of bees
Ten tens of cows one hundred cows
Ten cowherds for a cattle raid
Ten bitches easily loosed
From findruine and silver chains
Onto herds of wild deer

Here, apart from silver, we see findruine used in elite dog chains. Findruine is a silvery metal which is often used to describe high value items in Irish texts, given a value of ‘less than gold’. For a long time it has been a puzzle what findruine is exactly, but the Peebles hoard, dated 1,000 and 800BC, has provided a plausible answer, that it is bronze with an enriched surface of tin, which gives bronze the appearance of ‘glowing’ silver.

This also raises the suspicion that some of the handful of bronze chains with double links which have been found may also have originally been made of findruine. Although these bronze chains are much smaller than the silver chains, few in number and widely scattered, it may be that we shouldn’t yet dismiss them as unrelated until we understand more. (See Scotland’s Early Silver pp 102-103).

Dogs in their silver chains

What we have in the Irish texts is quite consistent and persistent, that elite hunting hounds (mílchú dil.ie/32231) have silver chains (slabrad dil.ie/37788) . The presence of a silver chain appears to be an exceptional statement, as other dogs on chains are spoken of without silver being explicitly mentioned. Another word can be used for a dog chain, cona árach ‘dog leash’ dil.ie/3926, but this doesn’t occur in the context of a ‘silver chain’.

• A ticklish comment on the massive silver chains by user fickle-tickle: Those are no necklaces, they’re what we used to keep the werewolves in line.

Here’s a medieval picture of dogs on a chain, although at this point it’s not yet clear whether the silver chains are collars, or part of the leash. From the Aberdeen Bestiary Folio 18r:

Lothian, Lleu, and Lleu’s chain

Now, even if what is said so far is valid, that the silver chains are dog chains, and that the function of these massive silver chains is to guard the entrances and centres of Lothian, that still leaves us with the question of how it is that a silver dog chain has such a huge ritual significance to warrant deposition of such extraordinary silver chains by the Gododdin.

It may be that part of the answer lies in the language. Early Lothian was called Lleuddiniawn in Modern Welsh, ‘country of the fort of Lugus’. Which tells us two things. First, that the region is named for the deity Lugh (Irish), Lleu (Welsh). Which is hardly surprising given the pan-Celtic occurrences of his name on places.

And secondly, that there was originally another Lugdunum within Lothian. John Wilkinson’s article here is well worth a read on this topic. Traprain Law itself seems the most likely candidate, with its tales of King Lot or Lleuddun (W.). In Welsh texts it is Edinburgh that is mentioned as the seat of Lleuddun Llwyddog, but as there is already evidence for an early name of Edinburgh, this is more likely to be a later confusion for Traprain Law, long deserted in the mid-400s.

The linguistic connection here may be two-fold. In one of my latest blogs ‘Lugh and the Pictish dog symbol’, I discussed the possibility of Lugh/Lleu having an ancient aspect as a big cat, which in Celtic times had morphed to a large dog. If this turns out to be correct, then Lugh, in his aspect as the fierce guardian dog, was being invoked through the silver dog chain to guard his named territory.

Note that now we have moved past the issue of why the Gododdin silver chains are so much bigger than the earlier bronze ones - because they aren’t necessarily made for ‘real’ dogs, they are made for a deity in his aspect as a dog, and so the bigger, the more expensive, makes the symbolism the more powerful.

Lugh’s chain

Which brings me to an odd item. Charles Squire recorded that the Milky Way was called ‘Lugh’s chain’. Unfortunately, Squire didn’t tell us what the original terms were, nor where he got them. But why Lugh would have a ‘chain’ at all has long been one of those nagging puzzles. It may be coming full circle here, that Lugh in his aspect as a fierce dog of course had a glorious silver chain, just like the other elite hounds in their silver chains.

The dogs of the Severn Estuary

So far, we have dogs that wear silver chains in Ireland, dog symbols and silver chains in Pictland, silver chains in Lothian the region of Lugh, even a silver chain at Traprain Law which was likely another Lugdunum. But, can we find any similar trace in southern Britain? There are four locations that might be relevant, Caerleon, Nettleton Shrub, Lydney, and Pagans Hill, all in a fairly limited region around the Severn Estuary. (My thanks to Bill Patterson for help with this section!)

The Hounds of Caerleon and Pagans Hill

On the north side of the Severn estuary lies Caerleon, in its day a massive permanent Roman trading fort. And there is a carving here of a ‘mastiff’ with a chain around his neck.

And at the Pagans Hill sanctuary on the south side, a statue of another sitting mastiff was found, again with what appears to be a chain around his neck.

And, here’s the exciting moment. At Caerleon, a bronze double-linked chain, was also found in the 4^th century destruction deposits of the civil settlement of the Roman fort. And it is likely the same kind of chain as the Caerleon mastiff is depicted wearing. This is the only instance that I’m aware of in southern Britain of the same type of bronze chain as the massive silver chains, and is not far off contemporary with the silver chains.

This bronze chain is smaller, its length roughly a third as long as the silver chains, although I gather it isn’t complete.

Caerleon was the site of a massive Iron Age fort founded in the 5th century BC. This was followed by a huge permanent Roman fort built around AD 78 called Isca Augusta, making it a major military and trading hub. The name Caerleon is considered to derive from ‘Fort of the Legions’, but given the regional focus on the dog with his chain, it does make me suspicious that the native fort was originally yet another Lugdunum ‘Fort of Lugh’.

The fascinating thing about finding an actual depiction of a dog wearing a chain, is that this is a particularly Celtic depiction, as normally Roman dogs had simple collars. Here’s what Ralph Haussler wrote : pp 72-73

For both Caerleon and Pagan’s Hill, one has argued that the collar might be an indicator for a domestic dog, but there is a problem: first, this does not look like the typical dog collar we know from Roman art which is usually of much simpler fabric (cf. already Boon 1989: 205). Was it a particular chain to demonstrate control over a (mystical) wolf? It is also possible this it represents some form of symbolic decoration: statues in temples and sanctuaries are frequently decorated in various forms, and this collar (or jewellery) might be a sign of status and prestige.

Here at Caerleon and Pagans Hill, we have the sacred dog with his chain depicted , to add to the Irish stories of the hounds in their chains, and the actual bronze and silver chains used in ritual deposits.

Cunomaglos ‘Hound Lord’

At the Romano-Celtic sanctuary of Nettleton Shrub, not far from Pagans Hill and Caerleon with their mastiffs, there is an altar dedicated to Apollo Cunomaglos, ‘Dog Lord’. It’s quite a significant, complex sanctuary, including a hostel, dated to before AD c.300 to 370. Apollo was the Roman god of the sun, but the epithet here is purely Celtic.

When the Roman deities were equated to Celtic deities, a practice promoted by the Romans, there was a problem that the deities often didn’t match up exactly, so people took features of one deity and just lined it up with another as they saw fit. This was particularly problematic for Apollo as the Roman sun god, as the Celtic sun deity was female. So while there’s nothing explicit to say this Nettleton Apollo was being equated to Lugh, it’s possible that the equation with Apollo was thought reasonably close given that Lugh is a god of light and midsummer.

There is another feature in common to consider. Apollo was born with Artemis his sister when their mother Leto had turned into a wolf. In Greece he was called Apollo Lykaios ‘Apollo of the Wolves’. Haussler suggests that Cunomaglos should be read as ‘Wolf Lord’. The Celtic word cu, is a general ‘dog’ term that covers all sorts of dogs, including hounds, mastiffs, wolves, so I’m not sure it’s safe to limit it here to ‘wolf’. The Pictish dog symbol is a large mastiff, definitely not a wolf, and we have two beautiful portrayals of actual Pictish wolves for comparison. The Irish milcu wearing silver chains are described as hunting hounds. Also, Celtic does have a particular term for wolf, fáel, which they could have used if they were just translating from the Greek Lykaios.

We can’t go past this without mentioning Kentigern, the saint from Traprain Law, ancient Lugdunum with its silver chain, whose grandfather was Lleuddun. His name has been a headache for linguists, but one derivation may be from *Cuno-tigernos, again meaning ‘Dog Lord’. Not the same term as in Cunomaglos, but with the same meaning.

Finally, I started this series of blogs about the Pictish dog and Lugh, pointing out a persistent ambiguity between lions or lynxes and these mastiffs and hounds, based on the association of Lugh with the Lughnasad constellation of Leo which in Roman terms is a lion, but in Celtic terms was a cu ‘dog’. And here at the Romano-Celtic shrine of Nettleton we have this same ambiguity, a dedication to a Dog Lord, but as Haussler reports:

Among the Roman pottery from Nettleton Shrub, it is surprising how many depict animals in various forms, notably panthers and lions (Simpson 1982)

The Lydney dogs

Probably the biggest difficulty with looking at Celtic dogs is that they occur with a number of deities and in a variety of contexts. They figure large in everything Celtic, real or mythical. So we do have to be wary of just putting all dogs into the same philosophical bucket.

Another temple on the Severn Estuary to the east of Caerleon, is the famous temple of Nodens at Lydney, which appears to have primarily served as a place of healing. And it is notable as a place with lots of dogs. Below are two of the votive dogs, but although one hound has a thick collar, neither has a chain collar, nor are they the thick-bodied mastiffs of Caerleon and Pagans Hill. Their presence continues a broad emphasis on dogs in this Severn Estuary region, but at Lydney they are a different kind of dog, with a different function, they don’t wear chains, and they are associated with a different deity, Nodens (Ir. Nuada).

The basic concept of a silver chain

Here then is the proposal coming out of this analysis. The massive silver chains, and at least some of the earlier double-linked bronze or findruine chains, are symbolic of the chain of Lugh, in his aspect as the fierce guardian hound. The chains formed a ritual deposit in key places that required guarding, at the centre of power, and at routes and junctions entering his own named region, Lothian, the land of Lugh.

As a final observation to wonder about, here is the famous silver cup of Lugdunum, modern Lyon. The deity on the cup is perhaps a conflation of Lugh and Mercury, being the patron deity of the city. And sitting there is yet another mastiff similar to that of Caerleon and Pagans Hill, again with a collar, although here not a chain as such.

Subscribe now

No-one knows what they actually were or what their function was, no-one knows what their significance was, no-one knows why they were buried at various locations in Scotland. We can’t even say where all this silver came from. All we can say for sure is that they were obviously extremely important items, the object of an exceptional ritual deposit, and, we’d really expect we should have some inkling what they’ll all about.

A lot has been written before about these massive silver chains found in Scotland, and I’m not going to repeat all that here. There is a chapter with good coverage in ‘Scotland’s Early Silver’, and here is a short video by the National Museum of Scotland from 2018 in which you can get an idea of just how massive these chains are.

In this blog, I’m just going to discuss the silver chains found in Gododdin territory below the Firth of Forth, and I’ll leave the Pictish silver chains for another day. I’m going to look in more detail at where these chains were found, in the hope that the deposition sites can provide us with clues. And in my next blog, I’ll take what we’ve learnt here and see if we can figure out what ideas lie behind a silver chain.

Mapping the chains of the Gododdin

I’ll list the details of each chain’s findspot at the bottom of this blog, but here I’m going to summarise those details, and present the overall picture.

Looking closely at the find spots of the chains, a pattern can be discerned – these chains are marking out a territory, roughly equivalent to today’s Lothian.

Firstly, silver chains are found at the two known Gododdin centres, Traprain Law, and Din Eidyn (Edinburgh). From this, we are left in no doubt about the symbolic importance of the chains to the Gododdin. Another chain was found at Haddington, close to Traprain Law, possibly near the junction of a road. (When I say ‘near’, I don’t mean right on top of the junction, but in a low-lying place nearby.)

All other chains were deposited in a low-lying place, often boggy, near a significant route, sometimes a junction, on the roads heading north into Lothian, towards the power centres of Edinburgh and Traprain Law. They appear to mark the southern entrance on routes into Lothian. The area marked out is only slightly different to today’s region of Lothian. (This marking of routes is also the case with the silver chains inside Pictland.)

This period of the early 400s has so little known history, but most people these days assume that the territory of the Gododdin extended much further south into Northumbria than just Lothian. Either this is not the case after all, or, the chains were specifically relevant to the area of Lothian.

The silver chains were likely deposited in the early 400s, which does raise a question. This area is generally considered to be the territory of the Votadini, later known as the Gododdin. However, around AD 400, according to Gildas, the Picts and Scots had seized most of the land down to Hadrian’s wall, or according to Nenius, the River Tyne. Whether this meant that the Gododdin temporarily lost their base at Traprain Law is unclear, but possible. So, while it seems that logically the owners of the chains were the Gododdin, it’s vaguely possible this was not the case. On the other hand, the massive amount of silver invested in the chains does fit with the huge amount of Roman material entering places like Traprain Law, so I think it’s fairly safe to say the chains belonged to the Gododdin.

So, the general picture as presented by the silver chains, is as a hugely important ritual deposit, at their strongholds, and at routes and junctions which are the entrance points to the territory of Lothian.

In functional terms, they are guarding the territory of Lothian.

This key function, on which the safety of a territory might depend, starts to explain why such a huge amount of silver was put into a chain, which was then ritually buried. Nothing should be spared to ensure the safety of the land.

But the next question is, what has a silver chain got to do with a guarding function? What is the connection? And why is it so important? I will attempt to delve into these curly questions in my next blog on the silver chains: “What is a massive silver chain?”.

The Whitecleugh silver chain (512432)

Before I list the details of the find places of the silver chains, all of which are either at power centres of the Gododdin, or at strategic entrances to Lothian, I need to talk about the Whitecleugh stone, as it seems to present as special case.

The Whitecleugh silver chain has always been extra puzzling. Firstly, it has two Pictish symbols on one of its clasps, a long way south of where Pictland is generally considered to be, at least north of the Forth if not north of the Tay.

Secondly, it is also further south than the other silver chains, and was found by accident in a wild upland setting in the Lowther hills, and unlike all other chains, far from any known settlement.

But that said, its position is intriguing. It may well mark the central point that symbolically divides four territories. Today it sits on the southern border of Lanarkshire. In older maps, it sits near the conjunction of Clydesdale, Nythsdale, and Kyle, again showing a boundary position. But it also sits at a watershed near the source of a stream called Duneaton Water, which flows a short distance north-east until it falls into the River Clyde near Abington. The actual source of the Clyde today is considered to be just to the south of Abington, but it’s possible that in earlier centuries Duneaton Water was thought to be the first part of the Clyde, putting this chain location at the source of a major river of the land.

Given this watershed and boundary location, it may be that the silver chain was sitting at the spot which divided ancient kingdoms, as I’ve attempted to give a rough indication of below.

The source of silver

The source of so much silver that went into these massive silver chains has always been a mystery, usually just written off as ‘Roman’. Perhaps the Gododdin were trading with the Romans, hiring themselves as mercenaries, or raiding in the south. Still, it’s an awful lot of silver, no matter how they came upon it. But here, in the land around Whitecleugh are lead mines, and Roman forts and roads running through the area. And silver is found in this lead - there certainly has been tonnes of silver found in this lead in recent centuries. It does raise the suspicion, despite a lack of any archaeological evidence so far, that the Romans and the Gododdin were actually mining here and a lot of their silver is coming out of this desolate place in the hills.

DETAILS OF PLACES WHERE A SILVER CHAIN WAS FOUND

All the massive silver chains were found about two centuries ago, during a phase of agricultural and industrial expansion. Consequently, their exact findspot and their burial context are only roughly known. Nonetheless, most findplaces can be ascertained to a general location, and that is sufficient to provide a general context from the surrounding landscape.

It’s still likely that there are other silver chains out there to this day, but in places that are unsuitable for agriculture, perhaps too boggy. And it’s possible that other chains have been found in the past but not turned in.

This list doesn’t include the silver chains inside Pictland (another day another blog), but their findplaces are similar - on routes and junctions.

My thanks to Bill Patterson for his patient persistence and deep scholarship that has gone into this study of his home ground. All errors of course can be lain at my door!

Traprain Law (56385)

The volcanic hill of Traprain Law (56374 ) “dominates lowland East Lothian”, standing out over the surrounding region; “it remains impressive and craggy, an almost unassailable fortress with a known occupation of at least 1000 years”. Presumed to be the capital of the Votadini, later known as the Gododdin, it was probably abandoned in the mid 5^th century, when power was transferred to Edinburgh.

The silver chain was found on the top NE point of the hill. It wasn’t part of the Traprain Law silver hoard. I can’t find any direct dating context for it.

This silver chain is unique in that it is the only chain found in such a location, all other chains being found in lower or marshy areas. It suggests that this chain may have been a closing deposit as the Votadini left the hill in the mid 400s, around the same period as the other huge Traprain Law silver hoard (56384) was also deposited c.410-425 AD. If so, it gives us a reasonably good estimate of the date of the chains, as the early to mid 400s.

Queen’s Park, Edinburgh

A silver chain, since lost, was found in Queen’s Park beside Holyrood House. Like Traprain Law, Edinburgh was a royal centre of the Gododdin, probably until besieged and lost to the Anglo-Saxons in AD 638.

The exact findplace is uncertain and the chain now lost. But there are three small lochs at the base of Arthurs Seat, which may be the find location, especially as this would place the chain effectively at the base of the hill of Edinburgh castle.

For some reason, this chain is not counted in the list in ‘Scotland’s Early Silver’. It is found in the Old Name Book of the Ordnance Survey, but Andrew Breeze could not locate any further publication (Breeze, Andrew. “Pictish chains and Welsh forgeries.” Proceedings of the Society of Antiquaries of Scotland. Vol. 128. 1999).

Haddington (56498)

The find location appears to be in an arable field to the south of the river. Although this location is in the loop of the river, it’s not obvious that it would have previously been a marsh, although this is possible.

There is nothing particularly striking about the location, though it was probably close to where there have been routes for millennia, especially the route (roughly following the A1) between Traprain Law and Edinburgh, the two historically-attested hill forts of the Gododdin. There is also another road that heads south, to meet up with Dere Street at Soutra Hill. Dere Street is associated with two other chains, Edinburgh and Whitlaw near Lauder, Soutra Hill being half way between them, and suggesting that another silver chain may have originally been deposited near there.

Hoardweel

The chain seems to be in or close to a moss, seen in earlier maps. On early maps there is both Devil’s Dungeon and Hell’s Cradle nearby. A number of forts overlook the road, including one of the southern ‘brochs’, Edin’s Hall. At Strait Leap there are copper mines, which may be why there is an important presence in this area.

Edin’s Hall (58777) was originally a small hillfort. Later, a substantial broch was built into the fort, with a substantial settlement in the western end of the hillfort. Copper ingots from the local mine had been buried in the floor of the broch. Dating however is still unclear, but seems to be of the Roman period.

Hoardweel is not far north of today’s B6355 which runs up to Haddington with its silver chain, near Traprain Law. This doesn’t appear to be a Roman road as such. Also near Edin’s Hall broch, is Abbey St Bathans, with a dedication to Baithéne mac Brénaind, the second abbot of Iona, which hints that this may lie on or near a route used to travel from Iona down to Lindisfarne.

Only 200m to the east of Hoardweel is another major north-south route A6112 / A6105 which leads down to Greenlaw with another silver chain. There is a large fort here perched on the top of the rise, which the road skirts around. This focus of activity in the area around Hoardweel is likely to relate to the presence of copper mines. Which is interesting given the Whitecleugh chain lies in an area of lead mines that produce silver.

Heriot’s Dyke / Greenlaw

This silver chain, since lost, was found at Heriot’s Dyke (79782). This Dyke, of unknown date, is a rounded bank with a ditch to the north, which once formed a boundary marker from east to west across Greenlaw Moor. Even today this place is wild moorland, with the loch of Hule Moss at its centre, a place internationally renowned for its overwintering pink-footed geese.

The moss is beside Greenlaw with its road junction, connecting the chains at Whitlaw near Lauder to the west, and Hoardweel to the north-east, basically along the line of the A697.

Whitlaw

The Whitlaw silver chain was found just to the north of Lauder. It lay close to the Roman road, Deer Street, which may have run just to the left of today’s road heading north, so very close to the chain. The road here follows the Leader Water.

Deer Street here ran north to Edinburgh, the power centre of the Gododdin, with its silver chain. From here, it is also possible to reach Traprain Law, following Deer Street to Soutra Hill, then branching off to following today’s B6368 up to Haddington with its silver chain.

Lauder is itself a road junction.

Above the findplace of the chain is Blackchester fort (56025) with several stone roundhouses inside it. While no clear dating is available, a 1st/2nd century bronze head of Hercules was found here. It’s possible both the fort and the silver chain are guarding Deer Street going north from Lauder.

Borland (2436500)

This chain was said to have been found on Borland farm, on the south side of the South Medwin Water, another low position near the flood plain. Across the river is Dunsyre Hill (48821), which is the end point of the Pentland Hills which stretch for 32 km up to Edinburgh. Between the chain and Dunsyre Hill is a wide river meadow through which the South Medwin Water runs for about 5 km.

The Roman road lies to the east, but this long low valley lying along the end point of the Pentland Hills is a significant point in the landscape which may have marked both a route along the south east side of the Pentland Hills, and another route around the base of the Pentland Hills.

However, there is confusion about when and where exactly this chain was found. Another claim is that the chain was found across the river at Todholes near an ancient henge when the last of the stones were removed. It may be that this happened when the railway was put through. Many ritual objects were deposited near or at ancient artefacts, so I suspect this is the case here.

A third claim is that this chain was originally found at West Linton. If it was originally from West Linton this would put it beside the Roman road itself.

Regardless of the exact find spot for this chain, the general location remains the same – the base of the Pentland Hills where routes needed to round the end of the hills.

The Whitecleugh silver chain

The details of this chain are included in the discussion part of this blog above.

Subscribe now

This linkage of Lugh to Leo on the surface seems tenuous. When we have an idea like this, one of the things we can do is to assume it’s true and from there test out how the idea stands up within the broader context. Most ideas if they are wrong will quickly fall apart under the pressure of this negative testing. If an idea still keeps standing, it doesn’t necessarily prove the idea is right, or fully right, but it is a good way to eliminate wrong ideas, so a good place to start.

Many people fail to do this testing of their idea before promoting it. And if an untested idea is comfortable to others it can push a field of study down the wrong path, sometimes for years. But this type of testing does come with a risk, it requires a good level of awareness of one’s own propensity for confirmation bias, especially difficult when the idea is your own. It means that you may well have to admit the idea was wrong, and change your mind, a hard thing to do at times.

When we test out an idea, it forces us to face the consequences, and see if any consequences still allow for the original idea, or negate it. If you find yourself suddenly jumping through hoop after hoop to try to maintain the original idea, then that’s a good indication the original idea is wrong.

So here we go, let’s see how well, or not, the idea that Lugh was represented by the constellation Leo holds up under testing. While doing this, I’m going to call in several ideas, the Pictish dog symbol, Viking weather vanes, the visualisations of constellations, and of course texts and myths.

(All pictures of Pictish stones are sourced from STAMS or CANMORE/TROVE.)

The constellation Leo in history

If we assume that Lugh was represented in the sky by the Celtic constellation equivalent to Leo, the first question to ask is - how was the constellation envisioned?

There is evidence in the Middle East going back to 4000 BC that the constellation was envisioned as a lion. Yet lions were not a feature of the northern European or the insular world. So unless we assume that the concept of a lion was imported from outside Europe, the Celts would not have used the picture of a lion for Leo.

Fortunately, in one medieval manuscript there is a zodiac depicted, and the constellation Leo is not marked as ‘Leo’ but as Cu ‘dog’ . It’s not clear how old or widespread this identification really was, and so far it’s the only instance found. But it’s something to start with. (And if you’re wondering why I haven’t given you a reference here, it’s because I’ve lost it and despite asking around no-one seems to be able to give me an answer. Please let me know if you know the source?)

If, in the Celtic worldview, Lugh was portrayed as a dog in the constellation Leo, then we need to test that out, see if it holds up. Whether it turns out to be right or wrong in one way is largely immaterial, as we’ll be sure to come across some exciting things along the way – that’s the wonder of Pictish research.

The Pictish dog symbol

The next question to ask, is there a Pictish symbol or icon that may represent Lugh as a dog or cat?

There happens to be a Pictish symbol of a thick-bodied dog - not a thin leaping hunting hound as we see in the hunting scenes, but a war dog or guard dog, the type of massive dog the British and Irish were famous for during the Roman years, one of their greatest exports. Could this dog be the Celtic ‘Leo’? Has this ferocious war dog been substituted for the fierce alien lion?

The shape of the stars of Leo lend themselves not to a thin hunting hound, but a robust thick-bodied war dog, lion-like. In this photo of a dog symbol, I’ve overlaid the main visible stars of Leo over the symbol, and they are an excellent fit. (This is a good example where a test does not of itself prove that the constellation Leo is this Pictish dog, but it would hopefully eliminate the idea if it had been wrong.)

This dog symbol occurs on the north-east coast of Pictland, on the Class II Christian stones of Golspie and Ulbster. A tribe called the Lugi were located in this region according to the geographer Ptolemy c. 150 AD. Here we have a tribe named for Lugh, and a war dog symbol on two of its important Pictish stones. (Alternatively, it has been considered whether the Lugi are named for ‘crows’, so let’s bear that in mind as we go.)

But, there is a catch. This dog symbol occurs in the region of the Lugi people, but also in the region called Cat, today Caithness. The word catt is also a Celtic word for ‘cat’. We suddenly find ourselves in a quandary, are we supposed to be looking for a dog or a cat?

The dog symbol is also found in Shetland on two stones from Bressay and another from Papil Burra. Shetland was called Insi Catt ‘Islands of the Cats’ in early Irish texts. Once again, we have an ambiguity between dog and cat.

In the southern region of Angus, the Pictish dog does occur on two CII stones near each other, Glamis1 and Eassie, The dog occurs on Glamis on one side of the cross stem, and on the other side as a symbol paired with a snake. The Eassie stone shows a stag, the wardog, and a hunting hound, which displays just how large and thick-bodied the wardog is compared to the hound. This southern dog doesn’t seem to have the up-curled tail though, rather it has its long tail curled around through its hind legs, just as the Pictish lions do, so in the end I’m not really sure if the southern ‘dog’ is the same as the northern ‘dog’, or even if it is meant to be taking on the aspect of a lion.

This dog symbol only occurs on Class II stones, which is unusual and means perhaps it may not have been part of the original set of Class I symbols. Its regionally restricted use may also indicate a late period change in usage of the symbols themselves, where symbols are used to designate a region.

It has been claimed in the past that this Caithness dog is a lion, as it is similar to the lion of Mark in the Book of Durrow, an early Irish manuscript. I’m putting the Golspie and Bressay dogs here beside the Durrow lion for comparison. There is a similarity in the typical Pictish body with internal curls and the tail curled above the body, but the Durrow lion has a distinctly larger and cruder face and long claws, which to my eye makes it a different animal.

The Picts also depict lions beautifully and realistically on other stones, for example on the Meigle stone and the St Andrews sarcophagus, so it’s hard to believe they would otherwise depict a lion in such a dog-like manner.

Nonetheless, it is interesting that once more we seem to be struggling with an ambiguity between a dog and a cat here. Durrow was a Columban monastery, so an artist could have been aware of the Pictish dog, and would have been well aware of the classical constellations, so is it possible the artist drew the Durrow lion similar to the Pictish dog because he knew they were two depictions of the same creature of Leo?

The picture of the constellation Leo depicted in the manuscript from St Gallen in Switzerland, has similarities to the Pictish dog symbol, notably the upcurled tail and general outline. But the claws and mane are distinctly lion-like.

So, we have the dog symbol running through both Caithness, the region of Cat, and Shetland, the Islands of Cats. Yet, these regions are named for the ‘cat’, not the ‘dog’. And we have a constellation portrayed as a ‘lion’ in eastern and classical astronomy, but perhaps as a war dog in Celtic astronomy. So which is it, dog or cat?

This is where it gets confusing, and I confess I haven’t really figured out the quandary here of how, or when, or if, the lion morphs into the Celtic dog.

Lions have not existed in western Europe or Britain in historic times. The Scottish wildcat is a timid forest dwelling cat, the size of a house cat, and rarely seen by humans. Lovely though they are, it is doubtful they would feature as an aspect of Lugh. But there is another candidate for the ‘cat’ of Caithness and Shetland, the lynx. The lynx only went extinct in Britain in AD 700, although it may have held on later in the north. It’s entirely possible that the Picts of Caithness named their region for the beautiful lynx.

Even if Caithness was originally named for the lynx, the Pictish dog symbol with its long upcurled tail is not a lynx with its notably short tail.

In Ireland, there has only been one mandible of a lynx found to suggest their historical presence. Even at the best times, the lynx would have been a rare sight in the habitation areas of Ireland, and today it has not been possible to reintroduce the lynx as the area of wilderness is too small to support a viable population. Under these circumstances, it may be possible that the war dog had been substituted for the lynx/lion of Leo at some point in time. But that implies that the ancient Celts, like the astronomers of the Near East, had originally conceived of Leo as a lion or lynx too. Just when this shift from cat to dog might have happened is hard to say.

If it is the case that a dog has been substituted for a cat in insular astronomy, we are in an even bigger quandary about Lugh himself. Was he originally a deity whose aspect was a cat or a dog?

Lleu and Llew

In my previous blog about Lughnasad and Lugh, I talked about the problem of how the name Lugh is derived. There are several linguistic ideas, but no firm consensus as yet.

One of the earliest ideas was that Lugh was derived from ‘light’, although that idea has linguistic problems. Yet ‘light’ would be a key aspect for a god celebrated originally at midsummer. The Welsh name Lleu is the same as the common noun for ‘light’ today, but is this just a happy coincidence?

In some medieval texts, post the 12^th century, Lleu is also written Llew. The common noun llew means ‘lion’. Another happy coincidence?

Now I’m no expert in Celtic etymology here, but my understanding is that deriving lleu from the Latin leo ‘lion’ has linguistic difficulties of its own. It is possible, but difficult.

So, we have an ancient deity name Lugos/Lug/Lleu which may have refered to ‘light’ originally, and whose name in the earlier Welsh period was likely viewed as the same as the word for ‘light’. But after the 12^th century it is sometimes written as Llew, which by then was the Middle Welsh word for ‘lion’.

Language is a complex thing, over time words morph and evolve following certain definable patterns. But they can also be influenced by what people perceive they mean or the connections they see. I can imagine a situation where the Welsh scholars may have retained the knowledge about Lleu’s connection to the constellation Leo, but also knowing that the classical constellation was a lion, this influenced them to slightly shift the name to Llew ‘Lion’.

A single coincidence is random noise. Coincidence repeated builds a case.

There is one other odd thing to mention here. Right at the start of the Pictish king list is a strange list of names, which I take to be deity-ancestors as half the names start with ‘Deo’. The first three on the list are Gilgidi, Tharain (Taranus?), and Morleo (Great Lleu/Lion?). There is also a later king called Brude Leo. No one quite knows what to make of this list, including me, but I’ll offer it here as another item to ponder.

And before I leave the linguistics, another odd one to mention is that one of the epithets of Lugh, found in the Duanaire Finn, is Conmac, where con is ‘dog’ and mac is ‘son’. This translates roughly as ‘dog-youth’. What the explanation for this is isn’t clear, as Lugh is a king and warrior, rather than a hunter with dogs, nor he is portrayed as particularly young.

To sum up, there are clues that Lugh may be represented as a dog by the constellation Leo. But the persistent confusion between a dog and a cat, makes it possible that the constellation was originally a big cat, a lynx or lion, which was later replaced in the insular world by the more relevant fierce war dog Cu.

This isn’t the first time I’ve come across a constellation that morphs like this, I’ve previously spoken about how the water-goat of classical Capricorn was morphed in Celtic lands into the far more present and important water-horse.

Failinis – the whelp of Lugh

Lugh is closely associated with two famous ‘dogs’, Failinis, and Cu Chulainn. Maybe in these stories his relationship to dogs might tell us something?

One treasure that Lugh possessed was a cuilén ‘whelp’. Unfortunately, this word is not specific to the young of a dog, just the young of 4-legged animals. In other stories though, this super-wondrous whelp is said to be a dog, whose name is Failinis, Shalinis, or Fermac.

Amongst a number of magical abilities through several different stories, such as raising his tail to make a magical wind and turning water into wine, Lugh’s whelp is said to be ‘more splendid than the sun in his fiery wheels’, and that all the beasts in the world would prostrate themselves before him. (The Wikipedia page on Failinis has a good coverage.)

If we assume Lugh is represented by the constellation Leo, the candidate for his splendidly shining whelp would be the star we know today as Regulus, the ‘King’ star, otherwise known as the ‘Heart of the Lion’, a briliant shining star in the constellation Leo, one of the brightest stars in the night sky.

Regulus is Latin for ‘king’ and remained a popular name through the Roman years and beyond, for example, Rule of St Andrews, and Riagail of Bangor. The Greeks called this brilliant star Basiliskos ‘Little King’, the Babylonian list of stars from c.1000 BC calls it LUGAL, ‘King’. The king star before whom all beasts in the world bow down.

Now, I know, I know, I’m not for a minute saying that Babylonian LUGAL has ended up as LUGOS, there really is such a thing as blind coincidence, but it does appeal to my sense of humour.

Around 3500 BC when Leo was the constellation of midsummer, the kingly star Regulus lay directly north. In Babylonian and later Persian astronomy, Regulus was one of the ‘royal stars’ that each marked a cardinal direction. It was used to measure the movements of the moon and planets, and played a major role in cosmology as guardian of the north.

Lugh’s whelp was a force to be reckoned with by day, but became a caer thened at night. The term caer has a broad meaning as a globular mass, sometimes a mass of molten iron. The term tened means ‘fire’, and caer thened is usually translated as a ‘ball of fire’ or a ‘thunderbolt’. A ‘ball of fire’ is also a great description of a brilliant star such as Regulus.

The term caer is often used in Welsh to describe stars and star groups, although it’s usually taken there to mean ‘castle’ or ‘fort’. But it’s also seen in Irish as the name of Oengus’ swan love, Caer Ibormeith, a story which has other star traces within it.

One version of the Failinis story says that it was a hound by night and a sheep by day, rather than a hound by day and a caer tened ‘ball of fire’ by night. This had me puzzled, until I realised that not only had the day and night forms been swapped, but the caer has been misread as caera ‘sheep’. A funny example of how stories can shift over time but still retain the basic elements.

Cu Chulainn – the other dog son of Lugh

Oddly enough, the hero Cu Chulainn, the son of Lugh, is also described in the same manner as Failinis, that all beings bowed down before him. Cu Chulainn gained his name after he accidentally killed the dog of the smith Culann and then took his place as ‘Culann's dog’. Culann the smith otherwise does not appear in any other story. Does the strange connections of words here suggest we are looking at two forms of an ancient tale, a dog ‘cu’ and a whelp ‘cuilén’, son and dog of Lugh, both also belonging to a smith, a smith called Culann, perhaps a mangling of ‘cuilén’.

To sum this up, Lugh has two famous ‘dogs’, his son Cu Chulainn, and his miraculous whelp Failinis. And there is good reason to think that their stories fit well with the picture in the stars, of the brilliant star Regulus in the constellation Leo. Does this also imply that Lugh himself has an aspect of a dog, as represented by the constellation Leo? For me, the case is building well but isn’t a given at this stage, but interestingly, if the original idea had been wrong, I would have expected it to have thrown up all sorts of difficulties by this stage, which it hasn’t. On the contrary, it seems easier to continue to fit data to fill out the picture, making it more likely to be correct.

Interestingly, here is a picture of a lion with upraised tail, and a small whelp – cat or dog? -from the Gelasian Sacramentary c750 AD. It’s hard to not view this as the lion of Leo with its whelp Regulus.

The whelp with the raised tail

One odd feature that keeps popping up is the ‘raised tail’. Lions themselves do not hold their tail aloft like this, nor do the more classical, and Pictish, pictures of lions have their tail held aloft, they usually curl their tail down through their back legs in a more realistic pose. But we have Failinis, the Pictish dog and manuscript lions, all with their tail raised aloft.

But, and thanks to Bill Patterson for pointing this out to me, there is another place where a lion holds his tail aloft, and that is on the spectacular golden weather vanes which were placed on the mask of a leading Viking ship. The jump from a Norse lion that moves with the wind, to the Irish dog that raises his tail to create a wind isn’t hard to make. One Irish tale even says that Failinis raised his tale and forced the men onto the sea.

Another pointer to this Irish/Norse connection, is that Failinis originally belonged to the smith of the king of Ioruaide, which may be a garbled form of name for Norway.

The difficulty here is deciding when and where this concept has come from. The Pictish dog symbol probably can be dated to the 700s, long before the arrival of the Vikings in Pictland. Perhaps then it is an ancient motif, held in common, back through the mists of time. Or some other great story waiting to be uncovered.

The lion on these Norse vanes seems to sometimes be identified as a ‘dragon’, which is confusing. One later Christian vane has the biblical King David killing the lion, and this lion too is often said to be the Nemean lion killed by Hercules as represented by Leo. But this later use of the lion motif points to the same being true of the earlier motif.

(If you would like to read more on the Viking vanes, here’s a page with pictures and explanations)

Below is a picture of a replica of one of the Jelling stones, which also has this same motif of the lion entwined with a snake. Note the lion’s mane, the big claws, and, the upraised tail.

The clues

Here’s a summary of the points above, do they build a telling picture? I think so, but only more research will tell.

1) Lughnasad runs over the period of the constellation Leo on the Coligny calendar, suggesting this constellation represents Lugh himself.

2) The classical constellation has always been a lion, but in later times the Celts knew it as cu ‘dog’.

3) The Pictish dog symbol occurs in the area of the Lugi tribe.

4) The Pictish dog symbol fits neatly over the stars of the constellation Leo.

5) The Pictish dog symbol occurs exclusively in both ‘cat’ regions of Caithness and the Cat Islands (Shetland).

6) The manuscript lions, first the lion of Mark in the Book of Durrow, and the lion of Leo at Basel, are broadly similar in shape, especially the raised curled tail over the back, but they have claws and manes.

7) There seems to be some strange yet persistent overlap between the Celtic mastif dog and the lion, which suggests that the Celts once viewed Leo as a big cat too, possibly a lynx in their context, but that in later years the war dog was substituted for the lion.

8) In medieval years, the Welsh began calling Lleu as Llew, ‘lion’, suggesting they too knew the crossover between the lion of classical Leo and the Celtic Cu, and the association of Lugh with the constellation.

9) The mythological start of the Pictish kinglist has a deity-ancestor Morleo, a name of unknown derivation.

10) Lugh has an unexplained epithet Conmac, ‘Dog-son’.

11) Lugh owns a famous miraculous puppy, a cuilen named Failinis, who appears in several stories.

12) Failinis, before whom all creatures bowed down, may be equated with the kingly star Regulus of the constellation Leo.

13) Failinis also has the shining aspect, bright than the sun, and a ‘ball of fire’ at night, which may be equated to the brilliance of the star Regulus, one of the brightest stars in the sky.

14) The Babylonian constellation lists calls Regulus LUGAL ‘king’.

15) The Pictish dog and the manuscript lions all have a raised tail, dissimilar to a real lion.

16) Failinis famously raises a magical wind by raising his tail, and a lion with raised tail is depicted on golden weather vanes for Viking ships.

17) Lugh’s son, Cu Chulainn, is also reported as having all beings bow down before him. He too often shines in splendid ways, and took the place of the dog of a smith Culann. He and Failinis may be different tellings of the same ancient tale, represented by the brillian star Regulus.

The timing of the festival is taken from my latest paper1 in which I was able to place the ‘base year’ of the calendar into the seasonal year (but bearing in mind that the calendar is Metonic, so base dates will swing around within a month).

The calendar shows that the Lughnasad festival runs for 8 days during the waning moon, stops for 5 days over the period when the moon is lost in the glare of the sun, then restarts for another 8 days during the waxing moon. In some ways, it’s much as what we might expect from traces in old texts. But in other ways, some details are surprising, and intriguing.

Lughnasad - The Festival of Lugh

To being with, the festival is kicked off on RIVROS Day 13 marked with DE[V/C]OR LUG. And this date in the ‘base year’ of the calendar is August 1. Well that’s a breath-taking discovery!

Before going on, it pays to remember that the calendar does not name any of its festivals, it merely marks each day of the festival with the term IVOS. So we don’t know what the Gauls called their festivals, but, as they are the equivalent in timing to the Irish festivals, they may well have had an equivalent name too.

The Irish term Lughnasad means the Festival of Lugh, or, the Funeral Games of Lugh, and we have Gaulish evidence that their festival too was all about this great pan-Celtic deity.

In Lyon, then called Lugdunum, not far to the south of where the calendar was found, the Altar of the Three Gauls was dedicated on August 1 in 10 or 12 AD, and from that time on, the concilia were held, during which laws, games, renewal of vows, feasting, competitions in eloquence were held. The calendar appears to be showing that the celebrations of August 1 were part of an ancient Gaulish tradition, overlain fortuitously by Roman celebrations.

This day, August 1, the kalends of the month Sextilis which had been renamed August for Caesar, was auspicious as the anniversary of his victory at Alexandra. Yet another example of a key ancient Celtic festival being overlain by incomers, here Roman, later Christian.

There is an amazing cup from Lugdunum (Lyon), which has maybe the only ancient depiction of Lugh-Mercury, which I have to put here just because it’s wonderful.

Lughnasad – a multi-part festival

This day marked with DEVOR LUG is followed by an IVOS ‘festival’ run of 7 days, making a total of 8 days. There is then a hiatus of 5 days, followed by a second IVOS run of 8 days. The Lughnasad festival is the longest and most complicated festival on the calendar.

The length of this festival on the calendar doesn’t come as a surprise, it was a gathering near the end of summer, moving into autumn, the best time of the year for games and partying. For instance, the Lebor Gabála Érenn claims that Lughnasad lasted for a half month (15 days) each side of the day of Lugnasad itself. While the calendar only shows 8 days each side, it still hints at a memory of a long festival in past times.

Lughnasad runs mainly through the calendar month called RIVROS, a name considered to be derived from a word for ‘fat’ according to Delamarre, a name which is wonderfully appropriate to this late summer season of plenty.

Here is a summary of this multi-part festival. It shows the calendar’s month days alongside the Base Year date, and the notations of each day. The two pairs of days which swap their notations are marked by arrows. Against the dates, key astronomical events are noted.

The second part of this festival, RIVROS 11a to ANAG 3, runs over the end of RIVROS into ANAGANTIOS, just like the other IVOS runs on the calendar which mark the quarterly festivals Beltaine, Imbolg and Samhain. So it is the first part of the festival, from RIVROS 13 to RIVROS 5a which is an additional and exceptional part of the Lughnasad festival. So why this exception? What’s peculiar about this time?

Lughnasad and the Moon

Normally, all the quarterly festivals take place during the waxing moon, as does the second part of the calendar’s Lugnasad festival. (Remember that the month started with the first-quarter moon.) But the first part of this festival period occurs during the late waning moon. This is oddly unsettling, given all the commentaries and folk tradition we have that the waning moon is considered an unlucky period.

Another thing of interest, which I don’t have space to go into in depth here, is that the festival has two unusual and rare examples where a pair of days of different day numbers swap their notations. This results in the first day of the new moon being swapped up to RIVROS 4, and the day of the dark moon when the moon completes its cycle being swapped down to ANAG 4. This festival month then has its moon start and finish marked out and changed – another indication that for some reason the moon has critical significance for Lughnasad.

Between these two IVOS runs is a space of 5 days centered on the three dark moon days at Days 7a-8a-9a, when the moon is lost in the glare of the sun. We know from lore even in Christian times that this period of the dark invisible moon was held in awe, well, it still is.

The calendar seems to be telling us that, for some fundamental reason, the moon is critical to the whole festival of Lughnasad. But this isn’t something that has been spoken of explicitly in ancient texts, nor in folklore.

It’s an instance where the calendar is telling us something new and critical. But how do we make sense of it?

Lugh, Tailtiu and Crom Dubh

The Irish texts tell us that Lugnasad was not funeral games for Lugh himself, but those set up by him for his foster-mother, Tailtiu, who had died after clearing the plains around Tara, suggesting she was a goddess concerned with agriculture. Because we have a festival with two separate runs of IVOS, and two deities involved, it is possible that one part of this festival was dedicated to Lugh himself and the other to Tailtiu.

Another interesting complication is that in Ireland the festival at Lugnasad has an associated day, the last Sunday in July called Crom Dubh Sunday. The word crom means ‘crooked’ or ‘bent’, and dub means ‘dark’ or ‘black’. In other words, Crom Dubh could potentially be translated as ‘black crescent’, and as such might be associated with this strange lull of 5 days focused on the dark moon.

So there seems to have been three distinctive personages involved in tradition, Lugh, Crom Dubh, and Tailtiu. Can we attempt to resolve this complex plot?

The astronomy of Lughnasad

At this point, other special positions on the calendar have indicated that we should look to the sky for an answer. It’s often bemoaned that we don’t know what the Celts called their constellations. By the time we have written records the names of classical constellations are being used. But what if the Celts had mostly the same or similar conceptions for their constellations, which is why they seem to be invisible to the record? And, I’ll even go one step further, what if a number of the classical constellations are actually based on Celtic astronomy – after all the Celts according to Roman commentators were renowned in their day for their science, maths and astronomy…

At August 1 in the period of the calendar, c. 200 BC – 200 AD, the sun lay in the heart of the constellation which today we know as Leo, the same place that it was at midsummer in the years around 3000 BC, which suggests that Lugh may be represented in the sky by the constellation of Leo, the ancient constellation of light at midsummer.

As we move to the second part of the festival, the quarter festival that runs over the end of RIVROS into ANAGANTIOS, the sun moves out of LEO into the constellation of Virgo. Virgo is classically conceived of as a woman with a sheaf of wheat in one hand, a grain harvest deity, as we might presume Tailtiu to be as well, clearing the plains for agriculture. In ancient times, Virgo would have represented the growing grain in late summer, but by the time of the calendar, Virgo occurs at harvest time. This suggests that Tailtiu may be a Celtic name for the constellation Virgo.

The traditions of Lughnasad

Folklorist Máire MacNeill made an indepth study of the lore around Lughnasad, and she found a common theme of Lugh fighting Crom Dubh for the grain of the harvest. This fight between the two gods also involved a female of agriculture, perhaps the personification of the grain.

This theme aligns nicely with Lugh as the deity of summer during the first part of Lughnasad fighting Crom Dubh, the dark crescent god, during the period of the dark moon, a fight which succeeds in the second part of the festival with a celebration of Tailtiu, goddess of grain.

Another theme found by Máire MacNeill was that Crom Dubh is said to own a bull, and the sacrifice of a bull sometimes occurred at Lughnasad too. So why would Crom Dubh be associated with a bull? The answer I think is fairly obvious, that the bull too has ‘crom’ – that is bent or crescent - horns.

In the saga of Cuchulain there are two bulls, one Finnbennach ‘Bright crescent’, the other Donn Cuailnge ‘Dark of Cuailgne’. It’s possible that these two bulls represent the bright moon and the dark moon. Bulls in continental art sometimes have crescents on their flanks. The moon with its crescent horns, Crom Dubh the ‘dark crescent’ god, bulls with their crescent horns, the association of like to like was a method of systematising the Celtic worldview.

The constellations LEO and Virgo

I imagine some readers are now objecting to the idea that Lugh and Tailtiu, as represented by the constellations now known as Leo and Virgo, can be traced back roughly 5000 years to the time of the great Neolithic monuments such as Bru na Boine and Stonehenge.

Well, these aren’t the only events on the calendar that seem to make sense by rolling them back to this time. It’s really not that long ago!

The names Tailtiu, Eithne, and Arianrhod

Unfortunately, I know some people will also object to this length of tradition, on the basis that the Irish language can’t have been present in Ireland at that period. But knowledge and culture doesn’t always align exactly with language – as the linguists constantly insist.

The name Tailtiu, like so many other deity names, is problematic. For instance, D.A. Binchy suggested that the name Tailtiu may be a back-formation from a placename meaning ‘beautiful’. Which may well be true, or not. If true, it could explain why the female associated in lore with Lughnasad is not called Tailtiu but is usually called Eithne, a name derived from ‘grain’.

Eithne is also the mother of Lugh. And unlike the name Tailtiu, we see many people in stories and history with this popular name, which suggests that Eithne may be the name of the constellation Virgo, rather than Tailtiu.

Suddenly, the story of Lugh grieving and rescuing Tailtiu, his foster-mother, has shifted to a story in which he is grieving and rescuing his mother, Eithne. Which puts a slightly different slant on the story.

Eithne in old texts is also said to be Boann, the cow goddess after whom the river Boyne is named. Another text asserts her ‘real name’ is Feada. It is only to be expected of ancient deities that they have many different callings. A mature culture will have deities that are complex and multi-faceted, playing key roles and performing complex functions.

If Eithne is the mother of Lugh, and is represented in the sky by Virgo, then how does this fit with Welsh lore? Here Lleu’s mother is called Arianrhod, so is this yet another name for Virgo? The story of Arianrhod starts with the great magician Math who must always have his feet in the lap of a virgin unless he is going to war. His virgin is Goewin, but she is raped and becomes ineligible as virgin foot-holder, so Math asks Arianrhod to take Goewin’s place as foot-holder. Some stories say that before this could happen Arianrhod stepped over his wand and gave birth to Lleu and Dylan, but other stories say that Arianrhod did become the virgin foot-holder. This is interesting as the constellation of Bootes/Arctophylax stands over Virgo, his feet in the lap of Virgo, the virgin.

I guess this just goes to show how complex these stories are. We can’t really take anything as a simple answer, not until we’ve done a lot more research to test and retest these ideas. So far, we seem to have Eithne and Arianrhod as mother of Lugh/Lleu, and maybe Tailtiu as foster-mother, the one grain goddess represented in the sky by Virgo. She may yet have other names to watch for.

Lugh and Light

The name Lugh also has always caused difficulties. Scholars have proposed a series of derivations, but the first and main one was a derivation from *leuk- ‘light’. However, lately it has been generally rejected on the basis that a Proto-Indo-European ‘*-k- never produced Proto-Celtic *-g-. Yet, a god of midsummer could naturally have an aspect of ‘light’, so I suspect this whole topic needs revisiting. So many deity names simply don’t follow strict linguistic rules, whether that is a result of fossilisation, or even a different originating language, so I’ll leave this mystery to the experts to sort out.

There is a theme attached to Lugh which hints at his association with ‘light’. Lugh is described as having a radiant face ‘as brilliant as the sun on a dry summer's day’.

Then arose Breas, the son of Balar, and he said: "It is a wonder to me", said he, "that the sun to rise in the west today, and in the east every other day". "It would be better that it were so", said the druids. "What else is it?" said he. "The radiance of the face of Lugh of the Long Arms", said they.

Another description of Lugh’s radiant face:

... they were not long there when they saw an army and a goodly host coming towards them from the East, and in the vanguard there was one young man high in authority over all; and like to the setting sun was the radiance of his face and forehead, and they were unable to gaze upon his countenance on account of its splendour.

This has in the past led scholars to claim him as a ‘sun god’, but we know that the sun is female in European lore as far back as we can see. It makes better sense to see him as a god of ‘light’, which in ancient times wasn’t directly thought to be part of the sun but a natural feature that roughly occurred with the sun, but preceded it at dawn.

The Welsh name Lleu is also today a common noun meaning ‘light’, ‘brightness’. Whether this means the name and the noun have evolved together over time from the same ancient term, or whether somehow the two have fallen together to appear the same today, is a question I will leave to the experts.

Lugh and Midsummer

Is there anything apart from ‘light’ which connects Lugh to midsummer? Or something that might hint at the ancient origin of Lugh being the deity of the constellation of Leo at a time when this sat at midsummer?

Oddly enough there is. In an Irish version of Lugh’s birth, he is said to have been born along with either three or nine seals, all of which leapt into the sea after birth. Similarly, in the Welsh version of the birth of Lleu, he is born along with a twin Dylan, who immediately jumps into the sea and swims as well as any fish.

Harbour seals in Britain and Ireland are born at midsummer. The implication is that Lugh was also born at midsummer.

Saint Moluag of Lismore, a name derived from Lugh, has a feast day of June 25, the summer solstice according to Roman and Christian time-keeping. In Ireland, saints are said to have their festivals on their date of death, but Pictland seems to be using an alternative system, perhaps one that make much older connections, such as for example, having the feast of the saint Brigid at the pagan festival of Imbolc.

Crom Dubh - who is he?

But, the big question here is why Crom Dubh must be fought by Lugh, and what is his relationship to Lugh’s mother Eithne?

Is Crom Dubh ‘Dark Bent’ the god of the moon in its dark and invisible phase? Or is he something even more? Does the dark phase of the moon allow the dark god to assert his power?

If Lugh is the deity of the light, are we seeing a fight for grain between light and dark.

Crom Dubh, or Crom Cruach, was said to have been propitiated by sacrifices of grain and milk, according to the dindshenchas. Which matches with the similar tradition of a bull being sacrificed at Lughnasad. He is certainly some sort of deity that has the power to prevent a plentiful supply of agricultural resources, but the picture of just who Crom Dubh is as a deity really isn’t at all clear.

It has been suggested the story of Crom Dubh and Eithne may have been a similar tale to Persephone who is kidnapped by Hades, the god of the underworld. But we should note that it is Persephone’s mother, Demeter, whose sorrowing for her abducted daughter causes the grain to lie dormant. And the complex picture of Demeter in all her many roles should warn us to expect the same level of complexity for the Celtic goddess of grain. It’s hard to know though how similar the Greek and Celtic stories may originally have been, or how much one has been influenced by the other, or which way the influence went, or if a common human experience of grain has resulted in two independent but similar stories.

The challenge for us here is to figure out how this story of Lugh, Eithne, and Crom Dubh fits together.

Eithne is either daughter or granddaughter of Tigernmas. In turn Tigernmas was killed by Crom Cruach at Samhain. Does this mean that there was an enmity between Eithne’s family and Crom Dubh? These types of close connections of certain parties is common in the myths, the details of the stories shift and morph over time, but the connections often remain tied relatively closely to each other.

The story so far appears to be one where the agricultural grain and milk in the form of Eithne were threatened by a dark god, Crom Dubh. Eithne died, but was reborn after being rescued by Lugh, the god of light and midsummer.

Where next?

Now that the Coligny calendar tells us exactly how the Celts organised and timed their festivals, it opens up all sorts of possibilities and hopefully will lead to many fruitful discussions. What I’ve said above may be right, or it may be wrong in all or part. Only a lot more research and debate will tell. But at least now, with the ancient Coligny calendar giving us hard data about the festivals, we can start from a firm foundation.

Subscribe now

McKay, H.T., 2024. IDENTIFYING THE SEASONAL PLACEMENT OF THE COLIGNY CALENDAR. Etudes Celtiques, 50, pp.37-270.

But the answer may be fairly obvious. A clue to the answer may lie in the DATE. And in this blog, I will be exploring what that date meant for Ecgfrith and Brude and their peoples.

I will largely be using textual evidence and ideas that have already been floated in academic circles, but I will be using this evidence to paint a different story behind Egfrith’s actions.

PREVIOUS IDEAS

Researchers have of course asked this question of Ecgfrith’s motives for the raid into Pictland many times before. The problem has been that while the battle itself and the reactions of people to the battle were written down for us, none of these early writers told us why Ecgfrith attacked Pictland.

One proposal that has been favoured is that Ecgfrith attacked Pictland to re-assert his waning power over the Picts[1]. The Picts had been under subjugation to Northumbria from the reign of Oswald of Northumbria in the 640s, that is for 45 years before the Battle of Dun Nechtain. This idea that the Picts were pushing back against the Northumbrians prior to the battle in 685, largely rests on an assumption that the Picts were regaining power under their king Brude (reigned 671 to 692). The sieges of Dundurn and Dunottar are generally attributed to Brude today, although the annals don’t tell us who was involved or who won. So, while we can conjecture, this may be a case of hindsight bias, for there is no written or archaeological evidence to confirm this idea.

But there is one good reason to find the idea unsatisfactory. If the Picts as a tributary people had been pushing back against their Northumbrian overlords, the Northumbrian elites and clerics would not have been averse to Ecgfrith doing what was his duty to do – maintaining his kingdom and its power over its neighbours. Instead, we see them adamantly and loudly decrying his vicious aggression, and attributing his death to the righteous vengeance of God.

This was a period of rising insular kingdoms, clashing against each other, kings made into saints for their battle leadership in the struggle for power over their neighbours. Most kings of this century died in battle - it was the rule rather than the exception. Ecgfrith’s grandfather Oswald has a story almost parallel to Ecgfrith’s. Oswald, aged 38, died in the battle of Maserfield against Penda of Mercia, a war in which Oswald was the aggressor. Before the battle Oswald claimed a vision in which Saint Columba urged him to ‘Be bold and manly’[2]. And Bede records that he died ‘in prayer’[3]. Oswald was never vilified by his own people, to the contrary he was praised and venerated as their patron saint. How different then is the picture of the veneration of aggressive Oswald attacking his neighbouring kingdom and dying in doing so, to the picture of Ecgfrith, the king who led his army ‘to ravage the province of the Picts’ and was utterly condemned for it. That these two similar events were viewed so differently suggests the motivation behind Ecgfrith’s actions was unusual and not judged as appropriate by his contemporaries.

Had the Picts been causing concern by uprising against Northumbria, we’d expect Ecgfrith to have been praised for his leadership in attacking the Picts, and the Picts would have been denigrated as the enemy, the trouble-makers. Yet, that’s not how Bede describes the situation, he makes no mention of any wrong-doing by the Picts, nor does he say Ecgfrith had gone to wage war on the Pictish king, instead he says that Ecgfrith ‘devastated their territories with most atrocious cruelty’.

I think with this contemporary evidence for Ecgfrith’s actions, we should be wary of the idea that he was attempting to re-assert his control in a targeted war against a rebellious tributary king, which he would have been considered within his right to do. His actions sound more like a vicious smash and grab raid.

But why would he do such a thing?

THE RAID ON BREGA IN THE PREVIOUS YEAR

Actually, this wasn’t the first time Ecgfrith had done such a thing. The year before, in 684, he had sent his army across the sea to Brega, a region in the mid-east of Ireland, where with ‘hostile rage’ they ‘lay waste Mag Breg, and many churches, in the month of June’[4]. The raid was devastating, smashing churches and monasteries, and taking 60 slaves. The insular world was horrified that this Christian king, founder and supporter of Northumbrian monasteries at their height, would commit such a heinous act against the church.

At this point, it is worth reading the full section from Bede[5] to gain a feeling of just how outraged he was about the Brega raid, and to see how he directly attributed Ecgfrith’s death to the vengeance of God for the raid.

In the year of our Lord's incarnation 684, Egfrid, king of the Northumbrians, sending Beort, his general, with an army, into Ireland, miserably wasted that harmless nation, which had always been most friendly to the English; insomuch that in their hostile rage they spared not even the churches or monasteries. Those islanders , to the utmost of their power, repelled force with force, and imploring the assistance of the Divine mercy, prayed long and fervently for vengeance and though such as curse cannot possess the kingdom of God, it is believed, that those who were justly cursed on account of their impiety, did soon suffer the penalty of their guilt from the avenging hand of God; for the very next year, that same king, rashly leading his army to ravage the province of the Picts, much against the advice of his friends, and particularly of Cuthbert, of blessed memory, who had been lately ordained bishop, the enemy made show as if they fled, and the king was drawn into the straits of inaccessible mountains, and slain with the greatest part of his forces, on the 20th of May, in the fortieth year of his age, and the fifteenth of his reign. His friends, as has been said, advised him not to engage in this war; but he having the year before refused to listen to the most reverend father, Egbert, advising him not to attack the Scots [the Irish], who did him no harm, it was laid upon him as a punishment for his sin, that he should not now regard those who would have prevented his death.

The Venerable Bede wasn’t the only one linking Ecgfrith’s death to righteous vengeance for the Brega raid. A poem attributed to the Irish monk Riaguil of Bennchor[6] has the same message, told from an Irish perspective. (This is an updated translation of the poem I did with the help of the Old-Irish-L linguists, with particular thanks to Prof David Stifter.)

Iniu feras Bruide cath, im forba a senathar  
Manad algas la mac De, conide ad genather
Iniu ro bith mac Ossa a ccath fria claidhme glasa
Cia do rada aitrige, is hi ind hi iar nassa
Iniu ro bith mac Ossa, las ambidis duba deoga
Ro cuala Crist ar n-guidhe roisaorbut Bruide bregha

Today Bruide makes battle for the patrimony of his grandfather
Unless the son of God does not wish it, tis he who will make reparation [for Ecgfrith’s attack on Brega]
Today mac Ossa [Ecgfrith s. Oswiu] has been slain in battle against blue swords
Although penance was said, it was penance too late
Today mac Ossa has been slain, whose drinks were black
Christ has heard our prayers, Bruide has freed Brega [of obligations].

So now we have two mysteries, two questions to answer. Why did Ecgfrith attack the churches of Brega in 684 with hostile rage? And, why did he attack Pictland in 685?

As Bede uses much the same language to describe both the Brega and Pictish raids, linking them in the same paragraph, it raises the distinct possibility that Ecgfrith was also intent on destroying Pictish churches, just as had been done in Brega. But if so, this would make his actions seem even more out of the norm, more deplorable.

Brega is a fertile, well-populated region, and has long been a symbolic centre of ritual and religion and power, both pagan and Christian, encompassing as it does the complex of ritual sites in Meath, Tara, Knowth, Newgrange and so on. By the time of the raid, Brega also had several significant Christian churches and monasteries, including Monasterboice, Rechru, Durrow, and Terminfechin, and in the Viking era, the Ionan church would establish Kells as an inland refuge.

Again, researchers have asked the question why Ecgfrith would have ordered such a devastating and shocking raid on Brega? A tentative suggestion has focused on Fínsnechta, the Irish king of Brega and High King of Tara. Fínsnechta certainly was a busy battle king of significant power in his day, but there is no mention of any battle between the two groups. Fínsnechta also has stories told of him in which Adomnan of Iona is acting as his principal saint and advisor.[7] Adomnan would be the person who later journeyed to the Northumbrian court to negotiate the release of the Brega captives, two years after Ecgfrith’s death[8]. But a regional power that was connected to Iona was common enough, so that alone hardly seems to warrant a risky attack.

Another suggestion is that Ecgfrith was raiding the Irish monasteries for their wealth and slaves, so as to give them to his own supporters in return for their service[9], much as the Vikings would later do. I think it’s hard to accept this idea, when Ecgfrith already had control of large regions and monasteries with fast-growing wealth, why he would risk a raid in a far-off place, against Christian churches, and under the nose of a powerful Irish king. I feel sure he had ready access to plenty of alternative avenues of wealth. But, this idea does demonstrate just how difficult it has been to come up with a clear reason behind Ecgfrith’s behaviour.

There is one mention in the Annals of Clonmacnois of the ravaging of the plain of Brega because of an alliance between the Irish and the Britons, but it makes no mention of how this alliance related to either Northumbria or the Picts. This has led to scholars attempting to propose a cascade of odd inter-kingdom relationships which may have led to Ecgfrith taking umbrage in some unknown way to an unknown alliance between other parties[10]. But the strange thing here is that this was not a battle with another king, nor a siege of a stronghold, this was a rage-filled, large-scale ‘ravaging’ with no political motive or outcome that is recorded. Indeed Bede explicitly says that the people of Brega were ‘harmless’ and twice asserts they were ‘most friendly’ to the English, a statement that confounds any modern proposal to claim this was an attack on an enemy, or a treacherous ally.

It’s difficult to go past the suspicion that the Brega raid was not targeted at Fínsnechta, or the Ionan church, but rather their goal was to do what they actually did – a devastating destruction of the churches of Brega.

Which leads me to ask again, why on earth would Ecgfrith do so? What could be so dire that, against the advice of his clergy, he would risk an infamous act that he must have known would be considered so heinous and sinful? And, why would he perhaps have repeated it the following year in Pictland with ‘atrocious cruelty’, again against the explicit entreaty of his churchmen? Up to this last year of his reign, Ecgfrith had been a successful king, ruling over a powerful and flourishing kingdom. He had managed to negotiate his way through inter-church rivalries that make your head swim. He had lost a few battles, but he’d also won a lot, and not gotten himself killed in doing so. His sudden change of behaviour in ravaging Brega and Pictland was a distinct change that needs an explanation.

People don’t do dreadful things in a vacuum, they act out of many impulses, and one big motivator is fear. But to see why terror may lie behind Ecgfrith’s actions, we need to go back twenty years, to the year he was nineteen. The year that his world shattered.

THE YEAR AD 664 : ECLIPSE, SYNOD, PLAGUE

The year 664 was momentous. Momentous in a good way for some, but mostly momentous in the worst way possible.

To start with, there was a full solar eclipse on May 1.

An eclipse was ominous enough, what dire events might it presage? But to occur on May 1, that must have rocked the insular world. For millennia, May 1, Beltaine, was considered the beginning of the year[11] (as displayed on the Coligny calendar), and, the beginning of a new era of history. A solar eclipse on May 1 would have been perceived as a very clear message – the world was starting a new age, things were about to undergo a massive change.

In mythology, we see this change of era for instance, as the Tuatha De Danann arrive in Ireland on May 1. Another instance that always makes me smile is Columba arriving on Iona on May 1, a clear statement that he brought with him the new era of Christianity. Did he really do this, or is it just a story with a message? I suspect it’s entirely possible he deliberately chose this day precisely because of its significance.

That same summer of 664 saw churchmen gather from Ireland and Britain for the Synod of Whitby, a synod which would go down in history. At stake were a few key points of disagreement between various church factions, including the way a monk should wear his hair, and the astronomical method for calculating Easter. Now these matters today may seem laughably trivial, but these arguments were fierce and deeply felt, and had been going on in various ways for centuries all over the Christian world, faction against rival faction. This wasn’t really about astronomy, the Romans had seen to that by disregarding that science to the point where neither the Celtic nor the Greek mathematics of astronomy was any longer taught or understood. This Easter debate was an excuse, propaganda disguising power struggles to see which church group could gain ascendancy over its rival. In the case of the Synod of Whitby, the Columban familia lost out in the argument with a rival British group claiming – falsely – to be aligned with the Roman tradition. Intriguingly, the person who decided this Whitby debate for once wasn’t a cleric himself, for clerics had spent centuries showing themselves completely incapable of giving way to a rival group to resolve this argument, no, it was the king himself, Oswiu, Ecgfrith’s father.

The synod didn’t end the dispute though, it merely made one group more powerful than the other. The intensity of this division would ripple down through many years to come, the Ionan church no longer in ascendancy in Northumbria despite it being so intimately involved in the founding of Northumbrian churches. Nor did it necessarily convert people’s beliefs to the opposite view point, Iona itself struggled for years to convert to the ‘Roman’ method, and one can only imagine the emotional turmoil of the losing side. Even amongst the Northumbrian churches not everyone agreed with the decision, some feeling so deeply that they relocated to Ireland to avoid changing. The important thing to note here, is the division and intensity felt by all about these matters in both Oswiu and Ecgfrith’s days.

Bede went to great length to write down the debates verbatim, for posterity to read, he considered these matters so critical. We might think that was a straightforward case of correctness for him, but life was suddenly to get far more complex, ambiguous, unstable, uncertain, an event that would throw a shadow of doubt over everything.

For the third momentous event of 664 was now to knock their world asunder. Shortly after the eclipse, and the synod, as Bede starkly wrote: ‘And the pestilence came.‘ One of the worst waves of the bubonic plague to ever roll across the world.

This wave would cause death on a scale that is unthinkable, the death notices in the annals of kings and bishops attest to that. Tuda, appointed bishop of the Northumbrians after the synod, was struck down. Amongst the list of dead were Cedd, bishop of the East Saxons, and all 30 of his monks save one small boy, gathered around the shrine of their saint, vainly praying for safety. Saint Cuthbert would be one of the few who miraculously recovered while most of Lindisfarne’s monks died[12]. Ripon, Ely, Gilling, Melrose, Jarrow, Barking, Lichfield, Lindisfarne, Wearmouth, the plague continued to randomly ravage the monasteries for 16 long terrifying years, killing the religious of one church after another. Ecgfrith’s first wife, Aethelthryth, lay among the dead, three days after having her ‘tumour’ lanced in an effort to save her life. Ecgfrith’s eldest brother Alhfrith, who had been a key proponent of the ‘Roman’ method at Whitby, disappears from the records in the same year 664, another likely family victim of the plague[13]. The list of dead in Ireland reads like a who’s who[14], one dead after another. In a world where plague was routinely viewed as reflecting God’s anger, the devastation of the monasteries and their rulers coming so close on the decision made by the synod, must surely have raised concerns that the decision had been wrong, that it had displeased God.

To make matters worse, the Ionan church, the principal ‘loser’ of the Whitby debate, claimed that the Dal Riatans and the Picts remained free of plague, under the protection of their saint, Columba[15]. If true, or even if claimed to be true, this protection of God through his Ionan church, will surely not have passed notice, a stark contrast perhaps as an indication of God’s view on the Whitby decision.

I’m not alone in thinking that God’s displeasure in sending the plague was likely linked to the Synod of Whitby’s decision being wrong. David Woods summarised his article on this topic[16]:

Adomnán does not seem to have believed that he had to look far to discover why God had spared the Picts and Irish in Britain alone when he had struck Britain, Ireland and the rest of western Europe with two great epidemics of plague c. 664 and c. 684. The coincidences in timing and geography pointed firmly in one direction: he had done so in order to reward them for their continued use of the right Easter table while he punished all the other peoples for their use of the wrong Easter table.

We’ve just lived through a 3-year pandemic that had a death rate of around 3 in every 100. Estimates of the death rate in 664 vary between 70 to 90 in every 100, and this wave would last in smaller outbreaks for 16 years. Northumbria may have even suffered outbreaks of the pneumonic version of plague with a kill rate of close to 100 per cent, as we hear at Lindisfarne the plague started in winter at Christmas and continued through the cold months for a whole year[17]. Plague kills a person within a day or three, it is sudden, random, unstoppable, people then knew no cause, and no cure. We must not underestimate the abject, helpless terror that the plague instilled in the hearts of victims and survivors.

The Justinian plague had started in the mid 500s, and it continued with outbreaks in small and large waves for the next two hundred years. Stories that were recorded still shock, the mother sitting wordlessly insane by her dead child’s bedside, survivors wandering in a stunned daze till they dropped and died from shock, the man who has lost his wife and children between dawn and dusk, scattered corpses left to rot with no-one strong enough to bury them, corpses shoved by their thousands into the city walls until the stench made the city unliveable.

With no known cause or cure, the church had to find a way to explain the devastation, why so many were dying, including the innocent children. They did this by saying that God had sent the plague as a dire warning to all people to keep to the path of righteousness and to stop their sinning. This one message is documented by writers all over the Christian world, with one notable exception, the Venerable Bede.

Bede does repeat the assertion of Gildas that the earlier mid 500s wave was divine punishment for the sins of the Britons. But on the major plague starting in 664 which Bede witnessed himself as a 15-year old, he strangely pulls back. He admits that the plague is ‘a blow sent by God the creator’, but does not focus that on the sins of a more specific target, unlike his contemporaries. It is thought that this was because he lived in Northumbria which in this century had seen the ‘happy’ establishment and flowering of Christianity with its churches and monasteries, which surely could only have been pleasing to God[18]. There is another possibility to consider though. All three events, the eclipse, the synod decision, and the simultaneous outbreak of virulent plague with so much death visited on the Northumbrian churches, may well have been viewed as linked together in cause and effect, because these types of beliefs were prevalent in the day, attempting to make logical sense of fickle fate. Even if Bede himself did not hold the view that the plague was God’s justice for an incorrect decision at Whitby, he may have avoided the topic as being too vulnerable to a suggestion that a wrong decision had been made. Things had become too chaotically uncertain.

Monasteries were undoubtedly implicated in the transmission of plague[19]. The ‘number and contiguity of their residents’ made monasteries Petrie dishes for infections to fester and explosively spread. Exponential expansion loves proximity. The outbreak of plague is recorded with the first deaths in July 664, nearly simultaneously in south and north England, and in Ireland by August 1. This extraordinarily fast spread of the plague in 664 was likely facilitated by movement between monasteries, being hubs of population and trade. It is even a distinct possibility that the gathering of so many monks traveling to the synod had in fact brought the plague with them and taken it with them as they left. The alternative, that ships from Gaul arrived with the plague at all ports at the same time, seems unlikely as the progress of later plagues has shown such scenarios to have a much slower rate of spread. In all, there were plenty of reasons to be concerned about the role of churches in the propagation of the plague, to say nothing of their vulnerability to mass casualties.

Irish records tell a different story about the plague of 664[20]. The story goes that the Irish king thought that the population had undergone such an expansion that it would inevitably lead to widespread famine and death – which it actually did cyclically. So he asked all the major saints of Ireland to bring down the plague to reduce the overpopulation. Most denied the request, but a handful agreed, they fasted against God and in response to the criticism he sent the plague. One of these saints was Fechin of Fore, who himself would die in the plague[21].

It may also be relevant that the outbreak of plague in Ireland is reported on a specific date, August 1, Lughnasad[22]. Great gatherings of people from far and wide were held at this time, in particular at Tailtiu, in Brega. So it is possible that this gathering did indeed cause wide and sudden transmission, just as we’ve recently witnessed gatherings do during covid.

So, the Northumbrians had available to them a story that the Irish church were implicit in causing the plague in 664, and perhaps another that a pagan festival was responsible. They were also hearing a claim that the Ionan church was being protected by God from the plague. Added to this was a coincidental link between the eclipse on an auspicious date, the synod’s decision, and the outbreak of plague in Northumbria, which may have been suspected as God’s judgement on a wrong decision by the synod.

Ecgfrith, as a nineteen-year old in 664, lived through these chaotic, terror-filled, uncertain, grief-stricken, often hopeless days that followed the Synod of Whitby. He had watched so many people around him die suddenly, inexplicably, in agony. He would not forget. Nor would his people.

The major wave of 664 continued in local outbreaks till about 680, when it waned for a few years. But, experience told them to expect another outbreak. It was just a question of when, and what would trigger it.

STOPPING THE PLAGUE : THE BREGA DEVASTATION of 684

Twenty years after 664, we come back to 684, the year of the Brega raid. The year that the next major wave of plague broke out.

Just as in 664, an eclipse had occurred on May 2 of the previous year. Although partial this time, it still would have caused trepidation of what was to come, given what had happened last time. And as if to confirm this worry, in the same year a major disease killed ‘all animals in common, throughout the world, to the end of three years, so that scarce one survived in the thousand of every kind of beasts’.[23]

In Ireland, several annals for 683 noted a children’s plague which would last for three years. The next year, 684, we have recorded ‘The plague of youths, in which all the chieftains and nearly all the young Irish noblemen perished’. While we can’t be certain that this plague of children was the bubonic plague, it is more than likely[24]. Larger waves of plague tended to happen when a new generation had been born without any immunity from previous exposure, making these later waves killers of children and ever more devastating.

Imagine the concern of everyone in Britain and Ireland, standing, waiting, watching, for the plague to arrive at their own place. Well, a lot of people didn’t stand and wait, they fled. They fled in terror to the farthest places they could get to. Inevitably, unfortunately, taking the plague with them. In Muslim countries, fleeing was made a capital offense, flee and we’ll kill you. In Ireland, we have a description of the Irish inlets thick with boats (presumably refugees from the continent) while the king Diarmait, in great fear of the plague, plead with saint Ultan to raise his hand ‘against the evil’[25].

But in 684 this was not an option for the King of Northumbria, now Ecgfrith in his prime. His role was to protect his people, however he could.

He had no science to explain the plague, no medicine to cure it, no vaccine to hope for, but he did perhaps have stories of how the Irish church was implicated in the promotion of the previous wave of plague. And he may have felt vindicated in his belief in their complicity if he thought what Adomnan reported was true, that the Dal Riatans and the Picts were exempt from the plague through the protection of Columba.

Here was something solid that he could do, he could direct his anger at the culprits. He could destroy the churches that were implicated in the spread of plague, or its non spread as the case may be. It may be no coincidence that he chose Brega to attack with ‘hostile rage’, as one of its main churches in easy reach of the coast was the monastery of Terminfeckin, the sanctuary for the ill, whose patron saint Fechin had reputedly caused the 664 plague’s outbreak. And for good measure, Brega was a region where Adomnan of Iona was the saint for the Irish high king.

STOPPING THE PLAGUE : THE PICTISH DESTRUCTION OF 685

The Brega raid unfortunately did not stop the next major wave of plague. It struck hard the following year in 685, the year Ecgfrith attacked Pictland.

We are not told that Ecgfrith’s Pictish ‘ravaging and laying waste’ was specifically targeted at Pictish churches, but the churches and monasteries were the only really obvious objectives to destroy in this period, just as the Vikings would discover a century later. It is often overlooked by historians that Ecgfrith did not just race into Pictland to meet the Pictish king in battle. Bede says explicitly that before the battle in which he lost his life, he had ‘devastated their territories with the most atrocious cruelty’.

It seems likely that Ecgfrith had tried to avert the coming plague by destruction of key Irish and Pictish churches, implicit as they were in their supposed power over the plague, which had seen their own regions protected, while Northumbria was badly hit. His targeting of churches would explain why his own churchmen had urged him against both the raids. If this scenario holds true, one of his main Pictish targets would have been St Vigeans, the church on its high mound at Arbroath, Angus, dedicated to the same plague-bringing Saint Fechin, only 10 miles from Dun Nechtain, the site of the battle. He wasn’t a king to wait at home made helpless with terror, he took what action he could.

At a time when written records were famously terse, we do find the terror of people about the coming plague leaking through. The fear and dread stand out clearly in the story of Cuthbert giving a sermon, in which he urged the monks to watch, remain stedfast in the faith, and let no temptation find them unprepared[26]. Bede in hindsight is citing this as Cuthbert having prescience about Ecgfrith’s death, but

‘some thought he said this because a pestilence had not long before afflicted them and many others with a great mortality, and that he spoke of this scourge being about to return.’

Bede then relates how Cuthbert told the story of how he had in the same way foretold the coming of the plague by urging the monks of Lindisfarne: “I beseech you, brethren, let us act cautiously and watchfully, lest, perchance, through carelessness and a sense of security, we be led into temptation.' Then, on returning to Lindisfarne the monks found that the plague had arrived as forewarned:

…they found that one of their brethren was dead of a pestilence; and the same disease increased, and raged so furiously from day to day, for months, and almost for a whole year, that the greater part of that noble assembly of spiritual fathers and brethren were sent into the presence of the Lord. Now, therefore, my brethren, watch and pray, that if any tribulation assail you, it may find you prepared.

The fear of the plague was every-present in the mind of everyone, even at the moment when their king had just been killed.

In a world of random pestilence, without cause, without cure, the only hope was the mercy of God through the intercession of his worldly saints. Ecgfrith was not alone in his belief that saints could prevent, or bring down, the plague.

Today, we, like the churchmen of his day, cannot condone or excuse Ecgfrith’s cruel destruction of the Irish and Pictish people and churches. But history is rarely a simple matter of black and white, history is hugely complex, and perhaps we can start to understand the deep, gut-wrenching terror motivating Ecgfrith to try to stop the plague before it could once again wreak its death and devastation on his own people of Northumbria.

[1] Fraser, J.E., 2009. From Caledonia to Pictland: Scotland to 795. Edinburgh University Press, p 212-216

[2] Adomnán, Life of Saint Columba (transl Richard Sharpe), Book 1, Chapter 1.

[3] Bede, Historia ecclesiastica gentis Anglorum, Book 111, Chapter 12.

[4] Annals of Ulster 685.2

[5] Bede, Prose Life of St Cuthbert, chapter 27.

[6] KBR (Royal Library of Belgium) MS 5301-20 p.30 https://www.isos.dias.ie/RLB/RLB_MS_5301_20.html#48   

[7] Whitley Stokes (ed. & tr.), “The Bóroma”, Revue Celtique, 13, 1892, pp. 36–117, 299–300.

[8] Annals of Ulster AU 687.5

[9] Pelteret, David, The Land of the English Kin, Chapter 10 The Northumbrian Attack on Brega in a.d. 684, p. 214-230 https://brill.com/display/book/edcoll/9789004421899/BP000013.xml?language=en&srsltid=AfmBOop1022YCP-HedRTlKU5VP-v7mwO8SmHPm39OhJ6wrhT7oYsQjK4#ref_FN000827

[10] Higham, N. J. (2015). Ecgfrith: King of the Northumbrians, High-King of Britain, p. 207-208

[11] McKay, H.T., 2024. Identifying the Seasonal Placement of the Coligny Calendar, Etudes Celtiques, 50, pp.37-270.

[12] Maddicott, J.R., 1997, Plague in seventh-century England, in Little L.K. (ed.), 2007, Plague and the end of antiquity: the pandemic of 541-750, Cambridge University Press, p.173-174

[13] Maddicott, p.179

[14] Dooley, A., 2007. The plague and its consequences in Ireland, in Plague and the End of Antiquity, p.221-222.

[15] Life of Columba, II.46

[16] Woods, David, 2011. Adomnán, plague and the Easter controversy. Anglo-Saxon England, 40, pp.1-13. https://www.jstor.org/stable/44479625

[17] Maddicott 1997, p. 185.

[18] Maddicott 1997, p. 180-182.

[19] Maddicott, 1997, p. 188-191

[20] This story is found in the Latin Life of Saint Gerald of Mayo, and in the Liber Hymnorum for the hymn Sén Dé by Colmán of the moccu Clúasaig.

[21] This story is found in the Latin Life of Saint Gerald of Mayo, and in the Liber Hymnorum for the hymn Sén Dé by Colmán of the moccu Clúasaig.

[22] Dooley, A., 2007. The plague and its consequences in Ireland. Plague and the End of Antiquity, p.220.

[23] Annals of the Four Masters, col. I, p. 290, xz,x 684

[24] Michelle Ziegler, The Mortality of Children, Ireland 683-685 (blog) https://hefenfelth.wordpress.com/2011/10/03/the-mortality-of-children-ireland-683-685/

[25] Felire Oengusso, 198-203, in Dooley, p 225.

[26] Bede, Prose Life of Cuthbert, ch27

In this blog I want to talk about some of the key findings about the celebration of Beltane, which have come out of discovering where the ancient Coligny calendar sits in the year. To read more about this analysis and how these conclusions were reached, you can find it in the latest issue of Etudes Celtiques1.

Before I start though, let me say that there is good evidence that this Celtic calendar was a scientific piece of time-keeping, the most magnificent luni-solar calendar ever designed by any human endeavour, one that was created within a deeply philosophical and mathematical mature culture that covered all of western Europe. Creations like this don’t happen in a vacuum, they take centuries of philosophical discussion and wrangling, mathematical understanding, deep knowledge of the skies, and a populace that consistently and persistently supported the institutional effort with hard-won resources. Let no-one tell you this is just a one-off, or just a Gaulish project, everything on the calendar can be found in all the Celtic language speaking regions, without exception. Including the festival of Beltane.

Placing Beltane in the year

So, where does the festival of Beltane sit in the original pagan year? The Coligny calendar tells us Beltane is exactly one lunar month (30 days) before the midsummer solstice.

The calendar itself does not explictly name the festivals, but marks each day of the festival with the notataion IVOS ‘festival’. But there is no reason as yet to not think that the festivals as named in Irish documents were called the same everywhere.

Now for the last few centuries, how these festivals originally fitted between the solstices and equinoxes has been a matter of huge debate. Today of course Beltane is celebrated on May 1, but that is a solar data from the Julian calendar, introduced by the Romans and later used by the Christian church, so we can be pretty sure these are at best approximations for the original pre-Julian dates.

These festivals are today often dubbed the ‘cross-quarter’ festivals, but that is not the case in any earlier documentation. But because of this false ‘cross-quarter’ notion, people have then tried to figure out how to measure this ‘cross-quarter’ moment – is it the day half-way between a solstice and equinox? Or when the sun is half-way along the horizon between a solstice and equinox? Or something else half way? The answer is none of these at all. Beltane is one lunar month before the summer solstice, and the other quarterly festivals are at subsequent intervals of three lunar months each.

This means that Beltane was a couple of weeks closer to the summer solstice than May 1. In what I called the Base Year of the calendar, when all the dates align on solstices and other significant moments, the date the festival of Beltane starts is May 20, with May 22 as the middle (most important) day, the first day of summer, the first day of MID SAMONI, the ‘month of summer’, equivalent to May 1 today. The festival ran for 5 days, which is relatively short, compared to the midsummer festival a month later of 9 days.

Beltane as a luni-solar festival connected to the summer solstice

While Beltane is pegged to the lunar month before the midsummer solstice, it is also a luni-solar festival, its central day on the first quarter moon. So, Beltane moved around in the solar year much like Easter does today, and that luni-solar variance can be up to a month.

One of the things I’m not too sure about as yet, is where in the Metonic cycle of 19 years the Base Year lay. But, if it was the first year of the Metonic cycle, then Beltane would have been celebrated at some point from May 20 up to the midsummer solstice on June 21!

That is a bit of a surprise, but the calendar itself is giving us clues that this is likely the case. The crucial central day of Beltane, SAMONIOS Day 1 has its notations swapped with those of DUMANIOS Day 1, the day of the summer solstice, linking these two days of Beltane and midsummer. These two months of SAMONIOS and DUMANIOS swap other days too, most significantly they also swap the days of their full moon. These two months are so intimately linked, almost as if we should consider them part of the same time event.

Beltane and the start of the Celtic year

Another question which has raged back and forth for the last few centuries is where did the Celtic year begin?

Let’s start with the obvious. The Coligny calendar starts with MID SAMONI, the month of summer, and has MID GIAMONI, the month of winter, six lunar months later. This breaks the year into two halves, the summer season and the winter season, just as we are told in the Irish manuscripts. And this division is emphasised by several of the notations whose patterns are similarly divided and mirrored from one half year to the other half year.

Beltane starts the summer month, MID SAMONI or in Irish céitemain ‘May’, which in turn starts the summer half of the year. The calendar also shows the festival of Samhain, at the start of MID GIAMONI, Mi Gam ‘November’ in Irish, starting the winter half of the year.

So far so good, as expected.

But many reading this will then object, don’t old manuscripts tell us the Celtic year started with Samhain? Well, no, they don’t. Certainly, certain activities were held or initiated at both Beltane and Samhain, the start of summer and winter respectively, but nothing in the documents explicitly says that Samhain started the year.

So we have no reason whatsoever to doubt that the calendar year starts with the summer season.

Now if there is one thing to say coming out of the Coligny calendar, is that the data and ideas that we have been thinking about and working with for the last few centuries turn out to be way way too simplistic. Too simplistic by a mile.

On the surface of it then, the Celtic year appears to have started with Beltane and the month of summer, MID SAMONI.

But the calendar then continues to show us this is too simplistic a solution in a broader sense. The calendar runs for 5 years, and it is broken into two halves, each of 30 lunar months (two and a half years), each preceded by an intercalary month (to keep the solar years aligned). And the second intercalary period starts with the festival of Samhain in MID GIAMONI, the winter season, the reverse of the first half starting with the summer season. And all the notations in the two intercalary periods are mirror images of each other, one equal and opposite to the other.

Then too the intercalary months have notations copied from various days, no doubt for purposes of prognostication. And both these intercalary months start off with an equal mix of both summer and winter days, as if the two are to be equally considered.

In other words, the calendar itself is continually showing a dance between two equal seasons of summer and winter. To say one comes before the other is to ignore the circularity of time itself, to ignore that summer and winter cannot exist without each other, that one is the mirror image of the other. Yes, the calendar does put the summer season at the start of its lunar year, but that is likely just a side effect of having to start somewhere when you are writing something down. Time is complex, complicated, and we live in this beautiful progression of moments, one to the next, each connected to what came before, and all the days to come.

Dating the festivals

Another question that modern researchers have long debated is just how old these festivals are – how far back do they go? Some have even gone so far as claiming they’re a fairly recent fantasy. But the important thing the calendar’s dating of the seasons and their festivals tells us with absolute certainty, is that these festivals are really ancient, at least going well back into the early Iron Age - and who knows how long before that. And in turn, it teaches us that we should take more seriously exactly what the ancient writers recorded. We need to treat the history of the past with professional respect and care, it belongs to real humans no matter how long ago they lived, no matter what they believed.

So Happy Beltane again! And hang on tight, the great festival of midsummer is coming shortly!

Subscribe now

This is the most consequential paper I will write about the calendar. It lays out how the calendar is placed in the year, by using particular notations that are unique and don’t just follow a general monthly pattern, and by expecting that all these points when taken together should sit at significant and identifiable points in the year, such as the solstices and equinoxes. Fortunately, there is only one layout that fits all the points. So, this analysis tells us how the Celts all over western Europe and the Isles, including the Picts, conceived of time, how they counted it, and organised around it. And this time, we’re not just guesstimating or making assumptions, this time – finally – we have the ancient pagan Gauls giving us a mathematically precise solution.

For those of you with access to a university library you should be able to get a copy soon, everyone else will have to wait the prerequisite two years for it to be put in the public domain. But then, as we’ve waited through 140 years of research into the calendar, I guess we can wait a wee bit longer.

I do expect some pushback. For some reason this topic is quite emotional for a lot of people, and over the last few centuries a number of ideas have become popular, some of which turn out to be valid, but some not. Like the idea that Samhain starts the Celtic year – which it doesn’t, or that the Celts counted from the dark times of the day, month and year – which they don’t, they only counted the day from sunset just as most of their neighbours did, nothing more. Or that their lunar month started with the full moon, which it doesn’t. Or that the notation TRINUX SAMONI refers to three days of Samhain - which it doesn’t.

From the moment I started working with the Coligny calendar, the only question people kept on asking me was – how does the Celtic year fit with the yearly cycle? And from this the related questions - How did they organise their festivals? Where does the Celtic year start? And so on. For a long time I brushed off these questions, because the way to solve a really big problem is to first identify the elements of that problem, find the data, analyse the data, understand that data and all its interactions – of which there are roughly 50,000 on the calendar. It’s a massive web of different notations, all with their own patterns, and all interacting with the other notations in a crazy, wonderful, mathematically mesmerising dance.

So I started from scratch. I took each notation, analysed its patterns, then compared my solution to other researchers’ solutions, discovering any discrepancies, analysing the differences, adjusting, testing, and re-testing, because this method not only provides a valid data set, but it also uncovers how researchers are conceptualising these issues. A lot of work had been done by previous researchers, but as time went on, I stood on their shoulders and added to the body of knowledge.

I’ve published three papers in this process, the first1 was to solve the most difficult and trying question of the accuracy of the calendar, to show that this calendar was indeed as early researchers had been sure it must, a beautifully accurate lunar calendar that kept each month in phase with the moon, and an extraordinary solar calendar using the Metonic cycle, with a variable day to keep in sync with the month and year forever.

The Coligny calendar is the most accurate, complex, and utterly beautiful luni-solar calendar ever created by mankind.

The second paper2 resolved the pattern of the ‘triple marks’, a complex web that is keyed to the fortunate and neutral lunar months over the 30-month intercalary period. The ‘triple marks’ divide the daylight hours into three periods – just as Irish stories describe.

The third3 paper was consequently able to resolve how the notation patterns of the intercalary months were created by a complicated system of copying notations throughout the 30-month intercalary period. A wonderful picture of the ‘borrowed days’ of Celtic lore.

I have to admit, I loved doing this work, I find it intensely satisfying. Especially when I am dealing with an actual historical problem whose resolution holds promises of access to crucial information.

And especially when it turns out that this amazing calendar reflects a beautiful mathematical sense that the Celts employed, often ideas that are alien to our thinking today. For instance, the most important point is neither the start or end of a sequence but the midpoint. Take for example the full moon. Historical events and folklore frequently tell us that the full moon was viewed as critical to the Celts – and it is, the calendar makes that quite clear as many notations focus and dance around this point in the month. But the full moon is in the middle of the first fortnight of the month – not the start of the month, and not the end, as we’d expect of a counting sequence today, but in the centre.

The calendar’s astronomy is also exceptional. Well the ancient commentators did tell us that the Celts were exceptional astronomers and mathematicians, and the calendar backs that up in a big way. Take for example the new moon – another point in the lunar cycle that the Celts and many other cultures view as a significant point in time. The new moon is just momentarily visible near the horizon, so spotting it (and declaring the event if that culture started their lunar month at that point) was always problematic, with weather and topography often thwarting the attempt. But the astronomy of the Celtic calendar doesn’t use unreliable ‘sightings’ but is calculated to the precise moment, because the whole lunar phasing is precisely configured. Stars too are marked for their rising with the sun, not relying on someone spotting a fleeting helical rising. Lunar months are organised to keep in sync with the phase of the moon with a tolerance of less than a day – forever. The whole notion of time has been lain out with mathematical precision and wizardry. Every insight is a joy and a wonder.

The calendar starts with the equivalent festival of Beltane, positioned at precisely a lunar month before the summer solstice. It also has other key points, like the full moon at the winter solstice. And that’s just the start of a cascade of wonderful revelations about Celtic time-keeping.

Placing the Celtic calendar into the year is just cracking the door open to all sorts of amazing surprises. It tells us how the festivals are organised, how long they were, what dates they were, and how many there were. It tells us what the Celts considered to be the most exceptional points in time, such as the full moon at the winter solstice. It tells us about the astronomy and philosophy and mathematics and organisational skills that all contributed to this magnificent and unparalleled piece of science.

But today, because it’s Easter, I’d like to tell you the story of St Patrick lighting his Pascal fire in defiance of the Irish king and all his elite who were gathered to perform rituals on the Hill of Tara, the Feis of Tara, on what one author called the ‘easter of the pagans’.

This is the story as found in Muirchú's 7th century Life of St Patrick:

It so happened in that year that a feast of pagan worship was being held, which the pagans used to celebrate with many incantations and magic rites and other superstitious acts of idolatry. There assembled the kings, satraps, leaders, princes, and the nobles of the people; furthermore, the druids, the fortune-tellers, and the inventors and teachers of every craft and every skill were also summoned to king Loegaire at Tara, their Babylon, as they had been summoned at one time to Nabuchodonosor, and they held and celebrated their pagan feast on the same night on which holy Patrick celebrated Easter. They also had a custom, which was announced to all publicly, that whosoever, in any district, whether far or near, should have lit a fire on that night before it was lit in the king's house, that is, in the palace of Tara, would have forfeited his life. Holy Patrick, then, celebrating Holy Easter, kindled the divine fire with its bright light and blessed it, and it shone in the night and was seen by almost all the people who lived in the plain.

For centuries now people have argued about this story. Is it an entirely fanciful religious tale of how Christianity claimed the souls of Ireland? And, was there ever really a special pagan ritual day at the same time as the Christian Easter?

Well the calendar tells us the answer is yes, there was. Not only that, it was one of the three most important days of the Celtic year. A day of magic and ritual.

Easter from very early days in the church is calculated as the Sunday after the full moon of the Spring Equinox (and all the various rifts and arguments were largely about how to exactly define the precise timing of each of these events). So Easter is a luni-solar date. And the Celtic calendar is also luni-solar, and its special day was on the full moon of the spring equinox. So the two dates could easily have fallen together if Sunday occurred within the 24 hours of the full moon – which it will frequently do.

Sometimes, history really is stranger and far more extraordinary than fiction. And not everything we are told can be dismissed as fanciful nonsense.

So Happy Equinox again, hope the spring puts a little happiness into your life! From the depth of cold dank winter eventually comes the light and warmth once more. Well done sun, well done moon!

Subscribe now

Share

A general pattern was identified, of a small figure flanked on the left by a large figure with a weapon over their shoulder, an axe or a sword, and another large figure on the right holding a shield or some other object above the head of the small figure.

All these small figure icons are found on a cross, not on the reverse with the actual battle scenes. It puts the icon in a context of both battle and religion.

All examples are found in a band running from Pictish Fife through to the region of Alt Clut around Govan and Dumbarton rock, all except Sueno’s stone in Moray. This distribution is restricted in place, and in time, likely the 700s. But, this small figure icon is not found anywhere else on carvings or manuscripts of their day.

Here again is the collage of the small figure icons.

What are we to make of this small figure icon?

Naturally, there have been several attempts to interpret the small figure icon over the years.

The earliest suggestion was by Mann, as representing a passage from the Book of Exodus. Another suggestion was the baptism of Christ1, but it is clear that weapons are involved here, not rivers and water. Nor would the Picts have depicted Christ as a small, that is relatively unimportant, figure.

Another biblical suggestion has been Samuel anointing David2, but David is usually portrayed as a full grown adult kneeling, rather than a child standing, and again there are no apparent items used in an anointing.

The sacrifice of Isaac has also been suggested3, with the figure on the left perhaps the angel protecting Isaac and preventing his sacrifice, although this is not how the Picts depict their angels. The usual depiction elsewhere has a small table altar with Isaac bending over it, and the ram in the bush behind, none of which is evident in the small figure icons. But most of all, it is hard to see how this biblical event could have a key importance within the explicit military context of these stones.

Of course, none of these suggestions allow for the fact that neither Pictish crosses nor the Alt Clut crosses appear to otherwise feature biblical scenes, with the possible exception of a strange Pictish version of David killing the lion. This era of art may be responding to the iconoclasm rampant in the contemporary classical world.

Geddes has recently put forward the suggestion that the icon represents the arming of David4. This does have the immediate advantage of placing the icon in a military context. However, if King David is the central figure, then it is surprising that he should be so small. The Picts elsewhere portray David killing the lion as a large warrior dressed in imperial robes despite him being a young shepherd boy in the bible when he smote the lion, so it seems improbable that as a warrior receiving weapons he is perversely depicted as a small child.

Nor does the arming of David occur as a popular biblical scene on other crosses, the only instance in the world being on one of a series of silver Byzantine plates found in Cyprus and dated between 613 and 630 AD, well over a century before, each plate depicting a different scene from the life of King David. These silver plates also feature David killing the lion by kneeling on its back, unlike the unique Pictish version where he stands beside the lion, making it very unlikely that these unique plates, or any unattested related template, are the source of the small figure icon.

Icons of David, usually him killing the lion, were indeed a popular motif from early on, with many examples in manuscripts and carvings all over the world, yet despite this popularity not a single manuscript instance of the arming of David is known. The other thing that tells against this interpretation is that the small figure has a shield seemingly held above him to protect him, the shield is not being handed to him, or rejected by him as the bible relates.

Sueno’s stone is all about a king in his prime boldly proclaiming his victories to the world, his sword held high. If I am right to identify this king as Oengus I, (in)famous for his many bloody battles, he came to the throne at a relatively old age, and was even older when attacking Dal Riata in the 740s. To then expect an icon which portrays a small boy rejecting armour seems perversely out of character.

The explanation that this Pictish king is modelling himself after David as he rejects real armour preferring God as his armour is antagonistic to everything in art, history, texts, that we know of the Celtic worldview, in which warriors are very much part of their armour which itself often has agency. Celtic stories are replete with tales of how warriors receive their armour, all are about celebrating this coming of age event and how important it is. Pictish men too are next to never without their weapons prominently shown, including the king and all his warriors on the Sueno’s stone itself. The Celtic worldview would not find a warrior going into battle without his weapons as explicable or culturally appropriate, no matter what one story in the Old Testament says.

Finally, another idea is that the small figure icon represents an inauguration scene5. There are several things that tell against this idea. First, we would not expect the king to be represented as a figure significantly smaller than the other figures as his subjects. Nor is there anything to suggest that the Picts ever placed a small child on the throne.

The bjannak

Which leaves us to find another interpretation for a key icon found in a battle context with serried ranks of warriors and trumpeters. An icon occurring with a cross, but seemingly not an important biblical scene that can be found on crosses or manuscripts elsewhere. An icon where a small figure is flanked on the right by a large figure carrying a weapon and another on the left holding a shield above his head in what seems like a protective pose. And, an icon which occurs with other Odin depictions related to battle – Odin arriving in Valhalla, the place where warriors go to after death, and Odin’s wolves who eat the battle dead. An icon that occurs on at least 6 stones in both Pictland and the kingdom of Alt Clut, possibly dating to the mid 700s.

That is a difficult question in the end. There is only one solution that I have found that might fit this scenario – the bjannak, the blessing by Odin of warriors going into battle.

Odin was so successful that he never lost a battle. As a result, according to the sage, men came to believe that "it was granted to him" to win all battles. Before Odin sent his men to war or to perform tasks for him, he would place his hands upon their heads and give them a bjannak ‘blessing’ and the men would believe that they would also prevail.6

Would the Picts really have depicted the god Odin blessing his warriors before battle? I suspect the answer is yes, because influences and trends in this 8^th century period are complex, fast changing, melding and morphing.

The word ‘blessing’ itself comes from earlier forms of the word first found as blētsian from around 725, later blessen in the Northumbrian dialect of 950 AD. The word derives ultimately from Proto-West Germanic *blōdisōn (“to sprinkle, mark or hallow with blood”). It is intriguing that once again we have a date in the mid 700s, and an Anglo-Saxon source.

The word used in later text for a blessing is bjannak, which comes through Celtic bennacht ‘blessing’, demonstrating the cross-linguistic melding of this powerful concept, Latin/Anglo-Saxon/Irish/Norse. Perhaps we are witnessing here an ancient pagan protective blessing of the warriors being transformed into a protective blessing of warriors by the Christian god and his saints.

Heimdall and Ragnarok

While the small figure icon only occurs in a limited region, the trumpeter or horn-blower does occur elsewhere.

At Gosforth on the west coast of Cumbria, a large, intricately decorated cross has a picture of a man with a horn, and in his other hand he holds a spear before the open menacing mouths of two reptilian creatures. It has been suggested that this is Heimdall7 with his magical horn Gjallarhorn, which he will blow to warn the gods that Ragnarok has begun. The idea of Ragnarok as the end of the world is similar to the Christian concept of the Last Judgement Day, and it’s thought that this is a melding of the two stories.

On the Isle of Man there is another trumpeter, similar to the Pictish ones, blowing a raised horn. Again this is considered to represent Heimdall, blowing his horn before Ragnarok.

These two stones show that the idea of Heimdall and his magical horn remain strong in Anglo-Saxon and Scandinavian areas, but the problem here is that both these stones are thought to be a century or two later than the Pictish stones. Nonetheless, it does raise the question of whether the story of Heimdall and his horn is behind the sudden fashion of trumpets in the insular world. Especially when Heimdall’s role as the messenger of Ragnarok is so easily melded with the new Christian biblical story that trumpeters will also herald the end of days.

The locations of the stones with a small figure icon

But, even should we identify this small figure icon as a protective blessing of the warrior, melding a Norse/Germanic pagan tradition with a Celtic tradition, and being transformed into a new Christian tradition, that still doesn’t answer one other key question – why does this icon only occur in these limited locations?

The locations of all but one of these stones with the small figure icon, or related trumpet stones, lie in a broad band from the Alt Clut region through Fife. The Cambusnethan, Barochan, and possibly Govan, stones are definitely outside Pictland, in the kingdom of Alt Clut, later Strathclyde. In other words, this icon is felt to be important beyond Pictland, yet it still does not occur anywhere else outside these locations in Ireland or southern Britain, implying it is an icon considered to be of particular relevance in just this limited region.

This limited distribution also suggests that this isn’t some biblical story of key importance, else we’d find it on other crosses in southern Britain and Ireland, which we don’t.

Sueno’s stone isn’t the outlier that it first appears to be. We’ve already seen the connection through trumpeters and ranks of warriors, as well as the small figure icon. But it is also related to the Barochan cross through its rare form of cross. This type of cross is the late Pictish form without a defined circular centre but still with circular inlets in the cross arms, which is also the type found on most of the recumbent graveslabs at Govan. But stunningly, Sueno’s stone and Barochan also have an outer ring, a rare occurrence on the large monumental stones (although some small Pictish memorials do have it). There are only three other possible monumental stones with crosses like this, all of them in Fife, at Carpow, Largo and Crail (blue and purple circles). This again fits into the same distribution pattern – from Fife to the Govan region. There is a distinct connection running between St Andrews and Govan as shown by these crosses. But the question then is, just what did that connection entail?

Dating the small figure icon and the battle trumpeters

Let’s now sum up some of the dating we’ve seen that may be associated with the stones with the small figure icon and the related war trumpet. This dating is definitely focused in the mid 700s, and likely represents a new fashion in the depiction and descriptions of trumpets. Each stone with a small figure icon may come from this period, or perhaps from later years, although contextual evidence for the early Viking period after 800 is lacking.

1. The real Irish trumpet, 750 AD

2. David with trumpeters, Vespasian psalter c.730 – 770.

3. Acca of Northumbria, leaves Hexham probably for St Andrews 732 AD

4. Establishment of the monastery of St Andrews by Oengus I in the 730s

5. Kinneddar has the same type of Pictish memorial stones, so mid-700s

6. Govan recumbent graveslabs also strongly feature the same form of late ‘Pictish’ cross, although on recumbents rather than upright stones.

7. The smiting of Dal Riata (the likely candidate for the Sueno’s stone battle) 741 AD

8. Northumbria defeats Britons of Alt Clut 750 AD

9. Northumbria and Picts under Oegnus I force terms on Alt Clut 756 AD

The key church centres of St Andrews in Fife, Kinneddar in Moray, and Govan in Alt Clut are all partaking in the new fashion of multiple small graveslabs over burials in a major churchyard. And their carvers, while expressing some regional differences, are also sharing icons and the form of their cross.

The locations of the small figure icons are part of this same limited group, and all are from the very late CII > early CIII era, probably the mid to late 700s. Which does present a problem as the Alt Clut stones are usually considered in art-historical terms to be much later than this.

Sueno’s stone recently had pits around it dated to the 800s8, showing that the stone was present at least by this date. If these pits are the remains of a gantry to raise the stone this may provide a date for the stone, however, that is an assumption that cannot be verified. The grouping of other related stones more likely places it in the mid to late 700s.

The next question is, what could have been happening in the mid 700s that might be relevant to such a military-associated icon? And what connects the Fife and Alt Clut regions through these battle icons?

Ango-Saxons and Picts at Alt Clut

The history of the Britons of Alt Clut is fragmentary to say the least, but if the Barochan stone in particular memorialises a battle, the dating of this group of stones and actual trumpets points to a battle in the mid to late 700s.

The Vikings attacked and destroyed Alt Clut in 860 AD. But in the years before this, the Northumbrian king Eadbert defeated these Britons in 750 AD, then, in 756 AD he followed this up by forcing terms on the defenders of Alt Clut, this time in an alliance with the King of the Picts, Oengus I. His luck turned however when his army was defeated as they were returning from this expedition – in an unidentified location.

It’s not known however what the outcome was for the Alt Clut relationship with the Picts, or what terms the Picts imposed on the Britons. Did those terms include asserting some sort of influence over the Govan church?

The connection to Oengus I is suggestive that these stones with the small figure icon are part of the story of his reign. Not only did he found St Andrews, but Sueno’s stone, if dated to his reign, is possibly memorialising the great battles in which Oengus all but destroyed Dal Riata in 741 AD.

Despite this, it may be that these battles between the Picts and Britons were the exception, that they usually lived as negotiated neighbours. During the 600s a number of Pictish kings had family connections to the Alt Clut dynasty, and this close relationship may have continued into the 700s. The Picts in particular would need access to the busy networks of the west coast through the Clyde Firth, so it would seem likely that the two groups largely cooperated to their mutual benefit. In which case we could be looking at a pattern of close and cooperative religious houses and their artists moving and sharing their art. And, the Barochan battle stone might still be standing testimony to some other battle fought here with an unidentified foe.

History is an iceberg at these moments, we see so little of what is happening on the ground, just a few titbits of battles and deaths in the annals. What we can take away from these titbits is that the 8^th century was the height of movement and interaction between regions, sometimes a battle making it to the annals, but mostly unseen and unknown interactions at many different levels of society.

The rather unexpected and quite amazing aspect of this story is just how close the regions of Fife and Govan actually were. There is a deep untold story here of the history between these two neighbours in the 700s.

Conclusion

There are at least 6 examples of the small figure icon, and another couple of stones with related battle trumpets.

All but Sueno’s stone are in a band running between Fife and Alt Clut, linking these two regions in some unspecified way.

This icon is restricted in space and time, and does not appear anywhere else.

Dating of these particular stones points to the mid to late 700s, and historically they might be associated with Oengus I, the famous battle king.

The iconography of this group of stones is focused on warriors, possibly commemorating a battle, or memorialising a battle or warrior. The small figure icon is one of these ‘battle’ icons, and on two crosses occurs with Odin-related battle iconography.

The small figure icon has no completely satisfactory explanation, the most likely being a form of the bjannak ‘blessing’ of the warrior, putting the warrior under the protection of the deity and making them invincible in victory, a concept melding both Celtic and Anglo-Saxon pagan tradition with the incoming faith.

Subscribe now

Share

This strange panel is at eye level and below a massive Pictish cross covered in intricate interlace. That this small figure icon is depicted along with the cross, and by its own without other iconography around it, presents a picture of something which must have been of key significance for the Picts. But, what?

Sueno’s stone is the biggest Pictish stone ever carved. Its reverse is covered by serried ranks of warriors, and scenes of beheaded corpses. While this small figure icon isn’t on the same side as the battle scenes, it still raises the question of whether it is also related somehow to a battle context?

The panel itself is extremely worn and its details largely unclear. But, there are several other stones with a similar small figure icon, and we may be able to further understand this icon by looking at these other examples.

The other stones with a small figure icon

GASK A similar panel may be on the Gask stone, overlooking the Dunning monastery, and close to Forteviot, a centre of political power at least by the early 9^th century. The lower LH side of the cross shaft has a leaning figure and another facing it, similar to the figures that form an oval surrounding the small figure on Sueno’s stone. Unfortunately, the surface of this stone is extremely eroded, so the details aren’t clear. But the leaning and encircling stance of the figure that can still be made out is otherwise unusual in Pictish art.

This stone has Pictish symbols on the reverse, although they are quite small and on the lower RH panel, as if in the process of losing their importance. It’s difficult to date a stone like this, but it is late in the closing CII period, in comparison to Sueno’s stone which has no symbols and has just crossed into the CIII period (while remembering these are artificial classes for ease of reference by modern scholars and in fact may not be completely sequential in dating). So it’s possible that both Gask and Sueno’s stone come from the same period, or one close to the same.

ST ANDREWS, FIFE There is another stone, much smaller this time, from Saint Andrews in Fife, again another key religious centre, established as a monastery by Oengus I in the mid-700s, and most of the Pictish stones here can probably be dated to activity in a few decades in the mid-700s. Here we have another small figure, flanked on the left by a man with a sword over his shoulder, while the figure on the right holds a Pictish round shield over the head of the small figure. The small figure is not being attacked by the figures with weapons. The position of the shield over his head appears to be protective.

Below this are two confronted beasts with a human head between their jaws and two birds above, the occurrence of which is unique in a Pictish art context, but if seen in an Anglo-Saxon or Norse context would no doubt lead to a discussion as to whether this is symbolic of Woden/Odin with his two ravens speaking into his ears and his two wolves, Geri and Freki, who roam the battle field "greedy for the corpses of those who have fallen in battle".

Odd though this solution may seem, we shouldn’t throw it out without further consideration, as there is a growing body of evidence for Anglo-Saxon, mainly Northumbrian, influence within Pictland during the 700s. St Andrews itself is considered to be founded by Oengus I with the assistance of Bishop Acca of Northumbria, so an Anglo-Saxon influenced piece of artwork isn’t out of the question. As well, the icon of the two ravens of Odin talking into his ears is so prevalent in both Germanic and Norse artwork and myth, that it would take special pleading to claim that someone created the same image but meant something entirely different by it.

In Celtic artwork, the raven is commonly used to signify that the person portrayed is dead, and in Celtic mythology ravens are also said to eat the battle dead. So it’s possible that both Germanic and Celtic lore are being melded here to indicate the same context – the battle dead.

The Alt Clut examples

The small figure icon is also found on another three stones, but this time in the Glasgow area, Barochan, Cambusnethan, and perhaps Govan itself.

It is definitely an expected turn of events to find an icon used both in Pictland and the Brittonic kingdom of Alt Clut, but it is even more unusual to not find even a hint of such a scene on any other cross in Britain or Ireland. All these crosses with their small figure icon can be linked through their uniqueness, linked in theme and concept, linked no doubt within a short period, but how might they be linked beyond that?

CAMBUSNETHAN The first is a fragment of a cross shaft from the church of St Michael at Cambusnethan, which has a small figure flanked by three larger figures, with their arms over the head of the small figure, although the details here are eroded away. The figure on the far left appears to hold a sword over his shoulder, similar to the St Andrews scene.

BAROCHAN The second cross near Glasgow is the major cross of Barochan, over the river Clyde to the south of the stronghold of the Britons at Dumbarton Rock. Here the small figure is once again flanked on the left by a man carrying a weapon, this time an axe as on Sueno’s stone, while the figure on the right holds some sort of item above the head of the small figure.

But it is what is on the panel above the small figure icon that gives pause, it is a single rider being greeted by a standing figure holding out a drinking horn. This quite common type of icon is neither Pictish nor Celtic, but Germanic and Norse, usually interpreted as Odin, or a warrior of Odin, being greeted by a Valkyrie female as he arrives at Valhalla, the great feasting hall of the warrior dead.

Below the small figure icon is a pair of confronted beasts with gaping mouths, similar to the St Andrews example but without the accompanying head and birds. (However, an earlier researcher, Mann, reported in 1919 that there was a head between the beasts’ mouths.) If this is the case, then once again we have Odin’s wolves under the small figure icon, just as on the St Andrews stone. But here the Barochan small figure icon is sandwiched between two Odin-related icons, Odin arriving in Valhalla above and Odin’s wolves below, and that strongly implies that the small figure icon too may well be related to Odin. And that throws the source of this small figure icon into question. Should we be looking for a Germanic/Norse source?

The other side of the Barochan stone has ranks of figures and a line of trumpeters, declaring its ‘prominent martial imagery’ according to Driscoll1. And it is this rare depiction of lines of battle figures that connects it to Sueno’s stone with its serried ranks of troops, something which is not seen elsewhere in Pictish art.

Neither the Barochan stone nor Sueno’s stone are located at a known church. Driscoll says this ‘may lend credence to the notion that its original context was primarily secular’. The primary candidate for this secular context could well be an actual battle of major importance.

GOVAN With two crosses in this region, Cambusnethan and Barochan, both with a small figure icon that stands out on these crosses as important, we might expect to also find one at Govan, with all its carvings at a key religious centre. And there may be, on the ‘upside down cross’. The entire icon is small and oddly situated on the side panel, but again it appears alone as if holding special significance in its own right. Unfortunately, it is too eroded to see the small figure clearly.

It’s not clear if this is meant to be the same small figure icon, as uniquely here the icon has a large seated figure to the left, rather than a standing figure with a weapon over his shoulder. The figure to the right once again holds an item above the head of the small figure, this time a small square. If the round item on the St Andrews stone is correctly identified as a round shield, then here on the Govan stone, this would be a small square shield. The Picts are portrayed with both small round and square shields on other stones.

But, the other faces of this cross have worn completely away, so there is no way to tell if this cross had other military aspects to it. For the moment, this may be a version of the same small figure icon, but then, may be not either.

The battle trumpeters

As well as the small figure icon, the Barochan stone has other unusual connections to Sueno’s stone.

On the other side of the Barochan cross is a line of trumpeters, blowing trumpets upwards, and above the trumpeters a row of men. Sueno’s stone also has a similar line of trumpeters in a scene of battle and beheading. The stuicc chatha ‘battle trumpets’ are mentioned in Irish texts, and an actual trumpet has been found in the River Erne, 58 cm long, made of yew with brass bands, and dated to 750 AD.

The Barochan stone was not located at a church, but on the main route leading north across the ford to Dumbarton Rock. Barochan Hill was the site of an Agricolan period Roman camp, which may have still been recognisable as such at the time the cross was carved. This has been identified as Coria of the Damnonii, a Celtic word for ‘hosting place’. Add to this its icon of a warrior arriving in Valhalla, and the wolves of Odin below, the lines of warriors, and then the trumpeters that we see involved in warfare on Sueno’s stone, all set out a strong military theme for the Barochan cross. So, the military contexts of both the Barochan stone and Sueno’s stone does support the idea that the small figure icon is also associated with battle. And that both stones were either hosting places or sites of actual battles.

DOGTON, FIFE Trumpeters are a rare sight on carved stones, and there are only two other stones apart from Barochan and Sueno’s. The first is a free-standing cross shaft, at Dogton in Fife. Unfortunately, the stone is very eroded so it hard to make out. But it too may have a rider with a figure before it as on the Barochan stone, and below that panel appears to be a couple of trumpeters similar to Barochan and Sueno’s. This time there is no small figure icon with these two panels (that we can now see due to its eroded state), but finding icons that go together on multiple stones does suggest that the Dogton stone is another ‘battle’ stone, possibly marking the site of an actual battle. Alternatively, the area, just inland from landing beaches in the Firth of Forth, may have been an ancient hosting place, a place to gather men in the face of invasion from the south.

ST ANDREWS (another one) There is yet another stone at St Andrews which may have a single trumpeter standing behind a half-size figure, a picture which is a little different to the other small figure icons but is possibly still part of this same theme.

The connections between stones with this small figure icon

Below are the stones with a small figure icon, grouped together for comparison.

All these icons focus on a small figure at its centre. In Pictish art the most important figure in a scene is usually slightly bigger, not smaller. There are two possible interpretations of this, that the central figure is a small child, or that the flanking figures are far more important, possibly deities.

The figure on the left holds a weapon, three times a sword (St Andrews, Govan, Cambusnethan) and twice an axe (Barochan and Sueno’s), with Gask now too eroded to tell.

The figure on the right either holds a shield on some item above the small figure’s head, or their body is bent around it as if to protect it.

On all stones this icon occurs with or on a cross, but most critically, it isn’t just one panel amongst many, it is usually alone, the implication being it holds a key significance.

Yet this icon isn’t a common theme on other crosses across Pictland, and it doesn’t occur outside these few locations.

The icon occurs in both Pictland and the kingdom of Alt Clut, but at places which may have an Anglo-Saxon connection or influence.

These stones have other features in common, such as warriors, trumpeters, and confronted beasts, which places this icon into a military context. On both the Barochan and St Andrews stone, accompanying icons appear to be of Woden, again indicating a military context.

All these stones come from the late CII into the early CIII period, which probably means they can be dated to the mid-700s into the early 800s.

Making sense of the Sueno icon

The small figure icon on Sueno’s stone is quite eroded, but given the pattern of the other instances, we can now try to redraw an interpretation. The flanking figure on the left is not holding the small figure’s arm, but holding the shaft of his weapon, a Pictish axe of form seen on Golspie and Aberlemno3. The righthand flanking figure is too eroded to be sure, but if the pattern holds, he will be holding a shield over the head of the small figure, not holding the arm of the small figure.

The next step

So far, I’ve collected the examples with a small figure icon, and found several related battle-focused stones with trumpeters and warriors. And I’ve noted the key facets of the icon - a central small figure with a larger figure holding a weapon over his shoulder to the left, and another larger figure to the left holding a shield in a protective stance above the head of the small figure. All of which has allowed me to try to redraw the Sueno’s stone icon, putting it into the context of the other examples.

In the next part of this blog, I will look at the potential significance of the small figure icon, the various theories, and try to find a solution that will fit the battle context of these stones.

Subscribe now

Share

The double-disc and crescent symbols, when without a rod, both have notches in them. However, the snake and notched-rectangle symbols, when without their rod, don’t seem to add an extra notch.

The double-disc with a notch

There are three CI examples of the double-disc with notches, Newton House, Westfield1, and Inchyra. The only other two examples of a double-disc symbol, Drumbuie1 and Fyvie2, are both no longer complete, so it’s not possible to say with certainty whether they would have had a notch or not.

The crescent with a notch

The crescent also has a circular inlet in its lower edge when losing its rod, as seen on the Daviot and Roseisle stones. There is only one other Class I crescent, at Breck of Hillwell, Shetland, however I have only seen a sketch of this badly worn stone, so I can’t myself confirm whether or not it too has a notch.

The rectangle symbol with a notch

A stone found at Dyce in 2018 has a small rectangle symbol with a notch in its upper edge. The rectangle has lost its dominance in the symbol pair as is shown by its small size in relation to the larger cauldron symbol below, plus the mirror+comb which indicates that dominance rests with the lower symbol on this stone.

This is the only case of a symbol that does not normally have a rod with floriate ends being given a notch when it loses dominance. However, the rectangle symbol usually has one, occasionally two, lines dividing it horizontally.

The examples of this symbol which are small compared to the accompanying symbol have lost their dominance in the symbol pair. These small non-dominant versions of the rectangle, for example the Old Leslie stone, do not have a horizontal line dividing them internally, so it may be that the dividing line which is seen on the full-size symbol was regarded as the equivalent of a rod.

This recent discovery at Dyce is not the only rectangle with an inlet. A while ago, another rectangle symbol with a rounded inlet, Ackergill2, was found, unfortunately without its accompanying symbol.

So for some reason, the rectangle symbol was considered appropriate to mark with a notch when not dominant on a stone. Which likely means we should think of this symbol with its horizontal dividing line as in the same group as the rodded symbols.

The ‘flower’ symbol

The square inlet in the Dyce rectangle may also be related to a square inlet in the base of the so-called ‘flower’ symbol (possibly meaning ‘fire’). Usually the ‘flower’ symbol has just a straight base, but on a couple of stones, Golspie and Park House, this symbol has a square inlet in its base.

On the Golspie stone, the notched ‘flower’ symbol is paired (by size) with a rodless crescent. The Park House stone has a mirror+comb, so the notched ‘flower’ is the upper symbol that is losing its precedence. Both these examples then are in a context where the symbol with a notch has lost its precedence.

Despite this, it’s not a clearcut case. Just a possibility to bear in mind. Other ‘flower’ symbols do not have this notch, and it’s unclear whether in that case they are all the dominant symbol.

Summing up the symbols with a notch

The symbol when losing its rod, or when drawn smaller, is indicating its loss of dominance to the other symbol in the pair, and such a symbol may have its subordinate status further emphasised by the addition of a small notch. This doesn’t occur in all instances of symbols without a rod, so doesn’t appear to be a requirement.

Notches only occur on Class I symbols, not on Class II, despite there being many more instances of symbols without their rods being used in the CII period.

It’s unclear if the notch has a meaning apart from emphasising the non-dominant status of the symbol. However, I can’t help but think that the notch may be ‘destroying’ the symbol, just as real objects are bent or broken to symbolically destroy them when used as ritual deposits.

Subscribe now

Share